Optimum placement and sizing of DGs considering average hourly variations of load

Highlights

• A new sensitivity index is proposed for optimal locations of DGs.

• Technical and economical factors are considered for optimal sizing of DGs using GA.

• Solar photovoltaic, biomass and wind power DGs are considered for day time operation.

• Biomass and wind power DGs are considered for night time operation.

• Annual saving is also computed due to placement of DGs in the distribution network.

Abstract

This paper presents a multiobjective technique for obtaining optimal sizing of Distributed Generation (DG) units considering both technical and economical factors of the distribution system. The technical factors include real power loss reduction, line load reduction and voltage profile improvement and the economical factors consider optimal DG investment cost. Three different Distributed Generation systems solar photovoltaic, biomass and wind system are considered for integration with the existing distribution system. Since solar photovoltaic system is not available at night time, only biomass and wind systems are operated and for day time operation all the three distribution generation systems are considered. A new sensitivity index based on voltage sensitivity and apparent load power is proposed for identification of optimal locations for DG placement. The optimum sizing of DG units operating at unity power factor and lagging power factor is obtained using GA for different load levels considering daily average hourly loading aiming at improving the technical performance of the distribution system with optimum investment on DG units. Simulation results are presented to show the advantage of the proposed methodology in terms of technical performance and annual economical savings of the distribution system.

Introduction

Distribution companies needs more efficient planning strategies in order to meet the forecasted load demand and to supply reliable power to the consumers. In the current deregulation environment the traditional expansion of additional substations to meet the load growth demand is not economical and utilities needs alternative technologies such as Distribution Generations (DGs). Due to the rapid depletion of fossil fuels and to meet the modern power system restructuring requirements, distributed generation became an immediate option. DG can be defined as “electric power generation within distribution networks or on the customer side of the network” [1]. Most commercially available DG technologies are wind power, solar photovoltaic, solar thermal systems, biomass and small scale hydro power. Distribution networks mainly designed for passive circuits to transport power from central power station to consumers and hence the rating and positioning of DGs to be installed should be considered with greater consideration. Appropriate location and sizing of DG units in distribution system improves the voltage profile, supplies peak load demand, minimizes line loadings and reactive power requirement, and reduces the system power losses. On the other hand improper allocation and higher penetration of DG units in the distribution system leads to increase in power losses and reduces the reliability of the power system. Therefore the incorporation of DG units in the power distribution system should be planned considering both technical and economical factors. The technical and economical challenges arise due to intermittent nature of renewable energy sources in the distribution system planning with DG incorporation should also be considered. The technical factors include the reduction of losses, line loadings and voltage profile improvement and the economical factors include optimal investment cost of DG units.

Several researchers proposed analytical and artificial intelligence based techniques for optimal allocation and sizing of DG units for distribution networks. In analytical methods many researchers derived mathematical expressions for obtaining optimal location and sizing of DG units. Wang and Nehrir [2] derived analytical expressions for optimal placement of DG with unity power factor for different types of load distribution along the radial feeder in power distribution systems to minimize the real power loss. Acharya et al. [3] developed analytical expressions for optimal location and sizing of DG for minimizing total real power losses in distribution systems. Gozel et al. [4] have proposed an analytical method based on the equivalent current injection without the necessity of Jacobian matrix for the determination of the optimum size and location of distributed generation to minimize total power losses. Lee and Park [5] developed analytical expressions based on generator power distribution factor for optimal allocation of DG and used Kalman filter algorithm for optimal sizing of DG. Hung et al. [6] have developed analytical expressions for determining optimal size of distributed generators supplying both real and reactive power for reducing real power losses in distribution systems. Aman et al. [7] have proposed analytical expressions for a novel power stability index based voltage sensitivity analysis for optimal location of DG and proposed a step by step iterative algorithm for optimal DG sizing to improve the voltage profile and minimize the losses. Many researchers have proposed heuristic methods based on iterative techniques and algorithms for optimum location and DG sizing for distribution systems. Celli et al. [8] have proposed a multiobjective cost function optimization methodology based on genetic algorithm and ∈-constrained method for optimal allocation and capacity of DG units into distribution networks. Keane and Malley [9] have proposed a linear programming based methodology to obtain optimal allocation and sizing of embedded generation considering technical constraints for accommodating maximum DG penetration on the distribution network. Atwa et al. [10] have proposed probabilistic generation load model based method for optimal sizing of distributed generation (DG) units in the distribution system to minimize annual energy loss considering various technical constraints. Abou et al. [11] presented genetic algorithm based optimization of multiobjective function approach to determine the optimal sitting and sizing of DG units with multi system constraints. Porker et al. [12] have proposed a software package based optimization model considering total power system and costs due to DG investment for optimal allocation and sizing of DG. Elnashar et al. [13] have proposed a heuristic optimization technique by considering the appropriate weighting factors of the performance indices for obtaining the optimum sitting and sizing of DG units. Moradi et al. [14] have used combined genetic algorithm and particle swarm optimization for optimal location and sizing of DG capacity considering system operation and security constraints in radial distribution systems. Sajjadi et al. [15] have proposed a memetic based optimization for simultaneous placement of DGs and shunt capacitors considering technical and economical factors for reducing active power demand, voltage profile improvement and reduction in power loss of distribution system. Shaaban et al. [16] have proposed a GA based multiobjective approach for optimal allocation of different types of DG units to minimize investment costs of system upgradation, cost of annual energy losses and to improve the reliability of the system. Prenc et al. [17] have also studied the DG power penetration based on average daily load and power production curves but in this method they have not considered any optimization technique. Kansal et al. [18] have used particle swarm optimization technique for the placement of DGs in a distribution network. Naik et al. [19] have proposed an analytical method for placement of DG units and shunt capacitors together for real power loss reduction in distribution system. Injeti and Kumar [20] have used simulated annealing technique to identify optimal access point and capacity of DG units. Murthy and Kumar [21] have studied the optimal allocation methods based on sensitivity approaches. Mistry and Roy [22] have studied enhancement of loading capability of a distribution network considering DGs. Hung et al. [23] have studied the optimal placement of DG units for minimizing energy loss by using an exact loss formula. Hamedi and Gandomkar [24] have computed energy not served index for every branch of distribution network to allocate DG resources for the improvement of power quality and reduction of real power loss. Ochoa et al. [25] have analyzed the performance comparison between static and time varying nature of both load and DG by using a multiobjective index with appropriate weighting factors. They have shown that the response of the distribution system is better when natural behavior of loads and generation is considered compared to the single scenario of maximum generation and maximum demand. Hedayati et al. [26] have used power flow continuation program based voltage sensitivity analysis for obtaining optimal allocation of DG units to improve voltage profile and reduce real power losses. Ettehadi et al. [27] have proposed voltage stability based DG placement method using continuous power flow algorithm and modal analysis. They also proposed a qualified load index to obtain a priority list of DG locations for overcoming reactive power shortages. Abri et al. [28] have proposed a method of locating and sizing DG units to improve the voltage profile and voltage stability margin of the network. Recently several authors have proposed artificial intelligent based techniques especially using genetic algorithm or swarm optimization techniques for obtaining optimum DG location and sizing for distribution systems to improve system performance and reliability. Aman et al. [29], [30] have proposed voltage stability maximization and loss minimization based approach using particle swarm optimization for optimum sitting and sizing of DG units, and maximum loadability based approach using hybrid particle swarm optimization for simultaneous placement and sizing of multiple DG units. Koutroulis et al. [31] have proposed a GA based methodology for obtaining optimum sizing of standalone photo voltaic and wind generation systems considering the objective of minimizing 20 year round total system cost which is equal to the sum of the total capital and maintenance costs. Zou et al. [32] have proposed an optimization framework based on TRIBE PSO and ordinal optimization for distribution system planning incorporating DG reactive power capabilities and uncertainties. Kayal and Chanda [33] have proposed a PSO based multiobjective approach for the placement of wind and photovoltaic systems for reduction of real power loss and voltage stability improvement of distribution systems. Hung et al. [34] have proposed analytical approaches for optimal sizing and power factors of DG units and a methodology for identifying best location for placement of DG units considering energy loss minimization. Arefifar et al. [35] have proposed a methodology for optimal sizing and allocation of DG units considering the probabilistic nature of renewable energy sources and the hourly variations in load profile in order to minimize the total annual energy losses for designing microgrids with optimum supply adequacy. Pisica and Eremia [36] have proposed GA based approach for optimum sizing and location of DG units in distribution systems.

Above literature survey shows that many researchers have made attempt to optimize the DG power penetration into the distribution network using analytical and artificial intelligence techniques for improving the performance and to reduce the burden on the distribution system [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36]. So far most of the researchers aimed at optimum DG sizing problem considering either only technical factors [10], [11], [29], [33], [43] or only economical cost factors [31], [32] for peak load operation of the distribution systems. The impact of DG penetration on technical and cost factors is discussed in [44]. The optimal DG sizing obtained considering only technical factors may lead to higher DG investments with slight advantage in technical performance improvement and may become financial burden to the utilities, and doesn’t suggest any choice of economically available DG units installation. The optimum DG sizing obtained considering only economical cost factors may not fulfill the technical performance requirements for the current and future load demand. In the present work the authors aimed at optimum DG sizing problem to improve the technical performance of the distribution system with optimum investment on DG units installation.

In this paper a, a new sensitivity index based on voltage sensitivity and apparent load power is used for identifying optimal nodes for DG placement and optimum sizing of DG is considered for various load levels considering average hourly load variations. A GA based multiobjective function considering both technical factors and economical factors is proposed to obtain optimum DG sizing with minimum investment costs and maximum technical benefits. In addition to that the annual economical savings due to the incorporation of DG units is also determined.