Electric Sector Capacity Planning under Uncertainty: Climate Policy and Shale Gas in the US∗

This research investigates the dynamics of capacity planning and dispatch in the US electric power sector under a range of technological, economic, and policy-related uncertainties. Using a twostage stochastic programming approach, model results suggest that the two most critical risks in the near-term planning process are natural gas prices and the stringency of climate policy. Stochastic strategies indicate some near-term hedging from lower-cost wind and nuclear will occur but robustly demonstrate that delaying investment and waiting for more information can be optimal under certain conditions to avoid stranding capital-intensive assets. These results are explained in terms of the optionality of investments in the electric power sector, leading to more general insights about uncertainty, learning, and irreversibility for utilities and generators. Additionally, I show that there is considerable value to limiting fugitive methane emissions from shale gas, which would allow more natural gas units to be built and operate under a larger range of climate policy and natural gas price scenarios. The stochastic solution is especially valuable if decision-makers do not sufficiently account for the potential for climate constraints in future decades, if fuel price projections are outdated, if discount rates are low, or if construction costs are expected to increase over time.