Reliability Constrained Optimal Sizing of Renewable Energy Resources and Capacity Credit Evaluation

Abstract

Renewable energy sources (RES) have become of paramount significance in mitigating global gas emissions as they are now increasingly used instead of conventional generation. The growing penetration of RES is changing its role from supplementary to alternative energy resources. If not properly planned, this transformation can significantly increase uncertainty due to the intermittent and non-dispatchable nature of resources such as solar irradiation and wind speed, potentially jeopardizing the reliability of the power supply. Unlike RES, conventional power plants are dispatchable. Consequently, RES is usually considered an energy rather than a capacity source. Assigning capacity value to renewable energy sources (RES) is a challenge faced in planning the integration of these resources with the grid. The capacity credit (CC) analysis evaluates the system’s actual power output compared with a constant capacity generator, i.e., conventional generator and determines an effective capacity to use for planning and operation. In chapter 2 of this dissertation, a multi-objective approach is introduced to simultaneously optimize reliability and cost. Also, to deal with multiple types of RESs, a new concept, cost credit, is proposed as a supplement or alternative to capacity credit. Cost credit is a parameter that can be used to quantify the cost during planning and increase the reliability of the system. The over-all objective is to combine and size the RESs, i.e., photovoltaic (PV), wind turbine (WT), and battery energy storage system (BESS), to meet the customer demand based on the total cost and reliability of the system. Two optimization methods, multi-objective particle swarm optimization (MOPSO) and non-dominated sorting genetic algorithm (NSGA-II), are explored for grid connected and stand-alone systems. The best combined size that gives optimal reliability and cost, is then obtained from their outputs utilizing a Fuzzy technique. Then, capacity credit and cost credit are estimated for the obtained optimal solution. Finally, sensitivity analysis is conducted to examine the impact of changing different parameters, purchasing/selling price, capacity of grid-connected, and swept area of RES, on the system size. Chapter 3 presents different factors that could affect the CC of a system. Two methods are proposed to determine the CC, namely equivalent firm capacity (EFC) and effective load carrying capability (ELCC). Since these methods are based on satisfying reliability criteria, daily loss of load expectation (LOLE), hourly loss of load (LOLH), and expected energy not served (EENS) have been employed as indices. To obtain the CC value, both methods apply two techniques: traditional and optimization. Genetic algorithm (GA) is the optimization approach used. Then, the two techniques have been compared, and the superior performance of the optimization approach has been demonstrated. Two hybrid systems, stand-alone (SA) and grid-connected (GC) modes, are proposed and used as case studies. The hybrid systems contain photovoltaic (PV), wind turbine (WT), and battery energy storage system (BESS). In this chapter, three different scenarios, system as a whole, only wind, and no batteries, are adopted to test the CC. Finally, sensitivity analysis is carried out to examine the impact of adjusting the wind speed, solar irradiation, and load. It is illustrated that the choice of reliability index plays an important role in determining the capacity credit and it is shown that EENS is a more comprehensive and consistent index of reliability. Chapter 4 aims to interconnect the RES with a battery energy storage system (BESS) to assist the system’s balancing. In case BESS could not handle the balancing, the demand response (DR) has been regarded as a virtual power plant to mitigate loss of load events. Moreover, multi-objective particle swarm optimization (MOPSO) is introduced to optimize the reliability and cost of the combined RESs such as PV, WT, BESS, and DR. Stand-alone systems with and without DR have been explored in this work, and the optimal solution is obtained using fuzzy logic. Since system planning is carried out based on the size of the combined system, capacity credit (CC) analysis of the optimal solution is obtained using genetic algorithm optimization (GA). Two approaches to estimate the CC, equivalent firm capacity (EFC) and effective load carrying capability (ELCC), are proposed. Different reliability indices are employed, namely daily loss of load expectation (LOLE), hourly loss of load (LOLH), and expected energy not served (EENS), to examine their impact on the CC.