A few months ago, a STEM and economics literate global decarbonization executive for a $4 billion annual revenue logistics business operating in most of the major trading companies of the world asked me "What drives this madness on hydrogen?" They were being peppered with irrational proposals for the molecule and its derivatives. I answered at length with the motivations that were creating cognitive biases among the self-supporting circle of fossil fuel firms, their financial firms, the countries with high oil rents and firms whose only intellectual capital is in technology which burns fossil fuels like internal combustion engine manufacturers.

This week I had an equivalent question from someone engaged by the Bulletin of the Atomic Scientists. The individual isn't an industry insider and didn't have a particular iron in the fire, but had been engaged to write a Bulletin piece on the perplexing enthusiasm small modular nuclear reactors (SMR) are seeing from an overlapping circle of advocates and firms.

This was, of course, in the week when NuScale inevitably imploded, a story I'll return to as I unpack some of the motivations behind those thinking a bunch of lab technologies that have been around for decades that depend on uranium from Russia, that don't have the physical characteristics for cheap nuclear generation and don't have the conditions for success for nuclear generation will be the saviours of the nuclear industry and a key wedge in fighting climate change.

The Bulletin, as a reminder, is the non-profit organization founded in 1945 by Albert Einstein and several of the scientists involved in the US creation of the atomic weapons detonated in Japan at the end of World War 2. Its founders realized that the weapons technology had the potential to destroy the human race and decided that informing the public and policy makers about the dangers was a good idea. Their most high-profile publication is the Doomsday Clock, a metaphorical meter of how close the human race is to destroying itself. They've branched out to a lot of science communication on other threats like climate change, AI and COVID, but still have a core focus on nuclear energy and weapons.

Some bona fides. I like nuclear generation. I know it's safe enough. I'm not concerned about radiation. I know it is capable of 90% capacity factors and limited load following. I've published assessments on nuclear vs renewables scaling in China that have garnered reasonable attention, material on the prerequisites for successful nuclear programs (and why those prerequisites don't exist in the 21st Century), a bunch of other basic nuclear grounding pieces and a fair amount about why small modular nuclear reactors are head scratchers. I just know that it doesn't have the conditions for success to be built and scaled economically in the 21st Century, and wind, water, solar, transmission and storage do.

The Bulletin representative undoubtedly saw something I'd written and put me on the long list of people to interview. We had that conversation this week, and it's worth writing down the perspective I shared with them.

Keep in mind the following. Everything you read after this point about why different groups are supporting small modular nuclear reactors and nuclear energy in general is my opinion. Many of the individuals or groups would deny vociferously that the motivations and logical fallacies that I'm going to attribute to them are accurate, even as they would feel a frisson of cognitive dissonance that tells them something is not quite right. In most cases, these people and groups believe that they are correct and that their positions are rational and carefully thought through.

And so to a tangled web that includes Bill Gates, Silicon Valley, desperate coal towns, desperate nuclear towns, the inability of the USA to build big infrastructure, the US Department of Energy's budget, magical thinking and more.

What are the seven conditions for success for nuclear energy?

There have been successful nuclear generation deployment programs in the world. The USA, France, Canada and South Korea managed it, more or less. Given the embarrassing track record of massive cost and schedule overruns that have led to nuclear generation deployments being 23rd of 25 categories in megaproject expert Bent Flyvbjerg's 16,000+ dataset of programs sorted by likelihood to succeed, what worked in the past?

They were national strategic programs. The strategic programs were aligned with nuclear weapons programs. The government picked and enforced a single design for all of the reactors. The reactors were GW-scale due to thermal efficiencies required for cost effectiveness. The government ran human resourcing. The programs ran for 20 or 30 years. They built dozens of nuclear reactors to maintain the teams and momentum and to share lessons learned.

This is obvious stuff looking backward from 2023. As I noted recently, nuclear energy and free markets aren't compatible. Nuclear programs are state programs with subordinate corporate partnerships.

As a note, China couldn't recreate the conditions for success despite having every ability to so. Their nuclear program peaked in 2018 with seven reactors achieving commercial operation but has been averaging three reactors a year since. This year the single reactor that's been connected to the grid may not achieve commercial operation. In my assessment, their industrial export strategy led them to build too many technologies and designs of reactors instead of rigorously enforcing a single design, hamstringing the deployment and scaling effort.

No country globally has the conditions for success for nuclear generation in the 21st Century. That was a Cold War era success story based on a hyperawareness of the threat of nuclear war which is vastly diminished in the age of trade.

What are small modular nuclear reactors?

The name kind of gives it away. The vast majority of the 450 or so nuclear reactors in the world that are generating electricity that gets put into the grid are gigawatt scale. That's because attempts in the 1950s and 1960s to generate electricity with the scale of reactors on nuclear subs and aircraft carriers found that they were really expensive at that size.

But gigawatt scale reactors are easy to turn into bespoke engineering megaprojects. Custom engineering for every site foregoes economies of manufacturing scale.

So the SMR crowd decided to ignore most of history's lessons about both the scale of reactors for commercial success and the conditions for success and lean into tiny reactors and lots of numbers. The hope was that Wright's Law — where every doubling of the number of manufactured items once in production manufacturing would bring cost per item down by 20% to 27% — would enable them to be manufactured and deployed cheaply. However, the doubling requires an awful lot of reactors, and only under the most unrealistically optimistic of scenarios are they in the price range of wind and solar today by 2040.

The last condition for success was repeating the deployment a bunch of times, but historically that's been dozens, not thousands. Rigorously repeating deployment of the same proven design with skilled teams in a relatively short timeframe with lots of national control and military rigour isn't a feature of the SMR bandwagon and proposals.

No, the SMR crowd think that there's a free market friendly version of nuclear generation. They mostly ignore the seven layers of overlapping security required for any commercial nuclear generation solution. They ignore the thermal and hence cost efficiencies of scale. They ignore the advantages of proven, simpler technologies in favour of novelty. They ignore the lack of military interest in a modern strategic nuclear program. They ignore that they don't have a major economy and geography backing a winner and forcing it to success, no matter what it costs.

And clearly they don't pay much attention to modern competitors like wind and solar energy.

There's almost nothing new in the SMR technologies and designs by the way. Some of them were operational in labs in the 1950s. Most of the technologies were never commercialized anywhere despite various attempts. Their claims about being safer are pretty meaningless as passive safety features exist on lots of the operational GW scale reactors and no one is building unsafe reactors anymore.

One thing that is relatively new is that most or perhaps all of the designs require high assay low enrichment uranium (HALEU). That would be fine, except that Russia had effectively been ceded a monopoly on the supply chain for the fuel. The US DOE (again) has created a consortium model to try to create a reasonably large domestic supply of the material, well above the scale currently available domestically for laboratory reactors.

So why are people paying attention to SMRs? That's the nub of it, and there are a lot of different answers, just as with hydrogen for energy. There's a circle of hope, delusion and despair that is grasping at SMR straws. Let's pull it apart.

2001 is calling

In the early 2000s, wind, solar and batteries hadn't seen massive global deployments and astounding cost reductions, and wind and solar especially had not been proven to be able to be managed to create reliable grids. Nuclear was the clear deployable low-carbon, low-pollution alternative. It was quite possible to have an informed, logical and defensible opinion at the time that nuclear would dominate low-carbon electrical generation. Memories of the successful nuclear deployments, if not why they were successful, were top of mind in people's heads.

Many people such as Michael Shellenberger and seemingly everyone involved with the Pritzer-funded Breakthrough Institute became convinced that lots of nuclear energy was required quickly and everything else was a distraction. There's no obvious connection between the Institute and Breakthrough Energy Ventures, but there's clearly an intellectual one.

Bill Gates, founder, funder and Chair of small modular reactor firm TerraPower founded the latter. He brought a set of other pro-nuclear billionaires along for the ride including Mukesh Ambani, Jeff Bezos who has money in fusion, Richard Branson who is a public advocate for nuclear, Chris Hohn whose hedge fund was invested in a Japanese nuclear utility whose assets included nuclear power stations, Vinod Khosla who blames environmentalists for nuclear's economic failures, Jack Ma who was instrumental in Chinese involvement in the UK's Hinkley Point C nuclear expansion and even Ratan Tata, former Chair of Tata Group who was instrumental in building nuclear energy in India.

Many of these people's hearts are in approximately the right place regarding addressing climate change, but billionaires are surrounded by a chorus of sycophants and so find it easy to end up having their opinions reinforced rather than challenged as data changes.

And by 2015 when Breakthrough Energy Ventures was founded, data had changed. Wind and solar were already being deployed at remarkable scales without running over budget or schedule. China's nuclear program was already struggling while its wind, solar and hydro programs were running well. It was clear to outside observers that the premise of Venture was flawed.

Shellenberger couldn't let go of his fixation on nuclear as the solution when data changed. So instead he started attacking renewables relentlessly, falsely and illogically. When that didn't work, and it was clear that renewables were the primary answer to decarbonizing electricity to address climate change, he became a climate change denier instead. It was impossible for him to let go of his public commitment and admit that he was wrong, and so he became even more wrong.

At that, he's done less harm than Breakthrough Energy Ventures. It continues to fund nonsense like Heirloom and AirLoom most recently, and Heliogen in the past. Gates' TerraPower is funded by US DOE money as much as anything else, and he goes cap in hand to Congress to demand that nuclear get more subsidies than it already does and that renewables be short changed.

The Ventures crowd are deeply influential. They have the ear of billionaires and national leaders. What they whisper major media outlets amplify. And they are simply wrong because they are stuck in the early 2000s and haven't updated their priors with modern results.

Their support for small modular nuclear reactors as something which is going to fix nuclear energy's modern inability to be delivered on schedule and budget in developed countries comes because they didn't analyze why nuclear programs were successful in the past and why SMRs won't fix that.

Silicon Valley energy 'disruptors'

Silicon Valley has done a lot of things in its innovation history. Early on, it did tremendous work with electronics and physical computers. More recently, it has exploited the internet mercilessly to both destroy and create value chains where data of any form was moving around or persisting. Quite a lot of what the Valley has done is destroy capital, but it has created some amazing firms and attracted a lot of talent. Some of the firms are even good things for humanity and many of the people try to have their hearts in the right places. And in the Valley, if something already exists there's a strong tendency to think it's obsolete and that novel things are by definition better.

Facebook and Google had stormed the world of data, media and entertainment. Moving fast and breaking things worked to make a lot of tech bros absurdly wealthy. Surely there must be another market ripe for disruption. And so they turned to energy. It was electrons. It flowed through wires. Obviously it was just as easy to disrupt as data!

Silicon Valley venture capitalists loved the idea of all the money sloshing around the world of energy. They saw the potential for all of the fossil fuel money flowing into their bank accounts and their kids' trust funds. A whole lot of nonsense technologies received a lot of funding, including a bunch of small modular nuclear reactor and fusion startups.

Yeah, not so fast. Electricity does not equal data. You can't compress or multiplex electricity. The grid is the biggest machine in the world, not a bunch of printing presses and hard drinking writers in New York.

The biggest machine in the world is like a very large crude carrier and the icebergs of wind and solar were already directly in its path and getting bigger every year. By not actually learning about energy, electricity and the grid, a bunch of Dunning Krugered up tech bros and women wasted a lot of venture capitalist money.

The actual disruptive technologies already existed for decarbonizing energy, so the tech bros discounted them at the expense of lots of wasted cleantech capital. The recent SPAC craze which has flung lots of money at bad ideas like Energy Vault and NuScale is just more of the same.

Conservative vs liberal tribalism and mental flexibility

Due to the relationship between nuclear weapons and nuclear generation, a lot of liberal, progressive people became opposed to nuclear generation. When the military was heavily involved in nuclear generation and nuclear technologies were being used for both fuel for generation and fuel for thermonuclear weapons, that wasn't an unreasonable position. The Bulletin was formed because nuclear weapons were clearly an existential threat to humanity.

It became a tribal conservative vs liberal thing, with conservatives favouring fossil fuels and nuclear power, and liberals favouring renewable generation.

On the liberal side, that meant that confirmation bias led them to accepting all sorts of nonsense about the dangers of radiation as fact, instead of the fiction it mostly was. On the conservative side, that meant that a deep opposition to renewables tended to spring up.

From a cognitive perspective, the studies are pretty clear. People with a liberal mindset are a bit better at changing their minds when new data comes along. People with a conservative mindset are a bit worse at that. So in the 2020s, there are a lot of people who were formerly opposed to nuclear generation who have stopped having palpitations about radiation and meltdowns, and instead are happy enough when nuclear plants get commissioned, but consider them more of a distraction.

Meanwhile, on the conservative side of things, there's still a lot of resistance to the empirical reality of massively successful, cheap and reliable renewables. And there's an inability to accurately diagnose why nuclear generation succeeded in the past because it conflicts with their biases. In the case of the billionaires, it interferes with their bias for the liberalized economics which have made them billionaires. It's very difficult to find an even fiscally centrist billionaire because tax cuts for oligarchs is just accepted as their due.

SMRs fall neatly into this void, promising that their favoured tribal technology will finally take off again with this one, simple change.

Conservatives kicking climate can down the road

Regardless of degree, conservative voters and politicians in every country in the world are less interested in actually doing something meaningful about global warming than progressive and liberal voters. The Republicans in the USA along with DINOs like Manchin don't even pretend to be interested, with the last two Republican Presidents walking away from global treaties on the subject and leaning into fossil fuels.

Small modular nuclear reactors have a lot of upsides for conservative politicians. They are all unproven, first of a kind commercial technologies so there's a decade of design, debate, development and regulatory work before any will be up and running. That's cheap money and keeps the fossil fuels burning longer. The promise of keeping dying or dead coal or nuclear towns alive wins votes. The promise of nuclear instead of renewables wins votes among their fans.

And they don't have to change anything. Actually delivering climate value is deferred so long that they can ignore the file for another decade if they insert the promise of SMRs into the mix.

And to be clear, if the conservative politicians were advancing a nuclear first policy with a clear understanding of the conditions for success and working hard to create those conditions in some country, I would be quite happy for that country. It's just that they aren't. It's delaying and deferring, not creating a national strategic program.

US Department of Energy thumb on the scale

A key point to remember about the US DOE is that 55% of its budget is related to commercial nuclear generation. The other 45% covers dams, geothermal, wind, solar, tidal, wave, biomass and biofuel energy.

There are two results from this. The first is that it's much easier to get big DOE money for small modular nuclear reactors than it is for any other form of generation, although hydrogen for energy plays are giving nuclear a run for its money right now. Gates' TerraPower and the just-failed NuScale both have received hundreds of millions of free money from the DOE in their lifespans so far. Like the continued efforts from the DOE around the perpetually failing technologies of perovskite solar panels, supercritical CO2 thermal generation and concentrating solar power, the DOE just can't help continuing to back bad ideas.

Which gets to the next problem. The DOE is a bureaucracy with all the pros and cons of that form of organization. It has internal politics. That 55% of budget for nuclear is jealously guarded and the source of lots of authority and influence. Bureaucracies persist and add to themselves. They don't voluntarily wind down.

And the people in that side of the DOE are watching a cliff looming toward them. Virtually every nuclear reactor operating in the USA is reaching its end of life in the next decade. What's coming is a century of decommissioning that's underfunded. That's not fun. That's not delivering lifeblood energy to a vibrant economy. That's cleaning up the mess it left behind.

They aren't stupid. They've watched the debacle of the Westinghouse AP1000 failures at Vogtle and Summer deployments. They've watched the European EPR program fail just as badly in Flamanville, Hinkley and Olkiluoto. They lived through the aftermath of the Fukushima disaster, which will approach a trillion USD in economic and clean up costs for Japan by the time the dust has finally settled decades from now.

They know that they can't get support to build more GW scale nuclear reactors. No one is buying that story anymore, although few are clear on the conditions for success. So SMRs are a big area of hope for them. 55% of the DOE's $150 billion budget pays for a lot of attention and air cover for a very low likelihood of success technology.

It pays for a lot of people to go to conferences with other like minded people where they agree with one another, usually politely but sometimes in heated ways, arguing about arcana that are irrelevant to the big problem SMRs face, that they just don't solve the industry's challenges.

Nuclear towns looking at the abyss

There are a lot of small towns with big nuclear plants in the world. Ontario's Bruce nuclear facility is located in the township of Kincardine, population 12,000, as one example. Bruce Power employs 4,000 people.

There are similar examples across the USA and Europe. The nuclear facility is the big industry in town. People who work there are paid a lot compared to rural averages. It's the major lifeblood industry. If the nuclear plant shuts down, the town withers and dies.

The town is looking at the looming cliff as reactors age out and are shut down. The county of 70,000 is looking at a huge part of its rural economy disappearing. Provincial politicians and bureaucrats are aware of the risk and don't like it because pockets of affluence turning into pockets of poverty is a headache. Conservative politicians especially are looking to preserve those rural votes because that's how they get elected.

So while the nuclear facility on the other side of Toronto, in Pickering with its population of 100,000 and lots of nearby employment opportunities in the manufacturing centres of Whitby and Oshawa and light rail for commuters into Toronto, is mothballed after a short extension, Bruce Nuclear gets a new SMR project that's going to 'save' the facility and the town.

To be clear, if SMRs made sense, existing nuclear power generation facilities are the place to build them. They are already at the centre of the seven overlapping layers of defence that nuclear generation sites require from the international, all supply and waste chains and the physical and electronic security of the facility itself.

Unfortunately, SMRs don't make sense, so it's false hope for the community. But the provincial Conservatives locked in Bruce County votes for themselves by starting to explore the design and permitting of a Brookfield Westinghouse SMR technology that's never been deployed. And they cemented their ties to Brookfield, a global powerhouse. And they threw scraps to Canadian firms SNC Lavalin — a perpetual receiver of corporate welfare which owns the CANDU nuclear assets as well as a history of corruption that's hard to believe — and some Canadian construction firm called AECON, not to be confused with the competent global firm AECOM. I'm sure that there's a story there too. The politics of the corporate ties are barely obscured.

Coal towns looking at the abyss

There are a lot of towns that used to have great work digging coal out of the ground near the town, processing it and putting into coal generation near the town. Delta, Utah springs to mind. It's a town of about 3,600 people in Millard County, population 13,000. About 400 people work in the coal plant.

Coal isn't facing a cliff, it's falling off the cliff already in the USA.

Irrational behaviour is ensuing. The first major loan that Jigar Shah, head of the loans program office at the US DOE was able to get through and announce with much fanfare was $504 million for the Delta coal facility. The premises of the loan make absolutely no sense. The town's electrical generation goes into transmission wires that go over the mountains to LA, 500 miles as the obsessive compulsive crow flies — seriously, did whoever created that idiom never watch any crows? — and the first premise and the only one with the tiniest bit of reality to it is that they are going to reuse that infrastructure.

Sure, but what's required to reuse that infrastructure? Well, they have to build big natural gas pipelines from somewhere into this tiny town in the middle of nowhere. They have to rip out the coal boiler and generators and replace them with "hydrogen ready" natural gas generation units. They have to build a lot of renewables around the town somewhere and transmission into the town to bring all of those green electrons to it. They have to build a massive industrial green hydrogen electrolysis facility in the town. They have to seal some natural caverns under the town and turn them into a huge geological hydrogen store that doesn't leak much.

In operation, they have to take the useful green electricity and throw away a bunch of it to turn it into hydrogen. When they put the hydrogen into the "hydrogen ready" gas generation units. This turns cheap electricity into expensive electricity.

This does preserve some power engineers and administrative jobs, but all the coal miners are out of work. All of the people who know how to run a coal plant have a lot of retraining before they can run a gas plant, and a lot more before they could run a hydrogen plant and storage facility. The timing suggests most of them will just retire instead of transitioning, to no human resource benefit whatsoever.

Why, by the way, do I keep putting "hydrogen ready" in quotes? Because most of the time natural gas burning units are hydrogen ready like your driveway is Lamborghini ready. That it's possible to park a fictitious future and very expensive Lambo in it does not in any way mean that you will be able to afford to do so.

What does all this hydrogen nonsense have to do with small modular reactors? Well, coal towns look at SMRs as a convenient technology to dump on the coal plant site. That actually makes more sense than the Rube Goldberg hydrogen nonsense, if SMRs were a solution. As Shah and others keep pointing out, coal plants already have big areas, security and transmission.

But coal plants don't have seven layers of overlapping security, with five layers inside the state. Fences at coal plants are to keep thieves, drunk people and Greenpeace protestors out, not much more. Nuclear facilities have federal oversight and security and military level ordnance onsite.

Nonetheless, expect lots of nonsense.

In the USA, the devolution of the ability to say no to the individual

Back to the US DOE. Why is it giving $504 million to a clearly ludicrous scheme whose only merit is keeping a tiny rural town with no future alive? In the USA, they've lost the ability to do big infrastructure. The political scene has devolved the right to say no down to the county and often the individual level.

As such, the only things that the US DOE can fund with loans are things which have already received that hyperlocal support. Bizarre assembled blocks of this and that like the Delta hydrogen play are the result, not rational regional transmission infrastructure for wind and solar, electrifying rail and an acceptance of sunk costs.

Instead of learning the primary lesson of the hollowing out of various parts of the USA due to the combination of China and automation eliminating factory jobs over the past 40 years, which is that paying people to leave dying towns in a combination of relocation funds, a social safety net, retraining and support in their new community, the USA is stuck with faint hope plays to keep the dying towns alive. For a good presentation of this, the great book Good Economics for Hard Times by Banerjee and Duflo has a good section on the very effective and very poorly funded Trade Adjustment Assistance program. Dying towns are a lot stickier than classical economics indicates, and how to deal with them is now well understood but politically unpalatable.

Compare this to the New Deals during and after the Great Depression. Massive dams. Massive transmission. The interstates. Hugely positive transformation programs run by the government. Not allowed in 21st Century USA. Of course, lots of inequity in those programs as well.

This is playing out in smaller rural centres around the developed world. Automation has radically reduced human labor requirements for extraction industries like mining and agriculture. There are fewer and fewer jobs in rural areas every decade. Cities are where people have significant economic, social and marital opportunities. Increasingly the people in rural areas are the ones without the ability to change and take risks, simply because anyone who can think remotely strategically and has the slightest degree of acceptance of risk leaves as soon as possible.

Combine the irrationality of dying towns and counties with the ability for those counties and the states that they are in to block actually sensible transformation, and the US DOE ends up giving loans that will provide hope but no other value.

Hence the madness driving small modular nuclear reactors

It's easy to see that assembling a group of stakeholders sufficiently large to create the appearance of movement and success out of these diverse organization isn't necessarily trivial, but is certainly possible. NuScale found willing partners in an overlap of this space. They had the US Department of Energy giving them lots of money. The US DOE also donated a bunch of space on a DOE nuclear-oriented national lab's grounds in neighbouring Idaho. It included the Utah Associated Municipal Power Systems (UAMPS) organization as a sponsor to get a bunch of small Utah cities and towns aligned to buy the electricity.

Utah and Idaho are both hardcore Republican states, so that was an easy sell. Utah Senator Mike Lee is so strongly opposed to climate action that he brought flip boards with images of Ronald Reagan riding a dinosaur to Congress to attack the proposed Green New Deal.

NuScale found some Wall Street SPAC money guys who were excited at the thought of pumping and dumping the stock of the company, which they succeeded in doing in August of 2022. NuScale's market capitalization is off 80%, but rest assured, the SPAC founders made out just fine, leaving the follow on retail investors holding the bag.

This is a coalition of deeply motivated thinkers and organizations piling hope upon fantasy upon irrationality upon ideology upon venality. The odds of NuScale succeeding were always slim to none. This is true for every SMR proposal out there.