Cost uncertainties and ecological impacts drive tradeoffs between electrical system decarbonization pathways in New England, U.S.A.

Advocates, policymakers, and researchers seek characterizations of tradeoffs from decarbonization pathways beyond those identified by grid operators or emerging from optimization models and robust quantification of uncertainty. We develop and apply to New England, U.S.A. an hourly-scale probabilistic model that accepts portfolio decisions and hourly demand as inputs and simulates direct and indirect (greenhouse gas and air pollutant emission) costs through 2050. The model supports diverse storage and generation technologies diverse accounting methods for Canadian hydropower, supports integration with models of ecological and other impacts, and draws on nationally available databases. In New England, direct monetary costs are minimized by new natural gas, but savings are vastly outweighed by excess damages from greenhouse gases compared to decarbonized pathways. Very different portfolios have similar monetized mean expected costs ($331 billion to $350 billion at 2% discount rate) but very different uncertainties (e.g., 90% CI ~ 23% of mean costs for new natural gas vs. 69% for small modular nuclear). Likewise, land use varies from negligible to 9,890 km2. Tracking uncertainties correlated across scenarios improves decision support (e.g., >90% confidence that constraining transmission with Canada represents ≥ $1.9 billion in added costs despite overlapping 90% CIs for absolute costs).