Randomized discrepancy bounded local search for transmission expansion planning

Abstract

Abstract-In recent years the transmission network expansion planning problem (TNEP) has become increasingly complex. As the TNEP is a non-linear and non-convex optimization problem, researchers have traditionally focused on approximate models of power flows to solve the TNEP. Existing approaches are often tightly coupled to the approximation choice. Until recently these approximations have produced results that are straight-forward to adapt to the more complex (real) problem. However, the power grid is evolving towards a state where the adaptations are no longer easy (e.g. large amounts of limited control, renewable generation) and necessitates new approaches. Recent work on deterministic Discrepancy Bounded Local Search (DBLS) has shown it to be quite effective in addressing this question. DBLS encapsulates the complexity of power flow modeling in a black box that may be queried for information about the quality of proposed expansions. In this paper, we propose a randomization strategy that builds on DBLS and dramatically increases the computational efficiency of the algorithm.

Grid expansion planning for carbon emissions reduction

“…To solve this problem we adopt the Randomized Constructive Heuristic (RCH) algorithm of [15] and for completeness it is redescribed here. This algorithm is used due to its ability to accommodate arbitrary complex models of S, its ability to generalize existing PGRP constructive heuristics and the high quality results it has achieved on other models of expansion planning [13], [15].…”

“…This algorithm is used due to its ability to accommodate arbitrary complex models of S, its ability to generalize existing PGRP constructive heuristics and the high quality results it has achieved on other models of expansion planning [13], [15]. The algorithm is discussed in Figure 1.…”

“…The RCH procedure is initialized with the best solution discovered and repeated until an improving solution is no longer found (line 9). In this paper, the function ORDER-VARIABLES is implemented using the constructive heuristic of [16], [17], [25], which has worked well in practice within the RCH implementation of [15]. More formally the function is defined by:…”

“…In this paper we adopt the Simulation-Optimization algorithm of [13], [15] to address this problem. The key idea of the approach is the encapsulation of the power model within a simulation black box.…”

“…• A generalization of RCHs to solve multi-scenario grid expansion planning problems • Case studies demonstrating that expansion planning can be used to achieve near best case carbon reductions when adding renewables • Case studies showing that adding renewable generation solely based on expansion cost can have unexpectedly low reductions in carbon output • A demonstration that the simulation-optimization approach of [15] can handle optimal power flow models, as was conjectured The key point of this paper is to assume renewable generation expansion decisions are pre-determined: the challenge is how to deploy them to maximize carbon reductions. Literature Review While literature on Power Grid Expansion Planning (PGRP) goes back many decades, it has only been recently, with a few exceptions [18], that there has been interest in how to use planning to reduce carbon emissions [9], [14], [19].…”