DOE's Nuclear Growth Proposal 'Untethered to Reality'

The US Department of Energy (DOE) is once again promoting large-scale reactors after spending a decade advancing smaller models. It boldly declares in a report last month that the domestic nuclear industry has the potential to “scale from ~100 GW in 2023 to ~300 GW by 2050 — driven by deployment of advanced nuclear technologies.” This is beyond absurd — it’s irresponsible. It’s absurd because the US no longer has the supply chain needed for large-scale nuclear projects — it can’t even forge a pressure vessel; it’s irresponsible because the cost of building 200-300 new reactors would be more than $3 trillion. Resources devoted to rescuing a dying industry are resources that wouldn’t be available for viable, less-costly strategies to achieve net-zero emissions in the power sector. More than that, the report reflects an energy agency still dominated by a nuclear-centric culture, and badly out of step with the times.

The March report, Pathways to Commercial Liftoff: Advanced Nuclear, is among three in a "liftoff" series intended to ”establish a common fact base with the private sector around the path to commercial liftoff for critical clean energy technologies.” (The other two cover hydrogen and energy storage.) Unfortunately, the nuclear report reads more like fiction than fact — the authors lay out a fantasy future for nuclear, based on a consortium of interests that coalesce around a unified reactor design, an initial order book of 5-10 reactors, costs eventually lowered by 30%-40%, manufacturers willing to invest in new supply chains, and a streamlined regulator. A third party (government or nongovernment) could provide “cost overrun insurance” and pay, say, up to 50% of a total cost overrun, thereby offloading the burden of risk that currently prevents US utilities from embarking on nuclear projects.

High-level nuclear waste (e.g. spent fuel) remains problematic — but let’s be optimistic: “While the current focus for DOE is on consolidated interim storage capability, the lessons learned from consent-based siting for a federal consolidated interim storage facility will be applicable for siting any permanent disposal facilities in the future.” Both New Mexico and Texas are fighting proposed privately owned interim nuclear waste sites in their states, although they remain under Nuclear Regulatory Commission (NRC) review. It's unclear how this might affect DOE's learning curve on consent-based siting for a federal facility.

The DOE puts the total cost of the build-out at "~$700 billion plus," but based on the current price tag of America’s only nuclear newbuild at Southern Co.'s Vogtle site in Georgia — roughly $17.5 billion per gigawatt — the cost would be $3.5 trillion-$5.25 trillion. That’s more than double the $1.5 trillion the US plans to spend over 30 years overhauling its nuclear arsenal, and in the neighborhood of the government's $6.9 trillion budget proposal for the entire nation in 2024. DOE argues that the high capital costs can be lowered by as much as 40% — after incorporating “lessons learned” from repeat deployments of a specific design and “investing heavily in upfront project planning and scheduling.”

We’ve heard it all before — unified design, industry consortiums and “lessons learned.” But persuading utilities to march to the same drum when it comes to reactor designs has never happened before — and the DOE itself is promoting a plethora of different designs through its small modular reactor and "advanced reactor" programs. DOE's assumptions about positive learning rates are far more optimistic than a host of other studies, including a 2020 MIT study covering five decades of nuclear plant construction in the US that found no evidence of positive learning. "Counter to expectation, nth-of-a-kind plants have been more expensive than first-of-a-kind plants," an abstract of that study states.

Getting to 200 GW by 2050 requires a rapid rollout — deployment must start by 2030, ramp up annually to 13 GW by 2040 and hit the 200 GW mark a decade later, DOE says in its report. That means building approximately 70 large-scale reactors between 2030 and 2040, or perhaps some combination of small and large. “There’s a role for small modular reactors,” the report assures, since “they can provide more certainty of hitting a predicted cost target and are likely to play an important role in the early scale-up of nuclear power.” That’s doubtful. For one thing, there are few cost targets even available for small reactors (50 megawatts to 300 MW). And where they exist they are anything but certain — witness the steep climb in NuScale’s cost forecasts. As of March, the company was predicting $9.3 billion for six 77 MW modules, or over $20 billion/GW and, prior to construction start, already more expensive on a per megawatt basis than Vogtle.

Manufacturing Constraints

Leaving aside financing, siting, public acceptance and regulation — all key to a successful nuclear plant rollout — there is the fundamental question of who’s around to build such a fleet. The US is essentially a service economy, and once vertically integrated companies like Westinghouse and GE that manufactured their own nuclear steam supply systems are now focused primarily on design, engineering and project management.

The DOE notes this problem in its report. “There is not enough N-stamp certified supply chain capacity to support 13 GW per year of added nuclear capacity,” it admits, referring to a special “nuclear” stamp required for many reactor parts. “There is no US large-forging capacity.” Expanding N-stamp capacity requires new investment and “can take over a year and cost hundreds of thousands of dollars.”

That’s putting it optimistically. Building heavy forging capacity to ASME (American Society of Mechanical Engineers) nuclear code probably would take five years, and cost in the tens of millions if not hundreds, according to one expert. Either way, it’s a classic chicken-and-egg dilemma. Suppliers won’t invest in new capacity without strong order books, and utilities risk delays without assured supply.

“You can’t get there today. There’s no vertically integrated company” in the US nuclear sector, says a former executive of a major US reactor vendor. “Everything’s offshored.”

What of 'Liftoff'?

“They just don’t seem to be tethered to reality,” adds another industry veteran. Besides the enormous cost, he questions whether there are enough potential sites with the “necessary heatsink, reasonable risk of natural disasters and full-throated public acceptance” for new nuclear plants. And without significant regulatory reform “there’s no way a private company could suffer the insufferable NRC processes.” Besides, he wonders, “doesn’t anybody [in the industry] but me think there will be multiple technical revolutions in competing energy production techniques?”

The DOE report downplays renewables, saying that “regardless” of the level of their deployment, the US will need around “550 GW–770 GW of additional clean, firm capacity to reach net zero; nuclear power is one of the few proven options that could deliver this at scale.” (The report includes a disclaimer saying it’s only intended as a basis for discussion and doesn’t reflect official policy.)

It's telling that nuclear was included in the first batch of "liftoff" reports ostensibly aimed at accelerating clean-energy technologies. It's authors are employed by an agency, which descends from the old Atomic Energy Commission (AEC), and has a long-standing reputation for grossly exaggerating nuclear's potential. When it was still the AEC and later after it became DOE (in the 1960s and '70s), the agency promoted "fast" breeder reactors, forecasting 1,000 such "advanced" reactors by 2000. None were ever completed.

During the more recent "nuclear renaissance" in the early part of the century, more than 30 new reactors were promised. Only two were ever built (at Vogtle).

If DOE is "untethered from reality," as the industry veteran suggests in his email, it's likely owing to the agency's history, organization and nuclear-centric culture. The DOE not only promotes civilian nuclear power, its “semiautonomous” National Nuclear Security Administration oversees the country's vast nuclear weapons complex. That's a toxic mix that clouds perspective and denies the country honest and realistic energy policymaking.

Stephanie Cooke is the former editor of Nuclear Intelligence Weekly and author of In Mortal Hands: A Cautionary History of the Nuclear Age. The views expressed in this article are those of the author.