By Brian Gitt
I work at an advanced nuclear tech company. I spend my days talking to developers of data centers, AI factories, and industrial sites—companies that hope advanced nuclear plants can meet their enormous and ever-increasing energy needs. Nearly every conversation comes down to two questions: How much will it cost? When can we get it?
Nuclear fuel has an outsized impact on both answers. To scale nuclear power, you need abundant fuel with a per-kWh cost that can compete with other energy sources like natural gas, and that’s available in a time frame that meets customer needs. You could have the best-designed plant in the world, but if its fuel costs too much or arrives too late, its power will be too expensive for customers or unavailable when they need it.
The challenge is that no fuel source checks every box: what’s available now may be expensive, what costs less may take longer, and what supports growth may require infrastructure that hasn’t been built at scale. The best approach isn’t to go all-in on one source. It’s to build a diversified fuel portfolio. Most people don’t put all their retirement savings into a single stock. That way, a bad outcome in one sector doesn’t threaten the entire plan. A similar logic applies to advanced nuclear fuel.
Advanced nuclear fuel differs from traditional fuel, which is typically enriched less than 5% and shaped into small oxide pellets. Advanced fuel has a higher enrichment level or a different form factor. It is more energy dense and efficient, but its supply chain is just beginning to grow.
There are three sources of advanced fuel: fresh fuel made from newly mined uranium, government-sourced material, and fuel recycled from nuclear waste. Using them together can reduce supply risk, control costs, and help scale advanced nuclear power faster than any single source.
The biggest advantage of fresh fuel is certainty: the industry knows how to make and transport it, companies know how to buy it, and regulators know how to review the plants that use it. But certainty comes at a cost. Fresh fuel is two to three times more expensive than government-sourced material or recycled fuel.
The supply chain is still small, and each link requires plants, permits, workers, equipment, and transportation. If one link is behind schedule, the fuel arrives late, and if the fuel is late, the power is late. Suppliers lack the firm orders needed to build capacity, while developers hesitate to commit before projects are licensed and financed. Each side is waiting for the other to move first. Relying on fresh fuel alone would leave the industry exposed to high prices, market swings, trade restrictions, geopolitical disruptions, and delays.
Governments have nuclear material left over from defense, research, and energy programs. Some can be processed into advanced-reactor fuel without all the steps required to make fresh fuel. Governments spend billions to store, guard, or get rid of this material. Turning it into fuel converts a costly burden into affordable clean power.
The downside is that we lack facilities to downblend the material and fabricate it into the form reactors need. Agencies must decide who can use and transport it and under what terms, so the delivery date remains uncertain. Stockpiles are also finite. They can support early growth, but they can’t fuel a reactor fleet forever.
Recycled fuel is the strongest long-term option. When most people hear the term “nuclear waste,” they assume it has no value. But waste is used nuclear fuel, and most of today’s reactors use only about 5% of their fuel’s energy before it’s removed and placed in storage. That means 95% of the fuel’s original energy is left behind. Leaving it untapped is like taking one bite out of an apple and throwing away the rest.
The right processes can convert used fuel into new fuel. Nuclear recycling could cut fuel costs by as much as 80%—a development that would change the economics of nuclear power. It would reduce the need for new uranium mining and enrichment, shield power producers from swings in fuel prices, and decrease dependence on foreign suppliers. Nuclear recycling would transform a long-term burden into a vast source of domestic energy.
Perhaps the biggest advantage of recycled fuel is its path to scale. The more reactors that operate, the more recyclable fuel they create, and the more recycling becomes self-sustaining. But that system can’t be built overnight. Not all advanced reactors can run on recycled fuel, and most countries lack commercial-scale recycling facilities. Those facilities must operate safely, reliably, and at a steady cost.
Fresh fuel offers the most certain near-term path. It’s more costly than the alternatives, but those alternatives need time to develop. Over the middle term, government material can provide a lower-cost alternative while the market grows. Recycled fuel provides the best scale and economics for the long run, but it will take the longest to develop.
Why not just stick with traditional light-water reactors? Their fuel is cheaper and their supply chains are already in place. Traditional nuclear should remain part of the mix, but it shouldn’t be the only thing. Advanced reactors can get more energy from less fuel, offer stronger safety features, and convert used fuel into a resource instead of a burden—addressing communities’ biggest concerns about new nuclear projects.
A well-established supply chain is one advantage of traditional nuclear, but much of it begins overseas. In 2025, about 93% of the uranium purchased by U.S. nuclear plant operators came from foreign sources. That dependence creates geopolitical risk while the United States already has a vast store of energy sitting in used nuclear fuel. We shouldn’t let that resource go to waste.
Diversifying nuclear fuel sources provides customers timely and cost-effective options. It enables us to unlock an abundant fuel solution for the next century, transforms two liabilities—waste and government material—into assets, and minimizes price and geopolitical risk. It is the surest way to secure energy independence and abundant, affordable clean energy for our future.
Brian Gitt is Senior Vice President and Head of Business Development for Oklo, Inc.
This article was originally published by RealClearEnergy and made available via RealClearWire.
Pretty fair summary, although the authors self interest is there, its not hidden, which is at least honest.
Nuclear was just getting started when publicPanic™ and cheap gas pushed it all onto the back burner.
It is time to revive it.
And HEU (high enriched uranium) will go critical in smaller masses and so is suitable for very small reactor modules. Which is something to think about, Portable reactors for ships, or to generate off-grid power.
And several countries are awash with recycled plutonium.
There are even breeder reactors that could be used to degrade medium level waste into something less noxious.
We are back in 1970. picking up where we left off.
France is on fire. Do you still think it was a good idea importing people from 3rd world countries?
I’m surprised the potential of LFTRs was not mentioned – Liquid Fluoride Thorium Reactors.
The Chinese have a trial rig in operation, and have refueled without shutting down, and the major obstruction for Copenhagen Atomics is . . . regulatory.
Supply?
Thorium is a waste product of monazite sand operations for rare earths.
And Th-232 is the only isotope of Thorium unlike U-235 (the naturally occurring fissile isotope) makes up less than 1% of mined Uranium and is costly to enhance it. You need enough U-235 to things started since Thorium has to have neutrons to make the actual fissile material (U-233) via decay of Th-233. The fuel cycle was proven in the commercial power plant, the St Vrain Nuclear Plant outside Platteville, CO, a few decades ago but it’s been closed for many years now.
Time is now to end the 50-yearlong nuclear hiatus with SMR leading the renaissance.
As pointed out, the US imports over 90% of the uranium oxide it needs. Major suppliers include Russia and Canada – oh dear!
I suppose it could always depend on its other good friends – Uzbekistan and Kazakhstan!
Only joking. Although the US administration acts like it doesn’t need outside help, reality indicates otherwise. Stomping round threatening the other 95% of the world with destruction if they don’t bend the knee to the US is possibly not the best tactic if the US are almost totally dependant on imports from them.
Once foreign suppliers get sick of US demands and threats, and find other trading partners, what then? Who knows?
To few people realize that Iran is processing its own ore .
The USA , under Israel , seems determined to not let it have any , even inspected commercial nuclear industry a all .
Australia has vast resources of uranium. African countries also have heaps.
Interesting article, sounds fine when you read it quickly. Problem is that the French recycle their waste but the reprocessed fuel is much moe expensive than newly made. The UK gave up on their THORP project as it was similarly uneconomic.
From the above article:
“But waste is used nuclear fuel, and most of today’s reactors use only about 5% of their fuel’s energy before it’s removed and placed in storage. That means 95% of the fuel’s original energy is left behind. Leaving it untapped is like taking one bite out of an apple and throwing away the rest.”
It’s no surprise to me that this old canard is still be touted as reason for recycling spent nuclear fuel (at a current cost per pound about 5 times higher than using newly-mined and refined U-238).
The main component of fuel for commercial nuclear power plant reactor is common, non-fissile U-238 enriched (mixed with) much more rare, fissile U-235 to achieve a beginning-of-life concentration of about 5% U-235. When this fuel reaches its end-of-life, the U-235 concentration has dropped to 3% or slightly lower. So, in fact the “most active” ingredient in the nuclear fuel, U-235, is being “consumed” to a level of about (5-3)/5 = 40%, but the major component of the nuclear fuel, U-238, has only been “consumed” by about 5-6% . . . that is, 94% of the original U-238 remains unchanged but this is the ingredient that is not fissile as is!
Two factors combine to define the usable life of nuclear fuel:
1) when the U-235 concentration falls below about 3%, the nuclear fission process becomes too inefficient to produce power in a stable fashion, and
2) when the U-235 and induced U-238 nuclear fission reactions have reached the ~3% lower concentration level of U-235, the accumulation of “neutron poisons” (fission products that absorb neutrons without causing further fission) choke off the baseline power chain reaction. The main “poisoning” nuclear byproducts in spent nuclear fuel (SNF) are Xenon-135, Samarium-149, and “reactor slag” (comprised of witches’ brew of Krypton-83 , Molybdenum-95, Neodymium-143, and Promethium-147). These are the chain reaction products that have to be safely removed—and correspondingly safely stored somewhere, somehow—in the process of recycling SNF.
So, bottom line, the claimed “95% of of the fuel’s original energy is left behind” is true only if looking at the energy content of the U-238 but that is not what really counts in producing nuclear fission power . . . without the replenishment of U-235 that energy will stay forever locked up.
Oklo has developed a fast-neutron reactor that will run off of the U-238.
Unfortunately, you apparently:
1) do not understand that current “slow neutron” fission reactors also “run off” (partly) U-238, and more importantly
2) Oklo’s nuclear reactor design is based on High-Assay Low-Enriched Uranium (HALEU) metallic fuel, which contains a higher concentration of uranium-235 (between 5% and 20%) than does traditional reactor fuel (typical 5% upper limit on U-235).
To repeat: U-238 by itself is not fissile, and cannot be used alone to produce energy in a self-sustaining nuclear reactor.
“I work at an advanced nuclear tech company”.
Let’s set the record straight:.
1) Only one nuclear plant has been under construction in the US recently — and its way over budget and delayed.
2) Nuclear is the most expensive option for power generation. Central solar and on shore wind are the least expensive, and fastest to build out
That it has been delayed by frivolous law suits and unnecessary and ever changing regulations is why it is expensive.
Exactly as the left intended.
No, it’s inherently a very expensive and complex technology. Utilities don’t like nuclear and aren’t building any in the US,
MN has 2 nuclear power plants built in the 70s. Still going strong.
Yes several were built in the 70s. It was the bad experiences with cost overruns and project delays that soured the utilities.
Why are we looking at the 70s. Why not look at Chinese built plants in Pakistan, Korean built plants in the UAE, and assess the quotes by Korean, Chinese and Russian suppliers in Egypt and Czechia. Russia has also bid on plants in India, Czechia, and Egypt, but the specifics have not been disclosed.
I’m not a nuclear expert,and leave it to the utilities to select the best power plant technologies for their needs. I’m agnostic as to what they pick, as long as they pick a zero emission technology.
Agnostic … as long as they pick what I prefer?
Wow. Idiocies.
It is complex which makes it expensive.
However it is much more expensive than it needs to be thanks to the actions of your fellow socialists.
I have no idea what you’re going on about. I’m a lifelong Republican who loves nuclear power. I wish it were cheaper so that more would be built.
Maybe the rest of the world has more money to spare than the US. For example, China, building new nuclear plants by the dozen.
Round the world, about 200 nuclear plants being built or scheduled to be built.
Who cares what Americans think?
Let them starve while they freeze in the dark, moaning about how everybody is stealing their intellectual property.
Only joking, but a country that has to find 3 billion dollars a week to pay the interest on its debt is hardly in a position to lecture other countries on how to conduct themselves, is it?
Socialists, communists, capitalists, right-wing, left-wing, fascists, democrats, autocrats – makes little difference in the long run.
Where do you see comments in which America lectures the rest of the world about nuclear?
Please define central solar.
Utility owned and operated solar plants that feed the grid.
Like Ivanpah, for example?
Ivanpah used an expensive older technology, whereas modern day conventional solar panels are far cheaper, and embody power plants that are less costly to build and operate than fossil fuel power plants.
I don’t know. Why don’t you tell me? Does America lecture the world about nuclear?
Nuclear what? You need to be specific.
“Who cares what Americans think?”
I, for one, will step up and say that I do.
Be that thinking mostly correct or—as is IMHO apparent most recently (say at least over the last 20 years)—largely incorrect.
You know, recent “main stream” America beliefs in “climate change”, that the Federal debt “doesn’t matter”, that the rule of law “doesn’t matter”, that the US Constitution “doesn’t matter”, that the US President usurping the powers given to Congress and SCOTUS via EO’s “doesn’t matter”, and now that democratic socialism is the future of America (i.e., the USA).
I don’t, of course, if the thinking doesn’t affect me directly. Why should I?
But more to the point, if some American policy results in a diminution of my quality of life, I might express displeasure at their actions, but I accept that Americans don’t care what I think, either.
I seem to be able to accept what I think I can’t change, so I’m content.
Somehow the Navy is able to build reactors with highly enriched U-235 quickly by using standardized systems. The reactors that were in subs starting in the late 1950s thru the 1980s were pretty much all the same power plants–S5W reactor plant with S3G reactor core. When the Los Angeles class came out the design was changed as the reactor core was larger but all the LA class boats had that same design. Same with the Ohio class missile-carrying subs.
So it isn’t the complexity as it is creating a standard and replicating it. That is one of the goals of the SMR plants where they can add and additional small reactor at the same location as demand increases.
I know about the Navy’s reactor plants having been a Reactor Operator in the Navy in the ’70s into 1981. I operated four different reactor plants: D1G, a prototype of the reactors used in the Navy’s Cruisers and Destroyers; S7G MARF, an experimental plant using pumps instead of control rods; USS Tunny, a fast attack sub; and USS Robert E Lee, a missile-carrying sub.
I hope SMRs succeed. But I don’t think the Navy precedent says much about the odds of commercial reactors achieving costs that are competitive with solar or wind, or even nat gas plants.
China. It is only fair to ask when they have a molten salt SMR and a pebble bed reactor running, now. Yes, they did the “standing on the shoulders of giants” plow to snatch our Oak Ridge Design and data, But there is more to this than generally acknowledged. They have a more responsive capitalism than we do.
How so?
Mao was an idiot. About the time our amazing crony capitalists realized they could make more money sending our manufacturing to lower labor cost third world nations, Deng Xiaoping, in December, 1978 realized that US wealth came from incentivizing individuals to work harder in their own interests. A command economy can incentivize keeping jobs at old, state owned industries, or can see the future and make hard work a national pride, while allowing many Chinese millionaires to organize Chinese production. It is estimated, with their vertical integration structures and “get it done” government fiat that they can get a major car factory producing in 2 – 3 years compared to our 4.
More than that, our crony capitalism can be slimmed down to their “get it running quick or maybe, you disappear.” Waiting until an American FEEED project can integrate all the vendors into the necessary “friend of the Governor” networks wastes money and time.
Is no one aware of the micro-reactor invented by Australian engineer Barry Gallagher, now being developed to prototype stage in the USA (because the Australian Govt is terrified of nuclear)? Once developed, these modules can be constructed in factories, resulting in much better quality control and, with economies of scale, bring down the cost. With this approach, we might finally achieve the safe, low-cost nuclear power promised for so long but never actually delivered.
Being small and impossible to explode, this should also address the safety concerns of the anti-nuclear mob but in reality nothing will satisfy them.
A couple things. A nuclear facility only needs to be refueled after years of use, that gives plenty of time to insure you have fuel ready when it is time to refuel. A coal facility has the fuel sitting right outside the door and you can continually add to it to insure fuel even in bad circumstances. Gas plants have fuel delivered to them continually but the fuel isn’t stored so if the delivery system is shut down the generators are shut down. The cheapest and safest route for AI would be coal. The best alternative to coal for a more dependable source of fuel would be nuclear since the refueling interval is so long. It is time to move on from the pissing match between current nuclear technology and future nuclear technology. Build the plants we know we can build now, make provisions for any additions to include the newer technology when they have become accepted. You already have the land and that land has already been permitted for nuclear generation. Which ever new technology wins is the one you choose in the meantime you have reliable, affordable, clean energy for your AI operations and you can sell excess generation if you choose.