Nuclear power plants tend to have high fixed annual costs, so they must run as much as possible to spread these costs over a maximum number of megawatt-hours. This makes nuclear plants inflexible, meaning they cannot rapidly ramp output, for economic reasons if not technical ones.
In a large market, this inflexibility is not much of an issue -- nuclear generators are price-takers and get paid the average annual locational wholesale electricity price. In this case, megawatt-hours are strictly commodities; nuclear power is cheap if costs fall below those of competing generators. However, in a more constrained market like those with lots of zero-fuel-cost wind and solar variable generation, nuclear power only competes as a commodity up to the minimum level of net load (load minus zero-cost generation). Any more nuclear generation would raise overall system costs.
Let's look at the specific case of Diablo Canyon Nuclear Power Plant in California. if California were better integrated into a regional grid, this larger grid could more easily accommodate Diablo Canyon’s inflexibility. If PG&E could competitively sell off excess nuclear generation to its neighbors, it could use the remaining power to replace its own dirtier gas generators and lower emissions. At that point, it might be worth tipping the scales toward re-licensing.
Because Diablo Canyon is generating somewhat expensive power into a shrinking power pool where policy and economics are creating demand for more flexible generation, saving it might not seem to be worthwhile.
Policymakers should only make a decision on support for existing or new nuclear plants after a pragmatic, hard-nosed analysis. And if they decide to support a plant in a wholesale market region, care must be taken to ensure the support mechanism does not undermine the market integrity or run afoul of federal law. Hopefully, we will maintain the plants that are currently most valuable at the mercy of an over-capacity grid. (GTM, 3/6/2017)