Community-level decentralized energy system planning under uncertainty: A comparison of mathematical models for strategy development

Abstract

Distributed energy systems renewable energy are one solution to the environmental and economic concerns of energy use. While energy planning and optimization have been conducted mainly as a mathematical exercise, practical approaches that incorporate the engineering realities and uncertainties are limited. Decision makers find challenges in community energy planning due to the lack of expertise, planning tools, and information. While a multitude of models and tools are currently available, there are no means of identifying the most appropriate or accurate methods, especially considering uncertainty. The main objective of this study is to compare and identify the strengths and limitations of various mathematical modelling techniques used in energy planning for grid connected renewable energy systems. As a case study demonstration, different multi-objective optimization techniques with and without uncertainty consideration (i.e. robust optimization, linear optimization, Taguchi Orthogonal Array method, and Monte Carlo simulation) were applied on a selected neighborhood in British Columbia. The optimization outcomes and the time and effort for evaluation were compared for the different methods. The findings indicate that robust optimization can be used to develop an uncertainty-based decision model. It significantly reduces evaluation time compared to the other methods. Although the presence of uncertainties can change the optimal configuration of a planned energy system, the assessment method itself does not significantly impact the outcomes. The findings of this study will enable the energy planners and researchers to compare different multi-objective optimization techniques, and to select the best for planning renewable energy projects, especially during the pre-project planning stage.