The variability of renewable energy sources, specifically for large scale integration into the electricity sector, has brought increased attention to the requirements for energy storage systems. To mitigate the intermittency of renewable energy, many different energy storage technologies, such as pumped hydro, compressed air energy storage, batteries, flow batteries, hydrogen energy storage, capacitors, and flywheels are available. Some of these energy storage systems (e.g., batteries, flywheels) are best suited for short-term storage and highly dynamic operation, while others (e.g., pumped hydro, flow batteries, hydrogen) can accomplish massive and seasonal storage of renewable electricity sources. The integration of high levels of renewable power will require the features of both of these types of energy storage systems. The current work focuses upon hydrogen energy storage as it may contribute to large capacity and seasonal energy storage. Studies show that hydrogen energy storage in areas like California, where the viability of very high levels of grid connected renewable energy generation is promising can highly increase the solar and wind market penetration. Hydrogen energy storage systems generate hydrogen from water using an electrolyzer dynamically powered by using inexpensive or otherwise curtailed renewable energy, storing the hydrogen, and subsequently using the hydrogen for various purposes (e.g., to produce electricity via a fuel cell for power generation or transportation applications).

In this work, the dynamics of hydrogen energy storage (HES) integrated with large-scale renewable power using the capabilities of the existing California natural gas infrastructure were investigated. The dynamics associated with the grid demand, renewable power, temperature, pressure, and HES capacity for one week of November and August were analyzed in detail while an aggregated analysis of the entire state for the whole year were simulated in MATLAB/Simulink.

First, the solar and wind resources, considering spatially resolved resource availability at each hour during the year, were analyzed to simulate large-scale implementation of combined solar and wind power resources sufficient to meet the entire annual California electric energy demand. At each hour the excess power from the renewable sources is directed to a model solid oxide electrolyzer (SOE) system to produce hydrogen. The hydrogen is pressurized through two stages of compressors and then stored in a model of the existing natural gas underground storage resources. When the grid demand is greater than the available renewable power, hydrogen can be dispatched from the storage resources to return power to the grid through a model solid oxide fuel cell (SOFC). In one scenario, the eleven California underground storage resources with a total working capacity of 9,579,527,832 m³ are dynamically filled and discharged to enable 100% renewable power delivery to the state for the entire year. This case required the introduction of solar and wind resources that are 11.16 times the existing capacity to meet California throughout the year. In a second simulation, constraints of hydrogen delivery to the storage resources through the natural gas system when it is produced next to the renewable energy resource (typically in the desert) are accomplished. It is found that the dynamics for transferring hydrogen to the underground storage resource through a long natural gas transmission pipeline and the dynamics associated with volume flow rate, pressure and temperature are reasonable for accomplishing HES at this local resource.