A multi-objective inverse (MOIn) energy system modelling method for guiding early technology development

When including emerging technologies in energy system modelling, uncertainties are inevitable in early-development technical estimations, leading to uninformative/misleading results. Knowledge of the worst, yet still viable parameter values, without extensive parameter variation, is desirable, to reconcile energy modellers’ system-level and technology developers’ perspectives. We develop a novel inverse optimisation approach, turning traditional input parameters of energy system models into optimisation variables and, hence, outputs. This addition of variables leads to a multi-objective optimisation problem, which models trade-offs between multiple interests at system vs technology level. We employ the augmented epsilon-constraint method to solve the resulting multi-objective inverse (MOIn) optimisation problem. MOIn determines worst-case technical specification targets, guides development pathways, informs about market potentials and determines technology roles. We implement MOIn in the energy system modelling framework Backbone. We demonstrate our implementation for Carnot Batteries as emerging technology in the Central Western Europe power system, inverting the capital expenditure (CAPEX) parameter, estimated from 20 €/kW/a to 913 €/kW/a in literature. We find the maximum CAPEX at which CBs are still endogenously invested is low compared to these benchmarks. As round-trip efficiency (RTE) and energy-to-power (EtP) ratios decrease, CAPEX for CBs to remain competitive becomes stricter. Increasing EtP ratios more efficiently affects CAPEX and total system cost than increasing RTEs. Further results highlight the high sensitivity of CBs’ market potential. At low CAPEX, CBs dominate as long-term storage, but are replaced by hydrogen with rising CAPEX, shifting CBs towards mid-term storage. Overall, MOIn uniquely integrates energy system modelling with technical parameter analysis, fostering collaboration. • Novel inverse optimisation method to turn energy model input parameters into outputs. • Multi-objective optimisation to model trade-offs between different interest groups. • Implementation of methodology in open-source modelling framework Backbone. • Informing and guiding early product development from a market perspective. • Case study reveals Carnot Battery CAPEX limits and sensitivities.