Digital Twin of Renewable Energy-Linked Power-to-Gas (P2G) Systems: Model Building and Continuous Update for Hydrogen Producing Electrolysis

Solar energy and wind energy, which are widely used renewable energies, are highly volatile and intermittent, and long-term large-capacity electricity storage is required to maintain the stability of the power grid. Power-to-Gas (P2G) technology is used to generate and store gas from surplus electricity produced from renewable energy. Most P2G systems use electrolysis to produce hydrogen: the produced hydrogen is stored in a pressure tank and can be converted back to electricity through a fuel cell or hydrogen combustion engine when needed.

Hydrogen has a wide flammable range and low ignition energy compared to existing main energy sources such as gasoline or natural gas, so it has a high risk of fire and explosion, and is a material with high risk of emitting high combustion heat in case of explosion. Therefore, it is essential to evaluate the risk of hydrogen processing and storage equipment. The stability of the P2G system is an important factor in system operation along with safety. In particular, when the introduction rate of renewable energy increases, it is essential to operate a stable P2G system in order to process more excess energy.

This study aims to build a digital twin for a 500kW alkaline water electrolysis system to be built in the Saemangeum Renewable Energy National Demonstration Research Complex in Korea. We improve the digital twin model of the water electrolysis system using real-time sensing data enhanced by the real-time continuous model update. Though the current digital twin technology is focused on design and operation, we expand the applicability of digital twin by implementing functionalities for safe operation and safety management, including dynamic risk assessment, monitoring and fault detection.

In this study, based on the first principle-based models of alkaline water electrolysis cells, a digital twin of unit cells is developed, and a digital twin for the entire water electrolysis system consisting of a stack of these cells is developed. The core models consisting of the digital twin are as follows: 1) thermodynamic model for calculating Gibbs energy related to the reversible voltage, 2) electrochemical model for calculating the cell voltage according to the temperature, pressure and current, 3) electrolysis temperature calculation model, and 4) hydrogen production performance degradation model. The system/physics and failure/degradation models are composed of the above models, and operation data are combined to tune parameters of models in real-time, which establishes a digital twin of the alkaline water electrolysis system.

The developed digital twin is validated for the entire P2G system along with each of the cell units, alkaline water electrolysis system, and hydrogen compression and storage system. As a result, the developed digital twin is differentiated from the existing operation support system as a unified model for optimum control and scheduling, monitoring and fault detection, and dynamic risk assessment of the electrolysis system. When electricity production changes due to weather fluctuations, the development system enables quick start-up and system stabilization based on more accurate prediction and simulation, and dynamic risk management through prediction of functional degradation and damage of the entire system over time. It will be expected to enable real-time optimal operation and accident prevention of the alkaline water electrolysis system across the entire lifespan.