GWO Based Optimal Reactive Power Coordination of DFIG, ULTC and Capacitors

Indonesian Journal of Electrical Engineering and Computer Science Vol. 11, No. 3, September 2018, pp. 805~813 ISSN: 2502-4752, DOI: 10.11591/ ijeecs.v11.i3.pp805-813  805 GWO Based Optimal Reactive Power Coordination of DFIG, ULTC and Capacitors 1,3 Mogaligunta Sankaraiah1 , S. Suresh Reddy2 , M. Vijaya Kumar3 Electrical & Electronics Engineering, JNTUA, Ananthapuramu, 515002, India Electrical & Electronics Engineering, N.B.K.R.I.S.T, Vidyanagar, SPSR Nellore district, 524413, India 2 Article Info ABSTRACT Article history: Wind is available with free of cost anywhere in the world, this wind can be used for power generation due to many advantages. This attracts the researchers to work on wind power plants. The presence of wind power plants on distribution system causes major influence on voltage controlled devices (VCDs) in terms of life of the devices. Therefore, this paper proposes grey wolf optimization method (GWO) together with forecasted load one day in advance. VCDs are on load tap changer (ULTC) and capacitors (CS), there are two main objectives first one is curtail of distribution network (DN) loss and second one is curtailing of ULTC and CS switching‟s. Objectives are achieved by controlling the reactive power of DFIG in coordination with VCDs. The proposed method is planned and applied in M atlab/Simulink on 10KV practical system with DFIG located at different locations. To validate the efficacy of GWO, results are compared with conventional and dynamic programming methods without profane grid circumstances. Received Mar 6, 2018 Revised Apr 28, 2018 Accepted Jun 10, 2018 Keywords: Doubly Fed Induction Generator (DFIG) Grey Wolf Optimization Algorithm (GWO) On Load Tap Changer (ULTC) Copyright © 2018 Institute of Advanced Engineering and Science. All rights reserved. Corresponding Author: Mogaligunta Sankaraiah, Electrical & Electronics Engineering, JNTUA, Ananthapuramu, 515002, India. Email: 1. INTRODUCTION Today the entire world focusing on Distributed generation because of non availability of input sources for conventional power generating stations and too many advantages of distributed generation (DG). Wind power is one of the best sources in DG, this attracts the research people to work on this [1]. In [2-3], Co-Evolutionary particle swarm algorithm and Artificial immune system are proposed for optimal placement and sizing of DGs. DGs are affecting the voltage stability of distribut ion [4-5]. These papers focused only on optimal placement and impact on voltage stability in the presence of DGs. Generally these DGs are directly connected to distribution system, which influences the power loss and switching operations of ULTC and capacitors, therefore the useful life of these devices are decreasing [6]. In [7] VCDs (ULTC & CS), DG and automatic voltage regulator (AVR) are coordinated, which reported that because of DG the switching operations of devices (SODs) are greatly increased almos t more than three times as compared with without DG. SODs are increased more than two times, when VCDs and DG coordinated by SCADA system [8]. In [9-10] VCDs are coordinated using two different approaches, first one is dynamic programming and second one is combined voltage control. In all these methods DGs are not included while dis patching the reactive power. In [11], synchronous machine as a DG and this reactive power is coordinated in the presence of VCDs. In [12], an autonomous system is taken including DG and real power of DG is coordinated together with power loss by optimal power flow approach. In [13-17], coordination done by TRSQP method, asynchronous and synchronous generators coordinated together with VCDs by voltage control, adaptive and Journal homepage: http://iaescore.com/journals/index.php/ijeecs 806  ISSN: 2502-4752 dynamic programming approaches are used for coordination respectively. All these methods are giving more importance for dispatchable DGs and the importance given for non d ispatchable DGs are very small. The objectives of this paper are reduction of power loss and switching operations of VCDs in the presence of DFIG. This can achieve by coordinating the reactive power DFIG and VCDs.This paper proposes grey wolf optimizer algorithm for reactive power coordination of DFIG, ULTC and Shunt capacitors in order to reduce power loss and switching operations of ULTC and Shunt capacitors. 2. MATHEMATICAL MODELLING OF DFIG Mathematical modelling of DFIG is very important, which affects the output of DFIG and therefore losses and SODs. Input to DFIG is wind, which is not constan t throughout a day or hour, so, the output of DFIG also changes. In mathematical modelling a relation is developed between input and output in terms of probability density function (PDF). This PDF describes the availability of wind based on that we can est imate the output of DFIG [18]. In generally wind speed of wind farm nearly similar to weibull distribution for particular time at a particular location [19]. Now the PDF can be written as: PDF(vel)  SF  vel    SCF  SCF  SF 1  expvelSCF SF SF  vel    SCF   WPDF(vel)  1  exp   (1) (2)   In Equations 1 & 2, PDF vel , WPDF vel , SF , SCF , vel , exp ,denotes probability density function, weibull PDF, shape factor, scale factor, wind velocity and exponential respectively. Based on Equations 1 & 2 the output of DFIG is characterised into three parts based on wind velocity. If wind speed is below cut in speed and above cut off speed the output of DFIG is taken as „0‟. If wind speed is above cut in and below rated the output of DFIG is written as 0.5  AD  RRB 2  MPC  vel3 . In remaining cases output is written as RP . Where AD , RRB MPC and RP represents air density, rotor blade radius, maximum coefficient related to performance and rated power respectively. Figure 1 shows the power availability of DFIG with respect to speed [20-21]. Figure 2 shows single line diagram of system with distributed generation. Figure 1. DFIG output characteristic Figure 2. Single line diagram of system with distributed generation 3. PROBLEM FORMULATION There are two main objectives of this paper; first one is reduction of SODs and second one is system power loss reduction. The objective function is modelled as a multi objective function; Figure 2 is used for this purpose. Where E represents voltage, suffix 1 represents grid, suffix 2 and 3 indicates sending and receiving ends respectively, suffix DFIG indicates DG as a DFIG, P & Q indicates real power and reactive power respectively. RL And X L indicates line resistance and reactance, suffix 2C and 3C indicates capacitor at sending end and receiving end respectively. Indonesian J Elec Eng & Comp Sci, Vol. 11, No. 3, September 2018 : 805 – 813 Indonesian J Elec Eng & Co mp Sci  ISSN: 2502-4752 807 The first objective, power loss is shown in Equation 3 is written by taking receiving end voltage as a reference. Power loss of (...truncated)