Mapping the historical deployment of onshore wind power: Implications for energy system modeling

Licentiatavhandling, 2026

Onshore wind power features prominently in modeled low-carbon energy futures. In practice, however, wind power installations are constrained by factors such as local opposition and lengthy permitting procedures. These factors are not well represented by bottom-up energy models, which tend to apply constraints that lack empirical justification. This risks misrepresenting the role onshore wind power can play in a low-carbon energy transition.

To address these limitations, we adapt the outside view from the technology diffusion and project management literature, benchmarking wind power assumptions in large-scale energy system models against historical deployment patterns in countries and regions with substantial wind power installations. Such patterns represent what deployment densities and siting outcomes have proved achievable given the range of drivers and constraints present across diverse settings, and can therefore be indicative of how wind power will be deployed in the future.

Paper A investigates the siting of turbines in relation to the regional wind speed distribution over time across 25 countries and regions, and compares it to the allocation of wind power by energy system models. The findings reveal that historical siting patterns in relation to wind speed are heterogeneous across regions but stable over time: year by year, turbines are sited at windier-than-average, though not necessarily the windiest, locations. Endogenous cost minimization in models allocates wind power to sites with higher wind speeds than observed in history, implying that such models may overstate wind power's cost-competitiveness.

Paper B examines how model results change when applying historical deployment patterns in the model instead of typical literature assumptions, drawing on the siting patterns from Paper A and on deployment densities observed at the municipal and county levels. Both parameters substantially increase system costs across diverse regions, and the empirical densities also reduce onshore wind generation considerably. These results indicate that prevailing modeling assumptions may overstate the contribution of onshore wind power and consequently understate the corresponding system cost.