MIT study: Existing US nuclear could be better run for climate and profit

Letting US nuclear bid on day-ahead reserves could bridge them to retirement and reduce system CO2 emissions

Michael Barnard
Jan 3 · 9 min read
Image of behind schedule, over budget Vogtle nuclear plant
Image of behind schedule, over budget Vogtle nuclear plant
Vogtle nuclear plant under construction years over schedule, billions over budget
Podcast version of this article

The US nuclear industry is having a tough decade. In 2010, it was common to see headlines in energy journals, think-tank press releases, and even the media that usually covers the Kardashians talking about a . Not any more.

The logic was sound in 2010. Nuclear remains one of the largest sources of low-carbon electricity in the world, a bit more than wind or solar, but far better than gas or coal. And reducing global warming requires lots of low-carbon electricity. At the time, wind and solar were as expensive as nuclear, and major analysts such as the IEA were projecting relatively low penetrations when the need was for much more.

Further, there had been a couple of examples of effective build-out of large fleets historically which indicated that it was possible, if all the conditions were right, to do it relatively economically and quickly. France is the poster child for this, and still has over 70% of its electricity coming from nuclear, although it plans to reduce the supply to 50%. China had a nuclear program that was going to turn on lots of new reactors. Watts Bar Unit 2 in Tennessee was under construction again and four Westinghouse AP1000 reactors in South Carolina and Georgia were entering the final phases of approval with construction scheduled to start in a couple of years. The AP1000s were going to be the standard technology used to avoid cost and budget over-runs by being standardized and manufactured as opposed to custom-engineered mega-projects. Chernobyl was firmly in the rearview mirror.

Around the world, nuclear aspirations were high. And then reality started rearing its ugly head.

On March 11, 2011, the Tōhoku earthquake triggered a tsunami that hit the Dai Ichi nuclear power facility in Fukushima, Japan. As documented elsewhere, total economic costs of the nuclear portion of the disaster alone likely are going to be close to a trillion US dollars. That’s $2 billion for every nuclear reactor on Earth, as Professor Mark Jacobson, lead of the 100% Renewables by 2050 Stanford research team, pointed out on Twitter. And that liability was almost entirely going to be felt by the taxpayers of Japan, not insurance companies, due to liability caps for nuclear.

Wind and solar started taking off, vastly exceeding IEA’s projections year after year, and plummeting in price. Gas generation soared as well, based on the massive and inexpensive supply of fracked gas within the USA. All of a sudden the price point of new nuclear was uncompetitive regardless of Fukushima.

And then the new US reactors started having problems. The AP1000s weren’t living up to the hype. Westinghouse entered bankruptcy, and Toshiba eventually sold it to Brookfield, which bought it for the long-term decommissioning revenue, not for building new plants. The Vogtle and V.C. Summers projects were both experiencing massive cost and budget overruns. V.C. Summers was finally put out of its ratepayers’ misery in March 2019, but the fiscal hangover will be felt for years. Vogtle has managed to secure additional billions in governmental financing and is proceeding, for now. Duke Energy’s 860 MWe Crystal River PWR was decommissioned early after installation of a new steam generator damaged it severely. Only Watts Bar made it into production, but that took over 30 years given that construction had been halted in 1985 and the new construction suffered budget and schedule overruns increasing its new cost to $4.7 billion despite being more than 60% complete already.

And existing nuclear plants were feeling the heat. Despite being a mostly-paid-for source of carbon-neutral electricity, they were competing with cheaper new gas generation and then in 2014 or so, cheaper new wind and solar as well. In 2014, 13 reactors were under threat of closure due to low power prices making them uneconomic. The cost of operating the plants had increased 28% to an industry average $36.27 per MWh. That number excludes any remaining debt payment, and wind PPAs were coming in under that halfway through the decade in the Midwest.

Nuclear plants started closing early. Dominion Energy’s 566 MWe Kewaunee PWR in Wisconsin in 2013. The two PWRs at San Onofre nuclear plant in California also in 2013. Vermont Yankee in 2014. 19 new reactors that were undergoing planning and scheduling were cancelled outright or postponed indefinitely.

Thoughts of a nuclear renaissance are firmly dead in most people’s minds, except for a subset of people like Michael Shellenberger who don’t accept global empirical reality about nuclear vs renewables as the path forward. They became convinced of the value of nuclear a decade ago, established a major public profile based on it, and now are pot-committed to the wrong technology. China, the great nuclear hope, is seeing nuclear not meet targets while wind and solar vastly exceed them, leading to a reduction in nuclear targets and a doubling of wind and solar targets for 2030. It’s clear that the USA will achieve much greater carbon reductions much more quickly and much more cheaply with renewables than by building new nuclear.

The nuclear industry is in many places going, hat in hand, to regulators and politicians, asking for additional funding for reactors to allow them to keep operating. In Ohio, the state government pushed through a $1.1 billion subsidy for their two operating nuclear plants to allow them to remain solvent a few years longer. The logic is somewhat sound, as they do exist and do provide low-carbon electricity. There’s merit in that. But if the choice is to build a mix of wind and solar for a lower cost than perpetuating an aging nuclear plant, it’s a tough argument to make.

But as the headline says, there’s hope for the 98 US reactors in operation today. When the various countries of the world were selecting their preferred technology for nuclear generation, the US swung to light-water pressurized water reactors (PWRs). They were deeply familiar with them as they were the same technology used on nuclear powered submarines and aircraft carriers, something the USA had which most other countries didn’t.

And PWRs can be used to follow load. It’s a much slower response rate than peaker gas or using SCADA controls to curtail or spin up wind and solar, but it’s viable. For much of their fuel cycle, they can drop or increase generation by 25% per hour, although when production is dropped they have to remain at the lower level for typically hours to allow xenon to dissipate before they can be increased again.

But unlike France, reactors are treated by grid operators as fixed, baseload generation, either on or off. There’s some limited seasonal load following related to hydro’s spring peak, but that’s about it. Part of that is purely economic. Load following with a nuclear reactor reduces the total GWh that are generated annually, and the only contracts they have are for committed baseload. If they were operated to follow load, they would lose money. Instead, renewables end up being curtailed when surplus baseload generation occurs. This is a somewhat reasonable approach, but it comes with an interesting wrinkle. Gas and coal reserve power is maintained for the nuclear plant and burn fossil fuels while wind and solar are curtailed.

So we have a technology that could load follow but isn’t allowed to by regulatory and contractual structures without economic penalty, and as a result lower-carbon forms of generation are curtailed while more gas and coal are burned. That’s an odd systemic choice in 2019.

Enter Jesse D. Jenkins of MIT and Zhi Zhou of Argonne National Laboratory, who led a team to model a Southwest alternative system management regimen, one which used real-world data from Arizona and New Mexico to project what would happen if Arizona’s PWRs were able to exploit their inherent technical abilities. They published their results in mid-2018 in the journal report . It’s a solid paper in a rock-solid journal, not a think tank puff piece. Their conclusions are very worth assessing.

Gas especially but also coal bid on day-ahead reserve markets and they are running as hot standby so there is fossil fuel being burned. The nuclear keeps spinning at its maximum and when there is surplus baseload generation, wind and solar are curtailed while the gas and coal continue to burn. That’s not how you would expect a rational system to operate and is the deep insight of the study about the system.

If nuclear were permitted to load follow to a limited extent, it could bid on day ahead reserve as well, knocking coal and gas out of the equation. As the nuclear would be running at perhaps 75%, wind and solar would not be curtailed for that 25% but could fill it. Southwest USA energy data for a year was processed with load following nuclear and found interesting results.

The big one is that if the PWRs were allowed to load-follow and bid on day-ahead reserve markets, the following would occur:

  • A lowering of wind and solar curtailment
  • A reduction of coal and gas being burned
  • Overall electricity costs drop due to not burning coal and gas
  • And a net revenue increase for the nuclear plants
Table showing increasing in gross margins if nuclear were allowed to bid on day-ahead reserves
Table showing increasing in gross margins if nuclear were allowed to bid on day-ahead reserves
Table courtesy Jenkins et al

This is a very good news story for the US PWR fleet and their owners. If they could convince regulators to allow this change in grid management and draw up new contracts to support it, they could keep more reactors running producing low-carbon electricity longer in the face of competition from cheap gas, wind, and solar. And the US overall would reduce its very high carbon pollution rate.

This would be a much better path forward than giving more public money in the form of subsidies or tax breaks to keep the nuclear reactors going. It’s non-trivial, as all things related to grid management and nuclear power are, but it’s viable. It would require regulatory changes, contractual changes, grid operation procedural changes and plant operation procedure changes, but that’s business as usual.

There are some wrinkles, both good and bad. As with many studies in this area, it assumes that there are no transmission constraints, which isn’t true in reality but is becoming more true. It also assumes a smaller regional grid without additional load balancing across a broader geographic region, something which is becoming less true with each passing decade. The load following ability for PWRs only applies for roughly the first year of their 18-month fuel cycle as in the last third they have to operate at or above 86% of capacity for technical reasons that are immutable. Only 20% combined wind and solar are assumed, so this is a next decade model, not a 2050 model. And for some reason they model in the PTC for wind energy despite that going away in 2020 and there being no possibility of this model being applied for any US region before then. This is very much a point-in-time approach and the basis for further studies specific to different plants in different regions as part of the assessment of viability and results, and the authors fully acknowledge this.

When I reached out, the primary study author, Jesse Jenkins, PhD, responded with a useful nuance:

This doesn’t provide a path forward for new nuclear of course. That’s still too expensive and too slow to build. But it gives a path to a more leisurely retirement for the existing plants and lower overall CO2e emissions for the USA.

Imagine a world in which nuclear, wind and solar were actually cooperating as the natural allies that low-carbon generation sources should be in the age of climate change. The USA has that option. Instead, the federal government and jurisdictions such as Ohio are simply throwing more billions at nuclear as renewable subsidies decline and disappear.

The Future is Electric

The Future is Electric is the house journal of TFIE Strategy Inc, a firm which assists global clients to future proof themselves in our rapidly changing world of business and technical innovation, and geopolitical and climate disruption.

Michael Barnard

Written by

Chief Strategist, TFIE Strategy Inc. Business and technical future-proofing. Top Writer Quora since 2013. CleanTechnica, Forbes, Quartz+ more. In 4 books.

The Future is Electric

The Future is Electric is the house journal of TFIE Strategy Inc, a firm which assists global clients to future proof themselves in our rapidly changing world of business and technical innovation, and geopolitical and climate disruption.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade