Abstract—The important of electric power distribution is to have centralized plants distributing electricity through Distributed generation (DG) which reduces the cost of maintenance on transmission a n d distribution station and also improve voltage profile. This research paper present the application of generation based on Biogas power renewable energy source to the Distribution network and how it stabilizes the network by normalizing the fluctuating voltage profile at the distribution end of power system. A Particle Swarm Optimization (PSO) model was performed and evaluation of the impact of the DG by stimulating the developed model in the system. A mathematical formulation and optimization algorithm was performed using the MATLAB/Simulink program. The results obtained were correction of the faulty buses voltages and stable power supply which is 29.4% better than the conventional one. The result shows the implementation of the optimisation technique has improved the energy efficiency of the distribution network.
Index Terms—Distributed Generation; MATLAB- SIMULINK; Particle Swarm Optimization; Voltage Profile.
I. INTRODUCTION
The power system is very complex due to high operational demand for electricity supply and load density.
The primary cause of instability in power system planning is the load demand which is generate due to the electricity supply variety need from a different class of customers [5].
Therefore, there is always voltage profile degradation, and the voltage profile reduced in the distribution system which moves away from the sub-station [23]. Thus, the ever increase of energy demand necessitated delivering electrical power over long transmission and distribution line to meet [15], [6]. Many conventional methods of electricity generation use primary sources of energy that are non- renewable and as such their thoroughness in nature. This calls for the use of distributed generation (DG) and the use of DG technology. DG system is a current smart grid concept which forms the backbone of a modern distribution system. The DG can be renewable energy source (RES) like wind turbines, photovoltaic, small hydro, biomass etc. or fossil fuel based sources such as internal combustion engines (IC), combustion turbines and fuel cells [25].
The Distributed Generation (DG) connection at the
Published on March 19, 2019.
Authors are with the Department of Electrical and Electronics Engineering, University of Abuja, Nigeria. (e-mail:
consumer side has given many opportunities for the network to minimize the existing power losses. When the DG units is implemented, the distribution system will no longer have a single supply system, which is from the transmission- distribution substation, but there will be multiple sources power in the network. While the other loads get the power supply from the main source, the DG units will supply to part of the local load. However, the size and location of the DG incompatibility will give an opposite effect to the distribution network such as voltage operating beyond the limit, power loss increase and others [2],[13],[17],[14],[19].
Literature Review
In solving an optimization problem in terms of optimal location and sizing of DG, there are several methods proposed, which can either be stochastic search algorithms or conventional. Example of the traditional technique includes using the analytical way which involves exact loss formula [12], a bus impedance matrix of study devoid [4], load concentration factor analysis [16], Power Voltage Sensitivity Constant method [22], etc. It is clear from the literature [3] that the conventional or classical method of analysis has the disadvantage of finding the optimal solution for the nonlinear optimization problems.
The analysis in [18] studies on optimal placement and sizing of DG units in the distribution system. The author used combination of two heuristic optimization methods which are Immune System and Genetic Algorithm in order to maximize the important of DG. Since the power losses and the cost of reinforcement are directly related to the size of DG, thus the authors reduce both factors in order to achieve the optimal DG output in the system. The authors have compared the method with other optimization methods which are GA, PSO, Immune Algorithm (IA), Ordinal Optimization (OO) and GA-OPF. However, the locations of DGs are diverse for each optimization methods and can cause the optimal power losses values to be different.
Reference [9] proposed the use of continuation power flow in order to determine the most sensitive buses to voltage collapse and used as the applicant for DG placement. Many other methods use heuristic techniques for DG placement in Distribution networks among which one of the approach [20], for optimizing the problem which is based on the concept of Pareto Optimality. In [10], optimal DG placement and enclosures based on reliability of the network was done by using Ant Colony Search (ACS) algorithm.
Application of a Particle Swarm Optimization for Improving Voltage Profile with Distributed Generation: A
Case Study of 33/0.415KV Abuja Airport Injection Substation
Stephen Oodo, and Felix S. Owolabi
II. DISTRIBUTED GENERATION
Distributed generation can be defined as an electric power source which is connected directly to the distribution network or on the customer site of the meter [1].
This definition of distributed generation is not analyzed and discussed the rating of the generation source, as the maximum rating depends on the local distribution network conditions, e.g. voltage level. Also, the definition of distributed generation does not define the technologies, as the technologies that can be used vary widely. Ackermann's definition is the most generic one, because there is no limit on the DG size and capacity and his definition covers the location of the DG. [11] Defines DG as "the generation of electricity by facilities that are sufficiently smaller than central generating plants so as to allow interconnection at nearly any point in a power system." IEEE compared the size of the DG to that of a conventional generating plant in their definition.
A more precise definition is provided by the International Council on Large Electric Systems [7] and The International Conference on Electricity Distribution (CIRED), which defines DG based on size, location, and type. CIGRE defines distributed generation as "all generation units with a maximum capacity of 50 MW to 100 MW, that are usually connected to the distribution network and that are neither centrally planned nor dispatched." CIRED defines DG to be
"all generation units with a maximum capacity of 50 MW to 100MW that are usually connected to the distribution network.
There are usually three operators in a typical genetic algorithm [8]: the first is the production operator (elitism) which makes one or more copies of any individual that possess a high fitness value; otherwise, the individual is eliminated from the solution pool; the second operator is the recombination (also known as the 'crossover’) operator.
III. PARTICLE SWARM OPTIMIZATION (PSO) PSO is an algorithm which based on emergent behavior and social dynamics that develops a population of individual to the space search. Edward and Kennedy formulated PSO in 1995. The social behavior of animals, such as fish schooling was developed by the algorithm. PSO is similar to the continuous GA in that it begins with a random population matrix. For each particle then adds that velocity to the particle position or values. Velocity updates are influenced by both the best global solution associated with the lowest cost ever found by a particle and the best local solution associated with the lowest cost in the present population. If the best local solution has a cost less than the cost of the current global solution, then the best local solution replaces the best global solution. The particle velocity is reminiscent of local minimizes that use derivative information, because velocity is the derivative of position.
The constant G1 is called the cognitive parameter [21]. The constant G2 is called the social parameter. The advantages of PSO are that it is easy to implement and there are few parameters to adjust [24]. The particle swarming becomes evident as the generations pass. The largest group of particles ends up in the vicinity of the global minimum and
the next largest group is near the next lowest minimum.
A. Proposed Algorithm
The PSO Algorithm used in this paper for analyzing the system is presented in the power flow technique and the optimization method which was programmed in MATLAB/Simulink.
The loads of this research paper was modeled as constant power, which describe real and reactive power absorbed by the system in the power injections at the buses. In order to keep a constant power injection, the loads are characterized by reacting to changes in the voltage or current. Loads at PQ buses are modeled as constant active 𝑃𝐿𝑘 and reactive power 𝑄𝐿𝑘provided that voltage at bus k is within acceptable limits as:
Loss Reduction
Voltage profile Improvement
The items are composed with constraints to obtain the proper objective functions. The main constraints are:
1) Load Losses
Before installing DG in power grid should be less than losses after installing of it. Loss with DG ≤ Loss without DG
2) Voltage constraint
V bus min ≤ V bus ≤ V bus max
The objective function, with composing constraints and goals, is determined as following:
n
i
W k
T k
n R k
F
1
3 2
1 1 max(0, max(0,
, 0 max[
max (1)
withoutDG withDG Voltage i i
Voltage
R % % (2)
m
i
withDG withoutDG Pi Pi
T
1
(3)
m
i
withDG withoutDG Qi Qi
W
1
(4)
First term in this expression, related to difference between average of voltage profile percentage in base case and other cases according to DG’s locations. By this way summation of active and reactive power losses difference are computed and objective function is established. Max operator is used for enforcing the constraints. The negative values influence is forbidden by this operator. Mentioned parameters are listed below:
withDG
i
voltage% : Voltage Percent in ith bus with DG resource.
withoutDG
i
voltage% : Voltage Percent in ith bus without DG resource.
withDG
Pi : Active Power Losses in jth branch with DG resource.
withoutDG
Pi : Active Power Losses in jth branch without DG resource.
withDG
Qi : Reactive Power Losses in jth branch with DG resource.
withoutDG
Qi : Reactive Power Losses in jth branch without DG resource.
K1,k2,k3: Emphasis or penalty factors n: Number of Buses
m: Number of Branches.
𝑃𝑘= −𝑃𝐿𝑘 (5)
𝑄𝑘= −𝑄𝐿𝑘 (6)
In the case that voltage limits are violated, PQ loads are modified for as:
𝑃𝑘=−𝑃𝐿𝑘𝑉2𝑘
(𝑉𝑙𝑖𝑚𝑘)2 (7)
𝑄𝑘 =−𝑄𝐿𝑘𝑉2𝑘
(𝑉𝑙𝑖𝑚𝑘)2 (8)
Where 𝑉𝐿𝑙𝑖𝑚 is depending on the case 𝑉𝐿𝑚𝑎𝑥o𝑟 𝑉𝐿𝑚𝑖𝑛. The overall efficiency of the Biogas is between 20-40%
of the distributed generation and its power rating varies between 0.3 to 7 MW in order to avoid waste in the system.
The advantages of Biogas energy are that it is a clean power source, and the cheapest technology compared to other types of renewable energies.
TABLEI:THE PARAMETERS FROM ABUJA AIRPORT (DISTRIBUTION NETWORK)
TIME KV1 KV2 KV3 I1 I2 I3 PH RG HZ MW KV1 KV2 KV3 7am 32 32 32 25 25 25 G 50 1.2 11 11 11 8am
9am 30 30 30 27 27 27 G 50 1.4 11 11 11 10am
11am 30 30 30 27 27 27 G 50 1.4 11 11 11 12noon
1pm 30 30 30 28 28 28 G 50 1.4 11 11 11 2pm
3pm 33 33 33 27 27 27 G 50 1.4 11 11 11 4pm
5pm 33 33 33 25 25 25 G 50 1.3 11 11 11 6pm
7pm 8pm 9pm 10pm 11pm 12M.N 1am 2am
3am 33 33 33 20 20 20 G 50 1.1 11 11 11 4am
5am 6am
The empirical data in Table I was used to run the load flow of Newton Raphson as shown in Fig 1. The reason of
doing that is to determine the faulty buses that did not fall within the range of 0.95 t0 1.05 per unit volts.
Fig. 2. Power flow result of evaluation of the voltage profile of the network without a distribution generation using per- unit system
Fig 3. Optimized faulty volts using particle swarm optimization
Fig. 3 shows optimized faulty volts using particle swarm optimization
The obtained particle swarm optimization result is 1.0316 that has to be used to normalize the faulty volts.
Fig. 4. Development of a model for the integration of the DG to the network without using particle swarm optimization technology.
Fig. 4 shows Development of a model for the integration of the DG to the network without using particle swarm optimization technology. Fig. 5 shows that the per unit volts
when particle swarm optimization is not used for the nine buses are 1.06, 1.06, 1.0561, 1.082, 1.064, 1.071 and 1.058 respectively that cause unstable power supply.
Fig. 5. Development of a model for the integration of the DG to the network with using particle swarm optimization technology
The model shows that the per unit volts when particle swarm optimization is used for the nine buses are 1.028, 1.028, 1.024, 1.049, 1.031, 1.038, 1.026, 1.023 and 1.05 respectively that cause stable power supply.
TABLEII:COMPARING FAULTY PER UNIT VOLTAGES WITHOUT PSO BU
S1 BU S9
BU S10
BU S11
BU S12
BU S13
BU S14
BU S15
BU S16
TIM E(S)
0 0 0 0 0 0 0 0 0 0
1.5 1.5 1.4 1.6 1.3 1.42 1.41 1.5 1.35 1
1 0.8 0.9 1.2 1 1.0 0.95 1 0.8 2
1.0 0.9 0.8 1.1 0.8 0.91 1 0.8 1 3
1.02 1 1 1.03 1.0 1.01 1.05 1 0.9 4
1.06 1.06 1.05 6
1.08 2
1.06 4
1.07 1
1.05 8
1.05 5
1.05
7 5
1.06 1.06 1.05 6
1.08 2
1.06 4
1.07 1
1.05 8
1.05 5
1.05
7 6
1.06 1.06 1.05 6
1.08 2
1.06 4
1.07 1
1.05 8
1.05 5
1.05
7 7
1.0 6
1.0 6
1.05 6
1.08 2
1.06 4
1.07 1
1.05 8
1.05 5
1.05
7 8
TABLEIII:COMPARING RECTIFIED PER UNIT VOLTAGES WITH PSO BU
S1 BU S9
BU S10
BU S11
BU S12
BU S13
BU S14
BU S15
BU S16
TIM E(S)
0 0 0 0 0 0 0 0 0 0
1.4 1.5 1.3 1.6 1.2 1.52 1.45 1.7 1.47 1
0.9 0.8 0.7 1.2 0.7 1.2 0.96 1.2 0.9 2
1.2 1.0 0.8 1.1 0.8 0.93 1.2 0.9 1.2 3
1.03 1.2 1 1.03 1.0 1.01 1.02
6 1.1 1.0 4
1.02 8
1.0 28
1.02 4
1.04 9
1.03 1
1.03 8
1.02 6
1.02 3
1.02
5 5
1.02 8
1.0 28
1.02 4
1.04 9
1.03 1
1.03 8
1.02 6
1.02 3
1.02
5 6
1.02 8
1.0 28
1.02 4
1.04 9
1.03 1
1.03 8
1.02 6
1.02 3
1.02
5 7
1.02 8
1.0 28
1.02 4
1.04 9
1.03 1
1.03 8
1.02 6
1.02 3
1.02
5 8
Fig. 6. Comparing faulty per unit voltages without PSO
Fig. 7. Comparing rectified per unit voltages with PSO
IV. DISCUSSION OF RESULTS
The modeling of the components discussed in this section is based on the assumption that the three phase system is balanced under steady state conditions. Using this assumption, per phase analysis can be done. With the result obtained after the load flow, it shows that the faulty buses that did not fall within the range of 0.95 to 1.05 per unit volts are buses 1 = 1.06 per unit volt, bus 9 = 1.06 per unit volt, bus 10 = 1.056 per unit volt, bus11 =1.082 per unit volt., bus 13 = 1.071 per unit volt, bus14 = 1.058 per unit volts, bus15 = 1.055 per unit volts and bus 16 = 1.057 per unit volts.
Fig. 6 shows that the buses 1,9, 10, 11, 12, 13,14, 15 and 16 do not meet to the per unit volt of 0.95 to 1.05 that gives stable power supply in the distribution network. The per unit volts of these buses at time 5s through 8s are 1.06, 1.06, 1.056, 1.082, 1.064, 1.071, 1.058, 1.055 and 1.057 respectively. At these per unit volts, it causes over current, over voltage, low power factor, short circuit that constitutes instability in power supply without using particle swarm optimization.
Fig. 7 shows comparing rectified per unit voltages with PSO. The result shows that buses 1,9, 10, 11, 12, 13,14, 15 and 16 meet to the per unit volt of 0.95 to 1.05 that gives stable power supply in the distribution network. At a time of 5s through 8s the nine buses meet with stable per unit volts of 1.028, 1.028, 1.024, 1.049, 1.031,1.035, 1.026, 1.023 and 1.025 that enhance stable power supply devoid of over current, low power factor, short circuit when particle swarm optimization is used.
V. CONCLUSION
This paper reveals the application of Particle Swarm Optimization integration with energy generation on the
distribution network to improve voltage profile and reduce loses with the operation of the voltage regulation and reactive power distribution network.
The application of distributed generation into the power systems has been analysed, In order to model a generation based on renewable energy, firstly energy sources have been studied; secondly, the technologies have been revised and finally the integration of distributed generation into the grid it has been carried out. The Power flow analysis has been programmed, by formulating main equations of electrical circuits in MATLAB enabling the study test system to be modelled.
In this paper, a standard system has been modelled based on load flow analysis in different days which shows the impact of distributed generation on Abuja AirPort injection Substation.
Various optimisation algorithms have been implemented based on the principle of natural selection to solve issues such as the location, the level of generation or control of the power factor of the connected generators.
All optimization algorithms and mathematical formulation have been performed using the MATLAB/Simulink program. It can be concluded that the implementation of the optimisation technique has improved the energy efficiency of the distribution network.
REFERENCES
[1] Ackermann T., etal (2001) "Distributed generation: a definition,"
Electr.Power Syst. Res., vol. 57, pp. 195-204.
[2] A. Uniyala and A. Kumar, "Optimal Distributed Generation Placement with Multiple Objectives Considering Probabilistic Load "
Procedia Computer Science vol. 125, pp. 382–388, 2018.
[3] A. Marimuthu, K. Gnanambal, R.P. Eswari and T. Pavithra, “optimal Location and sizing of DG units to improve the voltage stability in the Distribution system using particle swarm optimization Algorithm with time varying Acceration coefficient, “GRID Journals (Global Reasearch and Development journal for Engineering), international conference on innovation in Engineering and Technology,2016 [4] A. Tah and D.Das, ”Novel Analytical method for the placement and
sizing of distributed generation unit on distribution networks with and without considering P and PQV buses, “international journal of electrical power and energy systems, vol.78, pp.401-413, 2016.
[5] Abu-Mouti, F.S. El-Hawary, M.E, “Heuristic curve-fitted technique for distributed generation optimisation in radial distribution feeder systems”, Generation, Transmission & Distribution, IET, vol.5, no.2, February 2011, pp.172-180.
[6] Alemi, P., Gharehpetian, G.B., “DG allocation using an analytical method to minimize loss and to improve voltage security,” Power and Energy Conference, 2008, PECon 2008. IEEE 2nd International, vol., no., 1-3 Dec. 2008, pp.1575-1580.
[7] CIGRE, "CIGRE Technical Brochure on Modelling New Forms of Generation and Storage," CIGRE TF 38.01.10, November 2000.
[8] Chambers A., etal, (2001) Distributed Generation: A Nontechnical Guide: Penn Well.
[9] Feng Dong; Chowdhury, B.H.; Crow, M.L.; Acar, L.; “Improving voltage stability by reactive power reserve management," Power Systems, IEEE Transactions on, Vol. 20, no.1, pp. 338 - 345. DOI:
10.1109/TPWRS.2004.841241, Feb. (2005).
[10] G.P. Harrison, A Piccolo., P Siano., A.R Wallace, (2008) “Hybrid GA and OPF for distribution generation connections,” Electr. Power Energy Syst., Vol.78, pp. 392–398.
[11] IEEE Std 1547-2003,"IEEE Standard for Interconnecting Distributed Resources with Electric Power Systems", 2003.
[12] M.Mirzaei, J.Jasni, H.Hizam, N.I.A. Wahab and E.Moazami,”A combined Analytical method for optimal location and sizing of distributed Generation considering voltage stability and power loss in a power distribution system vol.25,NO. 1, pp29-42, 2007.
[13] M. Kashyap, S. Kansal, and B. P. Singh, "Optimal installation of multiple type DGs considering constant, ZIP load and load growth,"
International Journal of Ambient Energy, 2018.
[14] M. Kashyap, A. Mittal, and S. Kansal, “Optimal Placement of Distributed Generation Using Genetic Algorithm Approach," In V.
Nath and J. K. Mandal (eds.), Proceeding of the Second International Conference on Microelectronics, Computing & Communication Systems (MCCS 2017), Lecture Notes in Electrical Engineering 476, https://doi.org/10.1007/978-981-10-8234-4-47, 2019.
[15] M.F.Kotb, K.M.Sheb, M. El Khazenda, A. El Hussein, “Genetic Algorithm for Optimum Siting and Sizing of Distributed Generation”, The 10th MEPCON International Middle East Power Systems Conference, December 19-21, 2010.
[16] M. Shahzad, I. A hmad, W.Gawlik and P.Palensky, “Load concentration factor based Analytical method for optimal placement of multiple distribution Generators for loss minimization and voltage profile improvement Energies Vol.9, NO 287, 2016.
[17] M. S. Sujatha, V. Roja, and T. N. Prasad, "Multiple DG Placement and Sizing in Radial Distribution System Using Genetic Algorithm and Particle Swarm Optimization," Ch. Satyanarayana et al.,Computational Intelligence and Big Data Analytics, Springer Briefs in Forensic and Medical Bioinformatics, https://doi.org/10.1007/978-981-13-0544-3_3, 2019.
[18] Mrs.A. Lakshmi Devi and Prof B. Subramanyam, “A New Hybrid Method for Loss Reduction and Voltage Improvement In RadialDistribution System- A Case Study”, i- manager’s journal on Electrical Engineering, Vol.1, No.3, pp. 64 - 68, January-March (2008).
[19] N. Kanwar, N. Gupta, K. R. Niazi, and A. Swarnkar, "Optimal distributed generation allocation in radial distribution systems
considering customer-wise dedicated feeders and load patterns," J.
Mod. Power Syst. Clean Energy, vol. 3, no. 4, pp. 475–484, 2015.
[20] Rafael Cossent *, Luis Olmos Tomás Gómez, Carlos Mateo Pablo Frías” Distribution network costs under different penetration levels of distributed generation” European Transactions on Electrical Power Volume 21, Issue 6, pages 1869–1888. :7 OCT 2010, DOI:10.1002/etep.503, September (2011).
[21] Russell C. Eberhart, Yuhui Shi, James Kennedy, “swarm intelligence”, Morgan kaufmann; 1 edition April 19, 2001.
[22] S. Nawaz, A.K Bansal and M.P. Sharma, “A novel Approach for multiple DG Allocation in Real Distribution system, “international journal of Engineering and Technology. Vol.9, NO.2, 2017.
[23] T. Gozel, M. H. Hocaoglu, “An Analytical Method for the Sizing and Siting of Distribution Generators in Radial System”, International Journal of Electric Power System Research, Vol. 79, No. 6, June 2009, pp. 912-918.
[24] Wichit Krueasuk and Weerakorn Ongsakul, “optimal placement of Distributed Generation using particle swarm optimization”, in proceedings of power engineering conference in Australasian University, Australia 2006.
[25] Zhipeng Liu; Fushuan Wen; Ledwich, G.; Xingquan Ji; “Optimal sitting and sizing of distributed generators based on a modified primal-dual interior point algorithm”, Electric Utility Deregulation and Restructuring and PowerTechnologies (DRPT), 2011 4th International Conference on, vol., no., 6-9 July 2011, pp.1360-1365.
