Weak Project Systems Imperil Next Generation Nuclear Projects

Edward Merrow

I am among the many observers of our climate change crisis who believe that a robust nuclear power industry will be a necessary ingredient in any successful transition to a radically lower carbon environment. Despite all of the problems with the nuclear fuel cycle, it is difficult to see how a transition to a low-carbon electrified world can be accomplished without large quantities of reliable base-load power from nukes.

Necessary or not, new nuclear power projects are a rare bird indeed in the Western Hemisphere. The primary reason is simple: building a nuclear power station is one of the riskiest endeavors a power utility company could undertake. In the world of industrial megaprojects, nuclear stations have the worst track record of any other set of projects. Nuclear projects data from IPA’s capital projects database show a median (half above/half below) cost overrun of 110 percent! The median execution schedule slip is 65 percent. By comparison, unsuccessful $500 million-plus megaprojects in other industrial sectors (chemicals, oil and gas, refining, etc.) have median cost growth and execution schedule slip of only about 30 percent. Like most large, engineering-intensive megaprojects, nuclear power project outcomes tend to be bimodal. They tend to be very good or very bad. When they start to go wrong, they go very wrong.

So why are nuclear project outcomes so much worse than other industrial megaprojects? Often regulation and community opposition are cited as the culprits. In the early days of nuclear power, regulations were constantly evolving as more was understood about the technology. But that is largely a problem of the past, not the present. And community opposition is a fact of life for a great many megaprojects, many of which manage to succeed nonetheless by addressing concerns effectively.

The problem with nuclear power projects is they are just bad projects. They fail for the same reasons that many other megaprojects fail. They fail because too many of their owners are not strong enough project organizations to develop, define, and control projects of their size and complexity. These large complex projects really do not have to fail; we let them fail, and that should be considered good news because that means we can fix it.

Applying a generous set of criteria, about one-third of industrial megaprojects overall are successful; the corresponding number for nuclear power projects is about 10 percent. So what is the key difference? Successful megaprojects start with strong, integrated owner teams with all of the key owner functions represented. That team then guides the project through very complete definition of the project, which we call front-end loading. This entails understanding the site thoroughly and all of the regulatory requirements around the technology as it applies at the site. It requires progressing the engineering work to the point at which detailed design can be fully mobilized immediately after authorization. And it includes very complete project execution planning by the owner. Execution planning must be very complete so that effective controls can be established and maintained. Contractors do execution planning for their work, but not for the whole asset; that is owner work.

The ongoing National Academy of Engineering (NAE) study, Laying the Foundation for New and Advanced Nuclear Reactors in the United States, is focused on exploring the technical, regulatory, and economic outlook for future nuclear projects, with the added recognition that advanced nuclear technologies could provide a zero-carbon source of power in support of U.S. efforts to decarbonize energy production. It is my view that this mission is all but certain to fail unless the root causes of nuclear project failures are addressed.

The NAE panel is focused on future nuclear reactor technologies and how they might reduce the problems associated with the extraordinarily large and complex projects presented by today’s nuclear technology. The problem I see is that by the time those technologies are ready to deploy, time may well have run out on our ability to mitigate climate change. There is no equivalent of the FDA’s “emergency use authorization” for new nuclear power plant designs. So if we are going to succeed, we probably will have to succeed with today’s pressurized water reactors, not the advanced modular and micro-reactors of the future. That means in turn that we have to address the causes of today’s project failures.

The primary cause of today’s nuclear project failure is organizational: the utility companies in the United States and Europe are not strong enough megaproject development and delivery organizations to get the job done. Historically, only a few utility companies (e.g., Duke and TVA) had the project system strength needed to deliver nuclear power station projects. Most utility organizations will do one megaproject a decade, maybe not even that many. Therefore, they never really build the muscle that is needed within the organization to deliver large complex projects. Electric utilities have been traditionally very dependent on contractors working on generous EPC-lump-sum arrangements to deliver their projects. This strategy limited both the risk and the personnel requirements for the utilities. It is not a particularly cost-effective strategy, but it has been workable for less complex projects. But it is not a workable strategy for nukes because the risks to the contractors are far too high.

So, is there a path forward for nuclear power? Possibly, but only with significant changes. In the United States, I believe that only a national strategy can work, in which a few utility companies, perhaps one in each region, are designated as the nuclear power providers for that region. Those companies will have to build from the ground up the kinds of owner project delivery systems that can develop and execute successful megaprojects. They can be guided by the best project delivery organizations in the oil and chemicals industries, which do large complex projects much more often. The utility companies must see a payoff for the heavy investment in expertise they will have to make. That can only be provided by multiple projects over a substantial period of time. In Europe, an EU wide strategy would be needed, which seems very unlikely in light of Germany’s rejection of a nuclear powered future.

Without a significant nuclear power component, we may have already lost control of our future.


This paper is based on a presentation to the panel preparing the National Academy of Engineering’s (NAE) study, Laying the Foundation for New and Advanced Nuclear Reactors in the United States. Merrow’s presentation and recorded remarks are available at the National Academies committee’s meeting webpage.

Get more details

  • Read our Privacy Policy
  • This field is for validation purposes and should be left unchanged.