HomeMy WebLinkAboutJournal of Power Soruces Kotzebue PP modelENERGY EFFICIENCY EVALUATIONS
Model of the Energy and
Economic Value of a
Flow Battery in a Wind-
Diesel-Battery Hybrid
System
Based on the Premium Power Battery
Demonstration at Kotzebue Electric Association
Dennis Witmer, Dennis Meiners, Brad Reeve, Matt Bergin, Jessie Logan
8/18/2012
A MatLab/Simulink model has been developed to evaluate the possible impact of a large flow battery on
a wind-diesel-battery hybrid system. Based on this model, adding the battery to the system offers
significant operational and economic benefits.
Introduction
In Alaska, many remote communities have no connection to an electrical grid outside their communities
(building and maintaining transmission lines over the required distances to serve a small customer base
is not economically viable), and no roads from outside the community (most depend on summer barges
for delivery of bulk goods, and air transport for perishables). Some of these communities are served by
a small scale local hydro project, but most remain dependent on diesel electric generators.
While some advocates for renewable energy seem fond of disparaging the Diesel Electric Generator
(DEG), this form of power generation has proven to be well suited for Alaskan applications. Diesel
generators are efficient, reliable, have low capital and installation costs, and operate on a fuel that can
be transported and stored in quantities sufficient to support a community through long winters when
barges cannot travel. However, recent increases in the cost of crude oil have resulted in dramatically
higher costs of power generation in these communities. A typical small scale generator typically
generates 14 kilowatt-hours of electricity per gallon of fuel burned. When diesel fuel prices are at $1
per gallon, this translates to a fuel cost of just over $0.07 per kilowatt hours. As fuel prices rise, one can
estimate the fuel cost as the cost per gallon times $0.07—so a $5.00 per gallon fuel cost (typical
delivered cost to small villages) would represent a $0.35 per kilowatt-hour fuel cost. Total delivered
cost to a customer may be double this cost. These high energy costs are difficult for residents of these
communities, and form a drag on the local economy.[1]
Many of these remote communities have excellent wind resources, so using wind power to generate
electricity seems to be an obvious solution. In current large wind farms in the lower 48, electricity can
be generated for about $0.06 per kilowatt hours, far cheaper than energy from a diesel electric
generator. Wind turbines have been erected in approximately 16 communities, with more in the
planning stages. Typical large installed wind systems cost about $1400 per kW. These wind farms are
growing large enough that they are affecting grid power quality, leading to an interest in grid sized
energy storage, allowing frequency control, grid stabilization, and higher penetration.. Various models
have been developed to assess the economic benefits of energy storage in grid connected environments
such as the “ES-Select” tool developed byDMV-KEMA. [2]
Unfortunately, installed wind system costs in Alaska have proved to be far higher than in the lower 48.
These high costs are due to the need for foundations designed for bad soils (especially permafrost), and
the cost of use of the cranes needed to erect the turbines. The best construction season is late winter,
when the ground is hard frozen and vegetation protected by a blanket of snow, but heavy equipment
must be brought to the village during the summer barge season, meaning a crane must spend the better
part of a year on location. Recent costs for installing turbines have ranged from $5,600 per kW for
larger turbines installed in Kotzebue to more than $12,000 per installed kW for systems in smaller
communities.[3]
In grid connected systems, wind power typically represents less than 10% of the total electrical load, and
does not result in disruption of the normal operation of the grid (although this is changing, and is one of
the driving forces behind the search for energy storage on the grid). Initial wind systems installed in
Alaska, such as the wind farm at Kotzebue, were low penetration systems—1.17 MW of wind to support
a load of approximately 3 MW—but fuel savings in these systems are modest—a capacity factor of 30%
means that the estimated fuel savings reaches only about 12%. Given the high cost of diesel fuel,
achieving greater savings is a desirable goal.
Since most small communities have small electrical loads (0.1 to 3.0 Megawatts) compared to the size
of commercial wind turbines (1 Megawatt and larger), installing medium penetration (installed wind
capacity of 50-150% of average load) and high penetration (installed wind capacity of 200-500% of
average load) is both possible and attractive[4, 5]. Installing larger wind turbines could reduce the cost
per installed kW, provide additional diesel fuel savings for electrical generation, and potentially provide
energy for residential heating, providing additional savings for the community. However, a wind-diesel
hybrid system has some drawbacks, including the need to keep the diesel engines on at all time to
assure power remains on (especially during high wind events when turbines trip offline to prevent
damage), and the inability to provide any fuel savings during times when the wind is not blowing.
Adding energy storage to a Wind-Diesel hybrid system provides several obvious benefits. First, since the
energy storage system can store energy and deliver it at a later time, the diesel engines can be turned
off, since the load can be met with stored energy (peak shaving). If the battery is sufficiently large, the
wind energy can be stored for use after the wind event is over (load shifting), saving additional fuel. The
Wales demonstration project has demonstrated the positive effects of a wind-diesel-battery hybrid
system. [6]
Unfortunately, energy storage in conventional battery systems has proven to be quite expensive.
Conventional lead acid batteries have limited cycle lifetime and depth of discharge, making energy
storage with this technology uneconomical (although new lead acid batteries may change this).
Electrochemical losses also occur in real batteries, both in charging and discharging, and additional
energy is lost in the necessary inverters.
Several vendors have developed new battery technologies that claim to address the issues of energy
storage, and may prove useful for Wind-Diesel-Battery hybrid systems. Ultimately, the utility of these
systems must be verified in field demonstrations. However, some questions can be addressed through
development of models to simulate the possible operation of these systems. These questions include:
How does adding energy storage to remote community affect the cost of energy?
How much additional electricity can be delivered to a load by adding a battery to the system?
How much additional fuel can be saved by allowing diesels-off operation during wind events,
and though load shifting, that is, keeping the diesels off for additional time after the wind dies?
How much is a energy storage system worth?
Can a wind-diesel-battery hybrid system provide more cost effective energy than a straight
diesel system?
Modeling is one tool available to estimate answers to these questions. Some models are available,
including HOMER [7], a widely recognized tool to address these questions. [8-10], and Hybrid 2 [11] ,
both publically available. However, these models have some drawbacks, including the need for long
term data sets, the use of simplified assumptions for diesel generators (both use linear approximations
for fuel consumption vs. load) and batteries. Economic evaluations are based on annual fuel savings,
meaning that obtaining field data to verify the veracity of the model is data sparse (one data point per
system per year).
Other modeling efforts have focused on the electrical transients that occur in these Wind-Diesel hybrid
systems. Lone [12] specifically models the electrical performance of a system including a flow battery.
Takashi [13, 14] focuses on the use of flywheels in these systems. Sandhu [15] focuses on control
system strategies for optimized dispatch. None of these sources mention economic considerations.
Tay [16] reports on a Matlab model of a wind diesel hybrid system that includes an economic analysis of
a wind-diesel-battery hybrid system, but limits economic calculations to seven one day periods.
Based on the information found in the literature a model to evaluate the energy and economic
performance of a high penetration wind-diesel-flow battery configuration does not appear to be readily
available for public use. For this reason, a new Matlab/Simulink model was developed to analyze this
configuration.
Description of the Premium Power Battery and the demonstration at
Kotzebue Electric Association.
Kotzebue Electric Association (KEA) is a member owned cooperative utility, providing electricity to the
community of Kotzebue, Alaska, which is located 33 miles north of the Arctic Circle along the west coast
of Alaska, with approximately 3500 residents. The utility generates most of its power with diesel
generators, with an average peak load of 3 megawatts. KEA was the site of the first wind farm installed
in Alaska, with the first turbines installed in 1997, with additional wind turbines installed over the years.
In 2008, KEA proposed expanding the wind installation in Kotzebue from 1.17 MW to 2.9 MW by adding
two 900 kW turbines to the system, moving the system from a low penetration system to a medium
penetration system. At the same time, KEA proposed adding flow battery to allow more stable
operation of the system. Several vendors were approached, but eventually Premium Power of North
Reading, MA was selected as the battery supplier, and a purchase order for a Transflow 2000 unit was
negotiated. Specifications for this battery indicated that the maximum charge rate is 500 kW, the
maximum discharge rate is 500 kW, and the battery is capable of storing 3.5 MW-hrs of electricity. This
battery was to be installed in Kotzebue as a demonstration project, in to verify the performance of the
battery.
Given that the Kotzebue load is about 3MW, the .5MW output of the Premium Power battery is
insufficient to allow a diesel-off operation of the utility, so a larger storage system would actually be
preferable. Ideally, an energy storage system should be large enough to cover the total load for at least
a few minutes to allow starting a diesel engine, thus allowing safe operation of the utility in a diesel-off
mode. (Worth noting is the fact that in high wind events, wind turbines are designed to trip off to
protect themselves from damage, so the energy storage system must be able to handle a wind turbine
going from maximum power to zero in an instant.) However, it is possible to use the information
collected from Kotzebue to model a high penetration wind system for a smaller community, and to
assess the benefits of such a battery system in a high penetration wind system.
KEA has installed a Supervisory Control and Data Acquisition (SCADA) system that collects data at a one
second time rate. Since the battery performance depends on the current density which is proportional
to charge rate, using one second data during variable wind events allows a more precise estimate to be
made of energy stored and discharged by the battery.
For this model, one second load and wind farm output data collected during a wind event on February
19-20, 2008 was used. Approximately 300 wind data points were missing and recorded as zero wind
output—these values were replaced by the wind output value immediately preceding these records. A
data file of 85,000 seconds was used.
Figure 1 Data from Kotzebue SCADA system showing wind event
The cleaned load and wind output data used in the model is shown in figure 1. Note that at no point
does the maximum wind approach the load, indicating that the wind is a low penetration installation.
This, of course, also means that the wind is only displacing a small fraction of the total diesel fuel
consumed.
In order to model a high penetration system, the collected and cleaned data was modified to simulate
operation in a smaller community with a larger wind system. This was achieved by dividing the
Kotzebue load data by a factor of 5, and multiplying the wind data by a factor of 4 (to simulate the
addition of the two larger turbines, and a community with a stronger wind resource).
Figure 2 High penetration wind event simulated from Kotzebue data
Case 1. Diesel only operation.
The baseline case is a diesels only operation. While it is clear that this will result in the highest fuel
consumption, the installed capital and operations and maintenance (O&M) cost also affects the total
cost of produced electricity. For this model, diesel O&M is estimated at $0.02 per kilowatt-hour, and
diesel capital costs are calculated based on a $1000 per kW installed cost, a 5 year lifetime, and a ROI of
6% (utility rate).
Based on the size of the load, a Cat 800 kilowatt diesel engine was selected for this model. As can be
seen from Figure 3, the Cat 800 diesel generator is more efficient at high power levels, and much less
efficient at lower power, typical of most modern diesel engines. This means that during wind events,
the displacement of the diesel load by the available wind will push the diesel generator to a less efficient
part of the operating curve. The engine is sufficiently large to cover the entire load, providing spinning
reserves for the utility. In the diesel only mode, the engine will operate at an efficiency of about 34%.
Fuel consumption can be calculated from any power level, based on a gallon of diesel fuel having 39 kw-
hr of energy—so the fuel consumption of this engine is approximately 34% x 39 kW-hrs/gal = 13.26 kW-
hrs/gal. Using an estimate of $5 per gallon, this can be used to calculated fuel costs for the diesel engine
operation.
Figure 3 Efficiency curve for a Cat 800 Diesel Electric Generator
Figure 4 Diesel only operation
Figure 4 shows the output from the model showing the total energy output from the diesel engine as it
meets the village load. Costs for fuel (estimated at $5 per gallon, typical of fuel delivered to remote
villages), O&M and capital are summed for each one second cycle of the model, providing a total cost
for generating the power, which can then be converted to a cost per kW-hr of delivered electricity for
the system. Using the load data shown in Figure 2,a total of 12,547 kW-hrs of electricity are required to
meet the load. This requires a total of 925 gallons of diesel fuel, and results in Cost of Electricity (COE)
of about $0.43 per kW-hour, with most of this cost associated with the cost fuel.
Case 2. High Penetration Wind-Diesel Hybrid System Without Electrical Storage
If a renewable energy source is added to the system, this will result in a decrease in the diesel fuel
required to provide the electrical power, and reduce fuel costs. However, adding the renewable energy
is not free, as capital is required for the purchase and installation of the hardware, and additional costs
are incurred in operations and maintenance. For this model, a capital cost per second has been
calculated for each component added to the system (wind turbines and batter) based on total installed
capital cost, expected lifetime, and a utility ROI rate of 6%. O&M for wind systems is estimated at
$0.047 per kW-hour based on an estimate provided by the Alaska Energy Authority. O&M for the
Premium Power Transflow 2000 is estimated at $0.03 per kW-hour of energy stored or discharged,
although insufficient field experience is available to verify this estimate. Adding these costs to the
system on a one second basis allows the model to use a wind and load profile of any length and
calculate a cost of electricity for that period.
The simulated high penetration data shows a wind event that begins with calm winds, but rises to
produce wind energy well in excess of the electrical load. This means that sufficient energy exists to
operate the system in a “Diesel Off” mode—but this is not typically done. The problem with a wind-
diesel system without storage is simple—while it is apparent that the wind can meet the load over much
of the time, the wind can decrease to below the load, which could lead to system outages. While it
might seem that it is safe to turn off the diesel engines during the highest wind periods, wind turbines
are designed to protect against damage by tripping off line when wind speeds exceed their maximum
rating, resulting in rapid drops in power during high wind events. For this reason, in the absence of
electrical storage, the diesel engine must remain on, preferably operating at a relatively efficient range.
For this model, a minimum diesel load of 200 kW (25%) was used. However, it should be noted, based
on the diesel efficiency curve (Figure 3), that the diesel engine is operating in a less efficient part of the
operating curve, dropping from about 13 kW-hrs per gallon to about 10 kW-hrs per gallon, which
reduces the savings from the wind power.
Figure 3 Efficiency curve for a Cat 800 Diesel Electric GeneratorExcess wind power can be used for
heating loads, but this load is limited both by the installed controllable loads (using diesel generated
electricity for electrical loads is costly and inefficient—so the loads must be “dispachable” to be on only
when sufficient excess electrical power is available). The amount of wind that can be used is dependent
on both the size and number of the installed controllable loads, and the ability of the community power
lines to carry the additional energy. For this model, a total capacity of 1000 kW of dispachable loads is
assumed. Wind energy that is not used for meeting the electrical or heat loads is considered “excess”
and must be diverted to a controllable dump load.
Figure 5 Model results for high penetration wind system with no electrical storage.
The results from this model run are interesting—the cumulative wind produced (16,419 kW-hrs) is more
than the cumulative energy delivered to the load (still 12,547 kW-hrs, but not shown in figure 5 , but
still, more energy comes from the diesel generator (6,885 kW-hrs) than from the wind (5,662 kW-hrs).
However, the wind provides significant energy that can be used for heat. There is a definite decrease in
the diesel fuel consumed (524.9 gallons, at a cost of $2,924, a savings of 37%, or $1,702.47), but this
savings is is offset by the capital costs of the installed wind turbines.
The capital cost of the wind turbines can be calculated based on a 20 year lifetime, with the cost based
on the reported installed cost of the Kotzebue wind farm, of about $13,510,000 for a 2.9 MW installed
wind farm, using the 6% utility ROI, giving a rate of $0.03735 per second. (Note that this cost is on a
time basis, and not on a per kW-hr basis.) Multiplying by the 85,000 seconds in the run gives a total
capital cost of $3174.73. Note that the capital cost of the wind turbine system is greater than the fuel
savings listed above. When the Wind O&M is included, the total cost of electricity delivered to the load
is $0.60 per kW-hr.
Figure 6 Wind-Diesel Hybrid, no battery (detail)
In Figure 6, the effect of adding high penetration wind to a diesel system is shown. The diesel engine
must support the village load as well as a relatively small parasitic load that must be provided to the
wind farm. When the wind increases, the diesel load drops to the minimum load (200 kW), and remains
there for much of the run. As the wind increases, electrical energy becomes available for delivery to the
dispachable electrical heat loads. In this example, it is assumed that more energy is available than can
absorbed by these heat loads, and must be diverted to a dump load. As the wind drops, the loads are
gradually reversed, and the diesel engine resumes its role as the primary electrical supplier.
Case 3. High Penetration Wind-Diesel-Battery Hybrid System
Adding storage to a Wind Diesel System provides several advantages. Energy storage can be used to
“peak shave”, that is, store energy during gusty periods of wind, then inject this energy when the wind
drops below the load. This is the small system version of energy storage used for “frequency support”,
or “grid stabilization”, but in high penetration wind-diesel systems, this effect is much more pronounced
due to the fact that the variation in the wind energy is from a very small number of turbines, and very
large compared to the load. This function requires that the energy storage system be capable of rapid
changes in energy flows, and to store energy a number of seconds (flywheels are often proposed for this
application).
If the battery is capable of meeting the village load if the wind turbines drop to zero, the diesel engines
can be turned off, reducing diesel consumption. The minimum size of battery is one capable of
supporting the load long enough to bring a diesel engine on line—at least for a few minutes.
If the battery is sufficiently large to power the load for an extended period of time, the diesel engine can
remain off until the battery is discharged. All of these effects are beneficial, adding stability to the
system, reducing diesel consumption—but previously available commercial batteries have high costs
and limited lifetimes. Traditional lead acid batteries require careful management of charging rates and
depth of discharge in order to maximize system lifetime. Lithium Ion batteries can tolerate higher
charge rates so have high power ratings, but are considerably more expensive than lead acid batteries.
Several new utility sized battery suppliers have emerged in recent years that promise greater depth of
discharge, longer system lifetime without degradation, and are sized appropriately for small utility
applications. Most of these batteries are still in the “pre-commercial” development stage, which means
that purchasing and testing these batteries can be challenging, and there remains significant risk that
the batteries may not perform as advertised.
The battery properties used in this model are based on performance specifications provided by the
battery supplier. These properties (especially system lifetime, degradation rates, and O&M costs) have
not been independently verified. Electrochemical performance data and losses from the balance of
plant data were provided and incorporated into the model, and are calculated based on battery current
on a per-second basis, but these details were provided under a non-disclosure agreement. The battery
model was verified by comparing experimental charge-discharge cycle results during the factory
acceptance test with model predictions, with good agreement between the two values.
The battery module allows a maximum charge and discharge rate of 500 kW, and can store 2000 kW-hrs
of energy. The initial state of the battery is at a 200 kW-hr state of charge (10% of the full charge is left
in the battery to facility a “black-start” mode for the power plant). When a wind event begins, wind
power is first used do displace diesel generated power delivered to the load. If the wind continues to
rise, eventually the diesel load will decrease to 25% of full load, which is the minimum load placed on
the diesel. Excess wind energy is then sent to the battery, beginning the charging process, but no
discharge current is pulled from the battery if the diesel engine is operating. If the wind continues to
rise, eventually the battery will reach a 500 kW, at which time the battery can sustain the village load for
approximately 30 minutes, allowing safe operation in a diesel-off mode. If the wind continues to rise,
additional energy is diverted to dispachable heat loads, and if these are saturated, additional energy is
sent to a dispachable dump load. Eventually, the battery will be fully charged (2000 kW-hrs of stored DC
energy), at which time the battery remains off until the energy is needed to supply the electrical load.
Eventually, the battery will be discharged to a 200 kW-hr level, and the diesel engine will be restarted,
returning the system to the original state.
The model calculates the total energy provided by each of the components (diesel engine, wind
turbines, and power delivered by the battery to the load), and sums the cost of operation of each of
these components—diesel fuel, capital costs, and O&M costs.
Figure 7 Cumulative model results for Wind-Diesel-Battery system
The model results show that adding the battery to the system has resulted in several interesting effects.
First, more than half the total load is now being met by the wind system—8840 kW-hrs of wind energy
are directed to the load, or 66.5% of the required energy. The battery provides only about 5% of the
energy to the load, leaving 28.5% to be supplied by the diesel generator. ‘
Figure 8 Wind Diesel Battery hybrid results (detail)
One result worth noting is that the battery state of charge at the end of the model period is at
approximately 1613 kW-hrs of stored energy, which can be delivered as electricity to the load at some
future time beyond the modeling period. However, some of this energy will be lost in both
electrochemical and parasitic losses. When these factors are included, a total of approximately 1171
kW-hours of electricity are available for future loads.
Figure 9 Detail from Wind-Diesel-Battery hybrid model showing events during the increasing wind regime.
Figure 9 shows the events associated with the increasing wind event, with a decrease in diesel output as
the wind rises, eventually resulting in operation at the minimum level and the beginning of the charging
of the battery. Eventually the battery accumulates 500 kW-hr of electricity, and the diesel engine is
turned off. As the wind continues to rise, eventually energy becomes available for the heat load, and
some is eventually sent to the dump load.
Comparing Results for Diesel Only, Wind-Diesel Hybrid, and Wind-Diesel Battery Hybrid
Cases
Figure 10 Energy Delivered to Village Load
Figure 10 shows the sources for energy for each of the cases considered in this study. One very
interesting result is that even though the battery provides only a small amount of the total power
delivered to the system, it allows a much higher level of energy to be provided by the wind. This is due
in large part to the “spinning reserve” provided by the battery, allowing the diesel engine to be turned
off for most of the wind event.
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
Case 1 Case 2 Case 3
Energy Delivered to Village Load
Total Battery to Load kW-hr
Total Wind kW-hr elec
Total Diesel kW-hr
Figure 11 Cost of Electricity for each case
When costs are summed and compared, the impact of the battery on the system is apparent. For the
base case of diesel-only operation, most of the costs are attributed to diesel fuel. In the Wind-Diesel
case, fuel costs are lowered, but this is more than offset by an increase in the higher capital and O&M
costs associated with the wind turbines. Adding the battery to the system provides additional fuel
savings, and even has a small effect on the capital cost per kW-hr, as this case ended with an inventory
of energy remaining in the battery.
Conclusions and Discussion
This work describes a model created to assess the impacts of adding an appropriately sized battery to a
wind-diesel hybrid system in a simulated isolated village. Results indicate that the battery can increase
the amount of wind energy that can be delivered to the load, dramatically reducing the amount of diesel
fuel consumed by the system. Fuel savings are offset by higher capital costs, especially from the wind
turbines, but adding the battery results in electrical energy costs comparable with the diesel base case.
$0.00
$0.10
$0.20
$0.30
$0.40
$0.50
$0.60
$0.70
Diesel Diese & Wind Diesel-Wind-Battery
Cost of Electricity for Diesel, Wind-Diesel, and
Wind-Diesel-Battery Systems
Fuel
O&M
Capital
More significant than the results presented here is the demonstration of a new tool to evaluate the
potential economic impact of adding renewable or energy storage to isolated village systems. Other
models, such as Homer or Hybrid 2 attempt to estimate cost savings from adding renewable energy to
these systems, but do not incorporate features such as instantaneous diesel generator efficiency, the
need for spinning reserves, or the electrochemical properties of the batteries.
The model presented here uses one second data collected in Kotzebue as the basis for the simulation,
and is limited to a single day’s data. This model could be used to evaluate much larger data sets,
although data quality is often an issue. Ideally, a year’s worth of data should be used, or at least
representative 24 hour periods from various times of the year.
Model validation is always a major concern. One advantage of this work over other available models is
that this model can be tested against a system over a relatively brief observation period. For example, a
wind-diesel-battery system could be observed over a one day wind event—if load and wind power (or
wind speed) are recorded on a one second basis for that day, the model can predict total fuel
consumption (as measured by gallons added to the utility day tank) and total kW-hr delivered to the
load (as measured by change in total kW-hr on the utility switch).
The heat balance in these small village systems is also a major issue, and the model is designed to
evaluate this use of energy also. Additional information about heat loads and costs meeting those loads
is needed before it can be modeled.
This model can also be easily modified to evaluate the potential benefits of various other configurations
of small scale systems. Possible work includes : evaluating Wind-diesel-hydro systems (Kodiak Electric
Association has such a system), evaluating an Organic Rankine Cycle generator working on available heat
from a diesel generator (systems are being installed in several Alaskan communities), and evaluating a
device intended to increase efficiency at the low end of the diesel power curve (estimate fuel savings vs
capital cost for these devices).
References:
1. AEA. Alaska Energy Pathways. 2010 [cited; Available from:
ftp://ftp.aidea.org/2010AlaskaEnergyPlan/2010AlaskaEnergyReport.html.
2. DMV-KEMA, ES-Select. 2012.
3. Fay, G., T. Schworer, and K. Kieth, Alaska Isolated Wind-Diesel Systems: Performance and
Economic Analysis. 2010, Alaska Energy Authority.
4. Baring-Gould, E.I., Wind/Diesel Power Systems Basics and Examples, in Wind-Diesel Conference.
2008: Girdwood, Alaska.
5. Bering-Gould, I.a.D.C., Status of Wind_Diesel Applications in Arctic Climates, in Arctic Energy
Summit Technology Conference. 2007: Anchorage, AK. p. 17.
6. Drouilhet, S.a.M.S., Wales, Alaska High-Penetration Wind-Diesel Hybrid Power System--Theory of
Operation. 2002, National Renewable Energy Laboratory. p. 72.
7. Lilienthal, P. HOMER. [cited; Available from: http://www.homerenergy.com/.
8. Kansara, B.a.B.R.P., Modeling And Simulation of Wind- Diesel Hybrid System, in National
Conference on Recent Trends in Engineering & Technology, India. 2001. p. 6.
9. Lal, D.K., Bibhuti Bhusan Dash, and A. K. Akella, Optimization of PV/Wind/Micro-Hydro/Diesel
Hybrid Power System in HOMER for the Study Area. International Journal on Electrical
Engineering and Informatics, 2011. 3(3).
10. Shaahid, S.M., Impact of battery storage on economics of hybrid wind-diesel power systems in
commercial applications in hot regions. International Journal of Energy Research, 2012. doi:
10.1002/er.2930.
11. Manwell, J.F., A. Rogers, G. Hayman, C. T. Avelar, J. G. McGowan, U. Abdulwahid, K. Wu,
HYBRID2- A HYBRID SYSTEM SIMULATION MODEL: THEORY MANUAL. 2007, Renewable Energy
Research Laboratory, Department of Mechanical Engineering, UNIVERSITY OF MASSACHUSETTS.
12. Lone, S.A. and M.u.-D. Mufti. Integrating a Redox Flow Battery System with a Wind-Diesel Power
System. in Power Electronics, Drives and Energy Systems, 2006. PEDES '06. International
Conference on. 2006.
13. Takahashi, R. and J. Tamura. Frequency control of isolated power system with wind farm by using
Flywheel Energy Storage System. in Electrical Machines, 2008. ICEM 2008. 18th International
Conference on. 2008.
14. Takahashi, R. and J. Tamura. Frequency Stabilization of Small Power System with Wind Farm by
Using Flywheel Energy Storage System. in Diagnostics for Electric Machines, Power Electronics
and Drives, 2007. SDEMPED 2007. IEEE International Symposium on. 2007.
15. Sandhu, K.S.a.S.S., Power Quality Improvements in Wind Diesel Hybrid Systems Using Bacteria
Foraging Optimization Technique for Controller Parameter Optimization. International Journal of
Energy Engineering, 2012. 2(4): p. 145-150.
16. Tay, L.H., W.W.L Keerthipala L.J. Borle PERFORMANCE ANALYSIS OF A WIND/DIESEL/BATTERY
HYBRID POWER SYSTEM. 2002, University of Queensland, Australia: Perth, Au. p. 6.