| Abstract | The development of self-sustainable microgrids on a restructured power sector platform, with increasing rate of demand growth, motivated the research on new techniques for economic and efficient operation and/or management of the microgrid. Also, the awareness of environmental concerns and volatile oil prices have encouraged the use of renewable energy sources (RES) and related technologies in microgrids. Moreover, increased number of participants in the evolving power markets demand sharing of benefits and social welfare maximization. In consideration, this research work seeks to enhance the performance of energy management and scheduling in terms of 1) robustness to uncertainties 2) applicability in different market models and corresponding benefit maximization 3) variety of resources 4) computational efficiency and 5) Cost effectiveness. The detailed propositions are as follows. At the outset, an optimal power dispatch formulation is proposed for a distribution system operator (DSO) monopoly model by using stochastic weight trade-off particle swarm optimization (SWT-PSO) based modified backward-forward sweep (BFS) optimal power flow (OPF) method. The suggested approach reduces fuel cost, emission, node voltage deviations and settling time in overall implementation by means of renewable energy sources, grid power trade and prioritized load curtailment. The analysis also probes in to the benefits of online OPF over online economic dispatch (ED). The problem is then modified for a liberalized market model with two different strategies of the microgrid central controller (MGCC) viz. profit maximization and cost minimization, considering hourly bids, expected hourly profits and electric vehicles (EVs). Moreover, a multiagent benefit maximized scheduling objective is formulated and solved by using non-dominated sorting PSO (NSPSO) for prosumer consortium market to improve the individual benefits of agents viz. aggregator, GenCo and prosumers. Now, to analyze the effects of interval uncertainties of renewable energy and loads on optimizing microgrid benefits, a combination of interval arithmetic backward forward sweep (IA-BFS) and SWT-PSO is proposed. Further, to estimate sensitivities of active and reactive power uncertainties on power flow and cost intervals and to plan the spinning reserve accordingly, an efficient two stage stochastic optimal energy and reserve management approach using affine arithmetic (AA) is proposed. The first stage determines the power dispatch based on forecasts and the second stage uses an uncertainty versus reserve plot to dispatch the required spinning reserve from unused capacity of demand response, grid purchase and other dispatchable distributed energy resources (DERs). The solution is found to be better in terms of operational planning, real time computation and less conservative bounds of power flow & cost variables. But, since all the elements in the interval solutions obtained, are not significant in view of the probabilistic nature of statistical data, those elements which are significant with a desired confidence level are boxed using probability boxes (P-Boxes) to obtain further less conservative and more feasible intervals. Thus, the combined effect of interval & probabilistic uncertainties and sensitivities is addressed. In reality, there is an implicit cost incurred in the reserve market for compensating actual uncertainties. Financial risk of the aggregator can be manifested as this extra reserve cost. Thus, Sortino ratio, a tool used for investment portfolio optimization, is reformulated for microgrid power market so that the objective of aggregator becomes profit per unit risk. The higher the ratio for a dispatch, the better the dispatch performs (with respect to profit) relative to risk taken.
|