Compared to other optimisation frameworks, oemof.solph is typically used for energy system models with a higher level of physical detail. This recent study, implementing nonlinear properties and phase changes of seasonal ice storage is a perfect example. Those, who regularly attend our user meetings were able to observe the development from the beginning. We are happy that we also learned a lot about the technical aspects of combining complicated constraints and time series aggregation. Those insights will definitely be considered when polishing the time series aggregation support in our software.
🚨 New paper alert Balancing renewable energy supply over an entire year can be challenging due to heating and cooling needs. Seasonal cooling is one way to solve the problem and ice storage is emerging as a surprisingly effective option. Operating these systems and their technical details efficiently though can become a heavily computational task. In a recent article in Energies, Maximilian Hillen, Dr. Patrik Schönfeldt, Dr. Philip Groesdonk and Prof. Dr. Bernhard Hoffschmidt tackle this question. How can we optimise the operation of seasonal ice-storage systems in a way that is both detailed and computationally feasible for real-world planning? Their central result is an improved optimisation scheme implemented using #oemof that speeds up calculations while keeping results highly accurate. Applied to a business park in Germany, their approach reduced computation time by up to 80%, with only about 2.5% deviation in the main optimisation goal and around 9% in the Seasonal Energy Efficiency Ratio (a metric for overall efficiency across the season). Methodologically, the team uses Mixed-Integer Linear Programming (MILP), a standard mathematical optimisation technique, together with timeseries aggregation (TSA). TSA condenses long, high-resolution data (like hourly energy demand over a year) into a limited number of “typical periods”, preserving key patterns while cutting complexity. Their model captures the physical behaviour of ice storage in detail, including different temperature levels, the transition between sensible and latent heat (freezing/melting), and realistic charging and discharging efficiencies. Why does this matter? Because seasonal storage will be critical to matching variable renewable supply with real demand, especially for heating and cooling. Tools that are both accurate and fast to compute are essential if planners and engineers are to design robust systems at scale. Interested in the details? Read the full article here: https://s.dlr.de/9fsb4 What benefits or challenges do you see in focusing more on uncertainty when evaluating new green technologies and their optimisation tools? oemof association