Publication: Optimal planning of energy storage system for hybrid power system considering multi correlated input stochastic variables
No Thumbnail Available
Date
2024
Authors
ALAhmad A.K.
Verayiah R.
Ramasamy A.
Marsadek M.
Shareef H.
Journal Title
Journal ISSN
Volume Title
Publisher
Elsevier Ltd
Abstract
This paper formulates a mixed integer non-linear probabilistic optimization planning problem to determine the optimal location, power rating and capacity of compressed air energy storage system (CAES) for a hybrid power system that includes wind and photo-voltaic (PV) energy sources. The Quasi-Monte Carlo simulation (QMCS) method is adopted to generate multiple scenarios for a combination of wind, PV, load and electricity price uncertainties. Also, Cholesky decomposition is adopted to preserve the actual correlation coefficients among the generated input stochastic variables. Moreover, the QMCS method is combined with the probabilistic load flow (PLF) to track the actual output variables. Three constrained incompatible non-linear objective functions are to be minimized simultaneously including, the total expected planning and operation cost of all generation sources, total expected power losses and the total expected voltage deviation. This optimization problem is solved by the hybrid non-dominated sorting genetic algorithm (NSGAII) and the multi-objective particle swarm optimization (MOPSO). The IEEE 118-bus system is adopted as the large-scale testing system to assess the performance of the proposed methodology and the convergence capability of the hybrid algorithm in rejecting the disturbances in the system caused by the existence of 132 different input correlated stochastic variables. The simulation results show that utilizing bulk CAESs can decrease the dependency on the thermal generators by 15.0984 % and decrease the total investment and operation cost by 25.5026 % compared to the case without utilizing any ESS technology. Also, the hybrid NSGAII-MOPSO proved its capability to converge successfully and reject all the input disturbances which could affect its performance. Moreover, the results show that the voltage on each bus in all scenarios remains within the limits in the presence of large input disturbances. ? 2024 Elsevier Ltd
Description
Keywords
Compressed air , Compressed air energy storage , Electric energy storage , Electric loads , Electric power system planning , Genetic algorithms , Integer programming , Intelligent systems , Investments , Multiobjective optimization , Nonlinear programming , Operating costs , Particle swarm optimization (PSO) , Power quality , Pressure vessels , Screening , Stochastic systems , Compressed air energy storage system , Correlated stochastic variable , Multi objective particle swarm optimization , Multi-objective particle swarm optimization , Non-dominated sorting genetic algorithm (NSAGII) , Non-dominated sorting genetic algorithms , Optimal planning , Optimal planning problem , Planning problem , Probabilistic load flow , Quasi-Monte Carlo simulation , Stochastic variable , Storage systems , Monte Carlo methods