Reliability Evaluation Improvement of a Composite Power System Based on Artificial Intelligence Algorithms – Case study: Composite Power System in Latakia Governorate

Abstract

Reliability assessment is an integral part of the planning, designing, and operating studies of any electrical power system. Particularly, the reliability evaluation of the composite power system (generation and transmission) is very essential to assess the adequacy and security of the system to supply high-quality power to the consumers. The most difficult problems encountered in assessing the reliability of a composite power system are the large scale and complexity of this system and thus the time and computational burden required for such assessment. This article presents a proposed methodology for the reliability evaluation of the composite power system in Latakia governorate based on artificial intelligence algorithms. According to the proposed methodology, firstly, the daily, weekly, and seasonal peak loads are calculated based on the annual hourly load and their curves are plotted as well. Next, the forced outage rate of the generating units and the failure probability of the transmission line in this system are calculated based on the mean time between failure and mean time to repair. Later on, two artificial intelligence algorithms, namely, the differential evolution (DE) and the genetic algorithm (GA) are applied to improve the reliability evaluation of the composite power system from the time and computational point of view. DE algorithm was applied in the first stage, to generate the different possible state samples of the system in order to determine the failure states. In the second stage, and for each failure state determined by DE, GA algorithm is called to find the optimal minimum load curtailment without violating system operational conditions (voltage levels and maximum line capacity). The proposed load curtailment strategy in this work is based on curtailing more load form less importance loads compared to less load form high importance loads. Lastly, based on the failure states, the reliability indices for the whole system and for the load buses are calculated. To verify and validate the proposed methodology, it was applied to the composite power system in Latakia governorate. The obtained results showed that the proposed method is more accurate and much efficient in calculating the annualized reliability indices of the studied system compared to Monte Carlo simulation method. The results, also, showed that the proposed load curtailment strategy is highly efficient in maintaining the system operational conditions.