Risk-Constrained Optimal Energy Management of Multi-Carrier Microgrids Under Uncertainty

Abstract

The introduction of multi-carrier microgrids (MCMGs) is a paradigm shift in the decentralised energy system, allowing for coordinated dispatch of coupled electrical, thermal, and natural gas subsystems. Despite this, the economic challenges of MCMGs are quite sensitive to multi-dimensional uncertainties such as the intermittent generation of renewable energy, the volatile market electricity prices, and the multi-energy load demands. Traditional deterministic or risk-neutral energy management systems optimise average performance, but are not designed to buffer microgrid operators from operational gaps and tail risks with high impact from low-probability events. In this paper, a new risk-constrained optimal energy management approach for multi-carrier microgrids is presented, which aims to minimise risk by employing a twostage stochastic mixed-integer linear programming (MILP) approach. The Conditional Value-at-Risk (CVaR) is incorporated as a primary operational objective function to capture and mitigate financial exposure during worst cases. Extensive uncertainty profiles are computed using Latin Hypercube Sampling (LHS), while being downscaled by a fast forward scenario-reduction technique. A modified IEEE 33-bus electrical network and a 6-node district heating system are used for the validation of the proposed model. It is shown that the proposed framework strategically sacrifices a marginal increase in the expected daily operating costs while offering a significant decrease in the tail-risk financial losses through tuning the risk-aversion parameter.