Flexible climate impact emulation of thermoelectric power plant cooling constraints and buildings energy demand in integrated assessment modelling.

Integrated assessment models have long been used for systemic energy policy design and assessment, but they remain limited when incorporating climate impact feedback typically resorting to discrete SSP-RCP combinations with limited flexibility to evaluate different emission trajectories. Where climate impacts are incorporated, they typically use sector-specific ad-hoc methods, making it difficult to distinguish substantive differences across impact channels from artifacts of implementation. This is especially important as the compound effects of climate impacts and their cascading consequences become more salient. Here we bring forward a standardized abstraction for flexible climate impact emulation which allows for easy extension suitable for a general class of integrated assessment models and climate impact drivers. Our novel contribution is via the use of the Rapid Impact Model Emulator (RIME) which allows the emulation of climate impacts based on global warming levels. In conjunction with simple climate model MAGICC we can emulate impacts for two climate impact channels: reductions in usable thermoelectric power plant capacity due to rising temperature and buildings energy demand changes via reduced heating demand and increased cooling demand under warming. These reflect supply and demand side climate impacts. Emulation spans emission projections from a granular range of full-century carbon budgets, reflecting the diversity in mitigation scenario outcomes and allows for quantifications of small temperature differences in system costs. In isolation, the reductions in thermoelectric plant capacity due to changes in hydroclimatic conditions cause a 20% reduction in freshwater-based cooling technologies as well as a global 2% reduction in coal energy between 1.7C and 2.7C warming scenarios.

However, the joint impact of both drivers influences the technological choices with increased adoption of renewable energy sources with 15 EJ less coal capacity than under the effect of increased energy demand alone, between the same warming levels. This is a consequence of cooling constraints limiting the scalability of thermoelectric powerplants in years where buildings energy demand rises most. The first-best model response then takes account of infrastructure lock-ins engendered and drives the overall energy system into a different path with less thermoelectric power generation across the time horizon. This demonstrates the potential and importance of considering climate impact drivers as well as establishing the viability of flexible impact emulation in Integrated Assessment Models.