Planning and optimization of smart local energy systems for cost-effective decarbonization

Abstract

The energy transition in cold-climate regions calls for integrated and localized solutions that balance cost, emissions, self-sufficiency, and robustness. Current planning approaches for smart local energy systems (SLESs) often treat these factors in isolation, omit embodied emissions of technologies, or overlook operational uncertainty and variability. This dissertation presents an integrated methodological framework for SLES planning and addresses four research questions on (1) site potential, (2) multi-objective trade-offs, (3) enabling technologies and price thresholds, and (4) solution robustness in the Nordic context.

Methodologically, the framework combines linear multi-objective optimization, multi-criteria decision analysis, Monte Carlo simulation, and rolling horizon simulation, and incorporates cradle-to-gate emission factors based on life cycle assessment. The five case studies in Finland include three SLESs, one building, and one regional assessment.

First, infrastructure access was found to be the most important factor for SLES potential. Second, a cost-optimized SLES achieved 48% lower emissions and 17% lower costs than the reference system, with 46% self-sufficiency. Further emission reductions or self-sufficiency improvements beyond this cost-optimized system were highly cost-effective. Third, wind power, heat pumps, electric boilers, and thermal energy storage were central enablers for SLES performance, while battery energy storage was not viable at current prices. Last, the performance discrepancy caused by the perfect foresight assumption was high in scenarios that minimized emissions or maximized self-sufficiency. Negative effects of interannual variability were mitigated when multi-year data were used in optimization.

The research advances the development of affordable, sustainable, self-sufficient, and reliable SLESs. Importantly, the initial emission reductions and self-sufficiency gains can be very affordable, enabled by energy vector integration. The findings are likely transferable to cold‑climate regions with low grid emission factors and dynamic electricity tariffs.