Nuclear Plants Should Stop Making Power for the Grid
Since power markets and governments are not providing sufficient revenue to keep many nuclear plants open and the increasing adoption of wind and solar will only exacerbate this problem, merchant nuclear plants should focus on serving co-located industry rather than the grid. In fact, it may be the most economic way to decarbonize industry and balance a high renewables grid long term.
U.S. merchant nuclear plants have faced a number of premature closures in recent years due to falling wholesale energy and capacity prices . NEI reports that 16,490 MW of nuclear capacity responsible for 135.6 TWh of annual energy production (over 3% of U.S. electricity demand) have shutdown or announced shutdown since 2013. While most climate-concerned groups recognize the irrationality and economic cost of these closures, power market designers and regulators have been unable to respond quickly enough. A big part of the problem is that existing nuclear, despite being the largest source of zero carbon generation in the United States for the last 40 years, does not qualify for many of the government subsidies or Tech company PPAs because it is not “additive” zero carbon generation (it already exists).
Nuclear plants also suffer from their inherent economics: their operating costs are entirely fixed no matter how much you ramp them up or down or curtail them due to low energy prices (unlike a natural gas plant in which 80% of your operating cost is saved when you shut off the plant and stop consuming fuel). Similar to renewables, nuclear plants have zero marginal cost in the short run, so it is best to squeeze out every MWh you can from your sunk up-front costs. For this reason, nuclear plants tend to be price-takers and ride-through periods of low or even negative energy prices. From a technical and maintenance cost point of view, the light water reactor design of current U.S. nuclear plants also prefers steady-state operation rather than ramping the power and temperature of the core up and down during the fuel cycle. The following chart shows the break-down of typical costs at existing nuclear plants ($/MWh):
With the planning horizon afforded by a long-term offtake agreement and some further operational efficiency innovations, it would be reasonable to see these long-run average cost numbers approach or possibly even dip below ~$25/MWh for multi reactor sites. This is the total revenue such a nuclear plant owner hypothetically need to break-even in a world with no risk or uncertainty. It is important to recognize that nuclear operators do face some risks: equipment and nuclear fuel can fail, fuel prices can increase, extreme weather can occasionally force an outage, interest rates and the cost of borrowing capital could increase, the counterparty could not follow through with payment, a squirrel could short the transmission line. Therefore, it is reasonable to assume nuclear plants need some premium on top of this ~$25/MWh cost to compensate them for these risks (and let them capture some profit). One would need to spend time quantifying each risk, but a reasonable, possibly even conservative, assumption on this premium might be +$7/MWh, leading to a total revenue requirement of ~$32/MWh.
However, the price a co-located industrial customer would need to pay could be considerably less if they are willing to offer some combination of curtailment, profit sharing, equity, and financial hedging. Curtailment has the most immediate impact on reducing the industrial customer’s average cost per MWh. If power markets like PJM provide capacity payment for the full output of the nuclear plant, which they should as long as the industrial load agrees to be curtailed during extreme grid events (e.g. polar vortexes), the nuclear plant’s revenue requirement drops by ~$3/MWh (more if PJM capacity markets return to historically higher numbers). Further, if industrial loads also curtail to let the nuclear plant capture the highest priced hours of grid demand (say 3% of the highest priced hours in the year), this revenue requirement drops by another ~$3/MWh (more if renewables drive more wholesale price volatility in the future). Both types of curtailment bring the $32/MWh revenue requirement down to $26/MWh. If industrial offtakers offer profit-sharing, proof of hedged revenue that covers energy costs, or go so far as to buy equity in the nuclear plant itself, the $7/MWh risk premium could shrink or disappear entirely, bringing the final energy price for the industrial customer down below $20/MWh. In summary, co-located industrial loads could receive direct, unambiguous zero carbon power with 95–97% availability, and 100% reliability, at less than $26/MWh from nuclear plant operators (and less than $20/MWh if they bear more of the risk).
Such an arrangement could enable a much more economically palatable pathway towards the electrification and decarbonization of many industrial commodities (e.g. hydrogen, fertilizer, steel, electrofuels, indoor agriculture) while preserving the nuclear asset for grid balancing. With enough load attached to a nuclear plant, they can reliably recover their fixed operating costs, maintain steady reactor core operation, and become zero carbon peaker plants for the grid at a cost much lower than long-duration storage. Idaho National Labs has been studying this “Integrated Energy System” strategy for several years, but this transition to co-located curtailable load is economic and should start happening today.
Bitcoin miners are desperate for power contracts anywhere close to these terms and are willing to bear the additional cost burden of navigating the legal barriers and setting up the interconnection for a first-of-a kind power contract. We have already seen the first move with Talen Energy’s Susquehanna nuclear plant. With the growing criticism of bitcoin’s carbon and energy intensity, using bitcoin miners to preserve zero-carbon nuclear capacity would be a perfect marriage of two renegades of the energy world. Bitcoin currently represents over 4 GW of potential U.S. load and could pave the way for a broader industrial electrification movement. Nuclear plants could become WeWork’s for industrial load. Given the economic pressures and lack of regulatory action, one might go so far as to conclude that nuclear plants need to make this transition in order to survive.
