Hydrogen-based self-sustaining integrated renewable electricity network (HySIREN) using a supply-demand forecasting model and deep-learning algorithms

Research output: Contribution to journalJournal article – Annual report year: 2019Researchpeer-review

View graph of relations

Electricity generation from renewable resources such as wind and solar energy inevitably involve intermittency due to the variable nature of wind speed and solar radiation. In this study, a mathematical model of a hydrogen-based self-sustaining integrated renewable electricity network (HySIREN) employing a supply-demand forecasting model and deep-learning (DL) algorithms is developed. The proposed model is implemented as follows: an empirical model decomposition is applied to decompose historical renewable electricity supply-demand data into a number of sub-layers; DL models are utilized to predict renewable electricity supply-demand patterns using the disclosed sub-layers; the predicted surplus renewable electricity and the predicted renewable electricity shortage are explicitly divided through a comparison of forecasting renewable electricity supply-demand data; and according to the results from forecasting models, a smart hydrogen balance is designed by an integrated hydrogen production process encompassing the steam methane reforming process and electrolyzers. Finally, a self-sustaining energy system is constructed and the system flexibility is enhanced, where the predicted surplus renewable electricity is used to convert produced and stored hydrogen into electricity to satisfy predicted renewable electricity shortages. The suggested model was validated by a case study of Jeju Island in the Republic of Korea and the feasibility of the HySIREN model was evaluated. Approximately 64.5% of the total environmental costs were eliminated. The results of this study suggest it would be beneficial to construct environmentally benign strategies for self-sustaining energy systems based on renewable resources.
Original languageEnglish
JournalEnergy Conversion and Management
Volume185
Pages (from-to)353-367
ISSN0196-8904
DOIs
Publication statusPublished - 2019
CitationsWeb of Science® Times Cited: No match on DOI

    Research areas

  • Renewable electricity forecasting, Deep-learning algorithms, Integrated hydrogen production, Optimization, Jeju Island

ID: 168428856