Nvidia bets on Gates-backed nuclear startup to keep its AI ambitions from melting down Datacenter operators’ desire for cheap and clean energy to power their facilities has led to renewed interest in nuclear energy and small modular reactors (SMRs) – a tech Nvidia has just decided is worthy of investment. The GPU giant's venture capital arm, NVentures, this week joined Bill Gates and HD Hyundai in a $650 …
"massive water pressure reactors"
How about "massive, pressurized water reactors"?
A PWR or Pressurized Water Reactor uses, well, water, under pressure, something in the neighborhood of 2,000 psi, with a Tave ([Thot + Tcold] /2) of something like 500 degrees F. Pressurized means the primary coolant does not boil. So, a PWR is different from a BWR (Boiling Water Reactor) which does allow the coolant to boil (which simply means, for a given temperature, the pressure is lower), and may forgo the secondary loop entirely, feeding the "hot" (in more senses than one!) steam to the turbine. (Single loop plants are not recommended.)
Point is, the core is nuclear bits, coated with something like zircon, with pure, pressurized water running in a loop, between the core, to a steam generator, and back. Generally, a pump is involved.
The steam generator has U-Tubes (or something) which transfer heat from the primary coolant, to the secondary loop, turning soapy water into steam, which then goes through a turbine, which spins a generator (technically, an alternator), and goes through a condenser, feed pumps, and back to the steam generator.
Other types of plants may include molten salt, or liquid sodium as the primary coolant, but, most have a conventional steam plant attached.
IOW they can't match the thermal efficiency of a FPP of the late 1950's let alone one of the late 2010's.
"with pure, pressurized water"
Not really.
It's laced with "Chemical shim" because the control rods cannot control the design accurately enough. The SoA in "Chemical shim" is isotopically pure Boron, B10 at $10/gram at around 2000pm start of cycle ramping down to 10s of ppm at the end as the reactivity drops. a tank that's 5m in dia and about 15m long holds a lot of water.
That's basically borric acid. which is real good at eating through SA508 (but not it's SS internal cladding, which was all that was keeping most of the water in at one NPP).
'Last year the Institute for Energy Economics and Financial Analysis determined that, despite the hype, SMRs were "too expensive, too slow to build, and too risky to play a significant role in transitioning away from fossil fuels."'
Edit: 'Last year the shills for the oil industry "determined" that...'
There. Fixed that for you...
Nuclear proponents are great shills for the fossil fuel industry with its subsidies and hidden costs because what do they fall back when costs and timelines balloon out or when the nuclear projects get cancelled. Hence why Liberal/National Coalition in Australia is a fan of nuclear power because they know it'll prolong fossil fuels.
Personally I dont care what energy solution we transition too as long as it's not fossil fuel and we take a look at all the costs including hidden costs like environmental and health costs up front so we look at it with our eyes open and don't hide the costs on the public and future generations. If we looked at all the costs including subsidies and tax breaks fossil fuel is a losing proposition and we need to restrict it to where it is necessary in our energy mix at the moment and transition our economy and societies to a more sustainable future but that involves leadership that our societies seem bad at.
If you abandon fossil fuel and don't go nuclear the huge "hidden" or rather "ignored" cost is storage. Take for example last night where wind power was running at <750MW and being night solar wasn't there to help. Battery systems are generally quoted at 4 hour capacity so at least 2 orders of magnitude less than what's required for a prolonged weather event. Scaling batteries by 100 times is going to be very very expensive. Pumped hydro only has volumetric capacity for short term storage so the only real large scale storage alternative is electrolyzed hydrogen stored in salt caverns. This is also very expensive thanks to the infrastructure costs and it only having a 20-40% round trip efficiency.
The neat part of the TerraPower system covered by the article is that it doesn't need long term storage: just the short term storage built into its design that allows it to load follow.
TerraPower are promoting a type of "fast" reactor which uses a different type of nuclear fission reaction than is used in conventional "slow" or "thermal" neutron reactor, such as the SMRs under construction in Canada or the ones from Rolls Royce that the UK plan to build. The latter ones are just scaled down and simplified versions of the types of reactors which have been producing electric power around the world for decades. The innovation in these latter SMRs is in the assembly and construction methods, bringing to reactor assembly the sort of modular design and assembly innovations which have also revolutionized shipbuilding.
The ones that TerraPower are promoting are a different type altogether, although the basic principle is not new. While slow reactors derive most of their energy from the fissioning of U-235, fast reactors get most of their energy from using a fast neutron reaction to convert the U-238 (the main isotope of uranium) to plutonium, and then fissioning the plutonium. The basic intent is to be able to use all of the uranium to create energy rather than just the U-235 (with a minority also coming from a small amount of U-238 to plutonium to energy reaction) as is done in slow reactors.
Slow reactors use a "moderator" such as light or heavy water to slow down any fast neutrons and promote the slow reaction that the reactor uses. The moderator may also be the primary coolant, or the coolant may run in a separate circuit with the moderator not having contact with the fuel.
Fast reactors do not use any moderator, and their coolant therefore must be something which will not interact with neutrons. Most fast reactors have used liquid sodium for this, and this apparently is what TerraPower's design uses as well. They also have a molten salt heat storage bank to store heat for peaking purposes, but this doesn't form part of the reaction itself.
A number of liquid sodium fast reactors have been built in various parts of the world over the past few decades. All of them have proven to be overly complex, difficult to run, and expensive. I haven't seen anything that suggests that TerraPower's reactor will be any better.
Fast reactors were originally developed due to concerns about shortages of uranium, as they conserve fuel. The UK's plutonium stockpile for example was built up in anticipation of using it to fuel and start up a planned generation of fast reactors that were never built.
However, huge, high grade deposits of uranium were subsequently found in Canada, Australia, Kazakhstan, Namibia, and elsewhere, and uranium went from being scarce to being so cheap and abundant that a lot of older low grade mines were forced to shut down due to the glut of uranium on the market. Most people then lost all interest in fast reactors and their designers went back to the lab until a few tech billionaires decided that this was something new and shiny after all.
There is really no relationship between TerraPower's fast reactor and the conventional SMRs under construction in places such as Canada. I suspect that they are promoting their reactor as being an SMR in an attempt to have some of the latter's good publicity rub off on them and are downplaying the fact that it is a fast reactor due to the poor record of fast reactors in terms of complexity and high cost.
For those who are reading this and itching to go on about thorium, thorium uses a slow neutron reaction but undergoes a conversion reaction similar to that used by fast reactors, but thorium 232 to uranium 233 in this case. These reactors are much simpler than uranium-plutonium fast reactors, but there is limited interest in them because conventional slow neutron uranium cycle reactors have lower fuel costs.
Many people believe that it will eventually be cheaper to extract uranium from seawater (where there are huge amounts of it) than it will be to build fast reactors or use thorium.
I suspect that they are promoting their reactor as being an SMR in an attempt to have some of the latter's good publicity rub off on them
Mostly agree with your informative post, but the Natrium reactor is considered an SMR because it's small, 345MW, and modular, as in the reactor will be built in a factory and delivered to site, it's that simple. So I don't think they need an ulterior motive to promote their design as an SMR; the term does not define the technology of the reactor.
TerraPower are strongly downplaying the fact that it is a fast reactor, instead hyping the SMR aspect. I suspect this is due to the poor record of fast reactors in terms of cost and complexity compared to conventional slow neutron reactors.
If you had to state which is the biggest single technological feature which most distinguishes TerraPower's reactor from the vast majority of reactors in commercial operation today, it isn't it's size that singles it out, it's that uses a different nuclear reaction. This is in sharp contrast with other SMR designs on offer, which are basically scaled down and simplified well proven conventional designs which can use conventional fuel sourced from commercial suppliers.
I am willing to listen to what they have to say, but I am very skeptical about their ability to build and operate a fast reactor at the costs and reliability levels that they claim. There is a reason that most nuclear power development programs gave up on fast reactors years ago, and I haven't seen anything which suggests that TerraPower have overcome any of those hurdles. The history of fast reactors is a history of failed one-off projects and I suspect that TerraPowers will be again.
It'll probably be like the most mines where the clean up cost in escrow (if at all) is vastly insufficient. That's the traditional thing to do in big business overestimate the gains to the people and government underestimate the costs to the people and government and do some fancy financial engineering to make a buck in the short term and screw the long term.
This post has been deleted by its author
So "Fast spectrum" IE no moderator like water, heavy water or graphite.
That means the neutrons slam into the RPV walls with 10-100x the energy they do with thermal spectrum reactors ( which the vast majority of reactors that have been built, including the MSR and pebble types are)
"Sodium cooling" which the US, UK and France have experience of needs 3 loops between the sodium and the steam generator because radiactive sodium is a hard (19MeV+) Gamma ray emitter, and reacts violently with water. The slightest leak will admit air which will form sodium oxide, which is a highly effective abrasive, clogging valves, wearing away pipe walls etc.
Nuclear fuel prices are normally proportional to the enrichment level unless you've got a country with a load of bomb grade uranium they want to blend down.
Once that's gone the price for new fuel, starting from natural U at 0.7% U235,stripping about 0.4% of that out will need about 45-50Kg per Kg of new fuel (law of diminishing returns, stripping it down to 0% U235 is almost always prohibitively expensive, even for the US DoE), along with an enrichment site that can handle those levels (most are only safe up to 5%. Above that the whole layout of their plant has to be re-designed for criticality safety). Iran could handle it, but they may be going out of business soon.
Otherwise most of this design sounds kind of stupid and expensive. This corner of the nuclear power reactor solution space has been explored before and there's a good reason work on it ended.
Those reasons have not changed.
Iran could probably handle it.
OTOH the molten salt thermal storage loop is a good idea.
In the UK it could deliver the scale of and dispatchability that would upend the UK gas and electricity markets and end the UK's having the highest electricity prices in 28 countries, including all of the G7.
One day someone will sit down and answer the question "How do you build a nuclear power reactor for $4/W so it competes head on with fossil fuel stations and generates steam at their levels of efficiency (IE 540-600c, this thing is supposed to manage 350c. That's high by PWR standards but p**spoor by gas cooled reactor standards), allows on load refuelling, fissioning of Trans Uranic elements (Pu is a TRU but the others as well) and delivers a complete nuclear power solution inside the grounds of the NPP?
Today is not that day and this design won't get anywhere near that price.
The NRC commissioned Stone and Webster to look at how to build GW size PWRs cheaper and faster. Results here
And then Three Mile Island showed how a $1Bn asset could be turned into a $2Bn cleanup job.
BTW despite the huge advances in networking techniques and data multiplexing the amount of cable in a new PWR is actually greater than the figures in this report.
And it's recommendations don't take into account the discovery (by large scale ship builders) of the "1:3:8" rule that something that takes 1 hr of factory time will take 3 in an assembly and work area near the ship (or other large structure) and 8 hours working inside said structure, due to getting stuff to the work area, delays, different teams wanting access at the same time, etc.
Top tip. if a countries nuclear regulator allows it pre-cast, reinforced concrete looks a real winner, as does precision blasting for foundations (estimated to save months of time which means a lot of work can start months sooner).
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
