Soft-linking of improved spatiotemporal capacity expansion model with a power flow analysis for increased integration of renewable energy sources into interconnected archipelago

This present study offers a novel approach for the improvement of energy planning. This has become increasingly important as higher penetration of variable energy resources and increased interconnection between the different energy sectors require more detailed planning in terms of spatiotemporal modeling in comparison to the presently available approaches. In this study, we present a method that soft-linked the energy planning and power flow models, which enabled fast and reliable solving of optimization problems. A linear continuous optimization model was used for the energy system optimization and the non-linear problem for the power system analysis. The method is used to compare different energy planning scenarios; further, this also offers the possibility for implementation assessment of the proposed scenarios. The method was applied to interconnected islands for five different scenarios. It was determined that the detailed spatial approach resulted in 26.7% higher total system costs, 3.3 times lower battery capacity, and 14.9 MW higher renewable energy generation capacities installed than in the coarser spatial representation. Moreover, the results of the power flow model indicated that the highest voltage deviation was 16% higher than the nominal voltage level. This indicates the need for inclusion of implementation possibility assessments of energy planning scenarios.

Automated summary

This study introduces a method that soft-links energy planning and power flow models to compare different energy planning scenarios and assess implementation possibilities. Applied to interconnected islands, the method revealed that a detailed spatial approach can reduce battery capacity and increase renewable energy generation capacity, but also result in higher total system costs. Results from the power flow model showed that the highest voltage deviation was 16% higher than the nominal voltage level.