Stochastic generation scheduling with variable renewable generation: methods, applications, and future trends

Abstract

One of the most intricate optimisation problems in power systems is generation scheduling. It determines the schedule and dispatch of electrical power generation to meet the load demand under various technical and operating constraints. Generation scheduling is a vivid problem, particularly in recent years, due to the aggressive integration of renewable energy, with stochastic nature, into power grids. As the literature size has swollen substantially in the past several years, this study critically looks into uncertainty modelling and the formulation of various techniques that were implemented in stochastic optimisation-based generation scheduling. The strengths and weaknesses of existing methods are fully exposed. Market operation policies that significantly affect the scheduling of renewable energy generation in different timescales are elaborated. Potential applications and future trends in terms of modelling, computational performance, and incorporation of flexibility and resilience notions are thoroughly discussed.

On the dispatch of minigrids with large penetration levels of variable renewable energy

“…In this first case, the proposed model is applied to the IEEE 14-bus system without considering solar PV generation. The equations considered (1)- (16) and (18). In other words, equals zero in (18).…”

“…The equations considered (1)- (16) and (18). In other words, equals zero in (18). The total load demand and output power of each conventional generating unit are shown in Figure 2.…”

“…The data were taken from an online source of the University of Texas Solar Energy Laboratory from the National Renewable Energy Lab (NREL). The city is located at Latitude: 30.29° N and Longitude: 97.74° E [18]. The data are then fitted into a Beta PDF, as mentioned in an earlier section, which was found to provide the best fit of the given data.…”

“…Scheduling of the minigrid determines the hourly operation timetable of the committed DGs, including conventional generating units and RES, to satisfy the hourly load demand with the lowest operation cost over a 24-hour timespan [17]. To consider the uncertainty of RES in power system problems, stochastic optimization techniques are used; such as full stochastic programming, chance-constrained programming and robust optimization [18]. The traditional, and widely used, method to represent the uncertainty is scenario generation in full stochastic programming [19].…”

“…The traditional, and widely used, method to represent the uncertainty is scenario generation in full stochastic programming [19]. However, scenario generation techniques are computationally inefficient [18].…”