Salee 2018

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

15th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-NCON2018)

Optimal Siting and Sizing of Battery Energy Storage


Systems for Grid-Supporting in Electrical
Distribution Network

Samart Salee and Paramet Wirasanti


Department of Electrical Engineering Faculty of Engineering, Chiang Mai University
239, Huay Kaew Road, Suthep sub-district, Chiang Mai, Thailand, 50200
[email protected] and [email protected]

Abstract— Battery Energy Storage Systems can alleviate the environment and responding to the "smart grid" concept, is
problems that the uncertainty and variability associated with expected to solve problems [3] and [4].
renewable energy sources. The applications such as integration of
renewable energy into the grid, peak shaving in low voltage grid The BESS is applied to adjust the peak demand and reduce
and emergency backup. It could also be used to controlling the the electric bill by discharging a stored energy during load
power flow and grid minimising congestion in the distribution peaks. BESS units have been installed in the electricity grid for
network. In this context, the paper focuses on addressing the purposes of load leveling, stabilizing, load frequency
problems in determining the location and size of BESS by control and voltage control [5]. BESS has a number of features
considering network parameters that play an important role for that can be applied to provide a variety of grid support services
grid-supporting in Electrical Distribution Network. The paper to increase network efficiency and stability [6]. The
proposes the formulation of a problem that include the voltage unexpectedly of RESs could cause a collapse of frequency and
support of BESS to the grid and the total network losses, which voltage. Moreover, it has also effect to the electricity grids
the problems mentioned above is a complex problem and stability. Therefore a BESS was applied to shift time when the
requires the use of specialized methods to solve the problem more demand of decrease during off peak, BESS was used to store
quickly. Next section will present two step of analyzed methods. energy and apply when demand for electricity increases during
The first stage describes to use Genetic Algorithm for preparing peak periods. And BESS was convert to energy output to apply
placement and sizing of BESS. The variables are set to evaluate
with voltage level on transmission and distribution by
the objective function and determine fitness function of Genetic
transformer [7]. Since the consumption of energy storage in
Algorithm equation. The second stage is installing test the
placement and sizing of Battery Energy Storage Systems of the
power systems has emerged, it is expected that many battery
solution provided by the first part for evaluates the fitness by systems will be an important option to installation in low
AC-optimal power flow problem test by DIgSILENT voltage distribution systems to balance energy in accordance
PowerFactory programming. The IEEE 34 bus test feeder is a with the increased of photovoltaic systems [8]. Recent research
test case used to display and discuss the performance of the has shown that BESS has been used for load leveling and
proposed methodology. The test results show that the optimal voltage support. By presenting the results of an effective study
siting and sizing of Battery Energy Storage Systems could be of simulation based on real-life software, the optimal size and
used in solving grid management and grid upgrade on Electrical placement of energy storage devices is a key factor to consider.
Distribution Network integration effectively. BESS's behavior renders both charging conditions for energy
conservation and discharge when the power requirements are
Keywords— Battery Energy Storage System, Electrical met, with a pre-determined value for planning the grid
Distribution Network, Genetic Algorithm, DIgSILENT connection. While the interaction of BESS is controlled by
PowerFactory local primary control loops that try to maintain the voltage at
the pre-set point [9]. The general procedure for determining the
I. INTRODUCTION appropriate size of the BESS device has been verified and is
Integration of Renewable Energy Sources (RESs) into supported by [10] tested with the MATLAB program in the
electrical system requires the application of electronics devices isolated power system model. The results show that both the
and controllers to improve the quality of power generated. size of the device and the optimal operation using dynamically
RESs are difficult to predict to a high degree of accuracy adjustable state of charge limits. Integration of BESS with PV
because their physical processes are very complex [1]. To systems or wind turbines to address the stability issues in
achieve high penetration of RESs require many fundamental distribution systems has been monitored and focused on
changes in the ways that electric power systems and operation developing control methods for optimal BESS performance
to maintain reliable service and economically [2]. Battery [11].
Energy storage systems (BESS) is a new technology, so it When using BESS for grid storage, advanced modeling is
should play an important role in facilitating and adapting to this required to accurately operate strategy for providing the

978-1-5386-3552-0/18/$31.00 ©2018 IEEE 100


15th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-NCON2018)

optimal sitting and sizing. Therefore, this research investigated equipment. By compensating the power from the point where
the methodology and optimal operating strategy for providing a the peak demand is applied to the base load time.
sitting and sizing of BESS. The voltage and total power loss
were objective’s variables in this methodology. The Genetic C. Voltage control
Algorithm (GA) is analyzed by MATLAB. Including, the AC- BESS's ability to support voltage in electrical grids is to
optimal power flow (AC-OPF) of testing system operating store energy at high voltages and feed into the grid at low
strategy is based on DIgSILENT Power Factory. voltages to maintain the voltage profile to the required
standards to guarantee the power supply standard.
The analysis method is separated into two main parts, as
shown in Fig 1. The first part describes to use GA for preparing The following study in this research, BESS applications
placement and sizing of BESS. The variables are set to mentioned above were used for grid-support which consist of
evaluate the objective function and determine fitness function to manage the LV and control the power flow in Electrical
of GA equation in MATLAB. After that, the sitting and sizing Distribution Network (EDN).
values are returned and the fitness value is also received. The
main purpose of the second part is installing test the placement III. GENETIC ALGORITHM
and sizing of BESS. Values from the first part are calculated
for AC-OPF by DIgSILENT Power Factory program. Currently, GA have been widely used to apply for solving
the very complex problem. GA has been also solve the problem
Master Problem (GA algorithm) in the electrical system such as transmission infrastructure,
Determine the BESS site and size in MATLAB electric system flexibility and storage, impact on economic
program problem and load forecasting, reliability and resource
(Constrain of voltage and total power loss)
allocation [12].
In the problems of finding the maximum or minimum for
Fitness value Site and site of BESS the objective function are discontinuities search space and sub-
Analysis Optimal power flow in optimum peaks. This problem is more complicated when using
DigSILENT PowerFactory program
(Power flow constrain)
the calculus and derivative of the function to find the maximum
and minimum values. GA technique is the most valuable
Fig. 1 Proposed interaction of the MATLAB and PowerFactory program. process which use the random sampling search procedure. The
theory of evolution and natural selection of GA has been use to
The paper is structured as follows. Section 2 entails BESS. find the maximum and minimum of function by random a set
The theory of GA are explained in detail in Section 3. The of defaults through the GA process. After that maximum and
proposed algorithm is in Section 4. The case study and results minimum values were found in the search space at the same
which GA was analyzed by MATLAB programing. Including, time using the optimal value (fitness) converted from the GA
the AC-OPF of testing system operating strategy is based on function. This technique has the advantage and is the way to
DIgSILENT Power Factory programing are shown in Section find the minimum, from the first point to the second point, and
5. The simulation results are discussed in the last section. the point one at a time. And it is a technique that gives the
opportunity the most appropriate value (local optimal) [13].
II. APPLICATION OF BATTERY ENERGY STORAGE SYSTEMS GA process was studied in this paper which starts with the
IN ELECTRICAL DISTRIBUTION NETWORK initial population generated randomly (chromosome). The
Grid management at the distribution level uses position and size of the BESS are evaluated using the fitness
decentralized energy storage because of the dynamic behavior function. The next generation value will be redundant or cut off
of BESS, able to respond to fast and efficient power supply at their optimal value (fitness value). It determines the number
variations. Its can "smoothly discharge electricity" from of iteration to achieve the convergence of the constrain
various types of RESs to low-voltage and medium-voltage function. All fitness values were recorded and compared to the
networks and can also enabling for compensation of fluctuating lowest values. The value which is less than the best case
renewable generation to match with demand, mitigating scenario, is the best solution. If any solution is less than the
congestion, by controlling the power flow and maintaining basic case, it is not suitable for the BESS installation at that
voltage in an appropriate range [11]. BESS can provides a wide location and size. Then GA process will work again to
range of application for operating grid as follow below steps: investigate the problem. Comparison between all solutions can
be made to identify the best approach. And the next generation,
A. Peak shaving in low voltage grid population of site and size are generated according to the GA
In this functionality, BESS will be responsible for storing process. Eventually, the population of BESS optimal siting and
electrical energy at low demand and releasing electrical power sizing and highly efficient.
in times of high demand. This reduces the LV grids peak
demand and increases the electricity grid's capacity. IV. PROPOSED PROBLEM APPROACH
B. Load levelling in low voltage grid The objective of the master problem is the optimal location
BESS's ability to store and discharge electrical and size of BESS in EDN. Voltage and power control are
energy helps to reduce peak currents in the grid to reduce the important variables to consider for solving this problem. The
risk of overload that would have a serious impact on electrical purpose is to stabilize the distribution system and increase the

101
15th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-NCON2018)

capacity on the transmission line. Focus on the importance of Where:


determining the best location installation and the right size. To Voltage constrain : The voltage value of another bus in power
study the optimum installation factors by considering the AC- grid should be within voltage limit as set :
OPF. There are variables and conditions for establishing the
objective functions to be taken into account: voltage, power
and total loss of the system to manage the distribution network Vmin d Vi d Vmax
can be formulated as:
n
Vi : Voltage of bus i
 obj w1 ¦ (Vi  Vref ) 2  w2 PTloss   
i Vmax : Voltage maximum is 1.05 pu.
From Equation (1), the first part considers the consistency
of the deviation voltage at the nodes in the electrical network,
Vmin : Voltage minimum is 0.95 pu.
which is analyzed in terms of the quadratic equation that Total power loss constrains : The total power losses after
(Vi  Vref ) showing the voltage deviation arranged in installing BSSS in power grid should be less than before
installing BESS but if it increases not more than quarters of
accordance with sequence and the impact on the feeder the original as following
voltage profile. And in the next part will consider the total
power loss in a given system.
PTloss d 125% POri
Where :
Vi : Voltage of bus i
PTloss : Total power loss after install BESS
Vref : Magnitude of the voltage reference
POri : Total power loss of base system
PTloss : Total power loss
n : Total number of three phase nodes. The global constraints of the problem are [14]:
w1 and w2 : The Objective Function weights (power losses x Maximum number of BESS units installed in the
and Cumulative Voltage Deviation weights), subjected to: EDN.
x BESS operational and capacity constraints.
x Daily AC-OPF constraints.
w1  w2 1
A. Genetic algorithm used to solve the main problem. From the above mentioned problems, there are both integer
and non-integer variables. The characteristics of the equations
From the problem mentioned above, GA method used to are mixed, with both linear and non-linear equations. This
solve the correct siting and sizing of BESS. In this article will show you how to handle the structure of such
optimization problem the fitness value is the objective value complex equations.
from the AC-OPF. So, the fitness function of the GA are:
The placement and sizing of BESS are solved by defining
the fitness function and also optimized by constraints.
FGA objOPF (2) Reducing power losses, increasing voltage stability margin
and reducing cumulative voltage deviation can be evaluated
Where, FGA is the fitness value and objOPF is the by fitness function in this system. The different site and sizing
of BESS is processed automatically and proposes BESS size
objective value from optimal power flow analyze. limits. AC-OPF program is integrated with GA method until
B. Fitness Function the minimum solution is achieved for the suggest placement.
Function fitness or objective function retrieves the required Each of propose locations is repeated in this process.
data from the calculated value for estimate optimal sizing and An algorithm for improving the optimal sitting and sizing
location of BESS. In this research, cumulative voltage of BESS are presented in Fig 2. This procedure is divided into
deviation is selected with most importance that the voltage has three steps. First, setting data process, the initial value of
the most significant effect on the operation of the network. sitting and sizing were set, iteration and constrain,
The amount of power loss is chosen as the next priority. respectively. The calculation in this step was assessed by GA
Cumulative voltage deviation and power loss decrease and in MATLAB. Next step, DIgSILENT Power Factory and
efficiency also decrease. However, Fitness function is MATLAB program are integrated for receiving and
calculated when the values of BESS for the selected bus. transferring values. Received values are estimate for suitable
C. Constraints conditions. Then, the minimum values of sitting and sizing for
installing BESS are calculated. The results are the optimal
The optimization constrains in the method feature in the placement and sizing values of BESS for testing system in this
three part : study. Validation is tested several case studies, iteration values
1. Voltage constrains are tested until approach to accuracy value.
2. Total power loss constrains

102
15th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-NCON2018)

Start

Create initial site and size of BESS DATA


(GA on MATLAB program)

Define Iteration
N = Correct Value

Set Constrains
-Voltage
-Total power loss

S=0;
S=S+1;

Create New Chromosome

Import to DIgSILENT PowerFactory AC-OPF


Program

No
MATLAB get voltage and total loss value

No

Check Constrain

Yes

Evaluate fitness function

Get each chromosome

S=N

Yes

Compare fitness function for all chromosome


Calculation weights and choose minimum one Fig. 3 IEEE 34 Node Test Feeder in DIgSILENT PowerFactory.

This research sets out to prove that the method is able to


satisfy the constraint concerning BESS installation which, this
method was applied to solve the problem of voltage drop and
Select best site and maximum size value
active power loss in electrical system. Therefore, two case
studies were designed, i.e. without and with BESS installation
Remarks, only three phase buses were selected for setting to
End analyze an optimal placement and sizing of BESS.
The power flow solution is calculated by DIgSILENT
Fig. 2 Flowchart of GA and OPF programing PowerFactory program. It considers three main components of
the system were shown in Table I. For base case study (without
V. CASE STUDY AND RESULTS
BESS): 1) Total active power is 1.045948 MW 2) Voltage
An IEEE 34 Node Test Feeder, which is modeled in range is in the period of 0.94 p.u. to 1.05 p.u. and 3) Total
DIgSILENT PowerFactory, is shown in Fig. 3. It is used for a active power loss is 74.65 KW. From the above information,
case study. The based voltage of this case study is medium the voltage drop problem was encountered. Then, the
voltage at 24.9 kV. previously metrology was proposed approach, which the results

103
15th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-NCON2018)

of simulation in case study after install BESS found that the


optimal position for the BESS installation is the 834 bus and
the appropriate size is 0.7 MW. The value obtained by
analyzing the power flow of case study (with BESS) showed
the 1.00 p.u. to 1.05 p.u. and the total active power loss is
16.04 KW, which is noted that the voltage level increase and
the total loss in the network is lower than the value obtained
from the base case.
TABLE I THE POWER FLOW SIMULATION RESULTS
Simulation Results
Case Study Total Active
Total Active Voltage range
Power Loss
Power (MW) (p.u.)
(KW)
without BESS 1.045948 0.94-1.05 74.65 Fig. 4 Convergence characteristics.
with BESS 1.045998 1.00-1.05 16.04 The testing results in Fig.5 showed that after installing the
BESS at the correct position and size. It can fix the low voltage
The values of the optimal placement and sizing which were problem and change voltage profile in the overall system better
calculated by manual simulation, were equal to the value from when compared to the base case (Without BESS). Also, the
already mentioned method. On the other hand, the manual total loss of system has decreased as shown in Fig 6. In this
simulation method was used; it was expected to 260 cases for research, BESS was applied in test case which consist of to
completing analysis. Fig. 4 represent that if this method was manage the low voltage, total active power loss and to control
approach iteration at 25, a correct answer could be the optimal power flow. The overview results from these two
demonstrated. So, the GA performance when 25 replicates in comparisons showed the effectiveness of the optimal siting and
the scenario. The quality of the randomly generated population sizing of BESS.
will vary for every job. But the solution from GA is similar.
Therefore, this method can reduce the analysis time.

Fig. 5 Comparison the differences of Voltage Profile of Base Case (Without BESS) and after install BESS (With BESS).

Fig. 6 Comparison the differences of Active Power Loss (MW) in Lines of Base Case (Without BESS) and after install BESS (With BESS).

104
15th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-NCON2018)

[3] Ahmed S. A. Awad, Tarek H. M. EL-Fouly, Magdy M. A. Salama,


VI. CONCLUSION “Optimal ESS Allocation for Load Management Application” IEEE
trans. Power syst., vol. 30, pp.327-336, Jan. 2015.
This article has presented a methodology used to determine [4] C. Chen, S. Duan, T. Cai, B. Liu, and G. Hu, “Optimal allocation and
the optimal site and size of BESS. Impact on voltage level and economic analysis of energy storage system in microgrids”, IEEE Trans.
total loss in the EDN are proposed in this research. The Power Electron., vol. 26, no. 10, pp. 2762–2773, Oct. 2011.
analysis method concludes two processes. Using GA for [5] FO. Alexandre, C. Rachid, B. Antoine, “Sizing and Optimal Operation
evaluated location site and sizing of BESS by MATLAB of Battery Energy Storage System for Peak Shaving Application” ,
software. Next, installing test the siting and sizing of BESS for IEEE, PowerTech 2007, pp.621-625
calculated the AC-OPF by DIgSILENT PowerFactory [6] M. Pascal, C. Rachid, O. Alexandre, “Optimizing a Battery Energy
Storage System for Frequency Control Application in an Isolated Power
program. And evaluate the fitness of solution provided from System” , IEEE transactions on power systems, vol. 24, no. 3, pp.1469-
the overview of the two parts. 1477, Aug. 2009.
The test cases referring to IEEE 34 bus distribution test [7] I. Atteya, N. Fahmi ,D. Strickland, and H. Ashour, “Utilization of
Battery Energy Storage Systems (BESS) in Smart Grid : A Review”,
feeder are used to display and evaluate the performance of the Renewable Energy and Power Quality Journal (RE&PQJ), Vol.1, No.14,
proposed method and The AC-OPF test considers three key May 2016.
components of this system, i.e. total active power, total active [8] P. Fortenbacher, M. Zellner, E. Zurich, “Optimal Sizing and Placement
power loss and voltage range. Test results show that the of Distributed Storage in Low Voltage Networks” 21 Mar 2016
location and size of the BESS can be used to resolve [9] G. Celli, F. Pilo G, G. Soma D. Dal Canto, E. Pasca, A. Quadrelli,
distribution network management issues to support BESS “Benefit Assessment of Energy Storage for Distribution Network
installations. This method can be used to analyze complex Voltage Regulation”, CIRED Workshop - Lisbon 29-30 May 2012Paper
electrical systems better than manual simulation and faster than 0330
traditional analysis. [10] C.F. Lu, C.C. Liu, C. J. Wu, “Dynamic modelling of battery energy
storage system and application to power system stability”, IEEE
Proceedings - Generation, Transmission and Distribution, Volume 142,
ACKNOWLEDGMENT Issue 4, July 1995, p. 429 – 435
[11] EUROBAT, “Battery Energy Storage for Smart Grid Applications” ,
We would like to thank Department of Electrical Association of European Automotive and Industrial Battery
Engineering, Faculty of Engineering Chiang Mai University for Manufacturers, 2013
supporting the research laboratory and source of important data [12] Mohammad Shahidehpour, Hatim Yamin, Zuyi Li, “Market Operations
to analyze in this research. Moreover, we also would like to in Electric Power Systems : Forecasting, Scheduling, and Risk
thank the Graduate School, Chiang Mai University for grateful Management.,” New York, NY 10158-0012,2003, Appendix C.
the technical support. [13] Q. Zhong, N. Yu, X. Zhang, Y. You and D. Liu, “Optimal Sizing and
Sizing of Battery Energy Storage System In Active Distribution
Network”, 4 th IEEE PES (ISGT Europe), Copenhagen, pp: (1-5), 6-9
REFERENCES Oct. 2013.
[1] PSERC Executive Committee, “Challenges in Integrating Renewable [14] M. Nick, M. Hohmann, R. Cherkaoui and M. Paolone, "Optimal location
Technologies into an Electric Power System”, Apr. 2014. and sizing of distributed storage systems in active distribution
[2] International Electrotechnical Commission, “Electrical Energy Storage”, networks," 2013 IEEE Grenoble Conference, Grenoble, 2013, pp. 1-6.
2017.[Online].Available: www.iec.ch/whitepaper/energystorage.html

105

You might also like