User's Guide PVsyst Contextual Help

PVsyst SA 1994-2012

WARNING: About this manual. This is NOT a PVsyst User's Manual. This document is a printed version of the contextual help of PVsyst (you can call by typingF1 in the software). So, it is not organized as a common manual with chapters, sections and so on. It only displays a serie of independent pages suited for hypertext navigation Nevertheless, you can use this document for browsing the contextual help with hyperlinks, or print your own hard copy. This document is also available on our website www.pvsyst.com. A User's Manual of PVsyst is in preparation.

Contents

Chapter 1

1 Overview 1 General description of the PVsyst Software 2 Tutorials 2 Tips for beginners 3 Historical evolution of the software 14 Compatibility and Troubles

Chapter 2

14 Licensing 14 License rights and activation code 16 Asking for a license code 16 Payment conditions 17 Transferring the activation code on another machine

Chapter 3

17 Preliminary design 18 Grid System Presizing 18

Grid-connected system preliminary design

19 Stand-alone system presizing 19 20 20

Stand-alone system preliminary design Stand-alone system design Battery Voltage Choice

21 Pumping system preliminary design 22 Preliminary design: economic evaluation Chapter 4 22 Project design 24 Project design: tutorial 28 Project definition 28 29 29
Site and Meteo data in the Project Project Site and Meteo System definition

30 Plane orientation 31 31
Heterogeneous Fields Sheds mutual shadings

Contents
32 33 33
Shed optimization Sun Shields mutual shadings Orientation optimisation tool

34 Concentrating systems 35
Defining a concentrating system

36 PV array electrical behaviour 36 37

Arrays with characteristic's mismatch PV module / array with a shaded cell

37 Shadings: general organisation 38 40 44 48

Horizon - Far shadings Near Shadings: tutorial Near Shadings: main dialog Near Shadings: 3D construction

54 User's needs ("load") 55 55 56 57 57

Load profile: ASCII file definition Domestic User's needs User's needs: probability profile definition User's needs: daily profile definition User's needs: monthly values definition

58 Grid-connected system definition 59 59 60 60 61 61 63

Array voltage sizing according to inverter Amorphous degradation and array voltage sizing Inverter / Array sizing Grid Inverter sizing Different sub-systems, multi-MPPT inverters SolarEdge Architecture Normalised Grid voltages

63 Stand-alone system definition 64

Regulator Operating modes

64 DC-grid system definition 65 65 66

DC-grid load profile DC-grid line resistance DC grid: Overvoltage regulation

66 Pumping system definition 67 68

Pumping system sizing Pumping Systems: Generalities

Contents
74 80
Pumping system configuration Pumping Results: Examples

83 Array losses in PVSYST 83 84 86 86 87 87 87 88 88 89

Array losses, general considerations Array Thermal losses Array incidence loss (IAM) Module quality losses Array mismatch loss Array ohmic wiring loss Wiring loss optimisation AC ohmic loss from inverter to injection point Maximum current in wires External transformer losses

89 Module Layout 90 91 92
Construction and General description Subfields Secondary rectangles

92 Simulation 92 93 94 94 95 95 96 96
Simulation process: irradiance and PV array Simulation process: grid system Simulation process: stand alone system Simulation process: DC-grid system Simulation process: pumping system Export ASCII file On-line graph definition Simulation and comparison

96 Results 97 98 98 100 101 103 105

Loss diagram Normalised performance index Simulation variables: meteo and irradiations Simulation variables: Grid system Simulation variables: Stand alone system Simulation variables: pumping systems Simulation variables: DC-grid systems

106 Economic evaluation

Long term financial balance 107 108 Geographical and Meteorological data

Chapter 5

Contents
109 Meteorological data - Tutorial 113 Meteo database in PVsyst 114 Geographical locations / monthly meteo data 114
Importing monthly meteo values

115 Hourly meteorological data 115 116 116 117 119

Meteorological data averaging DRY or TMY hourly data Meteo tables and graphs Hourly meteo data quality check Time shift

120 Import from Meteorological data sources 122 125 126 127 128 129 129 131 133 133 134 135 136
Meteorological data comparisons ISM hourly data in the database Importing data from Meteonorm Importing PVGIS data Importing US TMY2 / TMY3 data Importing Canadian EPW data Importing Satellight data Importing SoDa-Helioclim Data Importing SolarGIS hourly data 3Tiers meteo data Importing WRDC Data Getting NASA-SSE data of a particular site Importing Retscreen Data

137 Import of custom hourly data ASCII files 137 138 Chapter 6
ASCII Conversion procedure PVsyst standard format

Conversion protocol 139 143 Tools and databases

144 Favourites 144 PHOTON database 145 Phovoltaic modules 145 148 149
Phovoltaic modules - Basic data Photovoltaics modules database PV Modules: Serie Resistance determination

150 Grid inverters

Contents
150 152 153 153 154
Grid inverters, main parameters Grid inverters, secundary parameters Grid inverters, efficiency curve Grid inverters, adjusting the efficiency curve Grid inverter database

154 Batteries 154 155 155

Batteries - Basic data Batteries - Detailed model parameters Battery Database

156 Back-up generator 156 Pump definition 156 157 158 158 159 159
Pump technologies Pump data: general page Pump data: performance curves Pump data: detailed parameter Pump data: current thresholds Pump: integrated power converter

159 Regulators for stand-alone systems 160 160 161

Converter: step-down technology MPPT or DC-DC converter Generic default regulator

161 Regulator for Pumping: parameters 162 162

Converter in the pump's definition Control device for pumping systems

163 Seller list 163 Array-Coupling Voltage optimisation 164 Comparisons between measured and simulated values 164 165 165 166 166 Chapter 7
Measured data analysis: general philosophy Checking the measured data files Predefinition of comparisons Transformations of data-files Data elimination in Tables

Cuts of erroneous data 166 166 Technical aspects

167 Updating Software and Databases 167 Languages

Contents
167
Special characters problems

167 Hidden Parameter 168

Default values and costs

168 Uninstall 169 File organisation 170 171 174 174 175 175
Export/Import of data files Directories contents File delocalization with Vista and Windows 7 Seing "hidden" files and directories in Windows Explorer Log Files Copy the data structure

177 Printing 177 Chapter 8

Print_Head

Copying Printer pages to Clipboard 177 178 Physical models used

178 Incident irradiation models 179 179 180 180 181

Meteo Monthly calculations Transposition model The Hay transposition model Diffuse Irradiance model Synthetic data generation

182 PV Module - model description 184 185 185 185 186 186
Phovoltaic modules - Model parameters Temperature behaviour correction PV Module shunt resistance Reverse characteristic of a cell PV Module reverse behaviour Model for Thin film and other new technologies

192 Batteries - Model description 195 Pump - Model description 197 198 198 199 Chapter 9
Pump model: General requirements Pump model from Head (and Effic) as f(FR), fixed U or speed Pump model from Head (and P) as f(FR), fixed U or speed Pump model from FR as f(Pp) for fixed Heads

Pump model from Ip and FlowR as f(Head) 199 199 Glossary

Contents
199 Albedo 200 Albedo attenuation factor 200 Altitude corrections 200 Autonomy and battery sizing (stand-alone) 201 Battery current efficiency 201 Battery maximum charge threshold 201 Clearness_Index 201 Climatic Distance 201 Dates read on the file 201 Diffuse attenuation factor 201 Domestic user's needs 202 Double orientation field 202 European Efficiency 202 Generic Year 1990 202 Grid current in tri-phased configuration 202 Incidence_Angle 202 Inverter: power overcharging 203 Irradiance loss 203 Longitudinal_Incidence_Angle 203 LOL "Loss-of-load" probability (stand-alone) 203 LOL "Loss-of-load" probability (Pumping) 204 LOL "Loss-of-load" probability 204 Metal resistivity 204 NOCT definition 204 Ohmic Loss Ratio (PV field) 205 Plane azimuth 205 Plane orientation 205 Plane tilt 205 Performance Ratio; PR, Losses

Contents
205 Profile Angle 205 Reference year: year value 206 Regulator voltage switch 206 Series Resistance (PV module) 206 Reverse characteristics (PV cell) 206 Summer Time - Daylight Savings 206 STC 206 Time Zone 207 Transverse Incidence Angle 207 Wind Velocity (Synthetic generation) Chapter 10 207 Validations 208 Validations of old versions of the program Chapter 11 215 References

Overview

This overview shows the general organization of the software and of the help system.
When beginning with the PVsyst software you are advised to have a look on the "Tips for beginners" and to follow the 3 tutorials.
Sorry, this Help is only available in English.

General description of the PVsyst Software

PVSYST V5.0 is a PC software package for the study, sizing and data analysis of complete PV systems. It deals with grid-connected, stand-alone, pumping and DC-grid (public transport) PV systems, and includes extensive meteo and PV systems components databases, as well as general solar energy tools. This software is geared to the needs of architects, engineers, researchers. It is also very helpful for educational training.

PVSYST V5.6 offers 3 levels of PV system study, roughly corresponding to the different stages in the development of real project: Preliminary design 17 : this is the pre-sizing step of a project. In this mode the system yield evaluations are performed very quickly in monthly values, using only a very few general system characteristics or parameters, without specifying actual system components. A rough estimation of the system cost is also available. For grid-connected systems, and especially for building integration, this level will be architect-oriented, requiring information on available area, PV technology (colours, transparency, etc), power required or desired investment. For stand-alone systems this tool allows to size the required PV power and battery capacity, given the load profile and the probability that the user will not be satisfied (("Loss of Load" 203 LOL probability, or equivalently the desired "solar fraction"). For Pumping systems, given water requirements and a depth for pumping, and specifying some general technical options, this tool evaluates the pump power and PV array size needed. As for stand-alone systems, this sizing may be performed according to a specified probability that the water needs are not met over the year. Project Design 22 : it aims to perform a thorough system design using detailed hourly simulations. Within the framework of a "project", the user can perform different system simulation runs and compare them. He has to define the plane orientation (with the possibility of tracking planes or shed mounting), and to choose the specific system components. He is assisted in designing the PV array (number of PV modules in series and parallel), given a chosen inverter model, battery pack or pump. In a second step, the user can specify more detailed parameters and analyse fine effects like thermal behaviour, wiring, module quality, mismatch and incidence angle losses, horizon (far shading), or partial shadings of near objects on the array, an so on. For pumping systems, several system designs may be tested and compared to each other, with a detailed analysis of the behaviours and efficiencies. Results include several dozens of simulation variables, which may be displayed in monthly, daily or hourly values, and even transferred to other software. The "Loss Diagram" is particularly useful for identifying the weaknesses of the system design. An engineer report may be printed for each simulation run, including all

Chapter 1 Overview

Overview
parameters used for the simulation, and the main results. A detailed economic evaluation investment conditions. 106

1
can be performed using real component prices, any additional costs and

Measured data analysis 164 : when a PV system is running and carefully monitored, this part (located in the "Tools" part) permits the import of measured data (in almost any ASCII format), to display tables and graphs of the actual performances, and to perform close comparisons with the simulated variables. This gives a mean of analysing the real running parameters of the system, and identify even very little misrunnings. In addition, Tools 143 include the databases management - for Meteo data and PV components - as well as some specific tools useful when dealing with solar energy systems: import of meteo data from several sources, tables and graphs of meteo data or solar geometry parameters, irradiation under a clear day model, PV-array behaviour under partial shadings or module mismatch, optimizing tools for orientation or voltage, etc.

Tutorials
There are presently 3 available tutorials, about the following topics Meteo data management Project design
24
109

(import from external sources)

40

(complete elaboration of a project)

3D near shadings construction

Tips for beginners

Help : you can get contextual Help from almost anywhere in the software, by typing F1, or very specific information are often available with little question mark buttons. Red dots : every time you have red dots on graphical views, you can drag them with the mouse to modify the involved parameter (examples: horizon line, plane orientation, near shading orthogonal drawings). Exporting tables : all result tables can be exported to other software by choosing "Export" in the menu: - either as text file (ASCII CSV format), - or by "Copy as text" into the clipboard (to be pasted for example in a spreadsheet software), - or by "Copy as image" into the clipboard. For some scrolling tables (solar parameter, meteo), you can choose the time period to be exported. Exporting graphs : all result graphs can be exported to other software by choosing "Export" in the menu: - either as image file (BMP color format) - or by "Copy" in the clipboard, which can be "Pasted" within any other software (MS-Word, etc.). Current window image: as in all Windows applications, pressing "Alt + PrintScreen" copies the current window into the clipboard. Printing tables, graphs or other components: 177 - allow for a double line comment in front of each print form, - are usually "intelligent" printings which hold complementary useful parameters, - offer a "preview" facility, - often ask for desired details about outputs, Screen resolution: PVSYST has been developed and optimised using the standard SVGA (800x600 pixels) screen resolution with small fonts. Using old VGA (640x400) is not advised: it will superimpose all dialog windows, hiding the "historical" tree of window labels indicating where you are presently located in the software. With higher screen resolutions; you are not advised using large fonts, which can produce unexpected display effects as the software was not quite tested for them. For changing the screen resolution, please open the msWindows tool "Display settings". You can usually reach it by right-clicking the windows main screen, and choosing "Properties".

Chapter 1 Overview

Overview
Historical evolution of the software

Of course any newly discovered bug (and bugs reported by the users) are repaired for each new version. Also the contextual "Help" system is continuously updated, either concerning new developments, or according to the numerous questions of users.

Version 5.3 (November 30th, 2010) by respect to Version 5.21 (September 3rd, 2010) 1. Help system: completely new CHM version, now fully compatible with Windows 7. A PDF version is available on the site. 2. New tool for the Meteo data quality check (Kt plots, clear days check) 3 4. 5. 6. 7. 8. 9. Implemented middle-interval shift for optimizing treatment of not-centered meteo data Importing meteo data from Global Incident (POA) measurements: bug fixed POA measured value now in the recorded data Concentrating PV module (CPV): definition with spectral corrections Concentrating systems: complete revision of the simulation process / variables Update of the loss diagram, also for concentrating systems. Inverter definitions, some bugs fixed: 3-voltage reading, bi-polar sizing, Heterogeneous fields: old files V 4.37 prevent simulations (0 everywhere) IAM calculation on diffuse: also with customized IAM function

Version 5.21 (September 3rd, 2010) by respect to Version 5.20 (August 3rd, 2010) 1. Direct link for importing PV modules and Inverters from PHOTON database 2. Shadings: define a new object (mansard, or roof-window) 3. Favorites choice in the main Database lists easier (by right click). 4. Directory Names now accept accents and some special characters (solving problems with Czeck Republic Windows XP installations). 5. Meteo data: the wind speed is now part of hourly values when defined monthly. 6. Inverters: still problems with 3-voltage efficiency definition (solved). 7. Sun-shields: bug during mutual shading calculations 8. Fixed littles bugs: printing of PR, soiling table in parameters, etc. Version 5.20 (August 3rd, 2010) by respect to Version 5.14 (June 30th, 2010) 1. File organization and localization has been changed Your working \Data\ structure is now in a writable area (avoiding delocalization of files written by PVsyst under Vista and Win7) 2. Module layout tool for the geometrical arrangement of your system 3. SolarEdge architecture: special option for decentralized architecture with Powerboxes. 4. Fixed bugs in Inverter definition, Projects, Latitudes over polar circle, etc. Version 5.14 (June 30th, 2010) by respect to Version 5.13 (June 25th, 2010) 1. On some machines, for unidentified reasons (firewall, proxy, ?) the AutoUpdate function induces crash at the opening of the software Version 5.13 (June 25th, 2010) by respect to Version 5.12 (May 25th, 2010) 1. Autoupdate: freezing at opening on some Windows installations (web access) 2. AC loss: now possible after inverter or after external transfo 3. AC loss: bug when identifying mono/tri situation 4. Heat loss default values: bug, not always possible to change value. 5. Stand-alone, economic evaluation: bug fuel consumption 6. Stand-alone, Available energy and Time fraction: bugs when very bad design 7. Simulation: Bug Hourly plots for some variables. 3

Chapter 1 Overview

Overview
Version 5.12 (May 25th, 2010) by respect to Version 5.11 (April 27th, 2010) 1. PV modules definition: - shows the apparent Series resistance (Rsapp, different from Rsmodel) - Efficiency plots vs. irradiance: display low-light values - Improved parameter definition in the help 2.Still bugs for the automatic update (freezed the program in some cases) Version 5.11 (April 27th, 2010) by respect to Version 5.1 (April 16th, 2010) 1. Thermal U value default definitions for some typical situations 2. Bug shed shadings, new feature according to modules 3. Load definition: lowered low limit to less that 0.05W. 4. Auto Updates should be operational from next version. Version 5.11 (April 16th, 2010) by respect to Version 5.1 (March 25th, 2010) 1. Bug when importing some meteo data (PVGIS and Helioclim) 2. Animation video file (*.avi) compatible with Windows media Player

Version 5.1 (March 19th, 2010) by respect to Version 5.06 (January 26th, 2010) 1. Automatic auto-update for new versions (doesn't work well until V 5.14) 2. Tool for the analysis of electrical effect of cell shading: Extended to several cells, in one or several sub-module (groups protected by one by-pass diode). 3. Generic (unlimited) shed shadings: electrical effect of shadings on the first cell row and bottom string. 4. Mismatch: histogram for the statistical study of loss distribution. 5. System design reference temperatures: now part of each project. 6. 2-axis tracking: shadings compatible with concentration option. 7. Help: tutorials for project design and meteo data. 8. F10 key for directly switching english <=> local language in most dialogs. 8 Helioclim data: updated tool according to the new web site data format. 9. Defined bi-polar inverters in the system design and simulation 10. Inverter: bug when efficiency not well defined 11. Heterogeneous fields: still bugs in area calculations and mixed fields. 12. Video recording of the shading scene now works 13. Stand-alone systems: bugs in regulator definitions and system verifications 14. Array voltage was not registred in the simulation 15. Export project tool: error warning, corrected Version 5.06 (January 26th, 2010) by respect to Version 5.05 (December 18th 2009) 1. PV model: Saturation current Io limit down to 0.1 pA (equation problems at low temperatures). 2. Helioclim data: the provider of these data has modified the site's format => readapted the program for a compatible easy importation 3. Shading calculations sometimes freezed. Improve reliability of shading calculations. 4. Near shadings, elementary objects, autorized tilt < 0. 5. System dialog: did not keep the defined parameters when re-entering the dialog. 6. Inverter database: terminated the update according to Photon Magazine 2009. 7. Vista and Windows 7 compatibility: Parenthesis were not allowed in the directory \program files (i86)\ proposed by Windows. Version 5.05 (December 18th, 2009) by respect to Version 5.04 (November 24th 2009) 1. Stand-alone systems: bugs in the Regulator dialog and the simulation process 2. Grid system sizing tool: still another deep revision for more conviviality bugs with master/slave definitions (sometimes divisions by 0) possibility of Strongly Oversized inverters (by modifying Hidden parameters)

Chapter 1 Overview

Overview
3. Inverter for 3 voltages: still some little improvements Version 5.04 (November 24th, 2009) by respect to Version 5.03 (November 10th 2009) 1. 2. 3. 4. 5. Hidden parameters were not modifiable (bug). Regulator definition had intempestive warnings, preventing using it. Heterogeneous fields: compatibility and warnings Orientation <=> Shadings Tracking frames with N/S frame: the tilt limits were not active Inverter database: partial update (about 30%) from Photon Magazine 2009

Version 5.03 (November 2009, 10th) by respect to Version 5.02 (October 2009, 26th) 1. Corrections in the Grid-system sizing tool (MPPT inverters, not yet perfect !). 2. Some background colors make things unreadable in Vista and Windows 7. Version 5.02 (October 2009, 26th) by respect to Version 5.01 (October 2009, 12th) 1. System definition freezed when defining multi-MWc systems. No more limit to the system size. 2. Improved the system sizing tool. 3. Corrected further bugs in the report (sometimes over-printing at head of the page). 4. Improved the ordering tool (sometimes e-mails were not well sent, and we did not receive your order). Version 5.01 (October 2009, 12th) by respect to Version 5.0 (October 2009, 6th) 1. We just discover a important bug: in some cases (synthetic generation without specified Diffuse monthly values), the Diffuse is very low, leading namely to over-estimated transposed values (GlobInc). Please reinstall this new version, and open the projects elaborated under V 5.0. If this occurs, the program will give a warning, re-calculate the meteo file, and you should re-simulate all your calculations for this project. Please discard the old inputs of such erroneous projects. 2. Help improved for system design and inverter sizing. 3. Bugs in the report of the Heterogeneous multi-orientation fields 4. Module database completed for all modules references in Photon Magazine 2009 (now about 5'300 modules) . Version 5.0 (September 2009) by respect to Version 4.37 (June 2009) This is a major modernization of the software. Many internal mechanisms have been improved since more than 2 years of development (in parallel to version 4.xx updates). Therefore there may be bugs which have not been detected during the development. Please be so kind as to report them carefully to the authors. 1. Multilanguage: the simulation report was already available in several languages, but now the software itself is (partly) available in English, French, German, Italian, Spanish and Portuguese. This is not yet a full translation: only the most used parts - especially regarding grid connected systems - were translated up to now. This is a very time-consuming job (more than 200 dialogs, and hundreds of information/ warning pieces to the user), and we will continue it progressively. On the other hand we don't intend to translate the Help at the moment. 2. Multi-fields: you have now the opportunity of defining several field types for a given project (with different PV modules or inverters, number of modules in series, etc). Their parameters are detailed on the final report, but the simulation results concern the whole installation. 3. Inverter: their definition includes many parameters which were not operational in the simulations up to now. Multi-MPPT devices: you can define a specific sub-field for each input. Possibility of Master/Slave operation: the Inverter's cascading is taken into account in the simulation process. For some models: power limitation when running under a specified input voltage, is now taken into account. Efficiency profile for 3 different input voltages. 4. Heterogeneous orientations: systems with 2 different orientations may now involve different sub-fields in each orientation, and/or a subfield for which the strings of a given inverter are distributed on both orientations (with mix of the I/V curve for correct calculation of the MPP). 5. Database management: the big lists of components stored as individual files, which took very much time to be loaded in the previous version (and sometimes caused bugs) are now replaced by centralized databases (CSV files). This results in an immediate access, and facilitates the updates of the DB of the software. Only the files you are creating or modifying by yourself will still be stored as individual files as before. This 5

Chapter 1 Overview

Overview
concerns: The PV modules (which should approach 5'000 modules in the DB of 2009), The inverters, The geographic site database 5of which the lis shows now the source of data).

6. Favorites: you can now define a list of your favorite components in the database, which drastically simplifies the use of big component lists. 7. Near Shadings: the full dialog and tool have been improved. You can now: Easily read and write a scene or a building directly from the 3D tool. Multiple selection allows to define groups of objects, that you can replicate or save on a file. Define/fix the characteristics of the view you would like to appear in the final simulation report. Register the shading's animation as a little video. 8. Special tracker devices with PV modules rotating on a tracking frame (with either N-S or E-W axis), are defined as special shading objects (with database of existing devices). 9. Shading of thin objects, like electrical wires or handrails, is now possible by weighting their effect on the "Module shading" part of the shaded areas. 10. Shading on strings: you can put a weight on the effect of the shading according to strings, in order to better approach the real shading effect on the electrical production (not only an upper limit) in the simulation. 11. Import Horizon profiles directly from Somletric SunEye, Carnaval software, Meteonorm. 12. Synthetic hourly data generation: the diffuse part may be renormalized to specified monthly data when available. This was not possible in the previous versions. 13. Improvement of the sizing tool for grid systems. You have now the opportunity of specifying either the nominal power, or the available area as starting point. The software indicates the required ranges for the number of modules in series and in parallel. A new powerful window shows all the constraints when sizing a field, i.e. The voltages of the operating array by respect to the inverter's specifications, Histogram of the waited power production of the array, compared to the inverter's nominal power. Estimation of the overload losses (and visualization of their effect on the histogram). This tool allows to determine precisely the ratio between array and inverter Pnom, and evaluates the associated losses. 14. The default losses management has been improved, especially the "Module quality loss" which is determined from the PV module's tolerance, and the mismatch on Pmpp which is dependent on the module technology. 15. Losses between inverters and grid injection have been implemented. These may be either ohmic wiring losses, and/or transformer losses when the transformer is external.

Version 4.37 (June 2009) by respect to Version 4.36 (April 2009) 1. Changed the program for the installation. The old one was no more compatible with the new ServicePack 3 of Windows. Version 4.36 (April 2009) by respect to Version 4.35 (March 2009) 1. CdTe modules: the module definitions of the old database are affected by the new recombination losses. => Recreated files in the database for all CdTe modules, including recombination term according to our recent measurements at the University of Geneva. Version 4.35 (March 2009) by respect to Version 4.34 (Feb. 2009) 1. Import of Meteo data: Bug in some new files of Meteonorm V 6.0 (containing leap years) Bug in some TMY3 files (date recognition) 2. CdTe PV modules: according to our detailed measurements: opportunity of defining Recombination Losses. 3. Shading factor calculation (especially diffuse): bug when the plane azimuth is very unusual (north in Nordern hemisphere). Version 4.34 (February 2009) by respect to Version 4.33 (Sept. 2008) 1. Database: PV modules from manufacturers (especially many amorphous) and new inverters. 2. Tool for analysing shading of one cell in an array: Improvements, shows now shaded I/V charact. and Pmpp 6

Chapter 1 Overview

Overview
loss. 3. Meteo data import: US-TMY3 Import implemented (1020 stations available) Satellight: Bug temperatures when importing 5 years at a time. PVGIS: Bug all months accounted as 31 days (overestimate 1.6%) PVGIS: Now uses PVGIS site new interface, much more convivial Meteonorm: Monthly files: still discovered a new format variant Temperatures: Possibility of importing NASA data, always available. 4. Tracking tilted axis: error when axis azimuth not south. 5. Measured and Meteo data: little bugs (extended available date formats). Version 4.33 (September 2008) by respect to Version 4.31 (July 2008)

1. Database: PV modules update from Photon Magazine 2008 and other manufacturer's data (over 3100 modules in the DB) 2. Loss diagram: still an error in the GlobShd evaluation (but doesn't affect the final results) 3. Meteonorm import: tolerant to another (not yet registred) file format Version 4.32 (July 2008) by respect to Version 4.31 (June 2008) 1. Database: Inverters update from Photon Magazine 2008 2. Loss diagram: incoherences in the shading and IAM relative losses (only display in this diagram - simulation results were correct) 3. Ascii Meteo importing tool: Date management improvement, Site names beginning by "New" are now possible. 4. Search Edit for easier choice in big component lists. 5. Print Preview: easier navigation through pages. Version 4.31 (June 2008) by respect to Version 4.3 (March 2008) Some bugs fixed : 1. Satellight data import: Temperatures were not well imported. 2. PVGIS import: copy/paste did not work with some internet browsers. 3. Site/Meteo choice for Projects: generated erroneous meteo files 4. Simulation report: erroneous tables overwrited parameters (namely IAM) 5. Import of Meteo ASCII files: improvements for daily data and date formats. 6. Long component's lists: edit box for direct access to a given item Version 4.3 (March 2008) by respect to Version 4.21 1. Import of meteo data from multiple popular sources (NASA, WRDC, PVGIS, RetScreen, Helioclim). 2. Comparison between several Meteo Data sources. Developments and results in the Help. 3. Import/Export of "Site" monthly data from/to EXCEL. 4. Improvement of the Meteo hourly files management (site and comment now editable/exportable). 5. Implementation of Tracking with vertical axis, also useable with positioning of modules on a "dish" 6. Module temperature calculation: revision (new parameters, absorption, etc). 7. PV model for amorphous: parameter determination according to a specified muPmpp value. Adjusment of all triple-junction module parameters in the database. 8. Bug Tracking: azimuth sign error in south hemisphere: the tracking was reversed ! 9. Bug Shadings, polygonal fields: rewritten the whole modules calculation. Version 4.21 (September 2007) by respect to Version 4.2 No new developments, only corrections of bugs. The main ones were: 1. Tracking two axis: returns to Azim=0 when sun over +/-90. 2. Near shadings: verification of interpenetration field-objects, some editing errors or improvements

Chapter 1 Overview

Overview
3. New PV modules didnt appear in the list. 4. Simulation: sometimes division by 0 with sheds. 5. Project situation dialog: improved copy of site <=> meteo. 6. Graphs: copy of the curve values to clipboard for exporting. Version 4.2 (July 2007) by respect to Version 4.1

1. Improvement of the navigability in the 3D construction tool, copy/paste of an object from one variant or project to another one. Automatic verification of the Field interpenetration (or tangency) with another object, which may prevent good shading calculations. 2. Backtracking strategy with all tracking systems, involves tracker width and distance definitions. 3. Sun-shields definition in the 3D tool; also with backtracking, which may considerably improve the sun-shield's performances. 4. High-Concentrating systems simulation, associated with 2-axis trackers; adaptation of the simulation process and loss diagram. 5. Long-term financial balance tool, including several Feed-in tariff strategies (day/night or seasonal variations, feed-in and self-consumption tariffs, etc) and system ageing. 6. Soiling parameter included in the simulation and loss diagram, with opportunity of defining monthly variations. 7. Direct import of meteo values from NASA-SSE database over the whole earth (by 1x1 cells), and practical procedure for importing meteo values from WRDC database, especially for countries where METEONORM data are scarce. 8. Improved the model for amorphous PV modules, especially safety of parameter boundaries, and behaviour presentation to the user. 9. Improved the Project dialog and choice of a site/meteo. Improved compatibility checks between the project's site and hourly meteo. Extraction/edition of the site geographic properties within an hourly meteo file, which was not possible up to now. 10. Database update, with PV modules and inverters of 2006/2007 (now around 2250 PV modules and 770 inverters). 11. Adaptations for Windows VISTA OS, especially concerning the visual interface. All other functions seem to be perfectly compatible. 12. Introduction of many "Logs" in the program, in order to facilitate the debugging of user's problems. Version 4.1 (January 2007) by respect to Version 4.0 1. MultiLanguage The Simulation output reports are now available in French, German, Spanish, Italian (useful for presenting the PV system characteristics to customers). Please contact the author if you wish to implement yourself a translation into your own language (you should have a good knowledge of the technical terms used in the PV technology). 2. Windows user's rights compatibility The DATA structure has been modified for compatibility with the user's rights protections in Windows. A user without writing rights can now copy the whole DATA structure for use in his own writable area. Data may be shared - or not - between different users of the machine. 3. Files and projects transfers Archiving or importing projects, as well as database updating tools have been improved. 4. Bugs in special shading parts Several "youth" bugs in newly developed features (often on special requests of users) have been fixed. Especially in the Shading part, concerning tracking mechanism as well as sheds with a tilted baseline or double-orientation. Version 4.0 (June 2006) by respect to Version 3.41 The main new feature in this PVsyst 4.0 version is the study of Pumping systems, which involves complex developments which may be not quite safe in this first issue. 1. Pump Model Development of a general and original pump model, suitable for use in PV applications.

Chapter 1 Overview

Overview

This should be suitable for any type of Pump or Motor useable in PV systems. This should describe the operating of the pump over a large Electrical and Hydraulic domain, encountered in PV conditions. Its parameter should be available from usual pump manufacturer's datasheets. For a given pump, the model may be specified using several kinds of original data. Its accuracy over the whole domain has been checked for some pumps using measured data. The defining dialog shows graphics of the behaviour of the model, as functions of all relevant variables. The pumps database is still limited to about 20-30 models; only one manufacturer has answered our request for datasheets 2. Controller / Converter device for pumping Model for a new controller for pumping systems, including the System Configuration controls and Power converter. A default controller is proposed for each system configuration, with parameters automatically adjusted according to the system for optimal operation. 3. Pumping systems Three system types are proposed: Deep well, Pumping from lake, pond or river, and Pressurisation system; For each type, several system configurations are possible: Direct coupling (with eventual improvements like booster, pump cascading or array reconfiguration), with Power converter (MPPT or Fixed V), or battery-buffered. Water needs and Head characteristics may be defined in yearly, seasonally or monthly values. 4. Presizing tool for Pumping systems As for stand-alone systems, a Pre-Sizing tool has been developed, which proposes a PV power and Pump power sizing, according to the location and meteo, user's needs and LOL requirements. This simplified model takes the pump technology and system configuration into account. It also proposes a very rough estimation of the costs. 5. Detailed Simulation The design of the pumping system - rather complex with such a number of pumps and system technological aspects - is assisted by numerous sizing propositions, and helps (advices, graphs, blocking of uncompatibilities, etc) when choosing the system layout and configuration. The hourly simulation accounts in detail for all features defined for the system, and is specific for each configuration listed above (direct coupling, with converter or battery). As for the other systems, a detailed engineer report explains all parameters and results of the simulation. All the losses and mismatches along the system are quantified, and visualised on the "Loss Diagram", specific for each configuration. 6. Help for pumping systems The development of this Pumping tool has brought a deep understanding of the PV pumping systems problematic, and the operating / efficiency limitations inherent to the numerous possible solutions. This Help describes in detail the implemented models, and sets a broad panel of the different available technologies, as well as delimits their implementation boundaries. 7. Heat transfer factor for thermal losses of PV array Some users has pointed out that the proposed parameter accounting for wind velocities was not correct and lead to underestimated thermal losses. There is a new detailed discussion on this subject in the Help, and the program now advices to use wind contribution only when the wind velocity is quite well determined (now default value is Kv = 0 W/mK / m/s) 8. Inverters Several parameter usually specified in the datasheet have been added to the inverter definitions. But none of them is used in the present simulation. Refinements of the inverter modelling are planned for a next version. These new data have been added in the whole database when available. Almost 300 new inverters were introduced, many also suited for US market. The database includes now more than 650 inverters. The 50/60 Hz frequency has become a choice criteria in the lists. 9. PV modules The choice list shows a nominal (MPP) voltage of each module for making the system design easier. 9

Chapter 1 Overview

Overview
The database includes now more than 1'600 PV modules. The PV module definition dialog was improved and some bugs fixed. Specifying the Voltage or Power Temperature coefficient is now possible also for amorphous modules. 10. Remarks in PV components An unlimited text editor is now available for giving detailed descriptions of all the PV components.

11. Miscellaneous improvements or fixed bugs - Projects are now sorted according to their system type in the list. - The "Archive Projects" tool has been debugged and improved. - Summer/Winter Hour may now be taken into account when importing Meteo Data as ASCII files. - Some little bugs concerning the simulation, especially of stand-alone systems (Wearing state not computed, display errors, etc). - Loss diagrams: complete review, some corrections for losses coherence. Version 3.41 (March 2006) by respect to Version 3.4 1. Automatic facility for importing "satellight" meteo data. "Satellight" data are real measured meteo data, available free from the web for any pixel of about 5x7 km in Europe, and for years 1996-2000. Their quality becomes better than any terrestrial measured data, as soon as you are far from more than 20 km of the measuring station. 2. Fix some bugs of version 3.4, concerning: - HIT PV modules model, - memory leakages and orientations in some shading special cases, - transposition safety when bad meteo values, - orientation choice dialog, - meteo ASCII conversion facility. Version 3.4 (July 2005) by respect to Version 3.3 1. New modelling procedure for the amorphous modules This procedure was established and validated using the results of a research project performed at CUEPE, funded by the SIG-NER fund (the SIG - Services Industriels de Genve - are the Geneva Electric and Gaz Utility). This project included detailed long-term measurements of 6 PV modules in real conditions. It also gave a quantified validation of the standard model for crystalline and CIS modules. 2. Extended component database Over 1'200 PV modules and 400 inverters are now referenced, with dates for identification of market availability or obsolete components. With such big lists, a mechanism for quick access time (and background process update) had to be implemented. 3. Near shading 3D tool View of the shading scene in realistic colors (settable by the user for each object) - instead of "iron wires" representation - improves the understanding of complex scenes, and gives a much more attractive image of the project for the final customers. Animation over a whole chosen day also clarifies the shading impact of a given situation. 4. Implementation of tracking planes in the 3D shading tool. Especially suited for the optimisation of heliostat arrays layout. 5. Review and improvement of the Simulation Process Clarification of the losses at any stage of the system, extension to battery systems with DC-DC converters, rewriting of the heterogeneous field treatment , etc. 6. Detailed loss diagram Gives a deep insight on the quality of the PV system design, by quantifying all loss effects on one only graph. Losses on each subsystem may be either grouped or expanded in detailed contributions. 7. Restructuration of the internal representation of the physical variables These two last improvements were made possible thanks to a very deep revision of the internal data structure, in

Chapter 1 Overview

10

Overview

order to obtain more flexibility when using the simulation variables. This reorganization is transparent for the user, but allows now many enhancements of the simulation process, namely easy adding of new variables when necessary, including them dynamically in the simulation process according to specific system configurations (for example defining regulator losses when used in a battery system), or inversely discarding other ones when they are not relevant. The old fixed variable set did not allow a coherent description of the system losses. Therefore simulation has to be performed again for getting the Loss Diagram on the old result files. Caution: In spite of intensive tests, these deep modifications may have produced some bugs which have not been detected by the author. We thank the users for carefully reporting any misrunning or strange behaviour of the software to the author. 8. Measured data simulation comparison Improvement and debugging of the Measured Data Importing Tool, and the comparison between measured and simulated values. Improvement of the break-down data eliminations. 9. Daily and Hourly Plots of the load profiles

Version 3.3 (February 2004) by respect to Version 3.21 1. Output results presentation and hourly plots When displaying simulation results, the standard printable result forms are now directly accessible (while they were only available through the "preview" option in the "print" dialog up to now). During simulation process, the program can store a sample of chosen variables in hourly values. This allows for displaying graphics in hourly or daily values, with several variables on the same plot. Thanks to a very fast and easy navigation over the whole year, this offers a powerful tool for visualizing the instantaneous behaviour of the system all over the year. This helps, for example, identifying unexpected behaviours of the system in some specific operating conditions (for example, SOC and regulation states in standalone simulations). 2. Stand alone systems: implementation of MPPT and DC-DC converters Up to now the stand-alone systems were only defined with a simple usual configuration (i.e. PV array directly connected to the load and battery through the regulator). It is now possible to include a MPPT or DC-DC converter between the PV array and the battery/load. This converter is part of the regulator definition. 3. New tool for optimising Fixed Input Operating voltage Shows the average power or efficiency of the PV array over a period (year, summer, winter), as a function of a fixed user's voltage. Shows ohmic and diode voltage drop effects. 4. Hourly profile for domestic use load definition In order to better estimate the battery behaviour and wear. Automatic placement of lighting and TV uses to evening hours. 5. Improved irradiance clear sky models for very high altitudes (up to stratospheric). For very special uses of the software. 6. Meteonorm input adaptation The outputs format of the Meteonorm monthly files has been changed with the new version 5, and we had to match the reading format accordingly. This is not yet quite fixed (namely the new Meteonorm files don't include the Station name nor the geographical coordinates, which have to be input manually). 6. Databasse update Added about 500 new PV modules, from the PHOTON magazine tables of February 2003 and February 2004. Added about 40 Inverters from PHOTON magazine, march 2003. Created a list of the main component manufacturers. This list is now selective, according to Manufacturers/ Retailers and Component type. Version 3.21 (November 2002) by respect to Version 3.2

11

Chapter 1 Overview

Overview

Shading factor calculations with partition in Modules chains: Some computation errors had been introduced with v3.2, fixed. Horizon definition with more than 20 points now possible. Minimization (iconisation) of the Program window now works. Corrections for compliance with Windows XP environment. Version 3.2 (July 2002) by respect to Version 3.11 1. Thin film modelling The main novelty is the special tools for the treatment ofThin film technology modules tripple junctions, CIS, CdTe).
186

(a-SI:H, with tandem or

Up to now the program used the standard one-diode model, although it is not well-suited for these technologies. There is no consensus up to now in the PV community on how to model these devices. The task of finding a general model is by far above our possibilities. It would require a big research project at the international level). Nevertheless we carefully measured a single device in great detail (a-SI:H tripple junction) and tried to find improvements of the "standard" one-diode model. We found and implemented two adjustments, which can improve significantly the performances of the model: - the Shunt resistance of the device is drastically increasing (exponentially) when the irradiation diminishes. - the temperature behaviour, which is fixed as a result in the standard model, can now be adjusted at any desired value (often given by manufacturer). For our test device, this diminished the modelling error (over a 4-month continuous measuring period) from about 11% to a very few (1 to 3%). But this doesn't take spectral effects into account (see the "Help" for further details). Be aware that these are available tools, but we cannot assess parameter values for any modules or any technologies. 2. Orientation optimising tools There is now an on-line tool showing the collecting performance as a function of tilt and orientation when choosing them. Also for shed disposition, a new graph shows the annual yield curves taking shadings into account, for optimising the tilt, shed spacing and collecting/ground area ratio. 3. Components Database tools Export and Import of PV components (PV modules, Inverters) between the PVsyst database and spreadsheets like MS Excel, allowing for displaying, input and correcting component data in tabular form. Improved and securised default values for the input of new PV modules in the database, only based on Manufacturer datasheets. The Excel sheet shows clear detailed information about required, optional and PVsystcalculated parameter. Inverters: Automatic build of efficiency profile, according to Maximal and "Euro" efficiency data. This allows for a much more easy input of inverter data from Manufacturer datasheets. Be aware that due to additional parameter, PV module and Inverter files written by Version 3.2 are not yet readable with anterior versions (but old files can of course be read by version 3.2 !). Update of the Database, which includes now more than 600 PV modules and 200 inverters of the market. 4. Miscellaneous Near shadings: several little bugs and practical improvements. Improved the tools for manipulating and zooming the scene on the screen. Implemented the display of shadings calculated by points, when standard polygon algorithm fails. Included a full example as tutorial for the "Measured Data Analysis" part, which allows for importing measured data in PVsyst, and closely comparing them with the simulated values. Revision and improvement of the "PV array behaviour" graphic tools. Included a detailed Help. "Perez" transposition algorithm (not the Hay transposition model proposed by default !) had a little bug which caused a discrepancy of the order of 2-3% on yearly results for vertical planes (and of course less for less tilted planes). Revision of the tool for defining Currency Rates, which had some bugs.

Chapter 1 Overview

12

Overview

Revision of the general displaying conditions when using screen settings with "large fonts". Many windows appeared not full developed, and had often to be resized. Also graphics and tables had sometimes very little fonts. Reading of files without "Archive" attribute (which is sometime removed by some file managers) is now possible. Version 3.11 by respect to Version 3.1 Meteo: generating meteo hourly files in CSV format for export. Meteo transposition tables: possibility of -180..180 scale. Horizon: automatic import of files from "Meteonorm" and "HoriZON" software. PV module: Pnom tolerance from manufacturer included as parameter. Simulation: The user can now define a PV module quality loss, by respect to the manufacturer nominal data on which is based the PV module model. Therefore the results can be adjusted if necessary for Energy Yield Warranty. Installations in Polar regions: Possibility of defining meteo, and performing pre-sizing and simulations for Arctic and Antarctic meteo (including months with zero irradiation). Version 3.1 by respect to Version 3.01, 3.02 and 3.03 Database updates will be periodically available on the WEB site www.pvsyst.com. A special tool allows to dispatch the files into the PVSYST data structure. Printed outputs: Printed results forms may also be exported as image, for pasting in other software like MS-Word. Allows to fully insert PVSYST results in documents (useful for example for sending them by e-mail). Import of Meteo data: Import from Meteonorm software now quite debugged (except for bugs of the Meteonorm software itself with monthly data: the program tells you how to come over). Direct import of US TMY files (240 US sites available free from WEB). User's needs definitions: Graphs and printings are now available. A new feature allows for importing Load Profiles in hourly or daily values from ASCII files (e.g. for example from EXCEL). Horizon definition: Diffuse and Albedo factors are newly introduced. Horizon treatment in simulation strategy has been changed: it is now equivalent to the near shadings treatment (i.e. included in the "shading loss" calculation). Near Shading tool: Further debugged. Undo facility (up to 10 levels), Very useful new tools for distances and angle "measurements" on the scene. Simulation: Losses: New parameter accounting for the real quality of PV modules. South hemisphere compatibility: Fully debugged, with new azimuth definition (negative toward east, clockwise). Sunpath diagrams are now from right to left (negative azimuts, i.e. sunrise, are on the right of the graph, as the sun progression ...). Therefore gives realistic image of the horizon drawing. Measured data: Importing widely debugged, Possibility for ASCII lines of more than 255 characters, Data import is now possible with input Daily Values. Measured data with orignal irradiation measurements in the collector plane. Graphics and comparison tool are debugged and improved. Version 3.0 (December 1999) by respect to Version 2.21 The PVsyst software has been entirely rewritten, using the Borland DELPHI 3 platform instead of the old Borland Pascal 7, gaining in user graphical interface quality and reliability, as well as in compatibility with most recent versions of the Windows operating system. The user's interface has been redesigned, and navigation in the software was strongly improved, with the collaboration of the LESO-EPFL team. Introduction of a "Green line" for guiding the user in the project development. Preliminary design: Implementation of this quite new sizing feature, for grid-connected and stand alone systems. Project design: The Project organization has been simplified. Parameter definition and results are summarised in one only "Simulation Version" file. Several valuable tools were added (including the sizing "expert" for building the system parameter).

13

Chapter 1 Overview

Overview
Several new Tools which help understanding more deeply many PV system behaviours. "Help" system, which provides a detailed contextual "help" by typing F1 anywhere in the program.

Compatibility and Troubles

This software is now available in several languages (English, French, German, Italian, Spanish, Portuguese). Additional languages could be included by filling the files "Texts.csv". But the languages using other Character sets than the standard ANSI may cause great difficulties. PVSYST V5.0 runs under any Win'95/'98, Windows NT or 2000, Windows XP, Vista (32 bits) and Windows 7. Most of the data files from PVSYST, versions 3.xx and 4.xx (projects, components, meteo) can be read with this new version 5.0. But the inverse is not true (upward compatibility). Importing Meteo Data: - Link for direct import from the Meteonorm 126 software (versions 4, 5 or 6). - Link for direct import from many popular meteo data sources 120 from the web (including NASA for the whole world, US TMY3, PVGIS, Helioclim, Satellight, Retscreen, etc). - Almost any custom "Hourly Meteo" or "Measured Data" ASCII file can be imported, whenever it holds one data record on one ASCII line. Most detailed data (hourly or daily data) produced by the software can be Output to CSV customised ASCII files (compatible with any spreadsheet program). Many Data Inputs or Output are possible through the clipboard (graphs or tables as bitmaps, tables as CSV-text images, allowing direct export to spreadsheet programs like Microsoft EXCEL). Many input files in ASCII format are accepted, i.e. for measured data, hourly load profile values, horizon profiles, etc. Troubles Many new features have been added in this versioon 5.0. These motivated deep changes in the internal organization of the software. Although it has been tested in some relevant conditions, it is impossible to check all the running possibilities after each change in the program. Therefore the early version 5.0 will probably show weaknesses in the first period. If you encounter some problem during the use, be so kind as to report them carefully to the authors. In the same way, if you have some suggestions for improvements or adding some useful new feature, please don't hesitate to contact the authors ! The "Pumping" part was a quite new development in version 4.0. Owing to its complexity, it was not at the top of performances in the first version 4.0. It should be progressively improved for the future versions, but we observed that as it is not used very intensively, we had very few returns of users about its problems.

You can install PVsyst from our website www.pvsyst.com, and install it. It will work during 30 days without any limitations, for evaluation. After that it will revert in DEMO mode, and you will need a licence and activation code for using it.

License rights and activation code

License Rights
For new customer, we request that each company, juridical entity, per country (hereafter named company), to purchase the PVsyst license rights prior to be entitled to purchase activation codes. This price is a one time fee per company. The set of the PVsyst license rights and the activation code(s) is considered as a group identified by a Customer ID.

Activation Code
After installation, PVsyst runs in evaluation mode (with full capabilities) during 30 days. Afterwards, it turns in DEMO mode , and you have to request an activation code to run the software in unlimited-time mode.
16

which will allow you

Chapter 1 Overview

14

Licensing

One activation code per workstation is required. The activation code is paired with the Local Number resulting of the installation of PVsyst on a given workstation. Requesting an activation code
16

Upgrade and Update

If you already own a previous version of PVsyst (with a lower major version number, i.e. V4.xx or V3.xx), an upgrade activation code is required to upgrade to the latest version 5.xx of the software. One upgrade activation code per workstation is required. Update of the software, i.e. change to a higher minor version number (e.g. V5.21 to V5.3) is free of charge . An automatic tool in the software checks for new version available each time you run PVsyst, and performs the installation upon user request.

Price list
PVsyst License Rights Activation Code(s) Upgrade Activation Code(s) (V4.xx or V3.xx) Discount for Academic and Educational Institutions 800 CHF/ per company 200 CHF/ workstation 100 CHF/ workstation - 20%

VAT
An additional VAT is charged for Swiss users only. No additional taxes nor shipping costs are required for all other countries (at least from the exporter point of view).

Transferring the software code to another machine

17

NB: The Activation Code is constructed using the "Local number" provided by each installation of the software on a given workstation. If you have to reinstall your "Windows" environment, the Local number will change and your code will no longer be valid. Therefore before reinstalling Windows, you should transfer the code 17 to another workstation, in order to keep your activation code valid. After reinstalling Windows and PVsyst on the original workstation, you will be able to retransfer back your code from the other workstation. NB: In special cases, people who need to dedicate the code to a given machine whatever the Windows installation (for example in classrooms machines where windows has to be reinstalled frequently), you can get another type of License code, based on the Hard Drive number. Nevertheless this code cannot be transferred to another machine.

Network
Installation on a network server is not recommended. Nevertheless, it is authorized only if each user computer has a valid activation code. Please note that this software has not been fully tested for this mode of operation. But you can share your data area in network with other users. For this please open "Files" / "Directories", and here you can copy your working space \ Data \ anywhere on your machine. If you have to put it on a network, please consider your network path as a disk, as PVsyst doesn't recognize the network \\: paths. Be aware that no check is performed for the simultaneous use of data.

15

Chapter 2 Licensing

Licensing
Warning

Neither the University of Geneva, nor the author take on any responsibility under any form concerning the database contents, the accuracy of the results, or for consequences whatsoever to their use.

Conditions for the supply of the DRY Swiss meteorological data,

prepared by the "Physics and Installations in Buildings" Section of the EMPA: Basic data are property of the ISM (Swiss Institut of Meteorology) and cannot be distributed. The EMPA has been authorized to distribute data in a prepared format. The EMPA prepares meteorological data from the ANETZ network of the ISM only for their use in software or computations in the field of Building Physics, Energy Management and Building Installations. Neither the ISM, nor the EMPA or employees of these offices accept any form of responsibility for the accuracy, the integrity or the ability to the use of these data, neither for consequences of their use. The delivery of prepared data to third parties is not authorized.

Asking for a license code

Requesting an activation code can be done in two ways: 1. From the software use the menu License / Order and Purchase, and choose the method Order by email. Then, provide your complete address (specially your email address) and fill in the order form with the type and the quantity of desired activation codes. Choosing Send by Email will send an email to PVsyst administration ([email protected]) containing your order form and the "Local Number" specific to the installation of the software on your machine. A copy of that email is sent to you at the same time. From the website www.pvsyst.com, choose Download/Purchase. Then log in to your user account with your email address and password (or create a new account) and complete the web form. You will be asked to provide the "Local Number" specific to the installation of the software on your workstation (the Local Number is found by clicking the menu License / Status and Activation). Payment by Credit Card 16 , Bank Transfer 16 or PayPal 16 is possible.

2.

Please note that the license code is closely related to the "Local number", which is created the first time you install the software. The Local Number is then closely related to the machine on which PVsyst is installed. If you order several activation code(s), you may specify the number of desired code(s) on the main order. After receipt of your order, an activation code will be sent to you by email within a few working days in order to run the software on your machine.

Payment conditions
Payment by Credit Card or PayPal
Payment by credit card is possible on our website www.pvsyst.com, via the secured website of Paypal. You will have first to log in to your user account (or create a new one). Once logged in, you can purchase PVsyst activation code(s) or pay an invoice using a credit card. When receiving your payment, we send the activation code, along with the corresponding paid invoice.

Payment by Bank transfer

If not possible by credit card, you can order your license and activation code asking a payment by bank transfer. In this case, we return an invoice, to be paid within 30 days. The detailed identification of our bank account is given in the invoice, but it is also possible to pay this invoice by credit card on our web site.

Chapter 2 Licensing

16

Licensing
The activation code will be sent as soon as we receive your payment.

Transferring the activation code on another machine

When running with a valid code number, the program provides a tool for transferring the software license to another machine. Note that performing this tool will turn the software on the initial machine in Demo mode . You could return to the initial state by performing another license transfer from the second machine. 1. First, on the second computer (the computer on which you want to transfer the license), install PVsyst by downloading the program from our website www.pvsyst.com. 2. In the main window, open the menu License then Status and Activation and copy the Local Number that appear in the Registration codes panel (you can use the button copy to copy it to clipboard). 3. On the first computer (the computer for which you have the valid license), open the menu "License" in the main window then "Transfer to another machine, and follow the steps described in the license transfer wizard. 4. When requested by the wizard, carefully report the Local Number of the second computer and your License Name. Then click Next. 5. The new activation code for the second computer will appear: please note or save it carefully (you can also send it by email) and report it for use in to second computer. 7. Be careful: after clicking "Close", the program will turn in Demo mode on the first computer, the present activation code will become invalid, and the software will not be able to run anymore with full capabilities. To come back, you will have to transfer back the activation code from the second computer. NB: It is not possible to transfer the code: - If you are not "Administrator" of your machine, - If the opportunity of transferring has been disabled previously on this machine (anti-theft), - If your user's code is of a special kind, based on the Hard Drive number of your machine. It is the pre-sizing step of a project. It is aimed to quickly define the general features of a planned PV system. In this mode the system yield evaluations are performed very quickly in monthly values, using only a very few general system characteristics, without specifying specific system components. A rough estimation of the system cost is also available. The procedure is straightforward: you just go over the three buttons "Location", "System" and "Results ". First click on "Location" button: you have to give a description of your pre-sizing project in order to identify it after saving. The pre-sizing projects are simple files which don't allow for several variants. Choose a location in the database. You can obtain the location details, or even create or import a new location from Meteonorm 126 or US TMY data 128 , using the "open" button. When necessary you can also define an Horizon
38

profile.

Click on "System" button. The pre-sizing procedure is then specific for each type of system: Grid-connected system Stand-alone system Pumping system
21 . 19 18

17

Chapter 2 Licensing

Preliminary design
Grid System Presizing

Pre-sizing is a rough estimation of the PV system energy yield, based on a few very general parameters and mainly dedicated to architects during an early evaluation of a site. You should not use this tool for the study of a system. The meteo input data are computed in monthly values (taking plane orientation and horizon into account) and applies efficiency coefficients according to a PV technology and other considerations. These coefficients may eventually be re-adjusted by an expert user for special conditions in the Hidden parameters 167 . The expected precision could be around 10% or more. More precise results will be obtained with the hourly simulation performed through the "Project Design" option, including realistic available components and detailed system perturbations. Especially the financial aspects are based on coarse hypothesis, which can widely vary from country to country. These hypothetic financial parameters can be adjusted by the user by choosing "Edit costs" in the economic results sheet.

Grid-connected system preliminary design

After defining the "Location" 17 the "System" button displays a first screen where you should first define the plane orientation (try dragging little red dots!). NB: a little tool helps for the choice of the optimal orientation choice when not optimal.
33

, or the amount of losses resulting of you

Then you have to choose if you want to size your system on the basis of: - Active area of the collector field - Nominal power of the system - Annual energy yield. Pressing "Next" gives a second screen for defining system properties, especially from the architect point of view: - Module type: "standard" (give also the module power), "translucide custom" (with spaced cells, you should define the filling ratio), or "not yet defined". - Technology: will determine the default efficiency, that is the needed area for a given power. - Mounting disposition: indicative, not used in calculations, - Ventilation property: will slightly influence the efficiency due to module operating temperature. Now you can open the "Results " which gives the Nominal Power, Area or Annual energy yields, as well as some result graphs, table and economic evaluation 22 (to switch from one to the other please use speedbuttons left). You can now play with the parameters and immediately see the results. You can print a report, or store graphs and tables in the clipboard to export it to another software. You can also save your project, and load another one for immediate comparisons.

Computation:
The evaluation of the available irradiance on the collector plane uses the Monthly Meteo 179 tool algorithms, and the system energy output computations are done using constant efficiency and correction coefficients according to the chosen system parameters. The accuracy is of the order of 10 - 20% (worst case for faade installations). If necessary the coefficients used for this tool may be modified in the Hidden parameters
167

Chapter 3 Preliminary design

18

Preliminary design
Stand-alone system presizing

Pre-sizing is a rough estimation of the PV system energy yield and user's needs satisfaction, based on a few very general parameters. It is aimed to determine the size of the optimal PV array power and battery pack capacity required to match the user's needs. The input solar energy is computed in monthly values (taking plane orientation and horizon into account), and requires only the monthly data provided by the "Sites" database. Besides the Battery voltage which is related to the overall system power and geographic extension (due to distribution ohmic losses), the two basic user parameters are: The desired system autonomy (in days), which determines the battery capacity, The required LOL, giving the required PV array nominal power. After sizing the PV system with this tool, it's real performances should be verified by performing a detailed hourly simulation (option "Project design"), using real components and taking all system perturbations into account.

Stand-alone system preliminary design

See also: Sizing of a stand-alone PV system
20 : general

considerations.

After defining the "Location" 17 the "System " button displays a first screen where you are asked to define the plane orientation. NB: the button "Show optimization" opens a little tool which shows the winter yield according to the plane orientation. For stand-alone systems, the plane orientation should usually be optimized according to the worst conditions, i.e; for winter irradiance. Pressing "Next" gives a second screen for defining the user's needs 201 , from a domestic use point of view. You have to specify every foreseen appliance, their consumption 56 and use conditions. Now you can open the "Results" which asks for - the required autonomy 200 in absence of sun - which determines the battery pack capacity, - the required "Loss-of-Load probability" 203 (P LOL) , - the planned system voltage 20 These parameters lead to the determination of the array nominal power (i.e. the installed STC according to the manufacturer specifications), and the battery pack capacity.
206

power

The first result graphs shows the potentially available solar energy, along with the user's needs. The second one gives the average state of charge of the battery (low values could lead to a quicker deterioration of the batteries), and PLOL monthly distribution. The table holds all monthly values, including then needed back-up energy. Finally the rough economic evaluation 22 gives an idea of the investment and energy price. You can now play with the parameters and immediately see the results. You can print a report, or store graphs and tables in the clipboard to export it to another software. You can also save your project, and load another one for immediate comparisons. Computation: The evaluation of the available irradiance on the collector plane uses the Monthly Meteo 179 tool algorithms, which calculate irradiation's monthly averages on the basis of instantaneous data for one day per month. This is not sufficient to manage the storage balance evolution from day to day, and the effective use of solar incident energy. Therefore the program generates a random sequence of 365 days, according to the algorithms of Collares-Pereira, renormalised to the monthly sums, and calculates the daily battery balance for three intervals in a day (morning, day and evening). The accuracy is of the order of 10 - 20% (worst case for very tilted installations). If necessary the general parameters (array overall efficiency, battery efficiency, battery low charge threshold) are user-modifiable through the menu option "Preferences"/"Edit Hidden Parameter" 167 .

19

Chapter 3 Preliminary design

Preliminary design
Stand-alone system design

When sizing a PV stand-alone system, the basic constraints are the availability of solar energy during the year, and the satisfaction of the user's needs. The problem to be solved is the optimisation of the size of the photovoltaic generator and the storage capacity, subjected to criteria which may take on different weights depending on the use: Reliability of the supply, which is very important, for example in decentralised telecommunication installations. But in a domestic installation, this may be overcome with a small back-up generator. This reliability is measured as the "Loss of Load" Probability ("P LOL"). Investment and maintenance costs, which should take into consideration the cost of the PV generator, the initial cost of the batteries, as well as that of their maintenance and replacement. The high price of the kWh used necessitates a highly detailed study of the real user's needs, and the use of specific appliances that are highly economical regarding to energy consumption. Durability: the cost of the batteries is closely related to the quality of the batteries chosen, as well as their longevity, which is itself dependent on the conditions of use (average state of charge, cycling, depth of discharge, temperature).

It is therefore observed that the optimisation of a PV system is a complex problem with several criteria, which has no ready-made solution. The preliminary design tool allows for "playing" with the main parameters, and rapidly observe their effects on the system performance, to optimise the desired characteristics by successive approximations. But the accuracy of these monthly results is not guaranteed and the behaviour of the chosen system will have to be verified by a complete simulation in hourly values, with real available components.

Battery Voltage Choice Battery Voltage Choice

In a stand-alone PV system with direct coupling to the user (without inverter), the battery voltage determines the distribution voltage. As now many DC appliances can be found as well in 24V as in 12V, this choice should be made according to system and/or appliance power, as well as the extension of the planned distribution grid to minimise the ohmic wiring losses. This choice should be done from the early planning of an installation, since the existing appliance voltage usually cannot be changed, and voltage translators will be expensive and not 100% efficient. The rated distribution values could be chosen according to the following criteria (inverter supposed directly connected on the battery pack): 12V: little systems for lighting and TV:

Appliance max power Corresponding current Inverter : about

24V: medium size, with fridge and little appliances, or wiring extension to more than 10 m.

< 300 W 25 A < 1 kW < 1000 42 A < 5 kW < 3 kW 62 A < 15 kW

Appliance max power W Corresponding current Inverter : about 48V: special industrial or agricultural use Appliance max power Corresponding current Inverter : about

Higher powers require either high DC voltages (special appliances) or AC feeding through inverter.

Chapter 3 Preliminary design

20

Preliminary design
Pumping system preliminary design
See also: Sizing of a pumping PV system After defining the "Location"
17 67

3
: general considerations.

the "System " button displays a dialog where you are asked to define:

On the right panel: the plane orientation. NB: the button "Show optimization" opens a little tool which shows the winter yield according to the plane orientation. For stand-alone or pumping systems, the plane orientation should usually be optimized according to the worst conditions, i.e. for winter irradiance. On the left panel: - The Water needs (in yearly, seasonal or monthly values). The nominal head at which it should be pumped (level difference between water outlet and source surface). The diameter and length of pipes (optional, for eventual friction losses). A pump technology (centrifugal for rather low heads, positive displacement for high heads). An array-pump coupling strategy, which strongly affects the system performances.

Now you can open the "Results" which asks for choosing: - either the tank volume, or the autonomy of the system in days. These parameters are coupled, according to the daily needs of water. - the "Loss of Load" probability (P LOL), i.e. the time fraction during which the operator will accept that the needs are not met (tank empty). These parameters lead to the determination of the array nominal power (i.e. the installed STC 206 power according to the manufacturer specifications), and the pump nominal power required. These are very rough estimations, as the pumping system performances are strongly dependent on the pump technology, head, flowrate, as well as the electrical matching between pump and PV array. The first result graph shows the potentially available solar energy, along with the user's water and energy needs. The second one (available by the speed buttons on the left) gives the average filling state of the tank, and the missing water (P LOL) monthly distribution. The table holds all monthly values, including an eventual required back-up energy. Finally the rough economic evaluation 22 gives an idea of the investment and water price. You can now play with the parameters and immediately see the results. You can print a report, or store graphs and tables in the clipboard to export it to another software. You can also save your project, and load another one for immediate comparisons. Computation: PVsyst performs a very simplified simulation, which runs over one year in daily values. The evaluation of the available irradiance on the collector plane uses the Monthly Meteo 179 tool algorithms, which calculate irradiation's monthly averages on the basis of instantaneous data for one day per month. This is not sufficient to manage the water storage balance evolution from day to day, and the effective use of solar incident energy. Therefore the program generates a random sequence of 365 days, according to the algorithms of Collares-Pereira, renormalised to the monthly sums, for calculating the daily balance from day to day, and the PLOL. This simulation is repeated with different array and pump size arrangements, until matching the input requirements (namely the desired PLOL). The program is able to propose: - The pump(s) size (power), - The PV array nominal power - A rough estimation of the investment cost and the cost of water pumped. If necessary the pre-defined parameters used (array, system matching and pump efficiency, etc.) are usermodifiable through the menu option "Preferences"/"Edit Hidden Parameter" 167 . Of course, this early layout proposition should be asserted by a detailed simulation, using real commercially available components, and taking all system features into account in an hourly modeling.

21

Chapter 3 Preliminary design

Preliminary design
Preliminary design: economic evaluation

Costs are given as specific costs for modules, other components, mounting, financial costs , taxes and so on. Modules and inverter costs are dependent on the quantities. Mounting and maintenance costs follow an exponential rule, referred to a 5 kWp system, with an exponential coefficient of 0.8: Cost = (Pnom / 5 kWp)**0.8 * Basecost (for 5kWp). These hypothetic default financial parameters can be adjusted by the user by choosing "Edit costs" in the economic results sheet. Please be aware that at this pre-design stage costs are based on very coarse hypothesis. They can widely vary from country to country, from time to time or from user to user (what costs are included here? customer or retailer costs? which interventions on the building? designer fees? taxes? etc). The economic evaluation 106 at the detailed simulation stage will offer a flexible and more precise tool for evaluating real costs according to the specific user's criteria. This part is aiming to perform a thorough PV-system design and performance analysis using detailed hourly simulations. These are organised in the framework of a Project, which essentially holds the geographical situation and meteorological hourly data. Optimisations and parameter analysis can be performed through different simulation runs, called variants. Procedure: NB: You have a step-by-step tutorial

24

for elaborating your first project.

After choosing "Project Design" and the system type in the main window, the procedure is the following: First define the Project through the "File" menu.
28

through the "Project/Variant" button. You can also retrieve an existing project

For one Project (including basically Geographic Location and Meteo, with eventual Albedo data), you can construct different system variants (as much as needed). For each variant, define the plane orientation Define the System
29 30

properties.

The program verifies the consistency of all parameters, and produces "Warnings" as Orange (acceptable for simulation) or Red (preventing simulation) LED's. When available (all parameters properly defined, that is only Green or Orange LED's), press the "Simulation 92 " button. Red buttons or warnings indicate bad definitions which prevent the simulation. When the simulation is completed, you will enter the "Results "Report" document.
96 " dialog, and

consult the main results on the

After simulation, each variant may be saved for further comparisons (please use "Save as" for not overwriting your previous variants). You are advised to define a significant description for each variant, in order to easily retrieve them in the list and to obtain a suited title in your final report.

For a given project, you are advised to first construct a rough variant keeping all parameters to their proposed default values. In a second step, you can define the required refinements: In the "System" definition panel, you can modify the "Detailed losses" resistance, module quality, mismatch, soiling, IAM). eventually define a Horizon profile
38 83

(temperature parameters, wiring

(far shadings),

Near shadings 37 , that is partial shadings of near objects, which require a rather complex CAO 3D construction of the PV-field environment.

The next table shows an outline of the project's organization and simulation process. 22

Chapter 3 Preliminary design

Project design
Hourly meteorological data
(Reference Year, TMY, Satellight)

4
Monthly meteorological data (Sites database) Generation of synthetic hourly values Hourly meteo measurements
(Custom, horiz. or in coll. plane)

Project
Specification of the site (geographic coordinates) Hourly meteorological data
Eventually Meteorological corrections (Albedo, Altitude, etc)

Simulation variant
(Many simulation variants may be defined for a given project)

Incident irradiance in the collector plane: Transposition from horizontal values to the collector plane
(fixed plane, tracking 1 or 2 axes, seasonal adjusment, heterogeneous fields) If "infinite" sheds or sun-shields: mutual shadings calculation

Far shadings (Horizon definition)

The far shadings affect the whole field at a time

Near shadings (partial shadings on the field)

Shading of near objects require a detailed 3D description of the system

Shading factor for beam component: "linear" or according to modules Shading factor for diffuse: spherical integral of the "linear" shading factor

User's load Necessary for Stand-alone and Pumping systems sizing

(Optional for Grid-connected)

Constant, monthly or daily profiles, or custom file data

PV array
Choice of PV modules (library) Number and interconnections of modules Specification of Losses (In a second step) :
Module quality, Mismatch, Thermal, Wiring resistance, Incidence angle (IAM)

System
Grid-connected, Stand-alone, Pumping, DC-Grid Choice of the components and configuration

Simulation of the system (by hourly steps)

Results for a simulation version: Complete engineer report - Economical evaluation
Customised Tables, monthly, daily and hourly graphs. - Detailed Loss diagram ASCII export file.

23

Chapter 4 Project design

Project design

Project design: tutorial

Return to the Project design
22

definitions

For your introduction to the elaboration of a project in PVsyst, we will construct step-by-step a full project, located at Marseille (France). The main stages, which you are advised to follow for each of your future projects, are: - Define the project, which holds mainly the geographic location and the meteo file used. - Define a first system configuration with the minimal parameter set, let PVsyst fix other parameters to default (reasonable) values). Simulate and save this variant. - Define successive variants by adding progressively perturbations to this first system. I.e. far shadings, near shadings, define specific loss parameters, etc. You can save each variant for further comparisons.

Definition of the Project

The project is the central object (file with *.PRJ extension) for which you will construct different variants of your system (files with the same name, and extensions *.VC0, *.VC1, etc for "calculation version"). - In the main page of PVsyst, choose "Project Design" / "Grid connected". - In the next window, choose "Project", and then button "New project". - Define the "Project's name". This will appear in the Project's list choices and on the final report. The other definitions in this dialog are only for memory, they are not used elsewhere in the software. You can put for example "Tutorial at Marseille". - Click the button "Site an Meteo". - Choose a site in the "sites" database (contains geographic location and monthly meteo data). We will chose "Marseille". NB: If you want to define your own site, which is not in the database, it is preferable to define it first in the database, using "Tools" / "Geographical Location)". - This automatically produces a synthetic hourly data file named "Marseille_Syn.MET". NB: At this stage you can choose another meteo file in the (*.MET) available data if desired. - Click the "Alb edo" button. Usually you will never modify the albedo factor. The value of 0.2 is a standard adopted by everybody. Nevertheless if you are for example in mountain weather with some snow, you can define one or two months with 0.8. - Click "OK". You will be prompted to save the project file. The description had already defined previously. Give a synthetic identifier for the file name, without special characters like accents. All the files referred to this project will have this name. You can define for example "Marseille_Tutorial". - Save the project.

Creating your first variant for this project

- In the Project's dialog, button "Back (Calculation)". In this window: a red button means parameters to be defined, or not acceptable. An orange button means acceptable, but not ideal parameters. A green lighted button means: Parameters defined, correct. A green out button, that this topic is not defined for this version. The "Simulation" will be enabled only when no button is red. - Click "Orientation" button: choose "Fixed tilted plane", Tilt = 25, Azimuth = 20 (toward east). NB: When defining the Orientation, you have a tool "Show optimization" for situating your choice by respect to optimum (regarding the transposition factor). The optimum is different for grid-connected systems (optimization on the whole year) or stand-alone/pumping systems (optimization on the more critical period). - Click "System " button: definition of the PV system. Will open the system definition dialog. - In the "Presizing Help" group, define "Availab le Area" = 125 m. - Choose a PV module. The list will show "All modules", or your "Favorites", or other categories according to 24

Chapter 4 Project design

Project design

availability. You can sort them by manufacturers, and then choose for example a module of Photowatt, say the PW1000/110W. - Choose an inverter. The comment in the "Design the array" will propose a suited power for the inverter(s). In this project we will choose 3 inverters of 4.2 kW, for example Diehl, Platinum 4800 TL. NB: the button "Show sizing" will open a window where all the sizing constraints are apparent. Here we see that the maximum voltage of the PV module (here 600V) is a strong limit for this system, far below the Inverter's VmaxAbs (880 V) or even Vmppmin (710V). - Now the warning panel doesn't show any warning (in orange: not optimal, in red: would prevent the simulation). Only by defining the size of the system and the components, PVsyst has proposed an acceptable design for your PV system. You can click "OK".

Executing the simulation

- You can now click the "Simulation" button, which is now activated as there are no red buttons. - Give a significant description of this variant (for example "First simulation"), - Press the "Simulation" button, and at the end "OK". - Click button "Results", and then "Report". Results This shows the report of your first simulation, which has now 3 pages: - First page: all the parameters for this simulation, - Second page: a reminder of the parameters, the main results, graphics of normalised values 100 , and table of monthly values, - Third page: an arrow loss diagram 97 , showing an energetic balance and all losses along the system. This is a powerful indicator of the quality of your system, and will immediately indicate the big sizing errors if any.

Saving your simulation

You should now "Save" this first variant for further comparisons. This will be saved as " Marseille_Tutorial.VC0". Be careful: use "Save As" for avoiding overwriting your previous variants if any. For opening previous simulations of the project, click "Load" just above.

Adding further particularities of your project

After this first "standard" simulation, you can progressively add the specificities of your particular project. You are advised to perform a new simulation at each step in order to check its effect and pertinence - especially by analysing the "Loss diagram".

Far shadings, Horizon profile

This is only suited for shading objects sufficiently far from your PV system, so that the shadings may be considered global on your array. The distance of the shading object should be, say, over about 10 times the PV system size. The Horizon Profile is defined as a curve, set of (Height, Azimuth) points . The Far Shadings operate in an ON/OFF mode: i.e. at a given time, the sun is or is not present on the field. When the sun is behind the horizon the beam component becomes null. But the diffuse component, supposed isotropic - coming from all points of the sky vault seen by the collectors with an identical intensity - is not much affected. Clicking the "Horizon" button will open a graph of the sun paths at your location.

25

Chapter 4 Project design

Project design

Defining horizon line by hand: The horizon (Height, Azimuth) values may be recorded on-site using a compass and a clinometre (measuring the height angles) or a land surveyor instrument. You can move any red point by dragging it with the mouse, or define accurately its values in Edit boxes. For creating a new point right-click anywhere. For deleting a point right-click on this point. You can of course save this horizon as a file fur further use in others projects in PVsyst. Importing Horizon from Solmetric "SunEye" instrument The "SunEye" records the horizon line using a fisheye camera, and provides results as files. You should choose the "Ob structionElevation.csv" file. Don't use the "Sky0x_PVsyst.hor" file, which was elaborated by Solmetrics for the old versions 4.xx of PVsyst. NB: You should correct the data if near objects are represented in the record. Importing Horizon from the "Carnaval" software Carnaval is a georeferred software (including altimetric data) able to establish an horizon line for any place given by its exact geographical coordinates - Latitude and Longitude - in France and neighbours. You should not use the near objects option in this software for establishing far shadings PVsyst masks. Carnaval produces a file named "YourProject.masque.txt". Please rename this file by removing the ".masque" characters as PVsyst doesn't accpt file names with 2 points. Importing Horizon from the "Horiz'ON" software The "Camera Master" tool is a specialized support for your Photo Camera, which allows to take several photographies with an horizontal reference (every 20 in azimuth). The software "Horiz'ON" gathers these photos as a single panorama on which you can draw the horizon line by mouse. This produces a file directly readable in PVsyst. NB: when established using a georeferred grid (like Carnaval or Meteonorm), the exact location should be carefully defined. This may be determined using GoogleEarth or a GPS. Please remind that a degree in latitude is 111 km, a minute is 1850 m and a second is 31m.

Near shadings, 3D construction

If you have near shading masks, you should construct your PV installation and the neighbours as a 3D scene (see the dedicated tutorial 40 ). The previous instruments (including SunEye) cannot help for this construction, which should be realized using architect's planes or equivalent, including heights indications. You will performe the simulation with linear shadings (irradiance deficit, lower estimation of the shading effect), and then "according to module strings" (upper bound of the electrical effect, higher estimation of the shading losses). Finally, for the definitive report to be transmitted to your customer, you can choose an intermediate value for the electrical effect, taking the by-pass diode recovery into account. For this you have to choose an intermediate fraction of electrical effect, which will depend on your system geometry. But we dont have well-established value 26

Chapter 4 Project design

Project design
to propose at the moment (60 to 80% ??? - Higher for regular shadings like shed mutual shadings !).

NB: The near shading loss doesn't cumulate with the far shadings: when the sun is behind the horizon, the beam component is null, and therefore there is no near shading contribution. Final layout of your system There is in principle no relation in the program between the definition of the system, and the definition of your 3D scene. Except that at the end of the process the program will check the compatibility of these different parts. Namely it will require that the plane orientations are identical, and that you have defined a sufficient sensitive area in the 3D scene for installing the PV modules defined in your system. But PVsyst doesn't check the real physical (geometrical) compatibility. In a second step, you should check the arrangement of your modules, and eventually modify the system definition (number of modules in series and parallel) in order to build a realistic installation. PVsyst doesn't offer specific tool for this try-and-error process in the present state.

Array losses
Finally there are several parameters which are fixed by PVsyst as reasonable default values for your early simulations, but that you should modify according to the specificities of your system. These are available in the "System" definition dialog, button "Detailed losses". Thermal Losses The thermal behaviour of the array is computed at each simulation step, by a thermal balance 84 . This establishes the instantaneous operating temperature, to be used by the PV modules modelling. The thermal balance involves the "Heat loss factor" U = Uc + Uv WindSpeed [W/mK]. In practice we advise not to use the wind dependency, as the wind speed is usually not well defined in the meteo data, and the Uv is not well known. Therefore we put Uv = 0 and include an average wind effect in the constant term. According to our own measurements on several systems, PVsyst proposes: - Uc = 29 W/mK for complete free air circulation around the collectors ("nude" collectors). - Uc = 15 W/mK for integration (back insulated), as only one surface participates to the convection/radiation cooling. - We don't have well-established values for intermediate situations with back air circulation. Our measurement on quasi-horizontal modules on a steel roof, 8 cm spacing and not jointive collectors, gave 18 W/mK; NB: up to the version 5.1, the default value was 29 W/m (free standing). From this version the default is set to 20 W/m as more and more installations are integrated. The thermal loss effect is visible on the array loss diagram. Wiring Losses The wiring ohmic resistance induces losses 87 ( R I ) between the power available from the modules and that at the terminals of the array. These losses can be characterised by just one parameter R defined for the global array. The program proposes a default global wiring loss fraction of 1.5% by respect to the STC running conditions. But you have a specific tool 87 for establishing and optimizing the ohmic losses (press "Detailed Calculation" button) . This tool asks for the average length of wires for the string loops, and between the intermediate junction boxes and the inverter, and helps the determination of the wire sections. NB: remember that the wiring loss behaves as the square of the current. Therefore operating at half power (500 W/m) will lead to only a quarter of the relative loss. The effective loss during a given period will be given as a simulation result and shown on the loss diagram. It is usually of the order of one half to 60% of the above specified relative loss when operation at MPP. This part also includes the definition of eventual losses between the output of the inverter and the injection point (energy counter). You have just to define the distance and the loss will also appear in the loss diagram. Module quality loss The aim of this parameter is to reflect the confidence that you put in the matching of your real module set performance, by respect to the manufacturer's specification. The default PVsyst value is half the lower tolerance of the modules. The result in the "Array loss diagram" is not exactly the specified value, at this is referenced to the STC when in the diagram value it is by respect to the previous energy. 27

Chapter 4 Project design

Project design
Mismatch loss

Losses due to "mismatch" are related to the fact that the real modules in the array do not rigorously present the same I/V characteristics. Remember that in a string the worst module drives the string's current. The button "Detailed computation" helps the understanding of this phenomenon, and gives indications on the loss parameter to be set for the simulation, according to your hypothesis on your effective module set. This parameter acts as a constant loss during the simulation. It is lower for thin film modules. It could be about null when sorting the modules according to their real performance (flash-test results provided by the manufacturer). NB: There is probably a correlation between these 2 last parameters. The Module quality loss is rather related to the average of the module's distribution, when the mismatch refers to its width. Soiling loss To our experience, the soiling effect is about negligible in middle-climate residential situations. It may become significant is some industrial environments (for example near railway lines), or in desert climates. The monthly definition allows for taking periodical cleaning or rains into account. This parameter may also be used for describing a snow effect (for example put 50% in winter month with 15 days coverage). IAM loss The incidence loss (reflexions due to the fresnel's laws) is sufficiently well defined by parametrization proposes by the "Ashrae" (US standards office). You will in principle never modify this parameter. NB: Assuming an isotropic diffuse, the IAM factor on the diffuse part is computed by an integral over all spaces directions, which include important low-incidence contributions.

Project definition
When creating a project you have to define: The project name, which will identify the project in the file list in your data library. The PV system customer address (facultative, not used elsewhere in the software), The geographical location The hourly meteo file
115 114

in internal PVSYST format (*.MET files).

29

NB: These are defined in the Project Situation and Meteo and meteo file may be of different kinds. -

dialog. The relations between Geographical site

199

Finally the "Next" button leads to the definition of the albedo

characteristics of the project site.

The project parameters are stored in a file with the extension .PRJ. All variants will have the same with extensions .VC0, .VC1, etc.

Site and Meteo data in the Project

The project is located at a geographical Site and meteo in Monthly values.
114

. Sites are described in *.SIT files, with geographical coordinates

The detailed simulation process requires the following meteorological hourly data: - Horizontal Global Irradiance - Horizontal Diffuse Irradiance (optional, may be constructed by a model), - Ambiant Temperature, - Wind velocity (optional, for module temperature calculation). These data are stored in meteo files
115

, written with the internal PVSYST format ( *.MET files).

When creating a project you will usually first choose a Site in the database; By default, if no matching meteo file (same name as the project site) is available in the library, a Synthetic Hourly Data 181 file will be automatically generated from the monthly meteo data defined for the Project site.

Chapter 4 Project design

28

Project design

But you can also choose an existing meteo hourly file in the library (for example, 22 Design Reference Years for Switzerland in the PVSYST library). You can also obtain Hourly Meteo files in several ways, all available in the "Tools" part of the software: Import an synthetic hourly meteo file from the Meteonorm For the US, import a US TMY2
128 126

software,
129

file (Typical Meteorological Year) available from the web (free), project (free).

For the whole Europe, import measured data from the Satellight

Convert any custom meteo ASCII file 137 (your own measurements, or Design Reference Year of other sources, etc.), using the PVSYST special converting tool.

Project Site and Meteo

When creating a project 28 the Geographical Site and Meteo may be defined in several ways. Please remember that a geographic site is attached to the project (your project's location) and another geographical site is defined within the hourly meteo file. These sites are not necessary identical. When different, the climatic distance will be mentioned; this should not exceed some dozens of km. Standard procedure - Choose a geographical site in the database. - If a corresponding meteo file (same site name) is available in the Meteo database, this will be automatically selected. - Otherwise a synthetic hourly file will be automatically created according to the monthly data of your site. Standard procedure with customised site name - Choose the nearest available geographical site in the database. - Modify the site name (which will appear on the final reports for the customer) by opening the site. This will of course keep the meteo monthly values of the original site. - You can also modify (not too much, say < 1-2) the coordinates (Latitude/Longitude). These values will be taken as basis for constructing the eventual synthetic meteo file, and in the solar geometry (namely transpositions) within the project. - If a corresponding meteo file (same site name) is available in the Meteo database, this will be automatically selected. - Otherwise a synthetic hourly file will be automatically created according to the monthly data of your site. Starting from a meteo data file - Choose the desired Meteo file in the database. - Copy Meteo Site => Project Site. - You may change the site name and coordinates if desired. In this dialog: - The opportunity of copying the site from a meteo file to the project, or inversely to generate a synthetic hourly file from the Project's site (normally performed automatically), are available as soon as the location coordinates are different. - If you need other meteo data , please define them in the database, using "Tools" / "Geographical site" or "Import meteo data". - When different sites, the climatic distance
201

between the sites is shown.

200

- If the altitudes are very different, altitude corrections project's site altitude.

may be applied to your Meteo data for matching the

System definition
The system definitions are primary aimed to define all the PV_system components necessary to fulfil the user's 29

Chapter 4 Project design

Project design
wishes.

In a second step, this part gives the possibility of modifying the PV-array specific loss parameters (thermal, wiring resistance, module quality, mismatch, IAM) (button "Array losses"). All these parameters are initialised at typical default values, so that first calculations give likely "average" results. In stand-alone systems the user's needs definition are absolutely necessary. They closely condition the system design. For grid systems, these can be defined afterwards in order to obtain the user self-consumed energy part and the energy reinjected into the grid. In DC-grid systems, especially for public transport, the instantaneous need could be lower than the produced energy at some instants, so that there will be some unused energy. Parameter definitions are of course different according to system types: Grid-connected systems 58 , Stand-alone systems 63 , Pumping systems 66 . DC-grid connected systems 64 ,

Plane orientation
PVSYST supports simulations with many plane orientation modes: Fixed tilted plane: You have just to define the Plane tilt
205

and azimuth

205

Seasonal tilt adjustment: the plane tilt may be adjusted with two values, for winter and summer chosen months. Tracking, two axes: the limit mechanical angles of the tracking device (in tilt and azimuth) should be defined, and are taken into account during the simulation. Tracking, two axis with frame: the collectors are fixed and rotating within a frame, itself rotating. Two configurations are available: a frame with North-South axis (and collectors with tracking tilt) or a frame with East-West axis (and collectors tracking according to the sun azimuth). Tracking, tilted axis: the axis's tilt and azimuth should be defined (the axis azimuth will usually be around 0, i.e. near the south in northern hemisphere). The rotation angle is called Phi (value 0 when plane azimuth = axis azimuth), with the same sign conventions as for plane azimuth 205 . Limits on the Phi stroke are required. Tracking, horizontal axis E-W: the orientation axis is defined as the normal to the horizontal axis. Stroke limits should be defined (here Phi = plane tilt), from lower limit (minimum -90 = vertical north) to upper limit (maximum 90 = vertical south). This configuration is indeed not suited for PV systems. Tracking, horizontal axis N-S: this is the usual configuration of horizontal axis tracking systems. You should use the "Tilted axis" option, with axis tilt = 0. Tracking, vertical axis: the collector is kept at a fixed tilt, but rotating according to the sun azimuth. This configuration may be used with "dish" arrangements, when a big rotating support holds several rows of modules; this particular case is made possible as the rotating axis of one row may displaced by respect to the collector. Tracking sun-shields: which may yield solutions to the difficult optimisation between sun protection and PV production. For full efficiency this should involve a Backtracking 51 control strategy. Double orientation: allows to define two collector planes with different orientations. You should define the fraction of the field allocated to each orientation. Warning: the PV modules of a same string should all have the same orientation. The electrical behaviour of modules connected in series and having different orientations is very complex and not functionally optimal (see the specific related tool) and is not supported in the simulation process. Unlimited sheds: 31 To be used when the sheds are very long by respect to their width. If the sheds are too short as one cannot neglect the edge effects, you should define sheds in the "Near shadings" CAO option instead. Unlimited sun-shields:
33

Same remarks as for sheds. The optimization of electrical yield of sun-shield

Chapter 4 Project design

30

Project design
systems is very difficult, and only suited for south faades.

Limitations
The tracking strategy in only computed using the solar geometry (so-called "astronomical" algorithms). A simulation using a strategy with irradiance optimization on the collectors is much more complex to implement, and is not yet available in PVsyst. Please note that the near shading calculations cannot be applied to Seasonal Tilt adjustment. For double orientation (heterogeneous fields), the shading factor is computed once for both fields at a time, so that it becomes not well applicable (induces errors) when the difference in orientation is too big. For sheds, please carefully see the special combination of "Orientation" option and "Near shadings" treatement
48

Also be aware that with tracking planes, the mutual shadings of several neighbour tracking units can become very important at extreme angles (see the "shed" optimization tool 31 with very tilted collector plane!). The collector's spacing should usually be very large, so that the horizontal space use is rather low. Also Backtracking 51 control strategy may help optimizing the electrical yield.

Heterogeneous Fields
It is commonly admitted that in a PV field, all the collectors should be perfectly identical and have the same orientation. This tool allows the phenomenological analysis of the effect of not observing this rule. It shows the characteristic of two fields, connected in parallel, which can be different in orientation and collector's number. A rigorous energy analysis is also possible in the framework of the simulation. In "Project design", and "Orientation" parameters, choose the Field type: "Doub le orientation". When connecting together the outputs of 2 different arrays, the resultant characteristic will depend on incident irradiances on each of the sub-fields. It is therefore necessary to introduce a model for the irradiance, in such a way as to be able to evaluate simultaneous irradiances under realistic conditions along the day. The tool uses a clear day profile, but with the possibility of modulating the global amplitude and the rate of diffuse irradiation, to approach the meteorological conditions of any day. The temperature of the modules is calculated according to the respective irradiances. The user may use the scroll bar to modify the time-of day, in order to evaluate the dynamic behaviour during the day when the orientations are different. The graph shows the respective I/V characteristics of each sub-field, and their resultant (current sum) when connected in parallel. The comment gives the nominal MPP value of each array, as well as their common value and the relative loss when connected in parallel. One can see that when the array voltages are comparable the power loss is usually low, even for very different currents (different orientations along the day, or different parallel strings). In this case the performances of each array are simply added together. But for different voltages the resultant characteristic shows two distinct maxima, with a serious loss of power. This could also induce the MPP tracking device into error, as it may "choose" the secondary maximum. When the arrays are expected to operate under different voltages (heterogeneous arrays, but also by partial shading effects), it is also very important to connect blocking diodes in each string. The dotted line shows the resulting behaviour if these diodes are omitted: the production of the higher array may flow into the lower one, inducing a feeding power into the "overvoltage" region.

Sheds mutual shadings

PVSYST offers a special tool for viewing and optimising the shed mounting (and also sun-shields). This is implemented in several places in the software: In "Tools" / "Tab les and Graphs of Solar parameters", you can define the geometry, optimise it with the mouse, and view the mutual shading effects of your choice on an Height/azimuth diagram. In "Tools" / "Monthly Meteo Computations": you can moreover have a quick meteo calculation for your site, and immediately estimate the irradiation losses over the year. During the definition of the project's parameters, the "Orientation" choice offers the opportunity of defining sheds or sun-shields. This will of course take the corresponding mutual shadings into account during the simulation process (on beam as well as diffuse component).

31

Chapter 4 Project design

Project design

With these tools, the mutual shading of sheds (or sun-shields) is performed using a simple geometrical computation, which gives the mutual shading fraction as a function of the Transverse Incidence Angle 207 or "Profile angle" in the shed orientation. This calculation is performed assuming sheds of "infinite" length (that is, it doesn't take the edge effects into account). The calculation may be performed for irradiance, i.e. the shading factor is the shaded area fraction of the full array (depends only on the pitch and tilt angle). This also slightly depends on the number of sheds as the first one is not shaded. This is what we call the "linear" shading. You can also estimate the electrical effect of these shadings. Remember that the current of the whole string (as seen from the inverter input) is limited to the current of the weakest cell, i.e. the shaded one. Therefore we suppose that the production of the bottom string becomes null (regarding beam component) if the bottom row of cells is fully shaded, and proportional to the shaded fraction of the cells otherwise. This is what we call shading "according to module strings" in the near shadings. Therefore this calculation requires the specification of the number of strings in the transverse dimension of the shed, as well as the size of one cell. Please note that with thin film modules with "long" cells of around 10 mm width, the cells should be placed in the transverse dimension of the shed, so that only a little part of each cell is shaded at a time. The three parts of the shed tool: geometrical design, shading graph and yearly yield graph, are meant for a better understanding when optimising 32 a shed layout. Please carefully see the limitations when using sheds with near shading scenes
48 .

Shed optimization
Shadings on Beam and Diffuse First of all, when optimising sheds, one should keep in mind that the shading losses are of two kinds: - The shadings which we usually visualize, are applied on the beam component. They apply especially on some periods of the day or the year, and often when the sun is not very "high" on the plane surface, therefore affecting hours of rather low production (there can be an exception with mutual shading of sheds, reinforced by the "electrical" effect of cell partial shadings). - The shading on the diffuse component, which applies permanently with the same value when assuming an isotropic diffuse distribution. This shading factor is related to the part of heavens which is "seen" by the collectors. Remember that in central Europe climates, the Diffuse fraction is of the order of 50% of the total incident irradiation ! In the shed disposition, the diffuse effect is particularly important as the heavens vault view is limited in front by the previous shed (affecting rather high incidence angles) and to the rear side by the plane itself. Due to its permanent effect, this is often the main part of the losses. This situation is even much more sensitive with sun-shields. General features and optimisation of sheds With shed disposition, we should be aware that: - The area occupation is strongly depending on the collector tilt. For acceptable shadings, the "Limit shading" profile angle 205 should be kept below about 18 to 20. With 30 collector tilt, this implies that you can only install a collector area limited to 45% of the total available area. - The mutual shading effect is also strongly dependent on the shed orientation: when not south, the morning or evening performances are much more affected. The software offers two complementary tools for well understanding these situations: - The beam shading graph, which shows the periods which are affected by the shadings. One can notice that with south orientation, the effects are rather limited to summer morning and evening. But as soon as you change the azimuth, the shading losses increase rapidly, and arise especially in regions where the incidence angles are high. - The yearly yield graph shows the relative gain (by respect of the horizontal layout), as a function of the module tilt. This factor is based on the yearly global useful irradiation (beam and diffuse) falling onto the collectors. The

Chapter 4 Project design

32

Project design

curves show the pure transposition yield (as if there were one only plane) and the curve with mutual shadings, either "linear" and for electrical losses. This graph can be drawn with two options: - Either keeping the Incidence limit angle constant. In this case we see that the collection/ground ratio is very sensitive to the plane tilt. - Or keeping the Shed spacing (i.e. a chosen collection/ground ratio) constant. In both cases, we see that the optimal tilt of sheds is lower than for a single plane. This is still more evident with non-south azimuths. From these tools we can observe that: With shed installations, choosing a rather low tilt is often a very good solution, which leads to acceptable losses of some few percent by respect to the optimum, with the following advantages: - The installable power is much greater on a given ground area. - The array orientation does no more affect the performances: you can install the sheds according to the building geometry, special "south-facing" arrangements are no more useful. - Module supports become more simple, cheaper and lighter, with less wind sensitivity. - Architectural impact can be much more acceptable. Nevertheless, the minimum tilt should be kept at some few degrees (2-3) for module cleaning by the rain. Moreover, frameless laminates should be used in order to avoid dirt and mosses accumulations on the bottom side.

Sun Shields mutual shadings

In the "Orientation" option and for near shadings, sun-shields are treated in a similar way as sheds (see Shed Mutual Shadings 31 discussion). Nevertheless, the user should be aware of two phenomena: With sun-shields, the accepted part of the diffuse irradiation is very limited: on one hand, there is at most only one half because of the rear wall, and on the other hand the upper sun-shield also cuts an important fraction of the remaining diffuse component. In regions with a high diffuse fraction (above 45 latitude, it is usually more than 50%), this gives a very significant loss. Please be very careful with orientations not exactly facing the south (or north in southern hemisphere). You can use the special tool "Shading graph" to visualise this dramatic effects.

For simulations, if you define near shadings in the "Orientation" parameters and you want to define other near shadings, please consider the collector plane in the Near shading scene as a vertical plane on the wall.

Orientation optimisation tool

When choosing the (fixed) plane orientation, an information panel indicates the corresponding Transposition Factor, the difference (loss) by respect to the optimum orientation, and the available irradiation on this tilted plane. Clicking the "Show optimization" button, you can see a graph of the Transposition Factor as a function of the plane tilt and azimuth. These graphs also indicate your actual choice by a violet dot on the curves, showing at once where you are positioned by respect to the optimum. The optimisation of the orientation depends on the planned use for the PV energy. For grid-connected systems, the energy is usually sold at a constant price all over the year. The relevant optimum is then to maximize the yearly energy. For stand-alone systems, the relevant solar yield for sizing the system is usually the winter months. For pumping systems, there is no general rule, depending on the final use of the water pumped: household (all over the year) or irrigation (some specific seasons, probably summer).

Therefore this tool gives the opportunity of choosing the optimizing period: Year, Winter or Summer (this option is fixed according to the system type in the presizing part). NB: This tool makes use of the Monthly Meteo calculations 179 , which perform quick transpositions from the Monthly Meteo values. The whole calculation for the curves is performed for several situations, in both directions from the point you have chosen (tilts at fixed azimuth, and azimuths at fixed tilt). 33

Chapter 4 Project design

Project design
Concentrating systems

Systems involving concentrating devices are not treated in whole generality in PVsyst. Some specific features have been implemented from version 4.2, for evaluating especially high-concentrating systems. But their accuracy is not quite established. The general study of concentration systems involves a detailed description of the irradiance distribution, which cannot be available using the present treatment of the Meteo, nor the limited information included in the meteo database (site database). Namely high concentrating performances require a good knowledge of the beam component. Then accurate models for achieving this evaluation would involve parameter like turbidity, water and aerosol contents of the atmosphere, which are not defined in our database. As an example, the eruption of the Mt. Pinatubo in 1991, had little effects on the global irradiance (less than 2%) but induced a very high loss of beam component: the pure direct was scattered by aerosols, resulting in a strong halo around the sun, up to 30% during almost 2 years. This had dramatic consequences on the productivity of the high concentration thermal plants all around the earth [Molineaux 1996] In PVsyst, the simulation of concentrating systems has to deal with 2 aspects: Acceptance of the Diffuse component The higher the concentration, the lower the acceptance angle, which implies a limited acceptance of the diffuse component. The maximum achievable Concentration Ratio CR is related to the acceptance half-angle CRmax (1axis) = 1 / sin CRmax (2axis) = 1 / sin (Kreith and Kreidler, 1978, p 248). In the present state, PVsyst is only able to treat 2-axis high concentration systems. General treatment of low concentration systems (especially 1-axis parabola or "Compound Parabolic Concentrator" CPC) would imply a very detailed description of the optical system, as well as a good knowledge of the irradiance distribution by any weather conditions, and would result in inhomogeneous irradiance on the PV receiver, which is very difficult to take into account when not one only cell. Therefore concentration system parameters are only proposed in the 2-axis tracking dialog, where you have to define: - The diffuse fraction to be taken into account in the simulation (usually near 0 with high concentration). - The acceptance angle for a full efficiency (half-opening angle, i.e. the angle between incident and optical axis). - The limit angle at which the efficiency falls to 0. Besides tracking errors (which cannot be taken into account), this will be useful when the array is reaching its tracking mechanical limits. The simulation will perform a linear decrease between the acceptance and limit angles, and will accumulate the corresponding tracking loss. Electrical behaviour of the concentration device In a 2-axis high concentrating system, the PV sensitive device is usually a single multi junction cell of some few cm2, with very high efficiency. This receives a flux of the order of 500x suns (50 W/cm), and therefore works in a domain where our PV one-diode model is not well attested. We can just notice that the logarithmic behaviour of the open-circuit voltage with irradiance - which is a straightforward result of the model - favours the efficiency at these running conditions. The device is mounted on a heat spreader (passively cooled), which ensures an acceptable operating temperature, of the order of 80C or less. Electrical modelling in PVsyst We did not develop a specific model in PVsyst for such a configuration. But some manufacturers of concentrating devices
35

as:

use to give performance data for their whole

Chapter 4 Project design

34

Project design

component - including concentrators - in a way comparable to usual photovoltaic modules. That is, when a set of concentrators with their PV devices is assembled as a module, they give the usual parameters Isc, Vco, Imp and Vmp, referenced to the irradiance on the aperture area (sometimes under 850 W/m instead of 1000 W/m ). The I/V curve is very sharp, with an excellent fill factor. Then we do the hypothesis that the standard one-diode model applies to this system, even though it has no real physical meaning. This is motivated by the sharpness of the I/V characteristics, and by the fact that our model allows to set a customised temperature coefficient, as required by the manufacturer. Therefore power behaviour according to irradiance and temperature - necessary for the simulation process - should be close to the reality. In this phenomenological model any optical aberrations are neglected. NB: By respect to standard systems, such high concentration systems suffer of two main loss sources, the negligible diffuse acceptance (diffuse is of the order of 30% even in most sunny regions, 40 to 50% in the middle Europe climates) and the full loss when reaching the tracking limits of the heliostats. These appear of course on the PVsyst Loss diagram. On the other hand, in the system definition, the heat loss factor Kc should be set according to the effective sensor temperature, reached under nominal irradiance and which should be specified by the manufacturer (some equivalent of the NOCT data).

Defining a concentrating system

A concentrating system is defined when involving a concentrating PV module (CPV). Defining a CPV module in PVsyst is done in the PV module definitions, part "Size and Technology". Only high concentrating CPV devices (with a concentrating factor of the order of 500x) are possible in PVsyst. These are usually equipped with very high efficiency concentrating cells (tripple junction GaInP2/ GaAs/Ge). The optical characteristics and real electrical behaviour of the cell cannot be modelled in detail. Therefore, as an approximation, the full CPV module is treated as it was a flat PV module, with a sensitive area equivalent to the optical aperture area. The thermal behaviour - with heat sinks for evacuating the heat of the cell - is supposed to behave in the same way as the flat plate modules, i.e. according to the usual thermal balance 84 equation with a heat loss factor U as parameter. Now there is no general rule for the determination of this this thermal factor, which depends on the heat sink properties. Therefore this should be specified by the module manufacturer. The CPV module is only able to use the beam component. A concentrating factor of 500x means an ideal acceptance angle of 2.5 at most (less than 1 in the reality) so that the diffuse contribution is completely negligible. This is the reason why the industry of CPV modules has fixed the STC irradiance value at 850 W/m instead of 1000 W/m for usual modules (but still with a cell temperature of 25C). When specifying a CPV module in PVsyst, you have the opportunity of defining specific corrections, namely for the spectral behaviour. This correction is named the CPV Utilization Factor. For applying this correction, you can define - in the PV module model parameters - a double-linear behaviour according to: - The Beam normal value (DNI) - The air mass (the CPV module will always operate in clear day conditions), - The ambient temperature (not for spectral response, but for accounting of optic's mechanical deformations). Again, these corrections are determined from long-term measurement in real conditions, and should be given by the manufacturers.

35

Chapter 4 Project design

Project design
PV array electrical behaviour

This option presents a set of graphic tools, intended for a better understanding of the electric phenomena involved when connecting several modules together in arrays. Electrical connections of non-identical modules are not trivial. They lead to complex I/V characteristics shapes and current flows, which are far from being intuitive. They often involve the reverse characteristics 185 of the cells in the module, which has to be modelled. In PVsyst, this reverse cell model is only involved in the frame of these phenomenological and didactic tools; moreover in the practise, modules are protected again severe cell reverse bias through bypass diodes. Therefore the exactness of this model is not crucial. The four graphical tools are: Reverse characteristics Array or Cells Mismatch
186

of cells or modules, effects and calculation,

37 , hot spots, effect of bypass

36

Array or Module with one shaded cell Heterogeneous arrays

31

diodes,

connected in parallel.

Arrays with characteristic's mismatch

This tool allows for the phenomenological study of the resultant I/V characteristic of a module or PV array, composed of non identical cells or modules. The program simulates the connection of any number of elements in series and in parallel - by affecting to the I/V model parameters of each element a random dispersion. The user can choose between a normal (gaussian) distribution, or a square distribution between 2 limits. The elements can be cells, assembled in a module, or modules, protected with by-pass diodes. The parameters that can be modulated are: - the short-circuit current Isc (analogous to a non-homogenous irradiance distribution), - the open circuit voltage Voc (which can also reflect temperature differences). The programme calculates each characteristic according to the standard model, and then adds up point-by-point the voltages of the elements in series and the currents from series in parallel. The user can visually follow these operations. He then obtains the overall resultant characteristic of the field, and the program traces the "mean" characteristic (corresponding to elements, all of which are identical) and two envelope-characteristics which can be chosen as 2-RMS values, or as extreme random values encountered in the sample. The program evaluates the Power loss at maximum power point, and at a fixed operating voltage, with respect to the nominal case. NB: The parameter dispersion being random, two successive executions of the same process will never give the same result ! You can choose the 3 following modes: Group of Cells corresponds to the behaviour of the chosen PV module according to its cell's dispersion. Usually in a module, all cells form one only chain (sometimes two or more), therefore only the current dispersion is relevant. For such a module, one can see that the resulting characteristics is strongly influenced by the cell with the worst current, resulting in the flattening of the current plateau just below the maximum power point. In such a figure a bad cell may work in its reverse characteristics region (that is with a negative voltage) on part of the current plateau. Remark: You will understand here the difficulty in exactly representing the operating of a real module with the help of usual models describing single cells, and that the use of too sophisticated cell-models (i.e. two-diodes models) will not improve the situation if they do not include this statistical distribution. Group of modules simulates a whole array. In this case the resulting figure looks quite different, with a "bumped" shape all along the plateau. This is due to the by-pass protection diodes, supposed to be always present in the modules. These give usually even better performances than nominal modules below half the nominal current, but degrade until the Maximum Power Point. One can see that the MPP power is much less affected than at fixed voltage operation below the MPP point. Remark: in the region of low voltages, some modules are operating in the reverse polarisation region. The by-

Chapter 4 Project design

36

Project design

pass diode blocks the reverse current which normally would flow (be consumed) through the cells. This is the reason why the performances are better. Without diodes, the characteristics would show a linear plateau analogous to the cells behaviour in a single module, near to the worst module characteristics. Group of modules with sorted modules. If we sort the PV elements to put them in increasing order of short-circuit current, in such a way that each series comprises modules with close characteristics, one can see a quite different behaviour. In this case, the diodes are no longer involved and the curve again becomes perfectly smooth. Each string behaves according to the average of its modules; but connecting them in parallel results in a characteristic very close to the average. This confirms that sorting the modules before mounting them in series can significantly improve the performances of an array, especially when working at fixed voltage.

PV module / array with a shaded cell

This tool visualises the behaviour of voltages and currents at the terminals of a shaded (or bad) cell, placed in a module or a group of modules. It evaluates the energy dissipation of this cell for various operating points, and its resultant temperature by making reasonable hypotheses for the thermal loss factor. It has to be noted that these behaviours are studied in a marginal situation - representing the most unfavourable case - where only one single cell is disturbed. The user can adjust the irradiance, the module temperature in normal operation (i.e. without dissipation), and the shading fraction of the considered cell. You can visualise three different cases, for a single module or arrays: Module without protection diode: the graph shows the degraded performance curve of the whole module (grey), to be compared to the normal operation (dashed line). But it also shows the dangerous reverse bias applied on the shaded cell, for example at the V=0 point (short circuited module), or at the Maximum Power Point operation of the whole array. The reverse bias voltage occurs along with a current greater than the nominal photocurrent, resulting in a serious power dissipation of the shaded cell, resulting in its temperature increase. This dangerous situation, named "Hot Spot", can be explained as each cell in the module is producing its rated current at bright sun, with its normal voltage, all of this power being dissipated in the shaded cell. (as they are connected in series, the current flow is identical in each cell, and the sum of voltages (shaded+non-shaded) is the module overall voltage, i.e. zero when short-circuited). Module with protection by-pass diode: this graph also shows the regulating action of the protection diodes connected in antiparallel to a module or two half-modules. In this case the reverse voltage at the module's terminals is limited to the diode voltage (about 0.7V), the excess current being drawn by the diode. In this case the power dissipation of the shaded cell is limited to the power production of the other cells inside the protected series. If too much cells in series, the module's manufacturer will distribute several diodes along the cells string. These behaviours help understanding that the effects of partial shadings on an array give rise to highly non-linear electrical behaviours, and that apparently negligible causes can lead to highly significant losses. This graphical tools should lead the engineer to carefully study the wiring of the PV fields, in such a way that the foreseen shadings simultaneously affect the smallest number of series possible. For example, in shed mountings, he should place the module series longitudinally, in order that the mutual shadings (on the lower part of the back-shed) be limited to just one branch of cells or modules.

Shadings: general organisation

We distinguish between two fundamentally different types of shadings: Far shadings are the horizon 38 effects. Far shading acts on the PV field in a global way: at a given instant, the sun is or is not visible on the field. Near shadings are partial shadings which affect only a part of the field. The shaded part changes during the day and over the seasons. We call shading factor the ratio of the illuminated part to the total area of the field, or inversely shading loss is its complement.

The treatment of near shadings is the most complex part of the PVSYST software. For beginners, you have a tutorial 40 which explains the main procedures on an example.

General procedure
Near shading calculations necessitate the reconstitution of the exact geometry of the PV field and its environment, 37

Chapter 4 Project design

Project design

in the 3D-space. You have first to build the global scene 48 of the PV system by assembling parametrised elements (PV fields, shading obstacles, buildings, trees) which can be adapted from template shapes. Assembly is done in a global perspective or plane view. Once this scene has been well established you can visualise shadows 53 produced for any sun position or time-of-the-year. You can also run an animation of the shadow evolution over a given day. During simulation, the calculation of the shading factor for each hour would spend too much computing time. Therefore the programme establishes a table of shading factors 45 as function of the sun's height and azimuth. During simulation, the hourly shading factor can be calculated very fast by interpolation. An informative tool is the iso-shading curves 46 , which are superimposed on a sun's paths height/azimuth diagram, allowing to estimate at a glance the shading effects according to the season and time-of-day. The shading factor is applied to the beam component. The program has also to calculates the shading factor for the diffuse component 45 (as well as for the albedo), which is independent of the sun position and therefore constant over the year. Simulation results include shading loss calculations for Beam, Diffuse and Global irradiation components. It is to be noted that the real effect of partial shading on the electrical production of the PV field is non-linear, and depends on the interconnections between the modules. The program gives the possibility of partitioning the field into rectangles, each of which supposed to represent a string of modules in series, and calculates another shading factor according to module cabling 46 . Altough not perfect, this second approach should give an upper limit for the real shading loss evaluation.

Horizon - Far shadings

We distinguish between two fundamentally different types of shadings: The horizon is describing far shadings effects. This is the simplest way for defining shadings, but it's use should be limited to obstacles distant of, say, twenty times the PV-array size. Horizon acts on the PV field in a global way: at a given instant, the sun is or is not visible on the field. Near shadings 37 are partial shadings which affect only a part of the field. The shaded part changes during the day and over the seasons. Near shadings management require a 3-D construction of the field and it's environment, and is much more complex to deal with.

Defining horizon profile is a very simple operation with the PVSYST graphical tool. The horizon is a broken line superimposed on the sun path diagram, which can hold any number of points. - To modify it, simply drag the red dots with the mouse (or define the desired value in the corresponding edit box). - To add a point, click anywhere with the right button. - To delete a point, click on this point with the right button. On-site horizon measurements (height and azimuth of some significant points) can be obtained with a theodolite, a detailed map, a fish-eye photography, etc. Importing a Horizon profile: You have the opportunity of importing horizon profile
39

files from some other tools or Software.

Saving Horizon profile: A horizon profile can be saved for reusing it in another project or meteo calculation. It is stored in the \Shadings\ subdirectory with an extension .HOR. NB: A file with PVsyst format is not an ASCII file and cannot be exported to other software. Treatment during the simulation process: The effect on the beam component is of the "ON/OFF" kind: at a given instant, the sun is or is not visible on the field. As meteo is computed in hourly time steps, the program determines the exact time when the sun crosses the horizon line and weights the beam hourly value before performing the transposition. The effect on the diffuse component is not so clear. Up to version 3.03, the diffuse fraction, as well as the Chapter 4 Project design
38

Project design

albedo, was not affected by horizon in PVSYST. From version 3.1, one admits that radiation from the back side of the obstacles is null, and therefore the diffuse attenuation is calculated as an integral of an isotropic radiation over the portion of sphere seen by the plane, above the horizon line. This is independent of the sun position, and therefore constant over the year. Albedo contribution is more difficult to estimate. For far horizons, some radiation may be reflected by the ground ahead of the collector plane. We consider the albedo to be linearly decreasing according to the horizon height (up to zero for horizon > 20). On the other hand, if the "horizon" obstacle is rather near, albedo should be considered as null. Therefore the user has the opportunity of determining which fraction of calculated albedo he wants to take into account, according to the distance of horizon obstacle. The reality is certainly very complex, and requires more experimental investigations to assess these hypotheses on diffuse and albedo contributions. Nevertheless, please note that these contributions (and their errors) are rather low for low plane tilts, since the horizon irradiation has a low cosine factor. They become more significant for very tilted or vertical planes. Horizon in Meteo data In the PVSYST sites database (Meteonorm source), basic irradiation values are usually defined for a free horizon. Nevertheless in the DRY swiss data of very mountainous sites - as probably in other DRY or user's own meteo measurements - the horizon effect of the measuring site is usually already accounted for in the data since it is the result of a measurement. Therefore it is not always clear if a horizon correction is to be applied or not, depending on the field horizon by respect to the measuring station horizon. Importing Horizon profiles
The horizon profile may be defined manually by a set of (Azimuth/Height) points in degrees. These may be from on-site measurements (using land-surveyors instruments like compass and clisimeter). They can also be imported from several sources:

Solmetric SunEye instrument. This is a computerized instrument using a fisheye-type camera for the
recording of the environmental masks. (www.solmetric.com , www.soleg.de). It provides an horizon height for each degree of Azimuth (i.e. 360 points). - The software produces several files, one of them being a specific file for PVsyst, named "Sky0x_PVsyst.hor". Don't use this file, it was designed for the old PVsyst version 4.xx, and is restricted to -120 to +120. - You should choose the file "ObstructionElevation.csv" file instead. The new Version 5.0 has been updated for importing this file in a clean way, by taking the highest height among 3 azimuth points, therefore reducing the data to 120 points from -180 to 180. This resolution is far sufficient for the hourly computations of PVsyst.

Carnaval is a free open source software, which may be downloaded from incub.energie.free.fr.
It is based on a geo-referred grid for calculating the horizon line at any place between longitudes 6W to 10E, and latitudes 41 to 52 N, therefore largely covering France, East of Spain, etc. It uses satellite data from the spatial US programme SRTM (Shuttle Radar Topography Mission), giving a grid of altitudes with a 3'' resolution (about 92 x 65 m). When using Carnaval (V 0.7), you should: - choose the Lat-Long WGS84 coordinate system (GPS compatible); you can identify the exact coordinates of your system by using GoogleEarth. - don't use the "near environment definition". The near obstacles - which produce partial shades on the PV field should be defined in the "Near shadings" 3D tool in PVsyst. - compute the horizon line, and then define your "simplified" horizon by following the line with the mouse. The exported values will only be the green points which you have defined here. - press the button "export the simplified line". The choice of the Azimuth scale doesn't matter as PVsyst will recognize any chosen option. - export/save as a text file. Carnaval will add ".masque.txt" to any text you type here. Therefore the final name will hold a points before "masque", which is not accepted in PVsyst. You should rename the file in order to eliminate the forbidden characters, i.e. the point before use in PVsyst. - Now your file is ready for direct import in PVsyst.

Meteonorm produces Horizon profiles, which you can also import in PVsyst. The filename of these profiles
39

Chapter 4 Project design

Project design
holds the exact coordinates of the site evaluated.

Horiz'ON software: this is to be used in conjunction with a specialized support for your Photo Camera, which
allows to take several photographies with an horizontal reference (every 20 in azimuth), and gather them as a single panorama on which you can draw the horizon line by mouse (cf www.energieburo.ch/web/produkte/ horizon ). This produces a file with extension *.HOR, which is directly readable in PVsyst. NB: when established using a georeferred grid (like Carnaval or Meteonorm), the exact location should be carefully defined. This may be determined using GoogleEarth or a GPS. Please remind that a degree in latitude is 111 km, a minute is 1850 m and a second is 31m.

Your own profile as a file: to be recognised as a valid horizon profile, you can also provide any text file with
the following characteristics: - ASCII file with comma, semicolon, TAB or space separator. - First line may be a comment. - One line per defined point. Each point defined as an Azimuth and a Height value, expressed in degrees. - Should hold the extension .HOR and be placed in the directory \Data\Shadings\ of the PVsyst data structure. For example, a horizon file edited in EXCEL with first column as Azimuth, and second column as Height, and saved as "CSV" file, will be valid after renaming it as "*.HOR"

Near Shadings: tutorial

The near shadings are one of the most difficult parts of PVsyst. Therefore we present here a full exercise for explaining the main steps, and tips/advices for an easier use of this tool. First, please remind that there is no possibilities of importing 3D shading scenes from other software like Autocad or GoogleSketch. The basic data structure in PVsyst is completely different, and it would be quasiimpossible to reconstruct it automatically from the data of these software. We will create a Farm, as constructed in the Geneve Demo project. This is built from the following "architect" plan:
35 m

10m

35m N 20

8m 10 m 10m Silo D=6m H=12m

Elvation : Above all walls : Eaves of 0.5 m

Roof tilt 25

H = 5m

Defining the 3D scene: - Open the button "Near Shadings" and then "Construction/Perspective". You obtain the main 3D window where you will construct your "scene". Constructing the building. The building 49 will be an assemblage of elementary ob jects 49 , gathered afterwards as one only object in the main 3D scene. - In the menu, choose "Ob ject" / "New..." / "Building/Composed ob ject"

Chapter 4 Project design

40

Project design

This will open a secondary 3D window, which is the referential of the building object. - In the menu, choose "Elementary ob ject" / "New ob ject". Here choose "Parallelepiped" and define the sizes (Width = 10m, Length = 35m, Height = 5m). - Click "OK", this will place the parallelepiped in the building object's referential. - In the menu, choose again "Elementary ob ject" / "New ob ject". Here choose "Parallelepiped" and define the sizes of the second wing of the farm (Width = 10m, Length = 25m, Height = 5m). - Click "OK", this will place the parallelepiped in the building object's referential, positioned at the origin. Positioning in the 3D scene You have now to position 52 this second wing in the scene. Please observe that for selecting an object, you have to click on its borders (remember: the objects don't "know" their interior ! ). The selected object becomes carmin. - Click the "Top View" button (the five buttons top left are for the positioning of the observer). - Perhaps you want to diminish the scale ("Zoom b ackward" button). - Perhaps you want to recenter your scene: click on the scene - but not on an object - and drag the scene's plane. - Be sure that the positioning tool is activated (button with 4 crosses on the left). This opens the "Object positioning" dialog. - Now, you can click and drag the red point for displacing the selected object with the mouse, and the violet point for orientating it. Position and orientate the object roughly at its place as second wing, perpendicular to the first parallelepiped. - The mouse doesn't allow to get accurate values. But after rough positioning, the dialog will give you the order of magnitude, so you can finely adjust the exact values according to the architect's plans. In this case you will put X = 10.00m, Y = 10.00m, and don't forget Azimuth = 90.0 NB: Please avoid the interpenetration of objects. This often makes problems for a correct calculation of the shades. Now you have to add the roof. - Main menu "Elementary Ob ject" / "New ob ject" and choose "Two-sided roof + Gab les". - Define the sizes: "Base width" = 11m, "Top length" = 30.5 m (for eaves), "Roof tilt" = 25, and "Gable 1 angle" = 45. - Click "OK". This will put the roof in the building's scene. You can position it with mouse and values as before (X = 5, Y = 5, and Z = 5, building heigth). - For the second wing you could do the same. You can also reuse this existing roof: "Edit" / "Copy", and "Edit" / "Paste". You will obtain a second instance of the selected object. - Position this object by mouse and values (be careful with the new azimuth, exactly 90). Now the 45 cut gable is not correct. For modifying the selected object, you can: - either choose "Elementary Ob ject" / "Modify", - or, more easy, doub le-click the object on its border. - Change -45 to +45 and click OK. Now your building is finished, you can include it in the main 3D scene by the menu "File" / "Close and Integrate". Adding the PV plane PV planes cannot b e integrated in building objects, as the PV planes elements (sensitive areas) are treated differently in the program. They should be positioned on the building only within the 3D scene. - In the main 3D scene, choose: "Ob ject" / "New..." / "Rectangular PV plane". - You have to define the sizes: "Nb of rectangles" = 1 (you could define several non-interpenetrating rectangles in the same plane), "Tilt" = 25, "Width" = 5.5 m, "Length" = 25 m. NB: There is no relation at this stage with the real size of the PV modules in your system definition. The program will just check at the end of the 3D definitions that the "plane" sensitive area is greater that the area of the PV modules defined in the system, without shape considerations. - Click "OK". The plane comes by respect to the origin in the 3D scene.

41

Chapter 4 Project design

Project design

- For positioning it, again click "Top view", position it globally with the mouse. Now you don't have rigorous references and you don't need to adjust the values, but be careful not to interpenetrate the other roof! And check the azimuth value (should be exactly 90). - Vertical positioning: now your field is on the ground. Click the oberver's "Front View" button, and position your plane on the roof by dragging the red dot with the mouse. Please always let some spacing between any active area and other objects (minimum, say 2 cm). If you put the plane below the roof, it will obviously be shaded permanently ! Completing with the silo and a tree - In the main scene: "Ob ject" / "New..." / "Elementary shading ob ject" / "Portion of cylinder". According to the plan: define Radius = 3m, Aperture angle = 360, Nb of segments = 16, Height = 12m. Click "OK". - In the main scene, be sure that the "Positioning" tool is activated, click "Top view" and position the silo with the mouse (if you don't know the order of magnitude or signs), and then with values (X = 18m, Y = 45 m). - You can also put a tree in the courtyard. "Ob ject" / "New..." / "Elementary shading ob ject" / "Tree". For determining the sizes and shape of the tree, you can use "Front view", and here play with the red points for adjusting the shape of your tree. And then you position it as you like in the courtyard (please remember: a tree doesn't have definitive sizes!). Positioning by respect to the cardinal points For convenience you construct the scene in the referential of the architect. After that the button "Rotate whole scene" will perform the final orientation of the global scene. - Select the reference object for the orientation (normally the PV plane). - In the "Rotate Whole Scene" dialog, define the new azimuth (here +20, west). This will orientate your whole scene. But each time you will have to re-position an object in the scene, it will be easier to come back in the original system's referential, i.e. to a plane orientation of 0 or 90 ! Shading test and animation Now your 3D scene holds obstacles and sensitive area, it is ready for a shading analysis. - Push "Shadow animation over one day" button. And in this tool "Play/Record animation". The shadows will be shown for the whole selected day. After execution you have a scrollbar for reviewing one or the other situation. For each time step, the date/hour, sun position and shading factor are mentioned on the bottom of the 3D window. You can try this for different dates in the year. If there is some shade that you don't understand well, you can click the button "View from the sun direction" on the top group. This way you will have a direct view of the shades and their cause. Colours You can now personalize the view of your scene by defining colours. - Click the 9th button from left "Realistic view". - The colour of each element may be defined in its definition dialog. - For example for the building: Double-click the building, this will open the building construction. - Double-click the roof, this will open the definition dialog. - Here you can define the colour of the roof, and the colour of the gables independently. - Please note: the colours are defined "at bright sun": choose them rather light. - If you define your own colours, store them as "personalized colour" in order that you can reuse them for other similar objects. Recording the scene If you do some bad manipulation, you have an Undo opportunity (2nd button left). As the 3D tool is not fully safe in PVsyst (sorry... this is not easy to program and there still remain some bugs), and you can also do bad manipulations, you are advised to periodically save your shading scene using "File" / "Save scene" as a *.shd file. Please note: your definitive scene (used in the simulation) will be stored along with your "MyProject.VCi" file. It doesn't necessitate a *.SHD file. Display in report This scene will appear on the final report. If you want to have a specific view of the scene in the report, you can request it by "File" / "Save scene view" / "Keep this view for the report".

Chapter 4 Project design

42

Project design
Use in the simulation
Your shading scene seems now to be ready for the simulation.

Choose "File" / "Close". You return in the near shadings dialog. Choose "Linear shadings" in the box "Use in simulation". Here the program checks the compatibility of your 3D scene with the other definitions of your system. - The plane orientation should match the one defined in the "Orientation" part. If not, you have a button for eventually correct the "Orientation" parameters according to the 3D construction. - The sensitive area should be sufficient for positioning the PV modules defined in your system definitions. This is a rough test, which checks only the area, not according to the real sizes and geometrical positioning of your modules. The maximum area value for warnings is much higher, to account for installations with spaced modules. - When everything is correct, the program asks for computing the Table 45 of the shading factors. The table is a calculation of the shading factor (shaded fraction of the sensitive area, 1 = no shadings, 0 = full shaded), for all positions on the vault of heaven "seen" by your PV plane. It allows the calculation of the shading factor for the diffuse and albedo 45 (which are integrals of this shading factor over the concerned spheric portion) . At each hour, the simulation process will interpolate in this table - according to the sun position - for evaluating the present shading factor on beam component. This also lets the construction of the iso-shadings graph 46 , which gives a synthetic view of the time in days and seasons where the shadings are problematic. The line 1% for example, gives all the sun's positions (or time in the year) for which the shading loss is 1%, i.e. the limit of shadings. Now clicking "OK" will integrate this shading effect in the next simulation. In the final loss diagram on the report, you will have a specific loss for the "Near shadings".

Electrical effect: partition in module strings

Now when a cell is shaded, the current in the whole string is affected (in principle the current of the string is the current in the weakest cell). There is no possible accurate calculation for this complex phenomenon in PVsyst. We will just assume that when a string is hit by a shade, the whole string is considered "inactive" concerning the beam component. This is an upper limit on the shading effect: the truth should lie between the low limit - which we call the "Linear shading" - representing the irradiance deficit, and this upper limit (see partition in module strings 46 ), representing the electrical effect. For this second simulation "According to module strings": - Go back to the Near Shadings definition, button "Construction/Perspective" - Click the button "Partition in module chains" on the left. - Here you can split the field into several equivalent rectangles, each representing the area of a complete string (not a module !). If several subfields, you should do this for every subfield rectangle. This is a rough estimation, for a rough computation. Perhaps you will not be able to represent the real arrangement of your modules. But you can try different configurations, perform the simulation and then decide which configuration is the best suited for your particular system. - When performing the shading animation, the partially shaded rectangles will now appear in yellow. The new shading factor is the sum of the grey+yellow areas, by respect to the field area. Use in the simulation - In the same way as before, in the "Near shadings", please choose "According to module strings" in the options "Use in simulation". This will ask for computing the tables, and then you can open the iso-shading graph for comparing the effect to the "Linear" one. - "Fraction for Electrical effect": this is the way how the yellow parts will be treated in the simulation. A 100 % value will withdraw the total electrical production of these areas in the simulation. This is the upper limit of the shading effect. Perform a simulation with this value. - For the simulation presented to you end-customer, you can fix a different value for better approaching the reality. But sorry in the present time we don't have means for a good estimation of this factor (perhaps around 60-80%, accounting for the by-pass diodes partial recovery ???). Horizon (far) and near shading cumulation

43

Chapter 4 Project design

Project design

During one step of the simulation, the program will first evaluate the beam component according to the horizon line (ON/OFF, full or zero), and then apply the near shadings factor on the beam component. Therefore when the sun is below the horizon line, there will be no near shading loss as the beam is null. In other words, potential near shadings for sun positions already concerned by horizon will not produce any additional losses.

Near Shadings: main dialog

See also: Near shadings, general organisation
37 .

This dialog gives the general commands when using near shadings. You have first to build the global scene
48

by clicking "Construction/Perspective" button.

You may also load an existing shading model from the library if available. The shading scene you build here is part of the present variant attached to the current project. Nevertheless you can save it in the library for other uses in another variant of this project or another project. The "description" comment is especially intended to specify your shading constructions in your model library. Once this scene has been well established (i.e. the scene is complete, PV-field area and orientation are compatible with the rest of the system), the "Tab le" button allows you to calculate the shading factor table This process also calculates the diffuse attenuation factor 45 , constant over the year. After that you will be able to see the Iso-shading diagram
46 45

and run the simulation.

If you have defined a partition in module strings 46 during your 3D construction, two tables will be calculated simultaneously: one with the conventional "linear" shading factor, and one with the shading factor according to the module strings. Please note that although these factors for beam component are different, the diffuse attenuation factor is unique, related to the "linear" factor, because diffuse irradiation is supposed to be rather uniformly distributed on the field area. Finally, the "Use in simulation" group allows you to run the same variant under different shading conditions (without shadings, linear or according to modules) without modifying your shading construction.

Near Shadings treatment In order to apply shading effects, we have to treat each of the three components (beam, sky diffuse and albedo) in the appropriate manner: The effect on the beam component is of "ON/OFF" kind: for a point on the field, the sun is or is not visible. For far shadings (defined by the horizon), the whole field is subject to this binary function at a given time. In the case of near shading, a shading factor representing the fraction of the field illuminated has to be established for each sun's position (height/azimuth table 45 ). The sky diffuse component is also affected by the near shading obstacles. For simplification, we assume that the diffuse sky irradiation is isotropic. At a given time, the shading effect on the diffuse irradiation can be thought as the integral of the shading factor over the visible part of the vault of heaven, that is the spherical dihedron between the collector plane and the horizontal plane. This is independent of the sun's position, and therefore constant over the year. The albedo is only visible from the collectors if no close obstacle is present till the level of the ground. This is why we have to integrate the shading factor at zero height, on the portion of the sphere "under" the horizon, included between the horizontal plane and the plane of the collectors. It is however to be remembered that for non-vertical planes, the energetic contribution of the albedo is weak in the global incident energy, and that errors in its estimation will therefore only have secondary repercussions. This is also independent of the sun's position, and constant over the year.

Diffuse Shadings factor according to module strings Spatial distribution of shadings on diffuse irradiation are smooth enough, so that we can suppose that they don't affect the electrical behaviour of the field. Therefore we will use the same diffuse and albedo shading factors (computed from the "linear" shading factor table) for evaluating the shadings according to modules 46 . Chapter 4 Project design
44

Project design

Reflexions : The reflections on near obstacles, usually specular, cannot be calculated by PVSYST. However, though they are spectacular, these effects have negligible energetic consequences: they are in general involved only for very special hourly periods, and in the presence of the beam component; moreover, their effect on the real output of a sizeable PV field remains negligible, when considering that to benefit from it, a complete chain of cells in series should be uniformly illuminated by this supplementary supply (the production of a chain is indeed limited by the production of the weakest cell). In the same way, the reflection on the back of a shed, on the one hand only intercepts a small efficient part of the beam component, and on the other hand only illuminates a non-homogenous band in the lower part of the collector. Note: Notice that the usual acception of the diffuse component on a tilted plane includes the sky diffuse and the albedo. Table of shading factor
The shading factor is a complex calculation which spends too much time to be performed during simulation at each time step. Therefore the program builds a table in sun's height (10 steps) and azimuth (20 steps). The simulation can then interpolate in this table. Shading factor computation details The geometric configuration of the shadow falling on the field, and the determination of the shading factor, are carried out in a purely geometric and analytical manner. For a given solar position, the programme first carries out a transformation of the co-ordinates of the whole system, so as to point the OZ' axis in the direction of the sun. Next, for each sensitive element of the PV field (sheds, rectangles, polygons), it projects each elementary surface of the system on the plane of the field being considered. The intersection of the field element with the positive projections (i.e. in front of the plane) of each element is then calculated. The reunion of these elementary shadows forms a polygon representing the global shading on the field element under consideration. The shading loss factor is the ratio of the area of the shadow polygon, to that of the sensitive element. This process is repeated for each sensitive field element (for example each shed). The greatest difficulty with this procedure resides in the calculation of reunions and intersections of polygons in the plane, in the general case. This operation has proved to be extremely complex to programme using polygons defined by their summits. The difficulties mainly appear when the summits or the segments are overlapping or very close, as it is the case in most of the object constructions, when each summit is a part of several elementary surfaces in the 3D space. Topological decisions depend on the proximity of points in space. It is therefore necessary to define distance criteria as functions of the resolution of the calculations of the machine, or topological criteria, etc. and the reliability of this procedure is not absolute. Thus, in some cases, the result may be erroneous. Often, the programme finds it out on its own, and again begins the calculation with a slightly different solar position (trials with modifications of 1 in height or azimuth). If it fails again , the shading factor is calculated in a completely different manner: the PV field is partitioned in about 2000 points, and the shading is calculated for each point. Although this method is an approximation, it always leads to a reliable result. Nevertheless some calculations may sometimes stay erroneous (the program doesn't detect the error by itself); but in practice, this error has usually very little influence on the global simulation over one year. During the elaboration of the shading table, the points (sun's positions) situated behind the plane of the PV field appear in blue.

Diffuse and albedo factors

The sky diffuse component is also affected by the near shading obstacles. For simplification, we suppose that the diffuse sky irradiance is isotropic. At a given time, the shading effect on the diffuse irradiance can be thought as the integral of the shading factor over the visible part of the vault of heaven, that is the spherical dihedron between the collector plane and the horizontal plane. This is independent of the sun's position, and therefore constant over the year. The albedo 45
199

is only visible from the collectors if no close obstacle is present till the level of the ground. This is

Chapter 4 Project design

Project design

the reason why we integrate the shading factor at zero height, over the portion of the sphere under the horizon, included between the horizontal plane and the plane of the collectors (below the ground). It is however to be remembered that for non-vertical planes, the energetic contribution of the albedo is weak in the global incident energy, and that errors in its estimation will therefore only have secondary repercussions. Diffuse and Albedo shading factors are computed from the "linear" shading factor table. Shadings according to module strings Spatial distribution of shadings on diffuse irradiation is smooth enough, so that we can suppose that they don't affect the electrical behaviour of the field. Therefore we will use the same diffuse and albedo shading factors for evaluating the shadings according to modules.

Iso-shading diagram
The iso-shading diagram is a graphical expression of the shading factor table. It shows lines of some given shading factors, superimposed on the sun paths. Blue lines also indicate the tangential limits of the plane (i.e. when the sun rays are parallel to the plane). This diagram gives a synthetic evaluation of the shading distribution according to the season and the time-of day during the year. The irregular look of the lines is due to the interpolations across discrete calculation points. Remember that this loss factor applies to the beam component reaching the PV plane. When the incident angle is high, even high loss factors will act on very low irradiance component, giving rise to reasonable effects on the overall efficiency.

Partition in module strings The real effect of partial shadings on the electrical production of the PV field is non-linear, and depends on the interconnections between the modules. In the PV array, the current of each cell string is limited by the current of the worst cell in the series. That is, when one only cell is shaded the entire string is strongly affected (which has also dramatic effects on the I/V characteristics of the whole array). This may be understood using the specific tool "Electrical behaviour of PV arrays 37 " in the "Tools " section). Even with by-pass protection diodes, this string does not participate more than marginally in the production of the PV array. This phenomenon is too complex to be treated in great detail - with real string distribution of modules in space - during the simulation process. Nevertheless, the program provides a simplified method, giving the possibility of partitioning the field into rectangles, each of which supposed to represent a string of modules in series. Then it calculates a "Shading factor according to strings", stating that as soon as a string is hit by a shadow, the entire string (rectangle) is considered as electrically unproductive. Although not perfect, this second approach should give an upper limit for the real shading loss evaluation. In practice, one often observe that (except for regular arrangements like sheds), this upper limit is not so far from the lower limit (that is, the linear loss). Fraction for electrical effect Since version 5.0, you have the opportunity of specifying the intensity of the real effect on the electrical production of the partially shaded strings. This should allow to present a final simulation with a better estimate of the shading loss; but the value of the electrical effect ratio is not easy to establish, even using the tool "Electrical behaviour of PV arrays 37 ". We cannot give a reliable value for this "Fraction for electrical effect" parameter in the present time. This is dependent on the shades distribution on the field and the electrical array configuration. For a shed arrangement (where the shades are very "regular"), it is probably near to 100%. When with more "distributed" shades like Chimneys, far buildings, it could probably be of the order of 60 to 80%, depending on the "regularity" of the shade (a diagonal-like shade has probably a lower impact as it concerns modules better distributed in the array). Chapter 4 Project design
46

Project design

NB: The realistic electrical effect calculation of shadings will imply the exact positioning of each module on the geometrical plane, using the "Module Layout 89 " tool, the identification of each electrical string in the array, coupled to the shading calculation: this is not yet available, and will be done in a next version. Please also see the effect of thin objects Procedure: In the global scene, click on the "Partition in module chains " speed-button. You have to define the rectangle number in length and width, or their sizes. Definitions are accepted even if the rectangle sizes don't match the field sizes. If the scene has several fields, you have the option of extending this rectangle size definition to all other fields. No checks are made by the program on the compatibility with the effective module chains defined in the PV array's system definition. Effect of thin objects
If the shading object is sufficiently thin, its shade will not cover a full cell. Even if it is rather far and produces a broad semi-shading (due to the sun's diameter), the irradiance loss should be considered as the integral of the shading figure, and will be the same as the effect of a well delimited thin shading of the same wire. This is the case of electrical wires above the array, handrails, etc. In these cases the current in the cell will be reduced by a factor of the order of the wire diameter by respect to the cell's size. This is the value which should be attributed to the "Thin object" electrical effect ratio parameter. Then during the simulation, the "Module" loss caused by this thin object will be reduced by the "Thin object ratio". The reduction is directly taken into account in the loss factor according to modules.
47

(like electrical wires, handrails, etc).

Near shadings Limitations

See also: Near shadings: general organisation
37

and PV Planes

50 .

There are several orientations situations in PVsyst for which the 3D shading calculations are either approximate, or impossible. As the shading calculation table cannot be "doubled" (in the present state of the program), the configurations with several orientations cannot be calculated accurately. This concerns: - The Seasonal Tilt Adjustment: the shading factor is established relatively to the plane orientation. The table cannot be valid for the summer and winter situations at the same time. Therefore near shadings are forbidden. - The Heterogeneous (double orientation): in this situation two independent incident irradiance calculations (for both plane orientations) are made during each step of the simulation. This would normally involve two different shading factors, one for each irradiance component, which is not supported yet. However the program allows calculation (approximated) if the orientation difference between both planes is not too different (the difference limit is fixed at 25, adjustable in the "Hidden Parameters" 167 ). In this case it applies the same shading factor calculated for plane #1, see below - to both planes. - The sheds with sloped bases (sheds perpendicular to a 2-sided roof) are a particular case of the doubleorientation situation. Here the 3D shading calculation is of particular importance due to the mutual shadowing of the sheds. But in this situation the global orientation differences will not be very marked, so that we fall in the approximated case mentioned above. NB: For 2 different plane orientations, the shading factor is "globally" correctly calculated when the sun is shining on both planes, but it is not quite correct to apply it identically on both irradiance components. Moreover in the present algorithm the shading factor calculation is of course not performed when the sun is behind the plane; and for this limitation the plane #1 is taken as reference... This will be improved for a next version. For tracking planes, the plane situation is well determined for any sun orientation, so that the beam Shading Factor may be computed without ambiguity at each time. But the diffuse shading factor - which results of a yearly average related to a given plane orientation - is not yet correct. From V4.1 PVsyst computes diffuse shading factor for 5 typical orientations (or 4x3 with 2-axis tracking) and uses the appropriate value at each time step.

47

Chapter 4 Project design

Project design
Near shadings and Sheds

In PVSYST detailed simulation, the mutual shading of sheds (or sun-shields) can be computed in two different ways: - By defining them in the "Orientation" parameters option. You have here to define general parameters of sheds (width, tilt, pitch, etc.) valid for the whole PV system, and the simplified computation is assumed to be "linear" (without electrical cells effects) and for unlimited length (that is, neglecting edge effects on both extremities of the sheds). - By explicitely defining a PV plane as sheds in the "Near shading" scene. In this case the computation accounts for shed edges, and a module partition can be defined. Please note that these two options should not be used at the same time, as the shadings will be accounted for two times !!! Definition by the "Orientation" parameter shed option This option is most suited when you have a field of numerous and little sheds (for example "one-module" wide sheds), sufficiently long as you can neglect the edge effects. Nevertheless, if you have to combine such an array with other surrounding shading obstacles in a near shading scene, you should define the basic array in the near shading scene as an horizontal plane, covering the sheds base extent (i.e. the whole area used for installing the sheds, on which the surrounding shadings will apply). This way the mutual shadings will be taken into account by the plane orientation shed algorithms, while the surrounding shadings will apply on the basis plane. This is of course an approximation, but the only way when the sheds are so numerous that the near shading complexity and calculation times become prohibitive. Shed definition in the Near Shading scene The "Near shading" shed construction should only be used when the number of sheds is less than, say, fifty sheds. The computing time and complexity of the shading factor calculation grows with the square of the number of elements. When you define sheds in the 3D scene, you have to choose a fixed tilted plane (not sheds) in the "Orientation" parameters, with the real values of tilt and azimuth of one shed. .

Near Shadings: 3D construction

Global scene building
See also: Near shadings, general organisation
37 .

The global scene gathers all the objects according to their relative position, in a referential 52 based on the cardinal points. In Construction mode, the shading obstacles are represented in black and the sensitive area of the PV fields in blue. First create your objects (Menu "Create") among the available parametrised object's library, i.e.: Five kinds of PV planes Elementary objects
49 50

: rectangular, polygonal, in "sheds", sunshields and tracking,

(a variety of 2D an 3D predefined shapes),

Building / Composed object 49 : an assembly of elementary shapes. Buildings can be saved as models for reuse in other shading scenes. They cant include PV fields, which should be added independently in the global scene.
53

The upper tool bar (blue icons) provides essentially means for defining the observer's point of view (perspective or orthogonal - top, front and side - views) and the Zoom. The left tool bar gives access to the following actions on the system and it's components: Displace the scene on the screen (you can do that at any time by clicking out of an object). Undo allows to retrieve up to ten last operations (Ctl-U)

Select an object: Each object can be selected by clicking on it. A selected object becomes purple. Copy creates a copy of the selected object (Ctl-C); this copy is kept "permanently" and may be passed to another scene.

Chapter 4 Project design

48

Project design
Delete suppresses the selected object (Ctl-V or Del),

Modify opens the edition dialog for modifying the selected object in its own referential. Double-clicking an object also opens its edition/modification dialog. Position allows to modify the position and orientation of the selected object. These can be edited as numbers. But it is much easier to set a plane, side or front view in order to drag the object in the scene with the mouse. Rotate the whole scene: very useful tool which allows to build the scene in its "natural" referential (parallel to the building according to architect's plans), and then rotate the whole scene according to the cardinal points. Partition in module chains 46 : creates a partition in module chains (i.e; rectangles) for non-linear shading calculations according to module strings distribution. Measurement tools scene.
54

allow to easily get real distances and angle measurements between points of the

Shadows drawing: with completed shading scene, allows to visualise the shadows on PV planes for any sun position or time-of-year. Includes quick dialog tools for incrementing the time-of-day (by +/-30 min) or the day-of-year (by +/- 10 days). You should click on the "Shadow drawing" button for refreshing shadows after each parameter modification. Shadows animation: sweeps the sun position every 15 minutes over a given day. Shows the shadings and draws the shading loss evolution, gives the overall loss on beam component over the whole day. After completion any individual position may be reviewed in detail. From sun view: In order to well understand which object is involved in the shading construction, you can choose to observe the scene from the sun position (button in the upper tool bar).

Realistic view: after construction, you can have a look on your scene in a more realistic way than the ironwire representation used during technical construction. You can define your own colors for each elementary object. This realistic representation helps understanding some complex constructions, and improves of course the presentation to a customer!

Most of these tools are also available either by Popup-menu (right button) or by keyboard accelerators.

Elementary objects
See also: Near shadings, general organisation
37 .

PVSYST proposes a library of elementary shapes, basic or usual in architecture: 2D shapes: Triangles (whatever, isosceles or rectangle), rectangles, trapezium, regular polygon, pseudocircle sectors. 3D shapes: Parallelepiped, square pyramid, triangular, hexagonal or octogonal prism, portion of cylinder. Building elements: House + 2-sided roof, Tree, Roof-like diedre, 2-sided roof+gables, 4-sided roof, prism chimney.

The "Elementary ob ject" dialog allows to build one elementary object at a time in it's own referential. The user chooses the shape and sizes either by using available parameters, or by dragging significant points in the orthogonal views (plane, side or front views). You can choose a custom colour for each elementary object. Remember that the chosen colour will be that at bright sun. In the global scene, surfaces not facing the sun will appear darker. More complex objects are obtained by assembling elementary objects. This can be achieved in the Building/ Composed objects 49 dialog. The "Elementary ob ject" is then positioned
52

in the general scene or in a "b uilding" element.

Defining a meaningful comment is not obligatory; it will help identifying the object afterwards in the global scene. Please note that internal representation of objects is built up with 3D-Points and Surfaces built with these summits, so that curves should be approximates by broken lines.

Building or composed objects

See also: Near shadings, general organisation
37 .

Although elementary objects can be readily integrated in a global scene, PVSYST allows to assembly several elementary objects to build a more complex one (for example a complete building), which will be manipulated as 49

Chapter 4 Project design

Project design
a whole in the global scene.

The construction takes place in a secondary perspective view, and is quite similar to the global scene construction 48 . Tools are the same (a subset) and the user can create and manipulate the elementary objects in the same way. As with other objects, the user is advised to give a meaningful description, and will have to position composed object in the global scene. However, "building" objects can be saved as models for use in other shading scenes. Please note that "Building" objects are intended to build shading objects; therefore they cannot hold PV sensitive planes. When necessary, these have to be created and positioned in the global scene, independently of the building object.
52

the

PV planes
See also: Near shadings, general organisation
37 . 52

PVSYST offers 5 kinds of PV-fields, which are created in their own referential, and then positioned general scene.

in the

The PV planes can be defined either by parameters, or by dragging the red points in the PV-plane view ("2D" button). Rectangular planes: you can define as many non overlapping rectangles as you want in the same plane, each with it's own height; length and position. Polygonal PV-plane: you can draw any polygonal shape in the plane. To add a point, click "Insert'" and then click on the desired segment. Planes in sheds: several identical planes. Each plane includes the sensitive area, as well as a frame for the shading on the next plane. You have to define the sensitive area sizes, the pitch and the tilt, and the inactive edge widths. You can also specify a lateral misalign between the sheds. Use orthogonal views to adjust tilt, pitch or misalign by dragging. Transversal slope: It is now possible to define a slope between sheds (for example disposition on a hill). Baseline slope: Special case of tilted baseline: if the sheds have to be placed perpendicularly on a tilted surface (for example a 2-sided inclined roof), the baseline of the sheds may be tilted. In this case you will define the PV-plane tilt by respect to the mechanical layout on this roof. But the real orientation of the shed planes will be different from that of your mechanical supports (this value appears in the definition dialog). With such an installation on a 2-sided roof, it is impossible to obtain a same orientation for the two sets of sheds, therefore you should use the Heterogeneous option in the Orientation parameters. NB: With heterogeneous fields, for a given simulation time-step the shadings factor is calculated only once and applied identically on both planes. Therefore when using near shadings the orientations should not be too different (the program set a limit). NB: When defining sheds in the Near shading scene, you should not use the Shed option in the "Orientation" parameters definition, otherwise the simulation will account twice for the mutual shadings! See also Near shadings and sheds 48 . Sun-shields: same definitions as for Sheds, except that the collectors are aligned on a wall. It has to be noted that the faade itself induces a significant irradiance loss, especially in summer when the sun has a course over the east and west azimuths. Faade slope: Although the faade is usually vertical you may define an inclined base (from one unit to the next one). Tracking planes: 51 the basic definitions of tracking planes are very similar to the sheds: several identical trackers, with a sensitive area and an inactive frame for mutual shadings. There are four kinds of tracking planes: two axis, tilted axis, horizontal (east-west), and sun-shields.
47

NB: Some special orientation configurations may induce limitations

on the 3D shading calculations.

The specific definition dialogs include 2D view in the PV plane (with graduated axis) , as well as usual 3D

Chapter 4 Project design

50

Project design
perspective and orthogonal views.

In its own referential, the PV-plane is defined with its tilt, but always facing the OY coordinate. Plane azimuth will be defined only when positioning the plane in the global scene. Please remember that the total sensitive area should roughly correspond to the total PV-modules area as defined in the "System" part. Defining a meaningful comment is not obligatory; it will help identifying the object in complex scenes. Very important: for proper operation, the sensitive areas should not be tangent or penetrate other shading objects. When positioning the PV fields in the global scene, please be careful for always leaving some centimeters between the planes and the other objects.

Tracking planes
The basic definitions of tracking planes are very similar to the sheds definitions with a sensitive area and an inactive frame for mutual shadings.
50

: several identical trackers,

There are four kinds of tracking planes in which the disposition may be different as the sheds. This is the reason why when creating a new tracking array, your first action should be to define the Tracking Parameters. This dialog is analogous to the one of the "Orientation" general parameters, it will define the kind of tracking, the mechanical stroke limits of your trackers, and eventual parameters for backtracking or concentration. The four tracking kinds are: - Two axis: by default, the alignment is defined along the east-west direction, with a given "pitch" between trackers. But you may also define a "misalign" value for a diagonal disposition, or even a set of trackers "one behind the other" by setting the pitch = 0. - Tilted axis: this corresponds to most of the one-axis tracking PV arrays, for which the trackers are usually aligned along the east-west direction. This option should also be used with horizontal "north-south" axis. This option offers the same disposition possibilities for pitch and misalign as above for two-axis planes. - Horizontal axis (east-west): this "old" option corresponds to a disposition analogous to parabolic linear concentrators in distributed thermal plants, with trackers "one behind the other" like in fixed sheds. It is not really suited for PV systems as the horizontal tracking takes advantage of the seasonal variations, but is not very efficient over the diurnal sun course. As above, the Pitch, Misalign and Transversal Slope parameters allow almost any configuration. - Sun-shields are taking a great importance with the architectural integration. But with fixed sun-shields it is very difficult to find a good compromise between an efficient sun protection and an acceptable PV yield. The shadings from one sun-shield to the lower one when the sun is high (i.e. during the best summer hours) is difficult to overcome, especially when the faade is not exactly south. This could be partially solved by using tracking sheds with a backtrackingNear_Shadings_BackTracking strategy. The construction dialog offers a tool for testing all possible rotations, in order to check the mechanical compatibility of your array layout (using the orthogonal plane or side view). The Shading animation over one day provides a powerful tool for optimising your tracking layout by trial-and-error. Heliostat array layout should be carefully optimised regarding the mutual shadings. Constraints are much more critical than for the sheds disposition, as significant yield may be waited even when the sun is very low on the horizon. The Backtracking 51 control strategy helps improving the electrical behaviour of tracking arrays.

Backtracking Strategy
See also Near shadings: Tracking planes
51 .

Heliostat array layout should be carefully optimised regarding the mutual shadings. Constraints are much more critical than for the sheds disposition, as significant yield may be waited even when the sun is very low on the horizon. The mutual shading problem is accentuated by the electrical behaviour of the strings under partial shadings. Identical shadings appear on each heliostat and may block many strings at a time. The backtracking strategy is now proposed by several manufacturers of tracking arrays: when the mutual 51

Chapter 4 Project design

Project design

shadings begin, the tracking angle does no more follow the sun, but it goes back in order that no shading occurs. In this situation, viewed from the sun, the trackers remain tangent to each other. Of course the incidence angle increases simultaneously. Therefore in the 3D shading scene, the shading factor remains 1 as far as the trackers do not reach their stroke limits. The position of the trackers evolves until coming parallel to the whole array when the incidence becomes near 90. The shading factor falls to 0 sharply, when the sun passes behind the plane of the array. It should be noted that the Backtracking doesn't increase the total irradiance received. It only improves the electrical loss effects of the shadings. The total irradiance reaching the modules is the same as if there were shadings: it corresponds to the total sun energy intercepted for this given sun direction, by the field area "seen" by the sun. Therefore a simulation with "Linear shadings" (not electrically realistic) and another one with backtracking should give the same results (if we neglect the IAM losses unavoidable with the Backtracking). The backtracking calculations are of course dependent on the mutual disposition of the trackers. For a given sun position, the tracker orientation should be determined using the pitch and the width of the frames/collectors. Therefore for defining backtracking, we need the same parameters as for the shed calculations, i.e: - Pitch between trackers, - Collector width (sensitive area), - Left and right inactive bands (of frames). These parameter are defined geometrically in the 3D scene, but should also be included in the "Orientation" parameters. This is the reason why we can define backtracking only in the framework of equidistant tracker systems. NB: As there are no shadings (except when the sun is behind the plane), the 3D construction is not really necessary. The "Orientation" dialog is sufficient for fully defining a Backtracking system for the simulation. The backtracking calculations are specific for each kind of tracking planes: - Two axis: the tracking mechanism is supposed to be a rotation around the vertical axis, and a variable tilt of the plane. In Backtracking conditions the tilt is facing the sun (90 - sun height) and the azimuth is adjusted for no shadings. Other geometries could be imagined, but would require additional parameters. - Tilted axis: the Phi angle around the axis is adjusted for no shadings conditions. This doesn't allow for misaligned arrays. - Sun-shields: When the sun is high in the heavens, the sun shields become highly tilted. It is also the case for east/west orientations when the sun is low on the horizon. The compatibility with comfort conditions has to be studied.

Global_Referential

Referential of the global scene

The referential of the global shading scene is based on the cardinal points: In the northern hemisphere: the X-direction corresponds to the WEST, Y to the SOUTH, Z to the ZENITH. As for the usual definition for the PV fields, the azimuths are given with respect to the South (OY), and positively towards the WEST, that is clockwise (antitrignometric direction). In the southern hemisphere, the X-direction points to the EAST, and Y to the NORTH. The azimuths are given with respect to the North (OY), and positively towards WEST, that is anticlockwise (trignometric direction). To facilitate the system construction, the user can build it's whole scene in a referential related to his system (for example parallel to his building), and then rotate the entire scene by the desired orientation angle.

Object positioning
Each object (PV plane, elementary object or building) is built in it's own referential, which has to be positioned in the global scene. Please choose the "Position" speed button (or use menu): this will show the positioning dialog. Positioning operations are greatly facilitated when using orthogonal views in order to drag the object in the scene 52

Chapter 4 Project design

Project design

(red dots). Plane view also allows to rotate the object (violet dot). Final fine tuning may be achieved by parameter edition. From a mathematical point of view, parameters are defined such that the object is first displaced in the main referential, then it is rotated by the given azimuth angle, and finally tilted around the new OX' rotated axis. This process is the same for positioning an elementary object in a "b uilding" object. Important notice: When positioning PV planes among other objects, please always leave a little space between the plane and the support surface. Indeed, shading calculations involve complex calculations of intersections and reunions between 2D projections of these objects. Confused points (and also points confused with a surface) often cause problems to these routines, and may sometimes lead to topological errors. On the other hand, module shading calculations consider a rectangle as shaded as soon as one point is shaded. When confused with its support surface, the baseline of the PV plane is calculated as shaded, and invalidates the lower rectangles.

Shadows drawing
When the global scene is completed, you can have a look on the produced shadows for any given sun position, or any time in the year, by clicking on the speed-button "Shadow drawing". Solar angles or time conditions may be easily modified to see the evolution during the year. After each parameter change, please click on the "Shadows drawing" button again for updating the shadow computation. NB: This "Linear" shadow computation is a very complex process involving polygon intersection and union calculations, which sometimes may fail, giving erroneous results (i.e; part of the field is over- or undershaded). If the program can detect the failure through it's internal checks, it uses an alternative computing method by distributing a grid of points on the PV area, and evaluating the shading state of each point. This second method is less precise, but leads always to reliable results. To minimize such problems, please always position PV planes with a little gap between them and their support surface. If you have defined a partition of your PV-field according to strings of modules, the partially shadowed string rectangles will also appear in yellow. The two shading factors will be displayed, indicating the lower and upper limits for the shading's real effect. You also have an animation of the shadows drawing over a whole day. This simultaneously draws the curve of the loss factor on the beam component, and calculates the overall daily loss on beam component for a clear day. Tip: Try the "View from sun direction" speed-button to deeply understand how the shadows are formed !

Observer

Observer's point of view

The upper tool bar (blue icons), as well as the "Ob server" menu, provide means for defining the observer's point of view: Perspective: by default, gives a standard view (height=20 and azimuth=30, with the observer at 200 meters). The observer's view angles may be changed in several ways: scrollbars on the image top, arrows keys on the keyboard, or by the menu. The observer distance can only be changed through the menu "Ob server"/"Detailed position". This could be necessary if you have a very big scene. Axonometry: this representation is equivalent to a perspective from infinite distance. Axonometry preserves parallel lines. Plane, front and side views are defined according to the referential axis. These are the only views which allow displacements or other operations in dragging mode (red or violet points on the objects). Forward and Backward Zooms are available by buttons and menu, but back to normal size is only in the menu. The Hand button allows to displace the scene on the screen. During the shadow drawing, you have the possibility of viewing the scene from the sun direction.

53

Chapter 4 Project design

Project design
Measurments

In the shading scene, there are four measuring tools, for real distance or angles determination. You define segments by clicking on two points of the scene. When clicking on an existing summit of the scene, the measuring point will stay attached to it. After defining them, you may always drag the existing points to other positions.

Distance measurement:
Click on two points to define the distance arrow. In perspective view, the distance is only defined between existing summits of the scene. Angle measurement: Only available in top, side or front views. Click to define a first segment, and then a second one (that is, you should define four points). The angle is given even if the two segments don't have any common summit. Nevertheless if so, the complete angle will be drawn. Azimuth measurement: This gives the usual azimuth direction by respect to south (or north in southern hemisphere). Define origin first, and then the extremity of the arrow. In perspective view, the vertical tilt angle (named "height") is also given. This allows for doing equivalences with the "horizon" representation on the "Height-Azimuth" sun path diagram. Tilt angle measurement: Only available in front and side views. Click on the origin first, and then the arrow extremity. NB: When attached to summits, these tools can help positioning the objects intuitively and with precision.

User's needs ("load")

The programme offers great flexibility to define the needs of the user (or the "load" of the system). You can choose among the following options: Unlimited load, which only concerns installations connected to a sufficiently "strong" grid to absorb all the power produced. This is the default option with grid systems. But this option obviously makes no sense with a stand-alone installation, where production and especially battery behaviour are closely related to electricity use. Fixed constant load, offers the simplest way to define user's needs. You just have to define it as constant power or yearly energy. Monthly values gives the possibility of defining monthly averages, which will be used by the simulation as constant over each month (no daily modulation). Values are defined using a special graphic tool 57 . Daily profiles in hourly values can be user-defined using a graphic tool 57 , as: Constant over the year: the same profile all along the year, Seasonal modulation: four specific daily profiles, for each season (defined as June-August, September-November, December-February, March-May). Monthly normalisations: one only daily profile, the amplitude of which being modulated according to given monthly sum values. Weekly modulation: a specific daily profile for "working days", and another one for "week-ends". Each of these staying constant over the year. The number of "Working days" in a week is user defined. Probabitily profiles are designed for grid applications, when the grid load is not illimited. It is the probability distribution of the power which can be absorbed by the load at a given instant. This is especially suited to DC-grid for public transports 65 . Possible power load is divided into 12 classes of given values; the user specifies the probability distribution values for each hour. This profile 56 can be defined: Constant over the year: the same profile all along the year, Seasonal modulation: one specific daily profile for each season (as above). Weekly modulation: a specific daily profile for "working days", and another one for "week-ends".

Chapter 4 Project design

54

Project design
-

Daily Household consumers is the default user's needs definition for stand-alone systems. It proposes a list of most usual domestic appliances, with unit power and daily use duration. Values 55 can be defined as: Constant over the year: the same daily load value all along the year, Seasonal modulation: one specific daily load for each season. Monthly definitions: a specific appliance distribution for each month. NB: With this tool a Week ly modulation option is independently available for each season or month, but only for putting some "week-end" days to zero or stadby. This appliances definition is mainly thought as a pre-sizing tool for stand-alone systems, in order to evaluate the load from a user's point of view. In presizing, the effective use of the defined values, during the simulation process, is equivalent either to a constant load, or to seasonally or monthly constant values. In detailed simulation, a superimposed daily profile may be defined. Load values read on ASCII file is the most flexible load definition: you can edit your own desired load profile (either in hourly or in daily values) in a spreadsheet editor or any ASCII file, and easily import it 55 in PVSYST.

NB: You can save your Load profile as a model for reusing it in another project.

Load profile: ASCII file definition

See also 'User's needs'
54

The most flexible way for defining a custom load profile is to input it as an ASCII file. The ASCII input source file can be edited in a spreadsheet program (like MS-EXCEL) or a text editor. It should hold one load value per ASCII line. Each record (line) may contain several fields, separated by semi-colon, comma, TAB or blank characters, and terminates by CR, LF or both. With MS-EXCEL, such a file is easily obtained by saving a usual sheet in 'CSV' format (Comma Separated Values). You can define either: - Hourly file , which has to contain 8760 hourly values ( [W] or [kW] ) from January 1st, 0h to December 31th, 23h (or eventually 8784 values if the meteo file used in simulation is a leap year). - Daily file , with 365 (or 366) daily values ( [Wh/day] or [kWh/day] ), from January 1st to December 31th. The file can have any number of head lines for titles. To import the file in PVSYST: choose the ASCII source file: its contents will appear for interpretation, define the separator, define the number of head lines (will be coloured in yellow), define the order of the "Load" field (column), define the time step and units.

Then, clicking "Read file" will import the values in the PVSYST "user's needs" (part of the simulation variant parameters), which can be visualised or saved as a model file.

Domestic User's needs

See also 'User's needs'
54

This is a simple tool aiming to facilitate the domestic load estimation from a practical ("end-user") point of view. This tool also clearly identifies the consumption of each appliances, and perhaps will help optimising the customer needs options (for example, bring attention on stand-by or washing machines real cost in PV systems). The user has to define the number of appliances, their unit power and daily use. The table shows the resulting daily consumption. The "Week -end use" option may be applied to each season or month independently, and acts by putting the load to zero during some days of the week.

55

Chapter 4 Project design

Project design

The last line defines the Standby consumption, which applies over 24hours by definition. When using "Week -end" option, you may choose leaving these appliances connected or not during the non-utilization days. When defining seasonally or monthly values, you have a facility for copying an already defined profile to one or several other seasons/months. For this please click on "Copy", choose the source profile and select the target seasons. Daily distribution When used in pre-sizing process, the daily load profile is assumed to stay constant over the day. That is, from the computation point of view, the Domestic User Load is equivalent to a yearly, seasonal or monthly constant profile (except if week-end use is defined). When used in the project design part (detailed simulation), it is possible to customise an hourly profile 56 , valid for the whole year (for example: enhancing the evening use of lighting, TV, etc). This can affect the battery use results and wear (owing to charge/discharge distribution). Details of the defined appliance profiles can be saved on a file for reusing in other projects; or printed, either in the load dialog or along with the simulation results. Consumption of some appliances Power / Consumption Fluorescent lamp Television (acc. to screen size) Video recorder Personal computer Fridge / deep-freeze : consumption Micro-wave oven Washing-machine (dishes, clothes) Consumption: DHW solar system pump + regulation Stand-by Consumption: 7 18 W 40120 W 2025 W 120 W 0.41.2 kWh/day 1 kW 2.2 kW 1.21.8 kWh/washing 30 W 5 W / app. 120 Wh/day

When using "standard" appliances with an AC distribution, add about 10% to account for the inverter efficiency. Pay great attention to standby consumption of a lot of modern appliances: 5W ribbon loss will consume the production of a 50 Wp PV module! Domestic use hourly distribution Defines an hourly profile over the day, in order to better match the battery behaviour (and therefore better calculate its wear and tear). Checking the "Auto evening use" box will automatically restrict the lighting and TV uses to the evening hours (i.e. from 19h to 23h). NB: In the framework of the "Domestic use", it is only possible to define one hourly profile, valid over the whole year. If you need more refined profile according to seasons, please use the "Daily Load" definition instead. NB: Hourly distribution cannot be specified in the Preliminary Design (pre-sizing) process.

User's needs: probability profile definition

See also 'User's needs'
54

When entering this probability profile definition tool, please first choose the "scale" or "bin step", i.e. the classes width matching your desired power range. Chapter 4 Project design
56

Project design

During utilisation in simulation, the real hourly load power will correspond to the midpoint of a given bin, randomly chosen according to the specified probabilities. At the first entry in this dialog, the probability values follow a default distribution. You can drag each probability on the graph, or define a given value by editor. Of course the sum should always be 100%, so that at each change all other values are modified in order to match this constraint. With a given range and a given probability distribution, the program calculates a "most probable" day sum or monthly sum. This is not an exact value, but only the statistical average when the number of days is increasing. If you are in "Seasonal Modulation" mode, you have to define a profile for each season. You have a facility to copy an already defined profile to one or several other seasons. For this purpose click on "Copy", choose the source profile and select the target seasons. The same holds for the "Weekly modulation" mode.

User's needs: daily profile definition

See also 'User's needs'
54

When entering this daily profile definition tool, please first choose the work ing units . At the first entry all values are null. You can use the operator facility to modify all hourly values at a time: First choose the operator, which can be: - Fix identical values for each month, - Add a given value to each month, - Multiply all values by a given value, - Renormalise all values to obtain a given yearly sum or average. Then define the Operator value, and click "Work out". You can also drag each hourly value on the graph, or define a given value by editor. If you are in "Seasonal Modulation" mode, you have to define a profile for each season. You have a facility to copy an already defined profile to one or several other seasons. For this purpose click on "Copy", choose the source profile and select the target seasons. The same holds for the "Weekly modulation" mode. In the "Monthly normalisations" mode, be aware that the daily sum has to match the pre-defined value for the considered month (as defined in the preceding dialog). So that when changing a single value, all other values are automatically modified accordingly. In this mode the best suited working unit may be % of daily load.

User's needs: monthly values definition

See also 'User's needs'
54

When entering this monthly definition tool, please first choose the work ing units. At the first entry all values are null. You can use the operator facility to modify all values at a time: First choose the operator, which can be: - Fix identical values for each month, - Add a given value to each month, - Multiply all values by a given value, - Renormalise all values to obtain a given yearly sum or average. Then define the Operator value, and click "Work out". You can also drag each month value on the graph, or define a given value by editor. NB : If you are in "daily profile" mode, the "Next" button will open the similar Daily values dialog
57 .

57

Chapter 4 Project design

Project design
Grid-connected system definition

The "system" is defined as the set of components constituting the PV-array, the inverter, up to the connexion to the grid. First rule: all the strings of modules connected to the input of an inverter (or a MPPT input), should be homogeneous: identical modules, same number of modules in series, same orientation. Exceptions may sometimes be acceptable - as far as only differences in the current of strings are concerned - for example strings of different orientations (cf Heterogeneous planes 31 ). PVsyst now allows the construction of heterogeneous systems with several different subfields
61

(up to 8).

For a given subfield: you have to define your requirements, and PVsyst will automatically propose a suited arrangement. The basic requirements for a sub-field (i.e. the parameters you should input) are: - The desired nominal power, or alternatively the available area for installing modules, - The inverter model, chosen in the database, - A PV module model, chosen in the database. Then the program will choose the required number of inverters, according to a pre-defined Pnom array/ inverter ratio of 1.25. It will then propose a number of modules in series, and a number of strings in order to approach the desired power or available area. The acceptable choices for the number of modules in series/parallel are mentioned on the dialog. They should meet the following requirements: - The minimum array voltage in worst temperature conditions (60C) should not be under the inverter's voltage range for MPPT, - The maximum array voltage in worst temperature conditions (20C) should not be above the inverter's voltage range for MPPT, - The maximum array voltage in open circuit (Voc at -10C in Europe) should not exceed the absolute maximum voltage at the input of the inverter, - The maximum array voltage in open circuit (Voc at -10C in Europe) should not exceed the allowed system voltage specified for the PV module. NB: The Voltage values calculated by PVsyst for Amorphous modules are the stabilized ones after degradation. The initial values may be up to 15% higher during the first months. This should be taken into account when sizing the system, especially concerning the absolute maximum voltages for the inverter input or the module insulation. The inverter power sizing is a delicate and debated problem. PVsyst proposes a methodology 60 based on the predicted overload losses. This usually leads to Pnom ratios far below those recommended by inverter's providers, but we think that they are closer to an economical optimum. All these conditions are explicitly displayed on a system sizing graph, (button "Show sizing"). You can now play with these parameters taking your own constraints into account. You can retrieve the automatic proposed values by clicking on the associated checkbox. Warning messages will be displayed if there is some incompatibilities between the chosen parameters. Red warnings are not acceptable (simulation cannot be performed) and orange warnings are indicative. These colours will be thrown back on the "System" LED's button. NB: All the sizing parameters mentioned above (array temperatures, Pnom array/inverter ratio, warning limits), may be modified in the "Hidden parameters 167 "). This is especially the case for sizing temperatures in different climates. If desired (in a second step of the development of your project), you can modify the PV-array specific loss
83

parameters (thermal, wiring resistance, module quality, soiling,

58

Chapter 4 Project design

Project design

mismatch, IAM) with the "Detailed Losses " button. - define a user's load 54 profile for determining the owner's own consumption and the injected (sold) energy to the grid (net-metering). You will usually define a hourly profile in order to take realistic daily production and consumption distributions into account.

Array voltage sizing according to inverter

The number of modules in series has to match the following conditions: The minimum array operating voltage (i.e. at max. module operating temperature, 60C by default) should be above the minimum inverter's operating voltage (Vmin of MPPT range). The maximum array operating voltage (i.e. at min. module operating temperature, 20C by default) has to stay below the maximum inverter's operating voltage (Vmax of MPPT range). The maximum array absolute voltage (i.e. Voc at min. temperature, -10C by default) has to stay below the absolute maximum inverter's input voltage. The maximum array absolute voltage (i.e. Voc at min. temperature, -10C by default) should not overcome the maximum system voltage specified for the PV module.

When the desired array configuration doesn't match these requirements, the system is usually not properly sized. Please see the diagram in the "System" definitions, button "Show sizing", which summarizes all these constraints. Design temperatures These conditions involve design temperatures, which are part of your project and may be changed according to your climate in the definition of the project, option "Site and Meteo" / "Next". The default values (for each new project) may be redefined in the Hidden Parameters, topic "System design parameters". These are: - Maximum cell temperature in operating conditions, default 60C, - Summer usual operating conditions, not used for sizing constraints, default 50C, - Winter minimum cell temperature in operating conditions, default 20C, - Absolute Cell lower temperature for determining the Maximum possible voltage of the array. The default is set to -10C for most European countries (best practice rule). For this limit, the cell temperature is considered as the ambient temperature (worst case when the sun suddenly appears on the field). NB: these parameters are used for design only. They are not involved in the simulation in any way. Amorphous modules For Amorphous or other thin film modules, the voltage values calculated by PVsyst are the stabilized ones after degradation. The initial values may be up to 10-15% higher during the first months. This should be taken into account when sizing the system, especially concerning the absolute maximum voltages for the inverter input or the module insulation.

Amorphous degradation and array voltage sizing

It is well-known that the amorphous modules suffer of a significant initial degradation during the two first operating months due to the Staebler-Wronski effect. This is sometimes specified in the data sheets, and the voltage can be about 10% over the stabilized value. As the electrical data used for the sizing of the PV array are the stabilized ones, dangerous voltages overcoming the absolute admissible voltages may appear just after the commissioning of the system. This may be taken into account in the Sizing dialog. Now when the commissioning takes place in Summer, the reference sizing temperature may be chosen much higher. With a usual temperature coefficient muVco of around -0.3%/C, and a reference temperature
59

Chapter 4 Project design

Project design

gain of 20C (+10C instead of -10C usual in middle Europe), this will reduce the Voc value by 6%. Otherwise, if your sizing overcomes the voltage limit, you can also suppress one module in each string during the first operating months.

Inverter / Array sizing

The inverter power sizing is a delicate and debated problem. Most inverter providers recommend (or require) a PNom array limit or a fixed Pnom (inverter/array) ratio, usually of the order of 1.0 to 1.1. But we have to notice: - The Pnom of the inverter is defined as the output power. The corresponding input power is PnomDC = PnomAC / Effic, i.e. about 4 to 6% over. - The Pnom array is defined for the STC. But in real conditions, this value is very rarely or never attained (the power under 1000 W/m and 25C is equivalent to that under 1120 W/m at 55C if we take a Pmpp = -0.4%/C). - The power distribution is strongly dependent on the plane orientation, - But the maximum powers are not very much dependent on the latitude: by clear day and perpendicular to the sun rays, the irradiance is quite comparable, only dependent on the air mass, - Most inverters accept a part of overload during short times (dependent on the temperature of the device). This is not taken into account in the simulation, and may still reduce de calculated overload loss, - When over-sized, the inverter will operate more often in its low power range, where the efficiency is decreasing. If the inverter is properly designed, when the Pmpp of the array overcomes its PnomDC limit, it will stay at its safe nominal power by displacing the operating point in the I/V curve of the PV array. Therefore it will not undertake any overpower; simply the potential power of the array is not produced. There is no power to dissipate, no overheating and therefore no supplementary ageing. In this mode the only energy loss is the difference between the Pmpp "potential" power and the PnomDC limit values. We can see on the power distribution diagrams, that even when the inverter's power is a little bit under the maximum powers attained by the array in real operation, this results in very little power losses (violet steps by respect to the green ones). The simulation - and the analysis of the overload loss - is therefore a very good mean for assessing the size of an inverter. Please see the power histogram in the "System" definitions, button "Show sizing". For an economical optimization, the final overpower losses are to be put in balance with the price difference with an inverter of higher power. These considerations often lead to very undersized inverters by respect to the manufacturer's recommendations (Pnom ratio of the order of 1.25 for optimal 30 south planes, up to 2.2 for a south faade).

Grid Inverter sizing

Please note that the inverter sizing should take into account the fact that: the inverter nominal power is defined as the device output power. The corresponding input power has to be increased by a factor 1 / Efficiency (about +4 to +8% at maximum power). the array nominal power is defined at STC 206 (1000 W/m, 25C). Under operating conditions the module temperature, mismatch and other losses decrease the effective array output power of at least 15 to 20% from the given nominal power.

Therefore, for proper operation an inverter nominal power about 20-25% below the array nominal power is sufficient. Moreover, an incident irradiation of 1000 W/m is rare on usual arrays. For example, faade installations in medium latitudes don't usually receive more than 800 W/m. Thus the inverter fine sizing requires a

Chapter 4 Project design

60

Project design
preliminary simulation to determine the Array output distribution.

Finally an economic optimisation can compare the price of the lost energy, to the economy realised when under-sizing the inverter. Such optimisations showed that for medium-Europe the inverter optimum size could be a factor of 1.2 to 1.4 below the array nominal power at STC, or even a factor of 2 for faades.

Different sub-systems, multi-MPPT inverters

You can define up to 8 different and independent sub-systems within a same project, with different inverters, different modules types, different numbers of modules in series. Provided that for a given subsystem (i.e. a given MPPT input), the PV-array is homogeneous. You may also define sub-systems for different orientations time.
31

, but only 2 orientations may be defined at a

During the simulation, each sub-system will be modelled independently, but only the energy and losses sums will be available as outputs. If you want to study the losses of a specific sub-array you should perform a simulation of this sub-system alone. This feature also allows the use of multi-MPPT inverters: you can define a subsystem for some MPPT input (s), and a different one for others. In this case the number of MPPT input should be used in place of the number of inverters. NB: Some new inverters (namely the Tripower series of SMA) have 2 MPPT inputs with very different powers. In practice this is very useful as you can define an array without much constraints about the module number on the main input, and one string with the remaining of your modules to be installed whatever their number - on the secondary input. PVsyst is indeed not foreseen for treating such assymmetric MPPT inputs. Nevertheles you can overcome the problem by modifying the number of MPPT inputs in the inverter definition (an save this as a new inverter). Please set here your total number of strings. After that, you can define one subfield with the (N-1) virtual MPPT inputs, and a second subfield with one string and the remainder of the modules. In the reality of your final wiring the (N-1) MPPT inputs will of course be one only MPPT input.

SolarEdge Architecture
The SolarEdge distributed architecture is based on a unique system design approach, characterized by a distributed DC-DC power optimizer for each PV module (or group of PV modules). These optimizers, with a current-driven output, are connected in series as strings, which are then connected in parallel to the input of a special (proprietary) inverter, operating at a fixed input voltage. Each power optimizer can manage 1 to 4 PV modules, and performs the MPP tracking at the module level. The output current is fixed by the inverter, and the output voltage is adjusted accordingly, in order to deliver the maximum available power from each module. In the simulation model, the efficiency loss of the power optimizer will be included in the inefficiencies of the global inverter. The Power Optimizers in a string operate completely independently one from the other, so that at a given time the module productions may be different without any affect on the system (shadings, mismatch, different orientations, etc). The main relevant parameters of a Power Optimizers are: Input parameters VmppMin, VmppMax and VabsMax, which determine the possible number of PV modules to be connected at the input, in the same way as for any MPPT inverter input. PPBmax: the maximum power admissible on the device. May be the Nominal power of the PV modules; or better for optimal sizing: the maximum attainable Power of these modules in this system, under real conditions. VPBoutMax : the maximum output voltage of the Power Optimizer (for example 60V). VPBoutMin : the minimum output voltage of the Power Optimizer (for example 5V).

61

Chapter 4 Project design

Project design
IPBoutMax . the maximum output current of the Power Optimizer. PmaxAC: The nominal output power to the grid. PMaxDC: the corresponding power at the DC input, i.e. VinvNom: fixed input operating voltage IDCMax: corresponding to the maximum power (= PMaxDC / VinvNom). PmaxAC / efficiency at PMax

The main parameters used for the sizing of the inverter are:

The string length limitations are Minimum string length: given as VinvNom / VPBoutMax (ex: 350V / 60V = 5.8, or 6 Power Optimizer). In practice we count at least one more module in order to ensure a good shading tolerance. Maximum string length: first condition is VinvNom / VPBoutMin (ex: 350V / 5V =70 Power Optimizer) . Maximum string length: second condition is for the max. power for the inverter: PMaxDC / PPBmax. Maximum string length: third condition is for the max. output current of the Power Optimizer.

For the sizing of the PV array, a good practice in PVsyst is to consider PPBmax as the maximum operating power in the real conditions of the system, established according to the incidence meteo data distribution on the collector plane. Power Optimizer parameters definition The available Power Optimizers and their parameters may be displayed in the SolarEdge inverter definition (specialized inverter parameters dialog). As the Power Optimizer are proprietary devices of SolarEdge, their parameters are not directly modifiable, but only defined in the original database of PVsyst. System sizing As with any usual system, you are advised to start by specifying the required power for your subfield (or the available area). After that you have to choose a PV module. When choosing a SolarEdge inverter, the system sizing dialog will change to a suited dialog for the SolarEdge architecture, and predefine the number of required inverters for your system size. You have first to choose the Power Optimizer to be used in your system (in the PV module group). Then in the Array design part, please define the Power Optimizer input configuration, i.e. the number of PV modules connected to each Power Optimizer (according to number of available inputs). Then you define the inverter input: The number of Power Optimizers in Series. The limits described above are shown on the right of the edit box. The nominal power corresponding to a whole string is shown, as well as the part of the inverter capacity (in percent). This very important information indicates how many identical strings you can connect on one inverter. For example if more that 50%, only one string of that length can be connected to each inverter. The number of Strings in Parallel. When one only string is allowed per inverter, this will be limited to the number of inverters. Below 50% capacity, this will be 2 times this number, or more

You are of course advised to use the "Show Sizing" tool for visually checking the sizing of this sub-field. Systems with different strings When you have strings with different lengths, you should define different sub-arrays, one for each length to be defined. In this case within a subfield, only a part of the inverter will be used for each string. Therefore you should define "Uses fractional Input" option, and define the inverter fraction to be used for this string. When several inverters are used, this will be the fraction for one inverter times the number of inverters. This will allow the use of the complement of each inverter within another subfield, with the suited fraction (for example 70% in one subfield and 30% in another one). Chapter 4 Project design
62

Project design

The total number of inverters defined in the whole system appears in the "Global system summary" table at the top right of the dialog. When inverter inputs are connected to strings of different lengths, the program is not able to check the full compatibility of your system. You should check by yourself that the defined fractions are compatible with the foreseen strings. Simulation With distributed SolarEdge architecture: there are no mismatch losses, the near shadings should be defined as "Linear", i.e. without string partition.

Linear shading is a good approximation for the shadings in SolarEdge architecture. As opposed to the option "according to module strings", that is used for the upper bound of electrical losses with regular inverters. The most exact way for SolarEdge simulation would be to define a string partition with rectangles corresponding to the number of PV modules in series at the input of one Power Box. These very little rectangles will reduce drastically the usual electrical shading losses observed with full strings, except in very regular cases like shed arrangement, where each module of the lower row becomes unproductive as soon as the bottom cell is shaded.

Normalised Grid voltages

The electric power is distributed using a tri-phased grid. In such a configuration, we distinguish two voltages: - The voltage between two phases Vpp, - The voltage between phase and neutral conductors Vpn, which is the usual connexion for the distribution in the house for most domestic appliances. The ratio between them is the square root of 3: Vpp = sqr(3) * Vpn. In European countries, the normalised low voltage in households is: Vpn = 230 V +6% / -10% (i.e. ranging from 207 to 244V) i.e. Vpp = 400 V +6% / -10% (i.e. ranging from 360 to 424V) In the US zone , there are several standards: Vpn = 120 V (no more very usual) i.e. Vpp = 208 V (some homes wired between 2 phases ?) or i.e. or i.e. Vpn = 236 V Vpp = 408 V Vpn = 277 V Vpp = 480 V (or 240 ? - Which range ?) (or 416 ? - Which range ?) ( ??? )

NB: Sorry, we don't have precise information about the situation in other regions.

Stand-alone system definition

First step: - You are asked to define the user's needs Second step: array and battery pre-sizing: - give the desired acceptable LOL probability - give the requested autonomy 200 , - choose the battery (system) voltage 20 ,
54 ,

proposed by default as a list of domestic appliance use. ,

203

The program will then perform a system sizing, in a similar manner as in the "Presizing" section.

63

Chapter 4 Project design

Project design
Third step: define the system components: - choose the battery model in the database => the program determines the number of units in series and parallel.

-choose a PV module model => the program determines the number of modules in series and parallel, according to the battery voltage. PVSYST cannot support mixing module types in a single system. Fourth step: Pass to the System configuration parameters ("Next" button): - the program asks for defining the regulator
159

This may be chosen in the database, with constraints specific to each commercial model (operating voltage thresholds, input and output currents, etc ), But for the first simulations of a project, it is recommended to use the "Generic Default regulator", which ensures a "standard" behaviour of the system, regardless of regulator constraints. In this case the regulator parameters are adjusted by default values corresponding to the actual system at the simulation time (for example, charging/discharging thresholds according to the battery pack configuration). This way, they don't produce compatibility warnings as the existing models do. - It is possible to use a power conditioning unit (MPPT or DC-DC converter). In PVsyst, these devices are part of the Regulator component definition. There are also "Generic default" devices for both MPPT and DC-DC operation. In these "Generic default" components, the parameters for efficiency and fixed voltage input value (DC-DC) may be modified inside the device. These values will be reported in the simulation variant parameter. NB: if you want to modify some disabled values in the Generic components, you can "save as" this component into a different file, and modify it as you like. Alternatively, you can permanently adjust the default values using the "Hidden parameter" 167 tool (expert users only!). - If desired you can also define a back-up generator
156

to obtain a stand-alone hybrid system.

You can now play with these parameters, and retrieve the automatic proposed values by clicking on the associated checkbox at any time. Warnings will be displayed if there are some incompatibilities between the chosen parameters. Red warnings are not acceptable (simulation cannot be performed) and orange warnings are indicative. These colours will be thrown back on the "System" LED's button. Fifth step: If desired (in a further step of the project), you can - modify the PV-array specific loss 83 parameters (thermal, wiring resistance, module quality, mismatch, IAM) with the "Detailed Loss " button. With a new simulation, the PV-array losses are pre-defined at standard "reasonable" values. These standard default values may be adjusted (by "expert users") in the "Hidden parameter" 167 section.

Regulator Operating modes

When using a power conditioning converter, this device is defined in the "Regulator" parameters part. When choosing a commercial regulator, this choice will automatically be updated according to the chosen regulator capabilities. With the "Default regulator" option, you have to choose the desired operating mode here, in order that the program can select the suited "Generic Default" regulator.

DC-grid system definition

This is to be applied to some Public Transport networks, which operate at DC voltage in the 600 - 1500V range. Upper voltages are difficult to carry out because of the maximum allowable voltage of the common PV modules. Such an installation doesn't require power conditioning, nevertheless very heavy safety disconnector devices should be foreseen for short-circuit protection in both directions. Please : Chapter 4 Project design
64

Project design
give the desired nominal power,

give the operating DC-voltage of the grid, the maximum allowed voltage on the grid, and line resistance 65 from the array's output up to the end-user (usually a vehicle). The line resistance from the conventional supply to the user should be defined only if a load profile is specified. chose the regulation behaviour safety voltage.
66

when the PV-array output voltage overcomes the maximum grid

choose a PV module model => the program determines the number of modules in series and in parallel.

You can now play with these parameters. The program cannot determine the exact optimum of the number of module in series. Therefore you should find this optimum by performing several simulation runs and compare the maximum system yearly efficiency. If desired (in a second step), you can define a user's load profile
65

modify the PV-array specific loss 83 parameters (thermal, wiring resistance, module quality, mismatch, IAM) with the "Detailed Loss " button.

DC-grid load profile

For reasons of security and ohmic losses in the overhead lines, the public transport networks are usually divided into little islands of relatively modest dimensions (of the size of a suburb). The power needs are therefore highly variable : while a starting tramway can absorb 1 to 2 MW within a few seconds, during the rest of the time, only a small number of vehicles (or often no vehicle) have to be supplied. The only "load" which is to be satisfied will then be the "ribbon-consumption" of the vehicle's heating systems and other services specific to the network. Due to the fact that the PV installation is unable to store energy, these very special characteristics require a preliminary and in-depth study of the on-site power demand, and its spread in time (power distribution analysis with a few-second sampling). PVSYST is able to treat this case through the "Hourly probability distributions" 54 load profile. We can assume that when averaging over a long period, an hourly probability profile should be statistically equivalent to few-seconds peaks with the same probability distribution. With significant PV installations, the PV-array power sizing should keep the over-energy (PV energy which cannot be used by the load) at a reasonable level. This could be a serious limiting condition to the system size when the ribbon consumption is low.

DC-grid line resistance

The PV-to-user line resistance is taken downstream the array output, up to the end-user (vehicle). It should not be confused with the PV-array ohmic resistance, which is distributed along the PV-array connexions (see shema). With undefined load profile (i.e. illimited load), the array output voltage is assumed to be the given "Main power supply" parameter. In this case the voltage at the user level is the nominal voltage minus the line ohmic loss; the supply-to-user line resistance is meaningless, as well as the overvoltage protection. When a load profile is defined, the "Main power supply" is the conventional supply nominal voltage. The program determines the user voltage (after the Supply-to-user line resistance voltage drop), and the PV output voltage (user voltage increased with the PV-to-user voltage drop). The PV current is then calculated according to the array I/V characteristics. The load current is the difference between the PV current and the user's need current. If the PV array cannot deliver it's current due to low load, it's voltage will increase according to the array I/V characteristics, and could reach the overvoltage limit. All these currents and voltages are interdependent, and have to be calculated from current balances in a loop procedure.

65

Chapter 4 Project design

Project design
DC grid: Overvoltage regulation

The output voltage at the PV field terminals is the end-user's voltage (bus, tramway), to which is added the line loss corresponding to the PV current through the line resistance PV-to-user. When the user's needs become too weak, this voltage will increase following to the PV array I/V characteristics, and may reach the array Voc value if there is no more load at all. It is therefore necessary to introduce a regulation system, avoiding the possibility that this voltage goes over the maximum admissible voltage for the network. This regulation can operate according to 3 different modes: - Global cut of the PV system, - Progressive decrease of the PV power by cutting off some strings of modules, - Progressive decrease by cutting off some modules in each string (which is not a very practical solution).

Pumping system definition

See also the Generalities
68

about pumping systems.

The complete definition of a pumping system involves several aspects, which are split in a "pipe" of 5 to 6 dialogs, which communicate through "Prev"/"Next"-like buttons. First step: Choose one of the three available systems: - Pumping from a deep well 69 , to a tank storage, - Pumping from a lake or river 70 , - Pumping into a pressurized tank 71 , for water distribution. and define the Hydraulic Circuit configuration (Storage Tank Second step: Define the water needs (may be yearly, seasonal or in monthly values). Define the pumping static depth if varying along the year (may be seasonal or monthly). Third step: Choose a pump model, taking the nominal Head into account (pumps are coloured in green for suitable, orange for not optimal, or red for not suited devices). Choose the number of pumps for reaching the required nominal power (all pumps wired in parallel
75 72

and Pipes

72 ).

Choose a PV module (also Green/Orange/Red), and a suitable PV array configuration (proposed by PVsyst). NB: This last choice is dependent on the System Configuration, which will be chosen during the next step. You will have to come back to this step for eventual correction after your Configuration choice. Fourth step: press "Regulation" Button Choose the Regulation mode 74 : again the Green/Orange/Red colours indicate the suitable choices, according to the previously chosen System type, Pump model and Pumps number. A collection of specific Warning messages explain the reasons of incompatibilities or poor design. The chosen Regulation strategy fixes the available set of regulating devices. There is a Default Regulating device corresponding to each strategy, with all parameters automatically adjusted according to the system (the systemdependent parameters will be re-evaluated just before the simulation). If "Battery Buffered" configuration, you still have to define the battery pack. In a next version of the program, it will be possible to specify an eventual back-up generator. Fifth step: open the Controller/Regulator Device All system running specificities are defined in the Controller/Regulation device. This includes namely the boundary operating conditions (Tank full, dry running, Power, Voltage, Current limits, etc.) In most cases the controller holds parameters specific to the system configuration strategy, which should be defined by the user (for example, Irradiance thresholds for pumps cascading or array reconfiguration, converter input voltages, etc).

Chapter 4 Project design

66

Project design
Back to Third step:

After that, you should return to Third step for checking the PV array definition. Please note that now the PV module list includes an indication of the MPPT nominal voltage, which makes easier the choice of a suitable module for a required operating voltage. The PV array sizing is particularly difficult when dealing with Direct coupling 75 configurations: in these cases you have to match the Array Voltage with the Pump Characteristics. The "I/V matching" button shows a specific graph which should help for this task. For MPPT or DC converters, the PV array voltage should usually be chosen as high as acceptable by the converter. This way you will have more chances to overcome the "Step-down" 160 voltage limitation. Ultimate step (after your first simulations): You may open the usual Array Losses define the wire sizes ).
83

dialog for modifying special array and wiring parameters (perhaps

Pumping system sizing

When sizing a PV pumping system, the basic constraints are the availability of solar energy during the year, and the satisfaction of the user's water needs. The problem to be solved is the optimisation of the size of the photovoltaic generator and the pumps, taking the head and the electric PV-Pump matching into account, as well as the chosen System configuration 74 . We give an idea of the procedure for a first, rough estimation: We start with the pump sizing. We first determine the Hydraulic Energy 73 needs over one day, assuming that Flow and Head are rather constant over the year (otherwise the day-by-day rough simulation provided in the presizing tool 21 is unavoidable). As a thumb rule, we can assume that on rather good days, the pump will run at its equivalent-full operation during about 6 hours; i.e; provide a flowrate [m/h] of about the daily water yield [m] / 6[h]. Assuming now a pump efficiency (usually about 50% for positive displacement, or 35-40% for centrifugal pumps), we can deduce the Pump Electrical nominal power suited for these clear day conditions. We should of course increase this value, accounting for the possible bad weather conditions over the year. Here the day-by-day simulation is necessary if we want to determine accurately the nominal power required to get a predefined LOL 203 . At this stage, we should emphasize that the system layout has a great influence on the pump nominal power; namely if the threshold losses are important, as it is the case with "Direct Coupling" simple systems. PV-array sizing Again as a thumb rule, we can choose the nominal PV STC power as about 20%-30% over the pump nominal power. Oversizing the PV-array will result in unused energy by clear weather. Undersizing it will operate the pump at lower powers, where its efficiency may drop or the thresholds dramatically affect the yield by cloudy, or morningevening conditions. When using direct coupling configurations, the array voltage is also essential. Detailed optimizations can only be given by a set of detailed simulations, using real devices. Tank sizing The tank size is simply determined by the required autonomy, using the daily consumption defined by the user, and assuming no water production. Other secondary characteristics of the pumping system should be determined in a second step: wire diameters between PV array and pump, pipe sizing, etc. These are involved in the Detailed Simulation Process. Moreover, the sizing may be subjected to criteria which may take on different weights depending on the use: 67 Reliability of the supply, and the consequences of no-delivery periods (may be overcome by a back-up generator), Investment and maintenance costs, which should take into consideration the cost of the PV generator, pump

Chapter 4 Project design

Project design

(s), regulation, and maintenance of the system. With battery buffered systems, also the initial cost of the batteries, as well as that of their maintenance and replacement. Durability: Quality of the pumps and regulators, ease of maintenance and replacement, special wearing conditions like sands or impurities in the water, etc.

Pumping Systems: Generalities

Solar Pumping System sizing and optimization is a rather complex task. Hydraulic and Electric requirements are strongly coupled, with highly non-linear interdependence which prevent an easy understanding of the behavior and performances of the complete system. Most pump manufacturers do indeed propose their own "standard" system configurations, or graphical tools for evaluating the production of their pump during a normalized day. But these dont give any information about the yield in real conditions, over the whole year, according to fluctuating needs or depth, and in a given climate. PVsyst deals with three types of pumping systems: - Pumping from a deep well 69 , to a tank storage; boreholes usually have some limitations (drawdown level depending on the flow pumped, static level variable over the seasons). - Pumping from a lake or river 70 ; here also the level may be seasonal. - Pumping into a pressurized tank
71

, for water distribution.

The general problematics is the following: the customer usually specifies: - Its water needs in volume, over the year, - The head (level difference) at which it should be pumped, - He may also require an autonomy storage duration (as the production is always dependent on the sun), - For a reasonable sizing, it is sometimes necessary to define a time fraction during which the owner will accept that the needs are not met by the system. ("Loss of load" probability). According to these requirements, and for a given site/meteo, PVsyst should allow to optimize the pumping system, and assess the results of a given configuration. As for the other systems, it proposes two complementary approach: - The Presizing tool 21 gives a quick but rough evaluation of the pump and PV array sizes, on the basis of very few parameters. - The Project Design 22 _Design part performs detailed simulation of the system, with many possible PVpumps coupling strategies. This provides namely a detailed analysis of the losses, useful for identifying the weaknesses of the design and optimizing the system.

Pumping: Borehole Modelling

See also Deep well system
69

for definition of the variables HD and HS.

If we consider the borehole as an impervious tube, when pumping the water level will drop as the flowrate Q [m3/ h] divided by the hole section area Aw [m]. On the other hand, the re-filling of the well from the surrounding porous medium is a diffusive process. One can admit as a reasonable hypothesis that the refilling flowrate is proportional to the stress, i.e. the drawdown dynamic head. Under these hypotheses the real level in the well (or HD evolution) will obey the following equation: dHD / dt = - 1/ * HD + Q (t) / Aw One can easily see that for steady-state conditions (dHD / dt = 0), this equation leads to a drawdown height HD linear with the flowrate. Indeed, compared to a reference case, we have for any flowrate: HD = Q * HDref / Qref Under this hypothesis, the ratio HDref/Qref is a characteristics of the well, which we will call the "specific drawdown" (expressed in [meter / m3/h]). This parameter is mainly related to the geologic properties of the surrounding ground (permeability, storage capacity), and the construction technique of the borehole. It may be measured rather easily, using a portable engine-pump and measuring the water depth and flowrate in stabilized conditions.

Chapter 4 Project design

68

Project design
Borehole parameter in PVsyst

As a matter of fact, a pumping test is often performed for measuring the borehole performance, which yields essentially 3 parameters: the static level (HS), a reference flowrate available from the well Qref, and the corresponding dynamic level (HDref). Navarte (2000) reports several results of such tests in Africa, of which we give some examples. HS
[m] Angola Rotunda 20 Chamaco 12 Lupale 20 Morocco Abdi 13 Ourika 17 Iferd 10

HDref
[m] 25 20 24 22 2 50

Qref
[m3/h] 7.2 6.9 5 21.6 10.8 36

HD / Q
[m / m3/h] 3.5 2.9 4.8 1.0 0.2 1.4

We can observe from these examples that the Dynamic contribution is not to be neglected ! The recovery time (corresponding to a 1/e re-filling) is easily calculated from the steady state conditions: = Aw * HDref / Qref

For example, in the case of a borehole of diameter 0.15 m in Rotunda, this is about 4 minutes. Therefore this dynamic model describes the short term behavior of the well. Medium-term (annual) variations are likely due to modifications of the phreatic water level along the seasons. They may be introduced in PVsyst by specifying a monthly profile of the static head HS. Long-term exhausting effects caused by an excessive water drain involve complex (and not sustainable) phenomena which are not modelled here in PVsyst. Finally, the simulation (as well as the real system regulation) should take the maximum head Hmax, i.e. the inlet level of the pump, into account for stopping the pump, avoiding dry-running.

Pumping: deep well system

See also: Pumping Systems:Generalities
68 .

Many solar pumping systems are of the "Deep Well" type, i.e. made of a submersible pump placed in the bottom of a borehole. The borehole is usually sunk by specific machines, with a diameter of about 12 to 20 cm. Special submersible pumps are designed for being inserted in such boreholes. They should of course lie below the water level, and are connected to the surface by a pipe for the water, and the feeding/control electrical wires. The water is pumped into a storage tank, according to sun availability. Remember that the Pressure or Head 73 is mainly related to the difference between the input and output levels. The pump has to provide a total head resulting of several contributions. In PVsyst we take reference to the ground level, we have (cf fig): HT = HG + HS + HD + HF where: HG = head due to the height of the outlet pipe above the ground (assuming that outlet pressure is negligible). HS = static head due to the depth of the water level in the well, in absence of any pumping. HD = dynamic "drawdown" head: in a borehole well, the effective water level is dynamically lowered by the water flow extraction (see below). HF = friction losses in the piping circuit, which depend on the flowrate.

69

Chapter 4 Project design

Project design

4
HG Storage Tank

Ground level

HS Static water level

Hmax HD

Water level when pumping

Pump inlet level

Pump

For this system, in the "Pumping Hydraulic definitions" dialog, you will be asked to specify: - The static depth. This may be also be given in seasonal or monthly values, in the "Water Needs" next dialog. - The maximum pumping depth, corresponding to the inlet aspiration level. The system will stop the pump when the dynamic level reaches this level, avoiding dry running. - The Pump depth, should be below the maximum pumping depth, - The borehole diameter (in cm), - The specific drawdown, expressed in [m / m3/h]: this is a characteristics of the borehole and the surrounding ground (see Deep well modelling 68 ). You will also define the Storage Tank
72

and Hydraulic circuit

72

parameters.

A little graphical tool shows the total head and its contributions, as a function of pump flowrate.

Pumping system: lake or river to storage

See also: Pumping Systems:Generalities
68 .

Pumping systems from a lake or river are similar to deep well systems, but with some technical simplifications: - The pump may be placed near the source (no more than 4-5 m above the water surface, less at high altitudes, for avoiding cavitation problems). - The pump is not necessarily of "Submersible" type, therefore much cheaper. On the other hand, it's maintenance is more easy. Remember that the Pressure or Head 73 is mainly related to the difference between the input and output levels. The pump has to provide a total head resulting of several contributions. In PVsyst we take reference to the ground level, we have (cf fig): HT = HG + HS + HF where: HG = head due to the height of the outlet pipe above the ground (assuming that outlet pressure is negligible). HS = static head due to the depth of the water level, by respect to the ground. HF = friction losses in the piping circuit, which depend on the flowrate.

Chapter 4 Project design

70

Project design

For this system, in the "Pumping Hydraulic definitions" dialog, you will be asked to specify: - The lake or river level depth, by respect to the ground. This may be also be given in seasonal or monthly values, in the "Water Needs" next dialog. - The Pump depth, may be at most 5 m over the source depth, You will also define the Storage Tank
72

and Hydraulic circuit

72

parameters.

A little graphical tool shows the total head and its contributions, as a function of flowrate.

Pumping pressurization system

See also: Pumping Systems:Generalities
68 .

This system assumes pumping from a generic water source (other storage, lake or river), into a tank which ensures a water static pressure allowing for distribution to customers. This is an alternative of "high" tanks like water towers. The pressurization is obtained by the compression of the air in the closed impervious tank volume when water level increases. The pump's problematics is the same as for lake/river, except that it's maximum head capabilities are usually higher: - The pump may be placed near the source (no more than 4-5 m above the water surface, less at high altitudes, for avoiding cavitation problems). - The pump is not necessarily of "Submersible" type, therefore much cheaper. On the other hand, being in open space, it's maintenance is more easy. Remember that the Pressure or Head 73 is related to the difference between the input and output levels, to which the pressurization should be added. The pump has to provide a total head resulting of the following contributions. In PVsyst we take reference to the ground level, we have (cf fig): HT = HG + HS + HF + HP where: HG = head due to the height of the outlet pipe above the ground (assuming that outlet pressure is negligible). HS = static head due to the depth of the water level, by respect to the ground. HF = friction losses in the piping circuit, which depend on the flowrate. HP = Pressurization contribution required for distribution.

For this system, in the "Pumping Hydraulic definitions" dialog, you will be asked to specify: - The source level depth, by respect to the ground. This may be also be given in seasonal or monthly values, in 71

Chapter 4 Project design

Project design
the "Water Needs" next dialog. - The maximum pressure in the tank (stop pumping), - The minimum pressure in the tank (stop feeding the users), - The pressurization (air) volume at minimum pressure. You will also define the Storage Tank
72

and Hydraulic circuit

72

parameters.

A little graphical tool shows the total head and its contributions, as a function of the pump flowrate.

Pumping: Storage tank As the solar yield is of course not constant, all pumping systems include a storage tank for receiving the pumped water when available, and distributing it to the final users. The storage tank bottom should usually be above the ground (or user's) level, in such a way that the static pressure is sufficient for distributing the water by gravity (water tower principle). With the Pressurized systems this requirement is of course not useful. The tank characteristics are requested from the user in the "Hydraulic Circuit Definition" dialog. The user should define: - The tank storage volume, - The tank diameter (if rectangular, an equivalent diameter for the same area, i.e: Diam = SQRT (Length * Width * 4 / Pi) - The water height in the tank, when full (related to Vol and Diam), - The alimentation mode, which may be: - Free output feeding by the top (usual situation). In this case you should define the feeding altitude (by respect to the ground). The outlet of the pipe is supposed without any other head loss. - Bottom alimentation: the pumping pipe output is at the bottom of the tank, requiring a non-return valve. This operation mode avoids the level drop between top and tank level, and may slightly improves the system efficiency, as the actual level for the head calculation is the tank level at each time. The required parameter is the bottom altitude. The tank volume should be sized according to the daily user's needs, and the required autonomy. Pumping: Piping circuit The pipes circuit produces friction head losses, which have to be kept at a reasonable value. The "Hydraulic Circuit Definition" dialog asks for the choice of a Pipe type (including diameter), and a total piping length. It allows inclusion of a number of elbows, as well as eventual other Friction Loss factors hydraulic singularities, valves, etc.
74

for diverse

The overall friction head loss appears on the little graphical tool (below the green line), as a function of the flowrate. It should be kept negligible at nominal flowrates of the system, by choosing the appropriate pipe diameter. Water needs The water needs may be specified yearly (constant value), or in monthly / seasonally values. Specifying needs in terms of hourly values (daily distribution) doesn't make sense, as most of the time the pumping system includes a storage for at least one day of consumption. The sizing part cannot take these variations into account and will be established using the yearly average. Of course, the detailed simulation will rely on these specified values at each time step.

Chapter 4 Project design

72

Project design
Static depth The static depth is defined with the deep well system definitions.

This represents the depth of the groundwater level, which may vary along the year. Therefore it is possible to re-define it as seasonally or monthly values. However the sizing part cannot take these variations into account and will be established using the yearly average. Of course, the detailed simulation will rely on these specified values at each time step. Pump motor technology

Pump motor technology

The motor technology is not crucial in PVsyst. This specification will be mentioned in some result outputs. Brushless DC motors seem to have the higher efficiencies. The only relevant information used by the program is whether the motor is AC or DC.

Hydraulic Power and Energy The mechanical power of a Hydraulic flow is basically the product of the fluid flowrate, by the head at which it is transferred. When using usual units in PVsyst: P hydr [W] or: = FlowR [m/h] * Head [Bar] * 1000/36

P hydr [kW] = FlowR [m/h] * Head [Bar] * 1/36

In the same way, the energy is related to the total water volume transferred E hydr [kWh] = Flow [m] * Head [Bar] * 1/36 or E hydr [MJ] = Flow [m] * Head [Bar] * 1/10 Head and Pressure units In solar pumping systems, Head is usually expressed in units of Level difference [meter or feet]. The pressure at the basis results of the water column weight. Physically, passing to pressure units involves multiplying the height by the water density (1000 kg/m) and the gravitation constant (g = 9.81 m/s). And for getting [bar] we have to divide by 100'000 [Pa/bar]. In summary, we have the following equivalences: 1 Pa 1 bar 1 bar 1 bar 1 bar 1 bar 1 bar 1 bar = 1 N/m (mksa basic unit) = 100 kPa (definition of the Bar) = 10.19 mWater (passing from level difference to pressure) = 33.44 ftWater (Idem) = 2088 lbs/ft2 (Pounds/Square feet) = 14.504 PSI (PSI = Pound/Square Inch) = 0.987 atm = 750.1 torr or mmHg

1 mWater = 0.0981 Bar 1 ftWater = 0.0299 Bar 1 PSI = 0.069 Bar

73

Chapter 4 Project design

Project design
Pumping: Hydraulic losses

In the losses flow (as shown on the Loss diagram) of non-battery pumping systems, we have chosen to account for all electrical losses even when the pump is stopped for hydraulic reasons. In this situation the ELowLev and ETkFull contributions are referred to the EPmpAv (Available useful energy at pump) energy. An alternative way would be to refer the hydraulic ELowLev and ETkFull losses to the EArrMPP energy, and to consider the system as "not running" during these situations. In this case the hydraulic arrow losses would come upstream of the other losses, with increased "virtual" energy hydraulic losses referred to EArrMPP. The results are equivalent, the account of the effective hydraulic losses being transferred from the Array and Converter to the EArrMPP quantity. NB: In Battery Buffered systems, the interpretation of the "Available energy at pump" is not well defined, so that we preferred the second alternative. Friction Loss Factors Some examples of the Friction Loss Factor values, for piping circuits (may be added for several singularities): 45 elbow, standard 90 elbow, standard 90 elbow, long radius Tee, along run, branch inactive Tee, used as elbow Gate valve, Gate valve, Gate valve, Gate valve, Diaphragm Diaphragm Diaphragm Diaphragm Butterfly Butterfly Butterfly Butterfly open open open open valve, valve, valve, valve, open open open open 0.35 0.75 0.45 0.4 1.5 0.17 0.9 4.5 24 2.3 2.6 4.3 21 0.24 1.54 10.8 118 10 70 7.0 15 6.0

valve, valve, valve, valve,

5 20 40 60

Non-return valve, disk Non-return valve, ball Water meter, disk Water meter, piston Water meter, Turbine

Pumping system configuration

Several system layouts and coupling strategies are possible: - Direct coupling 75 between the PV array and the pump(s), without power converter unit. This is of course only possible with DC motor pumps. Although often used owing to its simplicity, this layout requires a very careful electrical optimization, and cannot yield a good efficiency in any operating conditions. It may be improved by several special regulation modes: - Booster device 76 , electronic device for overcoming high starting current due to friction losses, - Cascading 77 when using several pumps, - Array reconfiguration 78 (half-array serie/parallel switching), not common but rather simple technique involving simple controlling device with relays. - Systems with Power conditioning units 78 , matching the voltage/current PV characteristics to the specific electrical needs of the pumps (DC motor behaviour, mainly current-driven) using modern electronic switching

Chapter 4 Project design

74

Project design

techniques. On the solar side, these can be either "Maximum Power Point Tracking" (MPPT), or with "fixed voltage" DC input. Such a device is of course necessary with AC motor pumps (inverter). - Battery buffered 80 systems, where a battery is used to regulate the pump running in time. The PV-Battery behaves like a standard stand-alone system, and the pump is always running in optimal conditions, at the nominal battery voltage. A strategy of switching the pump ON/OFF according to the irradiance level may be used to minimize the battery solicitation. Nevertheless the simulations show that a high charge/discharge rate is unavoidable in this configuration, leading to rapid wearing of the battery pack and a high rate of replacement. You can have a look on the simulation Results Controller By the way, even the simplest configurations (direct coupling) require the presence of a control device, which should at least assume the following functions: Manual Power ON/OFF Pump OFF when tank is full Pump OFF when aspiration level is below the pump inlet (preventing dry running), Eventually motor temperature protection, Protection against powers, currents or voltages which overcome the maxima specified for the pump (s).
67 80

for comparing the performances of these operating strategies.

The sizing constraints

are very dependent on the system layout.

Pumps: parallel and series connexions

Pumps in parallel and in serie

Either on the electric side or on the hydraulic side, you are advised to connect all the pumps in parallel. At the moment PVsyst accepts electrical connexions in serie only for centrifugal pumps with DC motor. Other configurations don't make sense. The reasons are: On the electric side, connecting two positive displacement pumps in series will prevent good starting conditions; after one pump has started, overcoming its peak starting current, the current will suddenly drop to the operating value; therefore the total current will be limited so that the second pump may never reach its own thresholds current. On the hydraulic side, it is probably not a good practice to connect two pumps in cascade for obtaining a higher head in the same flow, as non-linearities in the pump behaviour, or electric feeding differences, may lead to very unbalanced heads. This is especially true for positive displacement pumps. It is far better to choose a pump model which undertakes the nominal foreseen head.

Regulation: Direct Coupling

See also Pumping Configuration
74

, next

76

Regulation.

Direct coupling between the PV array and the pump is only possible with DC motor pumps. The simplified electrical layout is the following:

75

Chapter 4 Project design

Project design

The following figure shows a typical pump behavior, superimposed on the I/V array characteristic. An equivalent figure - with your real components - is available in the "System" definition dialog in PVsyst, when you are choosing your Pumps and PV modules.
PV array and Pump Characteristics
12 1000 W/m, 55C 800 W/m, 50C 600 W/m, 45C 10 400 W/m, 40C 200 W/m, 35C 8
Current [A]
Head = 40m

. Pump characteristics
. Head = 20m

DC-DC converter
0
0 10 20 30 40 50 60 70 80 90

Voltage [V]

Such a configuration implies in a very careful optimization. At any time, the operating point is the intersection of the two characteristics: PV production and pump consumption. If the pump curve is too high (array current undersized), the pumping threshold will be high, penalizing the low irradiances (low season, bad days and morning/evening). If it is too low, the full potential power of the array is not used during bright hours. The optimal sizing is therefore depending either on the irradiance distribution (i.e. location, orientation, meteo), and on the periods at which the water needs are the more important. Moreover, the pump characteristic is strongly dependent on the head, displacing the curve parallel to itself. Therefore, the sizing will also be dependent on the conditions of use, impeding using simple "thumb rules" valid for any system at any place.

Regulation: Direct Coupling with Booster

See also Pumping Configuration
74

, previous
75

75

and next

77

Regulation.

This is an improvement of the Direct coupling

configuration, often necessary with DC displacement pumps:

Chapter 4 Project design

76

Project design

Most displacement pumps require a significant peak of current (at low voltage) when starting, in order to overcome the internal friction forces. We see on the PV characteristics diagram that the array current is not able to provide the peak unless by waiting very high insolation, increasing the irradiance threshold. Help is usually provided by an electronic device named "Booster", which stores the PV energy in a big capacity and gives it back as a peak of current. This strategy is useable with a single pump system. When several pumps are involved in the system, the Cascade 77 configuration is best suited.

Regulation: Direct Coupling with Cascading

See also Pumping: Configuration
74

, previous

76

and next

78

Regulation.

If the system is equipped with several pumps, the regulation should switch them ON according to the available PV power, in order that each pump runs near its optimal efficiency.

This opportunity improves drastically the performances of the direct coupling irradiance threshold and improving the operating at high irradiances.

75

configuration, lowering the

But be careful: the determination of the irradiance threshold for starting the second pump is of great importance

77

Chapter 4 Project design

Project design
for the final performances (see the Results
80

4
analysis for details) !

For pumps with MPPT power conditioning units, the cascading cannot be used unless the MPPT algorithms are suited for "Master/slave" operation. Indeed, the operating point of the characteristics of the PV array cannot of course be driven simultaneously by two independent MPPT devices. Such an operation mode is not yet implemented in the present version of PVsyst.

Regulation: Direct Coupling with Array Reconfiguration

See also Pumping Configuration
74

, previous

76

and next

78

Regulation.

Direct coupling mismatch may be improved by performing a PV-array reconfiguration: if we consider two identical groups of PV-modules, at low irradiance all groups are connected in parallel, providing the high currents necessary to the pump starting. From a given irradiance level, the groups are connected in series, doubling the voltage and reducing the current of the PV array. This requires an electronic switch of rather simple technology (cf Salameh, 1990).

This strategy is not advised when several pumps are used: the Cascading operation in this case.

77

is probably more suited

But be careful: the determination of the irradiance threshold for commuting the arrays is of great importance for the final performances (see the Results 80 analysis for details) !

Regulation: Fixed DC input converter

See also Pumping Configuration
74

, previous

78

and next

80

Regulation.

Use of a DC-DC converter (Power Conditioning Unit) shows a much favourable figure than direct coupling. This cheap electronic device absorbs the power of the PV array at a fixed voltage, and behaves as a current generator for feeding the DC-motor of the pump.

Chapter 4 Project design

78

Project design

At the input side, the voltage may be chosen close to the maximum power point, and stays quite near for any irradiance. On most commercial DC-DC devices, the input voltage may be adjusted by hardware. PVsyst includes a specific tool optimisation 163 for determining the optimal DC voltage setting. Performances are only related to the array and meteo, they don't depend on the pump configuration.
Array average effic. by respect to MPP operation
120%

By respect to MPP operation

Average MPP operation (67 V)

100%

80%

60%

40%

Yearly array Yield at fixed operating voltage (maxi 99% at 65.5V)

20%

0% 40 45 50 55 60 65 70 75 80

Fixed operating voltage

At the output side, the power is supposed to be transmitted to the motor at the optimal current/voltage point corresponding to the available power. For AC pumps, a PCU (DC or MPPT converter) suited for a given pump is usually proposed by the pump's manufacturer. It is supposed to fit the inputs requirements (voltage and frequency) for proper operation. Efficiency Nowadays, the converter efficiency is usually of the order of 95% in the high powers region. This efficiency drops toward low powers as other similar devices like inverters. It is treated as such in PVsyst: an efficiency profile is constructed using the maximum and "euro" average efficiency, defined in a similar way as for inverters. By the way, efficiency drop often arises at powers which are below the hydraulic threshold of the pump; therefore it doesn't affect the normal running. Of course the DC-DC converter also plays the role of a "Booster". The starting high current is usually required under very low voltage, therefore low power. You can have a look on the Results 80 , which confirms that the Fixed Voltage DC technology gives performances approaching the MPPT, and is not very sensitive to the fixed voltage.

79

Chapter 4 Project design

Project design

Step-down technology We have to point out here a design constraint:: most of the DC-DC converters operate on "step-down" principle. This means that they cannot deliver a voltage greater than the input voltage. Therefore the PV-array MPP voltage should be over the maximum voltage required by the pump at the maximum desired flowrate. This means for example that when using standard pumps designed for 230VAC grid operation, the PV array should deliver at least 325V for obtaining undistorted sinus at the pump. This "step-down" limitation can be taken into account by the simulation only when the voltage behavior of the pump is well defined. With pumps specified only by power curves, it is neglected.

Regulation: Battery buffered configuration

See also Pumping: Configuration
74

, previous

78

Regulation.

This can be understood as a regulating device like a very big capacity, which operates over the time:

Conceptually, in this operating mode, the battery should not be sized for storing energy over a medium or long period - the water storage in the tank is far more efficient for this task. It should be meant for yielding a power complement when the sun's power doesn't reach the pump's power threshold, and also absorb the excess energy when it overcomes the pump's maximum power. This way the battery capacity may be reduced to a few operating hours. In practice the pump is connected to the battery, and operates at the fixed battery voltage at any time. It could be regulated in the same way as any other consumer in a stand-alone system, i.e. turned OFF according to the battery discharge threshold protection. But this would lead to a very intensive use of the battery, in a domain (low charge state) where the wearing is very important. It would be far better to turn the pump ON only when the sun already yields a significant power, but just not sufficient for activating the pump. This way the battery may be understood as a "power regulating" device. Therefore the regulating device should act according to an irradiance level sensor (in conjunction of course with the discharge protection of the battery), with a threshold carefully chosen, in such a way that it starts a little before the pump's threshold. The detailed simulations should help for determining this threshold in each given situation, in order to optimize both the water yield and the battery wearing conditions. Please have a look on the Results with power converters.
80 , which

indicate that the Battery Buffered performances approach the ones

Pumping Results: Examples

We would like to give here an example for comparing the performances of the different possible configurations of the system.

Chapter 4 Project design

80

Project design
This corresponds to the project given as Demo for pumping systems.

We have imagined a Deep Well system at Dakar (Senegal, 15 latitude), with a static level depth of 32m, and a tank feeding altitude of 6m. The water needs are set at 4 m3/day, all the year. The system includes a set of 2 pumps of 100 W each, positive displacement with DC motor, supplied by 4 PV modules of 60Wp, i.e. 240Wp (not ideal system of course: one only bigger pump would be better in this case, but this configuration allows to apply all the system configurations). We simulated this system will all configuration options available in PVsyst. This gives the following results:
Direct coupling Direct with Booster Cascade 500 W/m2 775 676 521 0 0.0% 35.0% 56.6% 26.2% 45.8% 72.0% 1161 290 782 5 0.6% 52.5 56.4 983 469 660 0 0.0% 44.4 56.5 27.3% 28.6% 55.9% Cascade 680 W/m2 1399 53 932 88 8.6% 62.6 56.9 5.9% 9.5% 15.4% Array Reconfig 680 W/m2 1436 23 946 108 10.2% 63.6 57.5 6.1% 9.9% 16.0% 1469 0 977 409 29.5% 65.6 57.1 1469 0 978 371 27.5% 65.7 56.9 1469 0 977 395 28.8% 65.7 57 1469 975 341 25.9% 65.5 57.1 MPPT Conv. 26V DC Conv. 28V DC Conv. 30V DC Conv. Battery 50 Ah 400 W/m2 1469 0 988 389 26.1% 66.4% 56.5% Battery 50 Ah 680 W/m2 1171 280 786 550 36.9% 52.8% 56.7%

Needs 4 m3/day
Threshold Water pumped Missing water Energy at pump Unused PV energy Unused PV energy System efficiency Pump efficiency Loss under pump starting Loss under prod. threshold => Loss under thresh

36.6% 36.6%

2.0% 2.0%

2.3% 2.3%

2.3% 2.3%

2.3% 2.3%

0.0%

0.0%

Pumping system performances acc. to configurations

1800

Pumped water Missing water Excess energy

Water [m3] or Energy [kWh]

1600 1400 1200 1000 800 600 400 200 0

Direct coupling

With power converter

te r

v.

Ar ra y

ire c

M PP

tw

tte r Ba

26

28

ire c

as c

as c

30

We can observe that all direct coupling configurations show lower performances than systems with power converters. The direct coupling yields half the needs only. Although it is very dependent on the PV array sizing, the performances of such a configuration are always deceiving and strongly depend on details of the system sizing. Introducing a booster device improves the situation; this overcomes the starting over-current set at 1A for this pump (in the lack of manufacturer's information, we have fixed this value arbitrarily; this corresponds to half the normal threshold current for this head). 81

Chapter 4 Project design

Ba

tte r

ad

ad

W /m y 68 0 W /m

fig

v.

v. on C C DC

W /m

W /m

pl in

Co n

on

os

on

Re c

Bo

ou

on y 40

v.

50 0

68 0

DC

ith

DC

tc

Project design

Cascading of the two pumps can be an efficient improvement (but in practice needs 2 wells - or a very wide one for passing output pipes near the higher pump). But be careful: the starting threshold of the second pump is a crucial setting! If too low, the starting current will stop the running of the first pump, while in a very efficient irradiance range.

The Predefined Graphs "Flowrate function of Irradiance" in the results is a suitable tool for understanding the behaviour and optimising the threshold ! Here for the 500 W/m and 680 W/m thresholds. The array reconfiguration option shows quite similar performances as pump cascading. Although suitable controllers seem to be not available on the market, it is very easy to construct without deep investment in sophisticated electronics. And it may be used of course with one only pump, which avoids the practical disagreements of the pump's cascading. As for cascading, the threshold determination is very important and should be carefully determined with the same tools. Power converters offer of course the best results. And as expected, MPPT converter is the more efficient. But we can observe that DC-DC converters with fixed input voltage are almost as efficient, and that the fixed voltage value is not critical. This could lead to cheaper converters, without MPPT algorithm implementation. It also allows to use standard converters, with standard (not solar) pumps. Finally, the Battery Buffered system gives equivalent results as power converters. At the condition that the Pump starting threshold (by respect to irradiance) is set sufficiently low. Otherwise the pump is not running sufficiently often, and the battery becomes often over-charged. But of course it involves the use of a (little) battery pack, which have to be replaced periodically. Water needs increase We tried to increase the water needs to 5 m3/day, the limit for not missing water with the MPPT. The results are given below: They are quit similar, and confirm that the converter technological options are rather equivalent. But the results of Cascading and Array Reconfiguration are a little worse by respect to converters.

Chapter 4 Project design

82

Project design
Pumping system performances acc. to configurations
2200 2000 1800 1600 1400 1200 1000 800 600 400 200 0

4
Pumped water Missing water Excess energy With power converter

Direct coupling

Water [m3] or Energy [kWh]

D

Array losses in PVSYST

See also: Array losses in PV systems
83 , general

In PVSYST, Array loss parameters are initially set to reasonable default values, so that modifications only need to be performed during a second step of the system study. PVSYST treats in detail the following loss types in a PV array: Thermal losses
84 87 86

Ohmic wiring losses Module quality losses Mismatch losses

87

Incidence angle (IAM) losses

Press the "Losses Graph" button for visualizing the effect of these losses on the PV-array I/V characteristics in given running conditions. In the simulation results, the effect of each loss will be available in hourly, daily or monthly values. They may be visualized on the Loss diagram 97 .

Array losses, general considerations

Generally speaking, array losses can be defined as all events which penalise the available array output energy by respect to the PV-module nominal power as stated by the manufacturer for STC conditions 206 . This is the philosphy stated by the JRC/Ispra European Centre recommendations, through the Normalised performance index 98 . Several of these loss sources are not directly measurable. Starting with incident irradiation in the collector plane (after taking irradiation shading effects into account), one can imagine that an ideal PV-array should yield one kW/kWp under an irradiance (Ginc) of 1 kW. That is, assuming a linear response according to Ginc, the ideal array will produce one kWh energy under one kWh

83

Chapter 4 Project design

co up lin g w it h C Bo as os ca te de r 5 C 0 0W as ca /m de 2 68 0 W /m Ar 2 ra y R ec on fi g M PP T C on 26 v. V D C Co nv 28 . V D C Co 30 nv V . D C Ba Co t te nv ry . (4 00 Ba W t te /m ry ) (6 80 W /m ) D ire ct
considerations.
86 .

ire ct

Project design
irradiance for each installed kWp (as defined at STC). This ideal yield is diminished by the following losses: -

Incidence angle modifier (IAM) 86 , is an optical effect (reflexion loss) corresponding to the weakening of the irradiation really reaching the the PV cells surface, with respect to irradiation under normal incidence. Irradiance Loss 203 : the nominal efficiency is specified for the STC (1000 W/m), but is decreasing with irradiance according to the PV standard model. Thermal behaviour 84 of the PV array. The standard test conditions are specified for a cell temperature of 25 C, but the modules are usually working at much higher temperatures. The thermal loss is calculated following the one-diode model 182 . For crystalline silicon cells, the loss is about -0.4 %/C at MPP. For fixed voltage operating conditions, the temperature mainly affects the I/V curve voltage, and effective losses are strongly dependent on the array overvoltage by respect to the operating voltage. The parameters available to the user (thermal loss factor) involve the cell temperature determination by respect to given external conditions. Real module performances 86 of the module by respect to the manufacturer specifications. PVSYST uses effective specification parameters to calculate the primary PV-array characteristics. The user may define a relative loss factor, which is related to the average effective module power at STC, and acts as a constant penalty during all simulation conditions. Mismatch losses 87 of the PV modules, which can be evaluated by a special tool, but is only taken into account as a constant loss during simulation. Dirt on the PV-modules, which is not taken into account in PVSYST, but could behave as mismatch losses, i.e. the string current will be mainly affected by the worst cell. Partial shading electrical effects, limiting each string current to the more shaded cell, are of course depending on the sun position. They are not explicitely calculated in PVsyst, but can only be roughly evaluated using the "Near shadings according to modules" 46 . MPP loss, i.e. the difference between the effective operation conditions and the maximum available power point. For MPP use (grid inverters) this loss is neglected in PVSYST. For fixed operating voltage, it can be quantified from the output simulation results (see EArrMPP, EArrUFix, MPPLoss 101 ). Ohmic wiring losses 87 , as thermal effects, essentially result in a voltage drop of the I/V-array characteristics. The real effect is different whether the array operates at MPP or fixed voltage. At MPP operation, PVSYST applies the wiring loss before computing the MPP. At fixed voltage, the effective losses are strongly dependent on the array overvoltage by respect to the operating voltage. Regulation loss is the energy potentially available from the PV array, but which cannot be used by the system. In MPP applications, this could be the array potential PV production outside the inverter input voltage limits, or during power overloads. This is usually accounted in "Inverter losses", that is in system losses. In stand-alone systems, it corresponds to the excess energy which cannot be used when the battery is full. In DC-grid installation, this is the potential current in excess by respect to the instantaneous load current.

In Normalised performance index 98 , all these array losses are accounted for in the "Collection Losses" Lc, that is the difference between Yr (the ideal array yield at STC 206 ) and Ya (the effective yield as measured at the output of the array). Nevertheless, unlike Ispra recommendations, in PVSYST the unused energy is specifically designed as Lu = Unused loss (see Normalised performance index 98 for details).

Array Thermal losses

Thermal Model
The parameters of the Thermal behaviour of the field are defined in the "Array Losses" dialog, available from the system parameter definition (see also Array Losses definition 83 ). The thermal behaviour of the field - which strongly influences the electrical performances - is determined by an energy balance between ambient temperature and cell's heating up due to incident irradiance: U (Tcell - Tamb) = Alpha Ginc (1 - Effic) where Alpha is the absorption coefficient of solar irradiation, and Effic is the PV efficiency (related to the module area), i.e. the removed energy from the module.

Chapter 4 Project design

84

Project design

The usual value of the Absorption coefficient Alpha is 0.9. It is eventually modifiable in the PV module definition dialog. When possible, the PV efficiency is calculated according to the operating conditions of the module. Otherwise it is taken as 10%. The thermal behaviour is characterised by a thermal loss factor designed here by U-value (formerly called Kvalue), which can be split into a constant component Uc and a factor proportional to the wind velocity Uv : U = Uc + Uv v (U in [W/mk], v = wind velocity in [m/s]).

These factors depend on the mounting mode of the modules (sheds, roofing, facade, etc...). For free circulation, this coefficient refers to both faces, i.e. twice the area of the module. If the back of the modules is more or less thermally insulated, this should be lowered, theoretically up to half the value (i.e the back side doesn't participate anymore to thermal convection and radiation transfer).

Determination of the parameters

The determination of the parameters Uc and Uv is indeed a big question. We have some reliable measured data for free mounted arrays, but there is a severe lack of information when the modules are integrated. What value should be chosen according to the air duct sizes under the modules, and the length of the air path ? One can observe that the heat capacity of the air is very low. Even with large air vents, the flowing air under the modules may quickly attain the equilibrium with the modules temperature at the end of the duct, leading to no heat exchange at all. Therefore for the top of the array the U value may be the fully insulated U-value; you can have big differences between the regions of the array near the air inlet, and at the outlet. PVsyst doesn't take this inhomogeneity of the array temperature into account. On the other hand, the use of the wind dependence is very difficult. On one hand the knowing of the wind velocity is extremely seldom (some programs construct synthetic hourly values from monthly data, but on which basis and with which models ?). On the other hand the "meteo" wind velocity (taken at 10 meter height) is not representative of the velocity at the array level (there may be a factor of 1.5 between them). In this respect the Uv value is obviously not the same for these two definitions of the wind velocity.

Default and proposed values

Up to version 3.4, the default values proposed by the program were Uc = 20 W/mk, Uv = 6 W/mk / m/s
208

values which had been measured on several installations coverage, with free air circulation all around.

, for modules mounted in sheds without back

These values matched very well the measurements (average deviation of the order of 0.5 to 1C), but were defined using Wind data recorded just over the array, in built environment. Such velocities are quite lower than the usual wind data available in the meteo files, recorded at a 10m level or more over the obstacles. Therefore with these parameters PVsyst calculated a very high dissipative coefficient and cool array temperature, leading to strongly underestimated heat losses. According to their own measurements, some users proposed, when using standard meteo values such those in the US TMY2 data (usually about 4-5 m/sec on an average), the following k-values: Uc = 25 W/mk, Uv = 1.2 W/mk / m/s

By the way, when the wind velocity is not present in the data (as it is the case for all synthetic meteo files constructed by PVsyst), PVsyst transfers the wind-dependent contribution into the Uc factor, assuming an average wind velocity of 1.5 m/s (or 5 m/s in the latter case). Since version 4.0, the default value is fixed for free-standing arrays, as : Uc = 29 W/mk, Uc = 15 W/mk, Uv = 0 W/mk / m/s Uv = 0 W/mk / m/s If you have fully insulated arrays, this should be halved: Concerned people agreed that this is an acceptable choice. NB: our recent measurements on an horizontal array of amorphous frameless modules, mounted not jointive at 8 cm above a steel roof, showed a value U = 18 W/mk. These values suppose an average wind velocity of around 1.5 m/sec at the collectors level. In very windy regions 85

Chapter 4 Project design

Project design

(larger average wind velocities), you can increase the values; but we cannot say by which amount in a reliable way. NOCT Values Some practicians - and most of PV module's catalogues - usually specify the NOCT coefficient ("Nominal Operating Collector Temperature"), which is the temperature attained by the PV modules without back coverage under the standard operating conditions defined as: Gincid = 800 W/m, Tamb=20C, Wind velocity = 1 m/s, Open Circuit (?). The NOCT factor is related to our loss factor U by the thermal balance (from the expression of the top): Alpha 800 W/m (1 - 0) = (U + U 1m/s) (NOCT - 20C).
c v

In the definition dialog, the user may define either the U factors or the NOCT. The program immediately gives the equivalence (using Alpha=0.9 and Effic = 10%, without wind dependence).

Array incidence loss (IAM)

The incidence effect (the designated term is IAM, for "Incidence Angle Modifier") corresponds to the weakening of the irradiation really reaching the the PV cells's surface, with respect to irradiation under normal incidence. In principle, this loss obeys Fresnel's Laws concerning transmission and reflections on the protective layer (the glass), and on the cell's surface. In practice, it is often approached using a parametrisation called "ASHRAE" (as it has become a standard in this American norm), depending on one only parameter bo: FIAM = 1 - bo (1/cos i - 1), with i = incidence angle on the plane.

For single-glazed thermal solar modules, the usually accepted value for bo is of the order of 0.1. But in a PV module, the lower interface, in contact with the cell, presents a high refraction index and our specific measurements on real crystalline modules actually indicate a value of bo = 0.05. The user is free to choose his parameter bo, or even to define any special profile shape as a function of the incidence angle. The custom profile can be easily drawn on the graph with the mouse.

Module quality losses

In the past, it was well-known that most of PV modules series didn't match the manufacturer nominal specifications. The real behaviour of modules by respect to the specifications was one of the greater uncertainties in the PV system performance evaluation. Now, with "guaranteed" power assertions and increasing availability of independent expertises, the situation seems going toward some clarification. Module series are sold with a given tolerance, final flash-test assertions, and actual powers usually may lie under the nominal specified power, but stay in the tolerance. PVSYST allows for accounting for this discrepancy by defining a loss factor, which is an energy loss at MPP, constant during the simulation process. For fixed voltage operation (battery, DC grid), the loss factor is supposed to be the same. The Module quality loss is a parameter that should express your own confidence to the real module's performance, by respect to the manufacturer's specifications. It is at your entire disposal: you can put it at any value (for example for expressing the LID "Light Induced Degradation" after some few hours, suggested by some manufacturers, or the long-term losses, or keeping some reserve on the production warranty, etc). By default, PVsyst puts it at half the inferior tolerance, meaning that you can have a module sample with average power between the lower tolerance and the nominal value.

Chapter 4 Project design

86

Project design
Array mismatch loss

Losses due to "mismatch" are related to the fact that the real modules in the array do not rigorously present the same I/V characteristics. A graphical tool helps for visualising the realistic behaviour of such an array, with a random dispersion of the characteristics of short-circuit current for each module. This tool allows for the quantification of power-loss at the maximum power point, as well as of current-loss (usually higher!) when working at fixed voltage. The simulation asks for a Mismatch loss factor (different for MPP or fixed-voltage operation), which is taken as constant during the simulation.

Array ohmic wiring loss

The wiring ohmic resistance induces losses ( R I ) between the power available from the modules and that at the terminals of the array. These losses can be characterised by just one parameter R defined for the global array The program proposes a default global wiring loss fraction of 3% by respect to the STC 206 running conditions (i.e; Vmpp/Impp, i.e. a resistance-like quantity). But note that the energy loss behaves as the square of the current (or the power), so that at half-power the ohmic loss is divided by a factor of four. The effective loss during a given period will be given as a simulation result. It is usually of the order of one half to 60% of the above specified relative loss when operation at MPP. The program offers a special tool
87

intended to optimise the wire diameters at each stage of the array.

This dialog also asks for the voltage drop across the series protection diode. The corresponding energy loss will also appear in the simulation results.

Wiring loss optimisation

The aim of this tool is to optimise the wire diameters in the array, given a predefined loss limit. First give a look on the array usual wiring schema, by pressing the "Schema" button. This identifies the different parts of the circuits by colors: - The "string" connections between all modules of a string, up to the wiring connection box (usually situated near the array). - The connexion between this boxes and the system (inverter, battery, ), - Optionally, when choosing "Groups of parallel string", the connection between the group boxes and the global array box. Secondly, you have to specify the average wire length for each circuit category (total length, i.e. minus and plus poles for each loop). The program shows the nominal current and the resistance of one branch, as well as their contribution to the whole array resistance (as seen from the array connections). It proposes a list of standardised wire sections, beginning from the smallest wire compatible with the actual loop current.

The "wire" button allows to see the standard wire properties (diameters, maximum current, resistivity), as well as their price, which you can define in order to optimise the wiring cost. The maximum wire currents are given by the European Norm, for isolated wires mounted in apparent or ventilated mounting ducts (CEI 364-5-523, 1983)
Finally, you can specify a maximum loss: - either in terms of % (at MPP) when operating at MPP, - or in voltage drop when operating at fixed voltage. The program will then optimise the wire sections matching these requirements, either minimising the copper mass, or the wiring cost providing you have defined the wire prices in the "wire" dialog.

87

Chapter 4 Project design

Project design
AC ohmic loss from inverter to injection point

The AC wiring losses may simply be defined by the distance between the inverter and the injection point. The program will determine the minimum section of the wires, and only propose suitable sections if you want to increase it. Inversely you can also define a loss fraction at STC, and if the section is defined the corresponding length will appear, as well as the voltage drop at STC. Securities with weak grids Please note that the sizing of the cables up to the injection point may be very important when the grid is "weak" (country regions). Indeed when injecting power into the grid, the grid voltage will increase due to line impedance. Then your inverter is equipped with a safety device, which will cut the production when overcoming a given maximum voltage. Therefore you are advised to minimize the voltage drop, at least in the parts where you have the opportunity of doing this.

Maximum current in wires

The maximum admissible current in conductors is specified namely by the IEC recommendations. The limits are due to temperature, and related to the dissipation of the joule heat, and therefore depend on the layout (environment of the wire, grouping of several wires, etc). PVsyst takes the minimum sections for the nominal currents into account, supposing the less constraining layout, i.e. "C" mounting mode, wires in a duct at free air. It is your responsibility to check the compatibility of the proposed sections with the specific layout conditions of your system. For example, ducts exposed at sun may encounter high temperatures, and should be sized as if the environmental temperature were 70C. This may lead to a reduction .of admissible current of the order of 40%. For more information about cable installation, please consult for example http://www.electricalinstallation.org/enwiki/General_method_for_cable_sizing The Current limit in PVsyst is calculated according to the norm IEC 364-5-523 (1983), following the formula: Imax = ImaxRef * Section [mm] ** Esect where: Copper: Aluminium: Exponent : ImaxRef = 13.5 A/mm ImaxRef = 10.5 A/mm Esect = 0.625 (where ** means Power)

Chapter 4 Project design

88

Project design

Admissible current in a single cable

10 9

Current density [A/mm]

8 7 6 5 4 3 2 1 0 0 50 100

CEI 364-5-523, Copper CEI 364-5-523, Aluminium

150

200

250

300

Cable section [mm]

Very high currents We can observe that for high sections, the specific current becomes very low (of the order of 1.6 A/mm for 300 mm, i.e. a diameter of 20 mm). For still higher sections the exponential behaviour leads to very low and unrealistic values. For very high currents we can admit that the conductors will no more be round, but become flat busbars or sets of several bars in parallel, which will have different heat removal properties. Therefore we have put a limit to the specific current, i.e. set a limit section from which the specific current will remain constant. This limit has been fixed to 300 mm2, but may be modified in the hidden parameters. Installation mode correction We have also defined a modifying parameter in the hidden parameters, allowing to modify the maximum current according to some Installing configuration specificities. This parameter is set to 1 by default.

External transformer losses External transformer losses

In some big PV installations (in the MWp range), the transformer is not part of the inverter, but an external device directly connected to the MT grid. The main losses associated with the transformer are: - The iron losses (mostly due to hysteresis and eddy currents in the core) are proportional to the square of the core flux, i.e. to the square of the voltage. As we have a constant grid voltage, this is considered as a constant loss. - The ohmic losses either in the primary and in the secondary windings. These may be represented by an equivalent resistance, and the loss will be computed as R * I during the simulation.

Module Layout
The Module Layout is a tool for the description of the PV module geometrical arrangement in your system, and their interconnexions. This is presently only a descriptive tool, independent of the electrical design and the simulation of your PV system. Therefore it is not mentioned in your final simulation report. In a next version, this construction will be linked to the shading calculations for an accurate evaluation of
89

Chapter 4 Project design

Project design
the electrical shading losses. This tool defines:

- a set of subfields 91 , areas which will contain the PV modules. Each area is a rectangular surface, on which you can eventually dispose sub-rectangles or triangles 92 representing inactive areas (window, chimney, etc). In the future version for shading calculations, each rectangular areas will be identified to a subfield area defined in the 3D scene. - The Layout of the PV modules. The program distributes the modules on the available area, taking mechanical constraints, spacings, etc. into account. - The attribution of each module to a string defined in the "System" design part. Therefore each module will be linked to a specific inverter or MPPT input. In the present time, this is already a valuable tool for providing a wiring schema. - In a next version, this will lead to a detailed calculation of the electrical effect of shadings for each inverter input, replacing (for not too big systems) the evaluation "according to modules" in the simulation. Informative / Pedagogic shading effects In the present version, you already have the opportunity of defining a shade on the subfield (by mouse), and observe the electrical effect on each inverter circuit.

Construction and General description

Available area definition In the first tab "Mechanical", you can define subfields we will name "sub-field area".
91 ,

i.e. surfaces for receiving the PV modules, which

Each sub-field area is a rectangle-based area, in which you can include "secondary rectangles 92 " of any size for defining areas "forbidden for modules". These secondary rectangles may be positioned on corners, or anywhere in the main rectangle (representing for example a window, a chimney with its shaded environment, etc. This rectangle may also be transformed into a rectangular triangle, for example for representing a four-sided roof. Within such an inactive "secondary rectangle", you may also redefine an active rectangle for receiving modules. You can define as much "sub-field areas" as desired. These mechanical sub-fields areas, defined "geographically" on your building, are not related and should not be confused with the "electrical subfields" of the PV system definitions, corresponding to a set of module strings connected to one or several inverters. For sheds arrangements, you will define one sub-field for each shed (or tracker if you have a tracking system). For several disconnected roof elements, one for each element. Module arrangement After defining your available area, you can push the "Set Modules" button, which will distribute the PV modules on this area. For this operation, you have several choices: Chapter 4 Project design
90

Project design
Modules positioned vertically or horizontally (portrait or landscape), Start filling from left border, right border or centred (idem for vertical), Spacing of the modules in X and Y.

Then the program will tell you the number of modules which may be installed on this area. In the "General PV Array" information box, you have also the total number of modules to be positioned according to the system definition, and the number of modules left to be defined. Now you can modify your inactivating rectangles, and the modules will rearrange accordingly. If you have too much modules, you can delete some of them with the right button of the mouse (or restore them with the left button). Attribution of modules to system strings The "Electrical" tab allows to define the electrical properties of your PV modules layout, either for defining a wiring schema, and for future shading calculations. First you should define the number of by-pass protection diodes (should normally be part of the PV module definition in the database, but often not defined or not reliable: please carefully check this information on the datasheets). And you should define whether the sub-series of cells (protected by one diode) are cabled in length or in width in the module (useful for detailed shading calculations). Now the program shows all the strings defined in your system (in all electrical sub-fields), and gives a representation of all modules in each string. These strings are numbered for the whole system, whatever the electrical sub-field. You should attribute each positioned module to a given string. For this, please choose one string in the electrical representation, and you can distribute the modules corresponding to this string on the positioned modules by the left button of the mouse. Use the right button for removing an attribution. The not-attributed modules in the electrical part remain white. There is an option for showing the attributed string number on each module. This figure may be used on site as a wiring schema. Electrical shading effects After defining a full and valid layout, the button "Shades " appears, where you can draw a shade on the subfield(s), and observe the electrical effect for each inverter circuit. The shading effects are computed according to each corner of each submodule (part of module protected by one by-pass diode). One corner shaded is interpreted as 1 cell shaded, 2 or 3 corners as 1/3 or 2/3 cells shaded, and the 4 corners as a full shading. The effect of one or several shaded cells is indeed very similar, so that this approximation is quite sufficient. The electrical shading effect is only applied to the beam component. In the reality, even at bright sun there is a residual diffuse irradiance of the order of 15% or more. NB: This tool has now just an informative/pedagogical objective. The link to the 3D construction and the implementation in the simulation process will be done in a next version. The Layout definition will be saved with your project. At the moment it doesn't have any implication on the simulation. It is not part of the final report: you can print it in the module layout dedicated dialog.

Subfields
A "sub-field area" is a (geometrical) surface element for receiving the PV modules. This should not be confused with the subfields or sub-arrays that you have defined in the "System" definitions, which concerns the electrical arrangements in strings ("homogeneous" for a set of inverter inputs). You can construct here sub-field areas of any shape, for receiving the modules of your system. You can define as much sub-field areas as necessary, by the button "Add", independently of your electrical subarrays. For example in sheds arrangements, you will define a sub-field area for each shed. For complex systems, you are advised to give significant names (can be typed in the ComboBox) to each

91

Chapter 4 Project design

Project design

sub-field. These names will namely appear in the wiring schema output of this tool, assigning each module to an electrical string.

Secondary rectangles
Each sub-field area is a rectangle-based area, in which you can include "secondary rectangles" of any size for defining areas "forbidden for modules" (button "Add"). These secondary rectangles may be positioned on corners, or anywhere in the main rectangle (representing for example a window, a chimney with its shaded environment, etc. either by mouse or by coordinates. These rectangles may also be transformed into a rectangular triangle, for example for representing a foursided roof. When defining a triangle, please choose a corner for the initial shape/position. After that you can modify it as you like by mouse and coordinates. Within an inactive "secondary rectangle", you may also redefine an included active rectangle for receiving modules. After distributing the modules on the area, you can still define or modify the inactive rectangles and the module layout should adapt themselves automatically.

Simulation
When all parameters are acceptable (LED's all green or orange), the program gives access to the hourly simulation. Simulation dates are based on the Meteo file dates, and can be restricted to a limited period. The simulation process 92 involves several dozens of variables, which are stored in monthly values in the results file, and will be available as monthly tables and graphs. Nevertheless the program cannot store all these data in hourly values. Data of interest for the user should be defined before the simulation, in order to be accumulated during the simulation process. PVSYST offers three ways for the output of detailed hourly or daily data: Accumulating Hourly values: the user may choose a set of variable of interest, to be accumulated in hourly values. Each variable will occupy about 18 kB on the output .VCi file. By default, the program chooses about ten fundamental variables. These will allow for displaying detailed plots of one or several simultaneously variables, with a very comfortable navigating tool over the whole year. Special graphs 96 : the user can pre-define four kinds of special graphs (time evolution, scatterplot, histogram and sorted values) for any variable in daily or hourly values. About ten specific and usual graphs are already defined with each new simulation. ASCII export files 95 : the user can choose any among the variables, to be written in daily or hourly values on an ASCII file for exporting to another software (spreadsheet, e.g. Microsoft Excel). The ASCII file is generated during the simulation process.
96

After completion, the simulation allows for opening the "Results"

dialog.

Simulation results for a particular "variant" can be stored, with all involved parameters, in a file named as the project's file, with the extension .VCi (i = 0..9, A..Z).

Simulation process: irradiance and PV array

The simulation involves about fifty variables
98

, which are all accumulated in monthly values.

When starting, the early parameter definition parts in the program have already verified the consistency of all input parameters. In a first step, the diffuse attenuation factor 201 should be calculated, by integrating simultaneously shading and IAM attenuation factors over the viewed part of the vault of heaven. The same thing holds for the albedo attenuation factor 200 .

Chapter 4 Project design

92

Project design
Then the hourly simulation performs the following steps, for each hour: Incident "effective" energy calculation -

Reading one hour data on the Meteo file (Horizontal global irradiation, temperature, eventually diffuse irradiation and wind velocity). If not available, computes diffuse irradiation with the Liu-Jordan correlation model, If necessary, applies the horizon correction on the beam component, => defined variables at this stage: GlobHor, DiffHor, BeamHor, T Amb, WindVel
98 .

performs the transposition (global, diffuse, albedo irradiations) in the collector plane, using either Hay or Perez model according to user's preference. This is done using solar angles at the middle of the time interval, calculated from project's site (not the site of the meteo file). With explicitely given meteo files (TMY or own measurements), this could give not very reliable results at early morning or evening, if the time properties of the meteo file are not properly defined (legal or sun time), or if latitudes of the two sites is very different. => defined variables at this stage: GlobInc, BeamInc, DiffInc, DiffSInc, AlbInc
98 ,

applies the shading factor (if near shadings defined) on the beam component, applies the IAM factor on the beam component. => defined variables at this stage: GlobIAM, GlobShd, GlobEff, DiffEff
98

This leads to the so-called "Effective incident energy", i.e. the irradiation effectively reaching the PV cell surface. Other secondary variables (essentially ratios of the above energy quantities) are available for displays: => Bm/Gl, Diff/Gl, DifS/Gl, Alb/Gl, Ftransp, FIAMBm, FIAMGl, FShdBm, FShdGl, FIAMShd 98 . Array MPP "virtual" energy Then the simulation calculates - the array temperature (energy balance between absorbed and heat loss energy), - the MPP operating point of the array (i.e. the maximum available energy), that is the "virtual" energy - as if the system was perfectly running - available at the array's terminals. Thus the MPP energy EArrMPP does already include the array losses, i.e. thermal losses, wiring losses, module quality, mismatch and IAM losses. => defined variables at this stage: Tarrray, DTArr, DTArrGl, EArrMPP
98

For double orientation fields, the whole meteo calculation is repeated for the second field orientation, output meteo variables are accumulated as averages between the two orientations, weighted by the field area ratio. Then both array characteristics are electrically combined, in order to search the real maximum power point. System energy The next simulation stages are system dependent : Grid connected system Stand-alone system Pumping system DC-grid system
95 , 94 . 94 93 ,

NB. All energies are calculated here as average power during one hour. They are expressed in [kWh/h] or [MJ/h], that is in a power equivalent unit. Therefore with hourly steps Power and Energy hold the same numerical values. Although most calculations are indeed related on power quantities, we will express them as energies for simplification.

Simulation process: grid system

Once the Incident Global Effective irradiation and the MPP running conditions are determined, the grid system has to take the inverter behaviour into account, that is: If EArrMPP doesn't reach the inverter input threshold, the array is considered as open circuit (I=0 and U=Voc).

93

Chapter 4 Project design

Project design
-

If the inverter output power exceeds it's nominal power, it behaves according to the definitions of the inverter characteristics, that is: Limitation: the inverter searches, on the array I/V curve, the running point corresponding to it's output nominal power, by decreasing the operating current. Cut: the inverter input is cut off, the array is considered as open circuit, the power is null. Cut until morning: once the cut-off has occured, the inverter doesn't run until the next morning. When the MPP voltage reaches the minimum or maximum voltage of the inverter's MPP window, the array voltage stays at this value on the I/V curve. Otherwise, in normal operation, the inverter drives the array at the MPP, it's input energy is EArrMPP, and output energy is calculated according to it's efficiency curve.

For each case, the energy loss by respect to the MPP ideal running is independently accumulated, as well as the array operating voltage and current. Finally if user's load has been defined (corresponding to the own consumption of the user), self-consumed and re-injected energies are accounted. The main output variables calculated here are: InvLoss, EArray, EoutInv Euser
100

Simulation process: stand alone system

Simulation simultaneously manages Array production, Battery, eventually Back-up production, and the user consumption. At the meeting point (battery terminals), all voltages are the same and simulation has to perform a current balance. For each component, the current is a complex function of the voltage: PV-array: search the operating point on the I/V array characteristics, (irradiation and temperature already known), paying attention that ohmic, module quality and mismatch losses have an action on the actual current, for a given voltage Battery: voltage characteristics of the battery model depends on state of charge (SOC), temperature and current, Load: Given energy, states current as function of the voltage, Back-up generator: Given energy, states current as function of the voltage,

so that balance has to be achieved by successive iterations. Once the currents are determined, SOC and battery voltage are calculated for the end of the time interval. Moreover, the system behaviour depends on the regulation state. They could be: PV-array disconnected when full battery, Load disconnected in case of deep battery discharge, Back-up generator eventually running (if defined in the system) according to battery voltage.

Due to battery voltage evolution, these operating conditions may change during the time step. In this case the program determines the exact time when a regulator threshold condition is met, evaluates the energies for this hour fraction, and starts again a balance loop according to the new operating conditions. Several variables 101 are computed during and after this process: array running characteristics, battery storage and ageing, load an used energies, etc.

Simulation process: DC-grid system

System calculation is rather different whether a load profile is defined or not. With undefined (illimited) load profile: When the load consumption (vehicle) is not defined, the delivery of the conventional power supply is irrelevant. The voltage at the output of the array is the "Main Power Supply Nominal Voltage", with a fixed value. The treatment is easy, since the array operates at this fixed voltage: current and power are directly deduced from the array I/V characteristics model. Voltage at the user's level is decreased by the "PV-system line resistance" voltage drop. In this case the safety overvoltage regulation does not operate, since the line voltage stays at a fixed value, and the PV current production is supposed to be always absorbed. With defined load (usually "probability" load profile): In this case the simulation has to resolve a balance current at the user's level. User's consumption is determined by the load profile definition requirement. Available solar current is coming from the PV array, the rest being

Chapter 4 Project design

94

Project design
delivered by the conventional power supply.

Available solar current has to be determined from the I/V array characteristics, taking both line resistances (PV-touser, and conventional supply-to-user) into account. Two cases can arise: Either the PV production is not sufficient: in this case the line voltage at the array output will stay around the supply nominal value, only influenced by the line ohmic voltage drops. Or the PV production covers the user's needs, so that the conventional supply current drops to zero. In this case the line voltage follows the PV-array I/V characteristics, and can raise up to the Voc value when the user's needs drop to zero. In this case the regulation overvoltage safety should apply if the allowable maximum voltage is reached.
105

Several variables

are computed during and after this process.

Simulation process: pumping system

See also Simulation Process
92

: irradiance and PV array

Once the Incident Global Effective irradiation and the MPP running conditions are determined, the system simulation is dependent on the Pumping System Type 68 and Configuration 74 . In any case, and for any running hour, the simulation has to determine the Flowrate furnished by the pump, as a function of the Head and the available electrical energy (from PV array, or eevntually battery). As the head at the pump is dependent on the flowrate (due to friction losses, and eventually drawdown level), this calculation always results of an iterative process. Also at any hour (night and day), some water is drawn from the tank by the users (this quantity is defined by the specified "User's needs", and supposed to be constant, over the day and even over the month). The simulation has to manage the situations where the tank is full (limiting the pump's flow at the user's draw, and stopping the pump during the rest of the hour), and when the tank is empty (the user's needs cannot be satisfied). For obtaining a consistent balance, all energies (below threshold, before producing flow, or potential unused energies when the pump is stopped) should be carefully accounted for, in any running situation. Along with the running losses (like inefficiencies, or I/V mismatchs by respect to MPP), these numerous energy contributions allow to construct a coherent Loss diagram 97 , which is a powerful tool for identifying the system weaknesses. Nevertheless this losses structure is dependent on the system types and configurations. The simulation variables 103 are specific to a system layout, or the order of the losses take place in a different order. The final relevant results include mainly the water delivered to the users, the missing water, the excess (unused) PV energy, and the system efficiency during the year (or performance ratio). And if economical features are defined, the global investment, yearly costs and cost of the pumped m..

Export ASCII file

Simulation Export file allows for writing chosen daily, hourly or monthly values on an ASCII file in order to be used in another software. It should be defined before the simulation, as the output file is generated during the simulation process. Procedure Define the output file name, which will be placed, by default, in the PVSYST \ DATA \ UserData directory. Choose the fields format (separated or fixed fields), and the separator. The date and hour format (including Microsoft EXCEL-compatible dates), The type of values (Hourly, Daily or Monthly), The parameters to be accumulated.

Models The user has the opportunity of saving these definitions in a library, to reuse it in other projects. Note Although your file parameter definitions will be kept over different simulation runs, you may choose if you want to effectively produce the file or not. By default, after producing the first output it will turn on "No Output", and you will have to explicitely ask for an ASCII output before running the next simulation. Tip: 95 for monthly values, you will have a very fast way of exporting data by using direct export of the monthly result

Chapter 4 Project design

Project design
tables, either through a file or by "Copy"/"Paste" in the clipboard.

On-line graph definition

Graphs of hourly and daily values have to be defined before the simulation process. You can define up to 30 different plots (including about ten already defined by default). Procedure Four kinds of plots are proposed: For each of them, you have to chose Daily or Hourly values, and the units. Time evolution, you will define the parameter to be accumulated, and the time interval (which may be different of the simulation period). Scatterplot allows to show the dependence of one parameter on another one (for example: Input/Output diagram). You must define the parameter for each axis, as well as the analysis period. Values distribution will plot values as an histogram. Apart from the physical parameter to be used, you have to define the starting, the width and the number of classes ("bins"). In the first class, you may impose a threshold to eliminate, for example, null values at night. These definitions depend on the range of values taken by the parameter, which has to be known priori. In case of doubt, you can confidently ask for a larger range (up to 400 bins) as the programme will automatically truncate the empty classes below and above. The accumulation can take place either in time (number of hourly or daily occurrences), or in energy. Ordered values are histograms in which each class accumulates all lower values, and therefore give the distribution of values lower than a given value. The necessary definitions are the same as those for histograms. Please also give a name, which will be the plot title in outputs. You can use the button at the right to produce an automatic plot name.

Models The user has the opportunity of saving definition of each plot in a library, to reuse it in other projects. Please note that the definitions of histograms will have to be updated according to the size of each system. Nevertheless the default graphs histograms are automatically updated.

Simulation and comparison

When all parameters are acceptable (LED's all green), the program gives access to the hourly simulation. Simulation dates are based on the Measured data file dates, and can be restricted to a limited period. Besides the "Online graphs" and "Output files" decribed for the usual simulation 92 , you should also define the "Comparison" 165 requirements, i.e. which simulation variable has to be associated to each measured data, and under which constraints. Then the simulation process, beyond accumulating graphs as for pure simulation, will also accumulate couples of comparable data in hourly or daily values. After completion, the simulation/comparison allows to open the "Results"
96

dialog.

Simulation results for a particular "variant" can be stored, with all involved parameters, in a file named as the project's file, with the extension .CMi (i = 0..9, A..Z).

Results
The simulation results are summarised in a printable "Report", which holds an exhaustive table of all parameters used during the simulation, as well as a short description of the main results. But many other results may be visualised and printed: The simulation involves several dozens of variables variant" file. These monthly results are available as:
98 , which

are all stored in monthly values in the "Simulation

Chapter 4 Project design

96

Project design
Pre-defined tables: several tables, grouped by parameter themes, are immediately available.

Custom table: you can build your own monthly table by choosing eight among any of the calculated variables. Custom monthly graphs: you can choose up to 4 variables to be simultaneously displayed. Be sure to choose comparable variable types (energy, irradiation, etc) as the graph holds only one common axis.

Moreover, hourly values are stored for some pre-choosed variables. These give rise to hourly and daily plots (even with simultaneous variables) with a comfortable navigation all over the year. This constitutes a powerful tool for observing and understanding the instantaneous system behaviour. Specialgraphs 96 should be defined before the simulation, in order to be accumulated "on-line" during the simulation process. About ten such "on-line" graphs of general use are present by default with any new simulation. They include a detailed energy loss diagram 97 , Monthly or Daily normalised "yield" indicators index 98 , performance ratio, input/output diagram, incident energy and array output distribution, etc. Economic evaluation After the simulation you can perform a detailed economic evaluation (for example nb. of modules, inverters) and results into account. Printing You can choose to print the following forms: - General simulation parameters, which summaries all the parameters involved in a "variant", - Detailed simulation parameters, such as Horizon (drawing+points table), Near shadings, detailed user's needs, etc, - A pre-defined form with the main parameters and main results of this simulation, - The detailed loss diagram, - Any specific result graph or table displayed on the screen, along with the main parameters, - The economic evaluation sheet. The Result dialog offers the opportunity of recalling other "variants" of the project, in order to perform quick comparisons.
106

of your project, taking the parameters

Loss diagram
The loss diagram provides a quick and insight look on the quality of a PV system design, by identifying the main sources of losses. Little buttons enable to group the losses into general topics (meteo and optical, PV array, system, energy use) or to expand any topic to detailed losses. The loss diagram is available for the whole year, or for each month in order to evaluate seasonal effects of particular losses. Please refer to Array losses, general considerations
83

for a general explanation of the losses in PVsyst.

The array losses start from the rough evaluation of the nominal energy, using the global effective irradiance and the array MPP nominal efficiency at STC. Then it gives the detail of the PV model behaviour according to the environmental variables. In stand-alone systems, the diagram gives a detail of the battery use, that is which part of the energy effectively transits by the battery. Minimizing the battery use is of some importance for the lifetime (number of charge/ discharge cycles). NB: Each loss is defined as percentage of the previous energy quantity. Therefore the percent values are of course not additive: when grouping the losses, the overall percentage is not the sum of the detailed values! NB: The accounting of the individual losses is far from being trivial problem! The simulation process and some variable definitions had to be deeply reformulated for obtaining a coherent figure. And some contributions are impossible to evaluate rigorously. For example in stand-alone systems, the ohmic losses are evaluated using the usual relation Ploss=R * I. But in reality the array resistance modifies the PV operating point and the whole circuit equilibrium, so that a more accurate calculation would probably be simulating the complete system with and without this resistance, and evaluate the differences. But even with this method, some loss contributions will be transferred to other ones. By the way, even if some individual losses are not quite well determined, the Energy values at each main step of

97

Chapter 4 Project design

Project design
the simulation are in principle correctly calculated. You can refer to the following pages for a detailed description of the individual variables: Meteo, irradiation and PV array Grid connected system Stand-alone system DC-grid system
105 101 100

98 ,

Normalised performance index

In order to facilitate comparisons between several PV installations, JRC (European Joint Research Center) introduced the following Performance Index. These indicators are related to the incident energy in the collector plane, and are normalised by the Pnom = Array nominal installed power at STC 206 , as given by the PV-module manufacturer [kWp]. Therefore they are independent of the array size, the geographic situation and the field orientation. In these definitions the yield energies are expressed as [kWh / KWp / day]. In other words, these quantities are numerically equal to the Equivalent operating time under a constant irradiance of 1 kW/m, that is, they can also be expressed as [Hours/day] when running at 1 kW/m, or [kWh/m/day] (see the remark below). We define the following quantities: Yr = Reference system Yield is the ideal array Yield according to Pnom as defined by manufacturer, without any loss. It can be understood as each incident kWh should ideally produce the Array Nominal Power Pnom during one hour. Yr is numerically equal to the incident energy in the array plane, expressed in [kWh/m/day]. Ya = Array Yield is the array daily output energy, referred to the nominal power [kWh / KWp / day]. Yf = System Yield is the system daily useful energy, referred to the nominal power [kWh / KWp / day]. Lc = Collection Loss = Yr - Ya, is the array losses 83 , including thermal, wiring, module quality, mismatch and IAM losses, shading, dirt, MPP, regulation losses, as well as all other inefficiencies. Ls = System Loss = Ya - Yf, include inverter loss in grid-connected systems, or battery inefficiencies in standalone. PR = Performance Ratio = YF / YR, is the global system efficiency by respect to the nominal installed power.

For stand-alone systems (or every system with limited load), we also introduce: Lu = Unused energy, the potentially available energy at the array output, which can't be used because the system is "saturated" (full battery, or limited load in DC-grid system). This should be determined during the simulation, and we have: Ya = Yr - Lu - Lc. In this case Lc is the collection loss, only when the system is able to use the produced energy.

Important remark about units There is often a unit's confusion with the quantity Yr, which may be understood - either as the incident energy (with units [Hours at 1kWh/m / day] or [kWh/m / day]) - or as the ideal array Yield according to Pnom (expressed as [kWh / KWp / day]). This numerical identity results of the STC definition kWp of electricity.
206

, i.e. one kWh/m of irradiance should produce one kWh/

The confusion comes from the fact that the kWh are not the same: in the former case it is incident irradiance, when in the latter case this is electricity !!!

Simulation variables: meteo and irradiations

The following variables are calculated during the simulation process Meteorological data: GlobHor as read on the meteo file. DiffHor read on the meteo file
92

, and available as results:

Horizontal global irradiation, Horizontal diffuse irradiation,

Chapter 4 Project design

98

Project design
BeamHor GlobHor-DiffHor. Tamb read on meteo file, Windvel read on meteo file monthly value, or default value (1.5m/s). Incident energy in the collector plane GlobInc BeamInc DiffAInc DiffSInc AlbInc Bm/Gl BeamInc / GlobInc Diff/Gl DiffInc / GlobInc DifS/Gl DifSInc / GlobInc Alb/Gl AlbInc / GlobInc Horizontal beam irradiation, Ambient temperature, Wind velocity If not present on file :

4
=

Incident global irradiation in the collector plane Incident beam irradiation in the collector plane Incident diffuse irradiation in the collector plane (usual "diffuse" acception including albedo) Incident diffuse irradiation (from sky) in the collector plane Incident albedo irradiation in the collector plane Secondary indicators : Incident Beam/Global ratio Incident Diffuse/Global ratio Incident Sky diffuse/Global ratio Incident Albedo/Global ratio = = = =

Incident energy on collector plane, corrected for optical losses GlobHrz Global on collectors, corrected for horizon (far shadings) GlobShd Global on collectors, corrected for near shadings only (indicative) GlobIAM Global on collectors, corrected for incidence (IAM) only (indicative) GlobEff "Effective" global, corrected for IAM and shadings simultaneously DiffEff "Effective" diffuse, corrected for IAM and shadings simultaneously "Effective" = irradiation effectively reaching the PV-cell surface. Secondary optical factors FTransp GlobInc / GlobHor FHrzBm BeamHrz / BeamInc FHrzGl GlobHrz / GlobInc FShdBm BeamShd / BeamInc FShdGl GlobShd / GlobInc FIAMBm BeamIAM / BeamInc FIAMGl GlobIAM / GlobInc FIAMShd on global GlobEff / GlobInc PV array virtual productions for loss evaluations Transposition factor Horizon shading factor on beam Horizon shading factor on global Near shading factor on beam Near shadings factor on global IAM factor on beam component IAM factor on global component Combined IAM and shading factors = = = = = = = =

99

Chapter 4 Project design

Project design
EArrRef Array Reference Energy for PR evaluation. Virtual energy produced at the manufacturer specification Pnom. Equivalent to the Yr normalised value. EArrNom Array Nominal energy at STC efficiency, starting point for loss diagram Virtual energy produced at TRef (STC: 25C) according to the PV model This differs from the preceding as it is b ased on the model parameters instead of PNom. PV array losses and MPP running GIncLss STC efficiency,

PV loss due to irradiance level Difference E(Glob Eff, T=25C) b y respect to calculation at i.e. decrease of the model efficiency with decreasing

irradiance TempLss calculated at Tmodule = 25C SpctCor ModQual

PV loss due to array temperature Difference E(Glob Eff, TMod) b y respect to model Spectral correction for amorphous Calculated from the Spectral correction Quality module loss fixed constant parameter,

189

model

MisLoss Module mismatch loss fixed constant parameter for MPP or fixed V operation, depending on system, OhmLoss Ohmic wiring loss calculated at each hour with the real array current EArrMPP losses), Array virtual energy at MPP (after wiring and mismatch

Virtual calculation independent of the system running (inverter, regulator) Tarray Average module temperature during operation DTArr Temperature difference between modules and ambient' DTArrGl DTArr weighted by "effective" global' irradiation TExtON Average ambient temperature during system operation. Further simulation variables are system-dependent: -Grid connected system 100 , Stand-alone system 101 , PumpingsystemPumping, 103 DC-grid system 105 .

Simulation variables: Grid system

The following variables are calculated during the simulation process Meteo and irradiation variables
98 92

, and available as results:

: see previous page.

PV array and inverter behaviour Earray Effective energy at the array output (taking inverter behaviour into account) Iarray Array current (taking inverter behaviour into account) Uarray Array voltage (taking inverter behaviour into account) InvLoss Global inverter loss IL Oper Inverter Loss during operation (efficiency curve) IL Pmin Inverter Loss due to power threshold' IL Pmax Inverter Loss due to power overcharging IL Vmin Inverter Loss due to low voltage MPP window IL Vmax Inverter Loss due to upper voltage MPP window Syst ON System operating duration Energy output and use

Chapter 4 Project design

100

Project design
EOutInv E Load E User E Grid SolFrac Available Energy at Inverter Output Energy need of the user if load is defined Energy supplied to the user if load is defined Energy re-injected into the grid Solar fraction EUser / Eload

Efficiencies EffArrR Array Efficiency: EArray / rough area EffArrC Array Efficiency: EArray / cells area (=0 when cells area not defined) EffSyR System efficiency: EOutInv / rough area EffSyC System efficiency: EOutInv / cells area. EffInvB Inverter efficiency Threshold loss included EffInvR Inverter efficiency When operating Normalised performance index
98

Yr Reference Incident Energy in collector plane GlobInc [kWh/m/day] Ya Normalized Array Production = EArray [kWh/kWp/day] Yf Normalized System Production = EOutInv [kWh/kWp/day] Lc Normalized Array Losses Yr - Ya Ls Normalized System Losses Ya - Yf Pr Performance ratio = Yf / Yr.

= =

Simulation variables: Stand alone system

The following variables are calculated during the simulation process Meteo and irradiation variables
98 92

, and available as results:

: see previous page.

PV array behaviour
EArrMPP and mismatch losses), voltage operation EArUfix voltage), or Battery reference voltage when PV-array disconnected. EUnused Unused energy (full battery) loss (EArUFix when Charging OFF) Array virtual energy at MPP (after wiring, module quality Virtual calculation independent of the system running and Array virtual energy at fixed voltage Voltage as calculated b y the b alance loop (real b attery

MPPLoss Loss by respect to the MPP operation (when charging ON) Earray Effective energy at the output of the array (when charging ON) IArray Array Current (accumulated in Ah) UArray Array Voltage (average when Charging ON) ArrayON State / Duration of the PV production of the array

If converter present: converter losses

CL Oper Converter loss during operation (efficiency curve) CL Pmin Converter Loss due to power threshold' CL Pmax Converter Loss due to power overcharging CL Vmin Converter Loss due to low voltage MPP window CL Vmax Converter Loss due to upper voltage MPP window CnvLoss Global converter losses OutConv Energy at converter output

Battery operation: storage, losses and ageing

101

Chapter 4 Project design

Project design
EBatCh Battery Charging Energy U Batt Average battery voltage, any conditions, UBatCh Battery Voltage during charging operation IBatCh Battery Charging Current (all currents accumulated [Ah]) ChargON Charging duration EBatDis Battery Discharging Energy UbatDis Battery Voltage during discharge operation IBatDis Battery Discharging Current (all currents accumulated [Ah]) DischON Discharging duration

ESOCBal Stored energy balance (according to SOCEnd - SOCBeg) SOCmean Average State of Charge during the period SOC Beg State of Charge beginning of time interval' SOC End State of Charge at end of time interval' NB: The SOC current calculations are referred to the actual capacity of the battery, which is defined at nominal current, but varies with the discharge current level and temperature. Therefore it is not quite well determined, and not reversible (i.e. it can be different when charging and discharging). EBatLss Battery global energy loss (EBatCh - EBatDis - ESOCBal) IBEffL Battery charge/discharge current loss (coulombic efficiency [Ah]) IBGass Gassing Current loss (electrolyte dissociation [Ah]) IBSelf Battery Self-discharge Current (depends on temperature [Ah]) EBattEff Battery energy efficiency (IBEffL * U Batt) EBGass Gassing Current energy loss (IBGass * U Batt) EBSelf Battery Self-discharge Energy (IBSelf * U Batt) NB: The sum of the detailed battery losses contributions appearing on the loss diagram should in principle match this Battery Global Energy Loss calculated above, i.e: EBattLss = EBattEff + EBSelf + EBGass But during the simulation, all these contributions are determined from the Currents balance of the system (PV array - Battery - Load), multiplied by the Battery Voltage, which is varying with currents, charge/discharge state, state of charge, temperature, etc. The resulting energies are therefore defined with some uncertainties. On the other hand, as explained above, the ESOCBal is also not well determined. Therefore the overall energy balance on the battery cannot be quite rigorous. WeCycle WeState MGass Wearing due to cycling Wearing state (cycling and age) Dissociated Electrolyte Mass per cell

System operating conditions E BkUp Back-up Generator Energy (UBatt * I BkUp) I BkUp Back-up Generator Current' (accumulated in Ah) BkUp ON Back-up Generator running duration FuelBU Fuel consumption of Back-up Generator Energy use E Avail Produced (available) Solar EnergyEffectively used energy E Load Energy need of the user (Load) Defined as Input data E User Energy supplied to the user Including back-up energy E Miss Missing energy Eload - Euser) SolFrac Solar fraction (= EUser / ELoad) T LOL Duration of "Loss of Load 204 " (Duration user not supplied) Pr LOL Probability of "Loss of Load" Efficiencies EffArrR Array Efficiency: EArray / rough area EffArrC Array Efficiency: EArray / cells area (=0 when cells area not defined) EffSysR System efficiency User / rough area EffSysC System efficiencyE User / cells area (=0 when cells area not defined)

(=

Chapter 4 Project design

102

Project design
EffBatI EffBatE Battery current charge/discharge efficiency Battery energy charge/discharge efficiency
98

Normalised performance index Yr GlobInc Yu Ya Yf Pr Lu Yr - Yu Lc Yu - Ya Ls Ya - Yf Lur Lu / Yr Lcr Lc / Yr Lsr Ls / Yr

Reference Incident Energy in collector plane [kWh/m/day] Normalized Potential PV Production (battery never full) [kWh/kWp/day] Normalized Array Production = EArray [kWh/kWp/day] Normalized System Production = EAvail [kWh/kWp/day] Performance ratio = Yf / Yr. Normalized Unused energy Normalized Array Losses Normalized System Losses Unused (full battery) Loss / Inc. Energy Ratio Array Loss / Incident Energy Ratio System Loss / Incident Energy Ratio

= = = = = =

Simulation variables: pumping systems

The following variables are calculated during the simulation process available as results.
95

of the pumping system, and are

The Irradiance and PV-Array variable 98 topic describes all the preceding simulation variables, which lead to the last quantity really independent of the system running conditions, the Array virtual energy at MPP. The set of variables involved in the pumping system simulation, as well as their signification and order, is dependent on the System Configuration 74 . The energies (in blue) at different levels of the system are visualised on the Loss diagram.

Direct coupling configuration

With Direct coupling 75 configurations, including improved solutions with booster, pump cascading or Array reconfiguration, the set of main variables is defined as: EArrMpp MPPLoss EArray Virtual available energy at the maximum power point. Loss by respect to the MPP running (EArrMpp - EArray) Effective energy at the output of the array, according to the real Voltage operating point. This contribution is also accounted for when the pump is stopped due to full tank conditions, assuming normal voltage of the pump, as if it were running. Corresponding current, instantaneous [A] or cumulated [Ah]. Corresponding voltage, instantaneous or averaged [V]. Cascading or array reconfiguration configurations: Operation duration with one pump / low voltage array Operation duration with two pumps / high voltage Operation duration with all pumps Energy loss under the starting current threshold (EArray when pump not started) (only for positive displacement pumps, without booster). Energy loss under pump producing threshold (EArray when FlowR = 0) (for centrifugal pumps, which should attain a given speed before reaching the useful head). Pump overload energy (EArray in excess of the pump's maximum power) Available useful energy at pump when running (EArray - EPStart - EPmpThr - EPmpOvr) (before taking the Pump stopping due to hydraulic constraints into account).

IArray UArray 1_PmpON 2_PmpON A_PmpON EPStart EPmpThr

EPmpOvr EPmpAv

103

Chapter 4 Project design

Project design
MPPT converter configuration
EArray IArray UArray CL_Oper CL_PMax EOutConv EPmpThr EPmpAv NB: Effective energy at the output of the array (normally = EArrMpp) (may be slightly different with step-down converter 160 voltage limitations. Corresponding current at MPP, instantaneous [A] or cumulated [Ah]. Corresponding voltage at MPP, instantaneous or averaged [V]. Converter efficiency loss during operation. Converter overload loss (acc. to the specified strategy, limitation or cut). Energy at the output of the converter Energy loss under pump producing threshold (EOutConv when FlowR = 0). Available useful energy at pump when running (EOutConv - EPmpThr).
74

The converter Voltage or Power threshold losses are included in EPmpThr. These losses are accounted even when the Pump is stopped for Hydraulic reasons

Fixed Voltage DC converter configuration

EArrMpp MPPLoss EArray IArray CL_Oper CL_PMax EOutConv EPmpThr EPmpAv Virtual available energy at the maximum power point. Loss by respect to the MPP running (EArrMpp - EArray) Effective energy at the output of the array, at the fixed converter voltage. Corresponding current, Converter efficiency loss during operation. Converter overload loss (acc. to the specified strategy, limitation or cut). Energy at the output of the converter Energy loss under pump producing threshold (EOutConv when FlowR = 0). Available useful energy at pump when running (EOutConv - EPmpThr).

For all of the above configurations: hydraulic constraints

These manage the Hydraulic commands of the pump. When the pump is OFF the losses listed above remain, and the lost energy is part of the EPmpAvail 74 . ELowLev Pump stopped due to low level aspiration (deep well, drawdown safety) ETkFull Pump stopped when tank is full EPmpOp Pump real operating energy (EPmpAvail - ELowLev - ETkFull).

Battery Buffer Configuration

The Battery-buffered configuration 80 has a quite different operating mode, as the pump is connected to the battery voltage, which is quasi-constant and independent of the PV-array production. The PV-battery-load simulation process is similar to the Stand-alone strategy 94 , with the pump as load. With this configuration we have chosen to account for the "hydraulic" losses due to "Low level" (drawdown limit) and "Tank Full" upstream the battery operating losses 74 , as we consider them as electrical losses, between the "Available PV energy" (at fixed nominal voltage) and the "Unused energy" when the battery is full. Indeed, there is no loss when the pump is stopped but the battery is not full: the available PV energy is simply stored into the battery. The necessary variables involved in the simulation are the following: EArrMpp MPPLoss ELowLev ETkFull EArray IArray UArray SOC_Beg Virtual available energy at the maximum power point. Loss by respect to the MPP running (EArrMpp - EArray at Vnom) Energy lost when Pump stopped due to low level aspiration (deep well, drawdown safety) Energy lost when Pump stopped due to tank is full These two above losses are accounted only when the battery charging is OFF due to full battery. Effective energy at the output of the array (at operating voltage) Accounted only when the charging condition is ON Corresponding charging current, instantaneous [A] or cumulated [Ah]. Corresponding charging voltage, instantaneous or averaged [V]. State of Charge, beginning of interval 104

Chapter 4 Project design

Project design
SOC_End UBatt IBatCh IBatDis IBEffL IBGass IBSelf EBatCh EBatDis ESOCBal EBatLss EEffLss EPmpOp NB: State of Charge, end of interval Average battery voltage Battery charging current [A or Ah] Battery discharging current [A or Ah] Battery Charge/Discharge current efficiency loss Gassing current (electrolyte dissociation when full) Battery self-discharge current Battery charging energy Battery discharging energy Stored energy, Balance between SOCEnd and SOCBeg Battery Overall energy loss (EBatCh - EBatDis - ESOCBal) Battery efficiency loss (EBatLss - (IBGass+IBSelf) * UBatt Pump operating energy

The balances of the battery energies can never be rigorous due to the very complex behaviour of the battery. For example its effective capacity, which strongly varies with the discharge current, the temperature, etc. If the current balances are well determined in the simulation process, the corresponding energies involve the operating voltage, which is also model-dependent and varies with state of charge, charge and discharge currents, etc.

Hydraulic part, for all configurations

Remember that the Hydraulic Energy
73

is the product of the Head and Volume pumped.

The last part of the Energy Loss diagram refers to Hydraulic energy. Implicitly, when it shows pumped water volumes, this is under a given Head. Inversely, the arrows for Dynamic Head Losses express a Head loss at constant volume. E_Hydro P_Effic H_Pump H_Stat H_Loss H_DrawD FlRate WPumped WStored W_Used W_Miss Pump hydraulic energy (energy to the fluid) Global pump efficiency (E_Hydro / E_PmpOp) Average total Head at pump (During Pump_ON) Static head requirement Friction head loss Well: drawdown head loss (Only deep well systems) Average flowRate when running Water pumped volume [m] Stored water in the tank Water drawn by the user Missing water, by respect to the user's needs.

A lot of further (secondary) variables are available for results, which are not described here.

Simulation variables: DC-grid systems

The following variables are calculated during the simulation process Meteo and irradiation variables PV array behaviour EArrMPP and mismatch losses), EArUfix voltage), or line reference voltage when PV-array disconnected. MPPLoss Loss by respect to the MPP operation EUnused Unused energy loss (over the user's demand) Earray Effective energy at the output of the array IArray Array Current Array virtual energy at MPP (after wiring, module quality Virtual calculation independent of the system state Array virtual energy at fixed voltage Voltage as calculated b y the b alance loop (real load
98 92

, and available as results:

: see previous page.

105

Chapter 4 Project design

Project design
UArray Array Voltage ArrayON State / Duration of the PV production of the array System operating conditions Esupply Energy from Conventional Power Supply LigLoss PV to User Line Ohmic Losses Energy use E Avail Produced (available) Solar Energy E Load Energy need of the user (acc. to Load definition) E User Energy supplied to the user SolFrac Solar fraction (EUser / ELoad) U User User Voltage Efficiencies EffArrR Array Efficiency: EArray / rough area EffArrC Array Efficiency: EArray / cells area (=0 when cells area not defined) EffSysR System efficiency User / rough area EffSysC System efficiencyE User / cells area (=0 when cells area not defined) Normalised performance index Yr GlobInc Yu Ya Yf Pr Lu Yr - Yu Lc Yu - Ya Ls Ya - Yf
98

Reference Incident Energy in collector plane [kWh/m/day] Normalized Potential PV Production [kWh/kWp/day] Normalized Array Production = EArray [kWh/kWp/day] Normalized System Production = EAvail [kWh/kWp/day] Performance ratio = Yf / Yr. Normalized Unused energy Normalized Array Losses Normalized System Losses

= = =

Economic evaluation
After simulation, an economic evaluation of the system may be performed on the basis of the defined parameters and the simulation results. The special economic tool is accessible in the "Results" dialog. Costs can be defined globally, by pieces, by installed Wc or by m. You can work with any currency, and pass from one to another by a selection in the listbox. The button "Rates" allows to adjust their relative parity. Choosing one of them as a reference, you can modify the exchange rates or add new currencies. The first data group determines the investment: The number and type of PV involved components (PV modules, inverters, batteries, etc.) are automatically updated from the simulation parameters. Prices can be defined for each component. When clicking the corresponding "open" button, you may define your own prices for the components used, either for one piece, or discount price for several pieces. These prices may be either saved in your component library, or just kept for the current session, without modifying the component database. After defining your component prices, you can get the calculated price in the economic evaluation sheet (check the "default" case), or modify it explicitely. For the module supports, the price "by piece" is related to one PV module. "Miscellaneous" allows you to define up to 5 additional labels of your own, which will appear on the printed output (for example "Transport" or "Engineering fees"). An eventual underworth corresponding to the substitution of an allowance (for example facade element

Chapter 4 Project design

106

Project design
replaced by PV modules) comes in deduction from the rough investment.

The net investment - for the owner - is derived from the gross investment by substracting eventual subsidies and adding a tax percentage (VAT). Choosing a loan duration and interest rate, the program computes the annual financial cost, supposing a loan pay back as constant annuities. The loan duration should correspond to the expected lifetime of the system. This procedure is justified by the fact that, as a contrary to a usual energetic installation, when purchasing a solar equipment the customer buys at a time the value of the whole energy consumed during the exploitation. The running costs depend on the system type. For a grid-connected system, usually very reliable, they are limited to an annual inspection, eventually some cleaning of the collectors and the insurance fees. Some Inverter suppliers provide a long-term payable warranty, including replacement, which are assimilable to an insurance. For a stand-alone system, one should add a provision for the maintenance and periodical replacement of the batteries. This last contribution is calculated by the program as a function of the expected lifetime of the battery pack, calculated by the simulation. Moreover, when using an auxiliary generator the program computes the used fuel cost. For pumping systems, there is also a provision for pump replacement, their lifetime being usually of the order of a few years. And of course for the batteries when used in the system.

The total annual cost is the sum of the annuities and the running costs. Divided by the effectively produced and used energy, it gives an evaluation of the energy cost (price of the used kWh). This is of course depending on the above hypothesis, especially the loan duration and rate, as well as the effective system production. For grid-connected systems, the long term profitability 107 may be estimated according to different consumption or feed-in tariffs conditions ("Financial Balance" button)

Long term financial balance

In the Economic Evaluation part 106 , the investment and annual costs (including loan refund) are evaluated according to the PV system financing. Now for a grid-connected system, the produced electricity may be either consumed by the owner, or (more likely) sold to the grid utility. In both cases the produced electricity has a financial value to be compared to the annual costs in order to evaluate the system profitability. This part performs an annual balance between costs and revenues, according to several possible sale dispositions. Simple feed-in tariff A now common procedure, especially in several European countries, is the purchase, by the grid utility, of the total energy produced. The feed-in tariff is set by a long-term contract (usually 20 years), at a level determined at the system commissioning time, and fixed for the whole contract period. Therefore the first (and simplest) option in PVsyst is to define a single feed-in tariff over a given contract period. It is also possible to define a connexion annual tax, as well as a progressive decrease of the system production (and also of the feeding tariff itself if necessary). Finally the residual feed-in tariff at the end of the contract period should be defined. As usually the contract period will correspond to the loan refund period, the annual balance after this period (even with lower tariff) will dramatically increase until the end of life of the system! Modulated feed-in tariffs In other cases (for example in the USA) the tariff is dependent on the hour of the day, or even on the season. PVsyst allows to define "Night" and "Day" - and even "Peak"- tariff levels, for specified hours of the day. These tariffs may be different in summer and winter (with specified months). In this case the final tariffs after the preferential contract period are all decreased by a same factor.
107

Chapter 4 Project design

Project design
Net-metering

The Net-Metering concept is defined here with a consumption tariff when the owner is able to use the produced energy for himself (economy on the electricity bill), and another tariff for selling the overproduction to the utility. This calculation requires of course that the user's needs are specified, and computed during the simulation, and that these data are stored in hourly values. These tariffs may also be modulated according to the day hour. The consumption tariff is likely to be increased during the next years; therefore an annual increase rate can be defined. PVsyst calculations Using the effective production given by the simulation, PVsyst shows the annual balance as well as the accumulated balance over the foreseen lifetime of the PV system, according to all these strategies. It also shows details on an annual table. But be careful: these balances result of differences of large quantities, and little perturbations either on the real production or the effective costs may result in dramatic deviations of the final profitability ! This is namely the case for the annual real irradiance variations, by respect to the meteo data used in the simulation. Failures of the system all over its lifetime may also significantly affect the effective balance. NB: In this part, when annual evolutions are specified in %/year, these are % of the initial value, not the preceding one. PVSYST clearly distinguishes between two kinds of Meteorological data: Monthly meteo data (*.SIT files), which are associated with the definition of every geographical site, Hourly meteo data (*.MET files), which can be DRY 116 in the database, imported 120 from known sources or custom ASCII files, or synthetic data generated 181 from monthly meteo data.
113

See also: Meteo Database Monthly meteo

, and Meteorological data sources

120

The geographic site 114 definitions (*.SIT files) holds the site and country names, the geographical coordinates, and should include monthly meteorologigal data , at least Horizontal global irradiation and ambient temperature (12 monthly values each). They can also hold diffuse irradiation and average wind velocity when available. These monthly data are easily introduced or modified manually, and there are tools for importing them from several sources. Hourly meteo The hourly meteo 115 files (.MET files) include an associated geographical site as defined above. They hold Horizontal global and ambient temperature , in hourly values. Diffuse irradiance and Wind velocity are optional. When not available in the original data: - the diffuse irradiance is calculated during the building of the internal file. When another irradiance (beam horizontal or normal beam) is available in the original data, this is of course used. Otherwise the program uses a model (Liu and Jordan correlation) for diffuse estimation in hourly values. - the wind velocity - only used for the calculation of the PV modules temperature 84 during operation, and therefore of weak importance in PVSYST processes - is taken as a default value, or from the associated site monthly data if specified. Due to the poor reliability of primary data, the use of wind velocity is not recommended for the array thermal loss evaluation. When not available for a given site, an Hourly Meteo Data set can be generated values obtained by a random process) from any monthly data "site". NB: By convention in PVsyst, generic years Chapter 4 Project design
202 181

(with synthetic hourly

like synthetic hourly data are always labelled as 1990.

108

Geographical and Meteorological data

Meteorological data - Tutorial
The meteorological data are the starting point of the evaluation of a project. But also the main source of uncertainty 122 .

PVsyst holds a limited meteo internal database of about 300 sites in the world, but gives an easy access to several public sources directly available from the Web. Data of other providers like national meteorological offices may also be imported using a flexible importing tool. This tutorial will guide you in different options, for discovering the management and organization of meteorological data in PVsyst, and the import from external sources. All meteo manipulations and visualizations are performed in "Tools".

Geographical coordinates and monthly meteo data : *.SIT files

The main database is given as objects including the geographical coordinates and associated monthly meteo data. These objects are stored as files with the name *.SIT, situated in the directory \Data\Sites\. Please open "Geographic sites". You can choose the country or region of interest and a given station. The definitions of a geographic site are gathered on two pages:

109

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

The "Geographic coordinates" page defines: - The name of the site, country and region, - The Latitude, Longitude, Altitude (which uniquely define the (x,y,z) coordinates of a given point of the earth), and the time zone. Ex: for central Europe, the Winter time corresponds to GMT+1, while the summer time is GMT+2. You can obtain accurate Latitude/Longitude coordinates from your GPS or GoogleEarth. In this dialog you can also: - see the sun paths corresponding to your site, - import/export the site data by "copy/paste" (for example in a spreadsheet like EXCEL), - print a complete form with the data of this site.

Chapter 5 Geographical and Meteorological data

110

Geographical and Meteorological data

The "Montly meteo" page holds the monthly meteorological data. Values for Global irradiance and Temperature are mandatory (for use by the simulation). Global diffuse and wind velocity are optional. They will be evaluated by models when necessary. - You can choose the units for the edition (you will find many kinds of units in other data sources). - If you define a new site (by its geographical coordinates), you can always obtain default Global horizontal and Temperature from the Nasa-SSE data, which are available for any point on the earth, by steps of 1x1 lat x long. - If data are given as lines or columns in a spreadsheet, you can "paste" entire columns at a time. - As for any database element you are advised to always mention the source of your data. After defining or modifying a site, the program will ask whether you want to keep your modifications, and if so it will modify or create a new site in the database (i.e. a new file in the \Sites\ directory).

PVsyst "native" database

The main PVsyst meteo database, including around 300 sites in monthly values, is based on the METEONORM 126 database. This software defines about 1'500 "Stations" - for which measured irradiances values are available. The Meteonorm data are usually 1961-1990 average values (the new Meteonorm V6.0 includes also some 1983-1993 averages, but usually not very different). In Meteonorm, data of all other sites are interpolated values between the 2-3 nearest stations. For all big European countries, all the measured stations available in Meteonorm are in the PVsyst database. But for many regions of the world, the measured "Stations" are very scarce, and Meteonorm has to use Satellite data for completing this information. But besides the original database, PVsyst offers tools for easily importing meteo data sources. See later in this tutorial.
120

from many other

NB: The original database, is stored in the file "MeteoDB.CSV", located in \DataRO\, which may be opened in a spreadsheet (for example in EXCEL). In this file each site is represented as a line. The year 1990 In PVsyst, we have adopted the convention to label all data which don't correspond to really measured data at a given time as 1990. This is the case, namely, of all Synthetic hourly data (see later) , or TMY/DRY datafiles. You will observe that recent data are really scarce in the "public" published data. This can be explained by the fact that analysing and assessing real measured data sets for sizing purpose is a very complex job, and the results of 111

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

these research groups are only available later on.

Generation of synthetic hourly values - *.MET files

Now the simulation process operates on hourly values. Therefore if we don't avail of really measured hourly data, we have to construct a set of hourly meteo data from these monthly values. The hourly data will be stored in *.MET files, and stored in the \Data\Meteo\ directory. For the irradiance, the synthetic generation 181 of hourly values from monthly averages is performed by using stochastic models, due to the Collares-Pereira team in the 1980's. This model generates a sequence of days, and then a sequence of hours in the day, using Markov transition matrices. These matrices have been established in order to produce an hourly sequence, with distributions and statistical properties analogous to real hourly meteo data measured on more than 30 sites all around the world. For the temperatures, we don't avail of a model predicting the temperature evolution by respect to the daily irradiation, as the temperatures are mostly governed by atmospheric circulations. Therefore the sequence of daily temperatures is mostly random (with constraints on the transition from one day to another one of course). Nevertheless within a day, the temperature profile is well correlated to the irradiance: this results in a sinusoidallike shape, with an amplitude proportional to the daily irradiation, and a delay of around 3 hours (the warmest hours are at around 3h - solar time). This involves that the average temperatures waited by the program are indeed the 24h-average measurements over the month (usual meteo data). Diurnal temperatures are not suited as the job is done by the model. Please note that this generation is a fully random process: two successive generations performed from the same monthly data will result in completely different years. When performing simulations of grid-connected systems, this may produce variations of 0.5 to 1% in the yearly result. For obtaining the synthetic hourly data from your monthly definitions: - Press the second button "Synthetic Hourly Data", - Choose the desired site holding monthly meteo data, - Press the "Execute Generation" button. There are some options that you will in principle never modify: - "Monthly Renormalization": the sum of the generated hourly data doesn't match the required monthly sum. This could be used for annual variations analysis, but in PVsyst we usually renormalize the data for obtaining the original monthly sums. - "Use monthly diffuse": in the same way the diffuse part is evaluated using the Liu-Jordan hourly correlation. At the end of each month, the diffuse values are renormalized in order to match the specified monthly diffuse. NB: This was not the case in the version 4. In middle Europe the Liu-Jordan correlation tends to over-estimate the diffuse. Now when the monthly diffuse is specified (usually lower), this increases the beam component and therefore the transposition results. - The temperature model was established on Swiss meteo data, with a detailed analysis of climate type dependence. But the differences between all these options are very low (slight coupling differences between irradiance and amplitude, or inertial shifts).

Visualization of the hourly values - *.MET files

In PVsyst, the hourly values files *.MET are binary files, which cannot be edited in a text editor. For visualizing their contents, you have to use the button "Meteo Tables and Graphs" in the "Tools" group. After choosing your meteo file, you have the opportunity of visualizing them as "Graphs" and "Tab les". Both may be in Hourly, Daily or monthly values.

Graphs of hourly values

In the "Graph" tab, please choose "Hourly", "Glob al" and "Diffuse" (default when entering), and click the button "Graph". This will open a graph with the hourly meteo values, and you can walk through your entire data using the Scrollbar on the right. The plot includes a blue line, which represents the Clear day model, superimposed on your data. Please observe a good day: it is very important that the data should be more or less centred on the blue line. This will always be the case for the synthetic data or the data imported from known sources using the "Import meteo data" PVsyst tool.

Chapter 5 Geographical and Meteorological data

112

Geographical and Meteorological data

But this may be different with personal imported data, when using the "Import ASCII meteo files" tool. If the data are not matching the Clear day model (i.e. are shifted toward morning or evening), this indicates that the hour stamps of your data don't match the PVsyst standard, and all the models using solar geometry will not work properly. When walking through the year, you will observe that clear conditions correspond to low diffuse component, and when the sun becomes hazy, the diffuse part increases. The difference between the global and the diffuse corresponds to the beam component.

Graphs of daily values

Have a look on the daily values. Each point represents the daily irradiance of one day, in [kWh/m/day], along with the Clear day model as an envelope curve. This plot gives a first check of the quality of the data. Now if some data significantly overcome the clear day model (say, by more than 3-5%), this indicates that the data are not good.

Tables of Hourly, Daily or Monthly values

Perhaps you want also show your data as tables. You can choose up to 8 values to be tabulated at a time, including the irradiance on a tilted plane (transposition model) or the normal beam component (for concentration) . As for each data table in PVsyst, you have the opportunity of: - Printing the table: this printed form usually includes the main parameters concerning the data, and for big tables (hourly or daily) you can choose the desired period. - Export / Copy as text: this will "copy" the full table in the clipboard, so that you can directly "Paste" it into an external spreadsheet (like msExcel). Please remember: in msEXCEL, the imported data will usually be gathered in one only column. For expanding the data to the cells you have to use the standard EXCEL options for importing data: menu "Data" / "Convert", and here you should choose "Delimited" / "Semicolon" separator. NB: The data will be copied with a decimal point. If you are using decimal commas (international preferences in Windows), you will perhaps have to change all points into commas. - Export / Copy as image: to be used for directly put the table "as such" in a report. - Export / Copy to file: will create a CSV file, which you can open in any spreadsheet program.

Importing Meteo data from many popular sources

Please choose the "Import Meteo Data" button. Here you have documented and easy-to-use tools for importing meteo data from many sources. Press F1 for obtaining a detailed description of all available meteo data. After choosing a given source, pressing F1 will open the detailed procedure for importing the concerned data. Please carefully follow it in detail, and pay attention on the advices written in red at the top of the screen during your progress. These data are not always completely comparable. The Help includes a comparison from the north to the south of Europe.
122

of these data for 12 sites

Meteo database in PVsyst

Monthly meteo (Geographical site) database The PVSYST library includes the Geographic Site's definitions (latitude, longitude, altitude and time zone 206 ), as well as monthly data of the global irradiation, temperatures and wind velocity for more than 330 sites over the world. These data are issued from the database of the software Meteonorm 126 (versions 4/5), which summarises well-established meteorological data of about 7'700 sites in the world. Nevertheless only about 1'200 of these Meteonorm sites avail of irradiance measurements. These are referenced as "Stations" in Meteonorm, and hold data as monthly averages over years 1960 to 1991. Meteonorm can also get data for any unreferenced place by performing interpolations (between 2-3 nearer

113

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

"stations"), taking altitude and region typology into account.

For the main European countries, the PVsyst database includes all the available "stations" with irradiance measurements. This represents about 10-15 "Stations" in each big country. Data of any other site issued from Meteonorm would be obtained by interpolation. You can use "Tools" / "Import Meteo data" for importing sources in the database. Hourly Meteo Database Detailed simulation requires meteorological data (Global horizontal and Temperature) in hourly values. Synthetic hourly data 181 files may be generated from any monthly meteo described above. This generation is automatically performed when choosing a site during the Project definition. The only hourly data originally included in the PVsyst package are meteorological hourly DRY (Design Reference Years) of 22 stations in Switzerland. These data proceed from 10 years of measurements in the frame of the ANETZ meteorological stations grid of the ISM 125 (Swiss Institute for Meteorology). Hourly data may be imported
120 120

data from Meteonorm and many other

from several sources ( Satellight, TMY2, Helioclim, SolarGIS, 3Tiers).

Geographical locations / monthly meteo data

A geographical site is defined by: it's name , country, and world region, its geographical co-ordinates: latitude, longitude, altitude and time-zone Monthly meteorological data. the monthly Global horizontal irradiation Monthly averages of the ambient temperature . monthly Diffuse horizontal irradiation monthly average of wind velocity.
113 206

To be used in the simulation, the minimum meteorological data should include:

The file may also include the optional data:

PVsyst includes a database from many sources.

of around 330 sites in the world, and you can import meteo data

120

Meteo values are displayed and easily defined on the screen. The units for capture and display are left to the choice of the user. It is to be noted that for the verification of rather "uncertain" data, the clearness index 201 Kt is also displayable, which is the irradiation actually received on earth, normalised to extraterrestrial irradiation in monthly values. The monthly average of Kt should usually lie between about Kt = 0.25 and Kt = 0.75 at any place. Limits for data acquisition are defined in "Preferences"/"Hidden parameters". Monthly meteo values can be used as a basis for the generation of synthetic hourly data
181

The specification of sites will be used in the management of hourly meteo files, project pre-design, the localisation of system design projects, the display of solar parameters, and many other tools.

Horizon
It is to be noted that in PVSYST, the basic irradiation values are usually defined for a free horizon. Taking the shading of the horizon 38 into account is worked out during the simulation, by cancelling the beam component when the sun is hidden below the horizon. If one is to use specific data recorded in the presence of a horizon, the correction is obviously not to be applied.

Importing monthly meteo values

In the "Site" definition dialog: First define the main parameters of your site:

Chapter 5 Geographical and Meteorological data

114

Geographical and Meteorological data

Name, Country, Region Latitude, longitude, Altitude, Timezone Then pass to the "Monthly meteo" tab: If you don't avail of monthly meteo values, please use the button "Default (from NASA-SSE)", Otherwise:
206

- Specify the origin of your data (source, year, etc). - Choose the units of your source list - Type in the monthly values in these units Or alternatively, if you have a list in another software, you can import monthly data from an external list of 12 values with any separator (for example from an EXCEL column or line). - Go to the external software and "Copy" (Ctl-C) your list of 12 values (Irradiations, Temperatures or Wind velocities) - Paste them in the corresponding column. Finally save your new site with a significant file name.

Hourly meteorological data

As meteo hourly basis, the programme uses binary files with a special format for PVSYST, and characterised by the extension ".MET". The meteo file includes a complete geographical site 114 object, followed by hourly values of the meteo parameter, that is horizontal global and diffuse irradiations, ambient temperature, and, if available, wind velocity. The data are recorded for whole days (0H .. 23H). Step labels are referred to the beginning of the interval (i.e. the 12 h label corresponds to the 12-13 h interval). A complete file for one year occupies about 70 kOctets (53 kOctets without wind-velocity). When importing own measurements, the meteo file can be restricted to limited periods, and even have full-days holes. In this case the date is included in each record on the file. The meteo files may be obtained in several ways: 22 DRY (Design Reference Years) for swiss locations are available in the database. ASCII data files 137 (official meteorological data like DRY or TMY, or measurements carried out directly by the user) can be imported in a great variety of formats, using a special programmable data interpreter. NB. In order to facilitate the analysis of existing systems, the programme also accepts irradiation measurements carried out in the collector plane. Generation of synthetic hourly values from monthly meteo values of global irradiation and temperature. The physical bases of this generator follow the directives of Meteonorm '95, using Aguiar/Collares-Pereira et al, and Scartezzini algorithms. Direct import from the Meteonorm 126 software, or US TMY data 128 , free from the web.

The characteristics and the contents of the hourly meteo files can be visualised and analysed thanks to the meteo tables and graphs 116 tool.

Meteorological data averaging

Meteorological data are often available for several individual years, and aggregating them for obtaining an average situation is not straightforward. Only Monthly meteo values may be averaged, as the seasonal distribution is not very different from one year to another one. Hourly or Daily values cannot be averaged. It would not make any sense to construct an average year by aggregating each day of different years, mixing sunny and cloudy days. This would result is a time series with only mean days, without clear nor bad days, which has no physical meaning. The right way for obtaining average conditions for hourly data is the construction of DRY obey statistical constraints and is a matter of specialists.
115
116

, which should

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

Therefore for obtaining the average behaviour of a complete PV system from several real years, you have two solutions: either you accumulate your meteo (hourly or daily) data in monthly sums, average the monthly values and generate a new Synthetic hourly file 181 . or you perform several one-year simulations with your different real meteo hourly files, and then average the results of the simulations.

DRY or TMY hourly data

Design Reference Years (DRY) or Typical Meteorological Years (TMY) are meteorological data files constructed on the basis of real measured data series. The elaboration of DRY or TMY obeys a sophisticated normalised procedure. It usually consists of a juxtaposition of selected months, chosen among 10 years or more of real measurements, according to several statistical criteria, and approaching average values. The month's extremities are eventually corrected for harmonious continuations, and these data sets should also include realistic extreme behaviours. They are constructed so that the whole year reproduces a typical meteorological situation for the design of energetic or architectural systems. The 22 ISM 125 (Swiss Institute for Meteorology) files of the PVsyst database were constructed from the ANETZ measuring network, according to a procedure elaborated by the International Energy Agency (IEA), task 9, sub-task E. For other European locations, DRY may sometimes be obtained at the addresses given below. These data are not necessarily prepared according to the same standards: some of them are referenced as "Test Reference Years" (TRY), and follow a similar method approved by the European Communities. German data are specified by the Free University of Berlin. The 1035 TMY3 files for the USA were prepared by the NREL. Here is a list of institutions where hourly meteo data may eventually be available (state 1994 - probably many of them obsolete !): Belgium: Denmark: France: Germany: Ireland: Italy: Portugal: Slovakia: Switzerland: Turkey: Institut Royal Mtorologique Cenergia ApS Direction de la Mtorologie Meteorological Service KNMI 3, av. Circulaire B-1180 Bruxelles 2, avenue Rapp F-75340 Paris Cedex 07 IRL - Dublin 9 I-00198 Roma.

Sct. Jacobsvej 4DK-2750 Ballerup

Berlin University University of Berlin Glasnevin Hill National Research Council P.O.Box 201 INMG Rua C - Aeroporta Via Nizza 123 P - 1700 Lisboa SL-813 68 Bratslava Ankara CH - 8600 Dbendorf Eastern Road, Bracknell Berkshire RG12 2SZ,

The Netherlands:

NL-3730 AE De Bilt

Katedre KPS

Stavebena Technika Universita

EMPA Ueberlandstrasse 129

Tbitak, Building Research Inst. 221, Atatrk Bulvari Meteorological Office, MET.0.1

United Kingdom: UK USA :

http://rredc.nrel.gov/solar/old_data/nsrdb/tmy2/ 1035 TMY data which can be directly imported

128

in PVSYST.

Meteo tables and graphs

See also generalities about meteo data
120

The meteo tables and Graphs tool displays the characteristics and contents of the meteorological hourly Chapter 5 Geographical and Meteorological data
116

Geographical and Meteorological data

files.

The geographical characteristics are stored in an object identical to the geographical site 114 definition, which may be edited and modified. Other characteristics are related to the data source (dates, recorded parameter, etc). The meteo file, written in a binary format, cannot be read using a text editor. This tool offers a powerful mean of visualising and analysing meteorological data in graphical or tabular form. The tool displays as well: - The basic data of the file: Global and Diffuse horizontal irradiance, ambient temperature and wind velocity; - Derived values like beam component, clearness index Kt 201 , or global irradiance on a tilted plane. All these data are available in hourly, daily or monthly values. Graphs can be obtained as time evolution (selected periods), values distribution histograms or sorted values (cumulated histograms). To be noted: a "clear day" model curve may be superimposed on the daily curves, allowing for checking the data timing quality 117 . Tables display hourly, daily or monthly values in a scrolling mode. Printings or data export (using clipboard as well as ASCII file) can be easily performed for any time period, allowing the use these data in other software.

Hourly meteo data quality check

The quality of the meteo data records is not always perfect, and may show several distortions due to sensor accuracy, data recording or processing, etc. In particular when importing custom data, you are advised to carefully check the resulting imported data . The main characteristics to be checked are: - The time stamps of each record. - The absolute values (sensor calibration, units, etc), - The aberrant values (especially by respect to the clear day model) PVsyst offers several tools for analysing this quality, located in the meteo hourly data visualizing tool "Meteo Tables and Graphs".

Graphs in hourly values

This is the first visual rough test to be performed. When comparing the measured values with the clear day model drawn in blue (especially by choosing the best days), you will immediately identify the main possible errors: - Time shifts of more than half-an-hour, - Irradiance amplitude for best clear days, - Aberrant values like night significant values or 'holes' in the data. This test may be completed by a look on the graph in daily values, for which the data of best days should be close to the clear day model, and never overcome it by more than, say, 5% to 10%.

Check Data Quality tab

This option directly shows a graph of the time shifts of the best clear day for each month. This allows to identify: - a systematic time shift in the imported data. Tthis may be corrected in the meteo file (you should define a Time Shift 119 correction). - a variation of time shifts along the year, which may be due to: - summer/winter time (or daylight saving time, DST) in the original data; the importing tool allows for taking the summer/winter hour changes into account, but the resulting meteo file will always be in

117

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

winter time. PVsyst doesn't support DST. - records defined in Solar time (very rare option, sometimes used for research time series), - a misrunning of the clock of your recording equipment, leading to random time variations along the year; In this case PVSYST cannot help, the time Shift correction is only defined as constant over the year. There are 3 further useful graphs: Hourly Kt morning/evening, which represents the hourly measurements of morning and evening Kt values in different colors. - This is a very sensitive test for the real time shift of your data: a scrollbar allows to modify the time shift value, resulting in very dissymmetric distributions between morning and afternoon points. The best timeshift is when the morning and afternoon top points are well balanced. - This also shows how the absolute values of Kt match the Kt corresponding to the Clear day model, as a function of the sun's heigth.

Monthly best clear days shows the detail of each monthly best day data by respect to the the clear day model, sometimes explaining why the value in the first plot is not well aligned. Best clear days Ktcc displays the sorted Ktcc of all days of the year. The Ktcc is the clearness index referenced to the clear day model (not to the extraterrestrial). This graph gives an idea of the calibration of the irradiance sensor: the best days of the data should be close (within 5%) of the clear sky model, i.e. Ktcc=1.

Chapter 5 Geographical and Meteorological data

118

Geographical and Meteorological data

Time shift
In the PVsyst convention, the time label accumulation up to the next record.
141

always refers to the beginning of the record, and concerns the

Now in measured data files, each record usually holds a time label. Depending on authors or data acquisition systems, this time may be referenced either to the interval beginning, or the interval end, or to the interval middle point. Or even at any other shifted time within the hour. For example, the data recorded for the time label 12:00 to 13:00 may be measurement accumulations from 12:20 to 13:20. When using these shifted data, PVsyst will keep its full-hour labels when displaying most of the results.

119

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

But all the calculations related to the solar geometry will be erroneous, as the solar geometry is normally evaluated for the middle of the interval (12:30), when it should correspond to the middle of the irradiance measurement period. Time Shift correction This is the reason why PVsyst defines a Time Shift correction (expressed in minutes) which may be associated with the meteo data file. This correction will be applied to the middle value of the hour for the calculation of the solar geometry (in our example, the solar geometry will be computed at 12:50). This correction concerns meteo calculations like diffuse model, transposition model, shading computations. This is particularly important in the mornings or the evenings. PVsyst offers several tools for the evaluation of the time shift correction clear sky model, or using the Hourly Kt morning/evening graph). Time shift correction when importing data Now the import of meteo values itself usually involves meteo models. Especially when importing GlobInc (POA) values, the retro-transposition involves a division by sin(HSol), which may be very low in the morning or evening. A time shift error in the solar geometry will result in very high errors for the GlobHor value (may attain several kW/m), which is of course quite unphysical, or even lead to POA significant values for HSol = 0. The diffuse evaluation is also depending on this time-shift, and this may lead to assymetric diffuse for clear days. Therefore when importing an unknown data set (especially with diffuse calculation or GlobInc/POA): - First import without precautions. The program will guide you to the graphs for the assertion of the correct time shift. - Evaluate the effective time shift of your data. If more than +/-30 min, you should modify the begin/end interval choice, or the time-zone 206 of your reference site. - Apply the observed timeshift in the format protocol, and re-import the data. NB: In PVsyst, the time defined around sunrise and sunset for solar geometry is the middle of the interval when the sun is over the horizon. NB: When importing data (especially from POA), PVsyst will limit the GlobHor to a reasonable value according to the clear sky model. This is the main cause of the discrepancy often observed between the POA original values and the transposed values.
117

(either by comparison to the

Import from Meteorological data sources

Besides the Meteo Database 113 included in PVsyst, There are many meteorological data sources available from the Web or by other means. PVsyst now includes a tool for easily importing the most popular ones. These are summarised in the table below, and we have performed a comparison 122 between their results. Sources of Meteo data, in hourly values Meteonorm 126 hourly values are not measured, but synthetic data constructed in the same way as the synthetic hourly values in PVsyst from monthly values. Constructing hourly data in Meteonorm instead of PVsyst may give slightly more realistic results as the models (especially for temperature) have been slightly improved, and wind velocities are also generated (with which reliability?) S 129 atelligh 129 t 129 Data provides real measured data in hourly values, over 5 years (1996-2000), for any pixel of 5x7 km in Europe. US TMY2/3data 128 (Typical Meteorological Year) are compilations of measured hourly data chosen among 1961-1990 (TMY2) or 1991-2005 (TMY3), for 1020 locations in the US. Canadian EPW data 129 (Typical Meteorological Year) are compilations of measured hourly data chosen among 1953-1995, for 72 locations in Canada.

Chapter 5 Geographical and Meteorological data

120

Geographical and Meteorological data

Helioclim-3 131 provides data in hourly values, measured by Meteosat, since February 2004. But these data are not free. Nevertheless the year 2005 is available for tests. SolarGIS 3Tiers
133 133

provides hourly data measured by satellites, recent, for any location on the earth. For pay.

also provides hourly data measured by satellites, recent, for any location on the earth. For pay.

... and in monthly values Meteonorm 126 monthly irradiance data are available for about 1'200 "stations", as averages of 1960-1991 (and also 1981- 2000 in version 6.0). All "stations" (i.e. with irradiance measurements) of the main European countries are referenced in the PVsyst database. Data for any other site may be obtained by interpolation (usually between the 3 nearest "stations"). WRDC 134 Data (W orld Radiation Data Center) provides monthly irradiance for 1195 sites in the world, averaged during periods between 1964 and 1993. Many of them are only over a few years. These data don't include temperatures, which should be obtained from another mean. NASA-SSE 135 Data (Surface Meteorological and Solar Energy Programme) hold satellite monthly data for a grid of 1x1 (111 km) covering the whole world, for a 10 years period (1983-1993). PVGIS-ESRA 127 Data give monthly values interpolated for any geographical location from average of 19811990 terrestrial measurements for Europe, and satellite 1985-2004 data (Meteosat) for Africa. Helioclim-1 131 are satellite data from Meteosat, given for each year 1985-2005 independently for Europe and Africa. The public access - for tests - is restricted to the 1985-1989 data. Retscreen 136 is a Canadian software which holds a complete database for any location in the world, optimised for using the best available data at each location from about 20 sources, the main ones being the WRDC and the NASA irradiance data. Temperatures and wind velocities are also provided probably with good reliability. NB: All these monthly data are imported as geographic site 114 files, and will require the construction of Synthetic Generated hourly data files for being used in the simulation. NB: Monthly values are often given as averages over several years. But due to the variability, it doesn't make sense to average meteorological data in Daily or Hourly values. Such data are always issued from real measurements at a given site (or eventually from satellite photographies). In the same way, it is also not reliable to interpolate hourly values between different sites.

121

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

Meteorological data comparisons

PVsyst now gives access to many popular meteorological data sources
120

These show that the available meteorological data are far from being an exact science! There are big discrepancies between these databases, and it is very difficult to estimate which one is the best suited for a given project or location, and what is the probable error. Chapter 5 Geographical and Meteorological data
122

Geographical and Meteorological data

We have performed a comparison between these sources, for several locations from north to south of Europe. Comparisons cannot be made rigorously, because of the variety of conditions: - All sources are not available everywhere. Some of them are for given locations, other ones perform interpolations or are for discrete grids of variable sizes. - Climate variability: the sources apply for measurements of given years, or averaged periods which differ from one source to another one (or even for one location to another one, depending on historical measurements availability). - Measurements: ground stations or satellite image data interpretations (involving sophisticated models). - Available parameter: many sources dont provide temperature measurements (or not reliable).

Comparison criterion
For the comparison, we have chosen as reference the annual available irradiation [kWh/m/year]. This parameter is relevant for PV grid systems, as the PV output is quasi-linear with the solar energy input. For other systems like stand-alone, the monthly distribution may also be of interest, but comparisons would require much more complex statistical methods. We dont show temperature results, which are of lower importance in PV systems. For most comparisons we refer to the Meteonorm data, which are the default data in the PVsyst database, and therefore likely to be used in any first simulation of a given system.

Global comparison between all sources

The next figure shows a comparison between all available sources, for 12 locations in Europe. All values are by respect to the Meteonorm values (the PVsyst internal database). Except the first green bin, which shows the difference of Meteonorm by respect to the average of all other sources (excluding Helioclim 2005). We can observe some trends: - We cannot say which source is the more representative of the reality (and which reality? - no one is of course able to foresee the future climate). - Meteonorm often gives lower values than the average. This means that simulations with default values in PVsyst will be rather conservative, and give prudent results for the final yield of the customers systems. - Except at Barcelona, very little differences between Meteonorm old (1961-1990) and new (1981-2000) data. - PVGIS seems to be one of the more compatible with Meteonorm data. - Also, when available, the WRDC data are well representative of the Meteonorm data. - Satellight data seem to overestimate the average by 5 to 10% (except in Berlin and Roma). - The Helioclim data response seems to be chaotic. Especially the 2005 values (Helioclim-3 hourly file) are very high over the other data, by a factor which is not always compatible with the 2005 irradiance (see below for Geneva).

123

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

Yearly GlobH Differences
20%

5
MN 6 / MN RetScreen / MN Helioclim 85-89 / MN WRDC / MN Satellight / mn Helioclim 2005 / MN

Meteonorm / Average NASA-SSE / MN PVGIS / MN

15%

10%

5%

0%

-5%

-10%

-15% Ham burg Berlin Caen Paris Geneva Lyon Toulouse Marseille Rom a Barcelona Madrid Sevilla

Climate evolution
We avail of a homogeneous sample of continuous measurements from the same source (ISM - Swiss Institute for Meteorology) for Geneva, from 1981 to 2007. This shows that at Geneva, the annual dispersion stayed far below 5% with only 3-4 exceptions during 20 years, but increased significantly since 2003. This is of course not necessary valid for other sites in Europe!
Geneva ISM measurements meteo data evolution
15%

By respect to Meteonorm 81-90 (1212 kWh/m2)

10%

5%

0%

-5%

-10%

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

2005

Comparison with other sources

These ISM data, which we consider as reliable due to the fact that they are meteo standard measurements, performed using calibrated pyranometers and corroborated with our own scientific measurements in Geneva since 25 years [P. Ineichen], are compared with the Satellight 129 (satellite) and Helioclim 131 (from terrestrial measurements) data. We can observe that the Satellight data overestimate the ISM data by around 5%, while the Helioclim results are more chaotic (over-estimate the good years and underestimate the bad ones). The Helioclim-2

2007

Chapter 5 Geographical and Meteorological data

124

Geographical and Meteorological data

2005 values are 7.5% over the ISM measurements.
Geneva meteo data evolution and comparisons
15%
By respect to Meteonorm 81-90 (1212 kWh/m2)

ISM measurements Helioclim (red) and Satellight (green)

10% 5% 0% -5% -10% -15%

1981 1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006
1997 2000

Satellight data
For other sites in Europe, the Satellight data are always far over the Meteonorm ones, with one exception in Berlin. This exception is not attributable to the Meteonorm value; as we can see on the global comparison plot above, the Satellight data for Berlin are significantly below the other Satellight data. We dont have any explanation for that.
Satellight: Yearly GlobH Differences by respect to Meteonorm
20%

Average 1998

1996 1999

15%

10%

5%

0%

-5%

-10%

-15% Ham burg Berlin Caen Paris Geneva Lyon Toulouse Marseille Rom a Barcelona Madrid Sevilla

ISM hourly data in the database

A meteorological hourly database of 22 stations in Switzerland is delivered with the package. These data proceed from 10 years of measurements in the frame of the ANETZ meteorological stations grid of the ISM (Swiss Institut for Meteorology). The DRY format were prepared by the EMPA (Eidgenssische Materialprfung- und Forschungsanstalt, 8600 Dbendorf). By contract, their use is reserved for building's physics and energy computations. The user should commit himself not to yield these data to a third person. These Design Reference Years 116 (DRY) were constructed according to a procedure elaborated by the International Energy Agency (IEA), task 9, sub-task E. These data are based on 10 years of measurements (1981-1990). The DRY is a juxtaposition of selected months, chosen among these 10 years
125

Chapter 5 Geographical and Meteorological data

2007

Geographical and Meteorological data

as function of several statistical criteria, and approaching average values.

The DRYs files of the ISM-EMPA are made up of the hourly values required for the simulation of PVSYST: horizontal global and diffuse irradiances, temperature and wind velocity. Temperature data are slightly corrected: their dynamics is enlarged in order that it includes the extreme values observed over the 10 years. But these corrections preserve the internal coherence between irradiance and temperature, and should not introduce mistakes on the predictive (average) values of the PVSYST results. Diffuse irradiance is not measured in the ANETZ grid: it is calculated according to the Perez-Ineichen model.

Importing data from Meteonorm

The Meteonorm software (version 3.0, 1997, version 4.0, 2000, version 5.0, 2003, Version 6.0, June 2007, 6.1, 2009) provides basically monthly meteorological data for any location on the earth. These are based on more than 7700 well-established meteo station data, from which about 1'500 (named "stations") avail of Irradiance measurements. Data are monthly values, averaged over the period 1961-1990 (when available). A second period (1981-2000 for irradiations, 1995-2005 for temperatures) may be chosen in the new version 6.0. Data for any location on the earth, specified by its geographic coordinates, may be obtained by interpolation between measuring stations, taking altitude and region typology into account. For regions with poor terrestrial measurements covering, help of satellites data may be used. In this software: - "Stations" design the meteorological stations which avail of irradiance measurements. - 'WMO/OMM" are the meteorological stations recording many parameters, but which don't avail of irradiance data. - "Towns" are a set of some main big cities, for which the data are interpolated. The PVsyst database includes only "Stations" with well-measured irradiations. For the main central European countries, all available Stations of Meteonorm are referenced in the database of PVsyst. Therefore choosing another site in these regions will lead to interpolation. Meteonorm provides its data in several ways: - In Monthly Values as ASCII files (named xxxxMON.DAT), which may be directly read by PVsyst for creating geographic site 114 *.SIT file. - in Hourly Values as ASCII files (named xxxxHOUR.DAT). The hourly data are built by a stochastic model, quite similar to the "Synthetic Hourly Generation" 181 performed in PVsyst. The Meteonorm process also produces stochastic hourly wind values, when PVsyst only puts the monthly average wind speed. These data may also be directly imported in PVsyst. Moreover Meteonorm proposes other prestations like irradiance computation on tilted planes, but these are not useful in PVsyst, which uses its own models.

Importing data from Meteonorm

The importing process in PVSYST is quite automatised. In the Meteonorm software: Choose a location and generate a file (will be stored in the " \MeteoTest\Meteonorm\Output" directory) : - Monthly values: just save it as a file (default format), which will be named xxxxMON.dat - Hourly values: choose the special PVSYST output format before saving the file (main menu option "Format" / "Output Format", and choose PVSYST). This will create a file named xxxxHOUR.dat. In the PVsyst software : - In "Tools", choose the "Import Meteo Data" button. - Select a Meteonorm data file (directory "\Meteonorm\Output") and follow the instructions. You just have to define the country and the region. If you select a monthly file, this will create a Site (*.SIT) file in the database.
126

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

If you select an hourly file, this will create both an hourly meteo file (*.MET) and a Site file (*.SIT).

Restrictions:
With Meteonorm V 5,.0 and 5.1 the format of the monthly output file is erroneous. The "*MON.Dat" files don't include the site name nor the geographic coordinates. With these versions you should use the import in hourly values, which works quite well. With Meteonorm V 4.0, the same problem arises when you ask the construction of hourly values. Please save the Monthly values before constructing the hourly ones. With Meteonorm V 6.0, the hourly files generated for PVsyst were erroneous at beginning. If your Meteonorm was installed before September 2007, you should import and install the "Patch version 6.0.1.4" (or upper) from www.meteonorm.com for importing hourly data properly.

Importing PVGIS data

PVGIS (Photovoltaiv Geographical Information System) is a research, demonstration and policy-support instrument for solar energy resource, part of the SOLAREC action at the JRC Renewable Energies unit of the European Communities (Ispra). You will find a complete description of this project at http://re.jrc.ec.europa.eu/pvgis/info/faq.htm#data The GIs database covers two regions in different ways: European subcontinent: - Geographical data: digital elevation model (1 km x 1 km for horizon evaluation), administrative boundaries, CORINE and Global land cover, cities, etc. - Spatially Continuous Climatic data: monthly global irradiation (from 566 ground meteorological stations, 1981-1990 averages from the ESRA project), diffuse/global ratio, air temperature, Linke atmospheric turbidity (useful for a future improvement of the clear day model). - Other computed values, not directly useful for PVsyst (optimal inclination angle for PV collectors, yearly irradiations on different inclined planes, PV yield, etc). Mediterranean basin, Africa and South-West Asia - Geographical data: elevation model (1 km x 1 km or 2 km x 2 km), administrative boundaries, Global land cover, cities, etc. - Spatially Continuous Climatic data: monthly global irradiation, from Helioclim-1 131 database (Ecole des Mines de Paris/Armines), based on METEOSAT images (1985-2004), with resolution of about 30x30 km. Air temperature, Linke turbidity. Several data sources of different kinds have been used to develop this database, which are listed on http:// re.jrc.ec.europa.eu/pvgis/info/faq.htm#data You will also find here an estimation of the global accuracy of the results. Owing to their continuous structure, the main results and parameters of this project are often presented as coloured maps at regional or continental level. New database Climate-SAF PVGIS These data are based on calculations from satellite images performed by CM-SAF (Geostationnary MeteoSat and Polar EUMetSat). The database represents a total of 12 years of data. From the first generation of Meteosat satellites (Meteosat 5-7), known as MFG, there are data from 1998 to 2005 and from the second-generation Meteosat satellites (known as MSG) there are data from June 2006 to May 2010. The spatial resolution is 1.5 arc-minutes (about 3km right below the satellite at 0 N, 0 W). The coverage extends from 0 N (equator) to 58 N and from 15 W to 35 E. These data are more representative of the last years climate, and show often higher irradiations thatn the "Classic" PVGIS. See http://re.jrc.ec.europa.eu/pvgis/apps4/databasehelp_en.html for further details.

127

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

Importing PVGIS data
Please open the web site http://re.jrc.ec.europa.eu/pvgis/

In the frame " Interactive access to solar resource and photovoltaic potential ", click on the "Europe" or "Africa" icon. You will obtain a GoogleMap-like tool for choosing your site. Typing a town name will put it at a suitable scale for navigating and easily find your exact location. On the right dialog, select "Monthly Radiation" tab, and check the following variables to be imported: - Horizontal Irradiation - Dif. / Global radiation (optional), - Daily Average Temperature (not available for Africa, will be imported from other sources) PVsyst does not make use the "Average Daytime Temperature", as the synthetic generation model for the daily profile includes the Day/Night effect, and waits for a 24h average as input. Then you choose "Web Page" output option, and click "Calculate" Select the whole text from "PVGIS" to the end of the table, and Copy (Ctrl-C) it into the clipboard. NB: If the data window doesn't appear after "Calculate", it is usually because your web navigator is blocking the advertising windows. In MS Internet Explorer, you can manage this in the "Options" of the main menu. NB: Up to the version 4.33, import of PVGIS data was performed using the old interface of the PVGIS site. This new interface is much more convivial. Come back to PVsyst, choose "Tools" / "Import Meteo Data" / "PVGIS", and follow the instructions in red.

Importing US TMY2 / TMY3 data

Data files of Typical Meteorological Year (TMY 116 ) are available on the web for 1020 stations in the USA, at the address http://rredc.nrel.gov/solar/old_data/nsrdb/1991-2005/tmy3/ . They can also be obtained as a single zip file or on a CD-ROM. These data sets are derived from the National Solar Radiation Database (NSRDB) and produced by the National Renewable Energy Laboratory's (NREL). A complete User's manual is available on the web site. Formerly available as TMY2 (239 stations, 1961-1990 data), since 2008 this database has been extended to TMY3 (1020 stations, 1991-2005 data). These new TMY3 are base on more recent and accurate data. TMY 116 s are data sets of hourly values of solar radiation and meteorological elements. They are juxtapositions of months or periods of real data, chosen in the multi-year data set in such a way that they represent a typical 1-year period. Their intended use is for computer simulations of solar energy conversion systems and building systems to facilitate performance comparisons of different system types, configurations, and locations in the United States and its territories. Because they represent typical rather than extreme conditions, they are not suited for designing systems to meet the worst-case conditions occurring at a location.

How to import US TMY2 / TMY3 data

NREL TMY2/3 data can be directly and very easily imported into PVSYST: - First download the desired file from the web site referenced above. TMY3: CSV file, 1.8 MB. TMY2: zipped files of 235 kB; choose the "DOS" format and store it in a temporary directory; then expand the file (these are self-extracting files, of the form "12345.EXE"). - The files are referenced as station code numbers. If desired you may rename them for easier identification. - Returning in PVSYST, choose "Tools "/"Import Meteo Data"/US TMY2/3, and follow the instructions on the screen. The site specifications (name, geographic coordinates) are automatically read from the TMY Chapter 5 Geographical and Meteorological data
128

Geographical and Meteorological data

file.

Importing a TMY file will produce a PVSYST hourly meteo file (*.met) with Global and Diffuse irradiance, temperature and wind velocity, as well as a geographic site file (*.sit).

Importing Canadian EPW data

These data files of Typical Meteorological Year (TMY 116 ) are available free on the web for 72 stations in Canada , at the address http://apps1.eere.energy.gov/buildings/energyplus/cfm/weather_data3.cfm/ region=4_north_and_central_america_wmo_region_4/country=3_canada/cname=CANADA Produced by Numerical Logics in collaboration with Environment Canada and the National Research Council of Canada, these Canadian W eather for Energy Calculations (CWEC) data were derived using a methodology similar to the TMY2. CWEC hourly files represent weather conditions that result in approximately average heating and cooling loads in buildings. Because they represent typical rather than extreme conditions, they are not suited for designing systems to meet the worst-case conditions occurring at a location. They are especially suited for the EnergyPlus program. The CWEC - EPW data are derived from the Canadian Energy and Engineering Data Sets (CWEEDS) of hourly weather information for Canada from the 1953-1995 period of record.

How to import Canadian EPW data

Canadian EPW data can be directly and very easily imported into PVSYST: - First download the desired file from the web site referenced above. After choosing the desired EPW file, you can simply download it by choosing "File" / "Save As" it in your web browser. - Then come back to PVsyst, option "Tools "/"Import Meteo Data"/Canadian EPW , and follow the instructions on the screen (i.e. simply "Choose" the file and press "Import"). NB: If the file downloaded by this mean is not suited for PVsyst (gives errors), this may be due to a transformation of the format by your browser. In this case you should download the ZIP file and extract the *.epw file. The import of EPW data will create a Geographic site (*.sit file) and an Hourly Meteo data file (*.met file) with Global, Diffuse, Temperature and Wind speed values.

Importing Satellight data

Satellight results from a European Project Team (cf www.satellight.com ), which has used data from the METEOSAT Geostationary Satellite, and provides (free of charge) detailed Irradiance (and also Illuminance - useful for natural lighting studies) time series in half-hourly values for 5 complete years (1996 to 2000) and for any pixel of about 5x7 km in Europe. These data have been carefully prepared, involving complex treatment of multiple Satellite photographies in several wavelengths. They exhibit an impressive accuracy, of around 20% (RMSE) without significant bias, far enough for the simulation needs in PVsyst. It has been established that in hourly values, as soon as a considered site is more than 20 km apart from a terrestrial meteo station, the satellite data are of better quality than the measured terrestrial ones (i.e. the RMSD - Root Mean Square Differences - are lower), while the MBD (mean bias differences) averages remain very close to zero. The following graph shows comparison difference RMSD, between terrestrial measurements of several stations of the ISM network (Swiss Meteo Institute), showing the effect of the distance between a site and the use of the data. This also gives a comparison with the Satellight confidence level, lying around 20% (which has become even better with the latest developments of the models).

129

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

Importing Satellight data in PVsyst

PVsyst offers an automatised link for importing Satellight data. - First Get a standard Satellight Data file "NomSite.tsv" from the www.satellight.com site, following the recommended procedure 130 . - In PVsyst, open "Tools" / "Import Meteo Data", and follow the instructions displayed in red, i.e.: - Choose your source file "NomSite.tsv" - The site and country names are automatically set if they have been defined in the Satellight file. - Define the region, which will be Europe in normal cases. - If the source file includes data of several years (usually 5 years), choose the year. NB: PVsyst meteo data cannot cover more than one year. By default, the year will be part of the PVsyst file name "NomSite_Year.met", so that you may create data for the same site and each year. - All Years: you can also import all years present in your source file at a time (usually 1996-2000). The program will create the five yearly files, and a "Site" file with the monthly averages. This may be used for creating a synthetic hourly file 181 representing the average of the 5 years. Please remember that averaging hourly values doesn't make sense: only monthly values are sufficiently aggregated to be averaged over several years. But in this case you loose the advantage of hourly measurements. - In the present state the temperatures - required for PVsyst simulation - are not present in the Satellight data. Therefore they have to be defined, in monthly values. The PVsyst procedure will invite you to define these values, which may be imported from any close site from the database. - Press "Import" for creating the hourly meteo file (*.met), along with a site file (*.sit) with monthly values for use in PVsyst. The importing procedure will then generate synthetic hourly temperatures (according to irradiances), using the usual synthesis model 181 , - Please carefully check the data quality, using the graphs of hourly values. Especially check the values of some clear days, against the clear day model. Getting Satellight data from Web The Satellight Meteo data may be obtained free from the web site www.satellight.com. Nevertheless you should register on this site. First, make sure that your browser accepts cookies, unless several features of the site will not operate correctly, and you will not be able to choose your site. Then you are required to register, or give your Identifier and Password. Press "Login" button. Then press "Site " in the main menu, and follow the Satellight procedure for choosing your site or location: 1. - Select the Region on the map Chapter 5 Geographical and Meteorological data
130

Geographical and Meteorological data

2. - Select your site .

If available, you are advised to choose your site in the Database, so that its name will appear in the file and be directly set in PVsyst. The available countries are grouped according to the previous "Region" choice. Pressing "Next" leads to the "Site selection" panel. If you are blocked for more than 20-30 sec here, please check that you are accepting cookies If not found in the database, you can also specify any location by its coordinates (latitude and longitude). NB: GoogleEarth provides an easy mean for the determination of the geographical coordinates of your system. 4. - Time Period: choose - From Sunrise to Sunset - Clock time - The tool in PVsyst is designed for using legal time, not solar time data. - All years or only specified ones (PVsyst will create files for one year at a time, but you will be able to choose any of the present years in this file when importing). - Always choose All months. 5. - Outdoor Solar Informations: 1. - Solar information Nothing to define 2. - Irradiances 2.1 - Surface type: Horizontal 2.2 - Parameters: Global and Diffuse 2.3 - Informations: Useful for printed outputs only. You have to choose at least one of these possibilities. 6. - Outdoor Daylight Informations: - Nothing to choose, concerns Illuminance data for natural lighting studies. 7 - Submit The site will perform the calculations, and send you an e-mail within a few minutes for downloading your results, as a zipped file "SiteName.tsv" file, or "Lat_Long_.tsv"). After Unzipping, this file can be readily imported in PVsyst 129 .

Importing SoDa-Helioclim Data SoDa: Solar Data portal

The SoDa Service offers an access to a large set of information relating to solar radiation and its use. It builds links to other services that are located in various countries. To answer a request, the SoDa service invokes several resources to elaborate the appropriate answer and ensures the flow and exchange of information between the services and itself, as well as with the customer. It provides meteorological data from different sources, mainly from the Helioclim database, managed by Mines PariTech. The data are computed from meterological geostationary satellite images, and are representative of real years. SoDa proposes also a climatic data bank in monthly values, with irradiance and temperature, averaged over the years 1990 to 2004. It is provided by Mines ParisTech.

The following services are available from the SoDa portal:

HelioClim 3: data for Europe and Africa, from 15 minutes time step to monthly values, for the horizontal global irradiance, the normal beam irradiance, and the global, beam and diffuse on inclined surfaces. These data are available for free for the year 2005. The same components are available for pay from 2004 to the current month, from a 1minute time step to monthly values. MARS data: this resource provides time-series of daily irradiation from 1975 onwards for 50km x 50km aeras, computed from spatial interpolation of time-series measured in meteorological networks; they are provided by the European Commission (UE). Nasa SSE: this resource provides time-series of daily irradiation from the 1st of July 1983 to the 31st December 2003 for cells of 1 degree by 1 degree.

131

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

-

Nasa-SSE + HelioClim-1: get the best of the NASA SSE and HelioClim Databases of Solar Radiation over the world. This service provides time-series of daily irradiation from the 1st of July 1983 to the 31st of December 2003. It automatically selects the database offering the best quality for the selected site.

These data don't provide temperatures, which have to be obtained by other means.

Methodology

(taken from P. Ineichen, 2011_A)

The Helioclim 3 data bank is produced with the Heliosat-2 method that converts observations made by geostationary meteorological satellites into estimates of the global irradiation at ground level. This version integrates the knowledge gained by various exploitations of the original Heliosat method and its varieties in a coherent and thorough way. It is based upon the same physical principles as for SolarGIS 133 , but the inputs to the method are calibrated radiances, instead of the digital counts output from the sensor. This change opens the possibilities of using known models of the physical processes in atmospheric optics, thus removing the need for empirically defined parameters and of pyranometric measurements to tune them. The ESRA models are used for modeling the clear-sky irradiation. The assessment of the ground albedo and the cloud albedo is based upon explicit formulations of the path radiance and the transmittance of the atmosphere. The turbidity is based on climatic monthly Linke Turbidity coefficients data banks. The Liu and Jordan (1960) model is used to split the global irradiance into the diffuse and beam components.

Importing SoDa-Helioclim hourly and monthly Data

When you buy Helioclim hourly data , you can ask them in the PVsyst standard format 138 (CSV file). In this way you can directly import them into PVsyst using the option "Import Meteo Data" and choosing the option "PVsyst standard format". For getting Helioclim data from their web site: Please open www.soda-is.com select Visitors: "complete list" select "Solar data for free" (or Solar data for pay if you are entitled, identical procedures). select Solar radiation data: "for free" or "for registered users" select in the horizontal plane (0nly 2005 for free): Either Monthly values Or Hourly values: Month, HC3 Hour, HC3

NB: Monthly averages from 1985 (HC1) are no more available. They are replaced by daily values, in conjunction with Nasa-SSE data. Reading of these data will be updated in a next version of PVsyst. define the site by search, map or coodinates if available, enter the site altitude select beginning and end date (if several years are selected only the first year is taken into account by PVsyst). when available, let "Tilt angle = 0", "Azimuth angle = 0", "Albedo = 0.2" (PVsyst only imports "Meteo" data on horizontal plane). if available, select the time step (hour, month is not very useful) if available, select the time reference = UT (Universal Time). output in Text File or screen: select Textfile (CSV-lik e) select output format: SoDaCSV press Execute the SoDa service Click (with left button) and you will obtain CSV-like data directly on your screen.

Chapter 5 Geographical and Meteorological data

132

Geographical and Meteorological data

select all the data (if (hourly values: up to 31/12) copy the selected data into the clipboard (Ctrl-C)

return to PVsyst and click "Import" (for "Pasting" these data), and then follow the instructions (the city name and country are not part of the results and should be filled manually). you will be prompted for defining monthly temperatures, that you may take from a nearby location (in the database or other sources). with hourly values, please carefully check the consistency of the data with the "Clear sky" model on the hourly plots.

Importing SolarGIS hourly data SolarGIS / ClimData http://solargis.info/index.html provides climate data for the study of Solar
Energy systems. It provides namely solar radiation required for the simulation within PVsyst, from 01/1994 up to present, with a spatial resolution 250 m: - Global horizontal irradiance (GHI or GlobHor) - Diffuse horizontal irradiance (DIFF or DiffHor) - Air temperature (T): time coverage: 01/1991 up to present - Direct normal irradiance (DNI or BeamHor) are also available for concentrating systems. The hourly data are available in the PVsyst standard format
138

, for a direct and easy import in PVsyst.

The SolarGIS data are managed by Geomodel Solar S.R.O., Bratislava, Slovaquia. They are only available for pay. The data elaboration uses the best available data sources (Meteosat EUMETSAT, ERA Interim ECMWF), NCEP GFC and CFS.

Details of the Method

(taken from P. Ineichen, 2011_A)

The irradiance components are the results of a five steps process: a multi-spectral analysis classifies the pixels, the lower boundary (LB) evaluation is done for each time slot, a spatial variability is introduced for the upper boundary (UP) and the cloud index definition, the Solis clears sky model is used as normalization, and a terrain disaggregation is finally applied. Four MSG spectral channels are used in a classification scheme to distinguish clouds from snow and no-snow cloud-free situations. Prior to the classification, calibrated pixel values were transformed to three indices: normalized difference snow index, cloud index, and temporal variability index. Exploiting the potential of MSG spectral data for snow classification removed the need of additional ancillary snow data and allowed using spectral cloud index information in cases of complex conditions such as clouds over high albedo snow areas.

3Tiers meteo data 3Tiers is a US-based company which proposes climate data for the study of Solar Energy systems.
(http://www.3tier.com/en/package_detail/3tier-solar-time-series) It offers a complete 12 to 14 year data file of hour-by-hour values for the whole world, with a 3 km resolution. - Global horizontal irradiance (GHI or GlobHor) - Diffuse horizontal irradiance (DHI or DiffHor) - Direct normal irradiance (DNI or BeamHor) are also available for concentrating systems. 3Tiers doesn't provide ambient temperatures, which should be obtained by other means for a near site (for example NASA-SSE).
133

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

The hourly data are available in the PVsyst standard format They are only available for pay.
138

5
, for a direct and easy import in PVsyst.

Details of the Method

(taken from P. Ineichen, 2011_A)

Satellite-based time series of reflected sunlight are used to determine a cloud index time series for every land surface worldwide. A satellite based daily snow cover dataset is used to aid in distinguishing snow from clouds. In addition, the global horizontal clearsky radiation Ghc is modeled based on the surface elevation of each location, the local time, and the measure of turbidity in the atmosphere. 3Tier opted to use a satellite-based, monthly time series of aerosol optical depth and water vapor derived from the Moderate Resolution Imaging Spectroradiometer (MODIS). This dataset was combined with another turbidity dataset that includes both surface and satellite observations to provide a turbidity measure that spans the period of our satellite dataset and is complete for all land surfaces. The cloud index n and the clear sky irradiance Ghc are then combined to model the global horizontal irradiance Gh. This component of the process is calibrated for each satellite based on a set of high-quality surface observations. Gh estimates are then combined with other inputs to evaluate the other irradiance components Dh and Bn.

Importing WRDC Data

W orld Radiation Data Center provides monthly global radiation data for 1195 ground sites from a 30-year (1964 - 1993) data set. Unfortunately this database does not include temperature values, which should be obtained by another mean (for example a near similar site from the internal database, or Meteonorm 126 , Retscreen 136 or PVGIS 127 values). The WRDC database is maintained for the W orld Meteorological Organization (WMO) by the Russian Federal Service for Hydrometeorology and Environmental Monitoring in St. Petersburg. The basic data are available as daily irradiation values (or eventually hourly) in a special unpractical format not readily readable in PVsyst ( http://wrdc-mgo.nrel.gov/ ).

You can find a list of available sites, with their geographical coordinates, on the NREL web site http:// wrdc-mgo.nrel.gov/html/sites.html. But several of these sites hold only partial data (sometimes only a few months or even no irradiance at all). The NASA and NREL provide an interface for getting either daily values for given "real" months (not useful here), or monthly averages over the whole available measurements of the concerned site.

Chapter 5 Geographical and Meteorological data

134

Geographical and Meteorological data

Importing WRDC Data
You can access the WRDC data, using this NASA site, and easily import them in PVsyst: - Go to the NASA site - Click http://eosweb.larc.nasa.gov/sse/ "Ground Sites ", and then "Monthly Plots ".

- You will be prompted for registration with your e-mail address and a password (free). - Here you have 2 possibilities: - Either you choose geographic coordinates , and the nearest site will be selected, - Or you select a country, and then a site in the country. - Select the Monthly Averages item (other ones are daily values). The required data will appear, along with the geographic coordinates. - Select the data from the Country name to the bottom of the table, and copy (Ctl-C) into the clipboard. Come back to PVsyst, "Tools" / "Import Meteo Data" / "WRDC", press "Import" and follow the instructions in red. As they are not included in the WRDC data, you will be prompted to define the monthly temperatures from another source.

Getting NASA-SSE data of a particular site

NASA SSE (Surface Meteorology and Solar Energy programme) are monthly data, average of 1983-1993 satellite measurements, provided for any cell in a grid of 1x1 over the world (111km x 111kmcos(Lat)). See http://eosweb.larc.nasa.gov/sse/ for further information. Also available from this database, but direct import not implemented in PVsyst: irradiances or temperatures in daily values for any period in the 1983-1993 range. In contrast to ground measurements, the SSE data set is a continuous and consistent 10-year global climatology of insolation and meteorology data. It is derived from several databases, including "Goddard Earth Observing Systems (GEOS-1), the International Satellite Cloud Climatology Project (ISCCP D-1), from data of the Geostationary and Polar Satellites for Environmental Observation (GOES and POES), the European Geostationary satellite Meteosat, and Japanese ones. etc. Although the SSE data within a particular grid cell are not necessarily representative of a particular microclimate, or point, within the cell, the data are considered to be the average over the entire area of the cell. For this reason, the SSE data set is not intended to replace quality ground measurement data. Its purpose is to fill the gap where ground measurements are missing, and to augment areas where ground measurements do exist. Accuracy The accuracy of Satellite measurements has been evaluated using numerous ground-based measurements. Although the reliability of these ground measurements themselves is not always well assessed, the NASA estimates that the RMS Error on monthly values is around 13-16%, and the Mean Bias Error (MBE) lies from -2% to 0.7%. Further information about the data generation: The solar energy data is generated using the Pinker/Laszlo shortwave algorithm. Cloud data is taken from the International Satellite Cloud Climatology Project DX dataset (ISCCP). ISCCP DX data is on an equal area grid with an effective 30x30 km pixel size. The output data is generated on a nested grid containing 44,016 regions. The nested grid has a resolution of one degree latitude globally, and longitudinal resolution ranging from one degree in the tropics and subtropics to 120 degrees at the poles. This, in turn, is regridded to a one degree equal-angle grid (360 longitudes by 180 latitudes). The regridding method is by replication, wherein any grid region that is larger than 1x1 degree is subdivided into 1x1 degree regions, each with the same value as the original.

135

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

Getting NASA-SSE data of a particular site:

- Identify the geographical coordinates of your PV system (you can easily do this using GoogleEarth). - Choose "Tools" / "Import Meteo Data" / NASA-SSE", fill in the geographical parameters and press the button "Import". The full available Irradiance/Temperature data of the NASA project are included in the PVsyst package, and are readily imported. Alternatively, you can also find the data directly on the NASA site: - Go to the NASA site http://eosweb.larc.nasa.gov/sse/ - Click Data Retrieval: Meteorology and Solar Energy - Enter Latitude and Longitude of your system (or choose on the map) NB: in the data the conventions are the same as in PVsyst: positive longitudes from Greenwich, toward east. - Choose: Parameters for sizing Insolation (average, Min and Max) Meteorology: Eventually: Air temperature (at 10m) Wind speed at 50m
84

NB: we do not recommend to use these wind data for estimating Module temperatures simulation.

in the
114

Then you can print these data, and you should enter them manually as a new Geographic location

Importing Retscreen Data

The RETScreen International Clean Energy Decision Support Centre (www.retscreen.net ) seeks to build the capacity of planners, decision-makers and industry to implement renewable energy and energy efficiency projects. This objective is achieved by: developing decision-making tools (e.g. RETScreen Software), and by training people to better analyse the technical and financial viability of possible projects. RETScreen International is managed under the leadership and ongoing financial support of Natural Resources Canada's (NRCan) CANMET Energy Technology Centre - Varennes (CETC-Varennes). For Renewable Energy Studies, the RetScreen software includes a climatic database including average measurements for 4'700 ground stations, compiled from over 20 different sources, and covering the period 1961-1990. These data are presented as a unique coherent database, and present numerous parameters including Irradiation, air temperature and wind velocities. Several sources are for one only country (CERES for Canada, SAMSON for USA). The main source elsewhere in the world are WMO/OMM, WRDC 134 when more than 5 years. The irradiance are completed by the NASA 135 Data data when ground measurements are not available or reliable.

Importing Retscreen Data

You should first download and install the RetScreen 4 software from http://www.retscreen.net. Open the software Retscreen 4 (will open in msEXCEL). In the main menu, choose "Retscreen" / "Climatic database". Choose your country and site. There is no direct mechanism for getting the Retscreen data. You have to "copy" and "paste" each item independently (site name, country, latitude, longitude (be careful with sign), altitude). Then you select the irradiance column, "copy" (Ctl-C) it into the clipboard, and "Paste" it in PVsyst. Idem for the air temperature column. You have still to define the origin of the data (ground or NASA).

Chapter 5 Geographical and Meteorological data

136

Geographical and Meteorological data

Import of custom hourly data ASCII files
PVSYST offers a very general programmable interpreter for importing ASCII files (either meteo or measured data files). Conversion procedure 137

The format of the source-file may be very varied; however it has to fulfil the following minimal criteria: Include the values of one time step (hourly or sub-hourly) per ASCII line. Each ASCII line ends with the characters CR,LF, or only LF. Sub-hourly records are gathered in order to give PVSYST hourly internal records. Comment lines at the beginning of the file, or of each day or month, can be skipped. Data are organised in fields, with a chosen separator (semi-colon, comma, TAB, blank, etc. ) or fixed widths. PVSYST accepts a great variety of date formats, and adjusts the source data physical units. The available parameters should include the minimal data which allow the calculation of the horizontal global and the ambient temperature . Since simulation also requires the horizontal diffuse irradiation, this is calculated, for example, from the normal global and beam data (usual measurements in the USA), or from the horizontal global and beam data. If not available, the diffuse irradiation is reconstituted using the Liu and Jordan correlation. With measured data, if only the irradiation in the collector plane is available, the program calculates the corresponding horizontal global and diffuse (inverse of the transposition). Nevertheless, due to edge effects, these data will only be valid within the original collector plane. If the wind velocity is not available in the source-file, it will have to be provided manually in monthly values. Usually the values are consecutive; but if values are missing on the source file (for example night values), the date should be explicitly mentioned on each line of data. Given that the internal file must hold complete days, the data for any missing hour will be completed with zero irradiation values, and with stable temperatures after the last available reading. Gaps of more than one day may be acceptable with the option of writing the date for each reading on the internal file. This will cause a gap of one or several full days in the internal file, but the use of this possibility is not advisable, as the normalisations of the display of simulated values in daily units have not been tested in great detail for these special cases.

It is also possible to automatically chain data monthly or daily files, provided that their name include exploitable identification code (date).

ASCII Conversion procedure

See also : Conversion of hourly data ASCII files Please follow each step of the conversion dialog: Give a significant name to the internal file to be created. This will identify the file in the meteo or measured data list boxes. Please carefully choose this title since you won't be able to change it after conversion. Choose the geographical site . If not available in the database, you can take a near site and change it's parameters (site name, coordinates, etc.). This site will stay internal in your meteo file, so that these modifications will not be thrown out to your site database (unless you explicitely save it as file). Choose your ASCII source file , which can reside anywhere on your disk. Choose the target internal PVSYST file, which will be in the \Meteo\ directory (meteo .MET files) or in the \Datameas\ directory (measured data .DAM files). Choose or define the Format Protocole
139 137

interpreter.

In some cases (depending on the Format protocole), the program will still ask for the beginning date or the year.

137

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

When ready, press the "Start Conversion" button.

During execution, a control executing window displays the contents of the source-file's line currently being processed, as well as the actually converted meteo values, which will be transcribed on the internal destination-file. You have the opportunity of examining the process step by step in order to ensure that the values are correct. The ASCII interpretation has been made as reliable as possible, regarding to all unexpected events which may arise on measured data files of any kind. For example, if the conversion encounters an extra comment or unreadable line, it will ask the operator for skipping it's processing. It also performs global checks (limits, consistency) on each value. After conversion, you are advised to check your file with the "Graphs and Tables " tool (either for meteo or for measured data files), and carefully check the time shift 141 of your data.

PVsyst standard format

With the tool "Import ASCII meteo file", you can import Hourly Meteo data in virtually any ASCII format. But if you have reliable and clean data , you can also manage them in a spreadsheet (like EXCEL) according to the format specified below. Before using the EXCEL sheet, you have to save it as a CSV (DOS) file. NB: you have an example in the database \UserData\GeneveSTD.CSV The advantage will be a simplification of the import process, as the Geographical site information is directly written in the file. But this tool doesn't include all securities and error recoveries present in the "Import ASCII meteo file" tool.

Description of the PVsyst Standard format:

ASCII file : one time-step on one ASCII line (CR+LF line separators), all characters in the file should be plain ASCII. CSV file : "Comma Separated Values"-like. Should be easily managed in EXCEL, saved as "CSV(DOS)" file format. 1. Separator = semicolon or comma (dep. on the Windows international settings), 2. Decimal character should be a dot. Horizon: the data should be horizon free. Header: All header and comment lines begin by # The header may have as many comment lines as desired. First line: acts as a file format identifier: #Meteo hourly data or #TMY monthly data Files not beginning with these tags are not considered as PVsyst standard files. The basic general information required for PVsyst in the header should have the form: First cell: Second cell: Third cell: #Tag Value May be anything (not used): comment, units, etc

The necessary fields will be #Site; Sitename #Country; CountryName #Data Source; Source #Time step; Hour #Year; 2010 #Latitude; 46.21 #Longitude; 6.78

free text, will appear in the PVsyst site / meteo lists only required when not specified in the data Value in decimal degrees Value in decimal degrees (- toward west)

Chapter 5 Geographical and Meteorological data

138

Geographical and Meteorological data

#Altitude; #Time Zone; #Time Shift #Plane tilt #Plane azimuth #Albedo 440 1 10 30 0 0.2 Value in m above seal level Value by respect to GMT (- toward west)

Eventual Time Shift 119 for the calculation of the solar geometry. Eventual plane tilt if import of DPI (POA) data. Eventual plane azimuth if import of DPI (POA) data. Eventual yearly albedo

These tagged lines may appear in any order in the header. You can add any additional comment defining your product, beginning by #. Data: 1. First line includes column headers (without # as first character). 2. Second line mentions units (without # as first character). 3. Third line and next hold data Dates: as in EXCEL, the date format may differ according to the Windows international settings; we require one column for each element of date. Columns with the following headers are required: For Hourly data: Year, Month, Day, Hour, (Minute), the time is the beginning of the data interval, the solar geometry will be computed at the middle of the interval. RMY, TMY: If the data are not really measured at a determined date, the year should be 2059 Required data (in any order): Header = GHI Horizontal global irrad. [W/m2 or kWh/month] Header = Tamb Ambient (dry bulb) air temperature [deg.C] Additional data (optional, in any order) Header = DHI Diffuse horiz. irradiance/irradiation [W/m2] Header = DNI Direct normal irradiance [W/m2] Header = DPI Plane of Array irradiance [W/m2] Header = WindVel Wind velocity (at 10m altitude) [m/sec] Other columns with different data may be defined, but will not be used by PVsyst. Leap years: - Real data (with real year definition) should include 29 th February when necessary. - Averages data (with no specific year definition, i.e. 2059) should never define 29 th February. In PVsyst, data which are not really measured (averaged months, synthetic generation) are marked with Year = 2059. Missing data: The PVsyst simulation requires complete data (for relevant monthly or yearly sums). Missing data should be -99, they will be replaced by modelled values (synthetic generation). Night data : The night data are mandatory

Conversion protocol
See also : Conversion of hourly data ASCII files
137

The conversion protocol window holds several tabs, and includes a sample view of your source file, where you can follow step-by-step the effects of your choices.

General Tab
Defines the general file organisation (separator, time step of your data, lines to be skipped).
139

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

When choosing the proper separator the display sample file will be organised in field columns.

Date format
Please choose among: - Reference year: your source data are well organised in a yearly sequence, from January 1st, not leap year. In this case the program does not need reading the dates nor writing them on the file. This is suitable for clean TMY or DRY files. - Sequencial dates, not read on the file: your source file is still a clean sequence of records, but not necessarily starting from January, 1st, and of any duration (one year maxi). Therefore the program can only memorise the beginning and end dates, it doesn't need to read nor to write dates on files. - Dates read on the file: If your data sequence is not very clean (some holes), the program should read the dates, and write them on the target file. This option displays a variety of date formats, including the field separator, as well as specific date separators (for example h, '/' or ':' for hour:minutes), which may all be meant as 'non numerical' characters. Dates can hold month and day numbers, or day-of-year. This option even allows to read specific measured data daily files, where the day number is not included in the record, but only in the file name (to be used in conjunction with the "Chaining" facility). When reading dates on the file, you should specify the corresponding fields on the record. The program asks for each date field, and displays the corresponding labels on the columns. The time base of your data will usually be Legal Time . The program leaves the possibility of solar time, but it is hard to imagine that someone will record data in this mode, as he will have to adjust the clock periodically. Finally you should pay attention to the Record Time Label 141 definition of your data . If not properly defined, the hourly records will be shifted of half-an-hour or one hour, and solar geometry will not act properly, especially for transposition. The final result can be checked using daily profile meteo plots 165 : on these plots you can ask for superimposing the "clear day" profile on your data. When choosing good days the model curve should be centered on your hourly data. If not, either the Time Label is not well chosen, or the recording data clock was out of order.

Meteo variables (meteo files)

This tab asks for available parameters on the data file (irradiations, temperature, wind). For each chosen parameter, you should specify: the field order in the record; the label will appear on the corresponding field on the sample file display. the multiplying factor to be used in order to obtain the PVSYST internal units, i.e. MJ/m. For example, specify 0.0036 if your irradiation data are recorded in [W/m].

With the fixed width fields option, you can displace the separator columns with the mouse. When using Global on tilted plane , you should specify the plane orientation.

Measured variables (measured data files)

For a given system type, the program proposes the parameters likely to be measured, each of them being of course part of the simulation parameters. As above, you should choose the available parameters, their field order and multilying factor.

Gathering fields (measured data files only)

In many data acquisition systems you can have several measurements corresponding to a single global simulation variable. For example: - when dealing with several PV-fields, the measurements are often recorded for each field. But as PVSYST will treat this as one only system, the currents and energies should be added, and the voltages perhaps averaged (not a very good idea). - if you have current and voltage measurements, for computing the power (energy) at each step. - for averaging several module temperature sensors. Therefore this tab allows to gather the data of several source channels in one only target internal variable. This can be performed using sum, product or average operator. After defining the number of gathered fields desired for each variable, the corresponding label will appear Chapter 5 Geographical and Meteorological data
140

Geographical and Meteorological data

several times in the "field order" list for associating different source fields.

Chaining files
One only file : the usual case. Manual chaining: at the end of the file, the conversion process asks the operator for the next file, to be written on the same internal data file. One file per day: the file names - DOS names restricted to 12 characters - should include day and month identifiers in a specified coded format. Then the conversion will automatically chain the daily files.

- One file per month: as above, for monthly files. Generalities The format protocole defines all characteristics governing the ASCII source file reading (Meteo or Measured data files). The ASCII file is assumed to hold one record (line) per time step. Time steps should be hours or fractions of hours. The Protocol defines Skipped lines, Separator, Variables to be read, Variables reading format. It performs physical unit conversions to match the PVSYST internal standard format [MJ, MJ/m, C, m/s]. Dates may be sequencial (read or not on the file), or can have "holes". Possibility of automatically chaining daily or monthly files with coded filenames. Date format separator

Data separator for date format

The standard separator | is the field separator you defined in the general characteristics. The / character may be one (or several) non-numeric characters. The values between parenthesis are optional. If months (MM) or year (YY) are not present in the data, put their field number to zero.

Sequencial dates The format protocol specifies "sequencial dates". i.e. data are assumed to be contiguous, without any holes. In this case the program does not read the date on the file, and the beginning date must be supplied by the user. Time label of measurements In the PVsyst convention, the time label always refers to the beginning of the record. That is, the time label 12:00 refers to an average or the sum (usually in legal time) in the interval 12:0013:00. - hourly data run from 0h00 (i.e. 0-1h interval) to 23h00 (23h-0h interval). - daily data identify the day (ex: 01/06/2010 0h00 corresponds to 01/06/2010, from 0H 00 to 23H 59'59"), - monthly data are identified by the first day of the month. Now in measured data files, each record usually holds a time label. Depending on authors or data acquisition systems, this time may be referenced either to the interval beginning, or the interval end, or to the interval middle point. Or even at any other shifted time within the hour (for example ISM Anetz data are recorded from 0:20 to 1:20). When importing the data, the "real" recording time of the data should correspond to the time of PVsyst, which is used for the evaluation of the solar geometry. The PVsyst time reference is assessed by the clear sky model, usually drawn on the hourly data graphs. A large discrepancy between data and clear sky model (over +/- 30 minutes) may be corrected by several means: - Adjust the "Interval beginning" or "Interval end" options in the format protocol,
141

Chapter 5 Geographical and Meteorological data

Geographical and Meteorological data

- Change the Time-zone 206 in the reference site definitions, - Adjust the "First line interval", or suppresse/add a line when using a reference year without date reading on the file. After this rough correction, you should also analyse the time-shift correction POA data. . .
119

, especially when importing

Data Multiplicator

Multiplying factor
This is the multiplying factor to be applied to your data, in order to obtain the PVSYST internal units, i.e. [ MJ/m2 ] and [MJ]. Examples: Irradiances given in [W/m2] => Multiplicator = 0.0036, Irradiances given in [kW/m2] => Multiplicator = 3.6, Powers given in [Wh] => Multiplicator = 0.0036,

Powers given in [kWh]

Energies given in [Wh]

=> Multiplicator = 3.6, => Multiplicator = 3.6,

=> Multiplicator = 0.0036,

Energies given in [kWh]

Variable definitions

Variables definitions
Please choose the available variables in the data, and define their field number and multiplicator. You may also gather several measured data channels in one only PVSYST variable (for example: add several arrays, or calculate energy from available voltage and current). To do this: First choose the target PVSYST variable in the list. Then choose "Gathering fields" tab, choose this variable and define the number of channels and the operator. This will duplicate the corresponding variables in this list, so that you can define a data channel for each one.

Data in Power or Energy

Power or Energy
If your source file is recorded with a time step less than one hour, you have to specify if the data are in Power (ex. [W/m2 ]) or in Energy (ex. [ kJ/m2 ], [Wh/m2]). With power measurements the interpreter performs an average over the hour. With energy measurements it will perform a sum.

Meteo Variable definitions

Meteo variables definitions

Please choose the available variables in the data, and define their field number and multiplicator.

Gathering Several Measured Variables

Gathering several variables

Please : - choose the preceding tab "Measured variables", and choose among the available PVSYST variables in the listbox, Then come back to this tab, select this variable and define the number of channels and the operator. This will duplicate the corresponding variables in this list of the preceding tab, so that you can define a data channel for each one.

Chapter 5 Geographical and Meteorological data

142

Geographical and Meteorological data

Daily Profile Generation

Daily profile generation

As imported data are in daily values, the program has to split them into hourly values. It may do this in two ways: - either by constructing a smooth distribution proportional to the Clear Day model (that is: a clear day with reduced amplitude according to the required daily sum), - or by generating a random distribution according to the Markov matrices, in the same way as when generating synthetic data. The second method is more realistic, but global simulation results are not very different.

Elimination codes Doubtful data may be eliminated from the simulation/comparison process and accumulations, either in comparison plots, or in the tables. A Flag code indicates how the elimination was done: Eliminations in the original measured data table: 0000 0001 : Hourly elimination, 0000 0010 : Daily elimination, Eliminations in meas-simul. comparison plots: 0000 0100 : Mouse elimination in hourly plot 0000 1000 : Mouse elimination in daily plot 0001 0000 : Limits on measured values 0010 0000 : Limits on simulated values 0100 0000 : Limits on differences meas.-simul. data 1000 0000 : Limits on ratios meas./simul. Data Yellow lines in tables indicate missing data in the original file. This third part of PVSYST gives access to the following topics: Meteo Database: 113

Geographic sites: 114 geographic parameters of about 200 sites in the world, including monthly meteo data (horizontal global and temperature). Synthetic hourly data generation: above monthly data.
181

to generate meteo hourly synthetic data from any

120

Importing Meteo data from external Databases most popular databases. Import ASCII meteo files: format.
137

: allows the use of meteo data from the

allows to import meteo hourly or daily data in almost any ASCII powerful visualising and analysing tool for hourly meteo data

Meteo Tables and Graphs: files.

116

PV component Database: PV modules Batteries Pumps

156 154 145

, ,

Grid inverters ,
159

150

, , for stand-alone systems

162

Regulators

Regulators (Controllers) Back-up generators Seller list

163 156

for pumping systems

(Gensets),

143

Solar tool box:

Chapter 5 Geographical and Meteorological data

Tools and databases

-

Graph/Tables of solar parameters: displays solar geometry and Clear Sky irradiation on planes. Electrical behaviour of PV arrays: 36 reverse characteristics of PV modules, mismatch, array with shaded cell, heterogeneous array. Monthly meteo computation: 179 quick meteo evaluation using geographical site database, with horizon, tilt, sheds and sun-shields, IAM effects. Transposition factor: field orientation optimisation and evaluation tool. Operating Voltage optimizationsation
164 137 163

. allows to import measured hourly (or sub-hourly)

Measured data analysis:

Importing ASCII measured data files: data in almost any ASCII format. File transformation
166

: technical tool for merging and cutting PVSYST measured data files.
165

Data tables and graphs: data files.

powerful tool for visualising and checking hourly measured

Measured Data Analysis: 164 parameter definition, simulation, and close comparisons between measured and simulated data.

Favourites
You can choose your favourite components for showing only limited lists in all choices. Your existing favourites are shown in green in the complete lists. You can define your favorites in the main database lists of the "Tools" part: Simple way: right-click the concerned element for selecting or deselecting it. For lists of components: - Click the button "Set Favorites", - Select the desired components in the general list, using the "Control" Key for enabling Multi-selection.

PHOTON database
You can directly import components from the PHOTON database. In the database management part ("Tools"), PV modules or Inverters, please press the button "Import from PHOTON". This will store the Web address of the concerned Photon database, that you just "Paste" in your web browser. After choosing your component, please select the whole text of the page (including "Supplier" on the top) and "Paste" it (Ctrl-C). Then return to PVsyst, and simply press "Import PHOTON" in the component's definitions. You may obtain a warning if some parameters are not understood or compatible by PVsyst. Please note: Many components are not fully defined in the PHOTON database. Some missing parameters (often Vco, Iscd, Vmpp and Impp) will prevent PVsyst for the import of this component. This is a problem of the PHOTON database, PVsyst cannot of course guess the missing parameters !

Chapter 6 Tools and databases

144

Tools and databases

Phovoltaic modules
All parameters related to a given PV module, as well as graphs of its behaviour, are available in the PVmodule dialog which includes several definition sheets: Basic data
145

, which holds the module identificators and main electrical characteristics,

Model parameters 184 , specifies some additional parameters necessary to the PV "one-diode" model 182 , and calculates the model unknowns. The "Model parameters" include the following subsheets (see Parameter summary 147 ) : - Rshunt - RSerie define these two basic parameters for the establishment of the model, - Rshunt Expon describe the experimentally observed exponential behaviour of the Rshunt value, - Recombination losses special modification of the model for amorphous and CdTe technologies (see modified "one-diode" model for thin films 186 - Spectral correction to be applied to amorphous technologies, but not CdTe, - Temper. coeff. allows to fix a required value for the muPmpp temperature coefficient.

Size and Technology Commercial data

146

146

, with physical and secondary characteristics,

182

, behaviour over a great

Graphs, a tool visualising usual and less usual graphs of the PV-model variety of operating conditions.
148

Photovoltaic modules database

: data source, adding new modules.

149

Photovoltaic modules database in Tabular form

Phovoltaic modules - Basic data

Module identifiers: - Model and Manufacturer will appear in the module choice lists. - Data source usually refers to the main parameter measurement source (most often Manufacturer, may be an independent institute or your own measurements). - File name should have the extension '.PAN'. - Nominal power is the rated power specified by the manufacturer at STC 206 . It could be different from the MPP model result. It will be used to determine the "installed power" of systems, which is involved namely in the normalised performance coefficients 98 . - Tolerance is the rated limits of the Nominal power dispersion, given by the manufacturer. The lower tolerance limit could be used in the Module Quality Loss specification 86 . - Technology gives choice for the main technologies available on the market. Except for Crystalline cells - for which the standard "One-diode" model is suitable - the "Advanced" button allow for correcting this model in order to match the "Thin films" and other special behaviours. The technology namely determines the value of the energy gap in the model (for example 1.12 eV for crystalline silicon, 1.7 eV for a:Si). Manufacturer Specifications or Other measurements These parameters are the main electrical module characteristics, available in any manufacturer data sheets. They are usually given for STC 206 , but the program accepts measurements performed under other conditions to establish it's model, allowing for on-site measurements in external conditions. These parameters therefore include: 145

Gref and Tref: the reference irradiation and module temperature conditions during measurements. Isc and Voc: module short circuit current and open circuit voltage at these given conditions.

Chapter 6 Tools and databases

Tools and databases

-

Impp and Vmpp: any operating point in the region of the maximum power point at these given conditions. uIsc : current temperature coefficient. This has only weak influence on the module behaviour. When not known, it can be taken as about 0.05%/C.

Internal model result tool After defining these basic parameters, the program still needs the definition of additional parameters - i.e. Shunt and Series resistance, as well as the number of cells in series - to establish the One-diode model 182 parameters. This special tool displays the module operating parameters calculated by the PVSYST model, for any given irradiation or temperature conditions. See also the PV module parameter summary
147

for a complete description of all parameters.

Phovoltaic modules - Sizes and Technology Important parameters: Module size should be defined properly, as it determines the module area and efficiency. The number of cells in series is also important for the model calculation. A very erroneous definition will give rise to a warning and prevent establishing the model. The number of cells in parallel has no influence on the model. When defined, the cell area allows the calculation of the cell efficiency. But this data is rarely available, and leaving zero will not be a problem. Specific parameters: The reverse characteristics parameters are only used in the "electrical behaviour of PV-arrays " didactic tools for the study of partial shading on cells or mismatch. These definitions are not involved during the simulation process. The quadratic coefficient Brev 185 of the reverse characteristic is as measured for a single cell, in darkness. The determination of this parameter is rough, as it may fluctuate from one cell to the other, and more importantly, as it is strongly sensitive to temperature, which obviously is highly variable in these extreme conditions of dissipation. As its use is mainly qualitative, the default value, measured for ARCO monocrystalline cells of 100 cm, will be sufficient in most cases. The number of by-pass diodes is often not fixed for a given module model, and can vary according to customers. It is usually not specified in the datasheets and given values in the database are to be checked in each case. Informative parameters: The maximum system voltage is an informative design constraint, specified only for some modules in the database. See also the PV module parameter summary PV components commercial data
147

for a complete description of all parameters.

The commercial data - especially the retailers and prices of the PV-components cannot of course be defined in the database, as they can vary from country to country. They are only meant to serve as a memo for the engineer. Nevertheless, you have the opportunity of defining the price of the components you are using, and to store the date when it was fixed. These component prices may be fed directly in the economic evaluation
106

of the "project design" part.

When using the "economic evaluation" tool, you can link the component price defined in your database, using the "default" checkbox. You also have a shortcut for defining or modifying the component price directly in the database. During this process, the currency in the database may be different of the actual working currency of the economic evaluation tool. Chapter 6 Tools and databases
146

Tools and databases

Phovoltaic modules - Parameters summary

We give here an exhaustive list of all parameters involved in the PV module definition, and their use.

Main parameters, to be defined for any module

Nom Power: Tolerance: rated power specified by the manufacturer at STC 206 . Should be close to Imp * Vmp. Lower tolerance specified for the Nominal power. If -0%, please put 0.1 % as the program interprets -0 as "value not available". Only used in PVsyst for establishing the default value of the "Module quality loss" factor (fixed at half the lower tolerance). Gives choice for the main technologies of PV modules available on the market. Acts on the model. Reference irradiance for the (Isc, Voc, Impp, Vmpp) specifications. Usually STC = 1000 W/m. Reference temperature for the (Isc, Voc, Impp, Vmpp) specifications. Usually STC = 25C, Short circuit current at (GRef, TRef) conditions, Open circuit voltage at (GRef, TRef) conditions, Max. power point current at (GRef, TRef) conditions, Max. power point voltage at (GRef, TRef) conditions, Temperature coefficient on Isc

Technology GRef TRef Isc Voc Impp Vmpp muIsc NCell serie NCells parall Cell Area

Number of cells in series, necessary for the determination of the model, for one cell. Number of cells in parallel, not really used by the model Area of one cell, will give sensitive area of the module, and allow the definition of a "cell" efficiency; Not necessary. Module Length Total length of the module. Necessary for the calculation of the module efficiency, Module Width Total width of the module. Necessary for the calculation of the module efficiency, Apparent length Tile modules or BIPV: if defined, will be used for the efficiency instead of total length, Apparent width Tile modules or BIPV: if defined, will be used for the efficiency instead of total width, Max voltage Max. voltage attained by the array in worst conditions Voc(low temp) - i.e. insulation voltage, Absorptivity Light absorptivity for the temperature evaluation through the thermal balance. Let as default. Nb of bypass diodes Not used by the simulation, definition not required. BRev Special parameter for reverse behaviour; not used in the simulation. Let as default.

Parameters of the "standard" one-diode model

Rshunt Shunt resistance at GRef. Measured as the inverse of the slope of the I/V characteristics around Isc at STC. If not specified, you can keep the value proposed by PVsyst. Intercept at G=0 of the exponential behaviour of Rsh according to irradiance, Parameter of the exponential. Series resistance involved in the one-diode model. For crystalline modules, you can keep the proposed value. Maximum possible value of Rserie. A higher value would not allow the I/V characteristics to pass through the 3 reference points. Apparent series resistance, as measured on the I/V curve around Vco at (GRef, TRef). This slope is the sum of the Rserie (model) and the exponential effects. Quality factor of the diode, involved in the "one-diode" model.

Rsh (G=0) RshExp Rserie, model Rserie max Rserie app Gamma

147

Chapter 6 Tools and databases

Tools and databases

muGamma IoRef muVco

muPmpp muPmppReq

Thermal (linear) correction factor on Gamma, modification of the "standard" model for obtaining a specified muPmpp if necessary. Diode saturation current, involved in the "one-diode" model. Thermal behaviour of the Vco. Related to Rserie, but the manufacturer data is usually not attainable with reasonable other parameters. This value is a result of the model, and will usually not be compatible with the manufacturer's data. Normally a result of the model. May be forced to a specified value by acting on the muGamma parameter. Specified value for muPmpp if required.

Parameters for the modifications of the one-diode model for amorphous and CdTe technologies
dmutau Specific parameter for the contribution of the recombination loss parameter. The validity domain of this value for finding a solution to the non-linear equations is strongly correlated with Rshunt and Rserie. It has implications on the thermal behaviour, especially of Vco. To our experience with the long-term measurements of several amorphous modules, the value of dmutau parameter should be rather near to its maximum (about 80 to 90%). Parametrized correction according to Air mass and Kt. Fixed correlation (from University of Loughborough), may be activated or not. Should ne be applied with CdTe technology.

Spectral corr

Photovoltaics modules database

The PVSYST database contains references to more than 6000 PV modules available on the European market. These earlier data came from a compilation of manufacturer's specifications, gathered by the TISO (Centrale di prova per componenti PV, Ticino, Switzerland). Some modules were really measured in real conditions and on a long period by this institution. Now the database is periodically updated using data published by the German Journal PHOTON Magazine . This PVSYST database has been limited to modules of power greater than about 35 Wp for Si-crystalline modules, best suited for grid-connected systems. Some new models submitted by the manufacturers are also used. Manufacturer's specifications, as well as these source databases contain the main necessary parameters for the introduction of new modules in the PVSYST library. Nevertheless, for definitive simulations, the user is advised to carefully verify the library data with the latest manufacturer's specifications; some technical data may be modified (especially care must be taken about geometrical dimensions!). We drop out any responsibility about the integrity and the exactness of the data and performances included in the library. On the other hand, the data (from manufacturers or TISO) are not always complete. Some parameters, necessary to the model, are often missing and should be fixed at realistic values. These are especially: The series and shunt resistances: never available in the catalogues, these values are sometimes given along with detailed measurements on modules. The shunt resistance may be estimated by a measurement of the slope of the I/V characteristics in the short circuit region (or the module resistance in darkness). The program default value 185 gives a value near to the measurements performed on several modules in our laboratory. This value has only little influence on crystalline silicon modules, but becomes significant with thin film modules. By default, the series resistance is then fixed in order to get a reasonable Gamma factor 149 . These choices may appear as arbitrary, but correspond to very close behaviours of the model, especially when restricting at MPP operation. The cells' area is involved in the efficiency estimation of the active area. If it is quite unknown, it may be put to 0 without danger, then the corresponding efficiency will be null. It may be important to define it

Chapter 6 Tools and databases

148

Tools and databases

with modules of special layout like tiles or spaced sells, for which the module area is not significant. The reverse characteristic 185 parameter is only used in the visualisation tools of array's behaviour with shaded or mismatched cells. Their influence is mostly qualitative, and depends strongly on the temperature. We don't avail of reliable indications except our own measurements on some modules. Therefore the database puts this measured value as default for any module. The number of protection diodes is usually not a basic characteristics of the module type: sometimes it has to be specified during the order. Data given in the database are not reliable. The temperature coefficient of the open circuit voltage muVoc is normally determined by the program (one-diode model) during the choice of the series resistance. Nevertheless it's variation's domain is restricted, and it is often hard to match the given value with an appropriate choice of Rs. Nevertheless, PVsyst gives the opportunity of adjusting the muVco value to a given value; but this is not recommended.

In version 3.4, some additional modifications 186 have been added to the Standard On-diode Model in order to better approach the amorphous and other thin film modules behaviour, on the basis of the more recent knowledge and our own experimental research. These involve special parameters that are not part of the manufacturer's data, which have to be fixed to default values according to our experiences on 6 modules. These additional parameters are: The Rsh(0) parameter (limit when Irrad=0, i.e. dark resistance), which governs the exponential shunt resistance behaviour as function of irradiance. For amorphous modules, the d/mteff recombination loss parameter.

Ideally these parameters - as well as the Rsh(STC) - should be part of the manufacturer's specifications in the future. It should be noted that the model used for modules is an approximation, sufficient for the use in the frame of the simulation, but which cannot represent exactly the complex behaviour of a PV module in all conditions: in particular, it is established for a single cell, and it can be extended to the whole module only under the condition that the cells are not rigorously identical. Nevertheless, the inaccuracies introduced by the model's imperfections or the approximation of some secondary parameters is widely dominated by the uncertainties coming from the variations between modules (or groups of modules), as well as those due to the methods of determining these parameters, which give not the same parameters during standard tests under simulator or in natural illumination. Photovoltaics modules database in tabular form The user has now the opportunity of exchanging all the parameter Module with an EXCEL file (or an ASCII text file), in tabular form (one line per module). The Microsoft EXCEL sheet "Components.xls" has been specially formatted for a good interpretation of the parameters (Required, Optional, Defined by PVsyst, etc.) When editing the PV module, the button "Export to Table" will store the data of the module in the clipboard, which you can "Paste" in EXCEL. Inversely, if you "Copy" a valid line in the EXCEL sheet, a button "Paste from Table" will appear in the PV module dialog, which allows for importing the data in the PVsyst program (and file). You can also export a set of chosen modules from the Module List in the "Tools" section on the software. When importing new modules without Rshunt and/or Rserie parameters, these will be set to default values by the software, in order to offer a valid model immediately useable. NB: The same feature is available for Inverters.

PV Modules: Serie Resistance determination

The Serie Resistance is one of the 4 unknown related (coupled) parameters when adjusting the one-diode model. The condition of passing through the three points (Isc, Mpp and Voc) will fix the 3 other parameter (Iph, Ioref and Gamma) as functions of a given Rserie value, which can be chosen between 0 and RsMax.
149

Chapter 6 Tools and databases

Tools and databases

The plot shows the corresponding I/V curve and its range of possible variations.

The Rshunt resistance is the inverse of the slope at the short circuit point (0,Isc). Its value is not critical, at least for crystalline modules with high fill factor. The value specified here is at reference conditions (GRef, TRef). It usually has an exponential behaviour according to the irradiance (see next page). The user can choose the RSerie value in order to obtain a specified Gamma (diode quality factor) value, or a required muVco value as specified by the manufacturer (which is not always possible). Alternatively, when clicking the RSerie Default checkbox - or when loading a new module - the program will search itself the RSerie value corresponding to a predetermined default Gamma value. Note that when at a given temperature the model results are not very sensitive to the Gamma value, it significantly influences the temperature behaviour of the model. From the junction theory, Gamma should lie between 1 and 2. With tandem technologies it will be between 2 and 4 and for triple junctions between 3 and 6 ! But our detailed measurements on some amorphous modules often showed higher values than expected. For Crystalline modules, we found that a Gamma value around 1.3 gives satisfying results. Therefore a Gamma default value is proposed in PVsyst for each technology: Si-Mono SI-Poly a-SI:H a-SI:H tandem a-SI:H tripple CdTe CIS AsGa Gamma = Gamma = Gamma = Gamma = Gamma = Gamma = Gamma = Gamma = 1.3 1.35 1.4 2.8 4.2 1.5 1.5 1.3

(measured) (unknown) (measured) (unknown)

As these values are not well established, they can be modified by the User in the "Hidden Parameter" facility.

Grid inverters
Grid inverters parameter are defined through a 5-sheet dialog: Main Parameters Efficiency curve
150

, the identificators and the fundamental properties of a specific inverter type,

152

Secundary Parameters
153

, the identificators and the fundamental properties of a specific inverter type,

, defines the detailed operating efficiency behaviour, ,

154

Sizes : physical sizes and weight, as well as comments about technology, Commercial data
146

The grid inverter database

includes about 30 inverters ranging from 0.8 to 100 kW.

Grid inverters, main parameters

This sheet includes the general data which are usually available in the manufacturer data sheets, as well as in some Inverter products databases (our main source is the Inverter surveys published every year by "Photon Magazine").

Inverter identifiers:
Model and Manufacturer will appear in the inverter choice lists. Data source: For most devices the PHOTON magazine yearly survey. But we have also many information coming directly from the Manufacturer. File name should have the extension '.OND'. Minimum and Maximum MPP voltages is the voltage window in which the inverter is able to search for the MPP. When sizing the array voltage (number of modules in series), this should be taken at operating

Input side (DC, PV array)

Chapter 6 Tools and databases

150

Tools and databases

conditions (around 50C or 60C). -

Absolute Maximum PV voltage is the absolute allowable maximum under any operating conditions. When sizing the array, it should be compared to 1000 W/m conditions, at the lower temperature (higher voltage) possible. This is fixed at -10C for middle Europe. This temperature condition may be adjusted in the "Hidden parameters" for other regions. Minimum voltage for PNom. Some inverters cannot deliver the full nominal voltage when the input voltage is too low (but still over the minimum MPP voltage). This corresponds in fact to an input current limitation. PVsyst takes this limit into account during the simulation, by displacing the operating point along the I/V curve, in order to respect this current limit. Nominal MPP voltage is sometimes specified by the manufacturer. In this case this may be an indication for the optimal number of modules in series. Not explicitly used in the sizing tool in the present time. Power threshold is the minimum input power needed to operate. It is admitted to be the own inverter power operating consumption. Sometimes referred to as "Starting production at ..." in the datasheets. If not known, you can take it at around 1% of the nominal power. This means that your inverter will begin to produce AC electricity from an irradiance threshold of 10 W/m...

The following parameters are often given by manufacturers, and sometime with a contractual constraint. But they don't have a real physical meaning as they depend namely on the plane orientation (please see the sizing tool). They normally don't present any danger for the inverter at running time as by overload, this device adjusts the power drawn from the PV array by displacing the operating point along the I/V curve. Nominal PV power is a usually specified parameter for inverters. It may be understood as the recommended nominal STC power of the PV array. Maximum PV power is sometimes specified by the manufacturers. It may be understood as the absolute maximal STC power of the PV array. If this is a contractual condition you have to check the "Required" checkbox, so that the sizing tool will warn you in red in case of excess. Maximum PV current is the absolute maximal current admissible at the input of the inverter, usually the ISC current of the PV array at STC. If this is a contractual condition you have to check the "Required" checkbox, so that the sizing tool will warn you in red in case of excess.

Output side (AC)

Although fundamental features of the inverter, the AC output parameters are not involved in the simulation results nor the system performance evaluation, which is only given in terms of Output Power (Exception: the Voltage/ Current characteristics at the AC side influence the eventual ohmic or transformator losses after the inverter). The mono or triphased mode: usually Mono- for inverters smaller than about 8 kWac, and Tri-phased for greater ones. Most grid distributors impose a limit on the phase unbalance, of the order of 3-5 kW. But of course you can distribute mono-phased inverters on each phase. The grid frequency is usually 50 Hz, and 60 Hz for the US zone. Many inverters are designed for accepting both frequencies, making them useable on all world markets. When choosing an inverter in PVsyst, you can ask for a selection of only 50Hz or 60Hz suited devices. The Grid Voltage parameter is also indicative in PVsyst. In the reality, the output voltage of the inverter will follow the grid voltage 63 at any time. The device is usually equipped with securities which will disconnect when the voltage goes outside a specified voltage range. But the effective Voltage thresholds specification is not part of the PVsyst parameters. By the way, PVsyst has no indication on the real grid voltage at a given time in the simulation process, and could of course not use this information. Nominal AC Power is the crucial parameter of the Inverter definition. It is the power the inverter can feed continuously. In PVsyst, when going over this power the limiting strategy is applied (displacement in the I/V characteristics, or cut). Maximum AC Power is often specified by the manufacturers, but its signification is not always clear. It is usually a power which may be produced for a short time, until the inverter's temperature becomes too high. As the device temperature is not managed in PVsyst (would involve number of unknown parameters), it cannot be taken into account in the simulation. Nominal AC current is often specified in datasheets. Its value is usually the current under Nominal power and Nominal grid voltage (see Grid current 202 for tri-phased), which you can obtain by clicking the associated Checkbox. Not used in the simulation. Maximum AC current is also an absolute rating specified by the manufacturer, not used in the simulation of

151

Chapter 6 Tools and databases

Tools and databases

PVsyst.

6
.

Efficiency
This panel indicates the maximum efficiency, as determined by the efficiency curve
153

It also computes from this curve the European efficiency, which is an average efficiency over yearly operating conditions in central Europe. Efficiency defined for 3 voltages: After recent tests and data publications about the efficiency as a function of the input voltage, there is now the opportunity of defining 3 different efficiency curves. The program will perform a parabolic interpolation between them at the simulation time.

Grid inverters, secundary parameters

These are not required parameters, defined for some models only. - String inverter: some manufacturers propose now inverters which may directly receive the string terminals without intermediary circuitry. The device includes all securities normally mounted in a junction box on the roof (fuses, non-return diodes, overvoltage protections). The number of "String" inputs should of course correspond to the effective number of strings. - Multi-MPPT: Some inverters now involve several MPPT inputs, which allow to connect PV arrays of different sizes, module types or orientations. This is now taken into account in the simulation. In the present state PVsyst assumes that all these inputs are equivalent, i.e. have identical input parameters. - Master/Slave: some inverters may operate in the "Master/Slave" mode, one of them ensuring the MPPT research, and transmitting this information to the other one(s). This allows to turn ON the required number of devices, improving the efficiency at lower powers. In this configuration the full array should be connected to all devices at a time, in parallel. For the Master/slave operation, you should define the Power threshold from which the second device will be turned ON. - Internal Master/Slave: many big inverters are now announced as "master/slave" devices. These are assemblies of standard units of a defined power, which operate internally in Master/slave mode. This operation mode results in a very flat efficiency curve (of the whole device). These should not be treated as "Master/Slave" in the simulation process as this behaviour is already taken into account in the efficiency. Other specifications give additional information about the configuration, features and securities of the device, not used in the simulation. These are: - Isolation Monitoring is a continuous check of the isolation of the PV array by respect to ground. DC Swich indicates if the device includes it, otherwise it should be foreseen externally on all DC inputs. AC Swich indicates if the device includes it, otherwise it should be foreseen externally. AC Disconnect adjust: the inverter should cut the AC connexion immediately when the grid voltage goes out of a given AC voltage range (in Europe, 230 V or 400V + 6% and - 10%). This feature allows to adjust these limits. ENS indicates that the cut-off security in case of grid defection work on the basis of the grid's impedance measurement. This system is mandatory in Germany.

Finally other parameter, related to old technologies, may be used for specific tests or analysis: - Operating mode: will always be "maximum power point", i.e. the input electronics is continuously searching for the operating point on the array characteristics, giving the maximum power (I * V). Although the fixed voltage mode is quite obsolete with modern inverters, it is still supported by PVSYST. Behaviour at nominal power (at AC output), can be the following: Power limitation: during overload (too much available energy at MPP), any modern inverter will limit the input power at the nominal value by displacing the operating point on the array I/V curve. This mode gives rise to usually low losses (only the excess virtual energy during very high productions), allowing to oversize the nominal array power 60 by respect to the inverter power (economic optimisation, depending on the system properties). Cut : when reaching overload, some old inverter models were simply cutting the PV production by safety (i.e. the array current was zero and the array voltage attained Voc). Cut up to evening : same case as above, but the inverter was not able to start again as long as the input was not null. 152

Chapter 6 Tools and databases

Tools and databases

Behaviour at Vmin/Vmax : -

Limitation: as above, in modern inverters the operating point stays on the limit voltage when MPP goes outside the window. Cut: in some old models the inverter production was stopped when encountering such conditions.

Grid inverters, efficiency curve Grid inverters, efficiency curve

The inverter's efficiency is characterised by a power transfer function during normal operation, depending on the instantaneous power. This transfer is usually given in terms of efficiency as function either of the input or of the output power. That is, it is represented by a non-linear curve, with a threshold input power which may be understood as the inverter's own consumption. However, it may be more convenient to transform this curve into an input/output power characteristic: the behaviour then becomes practically linear, with a pseudo-straight line cutting the abscissa at the working power threshold. The efficiency elbow at low powers is for a great part due to the mathematical transformation of this threshold. Inverter manufacturers try to improve the efficiency at low powers, giving rise to a hardly visible improvement on the beginning of the straight line (when represented as Input/Output graph). Defining the efficiency curve: The efficiency curve is rarely explicitly given by the manufacturers. For some inverters the database gives the efficiency curve measured by independent institutes. The programme offers the possibility of entering a profile characterised with at most eight points, in any one of these 3 modes ( Pout = f(PIn), or Efficiency as a function of Pout or Pin). Points may be entered by edition or mouse displacement. The particular behaviour which appears at nominal power (i.e. curve stop or sudden dropping discontinuity) is characteristic of the chosen overload power behaviour. Interpretation of the efficiency curve: In most of the specifications, the manufacturers or databases give the so-called "European Efficiency", which is intended to provide an average efficiency over yearly operating conditions. From these two values specified by the user - Maximum efficiency and European efficiency - , PVsyst constructs a default profile, with the following hypothesis: - Up to the power threshold (often named "Starting operating at ..." in datasheets), the input power corresponds to internal consumption for the internal needs of the device, and doesn't produce power. - From this threshold, the device is supposed to produce an AC power proportionally to (Pin - PThresh). - This production is penalized by an ohmic loss of the internal components (transformer and transistors), which increases quadratically with power (like R I). When translated into efficiency, this gives a maximum efficiency usually around 50-60% of the nominal power. PVsyst model for automatic profile In PVsyst, the contribution of the Resistive loss is fixed according to a normalized resistance factor, proportionnally to the difference between the given Max- and Euro- efficiency. Then, supposing the Max. Efficiency point at 60% of the nominal power, the effective threshold Pin_eff is adjusted in order to obtain the specified Euro efficiency. The normalized resistance factor commands the losses at high powers (above the 60% point), by respect to the losses due to Pin_threshold. We choose a usual value of 3.0 (arbitrary units) as default. The accuracy of this model is probably largely sufficient for representing most inverters. The less difference between Max- and Euro-efficiency, the less possible discrepancies of the model...

Grid inverters, adjusting the efficiency curve

If you avail of hourly measured data, and, in these data, of the the Inverter Output energy as well as the Array Energy parameters (measured with good accuracy) , this tool allows to compare the Inverter efficiency 153 defined for your device, with the experimental data. Moreover you can adjust your Inverter Device Specifications in two ways:
153

Chapter 6 Tools and databases

Tools and databases

- either you can perform an automatic best fit on the data, - or you can drag the reference points with the mouse in order to "visually" obtain the desired characteristics.

You are advised to save the inverter data with another name for your specific measured project, and you will then be able to use it when performing the comparison simulations.

Grid inverter database

The library includes about thirty inverters available on the Swiss and European market, in the range of 0.8 to 100 kW. Most of them have been measured by independent institutes (ISB - Ingenieurschule Burgdorf, Be, Switzerland, or GENEC, Groupement Energtique de Cadarache, France). The greatest inverters are often modular, or made to measure. It is not possible to propose current models, and we don't avail of measurements independent of the manufacturers. But these data will be easily introduced in the PVSYST's library with the help of the technical data from the manufacturers. There are several Inverter databases recently available. But these usually don't give an explicit efficiency curve behaviour, which is a necessary data for the PVSYST simulation process.

Batteries
We have given up to use the classical models (for example Shepherd's model), where a number of parameters are involved, which require practically a detailed measurement for each battery model used. We have tried to develop a two-level phenomenological model 192 , whose basic behaviour is simple and may be reproduced using the fundamental data furnished by all constructors, but to which specific disturbances are added; these being generally described by some manufacturers or battery specialists. For these secondary behaviours, when unknown, the user can do with the default values, specific to each type of technology, and proposed by the software. Therefore the PV-module dialog includes several definition sheets: Basic data
154

, the identificators and the fundamental properties of a specific battery type,

155

Detailed Model parameters Commercial data

146

, which gather the secondary behaviours given by default,

Graphs, a tool visualising the usual charge/discharge curves of the Battery Model electrical behaviour.
155

The Battery database

is unfortunately limited to a very few manufacturers on the European market.

Batteries - Basic data

Battery identifiers:
Model and Manufacturer will appear in the batteries choice lists. Data source usually refers to the main parameter measurement source (most often Manufacturer, may be an independent institute or your own measurements). File name should have the extension '.BTR'. Technology specifies vented or sealed (without maintenance) batteries, and tubular / plates / vehicle starting technologies. For the time being, the model only allows for the definition of lead-acid batteries. Ni-Cd batteries have quite a different behaviour, which has not yet been implemented.

Basic parameters:
The most important battery specifications present in any manufacturer data sheet. Number of elements (or cells), Nominal voltage: should be 2V for lead-acid batteries. Nominal capacity: as usually specified for a discharge rate in 10 hours (noted C10)and for a reference temperature of 20C. Capacity behaviour according to discharge rate and temperature will be defined in the next sheet. Internal resistance is considered to be constant (in approximation). 154

Chapter 6 Tools and databases

Tools and databases

-

Current efficiency (or coulombic efficiency), is the discharge/charge cumulated currents ratio, in [Ah]. For a working range below the overload, i.e., without dissociation of the electrolyte ("gassing" phenomenon appearing at approximately 85 to 90% of full charge), it is generally approximately 97% for lead-acid batteries. When the state-of-charge increases, the dissociation progressively appears, and is manifested on the one hand as an excess of voltage in relation to the normal charge curve, and on the other hand by a production of gaseous oxygen and hydrogen, which consumes a part of the charge current by electrolysis, thus disturbing the determination of efficiency. The real efficiency therefore depends on the working conditions and on the regulation. It will be quantified during simulation, and the current lost in electrolysis will give an estimation of the quantity of dissociated electrolyte.

In all of these definitions the voltage specifications may be displayed either by cell or for the whole battery.

Sizes
Specify the physical sizes and weight of the battery.

Batteries - Detailed model parameters

Allows to act on the secondary parameters used by the battery model 192 , which you should in principle not pay any attention to, except if you have specific data at disposal. The default values can be restored at any time by clicking on the "Default" checkbox.

Open circuit voltage

These parameters concern the modelling of the open circuit voltage, assumed to be linear up to the "gassing" region and down to the deep discharge beginning. These values, characteristic of the electrochemical Pb-H2SO4 couple, are drawn from a manufacturer's catalogue.

Miscellaneous parameters
Self-discharge current at 20C: behaves approximately as an exponential with temperature, doubling every 10C. The general temperature behaviour is specified as a dimensionless profile. Capacity vs discharge current, usually increases by a factor 25 to 35% by respect to the rated C10 capacity in solar use. Capacity correction vs temperature, increases by about 5% to 10% with temperature increases of ten degrees. Saturation charge current, charging saturation voltage at the reference temperature, i.e. limit for which the whole charge current is used for electrolyte dissociation. Gassing overvoltage profile, as a function of the state-of-charge. Phenomenological curve, not to be changed ! Lifetime vs depth of discharge, often given by manufacturers for solar batteries. The product of the depth of discharge by the number of cycles gives the total stored current during the battery life. Static lifetime at 20C, usually given by manufacturers for solar batteries.

Battery Database
The database is constituted of some battery models, available on the Swiss or European market, for different types and different technologies. Notice that though often used in small solar installations in holiday's houses, the car batteries (often called "starting" batteries), are not best suited to a solar installation running conditions. Built to stay charged most of the time, and to produce a big starting current, they are not able to be cycled, have an important self-discharge, and their lifetime is limited to about 4-5 years. Their use is economically justified in low-used installations (holiday houses or caravans). For industrial or professional uses, the special solar batteries (with grids or tubular, open or sealed without maintenance), are more expensive at buying time. But due to their lifetime the real stored energy cost is often lower. In the warm countries, where the overcharging and evaporation risks are important, the open batteries will be preferred to the sealed maintenance free models, thus the user can always complete the electrolyte by distilled water.

155

Chapter 6 Tools and databases

Tools and databases

Back-up generator
In PVSYST, the back-up generator is only considered as an alternator+rectifier element, ensuring the recharging of batteries when the solar energy is insufficient to satisfy the user's needs.

It is supposed to work only at nominal power. Its parameters are no more than the definition of the nominal power supplied, as well as a specific fuel-consumption (in volume per hour) to evaluate the back-up consumption. Provision is made for defining commercial available devices (sizes, weight, etc), but no one is proposed in the database.

Pump definition
General frame The pump device is considered as a black-box, with Current and Voltage inputs on the electric side, and Head and Flowrate values at the hydraulic side (cf Pump model 195 ). Technical features of the motor-pump aggregate are not needed in detail. In PVsyst, many Pump are associated with a power converter, which have to be included in the pump device definition. In these cases the input electrical variables are those of the converter. All parameters related to a given Pump, as well as graphs of its behaviour, are available in the Pump definition dialog which is composed of several definition sheets: General data
157

, which holds the pump identificators, and main electrical and hydraulic characteristics,

Detailed parameters 158 , specifies some additional electrical parameters, and the choice of a model according to the available operating data set. Data curves
158

tab(s) hold the definition of the operating data according to the model choice.
159

Current Thresholds

tab asks for the definition of the starting current thresholds, in a graphic-assisted way.

Size and Technology allow for writing a technical description of the pump features, as a text without limitations. Commercial data
146

Graphs and Running Conditions, offers a set of curves for visualizing the model behaviour, as well as a little tool intended to compute the pump operating state for any desired input or output variable.

Pump technologies Pump technologies

There are two classes of pump technologies: Centrifugal Pumps The water is moved with a rapid-rotating impeller. The pump should rotate at a sufficient speed for reaching the head required by the external system. The efficiency is mainly related to the flowrate. It shows an increasing curve (from zero efficiency at zero flowrate) until a maximum, which usually doesn't depend much on the Head. After this maximum the efficiency decrease is more marked for lower heads. The flowrate has a quadratic behaviour as a function of the power, with a power threshold depending on the Head; this corresponds to the minimum speed before reaching the external head. Centrifugal pumps are suited for systems with rather low heads and high flowrates. For extending the head range, many pumps use a multistage technique, that is implementing several impellers in series on the same motor axis, each ensuring a part of the required head. Positive displacement pumps In a positive displacement pump, the (uncompressible!) water is shut in an impervious moving volume, either with valves or with moving pieces with special geometric shape. Therefore some water is pumped as soon as the pump is rotating, and the flowrate is directly proportional to the pump speed. Power threshold is due to electrical losses in the motor before reaching a sufficient force for overcoming the torque. In many pumps, this starting torque is higher than the running torque (the friction losses are higher when the pump is stopped), requiring a starting over-current.

Chapter 6 Tools and databases

156

Tools and databases

There are several technologies:

- Piston pumps, where an alternating piston in a cylinder draws up the water from the inlet or pushes it out of the chamber to the outlet, using non-return valves. - Membrane pumps act in a similar way, except than the piston's imperviousness is replaced by a moving membrane. - Progressive cavity pumps use a special vis-shaped rotor in a cylinder, which imprisons a volume of water in the input chamber and pushes it along the tube to the output. - Rotating displacement pumps are made of rotor resembling to a paddle-wheel rotating in a cylinder with inlet and outlet openings. Positive displacement pumps are well-suited for high head systems. Their efficiency is usually rather constant for different flowrates. Surface and deep well "Normal" surface pumps are made of a motor and a pump aggregate, which are not necessarily integrated in a single case, giving the opportunity of coupling different motor types with different pump devices. The pump should be placed not too far from the water source (and at a maximum of about 5 mWater height for avoiding cavitation problems). There are no problems of accessibility for maintenance. But in many cases the fact of being placed above the water level requires a priming procedure, and some precautions for avoiding air entries. For deep wells, submersible pumps have to be placed at the bottom of the well. These should have of course a cylindric shape adapted to the well diameter, and the electrical part should be perfectly waterproof all over the life time. The technical constraints are more severe, and their quality should be much higher as the maintenance is not easy. Therefore the price of such pumps is usually much higher than for surface pumps. Moreover, it is technically very difficult (or impossible) to put several pumps in the same well. Nevertheless, there are now very sophisticated well immersed solar pumps on the market, some of them even including the power converter, and accepting a very large range of input voltages. These dramatically facilitate the system design.

Pump data: general page

Pump device identifiers:
Model and Manufacturer are identifiers which will appear in the pump choice lists. Data source usually refers to the main parameter source (most often Manufacturer, may be an independent institute or your own measurements). File name should have the extension '.PMP'. You can create a new pump device by changing the file name.

Electric side:
Definitions related to the input of the pump device, considered as a black-box: Define the motor type
73

, especially regarding AC or DC mode.

159

Define the power converter

type, if any.

Define the Nominal voltage (at nominal Head), and nominal (or maximum) powers at the three Head values specified in the hydraulic panel. These nominal or maximal powers are not quite clearly defined. They will be used during the simulation as nominal values (absolute maximum ratings should be defined in the "detailed" parameter). Some data sheets specify nominal values, and maximal ones only for exceptional conditions. NB: These voltage and power values are related to the pump motor itself in absence of converter, but to the converter inputs if present. if present.

Hydraulic side:
You have to define: The pump technology, especially whether centrifugal or positive displacement, The pump Layout, The Minimum head: usually the minimum head for which the manufacturer specifies data.

NB: A null Head at the pump with positive flowrate doesn't make much sense: it would mean that the flowrate is driven by other external forces. In this case, the efficiency is null.

157

Chapter 6 Tools and databases

Tools and databases

-

The Maximum head: usually the maximum head for which the manufacturer specifies data. This doesn't act as an absolute limit: if the system operating conditions require higher heads, the model results will extend to the required value. The Nominal head: this value is not well determined. It should correspond to the best suited head when using this pump. With centrifugal pumps, it could be chosen as the head with maximum efficiency. Otherwise it can be an intermediate value (rather toward higher values) between Min. and Max Heads.

When the model will be fully defined, the dialog will display the flowrates and efficiencies corresponding to these heads and their corresponding Nominal Power, as defined on the left "Electrical" panel. NB: One of the reasons for introducing the "Nominal Head" is to determine a more refined "function" for the Nominal Power as a function of any Head, useful in the model. With HeadMin and HeadMax, it would be a straight line between these two values; this adds an intermediate point.

Pump data: performance curves

This sheet is for the entry of the operating points defining the pump model
195

You may first choose the units of Head and Flowrate, according to your original data. Then for each curve you should choose the Parameter (Voltage or Head). Curve modifications act on the highlighted curve chosen by the "Set curve" box (red points). For constructing a curve, you are advised to put the data points at their approximate position using right-clicks of the mouse, and then specify their exact values in the editing boxes. As the curves are rather linear (and due to cubic interpolations), multiplying the data points doesn't bring more accuracy. Usually 4 or 5 points/curve and 3 or 4 curves are quite sufficient. In some cases Power thresholds (for flow production) are naturally defined by extrapolation of the data curve (for ex. in Flowrate=f(Power) curves). In other cases they are rather defined through the voltage threshold and the I=f (U) characteristics. The thresholds determination and behaviour around them is one of the more difficult features of the pump model. NB: After this definition, please have a look on the graphs for checking the general model behaviour. If irregularities or inconsistencies are observed, it is possible that one or another points is not well defined. We have sometimes observed that original data are not well defined (aligned) in the datasheets.

Pump data: detailed parameter

These are a complementary set of parameters.

Electric side:
- Motor type: reminds the choice of the preceding sheet - MPPT ou DC Converter (if any): asks for the model name (only informative, this doesn' refer to a device in the database). - Nominal Voltage: reminder of the preceding sheet. With DC converter: input voltage. In other cases: the most relevant operating voltage, often mentioned by the data sheets, even when the Voltage characteristics is not given. Other variables depend on the configuration: - Min/Max MPPT Voltage: the voltage windows for MPPT converter - Abs. Max Voltage, Abs. Max Current, Abs. Max Power: absolute maximum ratings at the input of the device (pump or converter), which should never be exceeded during the simulation. Corresponding protections should be specified in the Control Unit (either in the simulation, and in the reality !!!). - Maximum- and EURO-efficiencies, from which an efficiency profile will be constructed.

Hydraulic side:
Choice of the Data Set available in the data sheets. This will determine the model-type 74 used by PVsyst for simulating the pump behaviour, and makes available the corresponding sheet(s) for input of the data.

Chapter 6 Tools and databases

158

Tools and databases

Pump data: current thresholds

With Positive displacement pumps, and in absence of an integrated power converter, the motor will require an over-current before starting rotating. This panel asks for defining these over-currents for Head Min, Head Max, and an intermediate HeadMed at half distance between them. The final function at any Head in the model will result of a linear interpolation. Also the threshold voltage should be defined. This is the voltage at which the pump (i.e. Flow production) will stop. It usually corresponds to the elbow of the measured Current/Voltage curve. This elbow is not always well defined (not given in the data sheets). The model chooses it below the lowest specified running point. The I = f(U) behaviour between the last significant point and the origin (U=0, I=0) is approximated by a quadratic curve for completude of the model, but its exact values don't have a great importance during the simulation process.

Pump: integrated power converter

Necessity to include the converter in the pump definition When the manufacturer "imposes" a power conditioning unit device to be used with his pump, he usually specifies the converter electrical input and not the pump's input requirements as operating conditions. Therefore in these cases the component black box includes the converter-motor-pump set, and the PVsyst's pump model should act with the converter as input variables. This is namely the case with AC motor pumps, as PVsyst never manages the Pump input AC voltage nor its frequency. Only the Pump power input is a relevant input variable. This holds indeed for any pump for which the power is specified instead of the current and voltage (Flowrate=f (Power) model). Converter input The converter input corresponds - either to a fixed-DC voltage converter: the input voltage should be defined in the controller specifications. On most commercial devices, this is adjustable by hardware. At design time the controller dialog gives access to the Voltage Optimisation 163 tool for finding the best suited voltage. - or to a MPPT converter; the minimum and maximum tracking voltages have to be given in the Controller dialog. As for other converters, the Maximum- and EURO-efficiencies are required for establishing the efficiency curve. For simplification, there is no possibility for defining a customised efficiency profile. Converter output As it is usually specified by the manufacturer, the converter is supposed to match perfectly the input requirements of the pump. That is, it will be sufficiently well designed for being able to deliver the optimal voltage/current to the pump at a given time and for the required head, in order to use the available power with an optimal efficiency. The simulation will take into account the power limitation due to step-down technologie 160 , only when the pump's voltage behaviour is known in the model. This is the case when the pump model is given by a set of Flowrate=f (Power) curves, completed by additional I=f(U) curves on the pump input For Positive Displacement pumps, the converter replaces the booster, as the starting over-current is usually required at very low voltage, so that the starting power is not prohibitive.

Regulators for stand-alone systems

The regulator is at the heart of the working of a stand-alone system. It must ensure the protection of the batteries from overload (from the solar generator) and from deep discharge (by cutting off supply to the users). It can also control the starting of a back-up generator. Even though sophisticated regulation modes can be used in some photovoltaic systems (proportional regulators, partial cuts in fields, for example), the type of regulator currently available in PVSYST will only act on an "On/Off" basis. Its decisions are based on the modelled battery voltage. Each action is characterised by two thresholds starting and stopping - which can be given either in specific values by element, or for the whole battery.

Regulator parameters

159

Chapter 6 Tools and databases

Tools and databases

Besides identification as in other components (model, manufacturer, etc), the regulator parameters include: Technology and Data display specification. These are only indicative parameters, except that MPPT or DC-DC converters are specified by the "Technology" choice. Charging thresholds, for which default values proposed by the program are slightly dependent on the battery technology. Discharging thresholds are the regulator settings controlling the load disconnecting, ensuring the battery deep discharge safety. Temperature compensation (linear) is available with some regulators, especially recommended for use in warm countries. Some regulators have an internal sensor (they measure their own temperature), other have external sensor to measure the battery temperature. Maximum Input and Output currents are one of the most important real device characteristics. Before simulation the program will check the consistency of the regulator choice with the system characteristics. Associated battery pack: most regulators are suited for one -or sometimes two switchable - battery voltage. The simulation process needs knowing this switch position, and updates it at simulation time. As thresholds are slightly dependent on the battery technology, we also have to define it here. Back-up generator management is rarely available with little PV systems. When controlling the running of a back-up generator, some logic in the hierarchy of the thresholds should obviously be respected: for example, it would be stupid to define the generator's starting threshold as lower as the threshold of deep discharge security. The programme refuses erroneous configurations and produces an error message.

To decide on the adjustment of the working thresholds, it could be useful to have a look on the voltage curve of the battery according to the state of charge. One of the advantages of detailed simulation of PVSYST, is that it allows an in-depth study of the behaviour and the life of the batteries as a function of the thresholds chosen for regulation.

MPPT or DC-DC converter

Some PV systems are using a Power Converter 160 for a better match of the PV-array characteristics to the user voltage (usually battery voltage). In PVsyst simulations, the parameters defining this optional power conversion unit are part of the regulator device.

"Generic" Default regulator

When using commercially available regulators, the program checks if it is suited to the system properties. But in the early stage of the system study, the exact regulator strategy doesn't matter. That is the reason why we introduced a general purpose "generic" default regulator 161 , which automatically matches the system characteristics. This allows for performing the first simulations of a system without defining a specific regulation device.

Converter: step-down technology Converter: step-down technology

As a result of an electronic design constraint, most of the DC-DC (or MPPT) converters operate on "step-down" principle. This means that without transfo, they cannot deliver a voltage greater than the input voltage. Therefore in a pumping system, the PV-array MPP voltage should rise over the voltage required by the pump at the desired flowrate, at any time. Otherwise, even when the available power is sufficient, the pump operating voltage will adjust until reaching the corresponding (I, V) point on the PV array characteristics, reducing the effective useful power. This "step-down" limitation can be taken into account by the simulation only when the voltage behavior of the pump is well defined. With pumps specified only by power curves, it is neglected. NB: This means that when using standard pumps designed for 230VAC grid operation, the PV array should deliver at least 325V to the inverter for obtaining undistorted sinus at the pump (unless special step-up inverter, or inverter with transfo, is used).

MPPT or DC-DC converter

In PVsyst, the MPPT or DC-DC converter device - if present - is considered as part of the Regulator
159

Its operation and parameters are quite similar to the Inverter ones (in fact, the inverter is a MPPT-AC converter), i.e:

Chapter 6 Tools and databases

160

Tools and databases

- DC-DC converter: Nominal Fixed Input voltage, - MPPT converter: Minimum and maximum MPP tracking voltages, - Both : Absolute maximum input voltage, - Power threshold, - Efficiency profile according to power, - Maximum and eventually European efficiency, - Behaviour type at power overload or outside the voltage window.

As for inverter, the simulation process accumulates detailed specific energy losses during operating (inefficiency, under threshold, power overload, outside voltage requirements, etc).

Generic default regulator

When using commercially available regulators
159

, the program checks if it is suited to the system properties.

But in the early stage of the system study, the exact regulator strategy doesn't matter. That is the reason why we introduced a general purpose "generic" default regulator. When editing this regulator, most of the characteristics are preset at default values, and cannot be changed. These values (thresholds, maximum currents, etc) are automatically preset by the program according to the associated system configuration, in order that they are suitable for a correct simulation. Therefore, the first simulations may be performed with default regulation, without defining a specific device. NB: If you really have to change these default values: - For one specific system: please uncheck the "Universal Regulator" option; this allows for modifying the parameters; then you will "Save as" this component under another name. This way you will obtain a copy of the regulator, suitable for your actual system. - For all your future systems: the specific default values may be adjusted in the "Hidden Parameters"
167

part.

NB: With converters, the user may want to perform the simulation according to different converter efficiencies. With DC-DC converter, the user may also perform the simulation with several Input voltage values. Therefore the "Max efficiency" and "Euro efficiency" parameters, as well as the "Fixed input voltage" of the default regulator will be editable, and stored along with the "Simulation version" parameters.

Regulator for Pumping: parameters

Regulator for Pumping: parameters
See also general considerations about Control device for pumping systems
162

The Regulator dialog includes several pages, which are only available when pertinent. General Tab Reminds the main characteristics of any Pumping System Control Device, namely for system control and pump (s) operating safety. The parameter mentioned here are not all useful for simulation; but they will characterize the capabilities of each commercial device. - Main switch and Tank Full level sensor are in principle present in any control device. Pump inlet level sensor is required in deep well systems. - Electric safety limitations (Abs. maximum Power, Voltage, Current) will be set according to the pump(s) in the Default regulators. - If a Power conditioning Unit is used, you should decide whether it is part of the pump definition or defined in this component. This choice is automatic with Default regulators. - Some operational parameter for special configurations are defined here: Pumps cascading 77 : The irradiance threshold for switching the second (or third) pump ON. The program also gives the opportunity of defining pump Power thresholds for this purpose, but it is probably more difficult to implement in the reality. Array Reconfiguration
78 :

The irradiance threshold for commuting from parallel to series configuration. The irradiance threshold for turning the pump ON.

Battery Buffered system MPPT and DC-DC Converter 161

80 :

Chapter 6 Tools and databases

Tools and databases

Define the usual parameter for power converter units, that is: - MPPT: Minimum and Maximum MPPT voltage, the voltage range for tracking.

- DC-DC: Fixed input voltage, which may be adjustable on most devices. The dialog gives access to the Voltage Optimizing tool 163 for choosing the convenient value, according to the array definition. - Behaviour at Maximal Power, and Voltage tracking limits. In the Default Regulator definition, the above parameter are adjustable, but all other ones are fixed according to the system before the simulation process. - Maximum Array Voltage and Input Current, - Output (pump driving): nominal output voltage and power, maximal power. NB: When the converter is specified within the pump definition, the converter parameters of the pump are transferred into the Regulator device just before the simulation, overwriting the own regulator's definitions. Efficiency As for other converters or inverters in PVsyst, the efficiency curve may be either manually defined, or automatically constructed using the "Maximum efficiency" and "Euro efficiency" specified values. When the converter is specified within the pump definition, only the specification of Max and Euro efficiency is available (in the Pump definition dialog). Battery Management When using a battery, this panel defines the charge/discharge regulator parameters, in the same way as in regulators 159 for stand-alone systems. Commercial Data Identical to the other components
146

Provides an unlimited editing place for a detailed description of the device.

Converter in the pump's definition Converter in the pump's definition

When the pump's parameter are specified along with a Power Conditioning Unit, its characteristics will be automatically transferred into this Control device just before the simulation. Therefore in a real "Regulator for Pumping" device, you should decide whether the PCU is included in the pump's definition or not. This definition cannot be present at both places. That is, controllers with a PCU definition cannot drive pumps with a converter.

Control device for pumping systems Control device for pumping systems
In PVsyst, the control device is the heart of the pumping system operation. The PVsyst component definition, named "Regulator for Pumping" for convenience, includes several functions, which may be devoted to separate physical devices in the real system: - It defines the global operating configuration
74

of the system, and manages its specific parameters.

- It ensures the operating and safety limitations, namely the Power/Voltage/Current absolute limits, as well as the swiching OFF when the tank is full or the input level goes below the aspiration inlet (avoiding dry running). - It may include a power converter (MPPT or fixed V DC-DC, DC or AC output according to the pump device). When several pumps, there is usually one power converter for each pump, but in some cases there could be one only converter. NB: When the PVsyst pump device is defined along with its own power converter, the simulation will of course use the parameter of the pump power conditioning unit. See Regulator for Pumping: parameters Default Regulator When starting the study of a pumping system, the operating strategy choice is of course fundamental, but details
161

for a complete description of the component parameter.

Chapter 6 Tools and databases

162

Tools and databases

of the control are not relevant. Therefore, as for the stand-alone systems, PVsyst provides a Default Regulator for each configuration strategy. The secondary parameters of such regulators are fixed at optimal values according to the system definition (this is achieved just before the simulation process). But for each configuration, some operating values are basic choices of the user, which are a main part of the System definition parameters. These are for example Irradiance thresholds for Pumps cascading or Array reconfiguration, Input Voltage adjustment in DC-DC converters, etc. Real and Commercial Control devices In the reality, all functions of the PVsyst "Regulator" component will not always be gathered in one only commercial device. Therefore for your definitive simulations with real components, you will probably have to define your own "Regulator for Pumping" device, dedicated to your project. This may include the properties of one or several commercial components (for example existing control device + existing power converter), as well as some parameters specific to your project (for example Irradiance thresholds). NB: A specific Control device - as well as a power converter when needed - is often proposed by the pump manufacturers for each of their products. But they sometimes hold some "switchable" options, so that we have to define several corresponding components in the PVsyst database.

Seller list
The program proposes the frame for a manufacturer and retailer list, which is only meant to serve as a memo to the engineer. Each element of this database may be marked as "Manufacturer", "Retailer" or "Own contact". The concerned component type may also be specified. In this way, you can obtain specific lists, by clicking only the desired options. The PVsyst basic database currently includes some addresses of manufacturers of international interest (not exhaustive!). It cannot of course list retailers in each country. Nevertheless you can input your own contacts, and mark them as such, in order to easily select them.

Array-Coupling Voltage optimisation

This tool evaluates the PV-array performance as function of a fixed user's operating voltage. The MPP operating voltage of a PV array varies along the day and the year, according to Irradiance and Module temperature. Estimating the performances for a given fixed voltage is depending on: - The climate (will not be identical for tropical and medium-latitude). - The meteo distribution.(the optimum may be computed for a year, or a given season). - The PV plane orientation. - The Array composition (PV module, number of modules in series/parallel). - The protection diode voltage drop. - The Array wiring resistance: at the user level, the ohmic voltage drop will lower the nominal array output proportionnaly to the instantaneous current. The Relative average power yield (or efficiency) for different user voltage values is computed using meteo hourly data, over a given period. Observing the results may lead to the following remarks: - The Optimal fixed voltage is very near from the average MPP voltage. - Pure MPP operation gives a rather low yield increase of about 1.5 to 2% by respect to the operation near the optimal voltage. In practice, this implies that benefits of a MPP converter may be cancelled by the inefficiency of this device, when the operating voltage is well adapted. - The Efficiency is rather flat when modifying the user's voltage, but drops more sharply toward high voltages. - The diode and ohmic resistance voltage drops may be a significant contribution affecting the user's voltage match. - In middle latitude climates, the usual 36-cells modules match well the battery charging voltage (about 14V for a 12V-battery). In very hot climates, especially with long wiring circuits, this could require modules with more cells. - Special 33-cells modules are only suited for very simple installations without protection diode (not 163

Chapter 6 Tools and databases

Tools and databases

recommended).

Comparisons between measured and simulated values

The Comparisons between measured and simulated values is a very similar process as the Simulation process 22 . The main difference is that the original Meteo file of the project is replaced by the measured data file 164 , which is attached to each variant (so that the Project can treat several measured data files, for example for different periods). Therefore the measured data file should contain the meteorological data necessary to carry out the detailed hourly simulations. Regarding other aspects, the Project and parameter definitions organisation is exactly the same. Procedure: After choosing "Project Design" and the system type in the main window, the procedure is the following: First define the Project through the "File" menu.
28

through the "Project/Variant" button. You can also retrieve an existing project
30

For each variant, define the plane orientation Define the System
29

properties.

The program verifies the consistency of all parameters, and produces "Warnings" as Orange (acceptable for simulation) or Red (preventing simulation) LED's. When available (all parameters properly defined, that is only Green or Orange LED's), press the "Simulation 164 " button. When the simulation is completed, you will enter the "Results
96 " dialog.

In a second step, you can define if necessary : a Horizon profile

38

(far shadings),

Near shadings 37 , that is partial shadings of near objects, which require a rather complex CAO 3D construction of the PV-field environment.

Measured data analysis: general philosophy

The objective of this section is to closely compare on-site measured data with simulated values, either in hourly or in daily values. It has a two-sided function: On the one hand, it has helped us for the validation the software by comparing its results with carefully measured data in 7 installations. On the other hand, it constitutes a powerful tool for the analysis of the operation of PV systems in use, allowing for the detection and identification of even the smallest misfunctioning.

Procedure
This involves a much more complex process than the simple system simulation, which includes the following steps: 1. Importing the measured data: 137 . this is done by a programmable data interpreter, which accepts almost any ASCII file, provided that it holds records of hourly or sub-hourly steps, each one on a single ASCII line. It allows to choose, among the measured variables, those which suit the simulation variables. 2. Checking the imported data: 165 In order to verify the validity of the imported data file, a number of tables and graphs in hourly, daily or monthly values, may be drawn. Further, some specific graphs often used in PV data analysis (inverter efficiency, input/output diagrams, normalised performances parameters, etc.) are also available, allowing, at this stage, for using PVSYST as a complete tool for the presentation of measured data. 3. Defining the system parameters: 22 You have to define a project and variant parameters, exactly in the same way as for usual simulation. At this stage you should carefully introduce the real properties of your system.

Chapter 6 Tools and databases

164

Tools and databases

4. Comparisons between measured and simulated values: 164 after performing the simulation, you will obtain close comparison distributions for any measured variable. According to the observed discrepancies, you probably will analyse their cause and modify the input parameters accordingly. This offers a powerful way to exactly determine the real system parameters, as well as temporarily misrunnings. 5. Elimination of break-down events: 166 most of time measured data hold undesired records (break-down of the system or the measurement equipment), easily identifiable on the graphs. These can be selectively eliminated in order to obtain clean statistical indicators - mean bias and standard deviation - corresponding to normal running of the system.

Checking the measured data files

To check generated files in detail, you can use scrolling tables of monthly / daily / hourly values, as well as a set of control graphs for each parameter (time evolutions, histograms, etc.). Time interval definition check: The synchronisation of your data with the solar time is of great importance for solar geometry calculations, especially for the transposition model. Time definition checking 117 is performed through the same tool as for Meteo files. System data and running checks On the other hand, some specific graphs, usual when analysing measured data, are also available, which allow to use PVSYST as a complete tool for the presentation of data . These are: Input/output diagrams, which shows the system production as a function of the input irradiation, immediately gives indications about the misrunning days. Normalised performance 98 monthly and daily graphs, performances in a normalised way. summarising the system running

Inverter modelisation analyser 153 : this graphical tool allows to superimpose the theoretical efficiency profile of a library's inverter onto real measured data, and then to interactively modify its profile in such a way as to make it match the measured data. Within the limit of the accuracy of the electrical measurements recorded on the field itself (DC input and AC output), this gives a direct measurement and parametrisation of the inverter's performances in real conditions. You can then save the modified inverter for using it during the simulation process.

Predefinition of comparisons
As in the case of hourly or daily on-line graphs, the parameters to be compared should be defined before the execution of the simulation. So that during the simulation, the output file can store the couples of asked values (measured and simulated). For each comparison set, please choose: the parameter to be compared from the measured parameter's list the time interval and types of values to be accumulated (hourly, daily or monthly) you can impose some conditions for the accumulation: several conditions can simultaneously be specified: Horizontal or Incident global irradiation > 20W/m , in order to be limited to day-time values. Cuts of one, two or three hours after sunrise and before sunset. Selection of hours (mid-day). Eliminations of marked data (break-downs), explicitly eliminated 166 in the comparison graphs or the tables. Running of inverter for grid system, or production of PV-field in case of stand-alone system. Cuts according to the values of another measured variable, or a simulated variable chosen from among the 70 parameters calculated by the simulation. For example, this last option allows for the selection of the measurement points with a strong beam or pure diffuse irradiation (by asking "Diff/Glob Ratio"<25%, resp. >70%).

Then give a name to the accumulation (you can use the speed button on the right for pre-defined name).

After having defined as many comparisons as necessary, the simulation is started exactly in the same way as for the usual simulation. It follows the same process, on the basis of the meteorological data of the measurements 165

Chapter 6 Tools and databases

Tools and databases

file instead of those of the meteo of the project. At each time-step, and according to the specified conditions, it accumulates comparison data asked for above.

Transformations of data-files
The data files produced by PVSYST are not ASCII files, and have internal coherence constraints which imply that they should never be modified in a text-editor. But the programme proposes some useful transformations such as: Linking two files of same data structure over different periods. Linking is performed at a given fixed date, allowing to manage transitory differences in the accumulation format. For example: passing from summer time to winter time, which should be specified in the parameters of the site during the accumulation, and therefore giving rise to two distinct files during the conversion process. Be aware that PVSYST data files cannot run over 12 months ! Merging data, that is introducing one or several parameter data belonging to another internal data file. Example of use: introduction of meteorological data simultaneously measured by another acquisition system on a very near site. Cut of data according to dates. You can eliminate some given period (integer number of days) from a data file. Tip : You can also use the data elimination 166 tools for hiding faulty data in a file without definitively removing them.

Data elimination in Tables

You can mark (eliminate) break-down data - With the left button : elimination, - With the right button : to restore data.
166

in the tables by clicking on the desired line:

The eliminations performed in daily values will be thrown back to all hourly concerned values. You can also perform eliminations directly on comparisons graphs. These eliminations are of the same kind and will appear here if saved with the data file (.DAM file). NB. If you modify the data eliminations, the program will ask you for saving theses modifications on the data file.

Cuts of erroneous data

During a real data acquisition, the measured data often include erroneous measurements which correspond either to deficiencies in the measurement equipment, or to break-downs or disturbances in the PV-system itself. These non-significant values should be eliminated from the calculation of the comparison estimators, from the monthly accumulations, or even eliminated from graphs. PVSYST offers several means aiming to "mark" data for elimination: General criteria can be applied to hourly or daily comparison graphs: limits on the measured value, on the simulated value, on their difference or their ratio. Whenever possible, the graph explicitly traces the limits of the criterion. On comparison graphs, isolated aberrant points can be individually eliminated by a simple click of the left button of the mouse (or, contrarily, restored by a right click). On measured data tables
166

, single hourly or daily data can be eliminated with the mouse.

The eliminations performed on a comparison graph are carried over to all the other graphs of the calculated version. The eliminations of values in an hourly graph will therefore cause the elimination of the whole corresponding day on a daily graph. The eliminations performed on graphs can be saved on the file of the calculated variant (".CMi"), to be automatically displayed again in a later session. Further, the indications of eliminations of a calculated version can also be carried over to the original file of measured data (.DAM file). This allows, during another simulation, to reuse these break-down indicators as conditions for accumulation of new comparisons. This proves to be particularly useful for the elaboration of monthly comparisons graphs, where the elimination of break-down days is primordial. This chapter describes all technical aspects when using the software.

Chapter 6 Tools and databases

166

Technical aspects
Updating Software and Databases

Software and databases are periodically updated. To check for a new update of the software and/or the databases, choose "Web / Check for updates..." in the main menu. This will run the "Thraex Software's" AutoUpdater tool that automatically checks for new updates on our website www.pvsyst.com. The AutoUpdater tool wizard will open as follows.

Click "Next" to proceed with updates checking (if your internet connection needs a proxy configuration, click on "Connection"). If a new update is found, the AutoUpdater tool will inform you about the version number available, and then will automatically download the setup file. Click "Install" at the end to install the update. If the AutoUpdater tool fails to automatically install the updates, please visit our website and install the new update manually

Languages
Basically and historically, PVsyst was written in English. Some parts of the software are translated in other European languages: French, German, Spanish, Italian, but we could not translate the totality of the dialogs up to now. Virtually all options related to the design of grid connected systems are now translated, as well as all simulation reports (documents which can be directly presented to the final customers). Dialogs related to Stand-alone and Pumping systems are not translated. And the Help system has not been translated. The choice of the language is available in the main menu, or directly in the "Print" dialog (choice only valid for the concerned output). NB: If some user has competencies for proposing a translation in his own language, please contact the author. The job consists of filling a list of about 4000 words or pieces of text in an EXCEL-like sheet (see "Texts.CSV" in the \DataRO\ directory), and carefully checking the effect on the printed outputs. Nevertheless the present version is limited to languages which only involve a standard ANSI character set.

Special characters problems

PVsyst uses a standard "ANSI" character set for showing texts. This is quite standard for usual characters. But for some special characters, the conversion table provided by Windows depends on the international settings. This is especially the case for Asiatic users. Now PVsyst uses only a very restricted set of special characters within the Ansi set. These are essentially the square, cube or degrees (upperscripts), which may appear as local characters on some Asiatic machines. They may even lead to a crash of the program as these characters are stored in 2 bytes instead of one. The management of this ANSI table is quite confuse in the present Windows versions, and it seems to be quite impossible to choose a given table for use in a specific software written in DELPHI. Therefore in case of problems, the user has the opportunity of disabling the use of special characters, which will be replaced by standard ones (or equivalent expressions).

Hidden Parameter
Many parameter and physical values involved in the physical models and tools of PVsyst have to be predefined. These variables are initialised with reasonable values, as determined by the author at design time. But

167

Chapter 7 Technical aspects

Technical aspects
most of them may of course be subject to discussion, and therefore can be adjusted by the user, according to his own situation or hypothesis.

This is achieved through the "Preferences" / "Edit Hidden Parameter" option in the main menu. This will open a list of variables, with their effective initial value, modifiable by the user. Changes should of course be made by expert users only, no validity check being performed at this stage of the program. Each value is accompanied by a check box, indicating whether the value was modified, and allowing for retrieving the original PVsyst default value at any time. The variables are divided into the following categories: - Grid-Connected System Pre-sizing, - Stand-alone System Pre-sizing, - Pumping System Pre-sizing, - Specific Pre-sizing costs (all systems), - System design parameters, - Detailed Simulation Verification Conditions, - PV modules, - Regulators and converters, - Miscellaneous, The modified values can be either validated only for the present session, or be permanently saved.

Default values and costs

The pre-sizing default values (system efficiencies and costs) are determined at the creation of the presizing project. Pressing "Edit costs" allows you to modify these default values either for this session or for permanent use. Values of the presently displayed pre-sizing project will be updated after editing "Edit costs". Please be aware that at this pre-design stage costs are very coarse hypothesis. They can widely vary from country to country, from time to time or from user to user (what costs are included here? customer or retailer costs? which interventions on the building? designer fees? taxes? etc). The economic evaluation at the detailed simulation stage will offer a flexible and more precise tool for evaluating real costs according to the specific user's criteria.

Uninstall
The Uninstall tool of Windows is not available for PVsyst. At installation, PVsyst doesn't write anything outside of its own directory \PVsyst5\, nor in the Windows Registry. At first use (from V 5.2) , PVsyst will define a specific area for writing its data: - on Window 7 or Vista: c: \ Programdata \ PVsyst5 \ Data \

- on Windows XP and before: c: \ Documents and Settings \ All Users \ Application Data \ PVsyst \ Data \ When using the program, you may also have copied the whole \Data\ structure at another location of your hard disk. This is especially the case when you are not administrator of your machine. In this case PVsyst has created another directory \PVsyst_Data\ in your working space, as well as a little file named "DataPath.ini" in the \Documents and Settings\All Users\Program Files\ directory of your operating system. Therefore, when unintalling the software, you should:

Chapter 7 Technical aspects

168

Technical aspects
Perform the Transfer of your user's code to another machine rights, Export your old projects described above.
176

7
17 ,

in order to preserve your user

, or eventually save your whole \ Data \ or \ PVsyst_Data \ structure as

After that, you should simply delete the whole \Program Files \ PVsyst5 \ directory using the Windows explorer. Delete your PVsyst5 \ Data \ structure, as described above.

As stated above, there are no additional files nor registry keys left elsewhere on your machine. NB: Uninstalling is not urgent. Several versions of PVsyst may coexist on your machine without interference, as far as they don't share the same \Data\ structure (because some files of new versions may be incompatible with older versions).

File organisation
Installation
The PVSYST software can be installed in any directory of your choice (defined during the installation). This directory (the "root" of PVSYST, say \PVSYST5\ by default) contains the main programme and its executing files. It is usually placed in the \Program Files\ part of your operating system. No other files are put elsewhere in your system by the installation program (except under Vista or Windows 7, see below). PVsyst doesn't make use of the Windows Registry. The various help, image and data files are spread-out into sub-directories (also created during installation) according to the list defined below. Please note: the installation, as well as the first execution of the program, should be performed in a windows session with administrator rights.

New file organization and data positioning since version 5.2

Up to the version 5.13 the working area (`\Data\ directory) was located by default in your installation location, i.e. under \PVsyst5\Data\. But when you don't have administrator rights in your Windows session, Windows forbids writing in the \Program Files\ area. With Windows XP, this was an absolute conditions and when not administrator of your machine, you had to displace (copy) your working area to your personal working area (see below). Now with Vista and Windows 7, when the program is writing in these area, Windows delocalizes these files 174 to a special \VirtualStore\ directory. c:\Users \ User \ AppData \ Local \ Virtual Store \ Program Files \ PVsyst5 \ Data \ This is the reason why from version 5.2, PVsyst puts, by default, the \Data\ subdirectory in a location always writable by anybody: Under Vista and Windows 7: c: \ ProgramData \PVsyst \ Data \ Under Windows XP and older: c: \ Documents and Settings \ All Users \ Application Data \ PVsyst \ Data \

Customizing the location of your data

If you want to put your PVsyst \data\ structure elsewhere in your working area, you can copy the whole \Data\ structure, using the dedicated tool 175 (main menu "Files"/"Directories"). The customized location will take the name \PVsyst_Data\. NB: This opportunity of copying the \Data\ structure is also useful if you want to share your projects with other users on a network. But please note that PVsyst is not really meant for sharing data, and doesn't perform any checks for avoiding user's collisions. In this case: PVsyst doesn't recognize the network paths beginning by "\\". Please make your remote
169

Chapter 7 Technical aspects

Technical aspects
directories as a disk unit.

File organization and description

The Directory tree and file content is organised as follows: \PVsyst5\ \Help\ \Images\ \DataRO\ \Admin\ \Data\ \Sites\
172 172 172

(or any other "root" that you may have chosen during installation) contains the executable programme file PVsyst5.exe PVsyst5.chn Contains the Help file. *.bmp Some images used by the software. *.csv, *.dat Texts, hidden parameter, Original databases of Meteo sites/monthly data, PV modules and Grid inverters. Some basic files, including PVsyst.ini and *.Log files for debugging. All databases and writable user data, includes the following 10 sub-directories: Geographical site and monthly meteo definitions. Hourly meteo and other related files. ASCII meteo source files may be located anywhere in your computer.
173

172

\Meteo\

\Datameas\ \ComposPV\ \Projects\

173

In a similar way as for meteorological files, this directory stores internal files concerning measured data. All PV components of the database (each in a specific subdirectory) All the project elements (files of parameters and results) for your projects, including Preliminary and Project design. All shading definitions (Horizon definitions and shading scenes/Objects) Miscellaneous templates and user defined models. Will contain the user export files for other softwares. Namely the file "Components.xls", which can store some database elements (geoghraphical sites, PV modules or Inverter) data in tabular form, to be used with Microsoft EXCEL. Some special files.

173

\Shadings\ \Models\
174

173

\Userdata\

\Other\

Usually, the user will not have to manage or directly memorize the names of the files: each file in PVSYST includes an explicit header description, which will appear in all the choice-lists in the program, in addition to the file name. You are advised to carefully define these descriptions when available, in a unique manner, by a significant comment about the set of parameters concerned. When the parameters of a component, an element of a project, etc..., are modified, the program will remind the user to save the modifications carried out, either in the same file (button "Save"), or in a new file ("Save as") for creating a new component.

Export/Import of data files

The data structure 169 in PVsyst is well structured, and this structure should of course be carefully respected for proper operation. Several tools help managing (import/export) of external data. In the main menu "Files":

Copy the whole data structure

175 allows to displace your working \Data\ structure anywhere on your system. This is useful especially: - for users who don't have the writing rights in the directory of the main program (\program files\ in old versions of PVsyst). Since the version 5.2 the data are located in an area which should always made writable. - If you want to localize you data in a personal place (for example periodically saved), - If you want to share your \data\ structure with other users (network working). In this case make your

Chapter 7 Technical aspects

170

Technical aspects

network data path as a virtual disk. When choosing an external data location, PVsyst automatically remembers this in the little file DataPath.ini , which is located in the \Documents and Settings\All Users\Application Data\PVsyst\ directory.

Export whole Projects

allows gathering all files involved in a project 176 for exporting to an archiving place or sending to another user of PVsyst. These files are dispatched into an external directory (usually named \PVsyst_ExtData\ as a structure analogous to the one of PVsyst.
176

Import projects

176 restores whole projects either from exported structures, or simply from a set of the files 176 of a project (in a single subdirectory, without structure). This is especially useful for re-importing your old projects from a previous version of PVsyst. The concerned files are dispatched into the local data structure.

Export Database Components copies all or chosen database components created or modified by
the user. These files are dispatched into an external directory (usually named \PVsyst_ExtData\ as a structure analogous to the one of PVsyst. You can chose the desired categories and files to be exported as you like.

Import Database Components dispatchs a set of PVsyst database elements from an external data
structure to each convenient folder in the working \Data\ structure. The source data may be either a set of individual files, or an \Ext_Data\ structure created by the previous tool. There are sophisticated options for replacing only older files, or replacing the files only when minor modifications (date of availability, price) have been modified.

Export Log Files gathers the *.LOG files produced by your PVsyst runnings when they are requested
be the debugging center.

Directories contents
Since version 5.20, PVsyst places by default its \Data\ folder in a zone writable by anybody: - Under Vista and Windows 7; - Under XP and olders: Data \ c: \ ProgramData \PVsyst \ Data \ c: \ Documents and Settings \ All Users \ Application Data \ PVsyst \

and the Log and PVsyst5.Ini files in the parallel folder - Under Vista and Windows 7; - Under XP and olders: Admin \ c: \ ProgramData \PVsyst \ Admin \ c: \ Documents and Settings \ All Users \ Application Data \ PVsyst \

These folders should be writable by the user of PVsyst, from his Windows session. If not, they have to be made writable by the administrator of your machine (in an administrator session). NB: If you don't see these structures in your machine, you have to allow Windows for the display of Hidden files 174 .

171

Chapter 7 Technical aspects

Technical aspects
\DataRO\

This directory (DataRO pour "Read Only") is in the program's structure: c:\Program files\PVsyst5\DataRO\. It contains: Texts_5_xx.CSV Params_5_xx.DAT : Countries_MainBeg.CSV MeteoDB.CSV PVModuleDB.CSV InverterDB.CSV ReadMe.TXT : List of all the texts used in PVsyst for multi-language uses. Miscellaneous "hidden" physical parameters. Values can be edited and modified through the "Preferences"/"Edit hidden parameters" option in the main menu. Official list of the countries Basic database of the geographic sites and their monthly meteo data. Basic database of the PV modules Basic database of the grid inverters Last minute information about the package.

\Admin\

\Data\Admin\ directory
PVsyst5.INI : Initialisation file, containing, among other things: current directories, personalised variables, your user's code, etc... If you encounter any initialisation problem, do not hesitate to destroy this file as it will be automatically recreated by the programme with default values. (do not forget to note your activation code number before ! ) Params_5_xx.DAT : Currency.DAT : Miscellaneous "hidden" physical parameters. This file is originally stored in the \DataRO\ (read only directory) at installation. It is saved here in case of custom modifications. Defined currencies and exchange rates. You can update them in any dialog dealing with economical data.

yyyymmdd_hhmm.LOG Log files for debugging, created when they are activated in the "Preferences" dialog. To be sent to the author in case of problem. The program keeps the last five .Log files produced, and some older ones when an error occurred.

\Data\Sites\

\Data\Sites\ directory
Namesite.SIT : Parameters of geographical sites (latitude, longitude, altitude, time-zone), with their general climatological data in monthly values (global, diffuse irradiation, temperature, wind velocity). NB: The *.sit files include monthly meteo data, and allow to perform some quick but approximate meteo calculations. They may also be the source for generating synthetic hourly data (*.met files), necessary for the detailed simulation. NB: From version 5.0, the original database for sites is stored in the file "MeteoDB.CSV" in the directory \DataRO\. This directory will holde the files that you create (or modify) by yourself

\Data\Meteo\

\Data\Meteo\ directory
Meteo.MET : Internal file of hourly meteorological data, associated with a measurement site, and containing Global, Diffuse, and if necessary ambient temperature and wind-speed, in binary format. A year's complete file occupies about 70 kOctets. Specification of the format of ASCII source files for hourly (or sub-hourly) meteorological data import. They contain all the necessary indications for the recognition and the transformation of data when reading various ASCII formats. The main constraint is that the data of one timestep (hourly or sub-hourly) figure on one ASCII line ending with CR (and/or LF). Tables of transposition factors, calculated for all orientations (used in the orientation optimisation tool).

TypeFmt.MEF :

FTransp.TFT :

Chapter 7 Technical aspects

172

Technical aspects
\Data\DataMeas\

\Data\Datameas\ directory
TypeFmt.DAF : Specification of the format of source files for hourly ASCII measured data. Same characteristics as the *.MEF files for Hourly meteo data, but allows the user to choose the measured variables to be transcribed. They also offer the possibility for automatically chaining data file reading (daily or monthly files). Hourly measured data file. Normalised internal file, made up from your ASCII files using the .DAF interpreter. Even if the source is in sub-hourly steps, the interpreter accumulates the values in hourly steps.

DataMes .DAM :

\Data\ComposPV\

\Data\ComposPV\ directory
Each component type is stored in its own subdirectory. \PVModules\ *.PAN : \Inverters\ *.OND : \Batteries\ *.BTR : \Pumps\ *.PMP : \Regulators\ *.RLT : \RegulPmp\ *.RLP : \Gensets\ *.GEN : \Sellers\ *.REV : Parameters and characteristics of a PV module. Inverter for grid connection: operating limits, efficiency profile. Characteristics of a batteriy. Characteristics of a pump device Charge/discharge regulator, and possible back-up, "ON/OFF" operating , according to threshold levels of battery voltage. Pumping system controller, includes the regulation strategy, may include power converter or battery regulator. Back-up generator (Genset) for the battery-recharge. Addresses of manufacturers or other reseller.

NB: From version 5.0, the original database for PV modules and Inverters is stored in the files "PVModuleDB.CSV" and "InverterDB.CSV" located in the directory \DataRO\. This directory will hold the files that you create (or modify) by yourself

\Data\Projects\

\Data\Projects\ directory
Pre_Grid.PSG : Pre_SAlone.PSS : Preliminary design of grid connected systems. Preliminary design of stand-alone systems.

Pre_Pumping.PSP : Preliminary design of pumping systems. Project.PRJ : Project.VCi : Central definitions of a project for full simulation and/or measured data analysis (includes site, associated meteo file, albedo data, eventually altitude corrections, etc.).

Variant of a simulation version. Includes all specific parameters for one simulation (plane orientation, PV module arrangement, inverters, batteries or pumps, etc, loss factors, shadings, horizon, etc., as well as simulation results. These files have the same name as the project. The last letter of the extension determines the version number (running from 0..9 and A..Z). Variant of a simulation-measured data comparison, similar to the simulation variant, but including more the detailed comparison pairs in hourly or daily values.

Project.CMi :

\Data\Shadings\

\Data\Shadings\ directory
Horizon.HOR : Shading_Scene.SHD : Buildings .BLD : Shading_objects.SHO : Horizon profile definition. Complete shading scene, may also include the Shading Factor Table. Buildings templates created in the near shading scene. Collection of shading objects, which can be saved together using multiple selection.

173

Chapter 7 Technical aspects

Technical aspects
\Data\Models\

\Data\Models\ directory
Load.LOD : Graph.SPL : Expfile.SFI : Load definition or profile, explicitely given as file. On-line graph models which can be used as such, to be called before performing simulations. Format models for the creation of export files for the results, with a view to using them in other software.

\Data\UserData\
This is a place where you can put your own files regarding PVsyst. For example the files generated by the simulation process. NB: If you don't find the files your have generated here, it is possible that Windows has put them in a delocalized place 174 \VirtualStore\.

File delocalization with Vista and Windows 7

Valid only for old PVsyst versions before 5.20 With these new versions of Windows, when the program attempts to write in the \Program files\ subdirectory, (or \Program files (i86)\ for 32 bits programs in 64 bits OS), these new versions of Windows automatically redirect the writing to a virtual location situated at: c:\Users \ User \ AppData \ Local \ Virtual Store \ Program Files \ PVsyst5 \ Data \ Therefore all data that you have created by yourself (or modified) under PVsyst versions before 5.13 will be stored in this directory. You will also find here the LOG files or the PVsyst5.ini file. This directory should be destroyed when completely uninstalling the program. NB: in the Windows File Explorer, some of these directories are translated into the local language, for example in French: "Utilisateur" instead of "User". NB: The AppData subdirectory is often hidden. For visualizing hidden files or directories in the File Explorer, you should modify the parameters of this tool. For this: - Choose option "Organize" left under the menu), - Choose "Options of Directories and Search" / "Display" / "Advanced parameters ", and in the box "Hidden files and directories", check "Show the hidden files, directories and disk s ". NB: Since version 5.2, the working \data\ directory has been displaced in the c:\ProgramData\PVsyst\ directory, avoiding this problem. This is the new location where you should look for your files.

Seing "hidden" files and directories in Windows Explorer

Often some files or directories are not visible in the Windows file explorer. For visualizing all existing files, you should modify the parameters of the Windows file explorer: - Under Windows 7 and Vista: click "Organize" top left of the menu, choose "Options of the files and research". - Under Windows XP, this tool is in the menu "Tools " / "Directories options " on the right. In this windows tool, please choose "Display" / "Advanced parameters ", and in the box "Hidden files and directories", check "Show the hidden files, directories and disk s ".

Chapter 7 Technical aspects

174

Technical aspects
Log Files

The Log file records the main actions of the program, and some error messages. It is stored in the \Admin\ directory. PVsyst keeps the 5 more recent log files, and some special ones when errors occurred. Windows will write these file in the following directory:
- on Window 7 or Vista: c: \ Programdata \ PVsyst5 \ Admin \ - on Windows XP and older: c: \ Documents and Settings \ All Users \ Application Data \ PVsyst \ Admin \

where some subdirectory names (i.e. Users and User) may be translated into your language in the File explorer. Log files are intended for debugging only. In case of persistent error, please enable the generation of Log files (default option in the "Preferences") and send them to the author on request. You can retrieve the Log files (for example when asked by the debugging center) in the main menu "Files" / "Export Log Files".

Copy the data structure

This will copy the whole \Data\ structure (from the \PVsyts4\Data\ location) into the chosen location. The target location should be in a zone where you have writing rights (typically in your \My Documents\ area) If the target structure already exists, only newer files are copied. The target structure in renamed \PVsyst_Data\ for easier identification in your machine. This new location, associated with your own user Windows session, will be referenced in the little file DataPath.ini, which is located in the c:\Documents and Settings\All Users\Application Data\PVsyst\ directory. Importing/Dispatching data files
This tool is meant for importing individual PVsyst data files and dispatching them into their subdirectories. For the transfer of whole projects please choose "Import Projects". Please choose the "Source" directory using the "Browse" button. The source may be one subdirectory, or a root with several subdirectories; all PVsyst files in any of these subdirectories will be shown. The Target is of course the PVsyst data structure. Colours underline the status of the concerned files. If some file already exists in the PVsyst data, it also appears in the right panel, with specific colour: - Green: the file is new, doesn't exist in the PVsyst database. - Blue: the corresponding file in the PVsyst data is from original database (identified by a file date equal to 12:00), but the file to be imported is newer: should be copied. - Red: the file in the PVsyst data is from original database, and the corresponding file to be imported is an older one: in principle should not be copied. - Orange: the file in the PVsyst data is not from the original database: it has been modified by the user (for example defined component's prices). In the latter case: One of the main reasons of importing updated database files is when the availability of the component has changed (i.e. is no more available on the market). Therefore PVsyst gives the opportunity of performing an "intelligent" copy, which consists of maintaining your modifications and just updating the market availability dates in the component. You have choice between this intelligent copy (recommended for database update) or a normal copy which will overwrite your modifications. Taking care of these status, you can choose the files which you effectively want to transfer. Or ask for copying all the eligible files according to the above criteria. After transfer, the source files become grey (or disappear if "moved"), and the copied files are blue.

175

Chapter 7 Technical aspects

Technical aspects
Invalidating files
A bad file can disturb the normal running of the program and produce unrecoverable errors. Invalidating it puts a "$" as first file name's character. You will always be able to rename it when necessary.

Exporting projects
This tool gathers all files involved in a simulation project for export to an archiving place or sending to another PVsyst user. The source is of course the local data structure. It shows the Simulation projects (in green) and the Presizing projects (in maroon). You choose the desired projects to be exported. You have to choose the target directory (anywhere in a writable area on your machine), which could be named \PVsyst_ExtData\ by default. If not existing, a data structure analogous to the structure of PVsyst (i.e. 3 subdirectories \Projects\, \Meteo\ and \Components\) will be created. When clicking on the transfer button, the concerned files are dispatched to their corresponding subdirectories. The copied structure is reversible and could be re-imported "as such" by another installation of PVsyst (of same or posterior version). It may also be zipped for sending to another PVsyst user.

Importing projects
This tool dispatches all files related to a project to their proper directories. Please define the source directory, by choosing a project file among your external data. When available the program will show all transferable projects (green: simulation, maroon: pre-sizing projects). The external Project's files source may be either a structure analogous to the PVsyst data, or a "flat" list of all involved files in the same directory. Only the projects are shown here, not the associated files. NB: the external source directory may be the data structure of your old version of PVsyst, when Upgrading to a new version. The target is of course your present PVsyst data structure. Please select the projects you want to transfer. When clicking on the transfer button, the concerned files are dispatched to their corresponding subdirectories. Only the not-existing files are copied (i.e. the original files of the database are preserved). If you want to overwrite some files in your local data you should use the "Dispatch Files" option.

Files involved in a project A project (design project) involves several files, which should be placed in a well-defined directory structure. They are: in \projects \: ProjectName.PRJ, Variants ProjectName.VCi, Eventually comparison variants ProjectName.Cmi. in \meteo\: MeteoFile.MET, the hourly meteo file associated with the project Datameas.DAM, (eventually if comparisons with measured data), in \DataMeas\:

in \ComposPV\: Components files, each kind in its own subdirectory: \PVModules\ PVModule.PAN, \Inverters\ Inverter.OND, \Batteries\ Battery.BTR, \Regulators\ Regulator.RLT, \Pumps\ Pump.PMP, \RegulPmp\ PumpRegul.RLP.

The "Import" or "Export" project options helps transferring whole project structures, either from the local data or from remote data structure, ensuring that all needed files for a given project are present. If the target is not structured this tool creates the necessary directories, with a common root usually named \PVsyst_ExtData\ Note: Preliminary design projects are made of only one file, with extension .PSG (grid) or .PSS (standChapter 7 Technical aspects
176

Technical aspects
alone).

Printing
PVSYST offers the opportunity for printing several types of data: Preliminary design: all parameters and results are summarized on one form, Project design: one form for complete parameters, one for a results summary, and other ones for additional results when necessary. Any further specific table or graph can be printed independently. Tables and graphs: any table or graph output produced by PVSYST can be printed, or exported through clipboard or file. "Scrolling" tables (like daily or hourly meteo values, solar geometry) can be printed for selected periods in suitable format. Database components: Meteorological sites, or PV components (PV modules, Batteries, Inverters, Pumps, Regulators, etc.) can be output as "intelligent" forms including all relevant parameters.

All printer forms include one or two Head lines 177 which can be customized (including your Company name) and one or two Comment lines which can always be specified at the time of output Whenever possible in general table or graph outputs, the program produces an "intelligent" form which includes several parameters or references related to the printed table or graph. For example, solar geometry tables will include the geographical location parameters, as well as plane orientation or shed arrangement when applicable. Printer dialog: Change/Configuration: allows the definition of the properties of the printer, or to change the printer. Head comments: define up to two comment lines at the head of the form. These can appear bold if desired. - Options: Definition of the head of the form, and a few other formatting options; these definitions can be made either for the present form only, or stored as "Preference" for any future output. - Preview: performs a preview of the form(s) to be printed. - Print: sends the output form to the printer. - Copy to Clipboard: allows for copying the whole printed form as an image in the clipboard, for immediate paste in another software (like MS-Word, MS-Excel, Paint, or others). This allows for storing the whole printer page in documents, or sending results for example by e-mail. Please refer to "Options" 177 for format details. If several pages are to be sent, please perform the "Paste" operation immediately after the message for each page, before clicking the "OK" button of the message. NB: This output feature is different from the "Export"/"Copy to clipboard" tool present in all general table and graph menus, where only the displayed table or graph is exported. In several cases there are a few additional options when printing, for instance, the desired data periods. Also, if several items or pages are available for an output, specific parts may be selected for printing.

Print_Head
The printed forms always include a frame, with the "PVsyst V4.xx" information in the left corner. Printing date in the right corner is optional, with or without hour. The central first line is free for the general label of your choice (company, etc.). You can specify a second head line if desired (for example for your address/telephone). These customised specifications may be done either in the "Print"/"Options" dialog, or through the "Preferences" in the main menu.

Copying Printer pages to Clipboard

Copying whole printer pages is a convenient way to store final PVSYST results in documents, or to send them for example through e-mail. Using the "Copy to Clipb oard" button in the "Printer" dialog will store the whole printed page as an image. You 177

Chapter 7 Technical aspects

Technical aspects
just have to enter another software (for example MS-Word, Paint, etc) and paste the image.

Please note that with MS-Word, you are advised to use "Paste special" and then uncheck the option "Dissociate from text", as image frames are very difficult to manage in this software. Memory: resolution options But be aware that such an image of an entire page takes up quite a lot of memory: That is why the "Options" provide the opportunity of choosing the resolution. Please note that: - 200 DPI will give good outputs when printing, but consumes about 3.2 MB of memory. - 100 DPI will give coarse outputs, but occupies "only" 790 kB for each page. - 150 DPI give almost as good results as 200 DPI, but for 1.8 MB. After pasting in MS-Word, this image is compressed, and the resulting MS-Word .DOC file will be about 60 to 70 kB per PVsyst page image at 200 DPI. Color options The image may be stored either in Colors, or in Black and White. Of course, most of the external software accept images in color. But the different colors in graphs will not be reproduced on Black and White printers, nor will the colors be distinguished. Therefore when printing, PVSYST looks for the actual printer capabilities, and modifies the graphs accordingly: with color printers it keeps the colors displayed on the screen. With B&W printers it will adapt the graph lines (using variable line widths and dotted lines) or the bars fillings with grey levels. That is the reason why the final graph destination has to be precised at the time when PVSYST b uilds the image. The options give the opportunity of choosing Color, B&W, or leave PVSYST choose itself according to the actually connected printer. Main physical models used in PVSYST calculations : Meteorology: Irradiation computations PV system Components: PV modules Inverters Batteries Pumps
156 150 154 159 145 178

Regulators

PV system design: Grid system presizing

18 19

Stand-alone system pre-sizing Pumping system pre-sizing Simulation process Others: Pumping: deep well
68 92 67

Incident irradiation models

Irradiation on the PV-field We call " effective incident irradiation" Heff the luminous energy actually falling on the PV cells. It is obtained according to the following steps: If only monthly meteorological data available: Generation of hourly synthetic
181

meteo data (horizontal global

Chapter 7 Technical aspects

178

Physical models used

irradiance and temperature), If diffuse irradiance measured data not available: diffuse irradiance model
180

8
,

If horizon 38 (far shadings): calculation of the beam effective component (in this version of the program, the diffuse is considered as not affected by horizon).

At this stage, we have the Horizontal global, diffuse and beam components at disposal, with the relation: Gh = Dh + Bh. Computation of the so-called "incident energy" by a Transposition model on the PV tilted plane.
179

, i.e. calculation of the irradiance

At this stage, the plane irradiance is composed of global, diffuse, beam and albedo components, with the relation: Gp = Dp + Bp + Ap. Applying the near shading 44 calculations (shading factor on beam, diffuse and albedo components), either linear or according to electrical array connexions, Applying the IAM
86

(Incidence Angle Modifier factor),

this finally results in the Geff irradiance, the flux effectively useable for PV conversion. Heff will be the corresponding irradiation over a given time period. Note: we usually use G for designing irradiances (flux expressed in [W/m]) and H for irradiations (energies in [kWh/m].

Meteo Monthly calculations

This simplified computation performs quick meteo evaluations, using geographical site database only (i.e. monthly irradiation values), and evaluates horizon, tilt, sheds and sun-shields, as well IAM effects. This method takes advantage of so-called "average months" properties. With real meteo data of a given month, when constructing an "average day" by averaging separately the irradiances at each hour (i.e. all irradiances at 8:00, 9:00, 10:00, etc), we obtain an average profile which is very close to the "Clear day" model profile, with amplitude reduced in such a way that the day integral matches the monthly global value. This also holds for the diffuse component. Therefore, inversely, for the middle of each month, we construct a "clear day" with amplitude suited to the given monthly irradiation. We then assume that this "average day" is representative of the month, so that we can apply all mentioned corrections (transposition, shadings, etc) using the solar geometry of this middle-month day.

Accuracy
This procedure avoids constructing synthetic hourly values, and gives instantaneous evaluations with very acceptable accuracy. By respect to an hourly computation, monthly calculations for Geneva show that the transposition on tilted plane induces a yearly MBE of the order of: South plane, tilt SE or SW planes: E or W planes: tilt 0..90: tilt 45: < 1.3% 45: < 1.1%, tilt 90 : < 4.1%, < 2.7%, tilt 90 : < 11 %,

If only the global monthly values are known, the uncertainty on the monthly diffuse estimation correlation model (about 5%) may induce 1 to 4 % error more.

Transposition model
Transposition is the calculation of the incident irradiance on a tilted plane, from the horizontal irradiance data. PVSYST offers two transposition models: Hay's model, a classic and robust model which gives good results even when the knowledge of the diffuse irradiation is not perfect, Perez model (Perez, Ineichen et al.), is a more sophisticated model requiring good (well measured) horizontal data. The beam component involves a purely geometrical transformation (cosine effect), which does'nt involve any physical assumption.

Transposition is separately calculated for each irradiance component: -

179

Chapter 8 Physical models used

Physical models used

The two models differ by the diffuse component treatment:

In the Hay model, the diffuse irradiance is divided into an isotropic contribution, and a "circumsolar" proportional to the beam component. Through transposition, the half-sphere isotropic part is reduced according to the vault of heaven's solid angle "seen" by the plane (i.e; the fraction (1+cosi)/2, where i is the tilt angle). The circumsolar is transposed geometrically as the beam component. The specifity of the Hay model is the determination of the circumsolar fraction, which is chosen as the Clearness index Kb of the beam component. The Perez-Ineichen model introduces the "horizon b and" as a third diffuse component. It divides the sky into sectors, and parametrises the transformations of the circumsolar and the horizon band according to correlations established on the basis of data of several dozen of measurement sites, distributed all over the world. The albedo component is evaluated in the same manner in both models, as a given fraction (the "albedo coefficient 199 ") of the global, weighted by the "orange slice" fraction defined between the horizontal and the tilted plane extension (i.e. the half sphere complement of the "seen" vault of heaven), which is the fraction (1cosi)/2 of the half-sphere.

Validations During validations 208 of the software, we tested these two models with the data of each site. The comparison of their mean errors (MBE) presents a systematic difference of 1.8 to 2.2% depending on the (Swiss) sites, while the standard deviations RMSE are comparable in all cases. It appears therefore that the Perez's model, which is more complex and especially more sensitive to a realistic determination of the diffuse irradiation, is often not justified in the PVSYST software. Therefore, by default, the PVSYST programme uses the Hay's model. However, if the user can avail of good measurements of the diffuse irradiance, he can choose the most sophisticated model of Perez-Ineichen (main menu: option " Preferences " / " Preferences ").

The Hay transposition model

The Hay transposition model applies differently to the different components of the irradiance. The Beam component results of a pure geometrical transformation: BeamInc = BeamHor * sin Hsoli / sin Hsol The Diffuse component is supposed to be mainly constituted of an isotropic distribution, and a circum-solar contribution proportional to Kb DiffInc = DiffHor * [ (1-Kb) * (1 + cos i) / 2 + Kb * sin HsolI / sin Hsol ] The Albedo component is the irradiance reflected by the ground "seen" by the plane : AlbInc = * GlobHor * (1 - cos i) / 2

where i = Plane tilt Hsol = Sun height on horizontal plane Hsoli = Sun height on the plane (= 90 - incidence angle) Kb = Clearness index of beam = BeamHor / (Io * Sin Hsol) Io = Solar constant (depends on the day of year) = Albedo coefficient 199 (usual value 0.2)
The expression (1 + cos i) / 2 is the mathematical result of the spherical integral of a constant irradiance, from all directions "seen" by the plane (i.e. the orange slice between the plane and the horizontal).

Diffuse Irradiance model

When it is not explicitly measured, the diffuse irradiation should be estimated from horizontal global by a model. Chapter 8 Physical models used
180

Physical models used

We can mention two wide-used such models: - Liu and Jordan's correlation, which results from an experimental correlation of the D/G ratio by respect to the clearness index Kt 201 . - Perez model (Perez, Ineichen et al.), is a more sophisticated model taking hourly data sequence into account. In PVSYST the diffuse irradiation uses the "robust" Liu and Jordan's correlation. When applied to our data from the SIG (the only simultaneous measurement of global and diffuse irradiations we have at our disposal), this correlation gives good results with an MBE of 1.7% and a RMSE of 27% (with respect to the value of the diffuse irradiance), or 13% (referred to global irradiance). Note: The most sophisticated model of Perez-Ineichen has also been tested, but does not give significantly better results. From the opinion of one of the authors (P. Ineichen in our laboratory, who has also evaluated the Liu-Jordan correlation) this is especially suited for very well-measured data. Applying it to synthetic hourly data doesn't make great sense. This is the reason why we didn't implement it in the current version. Nevertheless we intend to offer it as an option it in a further version.

Synthetic data generation

Synthetic data generation provides a mean of constructing meteorological hourly data from only monthly known values. This is required since numerous simulation processes have to be computed as instantaneous values (or pseudo-instantaneous as hourly averages). This is the case, for example, with the transposition model which closely depends on the solar geometry. Irradiance generation For global irradiance, we dispose of well-established random algorithms (Aguiar et al), which produce hourly distributions presenting statistical properties very close to real data. The algorithm first constructs a random sequence of daily values, using a Library of Markov Transition Matrices (probability matrices) constructed from real meteo hourly data of several dozen of stations all over the world. Then it applies a time-dependent, Autoregressive, Gaussian Model for Generating the hourly sequences for each day. Temperature generation For temperature, such a general model doesn't exist. We used procedures adjusted only on Swiss meteo data (Scartezzini et al.), for which generalisation to any world climate is not proved. In fact the ambient temperature daily sequence shows only weak correlations to global irradiation. Of course the temperature should be continuous, therefore this sequence is constructed using essentially randomly daily slopes , with constraints on the monthly average. But daily profile can be much more related to the global irradiance. During the day, temperature behaves rather like a sinusod, with amplitude related to the global daily irradiance, and a phase shift of two to three hours. The corresponding correlation parameters (for amplitude and phase shift) have been quantified from several Swiss region typologies. One can accept that these can be generalised to analogous typologies for other places in the world. Notes: The region typology asked by the program is only used to refine these temperature daily profile parameters. The dependence of PV-system behaviour is not very temperature-sensitive (about 0.4%/C). Provided that the monthly average is correct, the global results of the PV production will not strongly depend on the temperature daily sequence. Please be aware that these temperature synthetic data are not reliable enough to be suited for building heating or cooling studies in any climate (which is not the goal of this software) !

181

Chapter 8 Physical models used

Physical models used

PV Module - model description

To describe the operating of a PV module, we use the Shockley's simple "one diode" model (primarily designed for a single cell), described, for example, in Beckman and al. This model is based on the following equivalent circuit for decribing a PV cvell:
I Rs
Photocourant Diode Utilisateur

Iph

Rsh

RL

The model was primarily developed for a single cell. Its generalization to the whole module implies that all cells are considered as rigorously identical. A more sophisticated model, implying 2 different diodes, is sometimes proposed for the very accurate modelling of a single cell. But in PVsyst, we think that small discrepancies in the cell parameters, inducing internal mismatch, as well as the moderate accuracy of our basic input parameters (usually from manufacturer), make no sense to use it. In the one-diode model the two diodes are considered identical, and the Gamma factor - ranging theoretically from 1 to 2 - defines the mix between them. This model is well-suited for the description of the Si-crystalline modules, but needs some adaptations for reproducing the thin film technologie module behaviour 186 . We observed that the CIS technology obeys quite well to this standard model. The main expression describing the general "one-diode" model is written as: I = Iph - Io [ exp (q (V+IRs) / ( NcsGammakTc) ) - 1 ] - (V + IRs) / Rsh with : I V Iph = = = Current supplied by the module [A]. Voltage at the terminals of the module [V]. Photocurrent [A], proportional to the irradiance G, with a correction as function of Tc Diode current, is the product Io [exp( Series resistance [ohm]. Shunt resistance [ohm]. ) -1].

(see below). ID = Io = Rs Rsh = =

inverse saturation current, depending on the temperature [A] (see expression below).

q = Charge of the electron = 1.602E-19 Coulomb k = Bolzmann's constant = 1.381 E-23 J/K. Gamma = Diode quality factor, normally between 1 and 2 Ncs = Number of cells in series. Tc = Effective temperature of the cells [Kelvin]

The photocurrent varies with irradiance and temperature: it will be determined with respect to the values given for reference conditions (Gref, Tref): Chapter 8 Physical models used
182

Physical models used

Iph = ( G / Gref ) [ Iph ref + muISC where G and Gref Tc and Tc ref muISC = = = (Tc - Tc ref) ]

effective and reference irradiance [W/m]. effective and reference cell's temperature [K].

temperature coefficient of the photocurrent (or short-circuit current).

The diode's reverse saturation current is supposed to vary with the temperature according to the expression: Io = Io ref where ( Tc / Tc ref )3 exp [ ( q EGap / Gamma k) ( 1 / Tc ref - 1 / Tc ) ]

EGap = Gap's energy of the material (1.12 eV for cristalline Si, 1.03 eV for CIS, 1.7 eV for amorphous silicon, 1.5 eV for CdTe).

Determination of the model parameters

Thus for a given temperature and irradiance, we have a model based on 5 unknown parameters (Rsh, Rsh, Iph, Io ref and Gamma,). The value of the shunt resistance Rsh, representing the inverse of the slope of the plateau I(V) for low V, will be independently treated (that is, fixed in the equations). As it is not possible to deduce it from the manufacturer's datasheet, PVsyst has to choose a default value 185 for processing the Database. In order to determine the 4 remaining parameters, we can write the 4 equations below, for the specified reference conditions Gref and Tc ref : I(V) at point V=0 (short circuit), I(V) at point I=0 (open circuit), I(V) at any other point, close to the maximum power point. The derivative muVco = dVco / dTc .

These equations are based on the following parameters (manufacturer's or measured data): ISCref = Short-circuit current at reference conditions. Vco ref = Open circuit voltage at reference conditions. Impref, Vmpref = Current / Voltage at any point (close to the MPP). muIsc = temperature coefficient of the short-circuit current. muVco = temperature coefficient of the open circuit voltage. Unfortunately, the resolution of this system sometimes leads to a solution whose values (Iph, IoRef, Gamma, Rs ) do not have a coherent physical meaning (Rs negative, Gamma out of its physical validity domain between 1 and 2, ...).

Model parameters in PVsyst

This is the reason why we preferred, in the programme PVsyst, to use only the first 3 equations and to propose to give the user, with the help of a graphical representation of various solutions, the explicit choice of the resistance Rs 149 , which will completely fix the solution. The last equation gives an explicit value for muVco. So, if he knows the value of muVco (often specified by the manufacturer), the user may chose the Rs value suitable for obtaining this required value (if possible). But for the automatic generation of the model (used for the whole database), we found more reliable to set the Gamma value at a reasonable level (i.e. Gamma = 1.35 for Si-crystalline), which determines the Rs and
183

Chapter 8 Physical models used

Physical models used

all other values.

NB: Model parameters with measured I/V characteristics

When we avail of one measurement of the complete I/V characteristics - not the case in PVsyst - the model parameters may be determined without ambiguity: Rsh is simply fitted on the data around V=0, while Rs is determined by minimizing the errors between the measured and the modelled values (one only free parameter).

Reverse part of the Characteristics

Some tools in PVsyst - namely in the study of Shading or Mismatch in arrays - require the knowing of the reverse characteristics, when a negative voltage is applied to the module (region V < 0). This model is not as well determined in PVsyst as the direct characteristics (see Reverse characteristics modelling 185 ), but this description is sufficient for understanding the behaviour figures in the involved tools. It is not used in the hourly simulation process.

Phovoltaic modules - Model parameters

Besides the basic electrical Isc, Voc, Impp and Vmpp data, the PV-module One-diode model 182 implemented in PVSYST, involves the following additional parameters, which usually are not given in data sheets: - Module shunt resistance Module_ShuntResistance Rsh, - Module series resistance Module_SeriesResistance Rs - Io ref, inverse saturation current of the diode model, - Gamma factor, "diode quality Factor", It also implies that you have correctly defined the number of cells in series, as the model basically applies to a single cell, which is "multiplied" according to the module wiring structure.

Model determination
According to a method proposed by Beckman and al., the known parameters (measured under given reference conditions, usually STC 206 ) allow to construct a system of 4 equations: The values of the I/V characteristics in 3 distinct operating points (Isc , 0), (0, Voc ) and (Impp, Vmpp), A fourth equation is provided by the temperature behaviour of the voltage Voc ,, ie., muVoc [mV/C], which can be measured relatively easily. with the four unknowns Rs, Io ref, Gamma and muVco. This is illustrated by the I/V curve shown on the screen: it has to pass through the 3 reference points mentioned above. The fourth equation should determine which curve is to be chosen between the two displayed limits, and therefore give the values for the 4 bound unknowns. But in practice the resolution of this system often leads to a set of not physical, or even very erroneous values for these parameters.

Alternative method
So that we preferred to offer the user a more "visual" method, allowing him to choose a parametrisation which corresponds better to the supposed physical reality. The programme traces the I/V characteristics corresponding to the 3 first equations (that is, passing through the 3 experimental points), for different values of Rs comprised between 0 and the maximum Rsmax value compatible with these 3 points. Then, by default, the program automatically chooses the Rs value corresponding to a reasonable Gamma factor 149 . Please note that the gamma factor, related to the dominant modes of recombination, should have a physical value between 1 and 2. Fixing it's value is a priori arbitrary, but it has been chosen as it seems to be the more stable of our four unknown parameters among several PV-modules. By the way it's exact value gives rise to very little differences on the final I/V curve behaviour. This way has the advantage of fixing automatically all the model parameters without any intervention of the user. Nevertheless, you have the opportunity of choosing any other Rs value within the authorised limits, for Chapter 8 Physical models used
184

Physical models used

example in order to better match your temperature factor muVoc if it is known. See also the PV module parameter summary
147

8
for a complete description of all parameters.

Temperature behaviour correction

In the standard one-diode model 182 , the temperature behaviour is essentially determined by the expression of diode saturation current Io, which is exponentially dependent on the Egap and the Gamma parameters. This expression fixes the muVco and muPmax temperature coefficient values, which are therefore a result of the model. Remember that the Gamma value is determined (supposed) in correlation with the Rserie choice, which can vary from 0 to Rsmax. But some manufacturers specify temperature coefficients muVco, which are not always compatible with this Rserie allowed range. It is the reason why PVsyst allows for modifying the temperature behaviour by introducing a linear variation of Gamma with operating Temperature: Gamma = Gamma0 + muGamma * (Tmod - Tref) The user specifies a required muPmpp, and PVsyst determines the suited "mugamma" correction factor: The tool shows a graph of the induced variations on muVoc, muVmpp and muPmpp. NB: This tool is made available for all modules, but it is not recommended (or only as a weak correction). Especially with the amorphous technology, the temperature coefficient is very sensitive to the D2MuTau parameter (see the specific sheet "Recombination loss"). According to our detailed measurements, cumulating this correction with the recombination leads to degraded performances of the model (that is, its ability to reproduce the data in any conditions).

PV Module shunt resistance

The shunt resistance Rsh is the inverse of the slope of the I/V characteristic in the neighbourhood of the shortcircuit point. When a measured I/V curve is available, it is easy to find it by fitting the data around V=0. But it is not possible to extract it from the usual manufacturer's data. Therefore PVsyst has to choose a default value. This is determined by calculating the virtual MPP conductance ( (Isc-Imp)/Vmp ), corresponding to the absolute minimum value for Rsh, and taking a given fraction if this quantity. With Si-crystalline cells (as well as CIS), the Rsh value is high, and its exact value has rather little effect on the general module behaviour (low shunt current loss). But with amorphous or other thin film technologies, it becomes significant, and should be ideally be specified by the manufacturer in the future. For a very accurate modelling of all modules, and especially for amorphous, the Rsh should be corrected according to the Irradiance 191 .

Reverse characteristic of a cell

Reverse Characteristics of a cell (i.e current behaviour when a reverse voltage is applied on it) is involved in all situations where the currents are not well balanced in a module array. This is the case namely in "mismatch" situations (of cells in a module, or modules in an array), partial shadings, or heterogeneous arrays (with different orientations, i.e different irradiances). Severe consequences of the Reverse Bias in arrays can result in so-called "hot spot" phenomena. These are the overheating of unbalanced (bad or shaded) cells, which can lead to their destruction. Bypass diodes mounted in the PV modules are intended to protect them against these dangers. PVsyst offers a specific tool 186 for visualising and understanding these special array behaviours. But they are not implied in the simulation process of PVsyst, which doesn't calculate the electrical array behaviour in detail at each hour. Therefore the reverse bias model exact determination is not crucial in PVsyst, as it is only used in the phenomenological array behaviour tools. Empirically, the behaviour of the cell's characteristic under reverse polarisation is quadratic with the applied voltage. This result comes from our own measurements, and is confirmed in Roger and al

185

Chapter 8 Physical models used

Physical models used

IRev = Iph + bRev (V + Rs I) for V < - Rs I

This expression could be valid till the avalanche zone (Zener), situated around V= -30V. But in reality, under irradiation (high photocurrent Iph), the dissipation, which varies with the cube of the reverse voltage, reaches a destructive limit well before this elbow. For example, the cells constituting the Arco M55 modules, dissipate about 18 W at a reverse voltage of -18V, and 25W at -20V, corresponding to a rise in temperature of the order of 100C. This is even more dangerous as the temperature's rise sharply increases the parameter bRev , and therefore the reverse current, leading to an unstable situation.

PV Module reverse behaviour

This tool presents the operating of a PV module when it is polarised towards negative voltages, as it can happen in the framework of an array or a module, when the cells are different, or if the irradiation is not homogeneous. The tool show three typical situations: "One single PV cell": we see that under bright irradiance, reverse bias of the cell rapidly involves high powers to be dissipated, as the current is already at least at the cell Photocurrent level. (See the reverse cell characteristics model 185 ). The relevant coefficient brev, defined along with the other parameters of the PV module in the "Components" part, can vary strongly from one cell to the other, and is highly dependent on temperature. But this behaviour is not critical in the qualitative evaluations of the PV array behaviour tools. If it is not precisely known, we can usually use the default value proposed by the programme. "PV module without by-pass diode": the solid line represents the whole module characteristics, that is, all cells in series. With identical cells, the total dissipated power is equally distributed on every cell. The blue dotted line shows an arbitrary operating reverse current (identical current in all cells in series). If one cell has a lower photocurrent - due to its quality or shading or a better BRev (flatter curve), then its voltage will follow its own reverse curve, and at the imposed common current it can produce a much higher power than for the other cells, therefore giving a rise of temperature (see the "Hot spot" phenomenon in the tool for one shaded cell 37 ). "PV module with by-pass diode": shows the resulting module characteristics when the module is protected with one or more by-pass diode, mounted in reverse bias. In this case the whole module reverse voltage is limited to the "forward" voltage of the diode (about 0.7V for one diode, 1.4V for 2 diodes, etc). The excess current is drawn by the diode, and cannot give rise to excess powers in the diode since the voltage stays very low.

Model for Thin film and other new technologies

Commercially technologies available on the market are now mainly a-Si:H (amorphous, including tandem and triple junction), CiS and CdTe modules. There is no consensus up to now in the PV community about the general modeling of these new modules. Several experimental works have observed significantly different behaviours of amorphous by respect to standard Crystalline cells. Mertens & al propose taking recombinations in the i-layer into account , resulting in a modification of the equivalent circuit, and a related modified "One-diode" analytical expression. Gottschalg et al 1998, Holley et al 2000 and Mertens et al 2000 report experimental analysis of model parameter dependencies as function of temperature and spectral irradiance contents in amorphous simple and double junctions. Betts &al have studied the spectral contents if the irradiance according to weather in central UK (Loughborough) and propose a correlation for correcting the response of amorphous modules. Modelling in PVsyst: Research project at CUEPE In order to clarify these observations and to establish an approached model useable in PVsyst - including the necessary proposal of default parameters - we have performed an experimental research at the University of Geneva, with the financial support of the SIG-NER fund (SIG - Services Industriels de Genve - is the Electricity and Gas Utility of Geneva). Details of this project may be found in the Final Report of this project, unfortunately only available in French at the moment. The study is based on detailed outdoor I/V measurements of 6 PV modules, every 10 minutes over a period of 3 months. This yields a data sample covering all environmental conditions, with irradiances ranging from 40 to 1000 W/m and temperatures between 0 and 70C.

Chapter 8 Physical models used

186

Physical models used

Methodology

We followed a phenomenological approach, by closely comparing the measured set of data with the model results. We used 3 indicators, easily identified on each I/V measured characteristics: the Pmax (MPP power), Voc and Isc values. All our modeling attempts start from the standard one-diode model 182 . The model main parameters are determined from one chosen I/V characteristics among the measured data. Knowing one complete I/V characteristics, the model parameters (Iph, Ioref, Gamma, Rs, Rsh) may all be determined accurately. With this set of parameters, we can now draw the error distributions for our 3 indicators. We assess that the precision of the model is mainly represented by the RMSE (named sigma) of these error distributions, i.e; the model response dispersion over all operating conditions. Around the optimum, the MBE (Mean Bias Error) values, noted mu) are rather related to the basic input parameter uncertainties (Vco, Isc, Vmp, Imp) at the reference conditions Gref and Tref (i.e. the position of the reference I/V characteristics inside the distribution). Standard model validation on Mono-Crystallie module One of our measured modules was a Siemens M55 monocrystalline We used it mainly as a calibration of our method and experimental set-up, and especially as a reference for the study of the spectral effects. As a by-product, this yields a validation of the standard model for this technology. Surprisingly, the pure standard model did not reproduce well our measurements. The model strongly underestimates the PMax data at low irradiances, giving a MBE of 6% and a RMSE of 3.9%. This could be very well corrected by applying an exponential correction for the Rshunt, as for the amorphous modules: with an appropriate Rsho value, the MBE on Pmax reduced to 1.1% and the RMSE to 2.3%. This correction also strongly improves the Voc behaviour, the RMSE passing from 6.0% to 1.0%. We should emphasize that our test module is very old (15 years), and presents significant corrosion around the collector grid. We cannot say to what extent this Rsh behaviour could be related to these module deficiencies. Nevertheless we strongly feel that the Rshunt exponential correction is a general feature which should be applied to any PV module for proper modeling. But the required parameter Rsho (and even the basic Rshunt value at reference conditions!) cannot be determined from manufacturer data sheets. Standard model applied to the CIS module With an analogous correction on Rshunt, our CIS module (Shell ST40) behaves quasi perfectly according to the standard model. The error distributions are far better than with our mono-crystalline module. With an optimal Rsho parameter, we observe on PMax MBE = 0.0% and RSME = 2.1%; and on Voc, MBE = 0.0% and RSME = 0.5% ! Measurements on a-Si:H triple-junction Our 4 other modules were amorphous: 2 identical "tiles" Unisolar SHR-17 with triple junction, a Solarex MST43MV (not yet produced) and a RWE Asiopak 30SG, both with tandem cells. One of the SHR-17 had already been tested during the whole summer in 2001, when the second one had never been exposed to the sun. The parameter and performances of these two modules were quasi identical, and very surprisingly we didn't observe any initial degradation on the "new" module! Our first observation was that for any I/V characteristics, it is always possible to find a set of the standard model parameter (Iph, Ioref , Gamma, Rs, Rsh) which exactly matches the I/V curve. The RSME between the 30 measured current points and the model is always lower than 0.3 - 0.4% of the ISC (i.e. 0.5 mA at 40 W/m, 4 mA at 800 W/ m). This means that the electrical (diode) behaviour of the amorphous junctions is quite similar to the crystalline modules. The problem is now to determine these parameter dependencies according to irradiation and temperature. For correcting the standard model results (which underestimate the power by about 8% to 10% when applied over all our 3-months measurements) we identified 3 dominant corrections: RSh exponential correction 191 . On all our data - including Si-crystalline and CIS modules - we observe an exponential increase of the Rsh when irradiance decreases. This correction has a moderate effect on technologies with high Rsh, but strong in amorphous. This is the main contribution of our corrections. Recombination losses
190

. This additional current loss takes place in the intrinsic layer of the amorphous

187

Chapter 8 Physical models used

Physical models used

junctions. The correction proposed by Mertens & al implies a distortion of the I/V curve, which cannot match our I/V measurements anymore, but dramatically improve the Vco voltage behaviour of the model. Spectral corrections 189 . Applying the spectral correction established at Loughborough UK for single amorphous modules (Betts & al.), improves the errors distributions by a factor of 10 to 20%. It is not clear whether this correction is quite suitable for our double and triple junctions. PVsyst doesn't take into account the the well-known initial degradation, due to Staebler-Wronski effects. PVsyst results are supposed to apply to stabilised module performances after 2-3 months of exposition to the sun.

NB:

Spectral characterization: Average Photon Energy (APE)

The "Average Photon Energy" is aiming to the characterisation of the energetic distribution in an irradiance spectrum. It is obtained by dividing the irradiance [W/m or eV/m/sec] by the photon flux density [number of photons/m/sec]. From detailed spectral measurements over one year performed at Loughborough, the CREST has deduced a parametrization of this quantity according to: - The relative air mass, which, in its simpler form is expressed as AM = 1/cos z, with z = zenithal angle. - The atmosphere transmission according to the weather, usually expressed with the clearness index Kt, ratio between the horizontal global and the irradiance outside the atmosphere. But as the clearness index is not independent of the air mass (depends on the sun height), CREST has chosen to use a clearness index normalised to clear sky conditions Ktcs . But while CREST determines the "clear sky" conditions by adjusting an exponential on the higher global values observed as function of air mass, in PVsyst we can use the "Clear Sky Model". This parametrization looks like the following:

APE parametrization at Loughborough (UK)

1.75 1.70 1.65 1.60

1.700-1.750 1.650-1.700 1.600-1.650 1.550-1.600 1.500-1.550 1.450-1.500

APE

1.55 1.50 1.45 1.40 1.35 1.30 0 0.2 0.4 3 2 1

STC: AM = 1.5 => APE = 1.6 eV

0.6

KTc

0.8

1 6

Air Mass

The "standard" spectrum AM 1.5 corresponds to APE = 1.6 eV. In the English climate, the annual distribution of APE is a bell-shaped curve, centered on 1.65 eV and with a half-height width of about +/-0.08 eV. By clear days, the APE diminishes (shift to the red) when the sun height decreases. Cloudy spectrums are rather characterized by spectrums shifted to the blues (more favourable for amorphous modules).

Chapter 8 Physical models used

188

Physical models used

Amorphous modules: Spectral correction

The reference irradiance used for the simulation (the meteo values) includes the whole spectrum from 305 nm (UV) to 2800 nm (IR). It is usually measured with pyranometers, which have a practical flat response over this whole interval. But each PV technology is characterised by a spectral sensitivity curve. When the SI-crystalline can use photons below 1'100 nm, corresponding to Egap = 1.12 eV, the photons should have a minimum energy of Egap = 1.7 eV (730 nm) for creating a pair in the amorphous silicium. Therefore, the photocurrent should be evaluated using a convolution integral between the incident spectrum and the spectral sensitivity. We will call "Utilisation Factor" (UF) the value of this integral, which represents the fraction of the spectrum effectively useable for generating photocurrent. Nevertheless, the spectral content of the solar radiation varies with the meteorological conditions and the humidity/aerosols of the atmosphere, etc. And of course we don't avail of spectral measurements in PVsyst. In order to estimate the Isc current at any instant, the CREST at the University of Loughborough proposes a procedure in two phases: first, characterising the spectrum using a suited parameter, which could be evaluated from available environmental parameters, and then determine a correlation between this parameter and the spectral sensitivity of the concerned technology. The chosen parameter is called "Average Photon Energy" (APE) 188 , and is obtained by dividing the irradiance [W/ m or eV/m/sec] by the photon flux density [numbre of photons/m/sec]. From detailed spectral measurements over one year, CREST has deduced a parametrization of this quantity according to air mass and "clear day" clearness index. Waiting for new measurements elsewhere, we can reasonably admit that this parametrization is valid at least for European climates. The second phase is to determine a correlation between the UF, calculated for each measured spectrum for a given technology, and the APE. It is found that they are quite well correlated, and lead to a simple quadratic expression. The final amorphous spectral correction UF is shown on the figure.

Utilization Factor a-Si:H

0.670-0.700 0.640-0.670 0.610-0.640

0.70 0.67 0.64

0.580-0.610 0.550-0.580 0.520-0.550 0.490-0.520 0.460-0.490

UF a-Si

0.61 0.58 0.55 0.52 0.49 0.46 0 0.2 0.4 0.6 4 3 2 1

KTc

0.8

1 6

Air Mass

It varies between about 0.5 (APE=1.45) to 0.065 (APE=1.70), i.e. a range of the order of 30%. It can be seen that the response of amorphous modules, by clear weather, decreases significantly when air mass increases (winter, morning and evening). But it remains rather good by cloudy conditions (lower Ktc). Finally it should be noted that the final spectral correction used in PVsyst has to be renormalized to the UF of the reference conditions when establishing the model (STC: AM 1.5 spectrum, corresponding to APE = 1.6 eV). This is the reason why the program will ask for the conditions in case of specifications based on outdoor measured data. 189

Chapter 8 Physical models used

Physical models used

8
Utilizing Factor Correction for a-Si:H

Four our 3-month measurement campaign, the Spectral correction calculated from this parametrization looks like the following:

1.20 1.15 1.10

Relative UF

1.05 1.00 0.95 0.90 0.85 0.80 0 200 400 600 800 1000

GlobP [W/m]

where we clearly identify the clear sky conditions, and the sensitivity enhancement for cloudy conditions (but acting on low-power hours!). NB: In the same conditions for crystalline modules, the UF varies between about 0.81 and 0.91 indicating a better use of the whole spectrum. But applying this correction to the measured data doesn't improve the results of the model. Back to Standard "One-diode" model
182

.
186

Back toAmorphous and Thin films modules

Thin film modules: Recombination losses

When it fits well the behaviour of Vco in Crystalline and CIS modules, the standard model fails to reproduce the amorphous modules voltage in any irradiance and temperature conditions. Amorphous junctions differ from other junctions by the presence of an "intrinsic" layer (p-i-n junction). J. Mertens 1 al. propose to take the recombination losses in this I layer into account, by adding a term in the general I/V equation. This term is equivalent to adding an element to the equivalent circuit, representing a current leak depending on the photocurrent and the voltage.
I Rs
Photocurrent Recombination Diode User

Rsh
Iph Irec (Iph, V)

RL

Modelling of this phenomenon leads, under some hypotheses, to the following expression for the recombination current: Irec = Iph di / [ where
eff (Vbi - (V + I Rs )) ]

di = Thickness of the intrinsic i-layer (of the order of 0.3 m),

Chapter 8 Physical models used

190

Physical models used

eff = Diffusion length of the charge carriers p and n: eff = 2 n n p p / ( n n + p p)

Vbi = Intrinsic voltage ("built-in voltage) of the junction. Its value may be considered as constant, and is about 0.9V for an amorphous junction. With this new term the general one-diode model I/V expression becomes I = Iph - Iph di / eff 1 / (Vbi - (V + I Rs )) - Io [ exp (q (V+IRs) / ( Ncs GammakTc) ) - 1 ] - (V + IRs ) / Rsh In our phenomenological study of 4 amorphous modules, we considered the quantity di / eff as one only parameter, and we sought the value which optimized the Vco response of the model. For all our modules we found that a value (di / eff) around 1.4 V gives excellent results and corrects quite well the Vco distribution, with a simultaneous improvement of the Pmax response. For example on our SHR-17 tripple-junction, the RMSE on Vco drops from 3.3% to 0.7%, and the MBE from 4% to 0.2% with this correction. Simultaneously, the RMSE on Pmax is improved from 5.8% to 4.1%. It should be noted that this new term doesn't modify significantly the procedure used for getting the model parameters. Simply the photocurrent value is now affected by a voltage-dependent correction in the equations. Nevertheless the resolution of the model gives a quite different gamma value, compatible with it "physical" limits.

Back to Standard "One-diode" model

182

.
186

Back toAmorphous and Thin films modules

Thin film modules: Fit on input values of Rshunt

This little tool allows for adjusting the exponential parameters, according to some known Rshunt values at different irradiances. The Rshunt value can be obtained on the basis of measured I/U characteristics of the module. Rshunt is the inverse of the slope around V=0 (i.e. the short-circuit point at Isc). If you avail of such data or measurements at different irradiances, you can put them on the plot. You can use the right button of the mouse for creating a new point, and then slide it to obtain the desired value. For deleting a point simply click on it with the right button. The "Fit" button will then give the adjusted exponential parameter. See also: Rshunt exponential correction vs Irradiance
191

Thin Films: Rshunt exponential correction vs Irradiance

The shunt resistance Rsh - corresponding to the inverse of the slope of the I/V curve around V=0 - is considered as a constant parameter in the standard one-diode model. But it is easy to observe on amorphous I/V curve families that this slope decreases with the irradiance. The next figure shows the measured Rsh behaviour over our measured sample:

191

Chapter 8 Physical models used

Physical models used

SHR-17 Old - R shunt function of Irradiance
1000

8
Measurements

R Shunt measured [ohm]

800

Parametrization

600

400

200

0 0 200 400 600 800 1000

Irradiance [W/m]

This distribution may be approximated by the following exponential expression: Rsh = Rsh(Gref) + [ Rsh(0) - Rsh(Gref) ] exp(-Rsh exp (G / Gref)) We observed a similar behaviour on all our measured data (including the Si-Crystalline and CIS modules). And curiously, all data may be rather well approached using a unique value Rshexp = 5.5. Therefore this value is set as default value in PVsyst. In this expression there are 2 unknown parameter left: Rsh(Gref), i.e Rsh in the standard model, and Rsh(0). which may be determined on measured data, but are not available in the usual manufacturer's data sheets. For processing the database, PVsyst has to propose default values, which are deduced from our own measurements on 6 modules, but cannot be of course very reliable. The Rsh default value 185 at reference conditions has already been chosen when establishing the model. For amorphous modules, PVsyst proposes a Rsh(0) value of 12 times Rsh. For Si-crystalline and others, we observed optimum values around 4 * Rsh, but this cannot be asserted for any module. Therefore PVsyst doesn't propose any exponential correction as default; this remains an optional refinement. In the future, these values should ideally be provided by the module manufacturers in their data sheets. NB: When two or several Rsh values are available, PVsyst offers a tool for adjusting the corresponding Rsh(ref) and Rsh(0) parameters 191 .

Back to Standard "One-diode" model

182

.
186

Back to Amorphous and Thin films modules

Batteries - Model description

Model choice justification
We have developed a non conventional battery model, trying to avoid the pitfalls which arise in a number of existing PV softwares: either an extreme simplification, which can only lead to rough evaluations of the system behaviour, or adjusted models based on numerous (often interrelated) parameters whose physical meaning is often not clear to the user, and practically necessitate a complete measurement of each battery used. We have therefore tried to fulfil the following criteria : On the one hand, the model should be presented to the user in a very simple manner, involving priori only "obvious" parameters specific to each battery: type of technology, voltage or number of elements, nominal capacity, possible internal resistance and Faradic efficiency (whose value is not critical and which can be taken at their default value). But on the other hand, it should be sufficiently detailed to satisfy the needs of the simulation of the PV system, where the charging current is practically imposed by the solar generator. In particular, its behaviour in voltage, not critical in the intermediary zones, should be realistic enough at the end of charge and discharge to make the regulator operate correctly. Further, it will be important to be able to
192

Chapter 8 Physical models used

Physical models used

estimate the ageing and the possible maintenance imposed by the conditions of use.

We have therefore opted for a " two-stage complexity " definition. A basic, very simple model describes the voltage with respect to the state of charge (SOC), the internal resistance and the temperature. It is to be noted that the SOC is rarely chosen as a basic variable by experimentalists for the batteries' modelling, as it is not directly accessible for measurement. But in the framework of our utilisation in system simulation, it is relatively well determined by the charge/discharge balance, providing that we model the losses from electrolytic dissociation which we will discuss later on. It is to be noted that this model is valid for lead-acid batteries. It will certainly be necessary to strongly adapt it for Ni-Cd batteries, which is much less frequently used in solar systems. This has not yet been implemented in this version.

Model description: open circuit as function of SOC

The basic linear model takes the simple form : Ubatt = Uocbase + Alpha SOC + Beta (Tbatt - Tref) + Ri Ibatt with: Ubatt = voltage for a battery- element. Uocbase = intercept of the open circuit voltage linear part at SOC=0. SOC Alpha Tbatt Tref Beta Ri Ibatt = state of charge (varies from 0 to 1). = slope of the open-circuit line (depends on the chemical couple Pb-SO4). = temperature of the battery. = reference temperature (usually 20C). = temperature coefficient (-5 to -6 mV/C). = internal resistance, assumed to be constant. = battery current (charge > 0, discharge < 0).

This model is then completed by a series of disturbances, whose values are predefined in the programme (modulated especially by the chosen technology), but which can of course be adjusted by the user if he has more specific data for his own battery at disposal.

Over-charging and deep discharge

The first two disturbances to be taken into account are the behaviours at the end of the charging and the discharging processes, which mainly affect voltage, and therefore regulation. When the battery approaches complete discharge , the voltage falls progressively, whatever be the current. We have fixed a fall of a quadratic shape starting from a SOC of 30%, which gives realistic results for most of the batteries. But this exact shape is of little importance for the behaviour of the system as a whole. At the end of charge, the problem is more delicate because of the apparition of the electrolyte dissociation ("gassing"). This phenomenon is rarely treated explicitly in the usual models, but it is still fundamental as it consumes a part of the charging current, and therefore affects the global operating of the system, in particular the real efficiency of the battery. We suppose that this phenomenon induces an excess voltage with respect to the linear behaviour depending on the SOC. The shape of this excess voltage is a predetermined "S"-curve, (ref.[22]). The "gassing" current increases exponentially with the excess voltage, and substitutes itself progressively for charge current. The "Delta" coefficient of the exponential has been measured by a German team for various batteries of different ages, and is established at about 11.7 V-1 (ref.[23]). The end of charge is another predetermined limiting curve, depending on the charging current (ref.[20]). It corresponds to the situation where the whole current is used for the dissociation, and allows to fix the Iogass parameter: Igass = Iogass exp ( Delta dUgass ).

Temperature dependence and polarisation

To put it more simply, the dependence of the voltage on the temperature is assumed to be linear in all
193

Chapter 8 Physical models used

Physical models used

operating conditions. In the same manner, the internal resistance is assumed to be constant; any possible dependence on temperature (studied in ref.[21]) is ignored. We also neglect the polarisation effects, or voltage drift depending on immediately preceding operating conditions; they show characteristic times less than one hour, and only affect the states of very low current.

State of charge determination

The state of charge remains to be determined. Fundamentally, it corresponds to the charging and discharging balance currents, with respect to the nominal capacity of the battery. But a number of corrections are to be made. First, the electrochemical conversion (an SO4-- ion - 2 electrons of current) is not perfect: the balance, called " coulombic efficiency ", is of the order of 0.97 for the batteries usually used in solar installations. This efficiency can be specified by the user, but remains constant in all the conditions of use. It should not be mixed up with the global current efficiency, including losses from "gassing" current, which depends on the operating conditions and appears as a simulation result. The coulombic efficiency is applied to the charging current. We then have the loss due to self-discharge current, which strongly depends on temperature, type and age of the battery. The temperature behaviour of the self-discharge current does not depend very much on the type of battery: it is close to an exponential, and doubles every 10C. In the programme, it is specified by a predefined profile (which increases steeply for uses above 20C). But its basic value at 20C is

asked in the module of the secondary parameters. Capacity corrections

Further, it must be noted that the nominal capacity of the battery is not a well determined parameter either. In practice, it is measured by carrying out a complete discharge (supposing that we can precisely define the states of full charge and the end of charge). But it depends on several factors of which the most important are age, discharge rate and temperature. In the lack of reliable data, the software does not explicitly take the age into account. We will suppose that the user will modulate his basic nominal capacity, according to its own criteria, for the simulation over one year. But we cannot neglect the effect of the discharge rate . The nominal capacity is usually defined, in the manufacturer's data, for a discharge in 10 hours (that is a current of one tenth of the capacity). In fact, these conditions are not realistic in a solar system, where the current is usually weaker and the real capacity can increase up to 150% (cf. curves). The model takes it into account by decreasing the SOC, at each discharging step, by an instantaneous amount weighted by the nominal capacity corresponding to this current value. But in order to preserve a correct balance, we also have to apply the same capacity correction to the recharging process; for this purpose, we will memorise an "average" capacity on consecutive discharge steps, in order to apply it to the next charging period. This procedure is not rigorous, and can sometimes lead to efficiencies higher than 100%; but on an average it should to be correct over a long period. Finally, capacity is reduced with battery temperature ; this is one of the reasons why the effective temperature of the battery should be know during operating, and should therefore be specified in the parameters of the PV system. The model evaluates the capacity in accordance with a predetermined profile, which is not critical and not specific to each battery.

Ageing and wear and tear

One of the advantages of the programme is also to evaluate the wear and tear of the battery depending on the running conditions, and therefore the investment to be planned for its replacement. Ageing is governed by two phenomena: on the one hand a "static" longevity, characteristic of the battery, whether it is used or not. This value is often given by the manufacturer at a reference temperature (20C). But it strongly varies with temperature, and we will agree with most of the manufacturers that it reduces by half for an elevation of 10C. And on the other hand , a deterioration due to use , depending on the depth of the discharge at each cycle, is given by manufacturers as a characteristic profile. For each discharging step, we therefore
194

Chapter 8 Physical models used

Physical models used

evaluate a wear and tear increment proportional to the current, but weighted by the actual depth of the state of discharge. The global wearing out of the time step is considered as the maximum value of these two evaluations.

Maintenance
The modelling of the "gassing" current also allows for the calculation of the quantity of electrolyte consumed, giving an idea of the frequency of maintenance (distilled water checks and addition). Unfortunately - or fortunately! - the problem is not all that simple as some sealed batteries (said to be without maintenance) contain a catalyst which allows the recombination of H2 and O2 gases to again give water.

Model use in simulations

Thus, PVSYST avails of a model which is certainly not perfect, but whose results seem to fulfil the exigencies of the simulation of a PV system. In particular, it will help explore the effects of regulation's threshold adjustments, either on the performances of the PV system, or on the ageing conditions of the battery. This model requires a minimum number of parameters to quickly become operational. It can be more finely adjusted if the user can avail of supplementary data about his own battery. The whole set of parameters appears on the standard sheet printed by PVSYST. The model gives rise to charge/discharge curves which can immediately be compared to those of the manufacturers in order to verify the conformity of the model.

Pump - Model description

See also the General Requirements Variables Let us define the following variables: Up, Ip = Voltage and current applied the pump. Pp = Up * Ip = Input power of the pump Uc, Ic, Pc = Voltage, Current and Power applied to the input of the power converter, if any. FR = Flowrate produced by the pump HT = Total Head, sum of the Static Head (related to the difference between input and output water levels), and Dynamic Head, due to friction losses in the pipes and system. Dynamic head is dependent on the flowrate, and will be computed by the system simulation process.
197

governing the elaboration of this model.

Model Structure The pump characteristics may be considered as a set of operating points represented as a surface in the 4-variable space, i.e. corresponding to the equation: (Up, Ip, HT, FR) = 0. This function will be defined on an operating domain, which is bounded by some limits (usually specified by the manufacturer): - Maximum voltage applied to the motor-pump, - Maximum electrical power, - Maximum current - Maximum Head (implies maximum current), and toward the low values: - Power threshold for starting operating (i.e. not null flowrate), which is a function of the Head.

195

Chapter 8 Physical models used

Physical models used

Boundary operating limits
10

8
Max. Power Max. Current

8 Head = 41 m

Current [A]

4 Head = 11 m 2

Max. Voltage

Threshold (Limit FlowR = 0)

0 0 10 20 30

Voltage [V]

40

50

60

70

80

Boundary limits of the operating domain and I/V measurements for pump Watermax BU (CIEMAT measurements) The equation =0 implies that only 3 of the 4 variables are independent. Therefore the model will provide relations allowing to calculate any one of the above variables, as functions of two others. The basic relations are: - Ip = f (Up, HT), the fundamental relationship which will be used for determining the operating point when directly coupled to a PV array. - FR = f (Pp, HT), completing the preceding relation for determining the corresponding flowrate. - Pp = f (FR, HT) will be used for example for sizing the PV array power, or for determining the efficiency. The other relations may be obtained by numerically inverting these 3 fundamental ones. As a complement, the model also provides functions for determining the Power, Voltage or Current threshold (i.e. the boundary where the flowrate drops to zero) as function of the Head. The Phenomenological Model The problem is now to determine this function . We would like to avoid references to technology-specific parameters, that is to physical models describing the motor or pump. Therefore our model is mainly based on the known performances, i.e. the operating points either specified by the manufacturer, or measured by other sources. If these points are sufficiently well distributed over the operating domain, they will completely define the pump behaviour. The informatic model has to interpolate between the given points; in practice, it will perform cubic interpolations between points, and linearly extrapolate the data up to the boundaries. Therefore this model is just a phenomenological one, without any physical contents. Physical assumptions will be necessary only if the data set is not sufficiently well distributed for allowing extrapolations within the entire operating domain. These very general assumptions will be established according to general behaviours observed when measuring a great number of pump technologies. These could probably be refined during our future works. Of course this lack of primary information in the basic data will result in lower accuracies of the model's predictions. In practice the manufacturers use to specify the performances of their products by giving different kinds of data sets. We identified 5 of them, which may be input directly in the PVsyst model. Each one has to be treated specifically in the model; that is, the algorithms of the basic functions mentioned above will be different for each kind: Given Ip and FlowRateas f(Head) for fixed Voltage
199

, usually for Positive Displacement Pumps.

199

Given Power and FlowRateas f(Head) for fixed Voltage point.

, equivalent set, as U is defined for each

Given FlowRate as f(Power) for fixed Heads 199 , usual for solar centrifugal pumps, but doesn't include Current/Voltage specification, therefore only suited for configurations with converter. Given Head and Poweras f(FR), fixed voltage or speed 198 is the usual way of defining standard centrigugal pumps for grid operation. This definition leads to the "Similarity Laws" model.
196

Chapter 8 Physical models used

Physical models used

Given Head and Efficiencyas f(FR), fixed voltage or speed being easily deduced for each operating point.
198

8
is equivalent to the preceding, the power

Pump model: General requirements

The pump model should match the following characteristics:

195

Input/Output variables
The model should describe the dynamic evolution of the output variable - usually the flowrate - as a function of the pertinent input variables, which are basically the head and voltage input, for any conditions within the admitted operating values. Indeed, when a given voltage is applied to the pump, this will run at an operating point characterised by a flowrate yield, as well as by a current drawn from the source. Therefore current is also a function of the Voltage and Head inputs.

Functions
The general model will give all the relationships between these 4 variables, i.e. calculate any variable from the 3 other ones. Therefore it will include the determination of the Current/Voltage characteristic of the pump, which is necessary to the calculation of the operating point when coupling the pump directly to the PV array.

Without I/V information: needs a Power converter

In many cases the motor is specified for use at a nominal voltage, and detailed I/V behaviour is not available. The Flowrate is then given as a function of Head and input electrical Power. These only 3 variables are in principle sufficient for characterizing the operating point, when the power input is fed through some power-conditioning unit, which will provide an adequate (unknown) voltage, and eventually frequency, to the pump motor. That is, a power converter is always required when the Current/Voltage are not specified.

Starting current
Besides these 4 operating variables, most positive displacement pumps also require special starting conditions, i.e. a starting peak current for overcoming the torque due to the friction forces when stopped.

Applicability to any technology

The model should cover any motor-pump's technology 156 , available on the market for use in PV systems: centrifugal pumps, positive displacement pumps (including piston, membrane or diaphragm, progressive cavity, rotating displacer, etc.). These pumps can be driven by diverse AC or DC motor technologies. The model will also apply to other standard pumps (not specifically designed for solar applications), with AC induction motors, driven by frequency converters (inverters). Ideally, the parameters necessary for achieving the modelling should be available from the manufacturer data sheets, in order that any user of the program can input its own pump model characteristics. In practice, manufacturers usually give performances for only a limited set of actual operating conditions. Details about the motor or pump technology and related fundamental parameter are usually not available. Therefore, the model will apply to the motor-pump group as a whole, without reference to intermediate values like torque or speed, which are highly technology-dependent and rely to unavailable specific technical parameters.

Model basis:
The model 195 will be a phenomenological one, based on the specification of some Operating points (FlowR, Head, Current, Voltage, Power) from which the operating properties will be interpolated or extended for any conditions.

197

Chapter 8 Physical models used

Physical models used

Pump model from Head (and Effic) as f(FR), fixed U or speed Pump model from Head (and Effic) as f(FR), fixed U or speed

Standard centrifugal pumps designed for grid applications are usually specified by one Head vs Flowrate curve, for nominal grid conditions (fixed voltage, 50 or 60 Hz). To be fully determined, the model will also need the specification of the efficiency for each given Flowrate(Head) operating point. For an operating point, the electrical power is easily deduced from the Efficiency, using the Hydraulic Power calculated as the product of the Head and Flowrate. Therefore this case is equivalent to the model from Head and Power as f(FlowRate)
198

73

Pump model from Head (and P) as f(FR), fixed U or speed Pump model from Head (and P) as f(FR), fixed U or speed
Standard centrifugal pumps designed for grid applications are usually specified by one Head vs Flowrate curve, for nominal grid conditions (fixed voltage, 50 or 60 Hz). To be fully determined, the model will also need the specification of the electrical power (or efficiency) for each given Flowrate(Head) operating point. This is not always provided by the manufacturers, as the power consumption is not a key parameter when using grid-powered pumps. Pumps without this minimum specifications cannot be used in PVsyst. Using this only information at nominal speed, centrifugal pump behaviour at other speeds may be very well described by the so-called "Similarity laws" model [see Abella, 2003, Suehrcke, 1997_1997]. These relations state that for two operating points at different speeds along an iso-efficiency line, one has: FL1/FL0 = P1/P0 H1/H0 1/ 0 = ( 1/ 0) = ( 1/ 0)

Using in addition a cubic-spline interpolation on the given Head(FlowRate) curve, solving these equations allows to determine any operating point from two given variables. Validation The results are impressive: The next figure shows a comparison between measured and modelled values, for the model established only on 3 operating points at nominal voltage, and for a wide range of operating conditions.
Model FlowRate (Pelec, Head) - Measurements
80 70

Head = 32 m Head = 22 m Head = 12 m

FlowRate Model [l/min]

60 50 40 30 20 10 0 0

10

20

30 40 50 FlowRate Measurements [l/min]

60

70

80

Model validation on a Solarjack SCS-160 pump, CIEMAT Measurements [Abella, 2004] A significant deviation is observed only at lower flowrates (powers). This is probably due to the converter and motor efficiency change with power (or speed), which are not taken into account here. Such pumps are most of the time driven by AC synchronous motors, and may be powered using a cheap standard Frequency Converter (FC) [see Abella, 2003]. When using a DC motor, the Ip / Vp characteristics is strongly dependent on the motor technology, and seldom known.

Chapter 8 Physical models used

198

Physical models used

Pump model from FR as f(Pp) for fixed Heads Pump model from FR as f(Pp) for fixed Heads

Grids of FlowRate (Power) curves for different Heads is the usual way of specifying the solar centrifugal pumps. These allow to get a very good determination of the Hydraulic/Power behaviour, but don't hold any information about the Voltage/Current characteristics. Such data are suited only when the pump is coupled to the PV system through a power converter. If needed, the basic function Ip (Up, Head) requires additional informations. These could be provided either by a set of Ip/Vp points for at least two head values (on which we can adjust curves), or by several parameters: Nominal voltage and Current at a given reference Head, V/ I at fixed Head (i.e. dynamic resistance) and I/ H at fixed voltage. When specifying such parameters instead of curves, we assume linear behaviours, which will of course penalise the accuracy of the Ip(Up) characteristics.

Pump model from Ip and FlowR as f(Head) Pump model from Ip and FlowR as f(Head) with fixed voltage
This is the common data usually specified for positive displacement pumps. Complete characterisation of the model requires such data sets for several operating voltages. It is unfortunately a common practice to specify such a data set for only one "nominal" voltage. This is of course not suitable for computing direct-coupled configurations. The extension to other voltages requires a strong hypothesis, which is (in this first status of the model) that the Pump Efficiency is constant when varying the voltage. The figure shows an example of measured efficiency profile. We can observe that within the reasonable operating range 30-64V, the variations are of the order of 15%
Efficiency vs Voltage
50 45 40 35 Head = 41 m Head = 32 m Head = 22 m Head = 12 m

Efficiency

30 25 20 15 10 5 0 0 10 20 30
Pump Voltage

40

50

60

70

If we avail of 2 voltage curves the efficiency figure is a linear interpolation, which improves the model's accuracy. Given Pp and FR as f(Head) for fixed Voltage This is equivalent to the preceding scheme, as for each data point the current may be easily determined from power, using the fixed voltage parameter of the curve. Some typical definitions

Albedo
The albedo coefficient is the fraction of global incident irradiation reflected by the ground in front of a tilted plane. This effect takes place during the transposition computation of the horizontal irradiation onto a tilted plane. The albedo "seen" by the plane is of course null for an horizontal plane, and increases with tilt. In the project definition, the albedo values can be adjusted each month in order to take any possible snowcover into consideration. The value usually admitted in the urban localities is of the order of 0.14 to 0.22, and can go up till 0.8 for a snow-cover. Ideally, the best value is obtained by a direct measurement on the
199

Chapter 8 Physical models used

Glossary

site! But in practice, except for vertical planes, this value does not take on any great importance as the albedo component is relatively weak in the incident global irradiation (this contribution can be visualised in the results of your simulation). The following table gives some usual values for the albedo: Urban environment Grass Fresh grass Fresh snow Wet snow Dry asphalt Wet Asphalt Concrete Red tiles Aluminium Copper New galvanised steel Very dirty galvanised 0.14 - 0.22 0.15 - 0.25 0.26 0.82 0.55-0.75 0.09-0.15 0.18 0.25-0.35 0.33 0.85 0.74 0.35 0.08

Albedo attenuation factor

The albedo is only visible from the collectors if no close obstacle is present till the level of the ground. This is why we have to integrate the shading factor at zero height, on the virtual portion of the sphere under the horizon, included between the horizontal plane and the plane of the collectors. As for diffuse, this factor is independent of the sun's position, and therefore constant over the year. It is however to be remembered that for non-vertical planes, the energetic contribution of the albedo is weak in the global incident energy, and that errors in its estimation will therefore only have secondary repercussions.

Altitude corrections
Altitude corrections
When associating a meteo file to the project site, the program can perform altitude corrections on the meteo data (irradiation and temperature). It is however to be noted that these altitude corrections, established for Switzerland according to Meteonorm studies in monthly values (Bremer 86), have not really been validated for an hourly behaviour, neither for local phenomena such as sheet of morning fog, which cannot be realistically taken into account. On the other hand, we do not apply this correction on diffuse irradiation, as the increase in irradiation with altitude mainly affects the beam component.

Autonomy and battery sizing (stand-alone)

Autonomy and battery sizing
In the Presizing process, the proposed battery pack capacity is determined according to the required autonomy of the system, given in days. The autonomy is defined as the time during which the load can be met with the battery alone, without any solar inputs, starting of course from a "full charged" battery state. With non-constant loads (seasonal or monthly definition, weekly use), this is accounted as the worst case over the year. The calculation takes the minimum SOC disconnecting threshold, and the battery "energy efficiency" into account. It also has to apply a capacity correction, as this mode of use usually occurs at a rather slow discharge rate (therefore corresponding about to the C100 capacity definition, i.e. with discharge in about 100 hours); when the nominal capacity of the batteries in the PVsyst database is defined as the C10 value. The C100/C10 ratio is usually of the order of 120% to 130%. These 3 sizing parameters are adjustable in the "Hidden parameter". NB: When defining a very large autonomy, the global system optimisation process will choose the "minimum" PV size which just meets the required "LOL". This may lead to a very low average State of Charge value over large periods of the year, which is quite damaging for the battery. In these cases, the actual PV-array sizing should be slightly increased by respect to the PVsyst proposal, in

Chapter 9 Glossary

200

Glossary
the definitive "Project" study.

Battery current efficiency

Battery current efficiency
This parameter is the Electrochemical conversion current efficiency, without gassing. The global coulombic efficiency, including gassing losses, depends on use conditions and will be a result of the simulation. The energy efficiency, including voltage losses between Charging and Discharging processes, will be a result of the simulation, too.

Battery maximum charge threshold

Battery maximum charge voltage threshold
The end-of-charge voltage threshold is slightly dependent on the battery technology. For vented lead-acid batteries it should be 2.23 V/element, while for sealed batteries with recombination catalysator it can be raised to 2.25 V/element (manufacturer recommendation). According to manufacturer data sheets, the end-of-discharge threshold may depend on the discharge current, due to internal resistance voltage drop. Nevertheless, this applies in back-up use with high currents (discharges in one hour or less). With solar use currents are usually far less and this effect can be neglected. PVSYST - as usual solar regulators - doesn't take it into account and proposes a fixed threshold voltage.

Clearness_Index
Clearness Index Kt
The Clearness Index Kt is defined as the ratio of the horizontal global irradiance to the irradiance available out of the atmosphere (i.e. the extraterrestrial irradiance multiplied by the sinus of the sun height).

The extraterrestrial irradiance is the Solar constant (1367 W/m) corrected by a yearly sinus function of amplitude 3.3% accounting for earth orbit ellipticity.

Climatic Distance
Climatic Distance

According to the Meteonorm definition, the "Climatic distance" between two sites is the quadratic sum of their horizontal distance and the altitude difference weighted by a factor of 100: DistanceClim = ( Dist Hor + (100 Diff Altit) ) Following Meteonorm Standard, it can be admitted that a meteorological site is representative of a given place if their climatic distance does'nt exceed about 20 km.

Dates read on the file

Dates Read on the File
The format protocol specifies "dates read on the file", but the Year is not read. In this case the year must be must be supplied by the user.

Diffuse attenuation factor

Diffuse attenuation factor
The sky diffuse component is affected by the near shading obstacles, as well as the incidence angular modifier (IAM). In order to take it into account, we suppose as simplification that the diffuse sky irradiation is isotropic. A shading factor on this diffuse irradiation (independent of the sun's position, and therefore constant over the year) can be calculated, by integrating the shading and/or IAM factors on the visible part of the vault of heaven, that is the dihedron between the collector plane and the horizontal plane.

Domestic user's needs

Domestic user's needs
The domestic user's needs are defined on a daily basis. 201

Chapter 9 Glossary

Glossary

These can be constant over the year, or seasonally / monthly modulated. You can of course copy values from one season/month to other ones. The Week-end option allows for concentrating the daily consumption on a specified number of days of the week only. This option may be specified independently for each season/month. Stand-by consumption may be extended to the non-utilization days if desired. When using an AC distribution with inverter, please define an additional consumption of around 10% to take the inverter efficiency into account. For simplifying data acquisition; you can save your definitions as a model for re-use in other projects.

Double orientation field

Double orientation field
It is possible to connect fields of different orientations on a single input (inverter or battery pack), so long as all modules of each string are kept in the same orientation. In this case the currents of different strings will be different (according to the sun orientation) but will mix at about the same voltage. This results in negligible mismatch losses. Putting modules in series in different orientations will yield complex I/V characteristics behaviour analogous to partial shadings effects. This can result in high losses and eventually hot spot risk. See the special tool "Electrical behaviour of PV arrays" in the "Tools / Solar tool box" part of PVSYST for visualising and understanding these phenomena.

European Efficiency
MPPT or DC-DC converters: European Efficiency
The converter doesnt' operate always at its maximum efficiency. Therefore, PVsyst defines a "European Efficiency" for converters, in the same way as it is usually defined for inverters. The "European Efficiency" is a calculated efficiency, averaged over a power distribution corresponding to middle Europa climate yearly operating conditions. If we denote by "E50" the efficiency at 50% of nominal power, it is defined as: EuroEfficiency = 0.03 x E5 + 0.06 x E10 + 0.13 x E20 + 0.1 x E30 + 0.48 x E50 + 0.2 x E100.

Generic Year 1990

Generic Year 1990
By convention in PVsyst, all data which are not relied to real meteorological measurements of a given year are considered as a generic year, labelled as 1990. This is the case for most of the monthly data in the database, which proceed from averages over several years. This also holds for synthetic hourly data or Design Reference Years (DRY).

Grid current in tri-phased configuration

In tri-phased configuration, when well-balanced, the AC current in each phase conductor is obtained by dividing the total Power by the Voltage * SQR(3) (voltage between phases): Power = Voltage * Current * SQR(3)

Incidence_Angle
The incidence angle is defined as the angle of the sun's rays, by respect to the normal to the collector plane. Incidence angle is zero for normal incidence, and 90 for rays parallel to the plane. The sun's height on the plane is the complement to 90 of the incident angle.

Inverter: power overcharging

Inverter overcharging
Behaviour at power overcharging (i.e. when the output power overcomes the nominal output power (AC-power): Limitation : the PV-field's running point moves along the I/V curve in order to limit the output power at the nominal value.

Chapter 9 Glossary

202

Glossary
Cut : the inverter output is turned off until input power at MPP matches the output nominal power. Cut until evening : the inverter output is turned off until the evening (some old inverter models).

Irradiance loss
Irradiance loss
In PVsyst, the evaluation of the "Losses" of a PV array (as for the definition of the normalised performance ratio), takes as starting point the energy which would be produced if the system worked always at STC conditions (1000 W/m, 25C, AM1.5). The loss due to operating temperature (instead of 25C) is well-known and referenced by everybody. It is strange that nobody tells anything about the loss due to the irradiance level, which is of the same kind. Please have a look on the graphs of the behaviour of a crystalline PV module (in the PV module dialog, choose "Graphs" / "Efficiency vs Irradiance"), you will see that the efficiency decreases for lower irradiances: this leads to the "Irradiance loss" (by respect to 1000 W/m). Therefore this Irradiance loss is a consequence of the "one-diode" model for PV modules. One of the advantages of some amorphous modules, is that this "loss" is lower. Unisolar has even published efficiencies which constant or increased efficiency towards 150 W/m (see the results from the modelling according to our own measurements of SHR-17 in the database). This is which explains (along with the temperature coefficient) why amorphous systems show a better productivity [kWh/kWp] than crystalline ones in middle Europe climates.

Longitudinal_Incidence_Angle
Longitudinal Incidence Angle (Beta)
Let us call Profile Plane the plane passing through an horizontal line perpendicular to a given azimuth, and the sun. We call Longitudinal incidence angle the angle formed by the projection, on the profile plane, of the normal to the collector plane (or the azimuth line), with respects to the sun's direction.

LOL "Loss-of-load" probability (stand-alone)

LOL "Loss-of-load" probability
This value is the probability that the user's needs cannot be supplied (i.e. the time fraction when the battery is disconnected due to the "Low charge" regulator security). It may be understood as the complement of the "Solar fraction" (although it is described in terms of time rather than energy). During the sizing process, the LOL requirement allows for determining the PV array size needed, for a given battery capacity. The LOL is calculated using a simplified and fast yearly simulation: the program splits the monthly meteo values into a realistic random sequence of 365 days (according to Collares-Pereira model), each day being divided into 3 periods: morning - day (with solar gains) - and evening. Then it performs a day-by-day balance, and reports the daily system state, in order to accumulate a realistic "LOL" yearly value. This process is repeated with different PV-array sizes, in order to find the exact PV size matching the required "LOL".

LOL "Loss-of-load" probability (Pumping)

LOL "Loss-of-load" probability
This value is the probability that the user's needs cannot be supplied (i.e. the time fraction when the tank is empty). It may be understood as the complement of the "Solar fraction" (although it is described in terms of time rather than energy or water volume). During the sizing process, specifying the required LOL allows for determining the required pump nominal power, and the corresponding PV array size. The LOL is calculated using a simplified and fast yearly simulation: the program splits the monthly meteo values into a realistic random sequence of 365 days (according to Collares-Pereira model). Then it performs a day-byday balance of the tank state, in order to accumulate a realistic "LOL" yearly value. 203

Chapter 9 Glossary

Glossary

This process is repeated with different pump and PV-array sizes, in order to find the "exact" devices matching the required "LOL".

LOL "Loss-of-load" probability

LOL "Loss-of-load" probability
This value is the probability that the user's needs cannot be supplied (i.e. the time fraction when the battery is disconnected due to the "Low charge" regulator security). It may be understood as the complement of the "Solar fraction" (although it is described in terms of time rather than energy).

Metal resistivity
Metal resistivity
The resistivity of wiring metals is strongly dependent on the temperature, which can widely vary due to dissipating currents. For pure metal, one has:

Copper: [Ohmm]
Default value: Aluminium:

Rho = 1.68 E-8 * (1 + 0.0068 * Temp [C])

Temp = 50C => 22 mOhmmm/m Rho = 2.7 E-8 * (1 + 0.0043 * Temp [C]) [Ohmm]

Default value:

Temp = 50C => 33 mOhmmm/m

NOCT definition
NOCT Definition
Some practicians - and most of PV module's catalogues - usually specify the NOCT coefficient ("Nominal Operating Collector Temperature"), which is the temperature attained by the PV modules with free air circulation all-around, under standard conditions defined as: Gincid = 800 W/m, Tamb=20C, Wind velocity = 1 m/s, Open Circuit (?). The NOCT factor is related to our loss factor U by the thermal balance: U (Tcell - Tamb) = Alpha Ginc (1 - Effic) which gives with the NOCT conditions: (Uc + Uv 1m/s) (NOCT - 20C) = Alpha 800 W/m (1 - Effic) Now in this definition of the NOCT, the operating state of the module (at open circuit or at MPP) is not clear. The definition probably concerns open-circuit modules (i.e. modules exposed at sun, but not really in use), in which case the Effic value will be 0. But if this definition is suited for working conditions, the electrical energy is drawn from the module, and this thermal balance is affected by about the efficiency value, of the order of 10%. In doubt of this definition, PVsyst proposes here both relationships, to the choice of the user. where U = Uc + Uv WindVel.

Ohmic Loss Ratio (PV field)

Ohmic Loss Ratio
The Ohmic Loss ratio is referred here to the PV array at standard conditions (1000 W/m, 25C), It is the ratio of the wiring ohmic loss Pwir = Rwir * Isc compared to the nominal power Pnom(array) = Rarray * Isc. Where: - Rarray = Vmp / Imp at STC - Rwir = global wiring resistance of the full system. This is computed for a given sub-array (an inverter MPPT input) as the resistance of all strings wires in parallel, in series with the cables from the intermediate connexion box on the roof to the inverter input. The global wiring resistance Rwir is obtained by putting all the sub-array wiring resistances in parallel.

Chapter 9 Glossary

204

Glossary
Use in the simulation

The "Global wiring resistance" value finally used during the simulation may be defined here: - as a Ohmic Loss ratio (the default value is 1.5% at STC) - or given explicitely in mOhm. The "Detailed computation" tool is only a help for this calculation. When you use it, its result is reported in the "Global wiring resistance" value by checking the "Calculated" checkbox. Of course the ohmic losses behave in a quadratic way with the array current (Ploss = R I), so that the ratio diminishes linearly with the output current. Therefore the average wiring losses are much lower during the whole running year.

Plane azimuth
Plane azimuth
In northern hemisphere, the plane azimuth is defined as the angle between south and collector plane. This angle is taken as negative toward east, i.e. goes in the antitrigonometric direction. Example: south plane, azimuth = 0, east plane, azimuth = -90. In southern hemisphere, the plane azimuth is defined as the angle between north and collector plane. This angle is taken as negative toward east, i.e. goes in the trigonometric direction.

Plane orientation
Plane orientation
If you don't have horizontal meteo data measurements, you can use the irradiance measured in the collector plane. This is the reason why you should define the orientation of the solarimeter here. PVSYST will recalculate the Horizontal Global and Diffuse irradiations, which are likely to produce this tilted plane irradiation (inverse transposition). During the data analysis and simulation comparisons, you will be able to use either the original values or a recalculated plane irradiation. At this stage, plane orientation may be slightly modified, but big changes are not advised as the tilted plane irradiation looses some information by respect to horizontal true measurements.

Plane tilt
Plane tilt
The plane tilt is defined as the angle between the plane and the horizontal.

Performance Ratio; PR, Losses

Performance Ratio
This is a quantity, defined namely by the European Communities (JRC/Ispra), which represents the ratio of the effectively produced (used) energy, by respect to the energy which would be produced by a "perfect" system continuously operating at STC under same irradiance (Incident Global in the plane). The PR includes the array losses (Shadings, PV conversion, mismatch, wiring, etc) and the system losses (inverter efficiency in grid-connected, or storage/battery/unused losses in stand-alone, etc). Unlike the "Specific energy production" indicator, expressed in [kWh/kWp/year], this is not directly dependent on the meteo input or plane orientation, allowing the comparison of the system quality between installations in different locations and orientations.

Profile Angle
Profile Angle
Let us call Profile Plane the plane passing through an horizontal line perpendicular to a given azimuth, and the sun. We call Profile angle, related to a given azimuth, the angle formed by the profile plane and the horizontal plane. This is the characteristic angle describing the shadows limited by a horizontal line (wall, shed, balcony).

Reference year: year value

Reference year
The format protocol specifies a "reference year". i.e. data are assumed to be regular, from January 1st to December 31th (not leap year). In this case the program does not read the date on the file, and the Year number must be supplied by the user 205

Chapter 9 Glossary

Glossary
(from 1981 to 2060).

Regulator voltage switch

Regulator voltage swich
Some commercial regulators offer a swichable battery voltage (usually 12/24V). The right switch position is not a characteristics of the regulator component, and will be chosen by the program according to the battery pack, before the simulation process. Checks on voltage compatibility - as well as input and output maximum allowed currents - are performed before the simulation and will produce a warning when not matching.

Series Resistance (PV module)

PV-module series resistance
The series resistance Rs is not directly measurable on the PV-module. It's effect is combined with the I/V diode model (exponential) slope: the lower Rs resistance, the sharpener I/V characteristics toward high voltages. Rs is one of the 4 unknown related (coupled) parameters when adjusting the one-diode model.

Reverse characteristics (PV cell)

Reverse characteristic factor
Empirically, the behaviour of the cell's characteristics under reverse polarisation is quadratic with the applied voltage. This results is from our own measurements, and confirmed in ref [15]: Irev = IL + Brev (V + RsI) for V < -RsI. This expression could be valid till the avalanche zone (Zener), but in reality the dissipation in the cell which varies as the cube of the reverse voltage reaches a destructive limit well before this elbow.

Summer Time - Daylight Savings

The internal PVsyst meteo data - in Legal Time - don't take the "Summer Time" hour shift (also called "Daylight Savings Time") into account: the hour is related to the original Time Zone (usually "Winter Time") all over the year. When importing meteo data, the source data could include such a shift, which has to be corrected. The shifting dates are never fixed, but obey to rules which may be different in some regions of the world. In the present state, PVsyst only "knows" the rules applied in Europe: the passage to summer time arises on the last Sunday of March, and to Winter time on last Sunday of October. For other conventions, or modifying the European standard, the user will be invited to define these dates at each conversion, as they will depend on the year. NB: Please carefully check the results on daily plots after conversion, i.e. the shift of the hourly data by respect to the Clear Sky model !

STC
The Standard Test Conditions for the specification of PV modules are normalised operating conditions when testing the module. They are defined as: 1000 W/m Irradiation, 25C Module temperature, AM 1.5 Spectrum, i.e. a normalised solar spectrum corresponding to the crossing of 1.5 times the "normal" atmosphere (vertical air mass at the sea level).

Time Zone
The Time Zone is the difference between the local Legal time, and the UTC or GMT UTC = Coordinated Universal Time, GMT = Greenwich Mean Time. PVsyst usually uses the time zone corresponding to the Winter local time. The summer time is one more unit. For example in Europe, the Winter time corresponds to GMT+1, when the summer time is GMT+2.

Chapter 9 Glossary

206

Glossary
You have a map of the Time Zones over the world at http://www.worldtimezone.com/

Transverse Incidence Angle

Let us call Profile Plane the plane passing through a horizontal line perpendicular to a given azimuth, and the sun. We call Transversal incidence angle the angle between the normal to a tilted plane of this given azimuth, and the corresponding profile plane. This is useful for mutual shadings of sheds or sun-shields calculations.

Wind Velocity (Synthetic generation)

The wind velocity values are not converted into hourly values, during the Synthetic generation, as we don't know any general rule for doing this and we don't have any algorithm at disposal . Therefore the simulation process will use the monthly value. Nevertheless the wind velocity is only involved in the estimation of the Array Loss Factor (U = Uc + Uv Vw [W/mk]) for the determination of the array temperature. Therefore the wind velocity has only a little incidence on the PV array production during the simulation process. Currently we don't avail of validations of the simulation results by third parties. Finally few people really perform results analysis by respect to their real data, or they don't make them public. This is not easy because: - The simulation results depend on many parameters, which may be set at any value in order to get the expected result... (for example: PV module quality loss, or mismatch, or soiling), - The real meteo values when running are rarely known (or sometimes not recorded with sufficient care) and operating parameters measurements are also subject to errors. - The real performances of the components used (especially the PV modules) is rarely checked in detail at the installation time. For getting reliable conclusions, the measurement conditions and the validation process should be clearly defined. Namely comparisons between measurements and simulation should be performed in hourly values. We present here validations performed with old versions of PVsyst (1996) on 7 Swiss installations. The yearly power was predicted with an annual accuracy of the order of +/- 5%, except with an installation involving amorphous modules (which were not well modelled in this early version). Recently we analysed very roughly several plants in Geneva, over 2-3 running years, usually designed with PVsyst using the straightforward simulation (i.e. with all "default" loss values), and renormalising the results to the real monthly irradiation the yearly results were within +/- 5%. We also analysed the data of a 10 kWc system of amorphous modules, and closely compared them to the simulation. The conclusion is that the calculation procedures are relatively reliable (within 1-2% MBE accuracy over one year). PVsyst tries to use the best models (or the most suited) for simulating each part of the system, and taking each behavior into account. But the parameter you put in the simulation are the main source of uncertainty (mainly the meteo data, usually known at 5-10%, and also the real behaviour of the PV modules by respect to the specifications). For the modules, PVsyst uses a PV model (experimentally studied by myself) with a very good accuracy, provided you put the good parameters... The meteo annual variability is around +/-5%, but the last few years are sometimes considered exceptional. Therefore any production warranty should always be done under condition of renormalizing the real results according to actual meteo data (which become more and more available from satellites data, but often not free).

207

Chapter 9 Glossary

Validations
Validations of old versions of the program
The seven tested grid-connected installations

10

We have carried out detailed validations of the programme, using the data from 7 grid-connected systems, whose detailed characteristics appear on table 1. These systems have been chosen, on the one hand for the quality of gathering data, and on the other hand as an attempt to represent a variety of different situations: sizes of 0.5 to 100 kWc, fields in sheds or integration (roofing, facade), types of collectors (Si-mono, Si-poly, or amorphous), types of climates (plain, mountain, etc.), possible near shadings. The validations have to be carried out carefully using measured data, with an hourly or sub-hourly time step, over sufficient periods (one year). According to the measurement's parameters available, we have tested the various stages of the simulation, giving special attention to those which involve the most delicate physical models.

Installation

N 13

Marzili

LESOSheds
Lausanne Sheds 45 South 12 kWc 111.6 m2 Solarex MSX 60 60 Wc 55.1 Wc TISO SI-poly

SIG

EIV

LESO-LRE

LESOUSSC
Lausanne Demosite 28 South 0.45 kWc 8.2 m2

Site

(Sw itzerland)

Domat-Ems Anti-noisew all 45 25 East 104 kWc 967 m2 Kyocera LA361 J48 48 Wc 48 Wc Ispra SI-poly

Berne Sheds 35 37 East 25.6 kWc 170 m2 BP Solar BP495Saturn 95 Wc 88 Wc Ispra SI-mono -0.9 % 3.2 % 7.8 %

Genve Sheds 35 9 East 7.6 kWc 61.5 m2 Arco-Solar M55 53 Wc 50.8 Wc Ispra SI-mono

Sion Sheds 45 South 3.2 kWc 31.7 m2

Lausanne Facade 90 South 3 kWc 28.6 m2

Field: type Tilt angle Azimuth Installed pow er Field area Collectors: manuf. Type Nominal Pow er STC Measured Pow er STC Measurements Technology Irrad. transposition: MBE (transpos.meas.) RMSE (daily val.) RMSE (hourly val.) Coll. Tem perature m odel Wind velocity measurement K factor (input param.) MBE (simulmeasure) RMSE (hourly val.) PV-Field DC energy Sim ul. Base: MBE (simulmeas.) RMSE (daily values) RMSE (hourly values) System AC output MBE (simulmeas.) RMSE (monthly

Photow att Flagsol USSC (USA) BPX 47500 Optisol/LESO 48 Wc 250 Wc 40.3 Wc 17 Wc TISO Manufacture LESO SI-poly r a-Si:H SI-poly tandem 9.3 % 5.5 % 10.4 % -11.3 % 7.7 % 11.4 %

2.8 % 5.5 % 11.7 %

-6.0 % 9.6 % 15 %

-2.2 % 2.9 % 5.1 %

No 29 W/m2K -0.3 C 2.1 C

No 29 W/m2K -0.5 C 1.5 C

No 29 W/m2K -0.03 C 2.7 C

Yes 20 + 6 vvent 0.8 C 2.1 C

No 29 W/m2K -0.7 C 3.8 C

No 13 W/m2K -0.2 C 3.7 C

No 23 W/m2K 0.0C 2.8 C

Plane irrad. 5.6 % 8.7 % 11.0 %

Plane irrad. 1.0 % 10.0 % 15.5 %

(field #3) Plane irrad. -0.7 % 2.2 % 5.2 %

Horiz. irrad. 0.7 % 5.0 % 9.8 %

Plane irrad. 3.1 % 3.4 % 6.5 %

Horiz. Irrad. 1.4 % 10.8 % 17.7 %

Plane irrad -13.6 % 8.9 % 13.4 %

5.5 % 5.4 % 8.3 %

1.0 % 4.5 % 9.9 %

-0.7 % 1.1 % 2.4 %

1.9 % 1.7 % 5.3 %

2.8 % 1.2 % 3.3 %

2.7 % 8.2 % 12.7 %

-12.8 % 7.5 % 8.5 %

Chapter 10 Validations

208

Validations
values) RMSE (daily values) RMSE (hourly values) 10.8 % 15.3 % 5.5 % 9.7 % 6.5 % 19.0 %

10
12.5 %

Table 1. - Summary of the 7 tested installations: comparisons between simulation and measurements. (A positive MBE indicates that the simulation overestimates the measured values) Incident irradiation models The first delicate stage is the treatment of the incident irradiation in the collector plane: it involves models for the estimation of the diffuse irradiation (from the global irradiation) and for transposition. These models have been otherwise studied [Perez et al], but proved to be the weakest link in the comparison process, with differences reaching more than ten percent for some of the data used. For the SIG installation at Geneva, the only site where measurements of the horizontal diffuse irradiation is available, the deviation does not exceed 2 to 3%. But other sites show differences going up to more than 10%. It should be emphasised that these results are highly dependent on the quality of the instruments used for the irradiance measurements (especially the calibration and sometimes the positioning). We supposed that the standard deviation in hourly values is a good indication of the performances of the model itself, and in this respect, the installation of Marzili (MBE=1% and hourly RMSE < 7.8%) confirms their validity. The mediocre results of the LESO can be explained by the considerable distance (several hundred meters) between the horizontal solarimeter and the measurement in the collector plane.

Fig 1a. - Measurement-Simulation Comparison for Incident Irradiation, Marzili installation, daily values

209

Chapter 10 Validations

Validations

10

Fig 1b. - Measurement-Simulation Comparison for Incident Irradiation, Marzili installation, hourly values Array temperature model The second model is the estimation of the temperature of the PV field using values of ambient temperature and irradiation. This temperature only acts as an auxiliary parameter in the calculation of the electrical production of the field, and its specification is not critical. Our model, resulting from a simple thermal balance, gives remarkable results. By using the default value suggested by the programme (k=29W/mk) for all the collectors without back-covering, and by adapting this value for the integrated installations, we obtain, in all cases, errors lower than 1C, with hourly dispersions of 2 to 4C at the most.

Chapter 10 Validations

210

Validations

10

Fig 2. Comparison of the measured temperature with respect to the model Marzili Installation. Array output calculation The electrical output, measured at the collector array terminals (DC energy), is calculated by the simulation on the basis of the incident irradiation (given the shading and non normal incidence corrections), the temperature of the modules, and the collector model (operating at MPP), keeping in mind the ohmic losses of the wiring and the module's mismatch. The excellent results obtained especially in LESO-sheds, the SIG or at the EIV, show that these models work perfectly, at least for silicium crystalline modules.

211

Chapter 10 Validations

Validations

10

Fig. 3a. - Simulation-Measurement Comparison for the output of the PV field (LESO-sheds installation, daily values)

Fig. 3b. - Simulation-Measurement Comparison for the output of the PV field (LESO-sheds installation, hourly values) Amorphous tandem technology The LESO-USSC installation, equipped with a:H-SI tandem amorphous collectors, has shown one of the limits of the collector's model used. Table 1 shows that the global results of the field are clearly underestimated by the

Chapter 10 Validations

212

Validations

10

simulation. Fig. 4 shows that the values with strong beam are very well reproduced (MBE=0.5 %), while the points of pure diffuse irradiation are clearly under-evaluated. This is attributed to the fact that this type of collector is more sensitive to diffuse irradiation, which cannot be reproduced by our model which does not take the spectral distribution of the incident irradiation into account.

Fig 4a. - Comparisons for the amorphous collectors of the LESO-USSC, strong beam

Fig 4a. - Comparisons for the amorphous collectors of the LESO-USSC, purely diffuse irradiation Inverter modelling 213

Chapter 10 Validations

Validations

10

Inverter's operation modelling does not give any special problems, except the exact determination of the power efficiency profile: the programme offers a specific tool to superpose the characteristic of an existing model (present in the component library) on the measured input/output data of the inverter. The exact profile corresponding to the data can then be adjusted interactively by the user, and it can be assumed that the residual errors of the simulation are not imputable to the calculation model of PVSYST, but only to the inaccuracy of the parametrisation of this inverter efficiency.

Fig 5a. - Inverter response, with standard available inverter specification (Marzili).

Fig 5b. - Inverter response, after manual adjustment (Marzili).

Chapter 10 Validations

214

Validations

10

Conclusion With the exception of amorphous collectors, we can assess that the PVSYST software is capable of simulating a large number of different grid-connected systems with excellent accuracy. By independently testing each of the algorithms, we have tried to identify the uncertainties related to measurement and parameter's determination, and those inherent to the modelling. Finally, it can be stated that the accuracy of the global results of the simulation is of the order of 2 to 3% (MBE). However, the specificity of the PVSYST programme will rather be the detailed and comparative study of special disturbing effects, and in this use, the quality of the different models allow us to hope for much higher precision. Stand-alone systems, and particularly the battery model, have not been validated up to now. We hope to find properly measured data, recorded in hourly values, to further analyse the software accuracy. General, system simulation Frank Kreith, Jan F. Kreider Principles of Solar Engineering McGraw-Hill, Hemisphere Publishing Corporation, 1978, ISBN 0-07-035476-6 John A. Duffie et W.-A. Beckman. Solar Engineering of Thermal process. John Wiley and Sons, N-Y. 2nd ed., 1991. P. Schaub, O. Guisan, A. Mermoud. GAP/CUEPE, Universit de Genve. Etude de l'installation photovoltaque de 7.7 kW des SIG, Rapport final. OFEN, 3003 Bern. Juillet 1993. P. Schaub, O. Guisan, A. Mermoud. GAP/CUEPE, Universit de Genve. Etude de l'installation photovoltaque de 2.2 kW des TPG, Rapport final. OFEN, 3003 Bern. Juin 1992. A. Mermoud, O. Guisan. Logiciels pour systmes PV : Analyse des besoins suisses, cahier des charges et proposition de structure. GAP/CUEPE, Universit de Genve / OFEN, 3003 Bern. Octobre 1992.. A. Mermoud, O. Guisan. Logiciels pour systmes PV : Recensement et valuation des logiciels de simulation photovoltaque existants. GAP/CUEPE, Universit de Genve / OFEN, 3003 Bern. Juillet 1993. Guidelines for Assessment of Photovoltaic Plants. Document A : Photovoltaic System Monitoring. Document B : Analysis and presentation of Monitoring Data. CEC/JRC Ispra, Issue 4.1, june 1993. Meteo and Irradiance models R. Perez, P.Ineichen, R. Seals, J. Michalsky, R. Stewart. Modeling Daylight Availability and Irradiance Component from Direct and Global Irradiance. Solar Energy 44, no 5, pp 271-289, 1990. R. Perez, P. Ineichen, E.L. Maxwell, R.D. Seals, A. Zelenka : Dynamic Global to Direct Irradiance Conversion Models. ASHRAE Transactions, Vol 98, Part I, #3578, 1992. P. Bremer, S. Kunz, R. Voltz. METEONORM . Vol 1 (Thorie), Vol 2 (Donnes), Vol 3 (Procdures de calcul), Vol 4 (Sources de donnes). Office Fdral de l'Energie, Berne (1986).

215

Chapter 10 Validations

References
Jan Remund, Esther Salvisberg, Stefan Kunz. METEONORM'95 : Energie solaire et Mtorologie. Manuel du concepteur - Notions de base, et logiciel METEONORM V2.0. Meteotest and Office Fdral de l'Energie, Bern (1995). METEONORM Version 3.0 (1997). METEONORM Version 4.0 (). METEONORM Version 5.0 (2004). METEONORM Version 6.0 (June 2007). www.Meteotest.ch www.Meteonorm.com David Bloom, Sara Dagher, Stuart Frye, Qiumin Hu, Marianne Baldwin, Brett Gouviea, John Duffy, Bill Berg. William Beckman and Ronald Stewart. International Solar Irradiation Database, Version 1.0 (Info diskette). University of Massachusetts, Lowell Photovoltaic program, USA

11

R.J. Aguiar, M. Collares-Pereira and J.P. Conde. Simple Procedure for Generating Sequences of Daily Radiation Values Using a Library of Markov Transition Matrices. Solar Energy Vol 40, No 3, pp 269-279, 1988. R.J. Aguiar, M. Collares-Pereira. TAG : a Time-dependent, Autoregressive, Gaussian Model for Generating Synthetic Hourly Radiation. Solar Energy Vol 49, No 3, pp 167-174, 1992. J.-L. Scartezzini, M.-N. Ferguson, F. Bochud. LESO-EPFL Lausanne. Compression of Multi-Year Meteorological Data. OFEN, 3003 Bern, Final Report, 1990. B. Molineaux, P. Ineichen, Impact of Pinatubo Aerosols on the seasonal trends of Global, Direct and Diffuse Irradiance in two Northern Mid-Latitude Sites. Solar Energy 58, no 1-3, pp 91-101, 1996. P. Ineichen, R. Perez, A new airmass independent formulation for the Linke turbidity coefficient, Solar Energy 73, no 3, pp151-157, 2002. P. Ineichen, A broadband simplified version of the Solis clear sky model, Solar Energy 82, pp 758-762, 2008. P. Ineichen, Interannual variability and global irradiance evaluation, 2010, Research report of the Institut of the Environnemental Sciences, University of Geneva. Available on www.pvsyst.com P. Ineichen, Five satellite products deriving beam and global irradiance validation on data from 23 ground stations (IEA). 2011, Research report of the Institut of the Environnemental Sciences, University of Geneva. Available on www.pvsyst.com P. Ineichen, Global irradiance on tilted and oriented planes: model validations, 2011, Research report of the Institut of the Environnemental Sciences, University of Geneva. Available on www.pvsyst.com P. Ineichen, Global irradiation: average and typical year, and year to year annual variability, 2011, Research report of the Institut of the Environnemental Sciences, University of Geneva.

Chapter 11 References

216

References
Available on www.pvsyst.com PV modules behaviour and modelling J.A. Roger and C. Maguin Photovoltaic Solar Panels Simulation Including Dynamical Thermal Effects. ?????, 1981. M. Camani, D. Chianese, S. Rezzonico Centrale di Prova per componenti PV. TISO, c/o STS, CH-6952, Canobbio. Rapporto Finale 1990-1993, OFEN-BEW, Bern.

11

M. Camani, N. Cereghetti, D. Chianese, S. Rezzonico DB TISO, PV modules: A) Database permanently updated, about 830 modules referenced (1998), available from internet on the site below. B) Booklet no 3, november 1998, extract. TISO LEEE Scuola Universitaria professionale, 6952 Canobbio Switzerland. Tel +41 91 940 47 78 - Fax +41 91 942 88 65 e-mail [email protected] http://www.supsi.ch/home.html Thin film PV modules behaviour and modelling R. Gottschalg, M. Rommel, D.G. Infield, M.J. Kearney. Test of parametric models for amorphous silicon modules and cells. 2nd World Conference on Photovoltaic Energy Conversion, Vienna, 1998. J. Mertens, J. M. Asensi, C. Voz, A. V. Shah, R. Platz, J. Andreu. Improved equivalent circuit and Analytical Model for Amorphous Silicon Solar Cells and Modules. IEEE Transactions on Electron Devices, Vol 45, No 2, Feb 1998. M.J. Holley, R. Gottschalg, A.D. Simmons, D.G. Infield, M.J. Kearney. Modelling the performance of a-Si PV systems. 16th European Photovoltaic Solar Energy Conference, Glasgow, May 2000. J. Mertens, J.M. Asensi, C. Voz, J. Andreu. Assessing the outdoor performance of single junction and double junction amorphous silicon solar modules. 16th European Photovoltaic Solar Energy Conference, Glasgow, May 2000. T.R. Betts, R. Gottschalg, D.G. Infield. Spectral Irradiance Correction fo PV system Yield Calculations. 19th European Photovoltaic Solar Energy Conference, Paris, June 2004. H.Al Buflasa, T.R. Betts, R. Gottschalg, D.G. Infield. Modelling the Effects of Spectral Variations on the performance of Amorphous Silicon Solar Cells. 19th European Photovoltaic Solar Energy Conference, Paris, June 2004. A. Mermoud Conception et Dimensionnement de Systmes Photovoltaques: Introduction des modules PV en couche minces dans le logiciel PVsyst. Projet SIG-NER, Rapport Final - www.cuepe.unige.ch (Mai 2005). Available on www.pvsyst.com A. Mermoud, T. Lejeune, Performance assessment of a simulation model for PV modules of any available technology Poster paper presented at the 25-th European Photovoltaic Solar Energy Conference and Exhibition (EU 217

Chapter 11 References

References
PVSEC), Valencia, Spain, 6-10 september 2010 Available on www.pvsyst.com Shadings

11

A. Mermoud, T. Lejeune, Partial shadings on PV arrays: by-pass diode benefits analysis", Mermoud, A. & Lejeune, T., 2010, Poster paper presented at the 25-th European Photovoltaic Solar Energy Conference and Exhibition (EU PVSEC), Valencia, Spain, 6-10 september 2010 Available on www.pvsyst.com Inverters H. Hberlin, H.R. Rthlisberger : Vergleichsmessungen an Photovoltaik-Wechselrichtern. Schlussbericht des BEW-Projectes EF-REN(89) 045, OFEN/BEW, 3003, Bern, 1993. J.M. Servant, J.C. Aguillon : Essais d'onduleurs PV raccords au rseau EdF. GENEC-Cadarache, 1993. Batteries C .M. Shepherd Design of Primary and Secondary Cells. and: An equation Discribing Battery Discharge. Journal of the Electrochemical Society, vol 112(7), July 1965, pp 657-664. J.B. Copetti, F. Chenlo and E. Lorenzo : Comparison between Charge and Discharge Battery Models and Real Data for PV Applications. 11th European Photovoltaic Solar Energy Conference, Montreux, Oct 1992, pp1131-1134. J.B. Copetti, F. Chenlo Internal Resistance Characterization of Lead-Acid Batteries for PV Rates 11th European Photovoltaic Solar Energy Conference, Montreux, Oct 1992, pp1116-1119. Watsun-PV - User's Manual and Program Documentation Watsun Simulation Laboratory, University of Waterloo, Waterloo, Ontario N2L 3G1, 1992. Bopp, Gabler, Sauer, Jossen, Hhe, Mittermeier, Bchler,Sprau, Willer, Wollny. A Systematic Effort de Define Evaluation and Performance Parameters and Criteria for Lead-acid Batteries in PV Systems. 13th European Photovoltaic Solar Energy Conference, Nice, Oct 1995. Pumping L. Navarte, E. Lorenzo, E. Caamao: PV pumping analytical design and Characteristics of Boreholes. Solar Energy 68, no 1, pp 49-56, 2000. J.A. Roger: Theory of the direct coupling between DC Motors and PV solar Arrays. Solar Energy 23, pp 193-198, 1979. Ziyad Salameh, A. K. Mulpur, F. Dagher: Two-stage Electrical Array Reconfiguration controller for PV-powered Water Pump. Solar Energy 44, no 1, pp 51-56, 1990. H. Suehrcke, J. Appelbaum, B. Reshef: Modelling a Permanent Magnete DC Motor/Centrifugal Pump Assembly in a PV system. Solar Energy 59, nos 1-3, 1997. M. Alonso-Abella, E. Lorenzo, F. Chenlo: 218

Chapter 11 References

References
PV Water Pumping Systems based on Standard Frequency Converters. Progress in Photovoltaic Research and Applications 2003, 11:1-13. M. Alonso-Abella, F. Chenlo, J. Blanco, D. Manso: Use of Standard Frequency Converters in PV pumping systems. 2nd World Conf. and Exhibition on PVSEC, Vienna, 1998. M. Alonso-Abella (CIEMAT), Detailed measurements of several pumps, 2004. Private communication, 2004. Andr Mermoud: Pump behaviour modelling for use in a general PV simulation software. 19th European PVSEC, Paris, France, June 711, 2004. Available on www.pvsyst.com Andr Mermoud: Pumping System Sizing and Modelling Tool. 19th European PVSEC, Paris, France, June 7-11, 2004. Available on www.pvsyst.com

11

219

Chapter 11 References

 PVsyst SA 1994-2012. All rights reserved. Product and company names mentioned in this manual may be trademarks or registered trademarks of their respective companies. Mention of third-party products is for informational purposes only and constitutes neither an endorsement nor a recommendation. The author assumes no responsibility w ith regard to the performance or use of these products. All understandings, agreements, or w arranties, if any, take place directly betw een the vendors and the prospective users. Every effort has been made to ensure that the information in this manual is accurate. The author is not responsible for printing or clerical errors. The product described in this manual incorporates copyright protection technology that is protected by method claims of certain U.S. patents and other intellectual property rights. This user manual w as created w ith Help & Manual.

Exporting Projects

176

3
3Tiers 133

F
Far shadings 37, 38, 39 Files 168, 169, 175, 176 Financial balance 107

A
Albedo 45, 178, 179, 200 Array behaviour 31, 36, 37, 163, 186 Array losses 83, 84, 86, 87, 97, 203, 204 Autonomy 200

G
Generator 156 Generic default regulator 161 Genset 156 Geographic site 108, 109, 114, 120, 122 Grid current 202 Grid inverters 60, 150, 152, 153, 154, 202 Grid voltage standards 63 Grid-connected system 18, 58, 60, 61, 93

B
BackTracking 51 Back-up generator 156 Batteries 154, 192 Battery capacity sizing 200 Building objects 49, 52

C
Commercial 146, 163 Compatibility and Troubles 14 Components 145, 150, 154, 156, 159 Concentrating 35 Concentration 34, 51 Converter 160 Copy Data 175

H
Head Units 73 Hidden Parameter 167 Historical evolution of the software Horizon 38, 39 Hourly Meteo 117 Hydraulic friction losses 74 Hydraulic power/energy 73 3

D
Data cuts 166 Databases 143, 148, 149, 154, 155, 167 Meteo Database 113 DC-DC Converter 160 DC-grid line resistance 65 DC-grid system 64, 94 Deep Well Pumping System 68, 69 Design Reference Year 116, 125 Detailed simulation 22, 24 Diffuse 45, 108, 115, 178, 179, 180, 201 Directories 168, 169, 175, 176

I
IAM 83, 86 Importing data 137, 139, 164, 175 Importing Meteo 126, 127, 128, 129, 130, 131, 133, 134, 135, 136, 138 Importing Projects 176 Incidence Angle Modifier 83, 86 Internationalization 167 Inverters 60, 150, 152, 153, 154, 202 Irradiance loss 203

L
Languages 167 License 14, 16, 17 Load 54, 55, 56, 57 Load profile 65 Losses 83, 84, 87, 97, 203

E
Economic evaluation 22, 106, 107 Exporting data 2, 95, 177

221

Index

Index

Pump model

156

M
Measured data 96, 164, 165, 166 Measured data analysis 1 Measurement tool 54 Meteo 108, 109, 114, 115, 120, 122, 137, 143, 179, 180, 181 Meteo Data 115, 116, 120, 122, 125, 126, 127, 128, 129, 130, 131, 133, 134, 135, 136, 138, 202 Meteo Data Format 138 Meteo Database 116, 120, 122, 125 Meteo quality 117 Meteonorm 126 Mismatch 83, 87 Models 44, 68, 145, 150, 154, 159, 178, 179, 180, 181, 182, 192 Module quality loss 86 Module strings 46, 47 MPPT Converter 160

Pump model 156, 195, 197, 198, 199 Pump 157, 158, 159 Pumping System Configurations 67, 74, 75, 76, 77, 78, 80, 161, 162 Pumping Systems 21, 66, 67, 68, 69, 70, 71, 72, 74, 80, 95, 103, 161, 162 PV array behaviour 31, 36, 37, 143, 163, 186 PV components 145, 150, 154, 156, 159 PV model 182, 185, 186, 189, 190, 191 PV modules 145, 146, 147, 149, 182, 184, 185, 186, 189, 190, 191 PV plane 30

R
References 215 Registration 14, 16, 17 Regulation loss 83 Regulators for stand-alone systems Results 92, 93, 95, 96, 106, 107 Reverse characteristics 185

159, 202

N
Near shadings 31, 32, 33, 37, 40, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54 NOCT 84

Satellight 129, 130 Screen 2 Shadings 31, 32, 33, 37, 38, 39, 40, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54 Ohmic loss 83, 87 Sheds 31, 32, 33, 48 One-diode model 182 Simulation 61, 92, 93, 94, 95, 96 Orientation 30, 205 Simulation variables 97, 98, 100, 101, 103, 105 Orientation optimisation 33 Solar geometry 143 SolarEdge 61 SolarGIS 133 Special characters 167 Partial shadings 37, 40, 44, 45, 46, 47, 48, 49, 50, 51, Stand-alone system 19, 20, 63, 94 52, 53, 54 Static depth 73 Plane orientation 30, 205 System design 1, 20, 29, 58, 60, 61, 63, 64, 66 Power Converter 160 System sizing 20, 29, 58, 60, 61, 63, 64, 66 Preliminary design 1, 17, 18, 19 Presizing 17, 18, 19 Pressure units 73 Printing 2, 177 Temperature behaviour 83, 84 Project 22, 24, 28, 29 Thermal loss 83, 84, 204 Project design 1, 20, 22, 24, 29, 58, 61, 63, 64, 66, 89, Thin film PV model 186, 189, 190, 191 92, 93, 94, 95, 96 Time label 141 Projects transfer 176 Time shift 117, 141 Pump Time Zone 206

Index

Index

222

Tools 1, 33, 36, 143, 179 Tracking 51

U
Uninstall 168 Updates 167 Updating database 175 User's needs 54, 55, 56, 57

V
Validations 208 Voltage optimisation 163

W
Warter needs 72 Windows Rights 169 Wiring loss 87

223

Index

Index