The NuScale Scam

The GKN is a firm believer in competition, most importantly among nuclear vendors. Let a 1000 flowers bloom. As such we avoid commenting on specific technologies. We want the market to decide. But the market is not deciding. Bureaucrats (literally desk potentates or deskpots for short) are and they are deciding it with our money. The DOE recently announced it is doubling down on NuScale. It is time for Congress put a halt to this dead-end, taxpayer ripoff.

On June 17th, the DOE released a statement that began with

As part of President Biden’s Investing in America agenda, the U.S. Department of Energy (DOE) issued a Notice of Intent (NOI) to fund up to $900 million to support the initial U.S. deployments of Generation III+ (Gen III+) Small Modular Reactor (SMR) technologies.

In DOE-speak GEN III+ means Light Water Reactors with passive safety features. SMR means a power rating less than 250 MW thermal per module. Given the specificity of the award, we can be confident that a large part of the 900 million will go to NuScale.

NuScale is DOE's poster child for new nuclear. Table 2 shows a timeline for NuScale's development.

I have found it impossible to figure out how much taxpayer money has been "invested" in NuScale. Table 1 is the best I've been able to do.¹

At this point, the taxpayer is out at least $560 million. Probably something like $700 million. Certainly about to get a lot more. What do we have to show for it? NuScale is defending two reasonably well founded suits. NuScale has an NRC Design Certification for a 50 MW module it no longer intends to build. It may get a Design Certification for the 77 MW module with additional taxpayer money as early as mid-2025.

A Design Certification does not mean you can build a plant. For that you need a COL (Construction and Operating License). The UAMPS project has crashed and burned, despite the taxpayer footing 45% of the bill. No one has yet to even submit a COL application for any NuScale plant. Meanwhile, the rules can change as they did for Vogtle's AP1000's which were also NRC Design Certified. If a NuScale ever gets built in the US, it will cost in excess of $20,000/kW, at least 6 times the cost of a Korean APR1400. The taxpayer will almost certainly pay for at least 40% of it, for which she will get bupkes/nada/zero equity. What is required to produce such a mess?

The Wrong Design Goal

It starts with a design philosophy based on the premise that current nuclear power is not safe enough. NuScale opted to sacrifice economy in favor of improved safety metrics. To do this, NuScale decided to rely on natural circulation in the primary loop. Instead of a pump, the core coolant circulation is driven by the density difference between the hot and cold legs. This is hardly a new idea. Chimneys are natural circulation devices. The Trident submarines, which go back to the mid-70's, have a low power mode in which they can shut off their main coolant pumps. This was done to make the submarine quieter at low speeds. But the plant can put out less than 10% full power in this mode. A natural circulation nuclear reactor must be huge for its power output.

To implement natural circulation, NuScale has to go to a very tall 23 meter high, very small scale (77 MWe), low power density, material intensive reactor. Natural circulation forces low fluid velocities to reduce pumping power. This sets up a vicious circle. Low velocity means low heat transfer rates which means more heat transfer area, which means still lower velocity and so on. The reactor vessel gets larger and you end up paying for an overly large heat exchanger.

To make matters worse, they had to reduce the steam pressure, which in a pressurized light water reactor means a reduction in steam temperature. NuScale uses pre-World War I steam conditions of 35 bar and 308C. As a result, Nuscale features a putrid thermal efficiency of 0.31. Low thermal efficiency means more steam and more fuel for the same power out; and it means means the reactor itself has to get still bigger.

NuScale is neither small nor modular in any real sense. It is huge and involves at least as much on-site work as a conventional Light Water Reactor. A 924 MWe Nuscale 12-pack is housed in an immense swimming pool, Figures 1 and 2. Near as I can tell, this structure is 125 m long, 55 m wide, and 60 m high. This would be a volume of 412,000 m3. Table 3 compares this number with other LWR designs according to reference \cite{peterson-2005}. The NuScale Nuclear Island structure is more than double the size of the much maligned AP1000 which produces 20% more power. The AP1000 is based on a whole bunch of rail transportable sub-modules housed in an immense, site built, concrete structure. NuScale is based on a whole bunch of rail transportable sub-modules housed in a immense, site built, concrete structure. The only real difference is the NuScale building is a hell of lot bigger than the AP1000.

Figure 1. Profile view of NuScale pool. Structure is about 125 m long and 60 m tall.

Figure 2. Plan view of NuScale pool. Structure is about 125 m long and 55 m wide. 6 pack eliminates 3 columns of modules, but right side changes very little.

NuScale Safety

NuScale is so safe that the NRC has effectively ruled that it requires no buffer zones.² In the jargon, the Emergency Planning Zone(EPZ) is the plant boundary. According to NRC rules, the NRC can consider reducing the normal 10 mile EPZ, but only if the plant's thermal rating is less than 250 MWt. Coincidentally, a NuScale 77 MWe module has a thermal rating of 77/0.31 = 248.4 MWt. NuScale brags they can put their plants anywhere, including urban areas.

Take another look at Figure 2. We've got a dozen of these 248 MWt modules packed together. Suppose something goes badly wrong with one of the modules, perhaps because of an earthquake that drains the pool. Will the other 11 modules be unscathed? NuScale claims a core meltdown will only happen once every 3 billion module/years.\cite{nuscale-2016}[19.1-49] This is a concocted number. It could be right, but it is almost certainly wrong. We don't know what the NuScale meltdown frequency is. Nobody does.

Buffer zones should be based on what could happen. In terms of what could happen, a NuScale 12-pack is a 2976 MWt plant. It must have a 3 GWt buffer zone. By UCert rules that's at least 2 km. Given the 4th power fall off in harm near the plant, even a 1-pack requires a 1 km buffer zone. The only thing that NuScale has done right is to put their first plant at a remote location at the Idaho National Lab.

Warped Incentives

In any remotely efficient economic environment, the outrageously expensive NuScale concept would never get on the drawing board, let alone off of it. But US nuclear is not remotely efficient, and the NuScale design is ideally suited for the twisted incentives that we have imposed on the US nuclear apparat.

This is not a coincidence. The two founders of NuScale worked at NRC and DOE respectively before moving to Oregon State. The goal was never minimizing the cost of electricity. The goal was minimizing the cost of getting an NRC Design Certification and then a license. They are counting on that cost being even higher for any competition. They are quite out front about this. It's the core of their pitch to investors. Here are just two quotes:

The vast majority of our competitors will fail not because their technology is not viable or possibly even better than ours, but because they won’t be able to stomach the $1 billion or more in development costs to bring their technology to market. [Roland Berger, NuScale Consultant]

The licensing process in the US is so onerous and costly that it is a significant barrier for the many under-capitalized companies trying to replicate what we have already accomplished. [Jay Surina, NuScale CFO]

But they've run into a problem. The design turned out to be so expensive; they've got no customers. So they have turned again to the taxpayer. The idea is to extract money from the tax payer, by combining the fear of global warming with the fear of radiation exposure. (The irony is exquisite. To protect and sell your over-priced nuclear plant, you need to spread bogus fears of radiation harm.) Their willing intermediaries will be the deskpots at DOE, who know the more money they have to dole out the more secure and powerful they become. These people will assiduously hype the fears and extoll NuScale's safety. It's all in place and it's working.

NuScale will never produce much electricity; but it will funnel billions of taxpayer money to this partnership, while pounding another nail into the US nuclear coffin. Congress must put a stop to this profligate waste of resources, and redirect the entire US nuclear enterprise.

1. Excludes value of GAIN awards (taxpayer funded work at DOE labs), loan guarantee, INL site.

2. The NRC accepted a worse case acute dose of 2000 mSv.\cite{nrc-2022} 2 Sv's acute will produce Acute Radiation Syndrome symptoms. The SNT Lost Life Expectancy to someone who got hit with 2 Sv's acute and nothing else is 1.2 years. His UCert compensation for that dose would be $153,000.