Amin, Amin, Petri, Ioan ![]() ![]() |
Abstract
The transition to local renewable energy communities offers a promising route to decarbonisation, energy security, and system decentralisation. Achieving high renewable shares necessitates highly integrated flexibility through combining multiple energy carriers and storage systems to manage variability and ensure reliability. This study explores the technical, economic, and environmental performance of a hybrid power-to-gas (P2G) system in a UK rural community, incorporating wind and solar generation, battery storage, and hydrogen production. A machine learning-based framework was developed for forecasting key energy system determinants. Among four approaches tested, artificial neural networks (ANN) and random forest (RF) demonstrated high accuracy, with ANN elected for real-time operation due to lower computational requirements. System simulations indicated that 92 % of the community’s annual electricity demand could be supplied by renewables, yielding average electricity cost savings of 54.3 %. Hydrogen blending at a 20 % volume-based scenario reduced gas demand by 6.3 %, while surplus hydrogen produced offered additional revenue potential. Over a 25-year lifetime, total revenues were approximately equal to the capital investment, with a 15-year payback period. These findings highlight the potential of hybrid P2G systems to support net-zero targets at the community level, considering supportive policy and system integration.
Item Type: | Article |
---|---|
Date Type: | Published Online |
Status: | In Press |
Schools: | Schools > Engineering |
Additional Information: | License information from Publisher: LICENSE 1: URL: http://creativecommons.org/licenses/by/4.0/, Start Date: 2025-07-28 |
Publisher: | Elsevier |
ISSN: | 1364-0321 |
Date of Acceptance: | 28 July 2025 |
Last Modified: | 08 Sep 2025 15:46 |
URI: | https://orca.cardiff.ac.uk/id/eprint/180988 |
Actions (repository staff only)
![]() |
Edit Item |