Multi-level optimisation models for transmission expansion planning under uncertainty

Abstract

The significant integration of renewable energy sources to electricity grids poses unprecedented challenges to power systems planning. Each of these challenges is implied by a particular circumstance faced by the system planner when devising the expansion plan, which should be tailored to address the needs and objectives of the system under consideration. Within this context, this thesis is dedicated to propose methodologies to address three timely situations that may arise when planning the expansion of the grid. In the first situation, we consider the case in which the system planner must meet established renewable penetration targets while complying with multiple deterministic security criteria. Renewable targets have been largely adopted as an important mechanism to foster the decarbonization of power systems. Hence, we propose a methodology that simultaneously identifies the optimal subset of candidate assets as well as renewable sites to be developed, while introducing the concept of compound GT n-K security criteria. In the second situation, we aim to minimize the regret of the system planner under generation expansion uncertainty. In many cases, e.g. the United Kingdom, the decision on the transmission expansion plan is taken by a market player that does not determine the future generation expansion. Within this context, we propose a 5-level MILP formulation to represent the minimization of the regret of the system planner in light of a set of credible scenarios of generation expansion while enforcing n-1 security criterion. Finally, in the third situation, the objective is to inform the optimal transmission expansion plan under ambiguity in the probability distribution of RES generation output. To do so, we present a methodology capable of determining the transmission plan under deterministic security criterion while accommodating a set of different probability distributions for RES output in order to integrate ambiguity aversion.