DNV BOS COMPARISON REPORT - TEXAS

Technical Note
LONGi Solar Technology (U.S.) Inc

Document No.: 10293644-HOU-T-01-G

Date: 14 September 2021

SAFER, SMARTER, GREENER

Contents
1 INTRODUCTION ................................................................................................................................................................. 1
Background ................................................................................................................................................................ 1

2 SCENARIOS ....................................................................................................................................................................... 1

3 PROCESS ........................................................................................................................................................................... 1
Design of the scenarios .............................................................................................................................................. 2

4 RESULTS ............................................................................................................................................................................ 3
Details of the results ................................................................................................................................................... 7

5 SUMMARY .......................................................................................................................................................................... 7

6 REFERENCES .................................................................................................................................................................... 7

Appendices
APPENDIX A – SCENARIO LAYOUTS

DNV Energy USA Inc. Page v

1 INTRODUCTION
LONGi Solar Technology (U.S.) Inc (“LONGi” or the “Customer”) has engaged DNV Energy USA Inc. (DNV) to provide
balance of system cost comparison services for three different modules and three separate layout configurations (“LONGi
BOS” or the “Project”). These scenarios were studied for hypothetical projects in Texas, United States.

Background
Large format modules are a newer module technology currently being developed by module manufacturers including
LONGi. These modules are larger than standard 72 or 144 cell modules and also have different electrical characteristics.
Due to these differences, the standard assumptions for how to optimally design a PV array may not hold true. Numerous
design factors are affected by the change to large format modules which also affects the overall balance of system costs for
a project. DNV took these differences into account and ran multiple scenarios to find the cost differences between three
different large format modules chosen by LONGi along with three different racking options.

2 SCENARIOS
Table 2-1 describes the modules that were used for the cost comparison analysis.

Table 2-1 Modules studied

Module Power (W) Voc (V) Isc (A) Length Width Thickness Area (m2) Weight (kg)
Type (mm) (mm) (mm)

182-72c 540 49.5 13.85 2256 1133 35 2.556048 32.3

210-55c 545 37.7 18.3 2384 1096 35 2.612864 32.6

210-60c 595 41.5 18.36 2172 1303 35 2.830116 35.3

The following scenarios were studied:

Scenario 1: A single-axis tracker system with one module high in portrait orientation (1P). Three layouts were
designed: one for each module type.

Scenario 2: A fixed tilt racking system with two modules high in portrait orientation (2P). Two layouts were designed
using the 182-72c and 210-60c modules.

Scenario 3: A fixed tilt racking system with four modules high in landscape orientation (4L). Two layouts were
designed using the 182-72c and 210-55c modules

3 PROCESS
Costs for various aspects of the project were calculated based on industry research, equipment manufacturer quotes, and
DNV experience. DNV used an approximately 3.7 MWdc power block as a basis for its calculations, designing multiple

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page 1

www.dnv.com
layouts for the racking and module scenarios. Once a layout was designed, DNV then found design parameters necessary
to calculate the balance of system costs requested by LONGi. All costs are calculated in 2021 United States Dollars.

Wood Mackenzie is an industry research company that aggregates data from hundreds of solar projects throughout the
United States and the world to come up with typical costs for specific project parameters. These costs can be filtered by
state, project size, and equipment type to accurately reflect the costs of a particular project design. DNV relied on these
figures [1] to find costs for particular project parameters and supplemented this information with quotes from equipment
manufacturers to come up with accurate final costs.

Design of the scenarios

DNV designed the scenario layouts based on a relatively flat, rectangular piece of land. The following sections describe
other factors considered.

3.1.1 Ground cover ratio

DNV determined the ground cover ratio for tracker and fixed tilt systems that would be appropriate for projects in Texas and
Spain. The spacing of the fixed tilt rows ensured no interrow shading between the hours of 9 am and 3 pm throughout the
year. As module dimensions vary between the studied modules, different row spacings are used for each scenario. DNV did
not optimize the tracker layouts for energy production as the primary goal of this analysis is a comparison between module
types and their associated balance of system. Row spacing is based purely on what would be reasonable for the area.

3.1.2 Racking material

DNV notes that the thickness and length of racking piles is highly dependent on the specific project site conditions.
Therefore, DNV has not calculated the tonnage of material per rack and has instead focused on the total number of piles
based on DNV’s experience for each racking type. Further, fixed tilt racking designs can vary greatly in how the support
structures are designed, with some utilizing multiple smaller supports as opposed to one larger one. DNV has assumed the
larger support option for this analysis.

3.1.3 Racking structure

The length and number of racks was chosen so that the layout of the site was rectangular shaped, with the optimal shape
being square. String lengths were also considered so that all strings were within a single row, with no strings being split
between rows.

3.1.4 DC combiner boxes

DC combiner boxes were placed strategically throughout the array to minimize the amount of string homerun wiring. The
number of strings per combiner box was determined based on typical combiner box ampacity ratings, the project layout, and
number of strings per rack.

3.1.5 Module stringing strategy

For modules in portrait orientation, a skip stringing design was used to minimize the length of dc homeruns to the combiner
boxes. When modules are laid out in landscape orientation, module leads are not long enough to allow skip stringing, so
standard stringing methods were used, resulting in longer module dc homeruns lengths. The number of modules per string
was determined based on the maximum voltage of the modules and inverter and for constructability.

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page 2

www.dnv.com
3.1.6 Inverter
Inverters were placed within the arrays as is typically seen in utility scale projects in the industry. Depending on the array
orientation, access roads may bisect a given inverter’s associated modules or a small segment of the array may be carved
out for the inverters. A twenty-foot-wide access road was assumed. The model of the inverter chosen for these designs is
capable of accepting the various dc capacities in each scenario.

3.1.7 Conductor sizing

Conductor sizes were chosen to limit the maximum voltage drop to less than 2%, limit the average voltage drop to less than
1.5%, and meet necessary ampacity requirements. As the 210 modules have higher currents than the 182 module, a larger
dc string conductor size was used for these conductors to keep the dc losses relatively similar to those of the 182 module.
DC conductors were assumed to be routed to dc combiner boxes and then buried underground and routed to the inverter
location.

4 RESULTS
The tables below list the total equipment necessary for each 3.7 MWdc block design in Texas along with their associated
cost per watt.

Table 4-1 Scenario 1 equipment totals and costs

Parameter Scenario 1: 1P Tracker

Module type 182-72c 210-55c 210-60c

Array design (27 modules per (35 modules per (32 modules per
string x 3 strings string x 2 string per string x 2 strings per
System Design
per rack) x 84 rack )x 98 racks rack) x 98 racks
racks

DC capacity (MW) 3.67416 3.7387 3.73184

Piles per rack 12 10 11

Pile spacing (m) 8.1 8.1 8.1

Mounting system Total number of piles 1008 980 1078

Racking Cost (¢/W) 7.27 7.50 7.50

Piles (¢/W) 2.35 2.29 2.52

PV string cable 2 way length (m) 21,302 14,329 14,850

PV string cable (#10 AWG Cu PV wire) 0.15 - -

(¢/W)
Cable and
($0.86/m)
combiner box
PV string cable (#8 AWG Cu PV wire) - 0.15 0.13
(¢/W)
($1.26/m)

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page 3

www.dnv.com
 Parameter Scenario 1: 1P Tracker

Combiner Box configuration 14 w/ 18 strings 14 w/ 14 strings 14 w/ 14 strings

Combiner Box (¢/W) 0.43 0.42 0.42

DC homerun cable 2 way length (m) 1,920 2,270 2,150

DC homerun cable (750 kcmil Al) 0.42 0.48 0.46

(¢/W)
($7.97/m)

AC equipment Inverter (¢/W) 3.68 3.62 3.63

Inter-row spacing (m) 4.39 4.43 4.21

Land (m^2) 52,803 54,451 55,950

Land Land (¢/W) 2.35 2.42 2.49

Note: this represents total 20 year
lease payments

Civil work (¢/W) 6.55 6.76 6.94

Labor DC (¢/W) 1.58 1.83 1.59

Labor Module (¢/W) 0.9 0.9 0.9

Labor
Labor Racking (¢/W) 1.7 1.80 1.80

Labor Foundation (¢/W) 1.79 1.74 1.92

Total BOS Totals (¢/W) 29.13 29.91 30.32

Table 4-2 Scenario 2 equipment totals and costs

Parameter Scenario 2: 2P Fixed Rack

Module type 182-72c 210-60c

Array design (27 modules (32 modules

per string x 8 per string x 6
System Design strings per strings per
rack) x 32 rack) x 33
racks racks

DC capacity (MW) 3.73248 3.76992

Piles per rack 24 24

Pile spacing (m) 5.1 5.1

Mounting system Total number of piles 768 792

Racking Cost (¢/W) 4.07 4.20

Piles (¢/W) 1.02 1.06

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page 4

www.dnv.com
 Parameter Scenario 2: 2P Fixed Rack

PV string cable 2 way length (m) 12,875 8,687

PV string cable (#10 AWG Cu PV wire) (¢/W) 0.09 -

($0.86/m)

PV string cable (#8 AWG Cu PV wire) (¢/W) - 0.09

($1.26/m)

Combiner Box configuration 16 w/ 16 16 w/ 12

Cable and combiner box
strings strings, 1 w/ 6
strings

Combiner Box (¢/W) 0.48 0.51

DC homerun cable 2 way length (m) 3170 3130

DC homerun cable (750 kcmil Al) (¢/W) 0.68 0.66

($7.97/m)

AC equipment Inverter 3.63 3.59

Inter-row spacing 6.10 5.82

Land (m^2) 38,936 39,415

Land Land (¢/W) 1.73 1.75
Note: this represents total 20 year lease payments

Civil work (¢/W) 6.39 6.47

Labor DC (¢/W) 2.04 1.9

Labor Module (¢/W) 0.8 0.8

Labor
Labor Racking (¢/W) 1.45 1.50

Labor Foundation (¢/W) 1.02 1.06

Total BOS Totals (¢/W) 23.42 23.58

Table 4-3 Scenario 3 equipment totals and costs

Parameter Scenario 3: 4L Fixed Rack

Module type 182-72c 210-55c

Array design (27 modules per string (35 modules per string
System Design x8 string per rack) x 32 x 6 strings pre rack) x
racks 33 racks

DC capacity (MW) 3.73248 3.77685

Piles per rack 24 25

Mounting system
Pile spacing (m) 5.1 5.1

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page 5

www.dnv.com
 Parameter Scenario 3: 4L Fixed Rack

Total number of piles 768 825

Racking Cost (¢/W) 4.52 4.66

Piles (¢/W) 1.02 1.10

PV string cable 2 way length (m) 35,113 28,218

PV string cable (#10 AWG Cu PV wire) (¢/W) 0.25 -

($0.86/m)

PV string cable (#8 AWG Cu PV wire) (¢/W) - 0.29

($1.26/m)
Cable and
Combiner Box configuration 16 w/ 16 strings 16 w/ 12 strings, 1 w/ 6
combiner box
strings

Combiner Box (¢/W) 0.48 0.51

DC homerun cable 2 way length (m) 3500 3020

DC homerun cable (750 kcmil Al) (¢/W) 0.75 0.64

($7.97/m)

AC equipment Inverter 3.63 3.58

Inter-row spacing 6.08 5.88

Land (m^2) 39,176 40,473

Land Land (¢/W) 1.74 1.8

Note: this represents total 20 year lease
payments

Civil work (¢/W) 6.43 6.64

Labor DC (¢/W) 2.26 2.06

Labor Module (¢/W) 0.8 0.8

Labor
Labor Racking (¢/W) 1.45 1.50

Labor Foundation (¢/W) 1.02 1.10

Total BOS Totals (¢/W) 24.35 24.69

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page 6

www.dnv.com
 Table 4-4 Summary of scenario costs

Scenario 1: 1P Tracker Scenario 2: 2P Fixed Tilt Scenario 3: 4L Fixed Tilt

Module type 182-72c 210-55c 210-60c 182-72c 210-60c 182-72c 210-55c

Total BOS 29.13 29.91 30.32 23.42 23.58 24.35 24.69

(¢/W)

Details of the results

When calculating the differences in costs between each layout and scenario, DNV determined a baseline value to scale the
associated costs by averaging the totals for that particular parameter. In some instances, averages were calculated based
on the racking technology (fixed vs tracking) as certain aspects should not be compared across all scenarios.

4.1.1 Labor
DC labor entails the work needed to trench, install all cabling, and install combiner boxes. Module labor is the work needed
to install modules on the tracker system. Racking labor is the work to install the racking components aside from the piles,
which is covered in the foundation labor costs.

4.1.2 Civil work

This includes the costs of grading the land, water flow and flood plain management, and roads within the arrays. It is
dependent on the size of the land used by the project.

5 SUMMARY
DNV determined the balance of system costs for three different scenarios in Texas utilizing industry data, equipment
manufacturer quotes, and DNV’s experience. Layouts of each scenario were designed to find the total amount of equipment
needed for an approximately 3.7 MWdc block of modules. Results of the analysis show the differences in costs between
various design parameters. There is little variation in overall prices within each scenario, meaning the module choice has
little effect, but it can be seen that certain racking configurations lead to different overall prices. The variation in two string
versus three string trackers also produces a noticeable cost difference. DNV notes that the results should only be used as a
comparison between different design choices and are not meant to be indicative of the total costs to build a project. The
calculated costs are not inclusive of any work done on the ac electrical side of the project.

6 REFERENCES
[1] Cox, Molly, Wood Mackenzie’s Interactive US Solar PV System Cost Model- H2 2020

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page 7

www.dnv.com
APPENDIX A – SCENARIO LAYOUTS

A.1 Scenario 1: 182-72c 1P Tracker

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page A-1

www.dnv.com
A.2 Scenario 1: 210-55c 1P Tracker

A.3 Scenario 1: 210-60c 1P Tracker

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page A-2

www.dnv.com
A.4 Scenario 2: 182-72c 2P Fixed

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page A-3

www.dnv.com
A.5 Scenario 2: 210-60c 2P Fixed

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page A-4

www.dnv.com
A.6 Scenario 3: 182-72c 4L Fixed

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page A-5

www.dnv.com
A.7 Scenario 3: 210-55c 4L Fixed

DNV Document No. 10293644-HOU-T-01, Issue: G, Status: FINAL Page A-6

www.dnv.com
About DNV
We are the independent expert in assurance and risk management. Driven by our purpose, to safeguard life, property and
the environment, we empower our customers and their stakeholders with facts and reliable insights so that critical decisions
can be made with confidence. As a trusted voice for many of the world’s most successful organizations, we use our
knowledge to advance safety and performance, set industry benchmarks, and inspire and invent solutions to tackle global
transformations.