Thesis PV Battery PSCAD
Thesis PV Battery PSCAD
A Thesis
presented to
the Faculty of California Polytechnic State University,
San Luis Obispo
In Partial Fulfillment
of the Requirements for the Degree
Master of Science in Electrical Engineering
by
Anthony W. Ma
May 2012
2012
Anthony W. Ma
ii
COMMITTEE MEMBERSHIP
AUTHOR: Anthony W. Ma
iii
ABSTRACT
installations. PV arrays are a unique source of power generation in that they are affected
by the location of the sun, shading, and temperature changes. These characteristics make
solar one of the most highly variable forms of renewable energy. In order to improve
primary focus of this paper is to determine if distributed energy storage systems can be
used to reduce the effect of solar intermittency. This paper examines the test data and
system are further studied using computer simulation modeling. This paper also shows
through computer simulations how a maximum power point tracker can increase a PV
arrays power output. The results of this thesis demonstrate that DESSs are capable of
smoothing out highly variable load profiles caused by intermittent solar power.
iv
ACKNOWLEDGMENTS
It brings me great pleasure to thank the people who helped to make this thesis
possible.
First, I would like to thank my teacher and advisor Dr. Taufik. Your passion and
enthusiasm for power engineering were the things that captured my interest in the
specialization early on. It was due to your leadership and guidance that I learned of this
project. Thank you for keeping my research interests in mind and thank you for all your
Next, I would like to thank Bill Torre, Mike Turner, and Steven Garrett at San
Diego Gas and Electric for proposing this thesis and for helping me with the data and
Finally, I would like to thank my parents, Brian and Lyleng Ma for their love and
support. This thesis is the culmination of many years of hard work and it would not have
been possible without their dedication. I am thankful for the sacrifices that you both
Disclaimer:
SDG&E does not endorse or accept the conclusions or results of this study.
v
TABLE OF CONTENTS
vi
6.1 Battery Management System .................................................................................... 71
6.2 Smart Inverter ............................................................................................................ 71
6.3 Single-Phase Inverter System Modeling .................................................................... 72
6.4 Harmonics and Power Quality Analysis ..................................................................... 73
6.5 Peak Shaving .............................................................................................................. 74
7 CONCLUSION...................................................................................................................... 76
BIBLIOGRAPHY ............................................................................................................................... 78
APPENDIX A: Grid-Connected PV Installations by State ................................................................ 80
APPENDIX B: Greensmith Power Vault 50 Datasheet .................................................................... 81
APPENDIX C: PowerGate Plus 50kW UL Datasheet ....................................................................... 83
APPENDIX D: International Battery IB-B-FHE-160 Datasheet ........................................................ 84
vii
LIST OF TABLES
viii
LIST OF FIGURES
ix
Figure 4-24: PSCAD Fixed Load Component .................................................................................. 45
Figure 4-25: Fixed Load Parameters .............................................................................................. 47
Figure 4-26: DC-Coupled Battery Backup System .......................................................................... 48
Figure 4-27: Complete Greensmith DESS PSCAD Model ............................................................... 49
Figure 4-28: Transformer Winding Voltages .................................................................................. 50
Figure 4-29: Three-Phase Voltage Source Signal Parameters........................................................ 51
Figure 5-1: Daily Solar Irradiance in San Diego on August 24th...................................................... 53
Figure 5-2: Nominal San Diego, CA Insolation on Aug. 24th from 10 A.M. 3 P.M. ...................... 55
Figure 5-3: MPPT Validation Simulation ........................................................................................ 56
Figure 5-4: PV Experimental and Simulation Power ...................................................................... 60
Figure 5-5: Power Measurement at System Equivalent ................................................................ 64
Figure 5-6: System Experimental and Simulation Data ................................................................. 66
Figure 5-7: Line-Line RMS Voltage ................................................................................................. 68
Figure 5-8: Real and Reactive Power at the Grid Interface ........................................................... 69
Figure 5-9: Long Simulation Run Example ..................................................................................... 70
Figure 6-1: Silent Power System at the Energy Innovations Center .............................................. 72
Figure 6-2: Inverter Output Current .............................................................................................. 73
Figure 6-3: Harmonic Spectrum of the Inverter Current (Fundamental 0.28 kA) ......................... 74
Figure 6-4: Harmonic Spectrum of the Voltage (Fundamental 7.43 kV) ....................................... 74
Figure 6-5: An Example of Peak Shaving ........................................................................................ 75
Figure B-1: Greensmith Power Vault 50 Datasheet Page 1 ........................................................... 81
Figure B-2: Greensmith Power Vault 50 Datasheet Page 2 ........................................................... 82
Figure C-1: PowerGate Plus 50kW UL Datasheet........................................................................... 83
Figure D-1: International Battery IB-B-FHE-160 Datasheet Page 1................................................ 84
Figure D-2: International Battery IB-B-FHE-160 Datasheet Page 2................................................ 85
x
1 INTRODUCTION
supplement a PV system. More specifically, this study examines if DESSs can smooth
out the effect of solar intermittency. Solar arrays are prone to intermittent periods of
generation due to shading from passing clouds. A battery system can be used to help
support the PV array during periods of lost PV power. PSCAD is used to model the
DESS, solar system, and electrical load according to the specifications and data that were
Due to the growing demand for renewable energy sources, the manufacturing of
solar cells and photovoltaic arrays has advanced considerably in recent years [1].
Starting from a small base, solar panel use has grown to a total global capacity of 40 GW
(40,000 MW) at the end of 2010. More than 100 countries use solar PV [2]. Solar
satellites. One of the most popular forms of PV installations is on the rooftops of homes
and buildings. One of the benefits of grid-connected solar electricity is that it can be used
US were approximately 7.2% [3]. Figure 1-1 shows the growth in grid-connected PV
installations from 1999 to 2008 [4]. Some of the other benefits of solar generation are the
environmental benefits and the purchase incentives for homeowners which are discussed
next.
1
Figure 1-1: Number of Annual Grid-Connected PV Installations (1999-2008)
generating energy and electricity today. Solar power systems do not generate air
pollution or emissions of any kind during operation. Currently the majority of electricity
generated today comes from burning coal. Recently there has been a lot of talk of clean
coal which is a term used to describe any technology that may reduce the emission of
greenhouse gases that develop as a byproduct of burning coal. However, clean coal
currently only exists as a concept and, even so, coal is still not a renewable form of
energy [5].
energy is controversial due to the fact that the reactors create dangerous nuclear waste.
Nuclear waste is accumulating at power plants across the world due to the lack of a long-
term storage solution [5]. With nuclear energy there is also the risk of a nuclear reactor
meltdown which would have devastating consequences on the people and the
2
environment surrounding the facility. The most famous examples of nuclear reactor
meltdowns are the Chernobyl disaster of 1986 and more recently with the Fukushima
nuclear power plants following the 2011 earthquake in Japan. Generating energy from
result in greenhouse gas emissions, but has a detrimental effect on the ecosystem around
Solar power generation is a proven technology that is ready for use now. As the
solar industry grows, PV systems are becoming increasingly more efficient and
affordable. Solar power does not create emissions or any other serious environmental
issues. Not only does residential solar generation have benefits for the environment, but
Under the California Solar Initiative (CSI) California home and business owners
can receive cash back for installing solar on their homes or businesses. These cash back
incentives combined with the reduction in the utility bill give property owners good
reason to install solar panels on their buildings. As a result, California leads the nation in
solar energy production. As of January 5th, 2012 there are currently 105,467 solar
projects in the state of California which combine to generate 1070 megawatts [6]. The
cost of installing solar photovoltaic systems is also becoming less expensive. Figure 1-2
shows the trend of average installed system costs over the life of the California Solar
Initiative [7]. It can be seen from the graph that the average cost to install a system is
3
Figure 1-2: Average Installed System Cost vs. Time
1.4 Capacity
All of these different factors have contributed to the recent proliferation of solar
installations and will certainly result in a greater number of PV installations in the future.
Data have shown that the capacity of solar power generation has steadily increased over
the past 10 years [4]. Figure 1-3 shows the growth in capacity of annual U.S.
photovoltaic installations from 1999 to 2008 in megawatts [4]. Figure 1-4 shows the
annual installed grid-connected PV capacity from 1999 to 2008 broken down by sectors
4
Figure 1-3: Capacity of Annual U.S. Photovoltaic Installations (1999-2008)
5
Figure 1-5: Non-Residential Grid-Connected PV Installations by Capacity (1999-2008)
installations from 1999 to 2008 while Figure 1-7 shows the average size of grid-
connected non-residential PV installations from 1999 to 2008 [4]. Both graphs show an
6
Figure 1-7: Average Size of Grid-Connected Non-Residential PV Installations (1999-2008)
Within the U.S., the state of California is a leader in the solar power movement.
Table 1-1 ranks the U.S. states by their 2008 solar capacity [4]. California is far and
away the leader of solar capacity within the U.S. with 62% of the market share in 2008.
Californias 178.7 MW of solar generation in 2008 was much more than second place
New Jerseys 22.5 MW. Also, California had one of the largest increases in capacity
from 91.8 MW to 178.7 MW (95% increase) from 2007 to 2008. Table 1-2 shows that
67% of all PV capacity installed in 2008 were in California [4]. Finally, Table 1-3 shows
that California is also the leader in per capita solar capacity [4]. The complete table of
one of the most progressive states in the U.S. and that shows with its strong adoption of
solar energy. The solar capacity in California is projected to continue increasing with
new PV installations in the future. On April 12th, 2011 Governor Jerry Brown signed
Californias Renewable Portfolio Standard (RPS) into law. The legislation will require
all of Californias utilities to source 33 percent of their overall electrical generation from
7
renewable sources by the year 2020 [8]. In order to reach this goal, utilities will have to
8
2 THE CHALLENGE
Solar arrays are a fairly simple source of direct energy generation. A basic
residential solar set up has the arrays mounted on the roof of the house. The power is
then regulated using a DC-DC converter and then converted to AC power using an
inverter. However, there are some unique challenges that exist with the use of solar
energy. As solar installations continue to increase, these challenges have the potential for
widespread impact.
Insolation and temperature are two of the main factors that determine solar generation.
changes in average temperature and insolation. Figure 2-1 shows the change in
temperature and the amount of daylight that is available throughout a year in San Diego,
CA [9]. Temperature peaks in the summer time and is the lowest in the winter time. The
maximum amount of daylight that is available is around June or July while the lowest
amount of daylight available during the year is around November or December. These
differences in temperature and daylight affect the solar cells power output and that can
present a challenge.
9
Figure 2-1: Average Temperature and Daylight in San Diego, CA
One of the biggest issues associated with solar panels is their high variability in
10
clouds, buildings, or trees it will see a drop in power generation. Figure 2-1 shows the
PV array power output of a 4.6 MW system in Springerville, Arizona for one day. This is
It can be seen from the figure that the PV output experiences periods of
intermittency throughout the day most likely due to shading. The challenge with
intermittent renewables is that they have unpredictable outputs that can cause system
instability and unreliability [10]. One of the ways to remedy this problem is to use a
There are a couple different methods used to boost the power output of a PV array
over the course of a day. One method is to use a solar tracking system. A solar tracker is
11
a system that orients the array panels towards the sun. As opposed to a fixed-mounted
solar panel array, a solar tracker system has its solar panels mounted on a motorized
system. The goal is to capture the most amount of direct beam from the sun and this is
achieved by making the sun as visible to the panel for as long as possible. Figure 2-3 is a
chart of the average daily insolation availability each month in San Diego, CA [11]. It
can be seen from the graph that the 1-Axis North South Tracking Array always has more
average daily insolation than its Fixed Array counterpart regardless of time of year or
latitude tilt. An even greater improvement is Two Axis Tracking which allows for West-
East tracking in addition to North-South tracking. The last row of Figure 2-3 shows that
Two Axis Tracking produces even more average daily insolation than the 1-Axis North
12
Another way of maximizing PV power is to make use of Maximum Power Point
solar cells at the knee of their I-V curves. The knee point is where the current and
voltage of the solar cell are both at their max. Increasing either current or voltage
integrated into the DC-DC converters control system. MPPTs are discussed further in
section 4.3.
13
3 SDG&E/EPRI ENERGY STORAGE SYSTEM
Installed at the SDG&E Skills Training Center is one of the first advanced energy
storage systems in California. The system is owned by the Electric Power Research
Institute (EPRI) and is on loan to SDG&E for testing. The system integrator is
Distributed Energy Storage Unit. Since this is one of the first systems of its kind to be
installed, every step of the process is being documented for posterity. Documentation
includes any lessons learned during the shipping, receiving, or installation process as well
The system is currently on an 18 month testing program where all its various modes of
operation will be exercised. Some of its operation modes include constant power
the engineers will also perform tests to make sure that the system complies with IEEE
14
Figure 3-1: DESS One-Line Diagram
The inverter in this system is based on a Satcon PowerGate Plus 50kW UL (PVS-
50-UL). It is a three-phase inverter rated at 50kW and 120/208 VAC. The system makes
use of a Maximum Power Point Tracking (MPPT) system to boost the PVs yield. It has
been modified by Satcon to offer full four quadrant power capability which allows it to
deliver both real and reactive power. This feature is not available in any off-the-shelf
Satcon inverters. The datasheet of the similarly sized Satcon PowerGate Plus 50kW UL
15
(PVS-50-UL) inverter is in Appendix C. The cabinet on the left in Figure 3-2 is the
Figure 3-2: Satcon Inverter (left) and Greensmith Energy Management System (Right)
On the right in Figure 3-2 is the cabinet of the Greensmith energy management
system. The cabinet houses the batteries as well as the batteries management system.
The Greensmith system allows for real-time battery module measurements of voltage,
current, temperature, cell capacity, State of Charge (SOC), efficiency, and State of Health
(SOH). It also has a system that will keep the cells balanced and it will protect the
batteries from under/over-voltage, temperature, and over current. The system also allows
16
the user to have control over the batteries operation including dynamic, real-time power
control.
Figure 3-3 is a picture of the inside of the Greensmith cabinet. The complete
battery module consists of 20 battery packs in series. The batteries being used are
Lithium Iron Phosphate (Li-Ion) and are manufactured by International Battery. Each of
the 20 battery packs is made up of eight IB-B-FHE-160 packaged cells. In total there are
160 rechargeable battery cells in series. Each of the IB-B-FHE-160 packaged cells has a
nominal voltage of 3.2 V and capacity of 160 Ah. The datasheet of the International
Battery IB-B-FHE-160 packaged cell is in Appendix D. When put in series, these cells
17
combine for a total battery system voltage of 512 V and capacity of 82 kWh. The
complete overall battery ratings are 50 kW, 82 kWh, and 512 VDC.
Since installation, SDG&E has collected data on the performance of the DESS
under various test cases. One of the modes of operation tested was the DESSs ability to
smooth intermittent PV generation. The SDG&E Skills Training Center has a small solar
setup of approximately 12 kVA, but it is not directly tied to the battery and is instead
availability of the solar array is very intermittent and it is often not functioning or turned
off. Instead of using data from the PV array, simulated PV data was entered into the
battery controller for testing purposes. The test took place on August, 24th, 2011 from the
hours of 10:00 A.M. to 3:00 P.M. Since the Skills Training Center is a commercial
business building, this time period is when the building is most active during the day. It
is also the time when the sun is the closest to the Earth and is therefore the time when the
most PV generation is occurring. Data is collected at every minute during that time
frame for the batterys SOC, voltage, temperature, instantaneous power, DC current, etc.
Some of the results of the test can be seen in Figure 3-4 below.
18
System Powers and Load vs. Time
100000 PV Power (W) Building Load (W) Battery (W)
80000
60000
40000
Power (W)
20000
0
8/24/2011 9:36
8/24/2011 10:48
8/24/2011 12:00
8/24/2011 13:12
8/24/2011 14:24
8/24/2011 15:36
-20000
-40000
-60000
Time
In Figure 3-4 it can be seen that the building load does not fluctuate much over
time. The building has an average load of about 66479 W during this time frame. The
PV output power on the other hand fluctuates greatly at certain points. The simulated PV
power data is based on a 50 kW array. The periods of power drop are most likely due to
shading from passing clouds. Otherwise the PV array produces an instantaneous power
of about 40 kW during this period. The last plot on Figure 3-4 is of the batterys power.
Positive values correspond to the battery charging and negative values correspond to the
battery discharging. When compared with the PVs power, it can be seen that the battery
closely follows the power output of the PV. More specifically, when the PV experiences
periods of intermittency, the battery discharges power to make up for the momentary loss
ready for new periods of intermittency. Figure 3-5 below is a plot of the combined PV
19
Combined PV and Battery Power
60000
50000
40000
Power (W)
30000
20000
10000
0
8/24/2011 9:36
8/24/2011 10:48
8/24/2011 12:00
8/24/2011 13:12
8/24/2011 14:24
8/24/2011 15:36
-10000
Time
Figure 3-5 shows how the battery is able to smooth out the variability in the solar
generation at every point. This result is also shown in Figure 3-5 below.
Battery Support
90000
80000 Net Load With Battery Load Without Battery
70000
60000
Power (W)
50000
40000
30000
20000
10000
0
8/24/2011 9:36
8/24/2011 10:48
8/24/2011 12:00
8/24/2011 13:12
8/24/2011 14:24
8/24/2011 15:36
Time
20
In Figure 3-5 the plot of Load without Battery is a plot of the building load after
PV power has been subtracted. Net Load with Battery is a plot of the building load
after the PV power has been subtracted and the battery power has been added. The Net
Load with Battery plot has a smoother profile than that of the Load without Battery
plot. The addition of the battery helps to smooth out the load profile.
21
4 PSCAD MODELING
The Greensmith DESS system parameters and test data from section three are
used to recreate the system using Power Systems Computer Aided Design (PSCAD
4.4.1.0). PSCAD is a power systems simulator that allows for the design and verification
of all types of power systems and power electronic controls. The PSCAD simulation
used for this report makes use of two previously created models.
4.2 PV Model
The PV model used in this report was developed by Dr. Athula Rajapakse of the
University of Manitoba, Winnipeg, MD, Canada and was provided by the PSCAD
technical support team. This model makes use of two custom library components. The
two custom components are the PV array and the Maximum Power Point Tracker
(MPPT). Figure 4-1 shows the PSCAD model of the two custom library components
[12]. The third component on the right is a thermal model for calculating the PV
temperature given solar radiation, wind velocity, ambient temperature, tilt angle of array,
surface emissivity, etc. Since the thermal time constants are much larger than electrical
time constants, in most EMT (Electromagnetic Transient) simulations use of constant cell
temperature should be sufficient [12]. Therefore, the thermal model is not used in this
report.
22
Figure 4-1: PV Array (left), MPPT Controller (middle), Thermal Model (right)
A solar cell can be modeled using an electrical equivalent circuit that contains a
current source anti-parallel with a diode, a shunt resistance, and a series resistance [13].
An example of solar cells electrical equivalent circuit model is shown in figure 4-2 [13].
The DC current that is generated when the cell is exposed to light varies linearly
with solar irradiance. Solar cells are characterized by their nonlinear I-V curve. Figure
23
The basic equation that characterizes the solar cell I-V relationship can be derived
considering the equivalent circuit shown in Figure 4-2. Kirchoffs current law of the
. (4-1)
After substitution of the equivalent diode current expression for Id and the shunt
branch current Ish, equation (4-1) becomes the following equation [12]:
(4-2)
.
Isc is the photo current and it is a function of the solar radiation on the plane of the
solar cell G and the cell temperature Tc. The photo current equation is as follows [12]:
(4-3)
IscR is the short circuit current at the reference solar radiation GR and the reference
cell temperature TcR. The parameter T is the temperature coefficient of photo current.
The current Io in equation (4-2) is called the dark current. It is a function of cell
(4-4)
IoR is the dark current at the reference temperature. The other parameters
appearing in (4-2), (4-3), and (4-4) are the electron charge q, the Boltzmann constant k,
the band-gap energy of the solar cell material eg, and the diode ideality factor n. The
24
composed of series and parallel connected modules and the single cell circuit can be
scaled up to represent any series/parallel combination [12]. Figure 4-4 shows the default
The Greensmith model uses simulated solar data based on a 50 kW array model.
Therefore, the PSCAD PV array parameters are set up to emulate a 50 kW model. Figure
25
Figure 4-5: 50 kW PV Array Parameters
Figure 4-3 shows that the optimum operating point of solar cells occurs at the
knee of the I-V curve. In order to extract the maximum amount of power from a PV
array, it is desirable to operate at the optimum operating point at all times. A Maximum
Power Point Tracker (MPPT) is a DC-DC converter that is placed between a PV array
and its load to ensure that the PV array operates at its optimum point despite varying
temperature, insolation, and load. The DC-DC converter is also necessary to regulate and
step-down the high voltage of the PV array. Figure 4-1 shows the MPPT custom library
component that was developed in PSCAD. There are a number of different MPPT
algorithms. A popular implementation is the Perturb and Observe (P & O) algorithm, but
it has limitations. The algorithm used in this model is based on the Incremental
26
Conductance (INC) method. The advantage that this algorithm has over the P & O
method is that it can stop and determine when the maximum power point is reached
without having to oscillate around this value. It can also perform MPPT under rapidly
varying irradiation conditions with higher accuracy than the P & O method [14].
However, a disadvantage of the INC method is that it can produce oscillations and can
computation time is longer than that of the P & O method due to the slowing down of the
sampling frequency resulting from the higher complexity of the algorithm [15]. The
PSCAD [12].
The controller used for the maximum power point tracking DC-DC converter is
27
Figure 4-7: DC-DC Converter Controller with MPPT
Figure 4-8 shows the MPPTs parameters. The parameters are the PV arrays ISC,
28
Figure 4-9 shows how the MPPT controller is able to track the optimum operating
point for varying solar radiation and temperature conditions [12]. The thick green line
indicates the variation of the PV array operating point during the variations of the solar
radiation and cell temperature. The MPPT controller accurately tracks the knee point
[12].
Figure 4-9: MPP Tracking Under Variable Solar Radiation and Temperature Conditions
The grid connected PV system is shown in Figure 4-10 [12]. The PV array is
connected to the input of a DC-DC converter. The DC-DC converter is a buck converter
that is controlled using the MPPT system. The output of the converter is the input to the
29
controller that is discussed further in section 5.2. Figures 4-11 and 4-12 show the
30
Figure 4-12: Firing Pulse Generation for Inverter
transformer [12]. Except for the transformer winding inductance and the smoothing
inductor, no additional harmonic filter is provided [12]. Modifications were made to the
system shown in Figure 4-10 in order to emulate the Greensmith system. The final
Energy Storage System in PSCAD. The battery model was provided by PSCADs
Technical Support Team and will be incorporated into PSCADs Master Library in the
future. There are many different types of batteries as well as many different factors that
31
affect their performance. There are four main cell chemistries in use for rechargeable
batteries: lead-acid, nickel-cadmium (Ni-Cd), nickel metal hydride (Ni-MH), and lithium-
characteristic. Hence for this system model, the charge and discharge characteristic can
be described by the same equation if the battery hysteresis effect is neglected. Figure 4-
14 shows the charge and discharge characteristic of a typical battery cell [16].
32
Figure 4-14: Charge Discharge Profile
2. Nominal voltage: The voltage of the battery under normal operating conditions
4. Charging rate (C rate): The amount of current that a battery can deliver for 1 hour
from fully charged to the end of life. For a 100 Ah battery, 1C means the
PSCAD based their battery model off a model described in [17]. The battery
component that was developed for PSCAD is based on Shepherds equation, which is
33
by a controlled voltage source in series with a constant resistance, as shown in Figure 4-
15 [16].
The equivalent circuit above is represented by the following equations which are
1 SOC (4-5)
Ebat E 0 K Q A e B(1SOC)Q
SOC
Where:
34
Rbat: internal resistance ()
1. The internal resistance is assumed constant during the charge and discharge
2. The models parameters are deduced from the discharge characteristics and
3. The capacity of the battery doesnt change with the amplitude of the current (i.e.
no Peukert Effect).
7. Charging and discharging history does not affect battery characteristics (i.e. no
hysteresis)
The main feature of this battery model is that the parameters can be easily
determined from a manufacturers discharge curve. Figure 4-16 shows the typical
35
Figure 4-16: Typical Discharge Curve
The three points that define the shape of the curve are the fully charged voltage,
end of exponential zone voltage, and the end of the nominal zone voltage [17]. The zone
occurring after the nominal zone is not generally useful and is therefore not covered here.
The internal voltage characteristic equation (4-5) is described by the sum of three
1. The exponential curve A e B(1SOC)Q represents the section from fully charged to
1 SOC
2. The nominal zone line K Q represents the middle section from the
SOC
3. The DC transition level of E0 is the value at the transition between the end of the
The discharge curve of Figure 4-16 is applied to equation (4-5) to show how a
discharge curve can be used to form an equation of the batterys internal voltage. The
36
A: Voltage drop during the exponential zone (V)
3 3
B 2.308( Ah ) 1 (4-8)
Qexp 1.3 A 1h
( ( )) ( )
( ( )) ( )
( ) (4-9)
For the fully charged voltage, the extracted charge is zero [18].
( ) (4-10)
( )
( ) (4-11)
37
This is a general approach to obtaining the model parameters that can be applied
to other battery types. The parameters of other common battery cells are presented in
The input and output signals of the battery model are the following:
Reset: The control signal used to control charge or discharge of the battery, input
signal
The dialog box for the battery model is shown below in Figure 4-18 [16].
38
Figure 4-18: Battery Models Available Configurations
There are five options for Battery Type. They are the following [16]:
User defined model: the characteristics of voltage vs. SOC and the internal
resistance vs. SOC are defined as the tabulated inputs directly. This allows for a
The other four options are electrochemical models based on the modified
o Lead Acid
o Nickel-Cadmium (Ni-Cd)
o Lithium-Ion (Li-Ion).
39
Nominal Voltage (V): The nominal voltage represents the end of the linear zone
Rated Capacity (Ah): The rated capacity is the rated capacity of the battery
Initial Capacity (Ah): It is used as an initial condition for the simulation and
Nominal capacity (Ah): It is extracted from the battery until the voltage drops
Voltage at exponential point (V): The voltage corresponds to the end of the
exponential zone
Figure 4-19 shows the PSCAD parameters of a 3.6V 1Ah Li-Ion battery [16].
40
In order to validate the battery model, a test case was developed in PSCAD. The
following test case was implemented in [16] and is used to demonstrate the charge and
discharge characteristics of the battery model. Figure 4-20 shows the circuit that was
used to test the charging and discharging characteristics of the battery model [16]. A
4.2V voltage source and a 3.6V, 1Ah Li-Ion battery are connected in parallel with
Figure 4-21 shows the Li-Ion batterys discharge curves with different C rates [16].
41
the discharge curves of Li-on
2.0C 1.0C 0.5C 0.2C
5.50
5.00
Battery terminal voltage (V)
4.50
4.00
3.50
3.00
2.50
2.00
0.0 2.0k 4.0k 6.0k 8.0k 10.0k 12.0k 14.0k 16.0k 18.0k
Time (sec)
Another test was conducted on this case to monitor the battery terminal voltage,
SOC, battery current, and load current through a charge and discharge cycle. The battery
controller is set up so that when the battery capacity is lower than 30% (SOC < 0.3) of
the rated capacity, the voltage source begins to charge the battery and supplies the load.
The voltage source stops charging the battery after the capacity reaches 80% (SOC =
0.8), and the battery takes over the load until its SOC < 0.3. After that, the charge-
discharge cycle repeats itself. The simulation results are shown below in Figure 4-22
[16].
42
Main : Graphs
battery terminal voltage (V)
4.40
4.20
4.00
y
3.80
3.60
3.40
SOC
1.00
0.90
0.80
0.70
0.60
y
0.50
0.40
0.30
0.20
Batttery current (A)
1.00
0.50
0.00
y
-0.50
-1.00
load current (A)
0.70
0.60
0.50
y
0.40
0.30
x 0.0 2.0k 4.0k 6.0k 8.0k 10.0k 12.0k 14.0k 16.0k 18.0k 20.0k ...
...
...
The battery starts off by supplying the load with 0.5C (0.5A when load resistance
= 7.2). At 5000 8600 sec, it is charged with 0.5C by the voltage source, and then
discharged at 8600 12200 sec [16]. These two tests illustrate the charge and discharge
43
The PSCAD battery model is used to model the Greensmith battery pack. In
order to accurately model the Greensmith battery, the parameters of a single IB-B-FHE-
160 battery cell were scaled up in order to be equivalent to the Greensmith battery. The
These parameters are used to define the Greensmith battery. The internal capacity
can be set as any value between 0 - 160Ah and represents the batterys SOC. Figure 4-23
shows these parameters entered into the PSCAD battery model specifications. These
44
Figure 4-23: Greensmith Battery Parameters
To model the load of the building, a three-phase fixed load was used. Figure 4-24
The fixed load component models the load characteristics as a function of voltage
45
( ) ( ) (4-12)
( ) ( ) (4-13)
Where:
V = Load voltage
In order to model a constant power load, set NP = NQ = KPF = KQF = 0. This will
(4-14)
(4-15)
46
( ) (4-16)
( ) (4-17)
For the purposes of this study, the loads were set up as constant power loads.
complete Greensmith DESS. The most common way of adding a battery to a PV system
47
battery-backup grid-connected PV system [18]. The DC-coupled battery is attached
simulation circuit. Figure 4-27 shows the complete PSCAD model of the Greensmith
system. The battery is DC-coupled between the DC-DC converter and the inverter. The
output of the inverter is attached to the equivalent feeder circuit. The three-phase fixed
load is attached at the output of the inverter which is on the secondary side of the step-
down transformer.
48
Figure 4-27: Complete Greensmith DESS PSCAD Model
Y-Y transformer. The secondary and primary line-line RMS voltages are 208V and
12.47kV, respectively. Figure 4-28 shows the transformers winding voltages in its
dialog box.
49
Figure 4-28: Transformer Winding Voltages
The primary side of the transformer is connected to the source of the system. The
systems source is a three-phase, 60Hz, AC voltage source with a line-line RMS voltage
50
Figure 4-29: Three-Phase Voltage Source Signal Parameters
In the next chapter tests are conducted on the PSCAD system to see if it
51
5 SYSTEM VALIDATION
In order to validate the PSCAD model, tests were performed on the system to see
if it could accurately represent the behavior of the Greensmith DESS. These tests include
One way to boost power generation from a PV array is to use a Maximum Power
Point Tracker (MPPT). In order for a MPPT to be useful it must be able to boost power
generation for any input conditions. In this test the model from [12] is tested on its own
without the battery or load attached. PVeducation.org has an applet that, based on the
latitude of a location and the day of the year, calculates the daily solar irradiance. The
latitude of San Diego, California is 32 N and date used was August 24th, corresponding
to the date of when the SDG&E data was collected. The generated daily solar irradiance
in San Diego on August 24th can be seen below in Figure 5-1 [19].
52
th
Figure 5-1: Daily Solar Irradiance in San Diego on August 24
graph. The data collected from SDG&E was only in the range of 10 A.M. to 3 P.M.
Therefore, to correspond with the data from SDG&E, the data points from Figure 5-1 in
the range of 10 A.M. to 3 P.M. were tabulated below in the first two columns of Table 5-
1. The insolation points in this range are plotted in the graph of Figure 5-2.
53
Table 5-1: MPPT Test Data
54
Nominal Insolation
1060
1040
Direct Radiaton (W/m^2)
1020
1000
980
960
940
9 10 11 12 13 14 15 16
Time (Hour)
th
Figure 5-2: Nominal San Diego, CA Insolation on Aug. 24 from 10 A.M. 3 P.M.
For every 15 minute point in the range, the corresponding insolation value was
entered into the control panel for the PV array. For this test the temperature is held
measure the instantaneous power from the PV array. The instantaneous power is
measured with both the MPPT on and the MPPT off at every step. The tabulated MPPT
on and off powers are shown in Table 5-1. The plots of these two powers are shown in
Figure 5-3.
55
MPPT Test
43000
42000 MPPT ON
MPPT OFF
41000
40000
Power (W)
39000
38000
37000
36000
35000
9 10 11 12 13 14 15 16
Time (Hour)
It can be seen in Figure 5-3 that at every time step the power generated with the
MPPT on is greater than the power generated with the MPPT off. This test validates that
the MPPT is able to boost the power output of the solar array.
In order to test the full system, the entire spectrums of data from SDG&E are
considered. In the 10 A.M. to 3 P.M. time frame, experimental data were collected at
every minute. For this simulation, the data at every 15 minute interval in that time frame
are considered. Table 5-2 shows some of the data that were collected from the
Greensmith system at every 15 minute interval. The data include the batterys SOC, cell
temperature, and power. It also has the buildings load as well as the amount of PV
generated power. The last two columns are called the Load without Battery and Net
Load with Battery and are formed from combinations of the other column values. Load
56
without Battery is the building load minus the PV power. Net Load with Battery is the
Recorded State of Cell Battery Building PV Power Load Without Net Load With
Time Charge Temp. Power(W) Load(W) Output(W) Battery (W) Battery (W)
8/24/2011
49 27.2125 35600 55970 37055 18915 54515
10:01
8/24/2011
58 28.80625 33100 54770 38225 16545 49645
10:15
8/24/2011
65 29.69375 17100 60580 38552 22028 39128
10:30
8/24/2011
68 30.075 6100 64370 39887 24483 30583
10:45
8/24/2011
68 30.425 1100 59790 41401 18389 19489
11:00
8/24/2011
67 30.7625 1100 62930 42396 20534 21634
11:15
8/24/2011
64 30.99375 2100 67810 43556 24254 26354
11:30
8/24/2011
63 31.1875 100 68850 41531 27319 27419
11:45
8/24/2011
61 31.35 100 72180 41290 30890 30990
12:00
8/24/2011
61 31.46875 600 73480 42130 31350 31950
12:15
8/24/2011
59 31.66875 -900 66910 40270 26640 25740
12:31
8/24/2011
57 31.81875 -12500 67480 24537 42943 30443
12:45
8/24/2011
51 32.19375 600 74860 42149 32711 33311
13:00
8/24/2011
46 32.49375 -22000 65170 17594 47576 25576
13:15
8/24/2011
45 32.425 4100 68940 39488 29452 33552
13:30
8/24/2011
44 32.575 -25800 67870 8318 59552 33752
13:45
8/24/2011
44 32.825 13600 68370 41510 26860 40460
14:00
8/24/2011
43 32.9875 100 63910 26974 36936 37036
14:15
8/24/2011
42 33.175 14100 61310 34991 26319 40419
14:30
8/24/2011
44 33.56875 -20300 64970 0 64970 44670
14:45
8/24/2011
38 33.825 -4400 60100 8408 51692 47292
14:59
The SDG&E data are used to define the specifications of the simulation model for
each data point. First, it is desirable to recreate the PV power as closely as possible. To
do this, the temperature of the 50 kW PV array model is set to the same temperature as
57
the Cell Temperature in Table 5-2. Next, the radiation level is adjusted until the array
produces a power output that is as close to the experimental value as possible. The
MPPT in the simulation is used at all times in order to replicate the actual model. Table
5-3 shows the input array temperature and solar radiation. The last two columns are the
power output.
8/24/2011
10:15 28.80625 748 38040 38225
8/24/2011
10:30 29.69375 796 38500 38552
8/24/2011
10:45 30.075 868 39860 39887
8/24/2011
11:00 30.425 904 41280 41401
8/24/2011
11:15 30.7625 904 42300 42396
8/24/2011
11:30 30.99375 892 43480 43556
8/24/2011
11:45 31.1875 844 41530 41531
8/24/2011
12:00 31.35 832 41580 41290
58
12:15
8/24/2011
12:31 31.66875 796 40290 40270
8/24/2011
12:45 31.81875 472 24400 24537
8/24/2011
13:00 32.19375 808 41970 42149
8/24/2011
13:15 32.49375 340 17810 17594
8/24/2011
13:30 32.425 760 39680 39488
8/24/2011
13:45 32.575 160 8366 8318
8/24/2011
14:00 32.825 796 41550 41510
8/24/2011
14:15 32.9875 520 27180 26974
8/24/2011
14:30 33.175 676 35290 34991
8/24/2011
14:45 33.56875 0 0 0
8/24/2011
14:59 33.825 160 8344 8408
Figure 5-4 is a graph of the experimental and simulation PV powers. The two PV
59
PV Experimental and Simuation Power
50000 Experimental
45000
40000
Simulation
35000
30000
Power (W)
25000
20000
15000
10000
5000
0
8/24/2011 9:36
8/24/2011 10:48
8/24/2011 12:00
8/24/2011 13:12
8/24/2011 14:24
8/24/2011 15:36
-5000
Time (Hour)
After the proper simulation PV power is achieved, the building load is the next to
be set up. The building load is modeled as a three-phase fixed load on the secondary of
the transformer. The fixed load is modeled as a constant power load. Since the three-
phase fixed load parameters are defined as the rated real power per phase, the
experimental building load is divided by three for this parameter. This is necessary to
represent the per-phase load. When all three phases of the load are combined they
represent the complete building load. Table 5-4 shows the complete building load and its
60
Table 5-4: Building Load Data
The final parameter that is varied for the simulation is the batterys state of
charge. In the PSCAD batterys parameters the state of charge is represented as the
initial capacity (Ah). Since the rated capacity of the battery is 160 Ah, the initial capacity
of the battery is the percentage SOC of the rated capacity. Table 5-5 shows the batterys
61
Table 5-5: Batterys Initial Capacity
Table 5-6 is a summary of all the input conditions used for the test simulation of
62
Table 5-6: Summary of Simulation Inputs
63
The instantaneous power flow is measured at the source of the equivalent system
using a multimeter. The power delivered from the source represents the loading of the
entire Greensmith system on the feeders circuit including the AC filter and transformer.
The real power and line-line RMS voltage of the multimeter were collected at
each point. The power measured is equivalent to the net load with the battery attached.
The power values are considered as the power delivered to the load side. The results are
64
Table 5-7: Simulation Results
Recorded Time Net Load with Battery (W) Line-Line VRMS (kV)
8/24/2011 10:01 20420 11.51
8/24/2011 10:15 22030 11.67
8/24/2011 10:30 24750 11.94
8/24/2011 10:45 27170 12.19
8/24/2011 11:00 27370 12.22
8/24/2011 11:15 26320 12.1
8/24/2011 11:30 24380 11.9
8/24/2011 11:45 23890 11.85
8/24/2011 12:00 23040 11.76
8/24/2011 12:15 23050 11.77
8/24/2011 12:31 22350 11.7
8/24/2011 12:45 21200 11.58
8/24/2011 13:00 20690 11.54
8/24/2011 13:15 18370 11.48
8/24/2011 13:30 20030 11.49
8/24/2011 13:45 16980 11.47
8/24/2011 14:00 19680 11.49
8/24/2011 14:15 17960 11.48
8/24/2011 14:30 18140 11.48
8/24/2011 14:45 16690 11.47
8/24/2011 14:59 17020 11.47
65
Experimental and Simulation Data
Load without Battery(Exp.) Net Load with Battery (Exp.)
Net Load with Battery (Sim.) Building Load (W)
80000
70000
60000
50000
Power (W)
40000
30000
20000
10000
0
8/24/2011 9:36
8/24/2011 10:48
8/24/2011 12:00
8/24/2011 13:12
8/24/2011 14:24
8/24/2011 15:36
Time
The Building Load is the load of the building on its own without any support
from the PV or battery. The Load without Battery (Exp.) is the building load minus the
experimental PV power. This graph would be almost exactly the same if the simulation
PV power were considered instead because the experimental and simulation PV outputs
are nearly the same. The large swings in this plots graph are due to the variability in the
PV power. The Net Load with Battery (Exp.) is the Load with Battery (Exp.) graph
with the experimental battery power added. This graph represents the net load that the
rest of the feeder circuit sees. By comparing the Net Load with Battery (Exp.) against
the Load without Battery (Exp.), it can be seen that the addition of the battery helps to
66
smooth out the load profile considerably. The large swings that were present in the Load
without Battery (Exp.) plot are reduced in the Net Load with Battery (Exp.) plot. The
last plot on the graph is of the simulation of the Net Load with Battery (Sim.). The Net
Load with Battery (Sim.) plot shows that the Greensmith simulation model achieves the
goal of smoothing out the variability in the PV power while also lowering the building
load. The reason that the experimental and simulation Net Load with Battery plots do
not exactly match is because the experimental system makes use of a battery management
system that regulates the battery output based upon the instantaneous PV power. The
simulation model of the system makes use of a battery controller that does not regulate
the battery output based upon the PV power. Instead, the battery controller monitors the
Even without the battery management system, the simulation is still able to reduce the
The voltage of the system is examined next. Table 5-7 has the simulated line-line
RMS voltages measured at the multimeter. Figure 5-7 is a graph of these voltages over
time.
67
Line-Line RMS Voltage
14
12
10
Voltage (kV)
0
8/24/2011 9:36
8/24/2011 10:48
8/24/2011 12:00
8/24/2011 13:12
8/24/2011 14:24
8/24/2011 15:36
Time
The simulation is able to maintain a voltage that is close to the rated voltage of
12.47 kV throughout. The variation in the voltage can be attributed to the inverter which
makes use of a simple inverter controller. The inverter is set to operate at unity power
factor, but during the transients, the reactive power output of the inverter changes [12].
Figure 5-8 is an example of how the PCC responds to transients created by the solar
variation [12]. The real power output changes proportional to the changes in the solar
radiation level [12]. The reactive power changes when the real power changes, but settles
back to zero in order to maintain a unity power factor. The PCCs nominal voltage is 11
kV [12].
68
Figure 5-8: Real and Reactive Power at the Grid Interface
charging and discharging of the battery, which can vary in its amount of power delivered.
Figure 4-10 showed that when a battery discharges, its voltage decreases. The opposite is
true for a charge cycle. Therefore, a change in battery voltage corresponds to a change in
the amount of battery power delivered or absorbed. Since the power is always changing,
the inverter is not able to settle to a unity power factor. The continually varying battery
power is why the Point of Common Coupling (PCC) sees a slight variation in voltage.
69
The simulations were also run for five second transient periods each time. If the
simulations were run for longer, the inverter controller would adjust the voltage to be
closer to its rated value. Figure 5-9 is an example of a simulation run that has a 30
second transient period. The secondary of the transformer has a line-line RMS voltage of
208.8 V while the primary of the transformer has a line-line RMS voltage of 12.24 kV.
Both of these voltages are close to their rated voltages of 208 V and 12.47 kV,
respectively. The disadvantage of a longer transient period is that it takes much longer to
simulate.
Since the fixed load is set up as a constant power load, variations in voltage do
70
6 FUTURE WORK
6.1 Battery Management System
In order for the PSCAD model to more accurately represent the behavior of the
System (BMS). The Greensmith BMS monitors the instantaneous PV power generation
and uses that information to make decisions on how to make use of the battery. If the PV
is supplying ample power then the BMS either charges the battery or has it held on
standby. However, if the PV power drops then the BMS discharges energy from the
battery to compensate. The current PSCAD battery model is controlled by the battery
control component. The battery control component controls the charging and discharging
of the battery based upon its SOC. A Battery Management System that monitors the PV
power would help make the PSCAD model behave more like the Greensmith DESS
The inverter controller used in this study is a simple P and Q regulator [12]. The
inverter is set to operate at unity power factor, but during the transients, the reactive
power output of the inverter changes. Fluctuations in reactive power can be minimized
by improving the inverter control, for example by using decoupled controls based in d-q
currents [12]. A better inverter design would be one based off a smart inverter. Some
important system monitoring and grid-interactive features for an inverter to have are over
71
6.3 Single-Phase Inverter System Modeling
This study made use of a three-phase inverter because the Greensmith DESS also
commercial size building. It is not uncommon for commercial buildings to require three-
phase power in order to run their machinery. However, in residential applications it is far
and small businesses only need single-phase power. For future studies the same tests that
were conducted in this study could be applied to a system that makes use of a single-
phase inverter. SDG&E has a second DESS system at their Energy Innovations Center
(EIC) that makes use of a single-phase inverter. The system is Silent Power Inc.s
output with 10 kWh of energy storage and makes use of a lead acid battery pack. A
picture of the Silent Power system is shown in Figure 6-1. The Silent Power system is
enclosed in the wooden structure and the EIC is the building in the background.
72
6.4 Harmonics and Power Quality Analysis
One of the biggest concerns with grid connected PV inverters is the possibility of
harmonic injection on to the grid. IEC standard 61727 Photovoltaic (PV) systems
limits at the Point of Common Coupling (PCC) [12]. At the PCC the current waveforms
were quite visibly distorted [20]. The inverter current waveforms can be seen in Figure
6-2 [20].
Since the current waveforms are distorted they lead to corresponding distorted voltage
waveforms. Figures 6-3 and 6-4 show the corresponding harmonic spectrums of the
73
Figure 6-3: Harmonic Spectrum of the Inverter Current (Fundamental 0.28 kA)
If the harmonic content of the system does not comply with the local utility then it
would be desirable to install a filter at the PCC of the PV system [20]. Harmonics are
also be used for peak shaving. Solar output peaks at around 12 P.M. with the rotation of
the Earth. However, residential load demand peaks at around 5-6 P.M. when most people
are returning home from work. Therefore the period of 5-6 P.M. is considered to be peak
74
residential load time. In most households nobody is home at noon and therefore there is
more PV power generated than the local load demands. Typically when a PV array
produces excess energy it sells the power back to the grid. However, with the use of
energy storage devices the excess energy could be stored in batteries to be later used that
evening. Figure 6-5 demonstrates how this works over the course of a full day [21].
The use of energy storage devices for peak shaving is something that can be
explored in future studies. Peak shaving has the potential to save the customer money by
75
7 CONCLUSION
supplies towards renewables. As one of the safest and most easily accessible forms of
history. With the increase in PV systems comes the need to examine some of the unique
issues that they present. One of these issues is the problem of solar intermittency. PV
arrays are subject to shading and temperature changes that make their power generation
difficult to predict. One possible solution is to use a backup battery to supplement the PV
array. The Electric Power Research Institute in conjunction with San Diego Gas and
Electric have been researching the feasibility of using Distributed Energy Storage
Systems to smooth out the power invariability from PV arrays. Test data collected on
their Greensmith DESS show that the battery system does a good job of smoothing out
the variability in the PV generation. The battery makes up the power to the load when
distributed energy storage system. More specifically, this study aimed to see if a DESS
Greensmith and installed at San Diego Gas & Electrics Skill Training Center was the
center point of this research. Experimental data from the system was examined along
with computer-simulated data of the equivalent system model. The experimental data
showed that the DESS is able to effectively compensate momentary drops in solar power.
The Greensmith DESS was recreated in PSCAD and simulation testing was conducted on
this model. Testing on the PSCAD model confirmed both the functionality of the
76
Maximum Power Point Tracker as well as the DESSs ability to smooth out solar
invariability. In both the real world experiment and the simulation, the use of a DESS
helped to smooth out the load profile of the PV system. The DESS does an effective job
future when solar energy becomes a substantial percentage of the electrical supply it will
be necessary to have DESSs more prevalent in Californias vast electrical utility system.
77
BIBLIOGRAPHY
[2] REN21, "Renewables 2011: Global Status Report," REN21, Paris, 2011.
[3] U.S. Department of Energy, "U.S. Climate Change Technology Program: Technology Options for the Near and
Long Term," U.S. Department of Energy, Washington D.C., 2003.
[4] L. Sherwood, "U.S. Solar Market Trends 2008," Interstate Renewable Energy Council, Latham, 2008.
[6] "Go Solar California - Homepage," California Energy Commission & California Public Utilities Commission,
[Online]. Available: http://www.gosolarcalifornia.org/. [Accessed 5 January 2012].
[7] "Welcome to California Solar Statistics," California Energy Commission & California Public Utilities
Commission, [Online]. Available: http://www.californiasolarstatistics.ca.gov/. [Accessed 5 January 2012].
[8] C. Meehan, "Clean Energy Authority," Clean Energy Authority, 14 April 2011. [Online]. Available:
http://www.cleanenergyauthority.com/solar-energy-news/california-makes-rps-into-law-041411/. [Accessed 4
May 2012].
[9] "San Diego, CA, California, USA: Climate, Global Warming, and Daylight Charts and Data," Climate-Charts,
[Online]. Available: http://www.climate-charts.com/Locations/u/US72290000477401.php. [Accessed 4 May
2012].
[10] "The Need for Electricity Storage," MegaWatt Storage Farms, Inc., [Online]. Available:
http://www.megawattsf.com/gridstorage/gridstorage.htm. [Accessed 4 May 2012].
[11] Sandia National Laboratories, Stand-Alone Photovoltaic Systems: A Handbook of Recommended Design
Practices, Springfield: National Technical Information Service, 1995.
[12] A. D. Rajapakse and D. Muthumuni, "Simulation Tools for Photovoltaic System Grid Integration Studies," IEEE
Electrical Power and Energy Conference, EPEC, 2009.
[13] A. D. Rajapakse, "Simulation of Grid Connected Photovoltaic Systems," PULSE, vol. October, no. 2008, pp. 6-7,
2008.
[14] Y.-J. M. Tung, A. P. Hu and N.-K. Nair, "Evaluation of Micro Controller Based Maximum Power Point Tracking
Methods Using dSPACE Platform," Australian University Power Engineering Conference, Auckland, 2006.
[15] Z. Yu and K. Ogboenyira, "Renewable Energy Through Micro-Inverters," Power Electronics Technology, 1 April
2009. [Online]. Available:
http://powerelectronics.com/power_semiconductors/power_microinverters_computercontrolled_improve_0409/.
[Accessed 4 May 2012].
[16] S. Jiang, "Battery Component in PSCAD/EMTDC," Manitoba HVDC Research Centre, Winnipeg, 2012.
78
[17] O. Tremblay, L.-A. Dessaint and A.-I. Dekkiche, "A Generic Battery Model for the Dynamic Simulation of
Hybrid Electric Vehicles," IEEE, 2007.
[18] O. Tremblay and L.-A. Dessaint, "Experimental Validation of a Battery Dynamic Model for EV Applications,"
World Electric Vehicle Journal, vol. 3, pp. 1-10, 2009.
[19] R. A. Messenger and J. Ventre, in Photovoltaic Systems Engineering, Boca Raton, CRC Press, 2010, p. 216.
[21] A. Rajapakse, D. Muthumuni and N. Perera, "Grid Integration of Renewable Energy Systems," InTech, 2009.
[22] L. Stoddard, J. Abiecunas and R. O'Connell, "Economic, Energy, and Environmental Benefits of Concentrating
Solar Power in California," National Renewable Energy Laboratory, Golden, 2006.
[23] C. M. Shepherd, "Design of Primary and Secondary Cells - Part 2. An Equation Describing Battery Discharge,"
Journal of Electrochemical Society, vol. 112, pp. 657-664, 1965.
79
APPENDIX A: Grid-Connected PV Installations by State
Table A-1: Grid-Connected PV Installations by State
80
APPENDIX B: Greensmith Power Vault 50 Datasheet
81
Figure B-2: Greensmith Power Vault 50 Datasheet Page 2
82
APPENDIX C: PowerGate Plus 50kW UL Datasheet
83
APPENDIX D: International Battery IB-B-FHE-160 Datasheet
84
Figure D-2: International Battery IB-B-FHE-160 Datasheet Page 2
85