DISTRIBUTED GENERATION PLANNING USING HYBRID OPTIMIZATION APPROACHES

Abstract

Objective of power system operation is to meet the energy demand economically and reliably. In the present environment, the justification for the large central-station plants is weakening because of economic, technical, and environmental concerns. In coming years, Distributed Generation (DG), a term commonly used for small-scale generation, will meet a large portion of electrical energy demand. As the penetration of distributed generation is increasing in the distribution network, it is no more passive in nature. Therefore, it is in the best interest of all the players involved to allocate them in an optimal way such that it could reduce system losses, improve the voltage profile, increase reliability, and reduce overall cost. Hence, the basic aim of this thesis is to develop efficient, economic and environment friendly methodologies for DG planning. Moreover, the thesis intends to propose hybrid optimization algorithms to solve these problems faster, accurate and efficient manner. Distribution utilities always strive to reduce power loss in their systems. Therefore, distribution loss reduction has always been one of the important objectives of DG planning. In view of this, the first contribution of this thesis is an integrated MINLP based approach for optimal placement of single and multiple DG units for loss minimization. To reduce the computational burden, two-tier model is proposed. Firstly, in Siting Planning Model (SPM), prospective candidate buses are shortlisted based on Combined Loss Sensitivity (CLS). This short-list of potential candidate buses is then passed to Capacity Planning Model (CPM). In CPM, the optimal locations and DG sizes are computed using MINLP based formulation. In this formulation, Sequential Quadratic Programming (SQP) and Branch and Bound (BAB) algorithms are integrated to handle discrete and continuous variables. This approach gives improved computational performance, strong convergence property, less solution time. It is observed that the proposed algorithm based on MINLP gives optimal solution due to its property of simultaneous placement of multiple DG units. The literature published in the last one decade has suggested many heuristic algorithms to solve the optimization problems of DG placement. These techniques are derivative free and simple to implement. However, they need several iterations to ensure converged solution and become computationally intensive. Convergence also depends on proper selection of tuning parameters. To ii overcome these difficulties, a hybrid optimization technique integrating Improved Harmony Search (IHS) and Optimal Power Flow (OPF) is presented for DG placement to minimize losses. The proposed formulation with few controlling parameters and embedded OPF shows strong convergence property and improved computational performance. The published literature reveals that environmental regulations, national policy of incentive and penalty for harmful emission plays a significant role in optimal DG planning. In India, National Action Plan on Climatic Change has set an ambitious target of 15% by 2020 for Renewable Purchase Obligation (RPO). To encourage renewable power generation as well as to meet RPO targets, a novel methodology to minimize annual cost, with Emission Offset Incentive (EOI), Generation Based Incentive (GBI) and penalty for carbon emission, is presented. The annual cost comprises DG capital, operation, maintenance, energy loss, grid energy and emission cost. Optimal solutions for different incentive schemes in terms of size, location, and types of DG are obtained. It is concluded that the appropriate incentive scheme can make cost intensive DGs such as SPV and wind viable. Furthermore, to solve the proposed formulation efficiently, a hybrid optimization approach is proposed by integrating Improved Harmony Search (IHS), and Teaching and Learning Based Optimization (TLBO). In DG planning with dispatchable DGs, peak load planning might lead to overestimation of DG size. Long term DG planning with multiple load levels may affect the optimal size, location, and time of adding new DG units at potential locations. Therefore, a novel formulation is proposed considering multi load levels to ensure system constraints within limits for all load conditions. It also considers simultaneous placement of DG and capacitor. It not only provides the optimal DG capacity, but also computes the optimal size for each load level and planning year. The increased DG penetration with proper planning methodology and efficient optimization algorithm would result in huge financial saving for utilities. This thesis has presented few such methodologies and hybrid optimization algorithms demonstrating their applications and usefulness.