ZN BR Battery Tech

FINAL REPORT
Zinc Bromide Flow Battery Installation

for Islanding and Backup Power
ESTCP Project EW-201242
SEPTEMBER 2016
Ryan Faries
Raytheon
Distribution Statement A
This document has been cleared for public release
Page Intentionally Left Blank
This report was prepared under contract to the Department of Defense Environmental
Security Technology Certification Program (ESTCP). The publication of this report
does not indicate endorsement by the Department of Defense, nor should the contents
be construed as reflecting the official policy or position of the Department of Defense.
Reference herein to any specific commercial product, process, or service by trade
name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply
its endorsement, recommendation, or favoring by the Department of Defense.
Form Approved
REPORT DOCUMENTATION PAGE OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the
data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing
this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-
4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently
valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From - To)
09-18-2016 Final Report 06/2012 – 06/2017
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Ryan Faries W912HQ-12-C-0027
Raytheon 5b. GRANT NUMBER
2000 El Segundo Blvd
El Segundo, CA 90245 5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER

Ryan Faries, Raytheon EW-201242
Mick Wasco, MCAS Miramar 5e. TASK NUMBER
Tom Stepien, Primus Power
Bob Riel, Dynalectric 5f. WORK UNIT NUMBER
Bob Butt, NREL
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT
NUMBER
Raytheon
2000 El Segundo Blvd
El Segundo
9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

Environmental Security Technology Certification Program ESTCP
(ESTCP)
4800 Mark Center Drive, Ste 17D08 11. SPONSOR/MONITOR’S REPORT
Alexandria, VA 22350 NUMBER(S)
EW-201242
12. DISTRIBUTION / AVAILABILITY STATEMENT
13. SUPPLEMENTARY NOTES
14. ABSTRACT
This Environmental Security Technology Certification Program (ESTCP) effort
demonstrates the energy security and cost benefits of implementing a Zinc Bromide
(Zn/Br) Flow Battery-based Energy Storage System (ESS) at the Marine Corps Air
Station (MCAS) Miramar. The effort integrates an innovative Zn/Br Flow Battery
and Intelligent Power and Energy Management (IPEM) technologies with the existing
MCAS infrastructure, providing energy security and islanding capability, while
reducing costs.
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION 18. NUMBER 19a. NAME OF RESPONSIBLE PERSON
OF ABSTRACT OF PAGES Ryan Faries
a. REPORT b. ABSTRACT c. THIS PAGE 19b. TELEPHONE NUMBER (include area
code)
108
310-647-9719
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std. Z39.18
Table of Contents
1 INTRODUCTION .................................................................................................................. 1
1.1 BACKGROUND ............................................................................................................... 1
1.2 OBJECTIVE OF THE DEMONSTRATION.................................................................... 1
1.3 REGULATORY DRIVERS .............................................................................................. 2
2 TECHNOLOGY DESCRIPTION .......................................................................................... 4
2.1 TECHNOLOGY OVERVIEW.......................................................................................... 4
2.2 TECHNOLOGY DEVELOPMENT ................................................................................. 7
2.3 ADVANTAGES AND LIMITATIONS OF THE TECHNOLOGY .............................. 20
3 PERFORMANCE OBJECTIVES ........................................................................................ 22
3.1 ISLANDED DURATION ............................................................................................... 23
3.2 BUILDING LOAD REDUCTIONS ............................................................................... 24
3.3 SWITCHOVER TIME .................................................................................................... 25
3.4 PEAK SHAVING ............................................................................................................ 25
3.5 ESS ENERGY STORAGE CAPACITY ......................................................................... 28
4 FACILITY/SITE DESCRIPTION ........................................................................................ 29
4.1 FACILITY/SITE LOCATION AND OPERATIONS .................................................... 30
4.2 FACILITY/SITE CONDITIONS .................................................................................... 31
5 TEST DESIGN ..................................................................................................................... 32
5.1 CONCEPTUAL TEST DESIGN..................................................................................... 32
5.2 BASELINE CHARACTERIZATION ............................................................................ 38
5.3 DESIGN AND LAYOUT OF TECHNOLOGY COMPONENTS ................................. 39
5.4 OPERATIONAL TESTING............................................................................................ 44
5.4.1 Test Configuration .................................................................................................. 46
5.4.2 System Integration and Checkout Tests .................................................................. 47
5.4.3 Grid Tied Mode Tests ............................................................................................. 50
5.4.4 Islanded Mode Tests ............................................................................................... 52
5.5 SAMPLING PROTOCOL............................................................................................... 78
5.6 SAMPLING RESULTS .................................................................................................. 79
6 PERFORMANCE ASSESSMENT ...................................................................................... 80
6.1 Islanding Duration ........................................................................................................... 80
6.2 BUILDING LOAD REDUCTIONS ............................................................................... 83
ZnBr Flow Battery Installation for

Islanding and Backup Power ii EW-201242
6.3 SWITCHOVER TIME .................................................................................................... 84
6.4 PEAK SHAVING ............................................................................................................ 86
6.5 ESS ENERGY STORAGE CAPACITY ......................................................................... 88
7 COST ASSESSMENT .......................................................................................................... 91
7.1 COST MODEL ................................................................................................................ 91
7.2 COST DRIVERS ............................................................................................................. 95
7.3 COST ANALYSIS AND COMPARISON to fully functional system ........................... 95
8 IMPLEMENTATION ISSUES ............................................................................................ 99
9 References ........................................................................................................................... 100
APPENDICES ............................................................................................................................ 101
Appendix A: Points of Contact ............................................................................................... 101
Table of Figures
Figure 1-1: Energy goals from the Secretary of the Navy. ............................................................. 2
Figure 1-2: Image of the US electrical distribution system taken from the FEMA website. ......... 3
Figure 2-1: Schematic of a traditional Zn/Br cell with two electrolyte flow loops. ...................... 4
Figure 2-2: Schematic of Primus’ approach to a Zn/Br cell using a solide titanium electrode and a
single flow loop of electrolyte. ....................................................................................................... 5
Figure 2-3: Early rendering of ESS located at MCAS Miramar. .................................................... 5
Figure 2-4: Illustration of Primus EnergyPod system complete with Zn/Br cells, power electronics,
and thermal management. ............................................................................................................... 6
Figure 2-5: Isometric view of the construction plans. ................................................................... 7
Figure 2-6: Photo of installed system at MCAS Miramar. ............................................................. 7
Figure 2-7: Developmental timeline for Primus EnergyPod system. ............................................. 8
Figure 2-8: Photo of Sandia National Laboratory 3rd party testing of Primus Powers EngergyCell
at Hayward CA. .............................................................................................................................. 8
Figure 2-9: Summary of results of Sandia 3rd party testing. .......................................................... 9
Figure 2-10: One line diagram for the MCAS Miramar microgrid circuit for the ESTCP project.
....................................................................................................................................................... 11
Figure 2-11: One line diagram design for test setup at NRELs ESIF facility. ............................ 11
Figure 2-12: Detailed single line diagram implemented at NREL for ESIF testing. .................... 12
Figure 2-13: Photos showing the live testing at NRELs ESIF facility with MCAS Miramar and
USMC personnel........................................................................................................................... 13
Figure 2-14: Photos of the various equipment that is part of the NREL testing. The top left image
is of Primus Powers 760kVA Parker grid tied inverter that is the main element in Primus’ power
electronics subsystem. The bottom left is of a 480 to 208 transformer, the bottom second from the

Islanding and Backup Power iii EW-201242
left is Primus’ EnergyBlock controller which manages Primus AC & DC busses. The bottom third
from the left shows the capacitor bank within the Parker inverter. The right most image shows the
AC breaker for the Parker inverter and the local HMI display. .................................................... 14
Figure 2-15: Photos of the IPEM controller and ancillary equipment used during the NREL testing.
....................................................................................................................................................... 15
Figure 2-16: Energy capacity timeline and scale up from Primus Power since the November 2013
briefing. ......................................................................................................................................... 16
Figure 2-17: Results of the Peak Shaving Test during the FAT. Note the peak charge power is
140kW and the peak discharge power is 225kW. ......................................................................... 17
Figure 2-18: Results of the Energy Capacity testing during the FAT. Note the total charge duration
is 6hrs and 20 min. The peak discharge power is 250kW and the total discharge energy recorded
is 390kWh. .................................................................................................................................... 17
Figure 2-19: Energy Capacity as a function of Discharge Rate shown on the DC output of the
ESS................................................................................................................................................ 18
Figure 2-20: Oscilloscope screenshot showing the connected logics of the EnergyPod and
PowerBox. The inverter was started in islanding mode. The image shows the bus voltage being
regulated continuously. ................................................................................................................. 19
Figure 3-1: Detailed interconnect diagram for the Islanding test setup ........................................ 24
Figure 3-2: Graph showing a very simple hypothetical load profile. The purple line represents the
historical load profile. The green line represents the load profile using an ESS to charge at night
and discharge during the peak time of day ................................................................................... 26
Figure 3-3: SDGE Time of Use Rate time periods. ..................................................................... 27
Figure 4-1: Birds eye view of MCAS Miramar. .......................................................................... 29
Figure 4-2: Map and aerial image of MCAS Miramar. ................................................................ 30
Figure 4-3: Map of the installation site at MCAS Miramar near building 6311. ......................... 30
Figure 4-4: Rendering of the ESS installation site next to the P196 inverter room under the carport
PV panels. ..................................................................................................................................... 31
Figure 5-1: Interconnect diagram of Zn/Br installation at MCAS Miramar. ............................... 32
Figure 5-2: Average daily load profiles by month for building 6311 at MCAS Miramar........... 33
Figure 5-3: Average daily power output of the P196 carport PV subsystem. ............................. 34
Figure 5-4: Simulated load and power output profile for the ZnBr installation during islanded
operation. ...................................................................................................................................... 35
Figure 5-5: Load profile for building 6311 including real, reactive and apparent power as a function
of time. The data was sampled at 15 minute intervals. The power factor is plotted on the secondary
axis. ............................................................................................................................................... 36
Figure 5-6: Plot showing example of how peak shaving can change the load profile of a facility
as seen by the utility. ..................................................................................................................... 37
Figure 5-7: Daily load profiles for building 6311 during November 2015. ................................. 38

Islanding and Backup Power iv EW-201242
Figure 5-8: Average hourly load profile for building 6311 during November 2015. ................... 39
Figure 5-9: Birds eye view of MCAS Miramar site and layout of system components. ............. 39
Figure 5-10: Photos of the Primus ESS being delivered to MCAS Miramar. The EnergyPod need
to have a large crane in order to lift it off of the delivery truck. ................................................... 40
Figure 5-11: Construction drawing schematic of ESS location. .................................................. 41
Figure 5-12: 3D CAD image of the ESS location........................................................................ 41
Figure 5-13: IPEM command and control suite. .......................................................................... 42
Figure 5-14: Photos of inside the P196 inverter room where the IPEM controller was installed.
The image on the left shows the location pre IPEM. The middle image shows some of the existing
fiber termination block and equipment moved to accommodate the IPEM controller. The right
image shows the IPEM controller and ancillary equipment installed........................................... 42
Figure 5-15: Screen shots of the IPEM HMI . .............................................................................. 43
Figure 5-16: Photo of IPEM HMI being utilized on 12/13/15 islanding demonstration. ............ 44
Figure 5-17: System configuration block diagram. ...................................................................... 46
Figure 5-18: Test configuration for PV Comms and Curtailment Test ....................................... 48
Figure 5-19: Test configuration for Energy Storage Charge/Discharge Test . ............................ 49
Figure 5-20: Test configuration for Islanded Mode Tests ........................................................... 51
Figure 5-21: Test configuration for Islanded Mode Tests ........................................................... 52
Figure 5-22: 3-Phase balance for voltage and current. ................................................................. 55
Figure 5-23: Plot of total real power (kW) and reactive power (kVAR) over time. Power factor
ranged from about 0.86 to 0.89. Peak 6311 load on Sunday was about 70 kW at 1515. ............. 55
Figure 5-24: Ground current measurement schematic. ................................................................ 57
Figure 5-25: Timeline for Islanded Operation with Battery Only test on 10/25/15. .................... 61
Figure 5-26: Photo of damaged H-bridge system. ....................................................................... 62
Figure 5-27: Summary of islanding data on the morning of 12/12/2015. ................................... 63
Figure 5-28: Voltage and current waveform of building 6311’s load provided by the ESS. ...... 63
Figure 5-29: Detailed load data taken from Fluke 437II that was attached to Building 6311 circuit
during 12/12/15 morning islanding test. ....................................................................................... 64
Figure 5-30: Photo of Chevy Volt used in islanding demonstration outside of building 6311. .. 65
Figure 5-31: Data from Fluke 1735 meter that was attached to the P196 PV circuit. ................. 65
Figure 5-32: Power data from the Fluke 437 that was attached to building 6311 circuit during
12/12/15 afternoon islanding. ....................................................................................................... 68
Figure 5-33: Power data from the Fluke 1735 meter that was attached to the P196 PV circuit
during 12/12/15 afternoon islanding. ............................................................................................ 69

Islanding and Backup Power v EW-201242
Figure 5-34: The AC waveform collected from the Fluke 437 during the final seconds of islanding
the system...................................................................................................................................... 69
Figure 5-35: ................................................................................................................................... 70
Figure 5-36: ................................................................................................................................... 71
Figure 5-37: Graph showing the theoretical relationship of building load, PV generation, and
minimum battery generation output and how it relates to PV penetration levels. ........................ 72
Figure 5-38: Screenshots of the IPEM HMI during the 12/13/15 islanding event. Screenshot
shows the initial loads of building 6311, ESS status and PV generation status. There is currently
no PV output but the status on Carport A inverter (shown in red square) indicates the inverter is
online and in its Startup/Bootup delay per UL1741 in the left image and in its Idle mode ready to
output power in the right image. ................................................................................................... 73
Figure 5-39: Screenshots of the IPEM HMI during the 12/13/15 islanding event. These screenshots
show the initial output power of the PV system and its relationship with the load and the ESS
output power. The image on the top is at 9:59AM when the PV penetration level was initial set
at 30%. The image on the right is at 10:34AM when the PV penetration level was set at 65%. 74
Figure 5-40: ................................................................................................................................... 75
Figure 5-41: Clouds forming in the southwestern sky. ................................................................. 76
Figure 5-42: Summary load and generation profile during 12/13/15 islanding demonstration test.
....................................................................................................................................................... 77
Figure 5-43: Screenshot of IPEM HMI at taken 2:34 when the ESS exhibited a fault and went into
its inactive mode. .......................................................................................................................... 78
Figure 5-44: Detailed schematic of the interconnection of the various subsystem components of
the installation. .............................................................................................................................. 79
Figure 6-1: Frequency measurements during 12/13/15 islanding demonstration test. Data was
taken from Fluke 437 power analyzer. Frequency was maintained at a very stable 60hz. .......... 81
Figure 6-2: Phase to phase voltage data during 12/13/15 islanding demonstration test. ............. 82
Figure 6-3: Voltage measurements on the EnergyCell string and DC output power from the Primus
central regulator going into the Parker inverter. At ~1:40PM one of the twelve EnergyCells active
during the test reached its end of discharge and took itself out of the string, dropping the string
voltage. The string voltage appears stable all the way to point the fault occurred. ..................... 83
Figure 6-4: Load profile from 12/13/15 islanding demonstration test highlighting load steps from
manually increasing HVAC and building loads. .......................................................................... 84
Figure 6-5: Outage time from 12/13/15 islanding demonstration test to enter islanding mode. Grey
box in L1N shows time recorded by Fluke 437 showing when the voltage was dropped and then
restored. ......................................................................................................................................... 85
Figure 6-6: Outage time from 12/13/15 islanding demonstration test to exit islanding mode. Grey
box in L3N shows time recorded by Fluke 437 showing when the voltage was dropped and then
restored. ......................................................................................................................................... 85
Figure 6-7: A graph showing a historical load profile of a building. .......................................... 86

Islanding and Backup Power vi EW-201242
Figure 6-8: Peak shaving test data. ............................................................................................... 87
Figure 6-9: Comparison of the load data collected on 12/12/15 to the data collected during the
peak shaving test on 1/12/16. ........................................................................................................ 88
Figure 6-10: Energy Capacity test conducted on 11/15/15. The discharge output of the battery was
set to 230 kW output power. ......................................................................................................... 89
Figure 6-11: Energy capacity test conducted on 11/17/2015. The discharge output of the battery
was set to 190kW output power. ................................................................................................... 89
Figure 6-12: Energy capacity calculated for 11/15/15 test. ......................................................... 90
Figure 6-13: Energy capacity calculated for 11/17/15 test. ......................................................... 90
Figure 7-1: SDGE Time of use schedule graphic. ....................................................................... 92
Table of Tables
Table 2-1: Summary of results from NREL testing. ..................................................................... 15
Table 2-2: Summary of advantages of the Primus Zn/Br system. ............................................... 20
Table 3-1: Summary of Performance Objectives.......................................................................... 22
Table 5-1: Table showing the relationship between PV Penetration, Load Reductions and its effect
on Islanded Duration. .................................................................................................................... 36
Table 5-2: Grid Tied Performance Objectives for ESTCP Demonstration ................................. 44
Table 5-3: Islanded Mode Performance Objectives for ESTCP Demonstration. ........................ 45
Table 5-4: List of equipment used in system test setup. .............................................................. 46
Table 5-5: Results of ground current measurements. ................................................................... 55
Table 6-1: Summarized IEEE1547.4 requirements pertinent to ESTCP demonstration ............. 80
Table 7-1: Cost Model for an Energy or Water Technology ....................................................... 92
Table 7-2: TLC cost for each year for a 20 year period............................................................... 94
Table 7-3: Cost Model for an Energy or Water Technology ....................................................... 95
Table 7-4: TLC cost for each year for a 20 year period............................................................... 97

Islanding and Backup Power vii EW-201242
1 INTRODUCTION
This Environmental Security Technology Certification Program (ESTCP) effort demonstrates the
energy security and cost benefits of implementing a Zinc Bromide (Zn/Br) Flow Battery-based
Energy Storage System (ESS) at the Marine Corps Air Station (MCAS) Miramar. The effort
integrates an innovative Zn/Br Flow Battery and Intelligent Power and Energy Management
(IPEM) technologies with the existing MCAS infrastructure, providing energy security and
islanding capability, while reducing costs.
Improving energy security and reducing consumption are key strategic objectives of the
Department of Defense (DoD). Achievement of these objectives is limited by commercial power
grid vulnerabilities and intermittencies of available renewable resources. Low cost, large scale
energy storage systems are needed to address these limitations.. Energy storage is a preferred
approach to enable off-grid ‘islanding’, improving energy security through grid-independent
operation. The ESS provides a reliable source of energy in the event of a cyber or physical attack,
natural disaster or technical malfunction.
This project started in the middle of 2012 and concluded demonstrations at the end of 2015.
Modifications to existing Miramar infrastructure were required to accommodate the ESS and allow
for islanded operations. The system design phase occurred from 2012 to 2014 and went through
changes in the supplier of the ZnBr Flow Battery. The pre-construction phase for the program
started in the fall of 2014 and the construction started in the spring of 2015. The new utility
switchgear and ESS were installed and commissioned at the end summer of 2015. The
demonstration phase of the project started in the Fall of 2015 and concluded at the end of 2015.
This project was intended to demonstrate that an energy storage system can be used as a
replacement for conventional diesel generators for emergency back-up power and demonstrate that
an ESS can function with renewable energy systems within a microgrid islanded operation to
enhance energy security. This project also intended to show that an ESS can be used for
economical benefits by changing the load profile of a building by charging and discharging the
battery according to a controlled schedule.
1.1 BACKGROUND
The Marine Corps Air Station, located at Miramar, CA has completed a significant study for
locating and sizing Renewable Energy (RE) generation in order to demonstrate progress towards
reaching Net Zero operation; e.g. a Military installation that produces as much energy on or near
the installation, as it consumes in its buildings and facilities. During the initial study, Energy
Storage Systems were briefly discussed, but not actively pursued due to constraints of previous
programs. MCAS Miramar has identified a need to manage the variable power generation of the
installed RE systems without adding additional sources. To improve a base’s overall energy
security, an ESS can bridge power gaps in the RE generation either by load shifting, peak shaving,
or arbitrage.
1.2 OBJECTIVE OF THE DEMONSTRATION
There were two main objectives of this project. The first objective is to demonstrate that Energy
Storage enables the use of existing renewable energy systems that normally are unavailable during

Islanding and Backup Power 1 EW-201242
a grid outage, to ‘Island’ a building circuit for 72hrs without a diesel generator. The genesis of
this objective lies with the current large deployments of renewable PV systems that have been
installed by the DoD. The majority of these PV systems were installed to meet renewable energy
goals without considering their interaction with microgrid and islanding energy security scenarios.
This means the systems have built in safety features such as UL1741 that shut the systems down
in the event of a grid outage. This project intends to demonstrate that an ESS can provide voltage
control capability in islanded operations that allow the functionality of existing PV systems in
microgrid mode at high penetration levels. This will enhance the energy security of the base in
the case of an extreme event (e.g. cyber attack, utility maintenance, or natural disaster),
demonstrate Energy Storage microgrids provide increased capability of existing PV installations.
The second objective is to demonstrate that an Energy Storage System can peak shave for demand
charge avoidance. Many DoD facilities have been attempting to reduce their operational energy
costs by implementing a variety of energy efficiency and renewable energy programs. One of the
biggest costs to many facilities is not in the cost of energy purchases but in the demand charge
issued to the facility based on its load profile. This project was designed to allow the ESS to be
programmed to charge/discharge according a defined peak shaving schedule. This is to show that
Zn/Br system can charge during off peak hours and discharge during peak hours reducing peak
load by the power output of the battery. The intent of this objective is to show that Energy Storage
can provide economic benefit in addition to improved energy security.
The field demonstration for this project was intended create operational scenarios for which the
two main objectives could be demonstrated. To demonstrate the energy security improvement the
project set up a scenario where power was interrupted to the microgrid circuit and the system
would provide back up power for the outage meeting the load requirements of the microgrid. To
demonstrate the peak shaving capabilities the project set the microgrid system up so that the ESS
could charge and discharge on a user created schedule and data could be collected on the systems
capability to peak shave during defined hours.
1.3 REGULATORY DRIVERS
The National Defense Authorization Acts 2010-2012 and Energy Independence and Security Act
of 2007 have shaped the Navy’s microgrid strategy. This has created five major energy goals
issued by the secretary of the Navy and shared in similar sense with the other branches of the
military. The five energy goals are listed in Figure 1-1 below.
Reduce Energy Consumption & • By 2020, USN will reduce energy consumption and
Intensity intensity by 50% from a 2003 baseline.
Power from Renewable • By 2020, 50% of total ashore energy will come from
Sources renewable sources.
• By 2020, 50% of installations will be net-zero
Net-zero installations consumers.
Reduce Non-Tactical • By 2015, reduce petroleum used in commercial vehicle
Petroleum Use fleet by 50% from a 2009 baseline.
• Provide reliable, resiliant and redundant power to
Increase Energy Security increase the energy security of mission critical assets.
Figure 1-1: Energy goals from the Secretary of the Navy.

The goal for increased energy security is one of the main drivers for this project. It is common for
our country’s military bases to get their power from their local utility companies. Utility
companies and their power distribution networks can be vulnerable to events such as extreme
weather or even cyber attacks. The San Diego area was subject to an 11 hour blackout back in
2011 due to an error during routine maintenance of the distribution system (Figure 1-2). This
creates the need for back-up generation systems and the current status-quo is to use diesel
generators. In the spirit of trying to meet the Navy energy goals the Navy is looking at ways to
leverage their renewable investments to replace diesel burning systems. The Navy is also looking
for creative ways that it can use microgrids and energy storage to improve its load profile to avoid
high peak charges and participate in economic incentive programs such as demand response.
Figure 1-2: Image of the US electrical distribution system taken from the FEMA website.
Energy storage can play a key role in meeting the energy goal and mission needs of our military
installations. The existing electrical distribution system was built around the production/use
principle that electricity must be produced when it is needed and consumed once it is produced.
This principle works when a generation network is in place that is monitored and controlled
predictably. The ability to control the generation has become more difficult with the increase of
renewable energy systems such as solar photovoltaics (PV) and wind. Both PV and wind systems
generate power based on unpredictable cycles of nature. At very low levels of renewable energy
penetration this can be handled through the existing generator network to keep the grid balanced.
The ability to store excess energy acts as a key enabler to increasing levels of RE penetration.

2 TECHNOLOGY DESCRIPTION
2.1 TECHNOLOGY OVERVIEW
Technology Description
The Zn/Br flow battery technology manufactured by Primus Power was chosen to be the storage
provider over other storage technologies for several reasons described below. The key determining
factors for islanding and renewables integration applications for Miramar are: low cost, energy
storage capacity, intelligent system control, transportability, cycle life, system lifetime, and safety.
The traditional Zn/Br batteries contain a solution of zinc bromide in two tanks. During charge of
the battery, zinc is electroplated on the anode and bromine is sequestered in a polybromide
complex that is stored in the electrolyte storage tank. On discharge, the polybromide complex is
returned to the battery stacks, and zinc is oxidized back into the electrolyte solution, forming the
identical Zn/Br solution the unit started with (Figure 2-1). This type of battery leverages many
years spent developing proper plating systems in a novel storage approach.
Figure 2-1: Schematic of a traditional Zn/Br cell with two electrolyte flow loops.
The traditional Zn/Br cell design uses carbon coated felt paper as the electrode surface. The cells
also have two separate electrolyte tanks for capturing the anolyte and catholyte separately during
charge and discharge. These separator membranes and carbon paper often are subject to
degradation and contamination and are a common failure mechanism amongst Zn/Br batteries that
requires reoccurring replacement. Traditional Zn/Br needs to be replaced after 1500 cycles which
would constitute replacement every 4.1 years if cycled daily.
Primus Power took a different approach to their Zn/Br cells. Instead of using carbon coated felt
paper for their electrodes Primus utilizes an activated solid titanium electrode for its Zn plating
surfaces. Using a titanium electrode provides Primus the capability to use a single flow loop of
electrolyte as opposed to dual flow loops as well as eliminate the need for an ion exchange
membrane, which is an early failure mechanism in tradition Zn/Br cells. This reduces the number
of tanks required and pumps for managing the electrolyte (Figure 2-2). The titanium electrodes
also provide better energy density when compared to traditional Zn/Br 3.1 kWh/ft2 vs1.7 kWh/ft2.

Figure 2-2: Schematic of Primus’ approach to a Zn/Br cell using a solide titanium electrode and a single flow
loop of electrolyte.
Primus’ ESS has three main subsystems that encompass the entire energy storage system. The
ESS has an EnergyPod which contains the ZnBr EnegyCells. The EnergyPod is connected to a
PowerBox that contains the power electronics of the ESS as well as the Battery Management
System. The system also has a chiller used to provide cooling to the ESS. The ESS was specified
at the onset of this program to be 250kW nominal power and 1MWh of energy capacity at a C/4
discharge rate. Primus uses a 30kW building block called an EnergyCell to build their EnergyPod
system. During initial tests of the EnergyCells it was determined by Primus that 14 EnergyCells
would be required to achieve the project goals for islanding and peak shaving. The EnergyPod
system was designed to be packaged in a 20ft container coupled with the PowerBox that was
housed in a 20ft ISO container. An early rendering of the system is shown in Figure 2-3, the
EnergyPod was then expanded into a 40ft container and the updated rendering is shown Figure
2-4, an image of how the system was intended to look during construction planning is shown in
Figure 2-5, and a photo of the complete system installed is shown in Figure 2-6.
Figure 2-3: Early rendering of ESS located at MCAS Miramar.

Figure 2-4: Illustration of Primus EnergyPod system complete with Zn/Br cells, power electronics, and
thermal management.

Figure 2-5: Isometric view of the construction plans.
Figure 2-6: Photo of installed system at MCAS Miramar.
2.2 TECHNOLOGY DEVELOPMENT
The developmental timeline for Primus’ Zn/Br technology is summarized in Figure 2-7 below.
Primus’ early development started in 2009. When Primus was put under contract in late 2012 for
this project they were operating at TRL4 and progressively matured their technology to TRL5 at

the conclusion of the 3rd party testing of their 30kW EnergyCell unit completed in the fall of 2013
by Sandia National Laboratory (SNL), then to TRL6 at the conclusion of this demonstration. The
following section will describe the TRL advancement of Primus’ technology leading up to the
demonstration.
Figure 2-7: Developmental timeline for Primus EnergyPod system.

At the beginning of 2013 Primus Power was in process of building the first EnergyCell building
block for their energy storage technology. They had already successfully built and tested a variety
of smaller cells that validated the technology at 1-2kW scale. In summer of 2013 the first 20kW
EnergyCell was finished being built and was ready for characterization testing. Primus chose to
have the EnergyCell 3rd party tested by Sandia. The test setup from the Sandia testing is shown in
Figure 2-8 below. Sandia performed a variety of performance tests on Primus’ EnergyCell.
Figure 2-8: Photo of Sandia National Laboratory 3rd party testing of Primus Powers EngergyCell at Hayward
CA.
The goals for the Sandia testing were to:
• Measure ability of EnergyCell to perform discharge at various durations
• Measure round trip efficiency of charge-discharge cycles
• Measure power rating of the EnergyCell
• Measure EnergyCell step response to command charge/discharge

The Sandia 3rd party testing results are summarized in and were briefed to the ESTCP office in
November of 2013. The testing validated that Primus’ technology provides the power, efficiency
and responsiveness needed for the demonstration. A couple of the takeaways from the Sandia
tester were:
• Power exceeds 20kW requirement (30+kW capable)
• Transient response consistent with demonstration application requirements
• Energy efficiency is consistent with expectations
The testing also showed additional work was required to meet demonstration islanding duration
objectives with reasonable load shedding. At 20kW EnergyCell was discharged at 20kW and
achieved an energy capacity of 25kWh at just over 1 hour of discharge time. At 3kW of discharge
power the capacity test resulted in 2+hr (43kWh) performance. In order to meet our system level
performance objective with the number of EnergyCells originally proposed each energy cell would
need to have an EnergyCapacity of 80kWh at 20kW. After the testing Primus showed that in-
house testing of smaller cells had exceeded 40kWh with altered additives and discharge controls.
Primus planned to continue developing electrolyte chemistry and charging modifications to meet
the 80kWh objective in their SOW.
Figure 2-9: Summary of results of Sandia 3rd party testing.

The results of the Sandia testing provided needed insight into the development and scale up
capability of Primus Powers first generation ESS. The EnergyCells performed very well in terms
of power density and responsiveness to meet the load following capabilities required for the
microgrid however the energy density performance was not meeting initial expectations at larger
scales. The original EnergyPod was envisioned to be in a 20ft ISO container with the 14
EnergyCells, however based on the energy density and capacity results from the Sandia testing the
size of the EnergyPod would need to expand to a 40ft container and further development would
be needed to improve the energy capacity to achieve the 750kWh threshold requirement. Primus
power committed to continue working on increasing their energy density and energy capacity of
its EnergyCells but needed more time to make improvements. Two options were discussed at the
November briefing with the ESTCP office.
Option #1: Deliver 250kW/500+kWh System in 40’ + 20’ ISO in August 2014
• Pros: Hold to original schedule
• Cons: Requires excessive load shedding to meet 72hr islanding objective, no opportunity
to improve performance/packaging density over current system, delivered system will
feature unique technology not common with other Gen1 system deliveries
• Risk Level: Confined to BoP. System will employ the same design as the SNL- tested
EnergyCell and 3-cell prototype
Option #2: Deliver 250kW/1000kWh System in 20-40’ + 20’ ISO in January 2015
• Pros: Meets original ESS performance specs to support 72hr islanding objective with
reasonable load shed, may meet original size/volume specs delivered system common with
other Gen1 deliveries
• Cons: Requires 6mo extension
• Risk Level: Mitigated through flexibility on deliverable form factor. If further electrolyte
development does not yield required performance improvements, Primus will deliver
required (250kW/1000kWh) capacity with existing technology
Raytheon and MCAS Miramar recommended Option #2 to the ESTCP office, based upon Primus
Power commitment to performance specifications and MCAS Miramar priorities. The ESTCP
office agreed with the assessment and the projected delivery of the battery was re-forecasted to
early 2015 with periodic assessments of system performance.
While Primus was working on scaling up and building their large scale system in 2014, Raytheon
orchestrated hardware in the loop testing of the Miramar microgrid utilizing the National
Renewable Energy Laboratory (NRELs) Energy System Integration Facility (ESIF) funded by
Raytheon outside the funding of this ESTCP program but provided was directly beneficial to this
project. The intention of the testing was to provide high-fidelity evaluation of the MCAS Miramar
microgrid in a simulated operational environment with real hardware in the loop testing with full-
scale/full power simulated sources and loads. The system testing reduced a lot of risk on
integrating the IPEM controller to manage the existing Advanced Energy PV inverters at MCAS
Miramar, the Primus ESS, and the various metering and control logic of the microgrid. The testing
at NREL was designed to re-create the designed Miramar microgrid at as high a fidelity as possible.
Figure 2-10 below shows a one line diagram for the Miramar microgrid circuit for the ESTCP
project.

12kV PV Power
Distribution Grid Power
Battery Power
Electric Connection
Communication
CT Link
PAD Mounted Switchboard B5-PS2T3
CT CT CT
IPEM
Energy
230kW Storage Unit
PV
Building 6311
(30kW PV Roof
Unit) Inverter/
DC Regulator
Figure 2-10: One line diagram for the MCAS Miramar microgrid circuit for the ESTCP project.
Figure 2-11 shows the one line diagram for the circuit that was designed to be used at NRELs ESIF
facility. The NREL system utilized the same PV inverters that exist at MCAS Miramar, a similar
main breaker point of common coupling, and the same inverter & BMS utilized by Primus’ ESS.
Figure 2-11: One line diagram design for test setup at NRELs ESIF facility.

The end result for the configuration used at NREL is shown in the detailed single line diagram
shown Figure 2-12 below.
Figure 2-12: Detailed single line diagram implemented at NREL for ESIF testing.
During the course of testing Raytheon hosted a demonstration of the system with MCAS Miramar
stakeholders and a representative from the US Marine Corps Headquarters (Randy Monohan).
Randy was the MCAS Miramar station energy manager when this project was originally proposed
and one of the earliest advocates in the project.

Figure 2-13: Photos showing the live testing at NRELs ESIF facility with MCAS Miramar and USMC
personnel.

Figure 2-14: Photos of the various equipment that is part of the NREL testing. The top left image is of Primus
Powers 760kVA Parker grid tied inverter that is the main element in Primus’ power electronics subsystem.
The bottom left is of a 480 to 208 transformer, the bottom second from the left is Primus’ EnergyBlock
controller which manages Primus AC & DC busses. The bottom third from the left shows the capacitor bank
within the Parker inverter. The right most image shows the AC breaker for the Parker inverter and the local
HMI display.

Figure 2-15: Photos of the IPEM controller and ancillary equipment used during the NREL testing.
The results of the NREL testing are summarized below:
Table 2-1: Summary of results from NREL testing.
Goal Result
The black start sequence and transition to islanding Demonstrated automated back start sequencing
work as anticipated within the 1hr time requirement
The ESS inverter and PV inverters power share Verified load sharing across operating range (0-200kW,
properly in islanding mode 0.1-1.0PF)
The UL1741 anti-islanding algorithms do not No issues observed
destabilize the ESS inverter in voltage control mode
The PV penetration be pushed to >50% without de- Successfully run up to 100% PV penetration (w/bi-
stabilizing the ESS inverter in voltage control mode directional power flow to ESS)
The system does not destabilize due to dynamic PV Characterized PV curtailment response timelines in
curtailment and the system can handle load step response to increasing and decreasing load changes
requirements for Miramar’s load
The system meets IEEE1547.4 requirements for power No issues staying within trip points
quality.

After the completion of the NREL testing at the end of 2014 Primus was finishing building up its
full scale system. Primus was ready to perform its Factory Acceptance Testing (FAT) of the
completed ESS at their Hayward facility in May of 2015. The purpose of the FAT was to assess
the performance and functionality of the system compared to performance objectives defined in
their statement of work.
During the course of the build of the ESS Primus Power was continuously trying to improve its
energy capacity capabilities of their EnergyCells based on the original assessment by the Sandia
testing. At the time of the FAT Primus presented their current state of the energy capacity available
with the configuration of EnergyCells that were to be delivered to MCAS Miramar. Figure 2-16
below shows the progression of meeting the targeted energy capacity as Primus was able to
manufacture more of its EnergyCells to populate the system.
Figure 2-16: Energy capacity timeline and scale up from Primus Power since the November 2013 briefing.
The FAT testing included the following test objectives and the results are summarized below
FAT Test 1: Peak Shaving
The objective of the Peak Shaving test was to demonstrate that EnergyPod is capable of storing
energy during off peak hours and push 250kW back to the grid during peak hours. The summary
of test results are shown in Figure 2-17 below and show that the ESS is capable of charging and
discharging at 250kW. It is important to note that the power output of battery is the net output
power of the entire Energy Storage System. This means the battery output power minus the
auxiliary power to the battery which includes: the control power to all the pumps, power
electronics, inverter, chiller and the heaters. It is important to mention this as various energy
storage systems have a separate auxiliary power requirements in their systems and don’t subtract
it from their output power when providing ratings.

Figure 2-17: Results of the Peak Shaving Test during the FAT. Note the peak charge power is 140kW and the
peak discharge power is 225kW.
FAT Test 2: Energy Storage Capacity
The objective for the Energy Storage Capacity test was to demonstrate the Energy storage capacity
capability in grid tie mode. The summary of test results are shown in Figure 2-18 and Figure 2-19
below. The energy capacity capability of the ESS was determine to be 390kWh during the FAT.
Figure 2-19 shows the anticipated relationship between discharge power and DC-DC energy
capacity. The curve shows the most optimal efficiency for the system is when the EnergyCells are
discharged at 10kW (140kW at scale with 14 EnergyCells).
Figure 2-18: Results of the Energy Capacity testing during the FAT. Note the total charge duration is 6hrs and
20 min. The peak discharge power is 250kW and the total discharge energy recorded is 390kWh.

Figure 2-19: Energy Capacity as a function of Discharge Rate shown on the DC output of the ESS.
During the course of investigating the energy capacity limitation of the EnergyCells it was
determined that the configuration of the flow channel frames within each of the EnergyCells was
causing non-uniform zinc plating across the plating area of the electrodes. The non-uniform
plating was leading to Zn inlet edge ridges that cause early shortening as the zinc reaches the top
of the opposing electrode earlier than the other areas of the electrode. Primus presented solutions
that provided better electrolyte flow across the electrode reducing the non-uniformity allow more
zing to be plated across the electrode surfaces improving the energy capacity. The new design of
the cell frames were still in their test phase and would require retrofits of all 14 EnergyCells that
were ready to be deployed further delaying the program.
FAT Test 3: Islanding capability (Black start)
The objective of the Islanding Capability test is to demonstrate that in islanding mode the Central
Regulator (CR) can regulate the bus voltage while the inverter creates the grid to supply power to
any load connected to the island. This test has two important aspects. The system needs to be able
to start the EnergyCells and boost the bus voltage in order to enable the inverter to create the island
by putting out 480vac 3ph output.
Test Procedure:
1. Open the main disconnect switch to the Grid and lock out tag out the disconnect switch
2. From the EnergyBlock GUI select Islanding operation
3. The system shall:
4. Turn the Aux. power to the EnergyPod and inverter
5. Send the EnergyCells into the discharge mode
6. Charge the bus voltage to 750vdc
7. Start the inverter
Summary of test results are shown in Figure 2-20 below.

Figure 2-20: Oscilloscope screenshot showing the connected logics of the EnergyPod and PowerBox. The
inverter was started in islanding mode. The image shows the bus voltage being regulated continuously.
At the conclusion of the FAT testing the ESS was demonstrated to be functionally operational
however still lacking in the desired Energy Capacity performance requirements defined in the
SOW. At this point in the demonstration Primus Power’s team had made tremendous amount of
progress and investment to get the system to function as required. As the program did not have
enough time or resources to continue developing the ability to increase the energy capacity any
further the system was accepted by Raytheon with agreement and understanding from MCAS
Miramar to deliver the system at the end of May 2015.

2.3 ADVANTAGES AND LIMITATIONS OF THE TECHNOLOGY
Primus’ Zn/Br flow battery approach provides advantages in cycle life, cost, and performance
when compare to similar technologies. The advantages are summarized in Table 2-2 below.
Table 2-2: Summary of advantages of the Primus Zn/Br system.
Primus’ Zn/Br battery offers higher current density when compared to similar technologies. Their
electrodes can operate at 200 mA/cm2 vs 50 mA/cm2 of traditional Zn/Br. Primus biggest
discriminator is that it eliminates two common failure mechanisms in ZnBr flow batteries (carbon
electrodes and separator membranes) by using a solid titanium electrode and not requiring a
membrane. This allows their cells to operate longer than tradition flow batteries without the need
for replacement. Component level testing of all of the ancillary equipment and stability testing of
their cells have predicted a 20 year lifespan.
Primus’s battery still uses a Zn plating mechanism for its batteries. The nature of the Zn plating
requires that the cells be completely discharged to prevent dendrite growth and maintain the health
of the cells. This requires that the EnergyCells be periodically stripped to properly clean and
maintain them. This is handled automatically by the Battery Management System and is
transparent to the user. However, this requires that an EnergyCell will be periodically taken
offline. The energy storage system will still operate however it will be operating less one
EnergyCell reducing its energy and power capacity during those times.
One major limitation of Primus’ current system is that when the system is in islanding mode the
ESS operates in voltage control mode. When operating in this mode the battery is currently not
capable of charging. This is currently attributed to adequate control of plating zinc on the
electrodes. Primus’ development and current algorithms for charging the battery depend on
optimal parameters for plating uniform layers of zinc on the electrodes in the EnergyCells. When
the system is in islanded mode controlling the parameters for plating zinc become more difficult
and Primus has not been able to analyze this functionality to include it in the current operation of

the system. Based on discussions with Primus’ engineers the capability to charge the system in
islanding mode is possible but requires testing the system and validating the techniques.
The foreseeable barriers to social acceptance would rely on the fact that an energy storage unit of
this size requires approximately a 60ft x 20ft of footprint to be installed. This means that the
locations that would like to use and energy storage system would need the space for the installation.
The containers that the system is installed with are traditional ISO containers so the visual look of
the system is not abnormal. The thermal management system and the pumps within the system
will make noise equivalent to an air conditioning system which is not out of the ordinary. The
electrolyte for the cells contains bromine which needs to be handled appropriately. The EnergyPod
contains multiple layers of spill containment which would abate many concerns for spill
protection.

3 PERFORMANCE OBJECTIVES
There are five performance objectives for this demonstration and are listed in Table 3-1 below.
The performance objectives were established based on early discussion with MCAS Miramar
personnel and to meet particular mission scenarios for improved energy security and operational
cost reductions.
Table 3-1: Summary of Performance Objectives
Performance Data Results

Metric Success Criteria
Objective Requirements
Quantitative Performance Objectives
Energy Security Performance Objectives
Islanded Duration Islanded Duration Meter readings from Building loads are Building loads
(hours) RE system, ESS, and met by ESS and PV were met by ESS
grid power feed for 72hrs under and PV for 5
controlled load hours 10 minutes
conditions meeting meeting power
power quality quality standards
standards of of IEE1547.4.
IEE1547.4 ESS is capable
of 7 hrs 10 min.
Building Load Delta Average Meter readings from Building loads can Building loads
Reductions kWh/day usage building 6311. be reduced by 50% were able to
through manual manually
changing of increased and
thermostats and decreased
lighting when increased by
compared to its 68% when
previous year’s compared to
average for that baseload during
given month. islanding test
Switchover Time Time (minutes and Clock timing from Time is less than Switchover from
seconds) command to go into hour Grid to Islanding
islanded mode to ESS was 4 minutes
discharging power
Operational Cost Reduction Performance Objectives
ESS Energy Energy Discharged in Meter reading of ESS is able to ESS was able to
Storage Capacity kWh energy discharged by discharge 1MWh of discharge
ESS energy during peak 390kWh in the
shaving cycle. lab and 290kWh
in the field
Peak Shaving Peak Demand Meter readings from ESS is able to store ESS was able to
Reduction (kW) RE system, ESS, and energy during off store energy
grid power feed peak time and during off peak
discharge 250 kW time and
during peak time to discharge
reduce peak load 100kW during

relative to historical peak time for 2
data over similar hrs and 45 min
time period.
Qualitative Performance Objectives

Ease of Operation Degree of ease of use Survey Satisfactory rating Survey to be
from survey results. issued before
final report
3.1 ISLANDED DURATION
Islanding is defined as being able to intentional isolate local facility circuit from the local electric
power system as defined in IEEE 1547.4. The circuit is then power by the operation of the ESS,
and RE. The Islanded duration will be the time that the system is commanded into islanded mode
to the time that the system can no longer sustain the loads of the circuit.
Purpose
The purpose of the Islanding objective is to demonstrate the applicability of an isolated utility
circuit going off-grid. This is useful in the case of an extreme event that could disrupt commercial
utility power supply. Emergency back-up operations can be maintained by operating off of RE
and an ESS if the load required is maintained within acceptable operation levels of the PV system
and ESS.
Metric
The metric used for the Islanded Duration objective is islanding time measured in hours and
minutes. The islanding time starts when the system is commanded to go into islanding mode from
the IPEM controller. The islanding time stops when the loads can no longer be met due to the ESS
being depleted.
Data
The data that will be required to calculate this metric is a multitude of measurements from various
sensors within the system.
1. Power output from the PV system
2. Load data from building 6311
3. Net Power output from the ESS (Battery output minus auxiliary power including: pumps, power
electronics, inverter, chiller, and heaters)
4. State of Charge of the ESS
5. Power quality measurements of the power provided to 6311
6. Clocked time showing the start of islanded operations to the end of islanded operations
Metering points for the islanding tests are shown in Figure 2-10 and in more detail in Figure 3-1
below. In addition to the metering data that is collected by the IPEM system, two independent
power analyzers were connected to the Building 6311 feeder breaker and the P196 Carport PV
system to collect detailed data for analysis to IEEE 1547.4 requirements.

Independent Power Analyzers B5-PS2T33 Pad
Switchboard 12000-
208Y/120V
Fluke Fluke 12kV
Distribution
437 1735 Grid
1000AT
MB1
900AT
800AT
700AT
MB1 Φ Rot.
CTRL Relay
A13 A12
Relay Relay
A4 A5 IO
Modbus Gateway Adapter
ZnBr Energy Storage System (ESS)
Energy Pod PowerBox

250kW 250kVA 1000A
480V-208Y
… Inverter Disconnect P196 Carport PV and Inverter Room
- 400A Fused PV Powered
~ Discconnect 100kW Inverter
- ~
~ 800A
Disconnect -
Energy
Block Chiller 232kW
Controller 400A Fused PV Powered STC
Discconnect 100kW Inverter PV Array
~
Building 6311 -
Rooftop Solar PV 30kW
SATCON Fused
30kW Inverter Discconnect I/O Adapter (Ethernet)
Rooftop
PV Array - Breaker Raytheon
P196 Inverter 1 I/O (Modbus)
Box RTU Adapter (Ethernet) Intelligent Power & P196 Inverter 2 I/O (Modbus)
~ Energy Management
Command & Control
ESS I/O (OPC) and HMI
SATCON Inverter I/O (Modbus)
Data Acquisition
Human
Machine Fiber
Ethernet Fiber Optic Underground Conduit
Interface Optic
(Client) Modem
Figure 3-1: Detailed interconnect diagram for the Islanding test setup
The time elapsed will be measured from when the system is commanded to enter into islanded
mode. While operating in islanded mode the various subsystems will be monitored and data will
be collected on the PV system, the ESS, and the building loads. Once the battery is depleted the
system will shut down until grid power is restored. After power is restored the various load data
and performance data on each of the subsystem will be collected and analyzed to assess the
behavior and stability of the circuit.
Success Criteria
The success criteria for this performance objective was that building loads would be met by the
ESS and PV for at least 72hrs under controlled load conditions meeting power quality standards
of IEE1547.4.
3.2 BUILDING LOAD REDUCTIONS
The building load reduction Performance Objective is defined as the percentage of load that has
been reduced during an islanded event as compared to the previous year’s average for that given
month during normal grid connection.
Purpose
The purpose of this Performance Objective is to characterize the amount of building load reduction
during an islanded event required in order to meet the 72hr islanded objective.

Metric
The metric for this Performance Objective is the percentage difference in kilo-watt hours per day
(kWh/day) of the load when operating in islanded mode compared to an equivalent load profile
for the same given month.
Data
The data required for this is the load data measured for building 6311. The following Equation 1
will be used in analyzing the data.
∫ 𝐿𝐿𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
1−� � ∗ 100 (1)
∫ 𝐿𝐿ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖
Where Lisland is the load data for the time during islanded and Lhistorical is the historical load for the
previous year’s average for the same given month. The result will be the percentage of load
reductions required during the islanded event.
Success Criteria
The success criteria for this objective is that building loads can be reduced by 50% through manual
changing of thermostats and lighting when compared to its previous year’s average for that given
month.
3.3 SWITCHOVER TIME
The Switchover Time defined as the time required to switch the system from its grid transition
mode (i.e. standby during grid outage) into islanded mode.
Purpose
The purpose is to characterize the timeline for islanded operations.
Metric
The metric used for this Performance Objective is time measured in minutes and seconds.
Data
The data required is the time recorded for when the system is commanded to go into islanded mode
and the time recorded when the ESS begins to discharge. The time recorded for when the ESS
begins to discharge will be subtracted from the time recorded for when the system was commanded
to go into islanded mode. The result will be the Switchover Time.
Success Criteria
The success criteria for this Performance Objective was defined to be less than 1 hour.
3.4 PEAK SHAVING
Peak Shaving is defined as being able to arbitrage power stored from off-peak to on-peak periods.
This allows a facility to load shift in order to reduce the facilities demand charges. The ESS is
charged and discharged in order to change its demand load profile seen by the utility company as
shown in Figure 3-2. This is useful for facilities that are on a tiered pricing scheme and/or are hit
with high charges of energy use during hours of peak operation. The ESS can charge during off

peak times at a lower cost and discharge during peak hours reducing the peak loads required by a
facility.
Figure 3-2: Graph showing a very simple hypothetical load profile. The purple line represents the historical
load profile. The green line represents the load profile using an ESS to charge at night and discharge during
the peak time of day
For many commercial and industrial facilities the cost of electricity can be heavily determined by
the amount of peak power that a facility uses during a billing period. The largest peak power
demand, typically for a minimum of 15 minutes, will dictate how much the facility is charged for
that billing period. Different utility companies have different demand charge rate structures. Some
utilities are so congested during peak times that they have a defined peak time period during the
day where they charge a higher demand rate then off-peak periods. Utilities that have this type of
rate structure also usually have incentive programs or mandatory demand response programs
where the facility can volunteer to participate or be directed to participate in load shedding during
seasonal peak times. Some utilities have a blanket demand charge that is based on the highest 15
min peak demand for a given billing month regardless of peak times. Controllable peak shaving
can provide a facility with flexibility to reduce its peak demand depending on its rate structure.
SDGE has different types of rate structures for commercial/industrial facilities (Figure 3-3). One
rate structure is a TOU structure that has two types of demand charges. The first is an On-Peak
Period Demand Charge which is based on the 15 minute average Maximum On-Peak Period
Demand. The second is the Non-Coincident Demand Charge which is based on the higher of the
Maximum Monthly Demand or 50% of the Maximum Annual Demand.

Figure 3-3: SDGE Time of Use Rate time periods.
For facilities that have TOU demand charge structure a basic peak shaving schedule is useful and
can change a facilities demand load profile. Also facilities that are subject to demand response
programs could benefit from this type of peak shaving mode. Currently Marine Corps pays a flat
rate to NAVFAC for each kWh of energy consumed therefore the Marine Corps does not have a
TOU rate structure however NAVFAC. NAVFAC is a direct customer to SDG&E and is subject
to demand charges and TOU structure which is aggregated across multiple bases in the southwest
and a flat normalized rate is applied to those facilities. While the Marie Corps is not subject to
SDG&E’s TOU structure it does influence the rate applied to them so it is utilized in our Peak
Shaving performance objective.
Metric
The metric used for the Peak Shaving objective is the difference in demand load between a relevant
historical load profile and a load profile seen by the feeder meter at B5PS2T3 switchgear when

using the ESS in peak shaving mode. The metric is measured in kW and represents the amount of
peak shaving achieved by using the ESS.
Data
There are two pieces of data required to calculate the Peak Shaving metric. The first is relevant
historical load profile data. This data was collected a couple days prior to using the ESS in peak
shaving mode. The second piece of data is the load profile when using the ESS in its peak shaving
mode. The metering points for the load was collected at the B5PS2T3 switch gear according to
the CT locations defined in Figure 2-10.
Success Criteria
The success criteria for this metric was originally determined to be that the ESS is able to store
energy during off peak time and discharge 250 kW during peak time to reduce peak load relative
to historical data over similar time period.
3.5 ESS ENERGY STORAGE CAPACITY
This Performance Objective (PO) measures the energy storage capacity of the ESS when operating
in grid connected operations. The purpose of this Performance Objective is to show that the energy
capacity of the energy storage system meets its rated 1MWh capacity.
Metric
The metric used for this Performance Objective is energy in kilo-watt hours (kWh) which is a
measurement of power over time. The value for this PO should range from 750kWh to 1MWh.
Data
The data required for this Performance Objective is power output of the ESS and recorded time of
the power output. This was captured on two different days of performing this test. The first day
captured was on 11/15/2015 and the second was captured on 11/17/2015. The measurement of
power over time will be analyzed and the energy capacity of the system will be the integral of the
graph from the beginning of discharge to the time that the power output of the battery reaches zero.
Success Criteria
The success of this Performance Objective was defined to be that the ESS is able to discharge
750kWh threshold and 1MWh objective of energy while in Grid connected mode.

4 FACILITY/SITE DESCRIPTION
MCAS Miramar in San Diego, CA has been selected as the host facility for this installation Figure
4-1. Miramar has a long history for installing renewable energy projects at its facility.
Figure 4-1: Birds eye view of MCAS Miramar.

A brief background of MCAS Miramar and its energy portfolio is listed below.
Electrical Utility Loads
• 14 MW Peak
• 7 MW Avg
• 5 MW Min
Renewables
• 3.2 MW Landfill Power Purchase Agreement
• 1.5 MW of Photovoltaic
• PV Parking lot lights
• 24 Solar Thermal systems including the Combat Training Tank (Pool)
Energy/Water Efficiency
• Area Wide Energy Management System (DDC)
• Advanced Metering Infrastructure
• $30M of HVAC/lighting retrofits in the past 2 years
• Reclaimed Water and smart irrigation control
• Replacement of over 1300 water fixtures in 22 buildings to low flow
Behavioral Awareness
• Unit Energy Managers
• Energy Star Portfolio Manger
For this demonstration we have selected the circuit feeding off of B5PS2T3 switchgear which
feeds building 6311 and the P196 Carport PV system.

4.1 FACILITY/SITE LOCATION AND OPERATIONS
The demonstration site at MCAS Miramar is shown in Figure 4-2. The specific location at MCAS
Miramar where the microgrid demonstration will occur is near building 6311 (Figure 4-3).
Building 6311 is mainly an office building for the energy manager, public works, and FEAD.
Since the building house the energy manager and staff, the ability to take the building offline
during the islanded scenarios is easier to facilitate. The base command has endorsed the project
as a major stepping stone in achieving a larger microgrid effort.
The project data communications is designed to be a closed loop system avoiding any
DIACAP/RMF and IT platform certifications. The data that is collected within the IPEM
controller and the ESS is stored locally at Miramar and can be downloaded on the base and
transferred for analysis.
Figure 4-2: Map and aerial image of MCAS Miramar.
Figure 4-3: Map of the installation site at MCAS Miramar near building 6311.

The installation site for the energy storage system is under the P196 carport PV system near the
inverter room as shown in Figure 4-4 below.
Figure 4-4: Rendering of the ESS installation site next to the P196 inverter room under the carport PV
panels.
4.2 FACILITY/SITE CONDITIONS
MCAS Miramar is located in a mild climate zone in southern California. The location provides
for provides good solar irradiance for the installed PV systems. Building 6311 is a perfect location
for this demonstration since it has its own switchgear with 230kW of PV attached to it. The
switchgear allows isolation of the circuit for islanding and the PV system allows the integration of
renewable energy into the circuit when operating in islanded mode.
Many southwestern installations have large amounts of PV installed on their facilities and many
are subject to similar Interconnect Agreements and UL1741 anti-islanding restrictions. This
demonstration at Miramar helps prove out the capability to use energy storage in a microgrid
application for integrating renewable energy systems when in islanded mode.

5 TEST DESIGN
This goal of this demonstration is to solve two main problems. The first problem is that DoD
facilities are vulnerable to grid outages due to extreme events and limited to non-renewable back-
up systems such as diesel generators which are regulated and can-not be used for cost reduction
applications such as peak shaving. The second is that the peak electrical loads of many DoD
facilities loads occur during higher rate periods incurring significant costs associated with demand
charges. The demonstration aims to answer the question: “How can an ESS, coupled with an
advanced control system, provide energy security while reducing overall facility energy costs?”
5.1 CONCEPTUAL TEST DESIGN
The ZnBr Installation is comprised of a ZnBr ESS integrated into the MCAS Miramar utility
infrastructure which includes a 230kW carport PV subsystem and a 30kW rooftop PV subsystem.
The ESS and the PV subsystems are controlled by the IPEM microgrid controller which also
controls and monitors the load demand and power quality required by the MCAS infrastructure,
the status and power generation of the PV system, and the State of Health (SoH) of the ESS (See
Figure 5-1).
Figure 5-1: Interconnect diagram of Zn/Br installation at MCAS Miramar.
The demonstration is intended to operate in two modes 1) Islanded and 2) Peak Shaving. The
islanding mode demonstrates the Islanded Operations performance objectives and the peak shaving
mode demonstrates the Peak Shaving performance objectives.

The primary mission for the ZnBr Installation is to provide emergency power in the case of a grid
outage. Maximizing the use of the PV and the ESS is crucial to extend the operational life of the
system. This allows MCAS Miramar to operate independently from the grid in the case of a
physical or cyber attack, or an environmental event that would otherwise shut down facility power.
The Installation is connected to a 230kW PV system that currently exists on the B5-PS2T3
switchgear. The carport PV inverters are UL1741 certified and therefore have built in safety
features that de-energize the inverters during a grid outage. This safety feature is enabled so as to
avoid inadvertently back feeding a circuit that may have a transmission wire down or a technician
working upstream on the circuit. In order to meet islanding duration goals in Islanding Mode these
inverters need to be active to supplement the ZnBr battery in providing power to the Miramar load.
To accomplish this, the PV inverters require a firm voltage source present in order to activate and
synchronize. The ESS provides this voltage source for the islanded system maintaining voltage
regulation of the circuit. During this mode, the circuit is isolated from the rest of Miramar’s
distribution system with the installation of a remote operated Main Breaker at the point of common
coupling, replacing the existing main breaker on the B5PS2T3 switchgear. The Main Breaker
opens and closes based on commands from the IPEM subsystem to isolate the circuit from the
grid, thereby meeting the guidance referenced in IEEE 1547.4. Since the ESS acts as the voltage
regulator for the system, it does not currently have the capability of charging while in Islanding
Mode as part of its current control software. Precise control of the zinc plating process is required
for the energy cells to operate efficiently. When operating in voltage control mode, Primus Power
has not fully developed the control systems and algorithms to monitor and maintain uniform zinc
plating that switches from charge to discharge quickly. While the basic principle of rapid discharge
and charge has been demonstrated the software and real time controller code has not been
developed therefore this is a current limitation of the system utilized in this demonstration.
The average peak load for building 6311 on the base is 61kW with a highest peak of 113kW.
Figure 5-2 shows the average load profile for building 6311 for each month of the year.
Figure 5-2: Average daily load profiles by month for building 6311 at MCAS Miramar.

The average daily PV Output for the P196 Carport PV subsystem is shown in Figure 5-3 below
for each month. Data for the month end of September was unavailable but is assumed to be similar
to the months of August and October.
Figure 5-3: Average daily power output of the P196 carport PV subsystem.
Based on the data the P196 PV subsystem generates more power output than building 6311
requires. Because the ESS does not currently have the capability to charge when operating in
voltage control mode and the PV system generates more than the 6311 load, control of the P196
subsystem is required in order to make sure that more power is not generated than is required
during islanding mode. Typical commercial PV inverters are not capable of being actively
curtailed, however the two Advanced Energy inverters that are part of the P196 subsystem were
capable of being enhanced to provide this capability. Raytheon had the two inverters upgraded
with new communication cards and firmware to add a curtailment function to their Modbus
interface.
During islanded operation, the IPEM microgrid controller controls the curtailment set point of the
PV Inverters in order to keep the power generated by the PV below the demand required by the
building. The ESS provides the remaining power delta between what the PV generates and the
required power to meet the load. An example of this behavior is shown in Figure 5-4 below.

Figure 5-4: Simulated load and power output profile for the ZnBr installation during islanded operation.
The green line in the plot shows what the normal power output capacity of the P196 carport PV
subsystem can achieve. The red line shows what the power output capacity of the P196 carport
PV subsystem is predicted to be once controlled by the IPEM subsystem. The power output is
controlled (or curtailed) to always remain below the load. Prior to starting the program it was
unknown as to the amount that the PV would need to be curtailed because it was dependent on the
capabilities of the power electronics within ESS, capabilities of the IPEM controller, the response
time of the AE inverters and the behavior of the Miramar load. Each one of these elements required
detailed modelling, analysis and testing to validate the proper functional behavior required to make
them work together. Early analysis of the microgrid showed that Islanding duration is directly
related to 3 main factors;
1. Battery Energy Capacity: A fixed value based on the amount of energy capable of being stored in
the ESS.
𝑃𝑃𝑃𝑃 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺𝐺
2. PV Penetration: Defined as the ratio of � ∗ 100�. This is limited by ability for
6311 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿
curtailment function in the AE inverters to respond to large drops in load and stability of battery
inverter
3. 6311 Load Management: The ability to be able
a. The largest loads within 6311 are due to Cooling and Interior Lighting (~54%)
b. Currently only method for reducing HVAC cooling loads is manual adjustments of
thermostats
In order to achieve the 72 hr islanding time duration then a combination of high PV penetration
levels and load reductions would be required.

Table 5-1: Table showing the relationship between PV Penetration, Load Reductions and its effect on
Islanded Duration.
The culmination of the design and analysis of the curtailment functionality was when the system
was tested at NREL in December of 2014. This was the first opportunity for bringing together all
of the major subsystems of the microgrid together.
The load profile for building 6311 at Miramar consists of both real and reactive power components.
The reactive component of the Miramar load is mainly due to motor loads from its environmental
controls (heating and air conditioning). A plot of building 6311’s load profile including the real
and reactive power components is shown in Figure 5-5.
Figure 5-5: Load profile for building 6311 including real, reactive and apparent power as a function of time.
The data was sampled at 15 minute intervals. The power factor is plotted on the secondary axis.
As result of the load consisting of a reactive power component power factor needs to be taken into
consideration when managing the PV load. The variable nature of PV production and motor loads
creates transient conditions that require accommodation by the power electronics of the ESS.
Therefore the amount of PV provided to the load needed to be balanced between the capabilities
of the ESS power electronics and the transient conditions of the circuit. During the course of
developing and testing the system it was also determined that the power electronics within ESS

require approximately 10kW of power output to maintain the control loops utilized in managing
the battery’s DC bus.
The second mode of the demonstration is to demonstrate the capability for an ESS to allow a
facility to reduce peaks in power usage by implementing peak shaving algorithms. This capability
is provided by a controlled charge and discharge of the ESS according to a programmed or
automated schedule. The result is that the load profile of grid purchases is changed in the favor of
the facility in order to avoid peak demand and transmission charges. An example of this is shown
in Figure 5-6 below.
Figure 5-6: Plot showing example of how peak shaving can change the load profile of a
facility as seen by the utility.
The independent variables that will be manipulated are defined below:
For Peak Shaving Mode
• Battery discharge/charge rate – The battery discharge/charge rate is the rated power that the Zn/Br
flow battery will charge or discharge during peak shaving mode.
• Battery discharge/charge time – The battery discharge/charge time is the scheduled times that the
Zn/Br flow battery will charge or discharge.
For Islanded Mode
• PV input curtailment – The PV input curtailment is the amount of PV that needs to be
curtailed when operating in islanding mode to avoid power instability. It correlates with
PV penetration of the circuit.
• Load Reduction – The load reduction is the amount of load that needs to be reduced when
compared to normal operations.

The dependent variables observed for this demonstration are:
• ESS SOC – This is the current state of charge of the energy storage system
• Boot-up time of ESS – This is the amount of time it takes the ESS initialize and boot up
into ready mode.
• Switch over time during blackout - This is the time it takes for the ESS to power up the
circuit when commanded to go into islanded mode.
• Successful switch to Islanding mode – This is the determination of whether or not the
switchover to islanding mode is successful
The controlled variables for the demonstration are
• Relative Building Load- The relative building load is the relative percentage of building
load the demonstration will hold to when compared to normal operations
5.2 BASELINE CHARACTERIZATION
The baseline characterization of Miramar’s building 6311 were taken in November 2015, prior to
the December demonstration. The data was collected from the Advanced Metering Infrastructure
(AMI) smart meters that are installed in the B5PS2T3 switchgear.
Figure 5-7: Daily load profiles for building 6311 during November 2015.

Figure 5-8: Average hourly load profile for building 6311 during November 2015.
5.3 DESIGN AND LAYOUT OF TECHNOLOGY COMPONENTS
This demonstration consist of four significant technology elements and they are; 1) The ZnBr ESS
2) The IPEM Microgrid Controller 3) The Switchgear 4) The PV Inverters. The locations and
layouts of each element are shown in Figure 5-9 below.
Figure 5-9: Birds eye view of MCAS Miramar site and layout of system components.

ZnBr Flow Battery
Primus’ ZnBr battery was delivered and installed on May 2015. Pictures of the battery being
delivered are shown in Figure 5-10 below.
Figure 5-10: Photos of the Primus ESS being delivered to MCAS Miramar. The EnergyPod need to have a
large crane in order to lift it off of the delivery truck.
The ESS location is positioned in the parking lot of building 6311 next to the P196 inverter room
and under one of the carport solar panel locations.

Figure 5-11: Construction drawing schematic of ESS location.
Figure 5-12: 3D CAD image of the ESS location.
IPEM Microgrid Controller

Data collection, analysis and system control forms the backbone of IPEM. Performance modeling
and simulation is performed on the system configuration to generate a baseline of data to reference
and optimize from. Optimized C2 code generated from simulation is loaded to the controller board.
The IPEM controller is the central process and decision making device which provides supervisory
control of all subsystems. The controller then uses the live data collected from the subsystems and
other ancillary sensors to dynamically optimize the holistic performance of the energy system.
Live and historical system status is provided to an operator via the HMI display. The HMI display

allows a user to switch between system modes and tailor the system performance based on user
desired parameters. The HMI provides a user with a high level system state display as well as low
level operating parameters for each component.
Figure 5-13: IPEM command and control suite.

The IPEM controller is located inside the P196 inverter room on the west wall. Images of the
IPEM controller as it was installed are shown in Figure 5-14 below.
Figure 5-14: Photos of inside the P196 inverter room where the IPEM controller was installed. The image on
the left shows the location pre IPEM. The middle image shows some of the existing fiber termination block and
equipment moved to accommodate the IPEM controller. The right image shows the IPEM controller and
ancillary equipment installed.
The IPEM HMI is utilized for system control and monitoring. Throughout operation, performance
against both technical objectives can be monitored via visual display of the system operation. This
dashboard style interface displays the live data from all of the various system components that the

controller is monitoring. Status of the system is easily identified through the use of indicators and
gauges (as shown in Figure 5-15).
Figure 5-15: Screen shots of the IPEM HMI .

Figure 5-16: Photo of IPEM HMI being utilized on 12/13/15 islanding demonstration.
5.4 OPERATIONAL TESTING
Testing of the microgrid including the demonstrations was divided into three phases of test: 1)
System Initialization and Checkout 2) Grid Tied Mode and 3) Islanded Mode. Each phase of
testing is described in more detail below.
System Initialization Checkout
System Installation, Integration, and Checkout is anticipated to consist of
emplacement/installation, interconnection (power and communication), and verification of
operation and communication of the equipment described in Section 7 prior to test start. This will
include verification of communications interfaces between various items.
Emplacement/installation and interconnect will be completed by NREL, Primus Power, or
Raytheon as indicated in Table 5-2. Communications interfaces between various items will be
verified by Raytheon, Primus Power and Advanced Energy (AE) in the week prior to test start.
Checkout is considered complete when each item is operational and communication between each
item has been established.
Grid-Tied Mode
The purpose of Grid-Tied Testing is to demonstrate the system is properly configured and
functionally capable of meeting the performance objectives in Table 5-2.
Table 5-2: Grid Tied Performance Objectives for ESTCP Demonstration

Performance Objective Metric Data Requirements Success Criteria
Peak Shaving Peak Demand Reduction Meter readings from RE ESS is able to store energy
system, ESS, and grid power during off peak time and
(kW) feed discharge 250 kW during

peak time to reduce peak
load relative to historical
data over similar time
period.
ESS Energy Storage Energy Discharged in kWh Meter reading of energy ESS is able to discharge
Capacity discharged by ESS 1MWh of energy during
peak shaving cycle.
Grid Tied Mode testing will achieve the following objectives:

1) Verify integrated system functionality and monitoring/fault detection functions of IPEM
Controller in the presence of real PV source and load characteristics
2) Validate scheduled peak shaving functionality in grid-tied mode in the presence of real PV
source and load characteristics
Islanded Mode
The purpose of Islanded Mode Testing is to demonstrate the system is properly configured and
capable of meeting the performance objectives in Table 5-3.
Table 5-3: Islanded Mode Performance Objectives for ESTCP Demonstration.

Performance Objective Metric Data Requirements Success Criteria
Islanded Duration Islanded Meter readings from RE Building loads are met by
system, ESS, and grid power ESS and PV for 72hrs
Duration feed under controlled load
(hours) conditions meeting power
quality standards of
IEE1547.4
Building Load Reductions Delta Average kWh/day Meter readings from building Building loads can be
usage 6311. reduced by 50% through
manual changing of
thermostats and lighting
when compared to its
previous year’s average
for that given month.
Switchover Time Time (minutes and Clock timing from command Time is less than hour
seconds) to go into islanded mode to
ESS discharging power
Islanded mode testing will achieve the following objectives:

1) Validate open transition/Black Start sequencing to commence operation in
islanded/microgrid mode
2) Demonstrate load following operation with ESS-inverter voltage/frequency control in the
presence of real load characteristics
3) Validate monitoring/fault detection functions of IPEM Controller in islanded mode
4) Validate PV curtailment functionality

5) Demonstrate load following operation with ESS-inverter voltage/frequency control in the
presence of real load characteristics and PV
6) Evaluate system power quality (e.g., voltage, frequency, harmonics) as a function of load
characteristics (e.g., transients, power factor)
7) Evaluate system power quality (e.g., voltage, frequency, harmonics) as a function of PV
penetration
5.4.1 Test Configuration
The test configuration is shown in Figure 5-17.
B5-PS2T33 Pad
Switchboard 12000-
208Y/120V
12kV
Distribution
Grid
1000AT
MB1
900AT
800AT
700AT
MB1 Φ Rot.
Telvent Sage 2300/
CTRL Relay Base Grid SCADA
2400
SCADA Hi/Low
Analog
A13 A12
Relay Relay
A4 A5 IO
Energy Pod PowerBox

250kW 250kVA 1000A
480V-208Y
- ~
~ 800A
Disconnect -
Energy
Block Chiller 232kW
~
Building 6311 -
SATCON Fused
Rooftop
~ Energy Management
Command & Control
Data Acquisition
Human
Machine Fiber
Interface Optic
(Client) Modem
Figure 5-17: System configuration block diagram.
Equipment Involved in Testing
Table 5-4: List of equipment used in system test setup.

Provided
QTY Equipment Installed By
By
1 Primus EnergyPod Primus Dynalectric
1 Primus PowerBox Primus Dynalectric
1 Primus Chiller Primus Dynalectric

Provided
QTY Equipment Installed By
By
2 AE100TX solar inverter 100kW Miramar N/A
1 Satcon 30kW Solar PV Inverter Miramar N/A
1 250kVA 480V/208V Wye Transformer Dynalectric Dynalectric
1 IPEM Controller Raytheon Raytheon
1 New B5-PS2T33 Switchboard Dynalectric Dynalectric
1 Fluke 437 Series II Power Analyzer Raytheon Raytheon/NREL
1 Fluke 1735 Data Logger NREL NREL
5.4.2 System Integration and Checkout Tests
The System Integration and Checkout tests encompass the installation and methodical testing of
the various subsystems as they are installed at Miramar and commissioned into a complete system.
PV Communication and Curtailment Functionality Test
PV Communication and Curtailment Functionality Test
The PV Communication and Curtailment test utilizes the two AE 100TX Inverters, Miramar
Utility Grid, and IPEM Controller to test the communication interface and curtailment
functionality of the AE inverters (Figure 5-18). The Miramar Utility Grid is required in order to
generate a stable voltage reference in order for the PV inverters to synchronize and operate.

B5-PS2T33 Pad
Switchboard 12000-
208Y/120V
Energized during 12kV
test Distribution
Grid
1000AT
MB1
900AT
800AT
700AT
Component MB1 Φ Rot.
Telvent Sage 2300/
used in test 2400
SCADA Hi/Low
The AC Breaker on the Inverter Analog
will be Locked Out and Tagged
A13 A12
Out During This Test Relay Relay
A4 A5 IO

250kW/1000kWhr
Meter
Energy Pod PowerBox Disconnects
250kVA
250kW Inverter 480V-208Y
……….. P196 Carport PV and Inverter Room
M Discconnect 100kW Inverter
~
Ctrl Mod 800A
~
Disconnect -
EnergyCell Ancillary
Equipment
Energy
Block Ctlr
- Chiller
400A Fused PV Powered
232kW
STC
~ Discconnect 100kW Inverter PV Array
~
Building 6311 -
SATCON Fused
Rooftop
~ Energy Management
Command & Control
Data Acquisition
Human
Machine Fiber
Interface Optic
(Client) Modem
Figure 5-18: Test configuration for PV Comms and Curtailment Test
Prerequisites
1) The IPEM controller software is up to date and the HMI is running.
2) The Miramar utility grid is active.
3) The two AE inverters and Satcon inverter are energized and no faults reported.
4) The ESS is in Standby Mode.
Notional Test Procedure

1) Verify nominal operation of inverters A & B using the default curtailment setting of 100%
2) Collect data (I,V,etc) from PV inverters A and B using CTs and PTs interfaced to the IPEM Controller
3) Collect data (I,V) using separate power quality analyzers
4) Reduce the curtailment setting to 10% in 10% decrements. Each setting will be maintained for a minimum of 30
seconds.
5) Increase the curtailment setting to 100% in 10% increments. Each setting will be maintained for a minimum of
30 seconds.
Results
The PV curtailment commands were properly sent and implemented on the AE Inverters and
verified by the CTs in the switchgear as well within the power quality analyzers.
Item Description of Desired Outcome Outcome of Test
1 Verify that the PV curtailment commands resulted in the expected Complete

changes in inverter output

Verify agreement between IPEM Controller and power analyzer acquired
2 Complete
data
Energy Storage Charge/Discharge Test
The Energy Storage Charge/Discharge Test utilizes Building 6311, Miramar Utility Grid, Primus
Power ESS, and IPEM Controller to verify communications and control between the ESS and the
IPEM Controller. The Energy Storage Charge/Discharge Test also verifies the general operability
of the simulated Primus Power ESS approach. .
B5-PS2T33 Pad
Switchboard 12000-
208Y/120V
test Distribution
Grid
1000AT
MB1
900AT
800AT
700AT
Telvent Sage 2300/
used in test 2400
SCADA Hi/Low
Analog
A13 A12
Relay Relay
A4 A5 IO
Energy Pod PowerBox

250kW 250kVA 1000A
480V-208Y
- ~
~ 800A
Disconnect -
Energy
Block Chiller 232kW
~
Building 6311 -
SATCON Fused
Rooftop
~ Energy Management
Command & Control
Data Acquisition
Human
Machine Fiber
Interface Optic
(Client) Modem
Figure 5-19: Test configuration for Energy Storage Charge/Discharge Test .
Prerequisites
2) The Miramar utility grid is active.
3) The ESS is in Ready State.

1) If ESS is less than 100% SOC then Command ESS to charge at 140 kW for until the ESS is at 100% SOC
2) Command ESS to discharge at 250 kW for 2 hrs
3) Command ESS to charge at 140 kW for 2 hrs until the ESS is at 50% SOC
4) Command IPEM Controller to step through charge/discharge profile as shown below:
kW Time (min)
10 1 Discharge

30 1 Discharge
50 1 Discharge
100 5 Discharge
150 5 Discharge
200 5 Discharge
250 5 Discharge
0 5 Hold
-140 1 Charge
-160 1 Charge
-180 5 Charge
-200 5 Charge
0 5 Hold
Results
1 Verify dataflow from IPEM Controller to Primus Power Energy Block Complete
Controller
Verify operation of Primus Power Energy Block Controller with Parker
2 Complete
GTI Inverter equipped with DC pre-regulator
Verify ability of IPEM Controller to charge and discharge simulated
3 Complete
Primus Power ESS upon command
5.4.3 Grid Tied Mode Tests
The Grid Tied Tests encompass a subset of tests to demonstrate the Peak Shaving capabilities of
the system and test compliance with IEEE1547 and IEEE1547.1 For Grid Tied tests, the complete
test setup is employed.

B5-PS2T33 Pad
Switchboard 12000-
208Y/120V
test Distribution
Grid
1000AT
MB1
900AT
800AT
700AT
Telvent Sage 2300/
used in test 2400
SCADA Hi/Low
Analog
A13 A12
Relay Relay
A4 A5 IO
Energy Pod PowerBox

250kW 250kVA 1000A
480V-208Y
- ~
~ 800A
Disconnect -
Energy
Block Chiller 232kW
~
Building 6311 -
SATCON Fused
Rooftop
~ Energy Management
Command & Control
Data Acquisition
Human
Machine Fiber
Interface Optic
(Client) Modem
Figure 5-20: Test configuration for Islanded Mode Tests
Scheduled Peak Shaving Test
The scheduled peak shaving test demonstrates the ability of the system to execute scheduled peak
shaving with the IPEM Controller directing charge and discharge of the simulated ESS
Prerequisites
2) The Miramar utility grid is active and the Main Breaker is Closed.

1) Check that the ESS System State is at Ready State
2) Transition system to Peak Shaving Mode via IPEM Controller
3) Verify the following sequence
a) At a pre-determined time, the IPEM Controller directs the ESS to charge at 167kW off of the Miramar Utility
Grid
b) The IPEM Controller monitors simulated ESS SoC
c) At a pre-determined SoC, the IPEM Controller directs the ESS to cease charging
d) At a pre-determined time, the IPEM Controller directs the ESS to discharge at 100kW
e) The IPEM Controller monitors the ESS SoC
4) The user returns the system to the Ready State via the IPEM HMI
Results

The system is able to charge and discharge autonomously to a pre-
1 determined schedule and power levels via automated peak shaving Complete
control from the IPEM Controller
The IPEM Controller monitors ESS SoC and starts and stops charge and
2 Complete
discharge operations in accordance targeted levels
5.4.4 Islanded Mode Tests
The Islanded Mode Tests encompass a subset of tests that step through the various stages of
islanded operations. The Islanded Mode Tests are divided into three subcategories: 1) Pre-Island
Conditions 2) Grid Transition and 3) Full Islanding. For the Islanded Mode Tests all equipment
in the test setup is used.
B5-PS2T33 Pad
Switchboard 12000-
208Y/120V
test Distribution
Grid
1000AT
MB1
900AT
800AT
700AT

Telvent Sage 2300/
used in test 2400
SCADA Hi/Low
Analog
A13 A12
Relay Relay
A4 A5 IO
Energy Pod PowerBox

250kW 250kVA 1000A
480V-208Y
- ~
~ 800A
Disconnect -
Energy
Block Chiller 232kW
~
Building 6311 -
SATCON Fused
Rooftop
~ Energy Management
Command & Control
Data Acquisition
Human
Machine Fiber
Interface Optic
(Client) Modem
Figure 5-21: Test configuration for Islanded Mode Tests
Pre-Island Conditions Test
The purpose of Pre-Island Conditions Test is to demonstrate that the IPEM Controller and other
monitoring and control equipment are functioning properly in order to assess the current state of
the test setup. The IPEM Controller communicates with and pulls status from the various
subsystems and provides that information to the user in order to make decisions about how to
operate the microgrid.

Prerequisites

1) Check/Set ESS System State to Ready State
2) Check ESS State of Charge
3) Check PV status on Inverters A, B
4) Check communications and status on Main Breaker
Results
This test was conducted on 10/24/15. The night before the ESS was brought to 0% SOC then
charged to 100% overnight. The IPEM controller adequately checked the status of all of the various
subsystems 1) ESS 2) PV Systems and 3) the Main Breaker
1 IPEM Controller acquires PV inverters, ESS and Main breaker status data. Complete
Data is made available to the system operator to assess islanding readiness
Island Transition Test
The purpose of the Island Transition Test is to test the behavior of the system when the Main
Breaker is open and closed prior to conducting Full Islanded Testing.
Prerequisites
4) Miramar Operations has been notified of the event.

4) Check communications and status on Main Breaker
5) Open the Main Breaker
6) Check the status and comms of the Main Breaker
7) Check the status and comms of the ESS
8) Confirm the ESS has shut down and gone into standby mode.
9) Check status on PV Inverters A, B verify that they have de-energized
10) Close the Main Breaker

13) Check status on PV Inverters A and B
Results
1 The IPEM Controller allows the Main Breaker to be opened and closed and Complete
acquires expected PV inverter, ESS and Main breaker status data.
This test was conducted on 10/24/15. The IPEM controller displayed the status of each of the
subsystems and sensors of the microgrid. The test however did not use the HMI interface to open
and close the Main Breaker, only to show status. The PV inverters were remotely via IPEM
Modbus put in disable mode (as opposed to opening the AC disconnect) per AE’s
recommendations to NREL. The Satcon inverter was disabled manually by opening the AC and
DC disconnects to the inverter. The Open/Close function for the Main Breaker was conducted
from IPEM but through a manual process of changing the state through a terminal. The Main
Breaker Opened as commanded and power was cut off to all systems. The UPS within the
switchgear maintained power on the Main Breaker and IPEM equipment within the switchgear.
The battery went into its back-up power mode. The PV inverters went offline and this was shown
on the HMI. The Grid Status was reported to the HMI as Inactive, the ESS showed it was in Ready
Mode, the sensors within the switchgear showed there wasn’t any power on the feeder circuits.
IPEM then sent a Close command to the Main Breaker remotely through a terminal interface and
the Main Breaker closed, picking up the load of 6311. Power was returned back to 6311, the ESS
and the PV inverters. The IPEM HMI showed the Grid Status change to Active and the power
levels on each of the feeder circuits as well as the status on the PV inverters. It was noted from
the HMI that one of the PV Inverters (Inverter B) did not establish comms after the normal 5
minute countdown. Once this was noted the inverter was visually inspected in the inverter room
and it was noted that Inverter B’s display did not show anything and there was no indication that
the Inverter had powered up. A similar issue happened to Inverter A in July of 2014 during a
planned outage and was attributed to failure of the auxiliary power supplies. Bob Butt at NREL
had a Fluke Model 1735 Power analyzer connected to the building 6311 feeder circuit and recorded
power quality from the system. The results of his recording are provided below.
Worst case sag lasted 8 ms and was 109 V. Worst swell was 127 V and lasted 291
ms. Transient events would obviously be much faster but the test setup captured
some short duration events, and none appeared to be very significant.
Voltage THD was 2% in Phase A. It will be interesting to see how it changes with
battery system connected.

Figure 5-22: 3-Phase balance for voltage and current.
Figure 5-23: Plot of total real power (kW) and reactive power (kVAR) over time. Power factor ranged from
about 0.86 to 0.89. Peak 6311 load on Sunday was about 70 kW at 1515.
During this and other tests, ground currents were observed at various grounding electrode
connection points, as shown in Figure 5-24. The results of these measurements are shown in Table
5-5
Table 5-5: Results of ground current measurements.

Measurement Date/Time Amps, AC/DC Notes
Location
1 10/24 @1030 320 mA AC
2 10/24 @1030 810 mA DC
3 10/24 @1030 200 mA DC

1 10/24 @1155 350 mA AC Inverter, Energy Pod Aux.
Power Only (chiller,
heaters, controls)
2 10/24 @1155 820 mA DC Aux. Power Only
3 10/24 @1155 380 mA DC Aux. Power Only
1 10/25 @0915 250 mA AC System Testing Underway,
chiller running
2 10/25 @0915 170 mA, 700 mA Testing Underway
DC
3 10/25 @0915 275 mA DC Testing Underway
4 10/25 @0915 265 mA AC Testing Underway

Figure 5-24: Ground current measurement schematic.
Islanded Operation with Battery Only Isolated from Circuit (Self Powered)
The Islanded Operation with Battery Only Isolated from Circuit (Self Powered) test is intended to
go through the sequence of islanding with the battery isolated from the B5PS2T33 circuit. This is
meant to exercise the sequence of commanding the battery to Islanding Mode from IPEM and
allowing the battery to power up and provide power to its own overhead lead. Building 6311 will
not lose power during this test. One of the AC interconnects to the battery will be Opened during
this test to simulate a loss of power to the battery and prevent feeding power onto the B5PS2T33
circuit.
Prerequisites

4) The ESS is charged to 100% SOC and remains higher than 90% prior to test
5) Outage approval has been granted by Miramar

5) Open one of the AC disconnect switches that feed into the Primus 208Y/480V transformer (TBR).
6) Verify that the battery has loss of power and goes into standby mode.
7) Set the ESS in Islanded Mode Manually through Primus EnergyBlock Controller Interface
8) Verify that the following sequence occurs
a) The ESS restarts its power electronics in its stand alone setting and sends a message to the IPEM Controller
that it is in islanding mode and provides the initial power to the load.
9) Monitor the load quality data being provided by the ESS through the IPEM HMI and power analyzers
10) Disable the ESS from being in Islanding Mode Manually through the EnergyBlock Controller Interface.
a) The EnergyBlock Controller commands the ESS to de-energize in Standby Mode causing power loss to the
System.
12) Close the AC disconnect switches that feed into the Primus 208Y/480V transformer.
13) Verify that power is restored the Primus system and shows it’s in its Ready Mode.
14) Repeat the previous steps as many times as required until sequence occurs smoothly then follow the sequence
below.
15) Open one of the AC disconnect switches that feed into the Primus 208Y/480V transformer.
16) Verify that the battery has loss of power and goes into standby mode.
17) Set the ESS in Islanded Mode Manually through IPEM
a) The ESS restarts its power electronics in its stand alone setting and sends a message to the IPEM Controller
that it is in islanding mode and provides the initial power to the load.
20) Disable the ESS from being in Islanding Mode Manually through IPEM.
a) IPEM commands the ESS to de-energize in Standby Mode causing power loss to the System.
22) Close the AC disconnect switches that feed into the Primus 208Y/480V transformer (TBR).
23) Verify that power is restored the Primus system and shows its in its Ready Mode.
Results
1 The battery successfully transitions to Islanding Mode and provides the Complete
initial load to its overhead systems with no faults.
2 The battery successfully transitions to Grid Tied mode after Islanding test Complete
is complete.

This test was conducted multiple times throughout the weekend 10/23/15-10/25/15. The entry
criteria was met and the battery side of the AC disconnect was opened at the SDG&E meter
cabinet. This shut down power to the ESS. The battery would register the power outage and put
the system in a Standby state reducing its overhead power while running on its UPS. The Primus
EnergyBlock controller then commanded the ESS to go into Islanding mode triggering a series of
events to power the EnergyCells to bring up the DC bus and then close the AC breaker on the
Parker Inverter. Going through this process the first couple of times the team detected multiple
sequences that needed to be re-coded as the flow of events required to go into islanding mode was
better understood. One of the concerns going into this test was the phase rotation of the AC output
of the Parker inverter during islanding mode. The phase rotation output of the Parker Inverter is
hard coded into the system for its Islanded mode output. To determine if the battery was able to
match the phase rotation of the 6311 load, it was noted that the ESS Chiller has the same phase
rotation settings as 6311. If the ESS was able to pick up the load of the chiller then the settings
for the phase rotation on the Parker inverter should be adequate for the 6311 load. When the ESS
was brought up in Islanded mode the chiller it powered up successfully, therefore it is expected to
adequately match the phase rotation required for 6311. For additional confirmation, the Ion 8600
meter at the main switchboard MSB1 displays phase rotation, and could be used to check utility
and ESS phase rotation before B6311 is energized by the ESS. After this test was completed the
ESS system was shut down and the system was deemed ready for Islanded Operation with Battery
Only Test. It was also noted that during one of the controlled outages on 10/24/15 the AE inverters
were disabled remotely via the Modbus interface then the AC power was interrupted. Upon return
of power on of the AE inverters had a power supply failure and did not power back up. AE
technical services was notified and repairs were scheduled however the AE Inverter B was not
available during the tests leaving only one functional 100kW inverter for islanding testing.
This test was repeated on 10/25/15 after attempting the Islanded Operation with Battery Only Test
to allow Primus the ability to determine the cause of why the EnergyCells MOSFETs were being
damaged and to iron out the sequencing for the ESS to be brought up and successfully brought out
of Islanding Mode.
After conducting the Islanded Operation with Battery Only test on 10/25/15 there were a couple
sequencing issues that needed to be investigated to avoid further damaging more EnergyCells in
the ESS. At this time Islanded Operation with Battery Only Isolated from Circuit (Self Powered)
was then revisited to understand the proper sequence of shutting down the ESS to avoid damaging
the EnergyCells. The weekend concluded with still running the Islanded Operation with Battery
Only Isolated from Circuit (Self Powered) test. During the tests the team was going through
iterations of the ESS code to reduce the voltage stress on the EnergyCells in a controlled manner.
Without this sequence working properly the EnergyCells were damaging the MOSFETs in their
power electronics which requires removal of the EnergyCell to repair. Therefore the Primus team
was prioritized to concentrate on this issue so as to avoid damaging any further EnergyCells. The
ESS needs at least 7 EnergyCells to provide adequate DC voltage and current for the load. At the
start of the weekend there were 13 out of 14 EnergyCells functioning. During the Islanded
Operation with Battery Only testing 3 more of the EnergyCells failed due to damaged MOSFETs.
After going back to conducting Islanded Operation with Battery Only Isolated from Circuit (Self
Powered) a couple more EnergyCells failed in the afternoon dropping the system to below 5
EnergyCells. It was at this time the Islanding demo was concluded to allow Primus to finish their
assessment of the proper shutdown sequence and repair the damaged MOSFETs within the faulted
EnergyCells.

Islanded Operation with Battery Only
The Island with Battery only test is intended to go through the sequence of islanding prior to
establish a performance baseline prior introducing the PV generation into the microgrid. The PV
will be manually disabled for this test.
Prerequisites

4) Manually de-energize the PV inverters by disabling them through their communications interface (e.g. Modbus).
5) Check PV status on Inverters A and B and verify they are offline
8) Enter Islanded Mode Manually through IPEM HMI
a) The Main Breaker is opened and the utility power to BS5PS2T3 is disrupted.
b) Upon loss of power the System transitions into Islanding Mode. During the transition to Islanding Mode the
following steps occur:
i) ESS de-energizes and goes into back-up power mode but still is in its grid connected setting and waits
for the islanding command from the IPEM controller.
ii) The IPEM Controller and HMI stay on-line powered by its UPS. The HMI displays that the Grid is
down, reports the loss of comms with the Inverters as it transitions into Islanding Mode.
iii) The IPEM Controller commands the ESS to enter Islanding Mode. The ESS restarts its power electronics
in its stand alone setting and sends a message to the IPEM Controller that it is in islanding mode and
provides the initial power to the load.
11) Disable Islanding Mode Manually through the IPEM HMI
a) The IPEM Controller commands the ESS to de-energize in Standby Mode causing power loss to the System.
The IPEM Controller and ESS operate off of back-up power and the System goes into Grid Transition Mode.
b) The IPEM Controller commands the Main Breaker to close returning grid power to the System.
c) The IPEM Controller commands the ESS to return to its grid connected setting.
d) The System is now in Ready Mode.
Success Criteria

1 PV inverter faults are identified and displayed to the system user via the Complete
IPEM HMI
The system maintains voltage, frequency, phase-balance and harmonics
2 within a pre-determined range while PV inverters are faulted and brought Complete
off line
3 The PV inverters can be brought back on-line and resume operation in Complete
accordance with the pre-determined PV curtailment levels
This test was conducted in the afternoon (~15:15) on 10/24/15. The PV inverters were disable
remotely via IPEM. The Phase Rotation on 6311 was inspected on the ION 8600 smart meters to
start. They read: Phase A – 0.0, Phase B -120, Phase C 119. At approximately 15:57 the Main
Breaker was opened via the MOXA IO device. Power was interrupted to B5PS2T3 switchgear.
The IPEM controller, ESS, and switchgear all stayed on on their appropriate UPSs. The team used
the Primus EnergyBlock controller to bring the battery up in Islanding mode. During the startup
process the ESS faulted and it appeared some EnergyCells reported faults as well. The test was
aborted at this time and IPEM reclosed the Main Breaker returning power back to the system. A
plot of the outage is shown in Figure 5-25 below. Upon further inspection it was determined that
fault in the EnergyCells was due to damaging the MOSFETs during the test. This was suspected
to be caused by the shutdown sequence of the ESS when exiting Islanding Mode. At the end of
the test the team needed time to analyze the shutdown sequence and software code with their
software engineers. The day was concluded until one of Primus’ Software Engineers flew in the
morning on Sunday 10/25/15 and could review the software code and sequencing.
Figure 5-25: Timeline for Islanded Operation with Battery Only test on 10/25/15.
This test was repeated on the morning of 12/12/15 after Primus Power was able to validate the
proper sequencing for the shutdown of EnergyCells when exiting Islanding Mode. Once validated
another islanding demonstration test was scheduled for 12/11/15 to 12/13/15. On 12/10/15 a
component failure occurred in one of the H-Bridges of one of the EnergyCells activated a smoke
detector within the EnergyPod. The audible alarms were triggered and MCAS personnel heard

the alarm and notified the Miramar Fire Department. The event did not result in a fire however
the event signifies the importance of having proper safety monitoring and fire protection
mechanisms in place. Due to the damaged H-Bridge the EnergyCell with that component as well
as an adjacent EnergyCell would not be available during the 12/11-12/13 islanding tests. Therefore
the system would only have 12 of the 14 EnergyCells available for testing.
Figure 5-26: Photo of damaged H-bridge system.

On the morning of 12/12/15 the Islanded Operation with Battery Only Test was repeated and the
system successfully islanded at 9:31AM. The microgrid was commanded via the IPEM controller
to enter islanding model. The Main Breaker was opened at 9:27AM and Building 6311 lost power.
The IPEM controller commanded the battery to start up in islanding mode and at 9:31AM the ESS
picked up the loads for 6311 (Figure 5-27 and Figure 5-28).

Figure 5-27: Summary of islanding data on the morning of 12/12/2015.
Figure 5-28: Voltage and current waveform of building 6311’s load provided by the ESS.
The system was allowed to run to verify stability in the ESS’ ability to manage the load on the
building. The ESS was monitored and showed it was regulating the voltage within normal
operating parameters (Figure 5-29).

Figure 5-29: Detailed load data taken from Fluke 437II that was attached to Building 6311 circuit during
12/12/15 morning islanding test.
Once everything was determined to be stable, at 9:43 AM a Chevy Volt was plugged into the
circuit to add a large battery load to the building to see if the ESS could handle the type of load.

Figure 5-30: Photo of Chevy Volt used in islanding demonstration outside of building 6311.
At 10:17AM one of the AE inverters was turned on to provide 5kW of constant power output to
check the ESS response to other inverter generation sources (Figure 5-31).
Figure 5-31: Data from Fluke 1735 meter that was attached to the P196 PV circuit.
For the morning microgrid the system ran until 10:55AM. At this time the team wanted to make
sure all of the safety interlocks were functioning that prevent the microgrid from reconnecting to
the grid with the ESS still in voltage control mode which could cause catastrophic failure. Within
the IPEM controller subsystem there are both electrical relays and software logic that prevent this
from happening. The team took the time to monitor all of the software code and subsystem status
to make sure the proper logic was followed for a reconnect. As an extra precaution the team had
the 12kV feeder breaker to the B5PS2T3 switchgear opened. The team then exited islanding mode
and attempted a re-connect with the 12kV feed de-energized. There is a phase rotation relay at

B5PS2T3 main breaker that determines if the Miramar distribution system is energized and in
phase. The IPEM controller will not allow the Main Breaker to close when attempting to exit
islanding mode if Grid is not present on the primary side of the Main Breaker. Because the 12kV
feed on the primary side was de-energized the IPEM controller should recognize this and prevent
the Main Breaker form closing. During the attempted re-connect the IPEM controller properly de-
energized the battery and the phase rotation relay prevented the Main Breaker from closing
showing that the safety interlocks were functioning properly. At this time the team closed the
12kV feeder breaker, re-energizing the primary side of the B5PS2T3 switchgear. At 11:23AM the
IPEM controller closed the Main Breaker and grid power was restored to Building 6311.
Intentional Island with PV Tests
The Intentional Island with PV Test exercises the Islanding Scenario with the PV system available
as a DR along with the ESS using load from Building 6311 un-altered. The purpose of the test is
to conduct an end to end island scenario to characterize the behavior of the system using the ESS
and PV to meet load demands while isolated from the Miramar Utility Grid. The PV system will
be curtailed at various PV penetration/power ratio levels to determine what levels of PV
penetration generate instability of the distributed generation outside of a pre-determined range of
conditions. Power quality will be characterized as a function of load power factor and in the
presence of load transients. Component (e.g., inverter) level faults will be introduced into the
system to verify the system’s ability to identify and recover from component fault conditions.
Prerequisites


2) Check that the ESS State of Charge is > 90%
4) Check communications and status on Main Breaker, the Main Breaker should be closed and the communications
status green
5) Enter Islanded Mode Manually through IPEM HMI
a) The Main Breaker Opens and reports this status to the HMI
b) The utility power to Building 6311 is disrupted
c) Upon loss of power the system transitions into Islanding Mode. During the transition to Islanding Mode the
following steps occur:
i) ESS de-energizes and goes into back-up power mode but still is in its grid connected setting and waits
for the islanding command from the IPEM controller.
ii) The PV Inverters A & B de-energize due to loss of grid presence.
iii) The IPEM Controller and HMI stay on-line powered by its UPS. The HMI displays that the Grid is
down, reports the loss of comms with the Inverters as it transitions into Islanding Mode.
iv) The IPEM Controller directs the ESS to enter Islanding Mode. The ESS restarts its power electronics in
its standalone setting and sends a message to the IPEM Controller that it is in Islanding Mode, and
provides the initial power to the load. The Islanding clock starts.
v) The IPEM Controller calculates what the load of the system is within the first 5 minutes and sets the
initial curtailment set point for the PV Inverters below 50% of the load. As the load changes the IPEM
subsystem changes the curtailment set point of the PV Inverters to remain 50% below the required load
amount.
7) Monitor the load quality data being provided by the ESS and PV through the IPEM HMI and power analyzers.
8) Repeat previous steps increasing the PV penetration in 5% increments. At each increment assess power quality
provided to the load for compliance with IEEE1547.4
9) Determine which PV penetration level the power quality to the load exceeds the requirements in
IEEE1547.4Error! Reference source not found.
10) Disable Islanding Mode Manually through the IPEM HMI
a) The IPEM Subsystem commands the PV inverters to disable.
b) The IPEM Subsystem commands the ESS to de-energize in Standby Mode causing power loss to the System.
The IPEM subsystem and ESS operate off of back-up power and the System goes into starts its transition
into Grid Tied Mode.
c) The IPEM Subsystem commands the Main Breaker to close returning grid power to the System.
d) The IPEM Subsystem commands the PV Inverters to enable and sets the curtailment to 100%.
e) The IPEM Subsystem commands the ESS to return to its grid connected setting.
f) The System is now in Grid Tied Mode.
Results

1 The system successfully transitions from grid-connected to islanded mode Complete
via direction provided through the IPEM HMI
The system maintains phase balance, voltage, frequency and harmonics
2 within pre-determined limits for all, or a subset of tested PV penetration Complete
levels
3 The system successfully transitions from islanded mode to grid-connected Complete
mode via direction through the IPEM HMI
This test was conducted on the afternoon of 12/12/2015 and through the day on 12/13/2015. On
the afternoon of 12/12/2015 after the successful completion of the Islanded Operation with Battery
Only Test in the morning the system was then tested with increased amounts of PV to further check
the microgrids ability to function with shared inverter based generation sources. For this test to
simulate a Grid power loss the 12kV feed to the B5PS2T3 was opened, causing power loss to
Building 6311. The 12kV feed was opened at 3:19PM. This was detected and shown on the IPEM
HMI. The system was put into Islanding mode via the IPEM controller. The ESS was enabled
and picked up the loads at 3:25PM0 (Figure 5-32).
Figure 5-32: Power data from the Fluke 437 that was attached to building 6311 circuit during 12/12/15
afternoon islanding.
At about 3:36 one of the AE inverters was enabled and manually curtailed through the AE inverter
Modbus interface at 5kW output. The auto-curtailment feature within IPEM was being updated
within and was not available for this test so manual curtailment set points were used through the
Modbus interface on the AE inverter. At 3:44 the curtailment was set to 10kW output, then 15kW,
and then ultimately curtailed at 20kW. Due to sun beginning wane over the horizon PV was maxed
out at ~19kW and slowly started to decrease as the sun continued to set dropping to 17kW. At this
time the team wanted to observe a sudden drop out of PV generation and see the response of the
ESS so the AE inverter was disabled at ~3:50PM. The ESS responded generate more current as it
considered it an increase in load. At 3:56 the AE inverter was re-enabled at the 20kW curtailment
setting and the maximum output it could generate was ~14kW. The profile for the PV generation
is shown in Figure 5-33 below.

Figure 5-33: Power data from the Fluke 1735 meter that was attached to the P196 PV circuit during 12/12/15
afternoon islanding.
As the AE inverter energized one of the EnergyCells was starting to lose voltage and took itself
out of current source. This in addition to interaction with the AE inverter started causing
oscillations in the voltage control of the microgrid. Small flickers were observed in the lights
within the building. The AC waveform during this time was recorded on the Fluke 437 and is
shown in Figure 5-34.
Figure 5-34: The AC waveform collected from the Fluke 437 during the final seconds of islanding the system.

Figure 5-35:
At this point the team decided to that it had enough data on the system interactions and decided to
end the islanding test. The team then gathered the data and made updates to the control logic of
the system for another round of islanding tests the next day. The IPEM controller disabled the
ESS and the Miramar operations crew closed the 12kV feed to the B5PS2T3 primary side. The
IPEM controller then closed the Main Breaker and power was restored to 6311.
On 12/1315 this test was repeated with the expectation of including the auto-curtailment
functionality from the IPEM controller to island the system and determine the maximum PV
penetration level achievable while still maintaining stability of the system. The ESS was charged
overnight starting at 3:00AM until ~8:00AM (Figure 5-36). The starting conditions for this test
were 12 out of 14 EnergyCells were fully functional in the ESS, one 100kW AE inverter was fully
functional, and full functionality of the IPEM controller.

Islanding Event
Charging Time
Figure 5-36:
At 9:19AM the 12kV feed to B5PS2T3 was opened shutting off power to building 6311. The
system was commanded to enter Islanding Mode via the IPEM HMI. The IPEM controller
commanded the ESS to boot up in Islanding Mode and after about 4 minutes the ESS picked up
the load on the building. Once the load was picked up the functioning AE inverter detected a firm
grid presence and started its boot up and 5 minute countdown to energize per its UL1741
requirements. At this time the team really wanted to determine the maximum PV penetration that
the system could achieve while in a real islanding situation. The ESS needs to provide a minimum
level of current in order for its current voltage control logic to remain stable. The minimum
required current equated to the battery needed to constantly output a minimum of 10kW of power.
The Primus team wanted to keep a comfortable margin so they suggested maintaining 20kW of
minimum power output of the battery. The team was targeting to get higher than 75% PV
penetration. Because of the battery needing to output a minimum of 20 kW the load on the building
needed to be at least 80kW, as shown in Figure 5-37.

Figure 5-37: Graph showing the theoretical relationship of building load, PV generation, and minimum battery
generation output and how it relates to PV penetration levels.
When the Islanding test started the building load was running about between 30-50kW. The team
needed to get the load on the building higher so the team increased the load by turning on all of
the AC systems, various space heaters, plugging in the Chevy Volt, and turning on all of the
computers in the building. As the load started to increase around 9:54AM the AE PV inverter was
enabled with using the auto-curtailment function (Figure 5-38).

Figure 5-38: Screenshots of the IPEM HMI during the 12/13/15 islanding event. Screenshot shows the initial
loads of building 6311, ESS status and PV generation status. There is currently no PV output but the status on
Carport A inverter (shown in red square) indicates the inverter is online and in its Startup/Bootup delay per
UL1741 in the left image and in its Idle mode ready to output power in the right image.
The auto-curtailment function automatically changes the output power of the PV system to stay
under a maximum PV penetration value. As the load increases the IPEM controller increases or
decreases the PV power output to stay within the set PV penetration limit. The PV inverter was
initially enabled at 30% PV penetration at 9:59AM and was steadily increased by 5% increments

until 10:32AM when the PV penetration was set to 75%. The PV penetration was then ramped
down at 5% increments until 10:37AM when it reached 50% and was left for 26 minutes (Figure
5-39).
Figure 5-39: Screenshots of the IPEM HMI during the 12/13/15 islanding event. These screenshots show the
initial output power of the PV system and its relationship with the load and the ESS output power. The image
on the top is at 9:59AM when the PV penetration level was initial set at 30%. The image on the right is at
10:34AM when the PV penetration level was set at 65%.

At 10:53 AM the PV penetration level was increased from 50% to 70% and then steadily increased
to 80% (Figure 5-40).
Figure 5-40:
At the 80% setting the PV system was maxed out due to the available sunlight during the day. At
this point the maximum PV penetration level was achieved at 11:04AM at 79%. At 11:30AM the
PV penetration was set back to 60% and then back to 50% at 11:50AM.

Just before noon clouds started appearing on the horizon heading towards Miramar (Figure 5-41).
At ~12:00PM larger clouds started passing over the carport PV system causing the PV generation
to drop very quickly.
Figure 5-41: Clouds forming in the southwestern sky.

As the PV generation dropped the ESS responded to control the voltage by provided more output
power to meet the load. This occurred multiple times through the day. A complete summary of
the load profile, PV generation, ESS charge/discharge for the islanding event is shown in Figure
5-42 below. The largest cloud transient occurred at 11:58:43 and the PV output went from 43kW
down to 0kW in two seconds (21.5 kW/s).

Figure 5-42: Summary load and generation profile during 12/13/15 islanding demonstration test.
At 2:34PM the ESS reported a fault which caused its central regulator to ramp itself down and go
into its inactive mode. The building lost power and the IPEM controller reported the ESS in its
Inactive state (Figure 5-43).

Figure 5-43: Screenshot of IPEM HMI at taken 2:34 when the ESS exhibited a fault and went into its inactive
mode.
At this point the 12kV breaker to the B5PS2T3 circuit was closed re-energizing the primary side.
The system was commanded to exit Islanding mode and reconnect to the grid via the IPEM HMI.
The Main Breaker was then closed and power was restored to the building. This concluded the
final islanding test of the system. Data was collected from all metering devices and the system
was set back into Standby mode. The system successfully islanded under multiple load conditions
for 5 hours and 10 minutes.
5.5 SAMPLING PROTOCOL
The sampling protocol during the various operational tests and demonstration are described below.
Data Description
• Sample Rate = (1-5 second intervals for IPEM controller, subsecond intervals for power
analyzers)
• Grid input
• PV input
• Building load and quality (PF, CF)
• ESS power level and direction (charges vs discharge)
• Data transmission (to and from IPEM, ESS, PV Inverter)
• Response time
Data Collector(s)

• Raytheon Personnel
Data Recording.
• Automated:
• The IPEM control unit will log all variables in its internal database
• Calibrated Power measurement equipment will be use to validate the IPEM data
Data Storage and Backup
• IPEM controller employs built in flash memory which will store all collected data
• Remote monitoring data storage unit
Data Collection Diagram:
Independent Power Analyzers B5-PS2T33 Pad

Switchboard 12000-
208Y/120V
Fluke Fluke 12kV
Distribution
437 1735 Grid
1000AT
MB1
900AT
800AT
700AT
MB1 Φ Rot.
CTRL Relay
A13 A12
Relay Relay
A4 A5 IO
Energy Pod PowerBox

250kW 250kVA 1000A
480V-208Y
- ~
~ 800A
Disconnect -
Energy
Block Chiller 232kW
~
Building 6311 -
SATCON Fused
Rooftop
~ Energy Management
Command & Control
Data Acquisition
Human
Machine Fiber
Interface Optic
(Client) Modem
Figure 5-44: Detailed schematic of the interconnection of the various subsystem components of the
installation.
5.6 SAMPLING RESULTS
Provide a detailed summary of all sampling results in terms of both temporal and spatial
dependence as appropriate. Liberal use of graphics and tables is encouraged. The Final Report
serves as the archived document for all data gathered during the demonstration. All results should
be reported in this section or summarized and provided in detail in appendices.

6 PERFORMANCE ASSESSMENT
6.1 ISLANDING DURATION
The success criteria for this performance objective was that building loads would be met by the
ESS and PV for at least 72hrs under controlled load conditions meeting power quality standards
of IEE1547.4.
During the final demonstration tests on 12/13/15 the system was able to successfully island for 5
hours and 10 minutes. This was assessed by calculating the time period that the building 6311 was
picked up by the ESS and when the load couldn’t be sustained anymore and the building lost
power. The data was also analyzed to determine if the quality of power met IEEE1547.4
guidelines. The IEEE1547.4 document describes many guidelines for meeting the load conditions
for the microgrid and is dependent on fully understanding the existing load conditions that the
microgrid will need to maintain. Ranges for meeting power quality standards are contained in
ANSI/NEMA C84.1-2006 and referenced in IEEE1547.4. A summary of the important
requirements listed in 1547.4 are shown in Table 6-1 along with the description of compliance
based on data collected during islanding testing.
Table 6-1: Summarized IEEE1547.4 requirements pertinent to ESTCP demonstration
IEEE 1547.4
Paragraph Requirement Description Compliance Description
No
Voltage and frequency were maintained to
The planned DR island system shall maintain voltage and frequency for the
4.2 ANSI/NEMA C84.1-2006 ranges during the
entire island system including the non-participating DR systems and loads.
demonstration.
L1 Ave = 118
In a planned island loads shall be balanced for each phase. [Calculation for
L2 Ave = 118
voltage balance is in C84.1 -2006 and should limit unbalance to 3%.
L3 Ave = 118
4.2 Example: with phase-to-phase voltages of 230, 232, and 225, the average is
Max Deviation from Ave = 0
229; the maximum deviation from average is 4; and the percent unbalance
(100x0)/118 = 0
is (100 X 4)/229 = 1. 75 percent.]
Data shown in Figure 6-2
The reactive power requirements of the DR island system during the island Voltages were maintained within
condition are important to consider. DR shall support real and reactive load ANSI/NEMA C84.1-2006 ranges under
requirements at an acceptable voltage level. The reactive power reactive power conditions.
5.1.2
requirements of the load during island conditions needs to be understood in
relation to the real power requirements of the load and the DR island
reactive power resources.
Reactive power resources shall be sufficient not only to address steady-state The ESS provided sufficient reactive power
reactive power demands, but also to address dynamic reactive power to address dynamic reactive power demands.
demands, such as those related to motor starting within the DR island HVAC units were utilized to create reactive
5.1.2 system. There are possible interactions between the customer’s and area power loads.
EPS’s power factor correction equipment and synchronous motors and DR.
There needs to be sufficient reactive power resources available when
operating induction or some inverter-based DR.
DR island systems shall be capable of starting and maintaining motor HVAC units within building 6311 were
operations. Motor-starting inrush current can exacerbate voltage drops in turned on repeatedly during testing to create
the DR island system. This voltage drop may result in a degraded ability to motor-starting inrush currents. The ESS was
start the motor or cause loss of generation. Extended motor acceleration able to meet these loads while maintaining
5.1.4
times may cause excess heating, which may reduce motor life and may voltage levels per ANSI/NEMA C84.1-
cause motor overcurrent protective devices to operate. Soft- start controllers 2006.
or reduced voltage starters on large motors can reduce inrush currents and
thus minimize their impacts.
The DR island system shall provide the real and reactive power Variable load conditions were created
4.4.3 requirements of the loads within the island and serve the range of load during islanding tests and they were all met.
operating conditions. [TBR – Using Miramar 6311 Load Data]
The DR island system shall actively regulate voltage and frequency within L1 Vmax = 121.63 / L1 Vmin = 111/18
4.4.3 & 6.1 the agreed upon ranges (e.g., as specified in ANSI/NEMA C84.1-2006 for L2 Vmax = 121.42 / L2 Vmin = 110.64
DR island systems that include the area EPS). Voltage regulation equipment L3 Vmax = 121.46 / L3 Vmin = 109.97

within the DR island system may need to be modified to meet the needs of Shown in Figure 6-2 and Figure 6-1
the DR island system. [TBR – 184Y/106V to 220Y/127V, 59.3 Hz to 60.5
Hz]
Transient load steps were created with
HVAC units kicking on repeatedly as well
During the island mode condition, transient stability shall be maintained for
4.4.3 as PV generation sources turning off during
load steps, DR unit outage, and island faults.
islanding testes. The system maintained
power quality throughout the demonstration.
Both ESS and PV power were utilized in the
If there are multiple DR units in the DR island system, their operation shall islanding demonstration. The PV and ESS
4.4.3
be managed and coordinated to effectively meet the needs of the island. were coordinated by the IPEM controller
adequately during the test.
Once the DR island system is paralleled to the area EPS, all DR shall return The goal for the project was to re-connect
4.4.4 to IEEE 1547 compliance within area EPS time requirements. [TBR 1hr in the system within 1hr and this was achieved
the Demo Plan] during the testing.
Figure 6-1: Frequency measurements during 12/13/15 islanding demonstration test. Data was taken from
Fluke 437 power analyzer. Frequency was maintained at a very stable 60hz.

Figure 6-2: Phase to phase voltage data during 12/13/15 islanding demonstration test.
After post processing the data collection and further investigation it was determined that the reason
the battery went into inactive mode was because there was a power supply failure in one of the
control boxes of an EnergyCell. This resulted in the loss of gate power to one of the H-bridges
which triggered a fast fault in the ESS causing the central regulator to ramp itself down and set the
battery in inactive mode. Therefore it was concluded that there was still energy capacity still
remaining in the battery when it went inactive. This is supported by voltage measurements
collected on the DC string voltage in the ESS and the DC power injected into the Parker Inverter
(Figure 6-3). The ESS discharged ~159kWh of energy during the demonstration. The ESS has
been calculated to have ~290kWh of energy capacity based on the energy capacity tests. This
would have left ~131kWh of energy remaining in the ESS. The average load from 6311 was
~64kW during the islanding demonstration therefore the Islanding demonstration should have been
able to run for another 2 hrs at the average 64kW load. This would have put the islanding time at
a theoretical 7 hours and 10 minutes for those load conditions.

Figure 6-3: Voltage measurements on the EnergyCell string and DC output power from the Primus central
regulator going into the Parker inverter. At ~1:40PM one of the twelve EnergyCells active during the test
reached its end of discharge and took itself out of the string, dropping the string voltage. The string voltage
appears stable all the way to point the fault occurred.
6.2 BUILDING LOAD REDUCTIONS
The success criteria for this objective is that building loads can be reduced by 50% through manual
changing of thermostats and lighting when compared to its previous year’s average for that given
month. Building load reduction capability was calculated to be 68% from manual changing of
thermostats and DDC control set points during the islanding testing. The data showing the
increased manual load steps is shown in Figure 6-4 and represented in 100*(25-78)/78 = -68%.

Figure 6-4: Load profile from 12/13/15 islanding demonstration test highlighting load steps from manually
increasing HVAC and building loads.
6.3 SWITCHOVER TIME
Switchover time is the time from when the system is commanded to enter islanding mode to the
time power is restored to building 6311 by the microgrid. The success criteria for this Performance
Objective is defined to be is less than 1 hour. During the 12/13/15 islanding demonstration test the
time it took the system to transition into islanding mode was recorded at 3 minutes and 47 seconds
and is shown in Figure 6-5. When the islanding event was over and the system needed to restore
grid power the time it took for the system to re-connect to the grid was also recorded and was 7
minutes and 1 second. The timeline for switching the system into islanding mode starts when the
system is commanded via the IPEM HMI to enter islanding. The IPEM controller disables the PV
inverters, sets the ESS in standby, checks the safety interlocks within the switchgear then opens
the main breaker. The IPEM controller then commands the ESS to enter islanding mode. This
reboots the Parker Inverter within the ESS in voltage control mode which takes under a minute.
Once booted successfully the ESS starts to ramp up the voltage on the DC bus. This takes a couple
minutes for each EnergyCell to be added to the DC bus. Once the DC bus is above 600V the
Parker inverter closes its AC breaker and power is provided to the microgrid. For switching out
of islanding the system is restored to grid power via the IPEM HMI. The IPEM controller then
disables the ESS and PV inverters if the they are still running, if not it sets then in standby while
they are in backup power mode. The IPEM controller then checks the status of the base Grid to
see if it is active through the phase rotation relay. If the Grid is present and everything is in standby
IPEM closes the main breaker restoring power to the building. This is shown in Figure 6-6.

1150
L1N (V) Half Cy cle Voltage Min
650
103.345
0m:0s
0.101 3m:47s 113.509
111.449
104.447
106.738
111.183
0m:0s Half Cy cle Voltage Max
150 Half Cy cle Voltage Av g
-350
1150
Half Cy cle Voltage Min
L2N (V)
650
0m:0s
92.586 110.845
107.909
111.014
-350
1150
L3N (V)
650
96.913
0m:0s 111.329
105.545
111.157
-350
0.232
NG (V)
0.182 Half Cy cle Voltage Max

Half Cy cle Voltage Av g
0.132
225
Half Cy cle Current Min
L1 (A)
125 Half Cy cle Current Max

25 Half Cy cle Current Av g
-75
225
L2 (A)

-75
225
L3 (A)

-75
45
25
N (A)
Half Cy cle Current Max

Half Cy cle Current Av g
5
-15
09:19:30 09:20:00 09:20:30 09:21:00 09:21:30 09:22:00 09:22:30 09:23:00
Figure 6-5: Outage time from 12/13/15 islanding demonstration test to enter islanding mode. Grey box in L1N
shows time recorded by Fluke 437 showing when the voltage was dropped and then restored.
3600
L1N (V)
1600 101.586 112.63

0m:0s 0m:0s Half Cy cle Voltage Max
-400 Half Cy cle Voltage Av g
-2400
3600
L2N (V)
1600 104.154 119.32

0m:0s 0m:0s Half Cy cle Voltage Max
-2400
3600
L3N (V)
1600 106.315
0.089 117.303
0m:0s 7m:1s 0m:0s Half Cy cle Voltage Max
-2400
0.265
0.215
NG (V)
Half Cy cle Voltage Max

0.165 Half Cy cle Voltage Av g
0.115
560
360
L1 (A)

-40
900
L2 (A)

-100
560
360
L3 (A)

-40
112
N (A)

12
14:34 14:35 14:36 14:37 14:38 14:39 14:40 14:41
Figure 6-6: Outage time from 12/13/15 islanding demonstration test to exit islanding mode. Grey box in L3N
shows time recorded by Fluke 437 showing when the voltage was dropped and then restored.

6.4 PEAK SHAVING
There are two pieces of data required to calculate the Peak Shaving metric. The first is relevant
historical load profile data. This data was collected a couple days prior to using the ESS in peak
shaving mode and is shown in Figure 6-7.
Figure 6-7: A graph showing a historical load profile of a building.
The second piece of data is the load profile when using the ESS in its peak shaving mode. The
metering points for the load was collected at the B5PS2T3 switch gear according to the CT
locations defined in Figure 2-10. The load data collected is summarized and shown in Figure 6-8
below.

Figure 6-8: Peak shaving test data.
Both the historical load data and load data used in peak shaving mode are compared to each other
to quantify the peak shaving difference achieved.
The success criteria for this metric was originally determined to be that the ESS is able to store
energy during off peak time and discharge 250 kW during peak time to reduce peak load relative
to historical data over similar time period.
Based on the energy capacity available in the ESS installed at MCAS Miramar and that the
demonstration wanted to perform against SDG&Es winter TOU Peak time period it was
determined that the ESS should be set to discharge at 100kW power output to achieve 3 hours of
required discharge. During the test the battery started discharging at 5PM at 100kW and was able
to drop change the profile of the Grid purchases at the main feeder metering point to export 46 kW
of power into the distribution system as shown in Figure 6-7. At approximately 7PM, after two
hours of discharging the battery started approaching 30% state of charge and the total output power
of the battery started to diminish less than 100kW and slowly lessened until the battery was unable
to provide power out any longer just before 8PM. The end result showed that the battery was
capable of peak shaving at 100kW for just under the 3hrs but not long enough to get through the
whole Peak time period of 3hrs. The ESS would need to be set to a lower power discharge output
to get through the entire 3hrs.
The data was then compared to the base line data collected prior to conducting the peak shaving.
Figure 6-9 below shows the comparison to the two load profiles. The two load profiles show
similar load characteristics. The base load of the circuit operates between 30-50kW. As people
get to work in the morning there is an uptick in load on the circuit as lights are turned on an people
start their workday in the office. Sun rise this time of year is between 6:45AM-7AM and it is
shown that the load starts dropping as the PV systems start to generate power. The real differences
occur at 10AM when during the Peak Shaving test the ESS was set to charge which is why there
is a sudden ramp in load. At ~3:25PM the charge was stopped and the ESS dwelled at until 4:50

where the ESS was set to discharge at 100kW output. From here the delta between the two load
profiles is shown to be 105kW validating the 100kW capability of reducing demand during peak
time. At 7PM the battery started to reduce its power output as it neared the lower end of its State
of Charge causing the Grid load to rapidly increase until the battery was fully depleted at ~8PM.
If this system was metered by SDG&E the peak demand measured would have been 36kW right
before 8PM negating demand reduction achieved between 5PM-7PM. This shows that the output
power of the battery would need to be reduced in order to discharge for the full 3hrs.
Figure 6-9: Comparison of the load data collected on 12/12/15 to the data collected during the peak shaving
test on 1/12/16.
6.5 ESS ENERGY STORAGE CAPACITY
The data required for this Performance Objective is power output of the ESS and recorded time of
the power output. This was captured on two different days of performing this test. The first day
captured was on 11/15/2015 and the second was captured on 11/17/2015 and is shown in Figure
6-10 and Figure 6-11 below.

11/15/2015 Energy Capacity Test
300
200
100
-100
-200
-300
-400
12:00 PM
1:00 PM
2:00 PM
3:00 PM
5:00 AM
6:00 AM
7:00 AM
8:00 AM
9:00 AM
10:00 AM
11:00 AM
B6311 Load Battery Charge/Disch Grid Purchases/Export PV Generation
Figure 6-10: Energy Capacity test conducted on 11/15/15. The discharge output of the battery was set to 230
kW output power.
11/17/2015 Energy Capacity Test

250
200
150
100
50
0
-50
-100
-150
-200
-250
6:00:00 AM
7:00:00 AM
8:00:00 AM
9:00:00 AM
10:00:00 AM
11:00:00 AM
12:00:00 AM
12:00:00 PM
1:00:00 PM
2:00:00 PM
3:00:00 PM
4:00:00 PM
5:00:00 PM
6:00:00 PM
7:00:00 PM
8:00:00 PM
9:00:00 PM
10:00:00 PM
11:00:00 PM
B6311 Load ESS Charge/Disch Grid Purchases/Export PV Generation
Figure 6-11: Energy capacity test conducted on 11/17/2015. The discharge output of the battery was set to
190kW output power.
The measurement of power over time was analyzed and the energy capacity of the system was
calculated to be the integral of the graph from the beginning of discharge to the time that the power
output of the battery reaches zero. The data collected on 11/15/15 and 11/17/15 was integrated

over the time that the battery was discharging to determine the total discharge energy from the
ESS. The summary of the data is shown in Figure 6-12 and Figure 6-13 below. The ESS achieved
281kWh of energy capacity when discharged at 230kW power output and 294kWh when
discharged at 190kW power output.
Figure 6-12: Energy capacity calculated for 11/15/15 test.
Figure 6-13: Energy capacity calculated for 11/17/15 test.

7 COST ASSESSMENT
This project is intended to demonstrate that an energy storage system can be used as a replacement
for conventional diesel generators for emergency back-up power and show that an ESS can
function within a microgrid during islanded operation to enhance energy security. This project
also intends to show that an ESS can be used for economical benefits by changing the load profile
of a building by charging and discharging the battery according to a controlled schedule.
7.1 COST MODEL
The cost model was updated from what was calculated in the Demonstration Plan. The
Demonstration Plan that was submitted earlier in the project utilized the performance objectives
for peak shaving and islanding time to calculate the theoretical savings if those objectives were
realized. The cost model was updated based on the demonstrated performance of the installed
system.
Putting a cost assessment to the energy security aspect of this project is very difficult. NREL has
come up with using a Customer Damage Function (CDF) which tries to determine interruption
costs as a function of outage duration (Giraldez 2012). The CDF function for Miramar was
calculated to be $725/kW peak in a non-emergency situation for the islanding duration objective
of 72hrs. Since the system was only able to achieve a maximum theoretical islanding duration of
7hrs that will be the number used to calculate the CDF. This puts the CDF at $120/kW peak for a
non-emergency situation. Building 6311 had a maximum peak of 130kW in 2012. Therefore the
CDF of building 6311 yields $15,600 of cost associated with an outage of 7hrs. According to
SDG&E records over the last 10 years there are two spikes of outages recorded that impacted
customers in 2003 and 2011; therefore it will be assumed that over a 20 year period of operation
the ZnBr ESS installation will be used twice for back-up operations, and assumed to happen at
year 1 and year 10.
Since the probability of an outage occurring is a rare occurrence, the peak shaving mode of the
system is meant to provide economical benefit to the end user. This benefit will also be used to
calculate the operational cost reductions when using the system in addition to abating the CDF
associated with an outage.
The annual savings for operating in peak shaving mode were calculated using load data from
MCAS Miramar and SDG&E’s 2014 AL-TOU rate sheet for energy calculations (Figure 7-1). A
model was used that controls the energy storage unit to charge during off peak times and discharge
during peak times. SDG&E has different peak times for winter and summer operations so the

energy storage unit was controlled differently during the winter and summer.
Figure 7-1: SDGE Time of use schedule graphic.

The ESS was commanded to charge during off-peak hours and discharge during peak hours. The
model was run for a year’s profile. For each billing month the non-coincidental peak, the on-peak
peak, and the energy charges were calculated for the normal load curve and the grid purchases
curve when using the ESS in peak shaving mode. The result of the model showed there was a
$37k savings in demand charges and energy charges when using the ESS in peak shaving mode.
The cost elements associated with this assessment are shown in Table 7-1 below.
Table 7-1: Cost Model for an Energy or Water Technology
Cost Element Data Tracked During the Demonstration

Hardware capital costs ESS $840k, IPEM $41k

Installation costs Primus Power $37k, Dynalectric Construction $519k
Consumables No consumables used.

$10k/yr operational cost savings when used in peak
Facility operational costs
shaving mode
Maintenance ESS requires annual maintenance at $30k/yr
Hardware lifetime ESS cells are designed to last 20 years
Operator training $30k for operator training

Removal of equipment is $67k and the salvage value is
Salvage Value $471k using Single Present Value calculation from NIST
Handbook 135.
Customer Damage Function
$15.6k two times over 20 years
(CDF) Abatement
Total Lifecycle Costs (TLC) for the system assumes a 20 year life and includes the following:
TLC = [Hardware capital costs] + [Installation costs] + [Operator training] + [UPV
Maintenance Costs] - [UPV* Operational Cost Reductions] - [SPV Salvage Value] – [CDF
Abatement]
UPV Maintenance Costs
UPV Maintenance Costs are calculated using NIST Handbook 135.
𝑈𝑈𝑈𝑈𝑈𝑈 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 𝐴𝐴𝐴𝐴𝑈𝑈𝑈𝑈𝑈𝑈𝑁𝑁
Where A = $30k
UPVN = 14.88 taken from Table A-2 in NIST Handbook 135 Annual Supplement
UPV Maintenance Costs = $446k
UPV* Operational Cost Reductions
UPV* Operational Cost Reductions are calculated using NIST Handbook 135
𝑈𝑈𝑈𝑈𝑈𝑈 ∗ 𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂 𝐶𝐶𝑜𝑜𝑜𝑜𝑜𝑜 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 = 𝐴𝐴𝐴𝐴𝑈𝑈𝑈𝑈𝑈𝑈𝑁𝑁∗
Where A = $10k based on using SDG&E AL-TOU Primary rate sheet and peak shaving
performance demonstrated for 40kW of peak shaving in the summer and 100kW of peak shaving
in the winter.
UPV*N = 20 taken from Table A-3a in NIST Handbook 135 Annual Supplement using a
3% increase in price.
UPV* Operational Cost Reductions = $200k
SPV Salvage Value
SPV Salvage Value is calculated using NIST Handbook 135.

𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 = 𝐶𝐶𝐶𝐶𝑆𝑆𝑆𝑆𝑆𝑆𝑡𝑡
Where C = $840k
SPVt = 0.554 taken from Table A-1 in NIST Handbook 135 Annual Supplement
SPV Salvage Value = $465k
CDF Abatement
CDF abatement consists of two values, an abatement assumed at year 1 and an abatement assumed
at year 10. The abatement at year one is $15.6k based on the CDF function described earlier. The
abatement at year 10 is calculated using SPV in NIST Handbook 135 Annual Supplement.
𝐶𝐶𝐶𝐶𝐶𝐶 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴 = 𝐴𝐴 + 𝐴𝐴 𝑥𝑥 𝑆𝑆𝑆𝑆𝑆𝑆𝑡𝑡
Where A = $15.6k based on using SDG&E AL-TOU Primary rate sheet
CDF Abatement = $27.2k
Using the formulas above and date from Table 7-1 yields the following results for TLC.
TLC = [$881k] + [$556k] + [$30k] + [$446k] - [$200k] - [$465k] – [27.2k] = ~1,221k
The Total Lifecycle Cost for this system is $1,221k over a 20 year period and is shown in Table
7-2 for each year. The cost model indicates that with the current performance of the system the
cost savings due to operating the system do not generate a full payback within 20 years. If the
system achieved the original performance objectives the cost model is described in section 7.3.
Table 7-2: TLC cost for each year for a 20 year period.
Year UPV* SPV Salvage UPV CDF TLC
Operational Value Maintenance Abatement
Cost Costs
Reductions
1 ($10,065) ($815,640) $29,100 ($15,600) $654,795
2 ($20,130) ($792,120) $57,300 $696,450
3 ($30,195) ($768,600) $84,900 $737,505
4 ($40,260) ($745,920) $111,600 $776,820
5 ($50,325) ($724,920) $137,400 $813,555
6 ($60,390) ($703,080) $162,600 $850,530
7 ($70,455) ($682,920) $186,900 $884,925
8 ($80,520) ($662,760) $210,600 $918,720
9 ($90,585) ($643,440) $233,700 $951,075
10 ($100,650) ($624,960) $255,900 ($11,606) $970,084

11 ($110,715) ($606,480) $277,500 $1,000,099
12 ($120,780) ($588,840) $298,500 $1,028,674
13 ($130,845) ($572,040) $318,900 $1,055,809
14 ($140,910) ($555,240) $339,000 $1,082,644
15 ($150,975) ($539,280) $358,200 $1,107,739
16 ($161,040) ($523,320) $376,800 $1,132,234
17 ($171,105) ($508,200) $395,100 $1,155,589
18 ($181,170) ($493,080) $412,500 $1,178,044
19 ($191,235) ($478,800) $429,600 $1,199,359
20 ($201,300) ($465,360) $446,400 $1,219,534
7.2 COST DRIVERS
For this particular project since the energy storage technology was scaling up its system for the
first time there were cost drivers associated with building the first large prototype. Developing a
scalable low cost manufacturing process takes time and investment. Primus Power was able to
balance the uncertain costs of building a first of a kind unit with the unknown costs that are
normally associated with developmental technologies. Because of anticipated delays in
manufacturing and increased costs associated with developing their manufacturing line Primus had
to deliver a system that was fully functional and tested however was at reduced performance levels
due to the high costs of their Gen 1 prototype. Going through the experience of building their first
full scale system has allowed Primus to understand the behavior and performance of their system
at scale. This has been taken and applied to a Gen 2 version that is capable of meeting the
performance objectives of the original system at the anticipated original costs.
Other cost drivers for this type of technology implementations are the siting and infrastructure
upgrades required to accommodate new generation assets on an older distribution system. One of
the large costs on the installation of this project was the upgrades to the switchgear and the
transformer as well as creating a concrete pad for the ESS to sit properly.
7.3 COST ANALYSIS AND COMPARISON TO FULLY FUNCTIONAL SYSTEM
This section describes the cost analysis for a fully functional system that is capable of meeting the
performance goals (like implementing the Gen 2 of Primus’ system).
The cost elements associated with this assessment are shown in Table 7-3 below.
Table 7-3: Cost Model for an Energy or Water Technology
Cost Element Data Tracked During the Demonstration

Hardware capital costs ESS $840k, IPEM $41k

Installation costs Primus Power $37k, Dynalectric Construction $519k
Consumables No consumables used.

$37k/yr operational cost savings when used in peak
Facility operational costs
shaving mode
Maintenance ESS requires annual maintenance at $30k/yr
Hardware lifetime ESS cells are designed to last 20 years
Operator training $30k for operator training

Removal of equipment is $67k and the salvage value is
Salvage Value $471k using Single Present Value calculation from NIST
Handbook 135.
Customer Damage Function
$94k two times over 20 years
(CDF) Abatement
Total Lifecycle Costs (TLC) for the system assumes a 20 year life and includes the following:
TLC = [Hardware capital costs] + [Installation costs] + [Operator training] + [UPV
Maintenance Costs] - [UPV* Operational Cost Reductions] - [SPV Salvage Value] – [CDF
Abatement]
UPV Maintenance Costs
UPV Maintenance Costs are calculated using NIST Handbook 135.
𝑈𝑈𝑈𝑈𝑈𝑈 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 𝐴𝐴𝐴𝐴𝑈𝑈𝑈𝑈𝑈𝑈𝑁𝑁
Where A = $30k
UPVN = 14.88 taken from Table A-2 in NIST Handbook 135 Annual Supplement
UPV Maintenance Costs = $446k
UPV* Operational Cost Reductions
UPV* Operational Cost Reductions are calculated using NIST Handbook 135
𝑈𝑈𝑈𝑈𝑈𝑈 ∗ 𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂𝑂 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝐴𝐴𝐴𝐴𝑈𝑈𝑈𝑈𝑈𝑈𝑁𝑁∗
Where A = $37k based on using SDG&E AL-TOU Primary rate sheet
UPV*N = 20 taken from Table A-3a in NIST Handbook 135 Annual Supplement using a
3% increase in price.
UPV* Operational Cost Reductions = $740k
SPV Salvage Value
SPV Salvage Value is calculated using NIST Handbook 135.
𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉𝑉 = 𝐶𝐶𝐶𝐶𝑆𝑆𝑆𝑆𝑆𝑆𝑡𝑡

Where C = $840k
SPV Salvage Value = $465k
CDF Abatement
CDF abatement consists of two values, an abatement assumed at year 1 and an abatement assumed
at year 10. The abatement at year one is $94. The abatement at year 10 is calculated using SPV
in NIST Handbook 135 Annual Supplement.
𝐶𝐶𝐶𝐶𝐶𝐶 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴 = 𝐴𝐴 + 𝐴𝐴 𝑥𝑥 𝑆𝑆𝑆𝑆𝑆𝑆𝑡𝑡
Where A = $94k based on using SDG&E AL-TOU Primary rate sheet
SPVt = 10 taken from Table A-1 in NIST Handbook 135 Annual Supplement using a 3%
increase in price.
CDF Abatement = $164k
Using the formulas above and date from Table 7-1 yields the following results for TLC.
TLC = [$881k] + [$556k] + [$30k] + [$446k] - [$740k] - [$465k] – [164k] = $544k
The Total Lifecycle Cost for this system is $544k over a 20 year period and is shown in Table 7-2
for each year.
Table 7-4: TLC cost for each year for a 20 year period.
UPV* UPV
Operational Cost SPV Salvage Maintenance
Year Reductions Value Costs CDF Abatement TLC
1 ($37,000) ($815,640) $29,100 ($94,000) $549,460
2 ($74,000) ($792,120) $57,300 $564,180
3 ($111,000) ($768,600) $84,900 $578,300
4 ($148,000) ($745,920) $111,600 $590,680
5 ($185,000) ($724,920) $137,400 $600,480
6 ($222,000) ($703,080) $162,600 $610,520
7 ($259,000) ($682,920) $186,900 $617,980
8 ($296,000) ($662,760) $210,600 $624,840
9 ($333,000) ($643,440) $233,700 $630,260
10 ($370,000) ($624,960) $255,900 ($69,936) $564,004
11 ($407,000) ($606,480) $277,500 $567,084

12 ($444,000) ($588,840) $298,500 $568,724
13 ($481,000) ($572,040) $318,900 $568,924
14 ($518,000) ($555,240) $339,000 $568,824
15 ($555,000) ($539,280) $358,200 $566,984
16 ($592,000) ($523,320) $376,800 $564,544
17 ($629,000) ($508,200) $395,100 $560,964
18 ($666,000) ($493,080) $412,500 $556,484
19 ($703,000) ($478,800) $429,600 $550,864
20 ($740,000) ($465,360) $446,400 $544,104

8 IMPLEMENTATION ISSUES
This program spans from inception back in 2011 all the way to the end of 2015. There were
multiple challenges in implementing this program however each one was meant to implement final
demonstration in December of 2015. A few standout implementation issues will be noted in this
sections.
New Technology Development

Some of the challenges of achieving the desired islanding duration can be attributed to working
with technologies that are in their final development phases. As part of the experience with
working in the energy storage space it is shown that it is very difficult to scale systems up to utility
scale. Fielding technologies that have been demonstrated in relevant lab environments is always
a challenge and require iterations and lessons learned to optimize designs. This was realized early
in this project when the original energy storage company that was proposed was not able to build
the required unit due to challenges that arose in scaled units that were initially fielded. This
prompted a change in ESS supplier after the program was under contract. Once a new supplier
was selected the team had to manage a supplier that had promising technology however there
systems were lower on the TRL scale than the previous supplier and the team had to manage
technology that was in development and scale up simultaneously. This challenged the team in the
decision making process as the team was continuously balancing performance and cost of the
project appropriately to meet the demonstration objectives. This was shown near the end of the
project where energy capacity performance of the ESS was demonstrated be near our objectives
after our deployable system was ready but was not able to be implemented in our final
demonstration unit.
Interconnect Agreement
Due to the fact that this program spanned multiple years, the process of obtaining the interconnect
agreement from SDG&E took some understanding and effort. The use of large scale energy
storage systems in microgrid capacities is new to the utility industry for behind the meter
applications. Thus the interconnect agreement process is changing real time for utilities to adapt
to how these systems will be deployed. This project was subject to some of the real time changes
as a few iterations of the application were required due to changing application requirements.
Ultimately the IA and permit to operate was granted due to hard work amongst multiple parties
however it is still unclear if there is a well-defined process for getting IAs in place for microgrids.

9 References
Cullom, R. P. (2010, September 16). Sustainable Energy and National Security. Cambridge.
Department of Defense. (2011). Operational Energy Strategy. Washington, D.C.: Department of
Defense.
Giraldez, J. (2012). NREL's Approach to Microgrid Planning and Assessment. Military and
Commercial Microgrids Summit (pp. 24-31). San Diego: Infocast.

APPENDICES
APPENDIX A: POINTS OF CONTACT
POINT OF ORGANIZATION Phone

CONTACT Name Fax Role in Project
Name Address E-mail
310.647.9719 Principle
Ryan Faries Raytheon
[email protected] Investigator
858.577.6150
Mick Wasco MCAS Miramar Energy Manager
[email protected]
510.342.7602 CEO & Project
Tom Stepien Primus Power
[email protected] Manager
858.712.4746
Bob Riel Dynalectric Project Manager
[email protected]
303.384.7455
Bob Butt NREL Project Manager
[email protected]


ZN BR Battery Tech

Uploaded by

Copyright:

Available Formats

ZN BR Battery Tech

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ZN BR Battery Tech

Uploaded by

Copyright:

Available Formats

FINAL REPORT

Zinc Bromide Flow Battery Installation

ESTCP Project EW-201242

6. AUTHOR(S) 5d. PROJECT NUMBER

9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

13. SUPPLEMENTARY NOTES

15. SUBJECT TERMS

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

1.2 OBJECTIVE OF THE DEMONSTRATION

ZnBr Flow Battery Installation for

1.3 REGULATORY DRIVERS

Figure 1-1: Energy goals from the Secretary of the Navy.

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

Figure 2-3: Early rendering of ESS located at MCAS Miramar.

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

Figure 2-6: Photo of installed system at MCAS Miramar.

2.2 TECHNOLOGY DEVELOPMENT

ZnBr Flow Battery Installation for

Figure 2-7: Developmental timeline for Primus EnergyPod system.

ZnBr Flow Battery Installation for

Figure 2-9: Summary of results of Sandia 3rd party testing.

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

Summary of test results are shown in Figure 2-20 below.

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

Table 2-2: Summary of advantages of the Primus Zn/Br system.

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

Table 3-1: Summary of Performance Objectives

Performance Data Results

ZnBr Flow Battery Installation for

Qualitative Performance Objectives

3.1 ISLANDED DURATION

ZnBr Flow Battery Installation for

ZnBr Energy Storage System (ESS)

Energy Pod PowerBox

3.2 BUILDING LOAD REDUCTIONS

ZnBr Flow Battery Installation for

3.3 SWITCHOVER TIME

3.4 PEAK SHAVING

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for

ZnBr Flow Battery Installation for