Andreas V. Olympios, Fanourios Kourougianni, Stefanos Aristodemou, Alexandros Arsalis, Charalampos Konstantinou, Andreas Kyprianou, Panos Papanastasiou, George E. Georghiou
This study develops a stochastic, risk-aware capacity-expansion and hourly-dispatch model for residential districts, co-optimising photovoltaics, batteries, heat pumps, and a hydrogen chain (electrolyser, storage tank, fuel cell) under electricity price uncertainty. The approach integrates a whole-energy system model with a bi-objective formulation that jointly minimises expected system cost and downside risk, quantified using Conditional Value-at-Risk (CVaR). Electricity prices are represented using a mean-reverting stochastic process with a time-varying mean, enabling representation of risk preferences in long-term planning. The framework is demonstrated through a case study of a residential district comprising 50 dwellings in Nicosia, Cyprus. The results show that, from a purely cost-minimising and risk-prone perspective, hydrogen is not selected. However, risk-averse planning reduces worst-case system cost (CVaR) from 1.34 to 1.29 M€ with only a marginal increase in expected cost (1.06 to 1.07 M€), and favours the adoption of green hydrogen as a complement to batteries and PV. When hydrogen is adopted, self-sufficiency increases from 78.0 % to 84.5 %. Sensitivity analysis shows the investment-cost and risk-preference ranges under which hydrogen becomes part of the optimal portfolio. The primary contribution is a transferable methodological framework that explicitly incorporates electricity price uncertainty into district energy system planning for assessing emerging energy storage technologies under uncertain future conditions.