•Study of hydrogen energy storage for microgrid using a reversible solid oxide cell.•Optimisation is used to choose and size components for simulated microgrid.•Case studies are considered for England and Texas.•Economics for such a system are challenging: payback generally at least 20 years.•Hybrid storage (hydrogen + battery) preferred to battery for high self sufficiency. Reversible solid oxide cells (rSOCs) offer the prospect of long term bulk energy storage using hydrogen or methane fuel. Solid oxide technology, whilst less mature than alkaline and PEM technology, offers superior conversion efficiency - especially for electrolysis. Furthermore, the possibility of using the cells reversibly means that separate ‘power-to-gas’ and ‘gas-to-power’ components are not needed, potentially reducing costs. In this work, we consider the suitability of energy storage using rSOCs and/or battery storage for a microgrid consisting of houses equipped with solar PV generation. An agent-based simulation model is developed to assess the performance of such a microgrid. The model enables the microgrid’s self-sufficiency to be quantified, and hence the possible cost savings through avoided imports of grid power. Sizing of microgrid components is optimised to determine the most cost-effective design capable of achieving given self-sufficiency ratio. Case studies are considered for England and Texas. Initially, designs are considered with hydrogen energy storage only; subsequently, hybrid energy storage is considered, with a community scale battery working alongside the rSOC. Results suggest that payback periods for pure rSOC systems tend to be unfavourable. However, if prices fall to levels foreseen in the literature, a system designed to achieve 50% grid-independence could pay back its investment costs within 20 years. Systems designed for Texas need relatively less storage, owing to the good year-round solar resource; as such, payback time in Texas is superior to the UK. Hybrid storage with battery + rSOC is found to be preferable to battery only systems when (i) high SSR is required and (ii) large over-capacity of PV generation is not possible.