Vertical integration in energy system modelling Understanding the effects of European energy trading on the Swedish energy system

Abstract

Understanding a country’s future energy system requires modelling, yet national energy system models typically treat countries as isolated systems, ignoring cross-border interdependencies. Fully resolving this limitation is computationally intractable, as adequately representing neighbouring countries requires their neighbours in turn, rapidly producing models too large to solve. This project addresses this by developing a vertical integration methodology that couples a national sub-model with a surrounding macro-level model. Cross-border trade flows are extracted from the macro-level model and applied as boundary conditions in the sub-model, allowing the national system to respond to external energy market pressures without requiring a fully resolved continental model. The methodology is implemented within the GENeSYS-MOD framework and evaluated through a Swedish case study, using a newly constructed dataset divided into Sweden’s four electricity bidding zones. The results demonstrate that vertical integration fundamentally reshapes Sweden’s optimal energy configuration in ways an isolated model cannot capture. Hydrogen export demand to Finland emerges as the dominant driver, cascading through the power system and redirecting electricity away from domestic industry, slowing its electrification. Access to imported natural gas further reinforces this pattern, as fossil-fired heating becomes more competitive than electric alternatives. Although net emissions remain comparable between the models, the vertically integrated model shifts toward a greater dependency on carbon capture technologies rather than emissions reduction at source. The sub-model also exposed a spatial resolution mismatch: trade patterns optimised at the European level can produce infrastructure investments that are impractical at the national scale. The study highlights that what is optimal from a European perspective may not align with national climate goals, a tension that isolated national models cannot reveal. Addressing this through more sophisticated vertical integration methodologies represents an important direction for future research.