In the modern era, nuclear power plants have almost always become more and more expensive over time. They have a “negative learning curve” — along with massive delays and cost overruns in market economies. This is confirmed both by recent studies and by the ongoing cost escalations of nuclear plants around the world, as I’ll detail in this post.
The cost escalation curse of nuclear power
“Ever since the completion of the first wave of nuclear reactors in 1970, and continuing with the ongoing construction of new reactors in Europe, nuclear power seems to be doomed with the curse of cost escalation,” read one 2015 journal article, “Revisiting the Cost Escalation Curse of Nuclear Power.”
In the United States, the cost of Georgia Power’s newest twin Vogtle reactors may top initial estimates of $14 billion and reach $21 billion, according to recent Georgia Public Service Commission testimony. Of course, the first two Vogtle Units begun in 1971 took 18 years to build (a decade over schedule) at a final price of $9 billion — ten times the original price tag. BloombergBusiness wrote last fall, “Even as sympathetic an observer as John Rowe [former chair of the U.S.’s largest nuclear utility] warns that the new units at Vogtle will be uneconomical when — or if — they’re completed.”
Even the French can’t build an affordable, on-schedule next generation nuclear plant in their own nuclear-friendly country. Their newest Normandy plant, which originally was projected to cost €3bn ($3.3 billion) and start producing power in 2012 “will not start until 2018 at a cost of €10.5bn [$11.3 billion],” the Financial Times reported last year.
The high and rising price of new nuclear power plants does not mean new nukes will play no role in the fight to avoid catastrophic warming, as I discussed in January. It does means that, barring a huge unprecedented and ahistorical price drop in next-generation nuclear plants, the role nuclear power plays will be a limited one — a very limited one in market economies especially if the industry can’t reverse decades of cost escalation. Certainly an R&D; breakthrough is worth pursuing, but adding even more policies to specifically accelerate deployment of new nukes makes little sense at this point.
Nuclear cost curves are in stark contrast to their zero-carbon competition — renewables, storage, and efficiency — which continue to see huge price drops. That’s a key reason renewables and efficiency will play by far the biggest role in preserving a livable climate — even in a best-case scenario the International Energy Agency did with the Nuclear Energy Agency in 2015. They project the renewable share of electricity generation will be nearly four times larger than the nuclear share in 2050:
Similarly, in newly-released findings, South Australia’s nuclear royal commission found that the price of electricity from new nukes greatly exceeded not only business-as-usual projections for electricity prices but also prices in a “strong climate action” case. The Commission concluded “it would not be commercially viable to generate electricity from a nuclear power plant in South Australia in the foreseeable future.”
The Commission explicitly looked at plausible electricity prices for a new reactor in 2030 based on both current designs and possible fourth-generation ones, such as small modular reactors (SMRs). The Commission estimated the cost for the most viable nukes at US$7 billion for a typical large 1125-megawatt reactor and $2.8 billion for two 180-megawatt SMRs. The smaller SMRs would be providing electricity for a whopping US$0.17 a kilowatt-hour!
A study done for the Commission found that both large nukes and SMRs “consistently deliver strongly negative NPVs” (net present values) for both 2030 and 2050 — even for the strong climate action scenario. The Commission Chair noted that given how Australia’s National Electricity Market works, renewables are “the first energy that goes into the market” because they have the lowest costs.
The Commission’s findings are consistent with a 2014 Energy Policy study, “The cost of nuclear electricity: France after Fukushima.” Using cost data released by the French government after the Fukushima disaster, the study found the cost of French nuclear plants steadily escalated over the past four decades. Further, it projects “the future cost of nuclear power in France to be at least 76€/MWh (US$0.084/KWh) and possibly 117€/MWh (US$0.129/KWh),” which “compares unfavorably against alternative fuels,” such as wind.
Nukes are cursed with cost escalation everywhere
Ironically, even a brand new study in Energy Policy by nuclear advocates, “Historical construction costs of global nuclear power reactors,” purporting to show that nuclear power does not always have a negative learning curve, actually finds the reverse. Once you get past the very early-stage development (mid-1950s to mid-1960s) with its super-high costs, nuclear power plants clearly have a negative learning curve in every country the authors examine. The only small exception is South Korea, which has a tiny fraction of the world’s nukes and for which the cost data is questionable.
Leading energy analyst Dr. Jon Koomey of Stanford University summed up the new study this way:
“With the exception of South Korea, whose nuclear cost data are not independently audited and are therefore of unknown quality, all available cost data paint a consistent picture: the investment risk of nuclear power is significant, costs almost always increased over time in the modern era, and cost overruns and lengthened construction times need to be considered carefully by investors and policy makers alike.”
Moreover, Koomey explains, “adding interest during construction would only strengthen these conclusions.”
What? Yes the authors of this paper — in an effort to downplay the staggering cost of nuclear power plants — make use of the extremely dubious “overnight cost” of nuclear plants, rather than their real costs. The overnight cost is literally the theoretical cost of a power plant if you could build it overnight, if you could built it without interest charges or cost overruns from delays.
For wind and solar projects, which have construction periods typically measured in months, the difference between an “overnight” cost and the actual cost to complete the project is usually not that significant. But nukes typically take many years to complete — and a decade is not uncommon. Cost escalations that occur during this long construction period, plus the financing costs during construction, may easily double the total cost of a project compared to its “overnight” cost. Trying to oversell nukes by focusing on the imaginary overnight costs is tantamount to selling someone a house with “teaser” initial mortgage payments and failing to make clear that the payments will later balloon to a much higher level.
Cost-overruns are too intrinsic to nuclear power plant construction to ignore. Indeed, a detailed 2014 study of some 401 “electricity infrastructure projects built between 1936 and 2014 in 57 countries,” concluded:
… hydroelectric dams and nuclear reactors have the greatest amount and frequency of cost overruns, even when normalized to overrun per installed MW, and that solar and wind projects seem to present the least construction risk.
Quelle surprise. This study, “An international comparative assessment of construction cost overruns for electricity infrastructure,” examined the cost history of 180 large nuclear reactors worldwide, and found, “these overruns afflicted more than 97 percent of nuclear projects and led to a mean cost escalation of 117 percent per project.
Nuclear plants have large cost overruns, and it appears to be an intrinsic feature of them. The authors of “Revisiting the Nuclear Power Construction Costs Escalation Curse,” had a very important conclusion in their analysis of escalating costs for French nuclear reactors:
Our last result says that those reactors with better safety performance were more expensive. Then achieving higher safety levels also helped to explain the cost escalation in the French nuclear fleet.
I’m pretty sure the vast majority of Americans, like the vast majority of French, would prefer we built reactors with the best safety performance. That doesn’t come cheap.
Bottom Line: For now, the nuclear industry has priced itself out of the market for new power plants in market economies. If the world pursues the emissions and energy pathway needed to keep total warming below 2°C path, and if the nuclear industry can avoid another major disaster while resolving a variety of issues, especially cost escalation, then nuclear power can make a modest but important contribution. But it remains increasingly clear that new renewable energy will play a far bigger role in the transition.