Novel modeline methods and evaluation frameworks for assessing and improving the reliability of renewable-rich power systems

Abstract

The share of non-conventional renewable energy, such as wind and solar, is gradu­ ally increasing in many power systems. However, integrating wind and solar power on a large scale would considerably affect the reliability of power systems due to the stochastic nature of renewables. Conventional reliability evaluation methods cannot efficiently and accurately quantify power system reliability when there is a significant amount of renewable power in the system. Therefore, novel reliability evaluation tech­ niques are required to assess the reliability of modern renewable-rich power systems. This work implements novel modeling methodologies and evaluation frameworks to quantify the reliability of renewable-rich power systems accurately. Different algo­ rithms and techniques are developed to evaluate the reliability of generation, compos­ ite generation and transmission, and distribution subsystems because the nature of the problem and its complexity are different in each of the three subsystems. Firstly, a novel method based on Kernel Density Estimation (KDE) is proposed to model inter­ mittency and both diurnal and seasonal variations of wind and solar power generation using historical renewable power generation data. The proposed KDE-based renewable power models are used with Non-Sequential Monte Carlo Simulation (NSMCS) to evaluate the generation system adequacy of the Institute of Electrical and Electronics Engineers (IEEE) Reliability Test System (RTS) . The diurnal and seasonal varia­ tions of renewables and the correlation between the load and renewable generation are modeled in the proposed renewable power models. Secondly, the applicability of con­ ventional MCS for composite system adequacy evaluation is investigated. It is found that a more computationally efficient reliability evaluation methodology is needed to evaluate the adequacy of composite systems. Hence, a novel population-based intelli­ gent search method called Evolutionary Swarm Algorithm (ESA) incorporated with DC optimal power flow analysis is proposed to evaluate the adequacy of renewable­ rich composite power systems. The main objective of the ESA is to find out the most probable system failure events that significantly affect the adequacy of composite sys­ tems. The identified system failure events can be directly used to estimate the system adequacy indices. The random search guiding mechanism of the ESA is based on the underlying philosophies of genetic algorithms and binary particle swarm optimization. The computational efficiency of the proposed ESA is significantly higher than that of traditional simulation methods. Thirdly, the proposed ESA is used to evaluate the reli­ ability of solar-integrated power distribution systems. AC optimal power flow analysis is used to assess the system failure events while considering the feeder voltage levels. The annual reliability indices of renewable-rich distribution systems can be estimated using this method, which is not addressed in the prevailing literature. Apart from eval­ uating the power system reliability, the impact of increasing wind and solar integration on the reliability of power systems is investigated at all hierarchical levels. Further, the proposed reliability evaluation frameworks are used to identify possible methods to improve the reliability of power systems. The outcomes of this research allow precise planning and operation of modern power systems in a time-efficient manner.