Design optimization for reliability improvement in microgrids with wind – tidal – photovoltaic generation

• Published in: Electric power systems research 2020-11, Vol.188, p.106540, Article 106540
• Main Authors: Recalde, Angel A, Alvarez-Alvarado, Manuel S.
• Format: Article
• Language: English
• Subjects: Alternative energy; Atmospheric models; Auto-regressive moving average models; Component reliability; Design optimization; Distributed generation; Electricity distribution; Meteorological data; Modelling; Monte Carlo simulation; Network reliability; Optimal power system planning; Optimization; Performance indices; Reliability; Reliability analysis; Renewable energy; Renewable resources; Simulation; Smart grid technology; System reliability
• ISSN: 0378-7796; 1873-2046
• DOI: 10.1016/j.epsr.2020.106540

•A design-framework approach for the Monte-Carlo based PSO optimal reliability planning of a wind-PV-tidal RE DGs which are deployed throughout a distribution network based on size and location that maximizes reliability at low cost, i.e. EENS minimization.•Utilization of an integrated reliability assessment based on the system's state within the optimization method. The system's state calculation considers the events’ stochastic nature and includes the evaluation of the states of RE DG and components via probabilistic models, as well as time-dependent models for RE, grid supply, and load demands. Distributed generation is essential for smart distribution systems. This prospect mainly depends on the efficient use of renewable resources. Therefore, it is compelling to provide an optimum design framework that allows a thorough modeling to select the most convenient size, location, and renewable energy combination that maximize system reliability on a distribution network. This paper presents an optimization-based framework to design a distributed generation system by incorporating an optimal reliability assessment with wind, solar, and tidal energies. The network planning exercise considers the stochastic nature of the network's state by including time-series models and hourly-based analysis to accurately determine the reliability indexes. Historical meteorological data has been used to model and deploy a set of renewable energy distributed generators which maximize reliability in a 37-bus primary-distribution network. Due to the probabilistic modeling of the system's components, a Sequential Monte-Carlo simulation is used to manage reliability evaluation at the network level. Although large reductions on energy-not-supplied are expected, it is shown that cost and other performance indexes do not follow the same trend, and project selection requires some compromise between cost and performance.