Herding, Robert; (2023) Optimal Operation of Energy Microgrids under Uncertainty. Doctoral thesis (Ph.D), UCL (University College London).

Text RHerding_PhDThesis_goesPublic.pdf - Accepted Version Access restricted to UCL open access staff until 1 January 2025. Download (3MB)

Abstract

Large-scale electrification of energy demand and growing shares of renewables in the grid generation mix pose significant challenges to power systems. Microgrids offer a decentralised alternative to traditional grid designs, and can improve reliability, resilience, environmental and economic aspects for participants they serve or the power system they may interact with. To harvest their potential, value streams must be created from operational management strategies for microgrids, e.g. bidding in energy markets, utilising flexibility and advanced control of microgrid resources. This thesis presents novel models and solution approaches for optimal bidding of grid-connected microgrids using stochastic programming and mixed-integer linear programming (MILP). A strategy for bidding in the day-ahead electricity market is proposed for a microgrid with power generator, photovoltaic (PV) system, battery and electricity demand. A novel two-stage stochastic MILP model incorporates an optimisation-based selection of bid prices for hourly bidding curves, while the number of breakpoints per curve strictly complies with market rules. Monte Carlo simulation demonstrates benefits over deterministic and stochastic bidding models from literature assuming uncertain electricity prices and PV power. The problem is then converted to multi-objective, with conditional value-at-risk (CVaR) serving as additional objective to control financial risk from microgrid bidding commitments. A solution procedure applies ε-constraint method and fast forward scenario reduction to obtain bidding solutions for joint cost and risk minimization in reasonable CPU times. Also, a novel stopping criterion for scenario reduction is suggested. Finally, virtual bidding is integrated into the optimisation model for microgrid bidding. Additional scenarios for the real-time market electricity price are generated, and constraints for implicit and explicit virtual bidding derived. Benefits of virtual bidding to mitigate financial risk or reduce cost are presented in computational examples, and the impact of upper bounds on virtually traded quantities is examined.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as CSVJSONXML

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item