Data-driven, metaheuristic-based off-grid microgrid capacity planning optimisation and scenario analysis: Insights from a case study of Aotea-Great Barrier Island

Small privately-purchased off-grid renewable energy systems (RESs) are increasingly used for energy generation in remote areas. However, such privately-purchased stand-alone RESs are often unaffordable for households with lower incomes. While considerable attention has been devoted to a range of off-grid microgrid sizing methods, leveraging the potential of data-driven, artificial intelligence-based metaheuristic optimisation algorithms is less well-explored. Importantly, data-driven metaheuristics have the potential to produce the nearest solution to the globally optimum solution in microgrid sizing applications, which have been recognised as non-deterministic, polynomial time-hard (NP-hard) problems. Furthermore, there is a general lack of electrified transportation interventions considered during long-term grid-independent microgrid planning phases. In response, this paper introduces a novel metaheuristic-based strategic off-grid microgrid capacity planning optimisation model that is applicable to associated integrated energy and e-mobility resource plans. The formulated general off-grid microgrid sizing model is solved using a competitively selected state-of-the-art metaheuristic, namely moth-flame optimisation. To test the effectiveness of the proposed model, three independent microgrid development projects have been considered for three communities residing on Aotea-Great Barrier Island, namely Tryphena, Medlands, and Mulberry Grove. The sites of interest have different demand profiles and renewable energy potentials, with consequent changes in the technologies considered in the associate candidate pools.
Comment: Electricity Engineers' Association (EEA) Conference 2022, Hamilton, New Zealand, 19-21 September 2022