Nuclear power is the safest energy source ever built.
The fear of it has cost more lives than the technology ever has.
Deaths per terawatt-hour of energy produced:
Coal: 24.6
Oil: 18.4
Gas: 2.8
Hydro: 1.3
Wind: 0.04
Nuclear: 0.03
Nuclear is safer than wind.
Safer than solar (rooftop installation deaths).
60x safer than natural gas.
Chernobyl killed 31 people directly.
The WHO estimates fewer than 4,000 long-term deaths, in the absolute worst nuclear accident in history.
Meanwhile, air pollution from coal kills 800,000 people every year.
If it isn't already abundantly clear, countries like China and India will secure as many pounds of uranium as they can
Case in point: "We would buy as much (uranium) as Cameco can produce” Patnaik, India’s high commissioner to Canada, told reporters"
U prices will keep rising
UxC now joins TradeTech in reporting $93/lb. LT U3O8. This is the highest LT price since April 2008. Another +$3/lb. takes it to an ATH.
Of course, inflation-adjusted ATH needs to breach $150/lb. That's coming this cycle.
If you're pro environment, but anti-nuclear, you're not pro environment
Nuclear uses little to no land relative to the amount of power it produces
It is very safe
It is reliable
It produces 0 emissions
The produced waste is minimal and all of it is captured and stored safely
⚛️ Italy is preparing its return to nuclear power.
Prime Minister Giorgia Meloni announced that her government will approve an enabling law this summer to create the legal framework for reintroducing nuclear energy, 36 years after shutting down its last reactors.
In just a few years we went from:
-We should close all nuclear reactors
-Alright maybe just a few
-Let's pause and assess
-How about some life extensions for existing reactors
-We should think about restarting some power plants
-Massive buildout plans
The CAGR just keeps rising
*Available for everyone*
Cameco posted some very interesting results, particularly on the Westinghouse/nuclear power front and combined with the conference call, there was a lot to get into
I just finished a 7 page report on it that you can find here ⤵️
https://t.co/2JkIa1y3Ev
BREAKING: Belgium is moving to nationalize its nuclear fleet. The gov signed a Letter of Intent with French operator, Engie, for a full takeover of all reactors and liabilities. Dismantling of shuttered reactors has been HALTED.
This idea that uranium demand is going to be replaced by thorium on any reasonable near term timeframe is quite frankly ridiculous, as is the sheer amount of times this has been brought up in conversations I had with people on nuclear power
Time to debunk it, strap in. Thorium appears to be a fan favorite 'this clean energy source will quickly take over and eat nuclear power market share' comment I see come across more often than I'd like. The way I see it, this entire narrative reflects a fundamental misunderstanding of where thorium technology actually sits on the development curve, what the supply chain for any kind of commercial thorium fuel cycle would actually require and frankly also how the engineering and regulatory realities differ from a slick PowerPoint slide showing a pretty molten salt loop, because trust me there are a lot of those as well.
The Chinese thorium program everyone keeps pointing to is a real thing and I am not dismissing it, so I just wated to mention this out of the gate. The TMSR-LF1 facility out in Gansu province has achieved criticality and has been running experimental campaigns and the Chinese Academy of Sciences has laid out a multi-decade roadmap for scaling the technology. The trouble is that the TMSR-LF1 is a 2 MW thermal research reactor, which is to say it produces roughly enough heat to run a handful of apartment buildings if you wired it up to a turbine, which incidentally they are not doing to the surprise of nobody. The next step on the Chinese roadmap is a 10 MW demonstration unit targeted for around 2030, and after that, if everything goes well and no fundamental materials problems emerge (which remains a big if when it comes to this tech), a 100 MW commercial-scale demonstrator somewhere in the 2035 to 2040 window. To put that in perspective, China is currently building dozens of conventional uranium-fueled gigawatt-class reactors, with a fleet target in the hundreds of GWes by 2035 and even more come 2050, and every single one of those reactors is uranium-fueled. From what I can tell, even in the most optimistic Chinese scenarios, thorium accounts for a low single-digit share of their nuclear capacity by 2050, which is to say that even China itself is treating thorium as a hedge and a long-term R&D play rather than a near-term displacement strategy and I don't blame them.
Time to put my uranium analyst hat on for a second, but if China really believed thorium was about to displace uranium, why are they simultaneously locking up mass amounts of supply (see my Bannerman research piece from when that deal was signed), and roughly five Kazakh joint ventures pulling north of 10 million pounds annually? Why are they running a uranium storage facility on the Kazakh border? Why are the Chinese utilities out there contracting aggressively as if pounds are about to become genuinely difficult to find? Why are they building a strategic uranium reserve that, depending on what source material you run with, sits at ~750+ million pounds? Of course the answer is that they know what the actual timeline is and they are positioning accordingly.
Now let me get into the engineering side, because this is where things get even more interesting and where the casual thorium evangelism tends to fall apart and it's about time for that to happen. Thorium itself is not fissile, as many of you will know. I think this trips people up more than just about anything else in the discussion. Ready for some vague left-field pub quiz knowledge that you will probably never use anyway? Thorium-232 is fertile, meaning it absorbs a neutron, decays through protactinium-233, and eventually becomes uranium-233, which is the actually fissile material that does the work in the reactor. This means every thorium reactor needs a fissile starter charge, typically enriched uranium or plutonium, to get the breeding cycle going in the first place, and the protactinium intermediate has a 27-day half-life that creates its own headache because if it absorbs another neutron before decaying it becomes useless from a fuel standpoint, which is why the molten salt designs require continuous online reprocessing to skim the protactinium out. That reprocessing loop, I believe as I am not a nuclear engineer myself, is the single most underappreciated technical challenge in the entire thorium story, because you are essentially running a chemical reprocessing plant attached to the reactor in real time, with all of the materials science, corrosion, and radiation challenges that implies.
There is also the U-232 contamination problem, which I find rarely gets discussed in the popular pieces. When you breed U-233 in a thorium reactor, you inevitably produce some U-232 alongside it, and U-232 has decay daughters including thallium-208 which emits a 2.6 MeV gamma ray that is genuinely nasty from a shielding and handling perspective (more vague pub quiz knowledge, I warned you all). The implication is that any thorium fuel cycle facility has to be designed for remote handling at every stage of fuel fabrication, transport, and reprocessing, and the existing uranium fuel infrastructure simply cannot be adapted to this without a ground-up redesign. That represents a multi-billion-dollar industrial buildout that does not yet exist anywhere in the world at anything remotely resembling commercial scale.
The supply chain comparison, of course to the surprise of exactly 0 people, is where the gap becomes pretty stark. Uranium has a fully mature global supply chain including mining, conversion, enrichment, fuel fabrication, reactor operations, spent fuel management, and in some jurisdictions reprocessing, all of which represents roughly seventy (had to spell it out for dramatic effect) years of cumulative industrial investment and several $ billions of installed infrastructure. The thorium side has essentially none of this at commercial scale. There is no dedicated thorium mining industry to speak of, with most current thorium production occurring as a byproduct stream from monazite processing in rare earth operations (why do I know this? Because I also spent an unhealthy amount of time mapping out global REE markets... I really need a new hobby), and that material mostly gets stockpiled or treated as a low-level waste because there is simply no buyer for it. There are no commercial thorium conversion or fuel fabrication facilities outside of small experimental setups. There is no established regulatory framework for thorium fuel licensing in any major reactor market, which is to say no NRC pathway, no CNSC pathway, no IAEA safeguards regime built out for the U-233 cycle. There are no transport protocols, no waste classification standards, no operator training programs at scale. To build all of that from scratch, even with significant state backing, is realistically a 20 to 30 year endeavor at minimum, and that is assuming the underlying reactor technology actually works reliably at commercial scale, which has not yet been demonstrated anywhere and that pathway has never exactly been known for being 'on time and on budget'... a bit like uranium mining now that I think about it.
Meanwhile, every single reactor currently under construction worldwide, every SMR design with a credible deployment pathway into the 2030's window, and every hyperscaler nuclear deal I and many others have been tracking runs on uranium. The AP1000s in China, the Hualong Ones, the VVER-1200s in RSA and Egypt and elsewhere, the EPRs in France and the UK, the CANDU refurbishments, the proposed BWRX-300s, the proposed Kairos units, the proposed X-energy reactors going to the Texas Gulf Coast, every one of them needs uranium and the associated fuel cycle infrastructure and that demand is not getting displaced by a Chinese research reactor running at 2 MW thermal and if you believe that it will, I have an oceanfront property in the Swiss Alps to show you.
The way I see it, the thorium narrative has become a kind of energy meme that is mostly disconnected from the actual industrial trajectory. I have no problem with thorium R&D continuing (As we have seen in country's like India as well), and I think it is genuinely possible that by the 2050s and beyond thorium plays a somewhat meaningful role in the global nuclear mix, particularly in countries with abundant thorium resources and limited uranium access (again, with India being the obvious example, although their three-stage program has been crawling along for decades and is still essentially in the first stage of the proverbial trilogy). The idea, however, that any of this affects the uranium demand cycle that is unfolding right now, or the structural deficit through 2040 that I have walked through extensively in prior coverage on the Codex for the better part of 5 years, or the price discovery process happening in term contracts that just keeps moving higher and higher (more details on that in Friday's extensive report, which you can find on the Codex), is simply not realistic. Thank you for coming to my Ted talk, now please focus on nuclear power and stop it with the thorium energy memes chatter. Have a good and healthy day folks.
Just wanted to share some thoughts on the Defense Production Act and how it relates to nuclear power in the US, because the moment that we are going to have to start adding potential new builds in the country to serious demand side modelling is inching closer and closer.
The convergence of several pressures hitting at the same time, namely AI capex pulling forward electricity demand at a pace nobody modelled, the obvious need to reshore industrial capacity in the wake of supply chain weaponisation and now a growing realisation that you cannot run modern refineries and processing facilities without enormous volumes of fresh water and power, all point in the same direction. That direction, the way I see it, is nuclear, and specifically the kind of nuclear buildout that the US has not seriously attempted since the 1970s.
Allow me to elaborate a bit more on this, because I think there are some dots that need connecting here. Hyperscaler capex for 2026 is running close to $1 trillion globally, and a meaningful share of that is going into data centre construction that requires baseload electricity at gigawatt scale. Wind and solar simply cannot deliver that profile, which is why we have already seen (time for a throwback) Microsoft sign Three Mile Island, Amazon take Talen, and Google commit to small modular reactor offtake. (For those of you keeping score at home, the same hyperscalers that were demanding 100% renewable matching three years ago are now lining up for 60-year nuclear PPAs, which I find quietly hilarious.)
Layer on top of that the industrial reshoring story, where domestic petroleum refining, critical mineral processing and semiconductor fabrication all share a common feature, namely (and forgive me for kicking in an open door here) that they consume electricity measured in sheer quantities the existing grid was not designed to deliver and then on top of that, there is the water angle, which I think most people are still sleeping on. Desalination is electricity-hungry, the regions where water shortages are most acute happen to be the same regions where refineries and chip fabs want to expand, and the cleanest, densest source of the kind of round-the-clock power that desalination plants need is, well, nuclear and its ~90% capacity factor.
The honest reality I have to acknowledge is that traditional gigawatt-scale nuclear builds in the US still run somewhere in the 10 to 15 year range from permitting to commissioning, and that is in a world where the NRC moves faster than it currently does (although that can be cut with adequate support from all sides, including political, public and monetary support and looking at the available data, nuclear power is getting an increasing amount of all three in the US). SMRs, with their factory-built modular design and significantly smaller footprint, could in theory compress that timeline, although no SMR has yet been commercially deployed in the US and we will need both traditional large scale reactors (which the US is planning to build in earnest), as well as said SMRs. The regulatory pathways are still being worked through, but the urgency to start making meaningful progress is absolutely there now and when you prepare for the adding of meaningful nuclear capacity, you also have to prepare for the associated fuel picture years in advance. More demand is coming down the pipeline and we are going to need higher uranium prices for longer to even have a chance of meeting said growing demand.
Retail watches spot prices while utilities quietly lock in long-term contracts with $150/lb ceilings. In the term market, security of supply is beginning to trump price. When the scramble for uncommitted pounds begins, it won’t climb – it’ll gap up.
The WNFC in Monaco brought forward some very interesting insights on the uranium and nuclear power sector
SMRs were frequently discussed and it appears that demand is already showing up in the fuel cycle and that could be a major demand tail event
More details in today's report
"Nuclear power plants take too long to build"
Once in operation they can produce clean energy for 80+ years at a ~90% capacity factor. In those 80+ years, you would need to replace equivalent solar and wind capacity up to 3-4 times
"But they're too expensive"
No they're not ⤵️
France’s nuclear fleet gives it one of the world’s lowest-carbon electricity grids—
France generates two-thirds of its electricity from nuclear power, making it the country’s dominant power source.
As the chart shows, that’s far more than the average across Europe, which is 20%, and the world as a whole, at 9%.
Nuclear power is a low-carbon electricity source, giving France a very clean electricity mix for decades.
Per unit of electricity, France emits far less greenhouse gas than its neighbors and has some of the lowest-carbon power in the world.
The global average, based on lifecycle emissions, is 472 grams of carbon dioxide equivalents (CO2e) per kilowatt-hour of electricity generated. In France, this figure is 42 grams.
(This Data Insight was written by @_HannahRitchie.)