Power system network expansion planning using hybrid heuristic method

Abstract

Transmission network expansion planning (TNEP) is a large scale, mixed integer, complex, non-linear and non-convex optimization problem. Its main focus is to find the optimal structure and least cost transmission investment alternatives of the forecasted load and generation configuration. In this paper, transmission network expansion which focuses on alleviation of transmission line congestions in the considered base topology network is proposed. The proposed methodology is based on sensitivity analysis where by the moment the thermal rating of a particular transmission line (existing or candidate ) is violated then an expansion is inevitable. Varieties of classical as well as heuristic algorithms can be employed to solve the network expansion problem. In this paper the hybrid heuristic method is considered. This is a combination of the forward and backward heuristic methods. The expansion plan will be done chronologically starting with the backward stage for the normal conditions and then the forward approach is applied for contingency conditions analysis. For all this the main aim is to minimize the total investment cost, but at the same time ensuring that the network is robust and stable under normal and contingency conditions. This expansion problem which optimize the total investment and operation cost is modeled using a multi-stage decision framework where by the 1st stage will be for the expansion of the network with the connected known loads and power generation and the 2nd and final stage will be the expansion of the network with the forecasted load (assuming 120% increase in load in next 10years). In this, the transmission expansion planning the location, type and number of extra transmission lines of the optimal network configuration are determined. For illustration purpose the resulting mixedinteger nonlinear programming problem tackled under hybrid heuristic method is developed and applied on the IEEE 30 bus test power system. The proposed model is implemented in Mat lab software.