New Nuclear: Three Projects, Three Problems

From MasterResource

By Kennedy Maize — March 9, 2023

“NuScale’s estimated ‘levelized cost of energy’ (LCOE) … jumped by one-half last December (to $89/MWh from $58/WMh). The total cost for the project is estimated at a bit over $9 billion, but $1.4 billion is offset by DOE funding, which could increase in the future.”

More news, more problems regarding the live projects being counted on as the beginning of a new era of nuclear power. Back in the 1950s/1960s, the expectation was that learning-by-doing and scale economies would bring parity with fossil-fuel plants. Today, that same goal seems distant.


NuScale Power’s small modular reactor project, designed to provide electricity to Utah Associated Municipal Power Systems, a joint action agency serving 50 municipal utilities in Utah, Arizona, California, Idaho, Nevada, New Mexico, and Wyoming, has survived another near-death experience. Facing a vote by the participants in its project for six light-water pressurized reactors, totaling 462-MW on the grounds of the Department of Energy’s Idaho National Laboratory near Idaho Falls. Twenty-six of the 27 utility participants agreed to continue with the “Carbon Free Power Project.”

The vote came in the face of a steep increase in NuScale’s estimated “levelized cost of energy” (LCOE), which jumped by one-half last December (to $89/MWh from $58/WMh). The total cost for the project is estimated at a bit over $9 billion, but $1.4 billion is offset by DOE funding, which could increase in the future.

The December hike triggered a vote on whether to continue the project. Of the 27 municipalities participating in the plan, Morgan City, Utah, withdrew its 1.398 MW participation; Parowan City, Utah, reduced its commitment from 3 MW to 2 MW; and Los Alamos increased its share from 2.1 MW to 8.6 MW. That left the total commitment so far about 120 MW, far short of what makes the project economically viable. The current in-service estimate is 2030.

The initial project was a 12-reactor, 720-MW station at a cost of $4 billion, in service in 2026. The cost estimate quickly rose to $6 billion, original participants started dropping out, and NuScale and its contractor Fluor Corp. reworked the reactor design and scaled back the project to 462 MW. Are more cost escalations ahead? With rising interest rates and the experience of Plant Vogtle #3 and #4, more surprises can be expected–and requests for more DOE help.

X-energy and DOW

X-energy and DOW have switched the planned location for their four-unit, 320-MW high temperature gas cooled reactor (HTGR) using TRISO billiard-ball-size fuel “pebbles”, from Washington state to an unspecified site on the Gulf Coast, close to a Dow chemical industry facilities.

X-energy of Rockville, Md., is a startup specialist in HTGR technology. DOW, based in Midland, Mich., is a venerable chemical company, founded in 1897. Their planned $2.2 billion nuclear project has $1.2 billion in Department of Energy funding for the power plant and a fuel fabrication plant in Oak Ridge, Tenn. The target is for 2028.

HTGR is an elegant conceptual approach to nuclear, with a mixed practical record. It uses inert helium gas as a coolant, so the plants require water for only boiler feed and makeup. They produce high steam temperatures. The uranium fuel and the graphite moderator are incapsulated in a ceramic material, making them extraordinarily strong. Their advocates bill them as “walk away safe” in a nuclear accident.

The first U.S. HGTR was Peach Bottom on the Susquehanna River outside of Philadelphia. It ran well from 1967 to 1974. It used a solid, “prismatic” fuel combining uranium and thorium with the graphite moderator. The 40-MWe plant produced steam at 1,000 F and 1,450 psig. with a thermal efficiency of 37% and 88% power availability. A small, 13-MW HTGR operated in Julich, Germany, known as AVR, connected to the grid in 1967 and ran until 1988, the first to use the spherical, “pebble-bed” TRISO fuel.

Scaleup of HTGRs has been a problem. In the U.S., Public Service of Colorado built a 330-MW General Atomics reactor. The plant operated fitfully from 1979 to 1989, beset by plumbing problems. It shut at a point where the problems appeared solved, but the economics had changed. Germany scaled up the AVR with the 300-MW pebble-bed THTR, connected to the grid in 1985 and closed in 1989, a victim of financial and political woes, including the nuclear skepticism in the wake of the 1986 Chernobyl disaster.

Scaleup problems may be irrelevant, as HTGRs have operated successfully at small sizes.

Virginia’s SMR Plan

Virginia Gov. Glenn Youngkin’s SMR plan hits a roadblock, NBC television station WCBY-TV reports. The Republican governor launched legislation for an SMR for southwestern Virginia, the heart of the state’s struggling coal industry. The Bristol, Va., TV station is in the middle of the coal-rich region. The Washington Post reported, “Virginia Gov. Glenn Youngkin’s (R) energy plan calls for Southwest Virginia to build the nation’s first commercial small reactor. The governor was in Wise County in October promoting the plan at an abandoned mine site. Virginia is among at least eight states pursuing a small reactor.”

Youngkin’s vision crashed in the legislature, which ended its session last week without authorizing SMR provisions. “The actual SMR bill itself, did not get out of conference committee,” Republican Del. Israel O’Quinn told WCBY. “Just couldn’t get an agreement on exactly how to get that out of there.”

Last Word: Vaclav Smil

The familiar process of overoptimism and creeping reality brings to mind the wisdom of Vaclav Smil, who wrote in his most recent book, Invention and Innovation: A Brief History of Hype and Failure:

In light of the past experience with nuclear promises, the only sensible attitude is to wait and see how many of these announced plans will, even with the added incentive of accelerated decarbonization, become actual working prototypes, and then how many of those will make the second cut to lay the foundations of future commercial opportunities. In any case, no nation has announced any specific, detailed, and binding recommitment to what would have to be a multidecadal program of reactor construction.


Ken Maize is Proprietor, The Quad Report: Energy Policy and Politics at Sawmill Creek Communications in Knoxville, Maryland. Before then, he was executive editor of POWER magazine. He edited and wrote for Electricity Daily for 20 years, before which he held the same position for Energy Daily.

5 13 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
David Wojick
March 10, 2023 10:09 am

More by Maize on nukes:

20 something years ago I was a stringer for Ken’s Electricity Daily. He taught me how to write news and OpEds.

March 10, 2023 10:34 am

#1. “Scaleup problems may be irrelevant, as HTGRs have operated successfully at small sizes.“.

It does seem reasonable, if you are aiming for SMRs, to build and use –S-MRs.

#2. “In light of the past experience with nuclear promises, the only sensible attitude is to wait and see how many of these announced plans will, even with the added incentive of accelerated decarbonization, become actual working prototypes, and then …”.

Same for wind and solar. Try them out first. eg. have a town or region volunteer to test a wind- and solar-only supply, then wait and see how cost-effectively it works. That would be so much more sensible than pumping vast quantities of working people’s money into nationwide operations and demolishing reliable generators before you know if what you are doing has any chance of working.

Dave Burton
Reply to  Mike Jonas
March 10, 2023 11:06 am

Yes, it is not as bad to have only one town predictably freeze in the dark on windless winter nights, than to have the whole nation do it. Perhaps the good citizens of Rio Linda, CA would volunteer.

Last edited 17 days ago by Dave Burton
Reply to  Dave Burton
March 10, 2023 4:50 pm

Is that where California’s governor lives?

The governor and the legislature should be where the renewable energy test occurs. With a strict ban on backup generator usage.

March 10, 2023 11:06 am

All of the reactors in the article still use solid nuclear fuel, although in different configurations. To me the exciting developments in nuclear reactors involve liquid fuels (molten salt reactors) and not solid fuels. A molten salt demonstration reactor was successfully run at Oak Ridge for ~20,000 hours. Thorcon is building a new reactor based on the Oak Ridge design for Indonesia at the commercial scale. Exciting things may be closer than you think.

Curious George
Reply to  Deadrock
March 10, 2023 11:18 am

The Oak Ridge Thorium reactor actually ran on U233, was not designed to generate power, nor did it generate any. Success, indeed.

The Dark Lord
Reply to  Curious George
March 10, 2023 2:10 pm

thats why he said a demonstration reactor … not a power plant …

Reply to  The Dark Lord
March 11, 2023 8:40 am

Ya’ll would do well to see what this Experiment achieved…

Eric Vieira
Reply to  Deadrock
March 11, 2023 12:34 am

A liquid fuel system also allows for a configuration with moderated and non-moderated zones and allows for continuous elimination of impurities from the fuel. The volume change also allows for autoregulation of the fission reaction, and a fusion plug assures inherent walk-away safety. The main issues here are corrosion with fluoride salts, although progress is being made. There’s also a “dual fluid reactor” in development using this concept.
The costs for a prototype are always higher (development costs), but if everybody waits until someone else does it, then nothing happens.

Last edited 16 days ago by Eric Vieira
Rud Istvan
March 10, 2023 11:45 am

I wrote my summa thesis in 1972. It expanded Leontief static ‘snapshot’ input/output analysis to be dynamic over time, then used the new tool to compare coal, gas, and nuclear generation economics (capital, fuel, operations). Nuclear was never economic under any scenario. Too capital intensive. It appears not much has changed since.

Beta Blocker
Reply to  Rud Istvan
March 10, 2023 1:08 pm

In a truly competitive energy market, going with nuclear is strictly a public policy decision made for purposes of gaining some measure of energy security and reliability — for which one pays a premium in comparison with coal or natural gas.

The question which public policy decision makers must answer is this: how much of a price premium should energy consumers be willing to pay for the energy security and reliability nuclear provides?

The real cost of doing any large industrial construction project in the United States has approximately doubled over the last thirty years. The cost of doing nuclear construction has approximately quadrupled.

At the speed which inflation is moving, it’s not unthinkable that the costs of all types of large-scale industrial construction done in the US will double once again, but in a period of a decade or less.

Competition for the industrial commodities needed for the Net Zero transition will play a significant role in the future cost inflation of all types of industrial construction.

Last edited 16 days ago by Beta Blocker
Leo Smith
Reply to  Beta Blocker
March 12, 2023 6:14 am

That is simply not true. It is regulations that has driven up the price of nuclear power, regulations deliberately encouraged by its compeitors.

In principle building a reactor is no different from building a coal furnace and takes no more materials, and the fuel is *cheaper*.

But the paperwork runs to a whole library.
SMRs are shooting for around $10bn per GW, which amortized over 60 years with a putative 7.5% interest rate even assuming no capital is repaid in that time and with a 90% capacity factor as baseload works out as follows:

$10bn actual cost/ 60 = $166.7m per year capital depreciation
$ 750m per year interest charges

So $916.7m per year total cost of capital .

This shows that the capital cost is both dominated by the interest rates on the bonds raised, which is a function of how secure the market thinks nuclear is, and on the reactor lifetime.

But our reactor will average 9GW at baseload and so generate 9 x 24 x 365 GWh or 76TWh per year

$11.627 per MWh. That is the amortized cost of capital per MWh.

Now we can add another 10% for decomissioning, 15% for fuel and 10% O & M, and we get to $14 per MWh.

About one tenth of renewable energy costs.

And if the government underwrites it, the bonds become effectively gilt edged and you might halve that interest rate.

But the crucial factor is how much capital needs to go in

EDF’s latest abortion the EPWR is now reckoned to have a final bill of $40bn, for 3.2GW at Hinkley Point.

So $10bn per GW is not far off. Of course they will be paying out interest on the loan ten years before getting any return due to prolonged build times.

EDF have contracted the power at about $100/Mwh

Rolls Royce are shooting for around $6bn/GW and half the build time with their SMR.

Now they claim an operating cost overall of around $60/MWh so its clear that the capital cost is not dominating the equation. In neither reactor.

And we know that the fuel costs are negligible so it has to in fact be the O & M costs that are the highest.

Reply to  Rud Istvan
March 11, 2023 8:43 am


I respect your knowledge and experience so a serious question.

When reactors vessels can be built to last for 100 years or more, and if more highly enriched fuel is used so that refueling and the associated time offline is reduced, would not nuclear then be cost effective?


Leo Smith
Reply to  Drake
March 12, 2023 6:17 am

See my reply. The costs being levied by nuclear operators are between one half and one third of ‘renewable energy’ and I cannot make good Ruds claim that the capiital cost is outright the killer.

More like te protracted build times are.Which is why SMRs with one third to one half the ‘go, to grid’ times are infinitely preferable.

Leo Smith
Reply to  Rud Istvan
March 12, 2023 5:36 am

At todays prices it is deeply profitable.

As evinced by those running it.

John Hultquist
March 10, 2023 12:02 pm

A short few years back a “colonel mosby” (?) commented on every post
that had anything to do with energy that these small reactors were
going to save the earth and cure acne. That person seems, like Rip Van Winkle,
to be taking a nap. 😊

March 10, 2023 12:13 pm

Thorium liquid salts cooled reactors are covered on YouTube…simply search it……and “story tip”.

March 10, 2023 1:12 pm

We need some clarification, is this article only about SMR? Number two we have had more than a couple articles here at WUWT pointing out that levelized cost of electricity is not the appropriate measure for cost, so why should it mean anything to us in this case? We know how to build nuclear power plants, we have been using them for decades. The newer plants are superior to the older plants. I fail to see the relevance of this article.

Beta Blocker
Reply to  Bob
March 10, 2023 2:32 pm

Going with nuclear power is strictly a public policy decision. As is going with wind and solar. Left to its own devices, the power market in the US would shift decisively towards gas-fired generation. No wind, no solar, and no new-build nuclear.

The article is relevant because the public policy decision makers who make the energy resource decisions in the US pay close attention to what new-build nuclear costs. They don’t pay all that much attention to what new-build wind and solar costs.

The factors driving the rising capital costs for SMRs are the same ones driving the rising costs of all large-scale industrial construction projects.

For example, the UAMPS customer for NuScale’s SMR design says that increases in the producer price index in the past two years have raised the cost of fabricated steel plate by 54%, carbon steel piping by 106%, electrical equipment by 25%, fabricated structural steel by 70%, and copper wire and cable by 32%.

UAMPS also notes that the interest rate used for the project’s cost modeling has increased approximately 200 basis points since July 2020. The higher interest rate increases the cost of financing the project, raising its total construction cost.

All of this price inflation will impact new-build wind and solar just as it is affecting new-build nuclear. As will inflation in the cost of money.

Last edited 16 days ago by Beta Blocker
Reply to  Beta Blocker
March 11, 2023 12:56 pm

Okay, we have many issues here. Number one is this article referring to SMRs? The answer is either yes or no. Number two why are we discussing Levelized Costs of Electricity when previous articles have clearly shown it isn’t relevant?

You say:

“For example, the UAMPS customer for NuScale’s SMR design says that increases in the producer price index in the past two years have raised the cost of fabricated steel plate by 54%, carbon steel piping by 106%, electrical equipment by 25%, fabricated structural steel by 70%, and copper wire and cable by 32%.
UAMPS also notes that the interest rate used for the project’s cost modeling has increased approximately 200 basis points since July 2020. The higher interest rate increases the cost of financing the project, raising its total construction cost.”

All of these costs would apply to any power generator not just nuclear. If nuclear isn’t cost effective how do the French manage and how do we manage here in the US and other countries? Yes it’s true nuclear has expenses other generators don’t but much of that is political.

Sorry you have not clarified the problems I have with this article.

Leo Smith
Reply to  Bob
March 12, 2023 6:25 am

Well I have to say that WUWT is not generally staffed by engineers or cost accountants. Of course levelised cost is what finally matters, the problem is that the renewable energy lobby do not quote levelised costs when THEY say levelised cost, they ignore all the externalities like backup, stability cost, grid increased costs, subsidy and the like.
The true levelised cost of nuclear is way lower than the $160-$300/MWh of wind and solar including externalities and subsidies.

Nuclear is currently coming in, in the $60-£110 /MWh range. The only externalities there are are fuel disposal and decomissionsing and those are priced in to the nuclear costs.

The great myth is that nuclear is ‘too expensive’ It isnt.

Traditional coal is profitable at about $50/MWh , and gas used to be aroumd $60/MWh until global gas prices went through the roof.

Now its being sold at $100-$200!

March 10, 2023 1:35 pm

These anti-nuclear types are so fanatical! Too bad they are so effective, as fission is by far the best energy option for the world now.

As for nuclear waste: you may have noticed noise about “forever poisons” these days. It finally occurred to some folks that poisons buried in landfills NEVER get better yet no one talks of posting guards for eternities around the burial sites. So nuclear waste decays over time and that’s a problem? It’s actually a feature! Unlike “forever” chemical waste, nuclear can be held for a few years to reduce it’s toxicity and then buried just as if it’s a “forever” chemical waste. Poison is poison and the volume of nuclear waste is tiny compared to the immense volumes of equally toxic chemical waste routinely disposed of.

Last edited 16 days ago by Kasmir
Eric Vieira
Reply to  Kasmir
March 11, 2023 12:48 am

Very often, nuclear costs are inflated for political reasons. For example in Europe, the dismantling of a nuclear power plant is at least ten times higher than in the US.
Why ? Because instead of waiting a few years till the radioactivity (via neutron activation) of the installation drops, which then allows for “normal” demolition procedures, they insist on dismantling it immediately with the ensuing stringent security measures, radioprotection, authorisations, red tape, yada, yada … Then, the greens can say: see how expensive it is !!! But when a wind turbine are decommissioned, they leave the 300 or so ton concrete foundations in the ground. They don’t restore the sites ! And the blades are untreatable waste.

Last edited 16 days ago by Eric Vieira
Leo Smith
Reply to  Eric Vieira
March 12, 2023 6:28 am

Not quite true, at least in the UK. Her we are letting the old Magnox reactors cool down for a decade or two before cutting them up BUT it is true to say that the sites are being restored to ‘greenfield’ – no industrial waste AT ALL. Suitable for agriculture or housing. Or new nuclear power stations one hopes

Kit P
March 10, 2023 3:29 pm

Not sure what the point is! I suppose every industry has armchair quarterbacks.

I suspect that we who build power plants would never have any problems if we had the same crystal ball.
SMR are a solution looking for a problem that does not exist like solar and wind.

There is a very good reason to build nukes with huge output: no coal, no gas, no oil no choice.

Beta Blocker
Reply to  Kit P
March 11, 2023 6:59 am

In the circles I travel in, the advocates of the large 1,100 MW reactors similar to the two AP1000s soon to enter service at Vogtle 3 & 4 are already saying that rapidly rising costs for industrial materials is an argument for abandoning the oncoming SMR technologies and going back to the traditional bigger-is-better for commercial power reactors philosophy.

The argument goes like this. We will be getting more megawatts per unit of industrial materials committed to the plant — the concrete, the fabricated steel plate, the carbon steel and stainless piping, the electrical equipment, the fabricated structural steel, the copper wire and cable, etc. etc. etc.

My response to this argument goes like this.

With completion nearing for Vogtle 3 & 4, we now have a very extensive record of costing information for predicting what it will take in time and money to construct another two-reactor nuclear power plant using AP1000 or similar size 1,100 MW reactors.

By all means, we should develop a new estimate for another two-reactor project using real world costing information taken from Vogtle 3 & 4 and then apply the necessary cost escalations to the final bill-of-materials which came out of that project.

Vogtle 3 & 4 cost roughly 30 billion dollars. My guess is that if a new estimate were to be done in today’s highly inflationary environment, another two-reactor project using 1,100 MW size reactors might cost in the range of 45 to 50 billion dollars.

There is only one way to know for sure, and that is to perform a disciplined and thoroughly honest cost & schedule estimate using the Vogtle 3 & 4 information as the basis.

Leo Smith
Reply to  Beta Blocker
March 12, 2023 6:41 am

Manufactuers of large reactors have an agenda. Material cost is the smallest cost of a new reactor. It is pretty much peanuts. All of te copst is in man hours dealing with regulatory hurdles, and actually bulding out the concrete.

Leo Smith
Reply to  Kit P
March 12, 2023 6:39 am

SMR are a solution designed specifically to address two serious problems that do exist:

  • Cost of building to regulations. The whole reason to build small units is that they can be pre-assembled in a factory under controlled conditions to a type approved specification. And they are under the limit of the size where passive cooling cannot be employed in a SCRAM situation. So emergency cooling if they get disconnected from the grid is not necessary, and that saves a huge amount of cost and complexity and regulatory approval. SMRs will simply not be able to do a Fukushima or a 3MI.
  • Time to build. Factory assembly means that a company needs merely to lay down a concrete pad, add some steam plant and turbines – all off the shelf kit – and drop in a completed reactor assembly. Long build times cost the interest you pay on the money you borrowed before the plant is generating profit.

Large plants are essentially hand built on site and require much greater safety kit to stabilise them.

SMRS are the way of the futire – mass produced at low cost. And able technically, if not politically, to be sited wherever there is need for them Integration with other technologies. Like aluminium smelters with say heated greenhouses attached. Or even in the middle of towns where the hot water provides free heat in winter.

Eric Vieira
March 11, 2023 12:24 am

An HTGR/TRISO system was shut down in Germany for political reasons (it was launched during the same time as the Tchernobyl incident), although it was working fine. It has the same inherent fault that fuel pellets have: the buildup of long half-life actinides which poison and stop the reaction (by absorbing neutrons), and which stay in the fuel pebbles. Although it
might be possible to configure the reactor to have a moderated and a non-moderated zone
with fast neutrons were the actinides could be fissioned.

March 11, 2023 8:21 am

Great writeup.
Sad considering the speed and scale of the 164 (full size) nuclear reactors China will build by 2050. One every 2 months vs. 1 every decade? or so in the US,,,

Kit P
Reply to  c1ue
March 11, 2023 2:31 pm

Someone does not have a clue that China and the US are different countries.

I was part of the new reactor building boom 50 years ago in the US. The US is now a leader in keeping 50 year old power plants running.

I was also part of the boom in China. My last power plant before retiring is
retiring was two 1600 MWe reactors in China. These plants were built because China could no longer produce enough slave labor coal.

China is done. China imports large amounts of energy and food using sea lanes kept open by the US Navy.

March 11, 2023 8:53 am

If ya’ll actually want to know why there’s no affordable nuclear today – see:




Gary Pearse
March 11, 2023 12:46 pm

If the world needs a reliable, low-cost, trouble-free reactor, that lasts 40yrs, can be refueled without shutting it down, it’s been available from Canada since the 1960s. The CANDU is modular (can be cranked out like automobiles!), uses heavy water and non-enriched yellowcake!

Moreover, it captures surplus neutrons causing part of the U238 to undergo fission adding to the thermal power. The latest improvement several years ago had fuel costs of 3 cents/kWh (USD 0.025/kWh! A nuclear accident is a worker spilling a cup of coffee!

The modular unit takes 3yrs to build at a cost (a couple of yrs ago) of $3.5 billion (USD 3B). There are no cost overruns! The world’s largest plant for 25yrs was at Bruce Point, Ontario, made up of 9 modules! Oh, and the large Pickering plant was is built in a suburb of Toronto in the 1960s! Only chauvinism stops the spread of this cuddly technology

There is a lot of excitement about new (?) Th salt reactors these days but never a mention of its Canadian inventors! The Chalk River nuclear research facility (which developed the CANDU) built the first one of it’s kind in 1947!!! It was finally shut down in the 1990s. They assisted Oak Ridge in the US to to build one in the early 50s. Check us out.

Leo Smith
Reply to  Gary Pearse
March 12, 2023 6:48 am

The fact is that any and all of these reactor technologies work just fine, don’t cost material wise much different, and are more than safe enough.

I liked the Candu, but the UK government were not prepared to support them.

SMRs exist to confront the artificial safety barriers that have been placed in the way of larger reactors.

We urgently need fast build reactors. Conventional 1.2GW reactors cannot be built in a factory and they all need active cooling under fault conditions. 500MW reactors or thereabouts can have passive emergency cooling and suit the standard 500-600MW turbines and generators that are standard issue, and smaller units are great for industrial plant that has high energy needs. Or portable use.

SMRS are needed.

Leo Smith
March 12, 2023 5:29 am

Contrast the investment in wind and solar which is ‘nameplate gigawatts, no matter what the levelised costs’ Currently standing at between $150 and $300/MWh.
Remember renewables are no threat to fossil fuel. Nuclear is, Big time. Which is why all the oil majors are overtly investing in ‘green’ power and not nuclear.

Hatter Eggburn
March 13, 2023 3:11 pm

Nuclear power has no future in democracies.
Democracies have no future.

%d bloggers like this:
Verified by MonsterInsights