Leslie Groves: Are you saying that there's a chance that when we push that button... we destroy the world?
Robert Oppenheimer: The chances are near zero...
Leslie Groves: Near zero?
Robert Oppenheimer: What do you want from theory alone?
Leslie Groves: Zero would be nice!
Yep I went to watch Oppenheimer…not Barbie though. Or not yet! Did I like it? Yes, it’s a good movie, but not one you go watch for the suspense. Nolan did a great job but I have to say that, unsurprisingly, the first moment that really grabbed my attention was when Oppenheimer asks Hans Bethe to recalculate the odds that they were going to blow up the whole atmosphere. Cut. Scene. Hans Bethe comes back confirming that the probabilities for an atomic bomb explosion to ignite the atmosphere and destroy the world are “near zero” (nothing is ever zero in quantum mechanics). Near zero…net-zero... If you work in the energy sector, I guess you know where I am going with this. Not yet though. There are a few turns I want to take and I’ll try to create a bit of suspense, or possibly disorientation, before making a point.
Near zero can mean a lot of things, but if we remain in the realm of possibilities, aka statistics, near zero can be thought as something whose occurrence is close to impossible. As Nassim Taleb would define it, a black swan: an event whose likelihood is infinitesimal but whose effects can be so impactful as to completely change the previous paradigm or, as in the case of the atomic bomb, to destroy life on earth. Regardless of the fictional nature of the dialogue in the movie, what kept me thinking is the abyssal difference between the quest to arrive at the atomic bomb first, even with an infinitesimal possibility of blowing up the globe, versus the current struggle we are facing in actually agreeing and converging towards a net-zero economy.
You may argue that you can’t compare WWII to the energy transition. Fair point. We tried to compare Covid to a war and it did not really work out well, did it? But building on the recent experience with Covid and the vaccines, I struggle to see how, in a society where public opinion holds its weight in the political discourse, ordinary people would agree to carry out an experiment that has a non-zero change of destroying the world.
I challenge everyone to find a person to consider that an acceptable risk. Committees, parties, pundits, the media circus, every form of conspiracy theory would force researchers to wait until they understood the science better and could be 100% confident that their project wouldn't ignite the atmosphere. And even then, there will still be groups of people, either for interest or simply due to a different belief, who will question that 100%. Apparently, we can't even seem to agree on the scientific definition of 'near zero'.
BANANA
Taking Covid aside, the most striking example of this is nuclear energy. After decades of operations, major safety improvements, and technological advancements, public opinion across great part of the Western world remains highly divided on the topic.
If we look at Europe1, the current energy crisis has flipped the balance between those in favor and those against nuclear energy: in 2016, 41% rejected the technology while 15% supported it. In 2022, the proportion of those in favor of nuclear power increased to 40% and while the share of those against it decreased to 15%.
In the US2, 76% of public opinion is now in favor of nuclear energy, against a 24%. However, as in the EU, the percent of people strongly in favored is below 30%, against a 5-15% strongly opposed. The majority of the public opinion can be considered fence-sitters, prone to changes.
The main arguments cited by those in favor encompass affordability, reliability, efficiency, and environmental benefits related to climate change. Energy independence and energy security, while dominating the policy discussion after Russia's invasion of Ukraine, are somehow less relevant themes, or at least not directly mentioned.
Public perception of new technologies is generally affected by perceived benefits, costs, risks, and the perceived ability of governments to reduce these risks. When the dimension of the project grows, and some of those attributes are not directly assessable, confidence drops. Most of us are now familiar with the acronym NIMBY (Not in My Backyard), which refers to the reluctance of the local population toward having things like power lines, power stations and pipelines being built near neighborhoods or businesses. It’s been trendy for decades but as all trends, it’s now being replaced by another catchier acronym, BANANA: Build Absolutely Nothing Anywhere Near Anyone. And while we may have found some good solution to solve NIMBYism, we may now need to understand how to peel the banana.
Regardless of Nimbyism or Bananism, which may be even less relevant for a nuclear power plants according to a recent study that call for a “reverse-NIMBY” (though there are also contrasting evidences), public opinion plays a critical role in driving policy choices and thus determining the viability of nuclear investments. But is public opinion the only obstacle to a new nuclear renaissance? Spoiler alert: it’s not.
Nuclear challenges
The OECD NEA published a report in 2020, “Unlocking Reductions in the Construction Cost of Nuclear Power Plants,” where they broke down the strategies for nuclear cost reduction into four categories: design and supply chain maturity, regulation stability and predictability, effective project management, and policy frameworks.
Design and Supply Chain Maturity: The construction of nuclear plants often begins before the design is complete, leading to design changes during construction that can cause delays and increased costs. Which is basically saying you start building the house while the architects are still arguing over the floor plan. Not ideal. The recipe is to use a consistent design for multiple reactors which result in cost savings due to learning-by-doing, like France did during the 70’s and the 80’s. A mature nuclear supply chain and an experienced workforce are crucial for cost-effective construction. It's the difference between building a Lego set with the instructions versus making it up as you go. Let’s remember the Lego because it will come back later on.
Regulation Stability and Predictability: Regulatory stability is vital for achieving design maturity and scale construction. Plant designs can change in response to evolving regulations, thus leading to higher costs and project delays. Predictable regulatory changes are necessary, with clear timelines and application rules. It’s a critical condition for every investment, but more so for big projects like a nuclear power plant.
Effective Project Management: Project management plays a significant role in keeping things on track, on time, and on budget. Project delays and rework can substantially increase expenses.
Policy Frameworks: In one way or the other, all points above are heavily influenced by government policy choices. Given the complexity, the uncertainty and the cost of nuclear power plant, the likelihood of having a merchant project is near zero. Government policies drive and influence the feasibility and competitiveness of nuclear power. But unlike regulatory bodies, governments tend to change more often. Every election represents a major source of uncertainty for nuclear power plant, whose construction time go well beyond electoral cycles.
And since western democracies have shown a tendency toward major political swings over the past 20 years, with lots of political parties primarily worried about chasing a very unstable public opinion, the question around the feasibility of nuclear investments seems quite relevant if we aim to have nuclear as part of the energy mix by 2050. It’s not a case that out of the 56 GW of reactor under construction, less than 20% are being built in Western democracies.
In July alone, China approved 6 new reactors for a total cost of $ 17billion, just a few billions more than the cost of one single nuclear plant in Europe. The Finnish Olkiluoto power plant that began operations this year, after an 18-year journey from the investment decision, cost roughly $ 12 billion. Meanwhile, the price tag for the French Flamenville 3 plant, which has another year of delay, has jumped to € 13.2 bln (roughly $ 13.5 bln) last time I checked (from the $ 3 bln initially budgeted).
Taking the Chinese numbers at face value, which one should probably not do given that the State-owned China National Nuclear Corp has signed a deal to build a nuclear plant in Argentina for $ 8.3 bln, the question on everyone’s mind is: why can China build nuclear reactors more affordably and rapidly than anybody else?
If we look at the playbook laid out by the OECD NEA, China ticks all four categories:
Firstly, public opinion does not play a relevant role in driving public policies, thanks to a heavily centralised and stable government. There’s no local committee that will block construction works or protesting against the removal of olive trees.
Financial conditions: nuclear reactors have enormous capex costs and financials matter. The Chinese government can build a nuclear reactor with a 1.4% rate. The same entity, China National Nuclear Corp, applies a 10% rate when it does that on the market.
Supply chain: equipment, component makers, and the whole industry that surrounds the nuclear industry benefits enormously by a long-term visibility on the demand side.
Regulation is stable and enforced centrally: no changes in new requirements or changes in how the market operates hinder the viability of those investments.
Is it gold all that glitters? Not entirely. Despite China's controlled conditions, data from the past decade reveals a surprising reality: the Chinese nuclear program seems to have peaked already between 2016 and 2018, making way for wind and solar. The chart below shows the yearly capacity additions of solar, wind and nuclear. As you can see, solar and wind are both on an upward trend, while nuclear isn’t. And I intentionally set the nuclear capacity on the right axis, with an order of magnitude lower than solar and wind (as everybody in the industry would expect given the higher capacity factor nuclear power plants have).
How come even the Chinese government, that has potentially no issues when it’s about minimizing the costs and the risks of nuclear power, has decided to give much priority to renewables? A possible answer lies in one of the favorite games played by children all around the world: Lego.
LEGO
Prof. Bent Flyvbjerg recently published a book called “How big things get done” where he and his team looked at roughly 16,000 projects worldwide comparing the initial costs, budget and time estimate with the final cost, schedule and outcome. Their findings are quite revealing: only 0.5% of the projects were completed on time, within budget and delivered the promised benefits. The remaining 99.5% incurred in cost overruns, major delays and fell short of delivering all benefits.
What is even more interesting is the distribution, which borrows the fat tail/thin tail difference Nassim Taleb talked about in his work. Fat tails represent events at the extremes of a typical normal distribution, with outcomes that have an exceptionally high impact. In this specific case, projects with fat-tailed outcomes are those most prone to going over budget, exceeding timelines, and falling short of delivering expected benefits.
As can be seen from the distribution below, solar power projects were the leader—not just among energy projects, but all projects - rarely straying off-budget. Wind power and electricity transmission lines follow.
And here comes the plot twist: at the opposite site of the spectrum, in the fat world, nuclear storage, the Olympic Games and nuclear power stations tend to break the bank. Why is that?
The argument provided by prof. Flyvbjerg has something in common with the OECD playbook: modularity. He uses the analogy of Lego: when a project can be broken down into modular units like Lego pieces, then both, from a supply chain perspective and from a project management perspective, stand a better chance of finishing on time and on budget. When that does not happen then there’s 100% certainty that the project will not deliver as planned.
In the case of nuclear power plant, in order to avoid that, we’d need to replicate as much as possible the French nuclear program of the 70’s and 80’s: building reactors with one unique design, thereby creating a dedicated supply chain with long-term order visibility, the capacity to scale equipment production, and an industry that can foster and maintain the needed know how and expertise on a constant basis. Moreover, we’d need to design the power market around the lack of flexibility of nuclear power plants, so effectively minimizing all other inflexible sources of generation such as solar and wind. Basically replicate the US Navy3.
Conclusion: is a Manhattan project the solution?
What transpires from the movie is that “Oppie” and the team were just too busy thinking about how to build the bomb to actually think about whether they should in the first place. And even though the war was about to end, and there was no need for the bomb anymore, the inertia built into the process (not to talk about the sunk cost fallacy) was such that they just carried on, even though there was still a non-zero risk of actually igniting the atmosphere. We all know how that ended and without spoiling the movie finale, our hero spent the last part of his life trying to reverse the chain reaction, albeit a different one, they actually ignited. Fast-forwarding 80 years, we find ourselves stuck on whether we should go net-zero in the first place, with the how being a secondary issue left to an endless debate between factions, with the risk of neglecting even the basic elements.
As the team did in the movie, I’ve been mulling over the topic of nuclear power for some time now and while I acknowledge the climate benefits related to having it as part of the mix. I’ve been struggling to see it as a real investment option for the near future. In the current political and technological framework, it seems more of a weapon of mass distraction, another topic you need to have on your programme, that fuels discussions and perfectly fits in the current hyperpolarized political debate.
Nonetheless, the events of the past few years are now spurring governments, including Western democracies, on a path to regain control over their economies. Alongside AI, political economy is going to be the topic businesses around the world will have to understand and deal with. And being energy the source of everything, that seems to be the natural starting point. Nuclear’s long development time, let’s say optimistically 12 years, makes it such that the market can't independently choose to scale nuclear to the level that would be required. So effectively we’d need another Manhattan project for energy, where you have different sites splitting tasks but all working together.
And possibly the solution starts from dropping the net from the equation and go back to the near-zero, accepting that there’s an element of risk in every decision and without looking at the system as a whole we would lose all the benefits that a complex system as in terms of generating new attributes, new dynamics and better performance. As Hartmut Rosa wrote in his “The Uncontrollability of the world”, the world can’t be reduced to a point of aggression in order to bring it under our control. “A successful relationship to the world aims not at making things controllable, but at bringing them within reach.”
Public support for nuclear energy in Europe is growing [Project Europe Research]
“The Navy has been building nuclear reactors for 70 years as part of its Naval Reactors program, and has built over 200 nuclear-powered ships and 500 reactor cores: that’s more experience building nuclear reactors than anyone in the world” - What does it cost to build nuclear power