back
152 comments
Part of the problem in building conventional nuclear reactors is that an $11Bn project like the conventional build mentioned in the article is of such a scale that it’s inevitable that something will go wrong that causes the budget to explode. The sales pitch for small reactors is they’re not $11Bn projects, so there’s a possibility they could actually be completed on time and somewhat near budget.

The author is also clearly anti-nuclear - that’s not to say they’re not right, but they’re motivated.

> is of such a scale that it’s inevitable that something will go wrong that causes the budget to explode.

It's not just the scale, it's also how few there are in planning in the west. You work out the kink (both building and operational) by building more of the things, and that's also how your builders skill up. At the height of its buildup, France had half a dozen nuclear reactors being built concurrently.

Because SMRs are smaller and you need more for the same total output, more would be built, and thus there are more opportunities to work out the kinks in manufacturing and setting up.

An other theoretical advantage of SMRs is they could be built in factories and craned in, and when the fuel is expended they get craned out, a new reactor is dropped in, and the old reactor is moved to a refueling facility (which can be nearby a reprocessing plant), the site doesn't need to be offline for refueling, and it gives higher opportunity for automation.

Indeed, if SMRs can serve as a way to get the industrial learning loop of nuclear working again, it may be worth it. The size of the current power plants were defined exactly due to the need of scaling the energy output to a level worth the auxiliary costs of lower output temperatures than coal, combined with increasingly expensive safeguards.

Why being anti-nuclear is even a legitimate stand is a good indicator that people still don’t take climate change (and general over consumption) seriously. It’s like being anti-chemicals because some of them can be toxic to ingest.

Why people are still pro-nuclear is a good indicator they don't take climate change seriously. CO2 emissions are reduced more quickly and more cheaply by going with renewables instead of pouring the money down the nuclear rathole. This is especially the case when experience effects on renewables and storage are included. The faster we spend on them, the faster they get cheaper, and the more quickly even existing fossil fuel capacity becomes cash flow negative.
https://en.wikipedia.org/wiki/M._V._Ramana doesn't look "anti-nuclear" to me. Could you elaborate?
The closing paragraphs of the article:

> The climate crisis is urgent. The world has neither the financial resources nor the luxury of time to expand nuclear power. As physicist and energy analyst Amory Lovins argued: “… to protect the climate, we must save the most carbon at the least cost and in the least time.”

> Expanding nuclear energy only makes the climate problem worse.

> The money invested in nuclear energy would save far more carbon dioxide if it were instead invested in renewables.

> And the reduction in emissions from investing in renewables would be far quicker.

These statements may wind up being correct, but they’re speculative, and they’re anti-nuclear.

To the point of the Wikipedia page, a couple selected publications - although in fairness I haven’t read the publications and am just judging from the titles:

> Nuclear power: Economic, safety, health, and environmental issues of near-term technologies, Annual Review of Environment and Resources 34, 2009, 127-152

> Beyond our imagination: Fukushima and the problem of assessing risk, Bulletin of the Atomic Scientists, 2011, 19 April

> Nuclear Power in India: Failed Past, Dubious Future, 2007, Available at www. npec-web. org/Frameset. asp

Again, all of these could be entirely factual and well defended, but the author is clearly not a disinterested observer.

I’m more pro-nuclear than the author, and mostly I’m skeptical about good scalable solutions to the storage problem. I’d prefer renewables + storage to be the answer, but I think we made a mistake when we halted nuclear development and I don’t think we’re doing ourselves any favors by not pursuing the technology.

Uh... I would say that page is the resume of an anti-nuclear ideologue, of someone who has built his career on being against nukes and nuclear energy.
Proponents assert that SMRs would cost less to build and thus be more affordable. However, when evaluated on the basis of cost per unit of power capacity, SMRs will actually be more expensive than large reactors.

That's not what affordable means. People are building smaller reactors because it's more likely that a project will be completed, not because of unit costs.

And because if you build a lot of reactors, then you get a learning curve where they get cheaper over time, just like any other mass-produced item. Some of the SMR designs can even be built in factories.

The article ignores this, and in fact complains about the cost of the first six reactors proposed by NuScale (which, fwiw, is just a smaller LWR, not one of the more innovative designs).

Before anyone talks about the negative learning curve of large reactors in the US, bear in mind that we mostly build those as one-offs, so no learning curve exists. We don't build many, regulations change pretty frequently, and sometimes the NRC requires design changes after construction starts.

Economies of scale don't work for infrastructure. I mean if it would work for nuclear plants why have people not done it for coal plants? There have been many of them build, and a nuclear power plant is 70% the same as a coal power plant (which incidentally also explains much of the cost for nuclear power generation, it's essentially a thermal plant with added complexity, so can't really be significantly cheaper than a coal plant).
Korea builds a lot of nuclear and if anything its only gotten more expensive.
There’s clearly some value to consistent power sources producing clean energy but if SMRs are in the $0.15/kwh wholesale range, they’re DOA.

Solar+battery gets you there today, is getting cheaper every day, and there’s almost zero project risk. If the install is small enough, you could likely beat the realistic costs for SMRs with solar + battery + diesel generators if you need a guarantee of 100% uptime.

> Solar+battery gets you there today, is getting cheaper every day, and there’s almost zero project risk.

Debatable. In Ontario nuclear costs 10¢/kWh while wind costs 15¢ and solar 50¢ (Table 2):

* https://www.oeb.ca/sites/default/files/rpp-price-report-2022...

And when wind goes to zero at night, then (natural/methane) gas generators are often spun up (generating carbon emissions).

Meanwhile the refurbishments of Ontario's nuclear plants are on-budget, and often ahead of schedule:

* https://canada.constructconnect.com/dcn/news/infrastructure/...

> Solar+battery gets you there today, is getting cheaper every day

Solar + battery cannot be sufficient for most Europe except Mediterranean countries, it's not a matter of cost, there's just not enough sun in Winter and you need months of electricity worth of storage which isn't happening in our lifetime.

In general, talking about energy price ($/kWh) only make sense when you have fossil fuel as a near majority of your mix (because you have practically unlimited power as long as you spend money on fuel), but cease to make sense without it, because nobody cares about energy (Wh), what you (and the grid) need is power (W). With either nuclear or renewable, energy is practically free, but power is what costs money, and as we move towards a decarbonized mix, we'll need to change how the economics work to adapt to the underlying changes (including how we price electricity to consumers and businesses), because when you don't align the economics with the how the supply works the system collapses (like it did in Texas as few winters ago).

>Solar+battery gets you there today

Are you extrapolating rooftop solar rates in California to a large, industrial solar installation in Canada, or where do these figures come from?

Solar + battery: will there be enough raw material available to make it a viable alternative at world scale? I am hearing that we won't be able to extract enough metals for a world transition based on renewables. Cost is secondary when physics gets in your way.
In terms of cost per ton moved, trains and semi-trucks outperform taxis. And yet taxis still exist. Turns out cost per vehicle (or per reactor) is important in some situations.
Would this not relegate this option to niche areas?

Elsewhere it needs to compete on price against other viable options.

Yes. Also, there really aren't any such niche areas save some military naval applications. Even there, most ships in the US navy are not nuclear, because burning liquid hydrocarbons is cheaper.
What do you mean? The lower unit costs is exactly the point of SMRs.
Unit cost of reactor not unit cost of power ?
Also, not every application of nuclear energy requires 1GW...
This needs to be recognized as an indictment of our society. An inability to accomplish anything in a collective and intentional manner is only one of the many disasterous outcomes of the wretched individualism promoted by the neoliberal agenda. And make no mistake; this is not a result of a so-called "free market". It is the result of top-down privatization. The US government, for example, has a much larger budget than ever before, accounting for inflation and so on. It's just being delivered to private firms for profit instead of on building big things or maintaining the big things we used to build. This lack of cooperation across society is also the primary source of cultural breakdown and the sense of "division" felt by all of us, both rich and poor.
In much of Western Canada, energy use is highest at night in deepest winter when there is no appreciable wind, meaning that the grid is completely reliant on gas-fired power. If we want to reduce gas-fired capacity, we need another source of on-demand power. Meaning either nuclear or a robust storage solution. There seems to be more enthusiasm for the latter, but there’s no buzz about any viable solutions on the horizon.
Nuclear isn’t great for ‘peaker’/infrequent usage. The large up front capital costs means they require running at 90%+ utilization to be economic, and the way the radioactive decay chains work means they tend to take awhile (hours) to stabilize at their high power outputs, and hours to days to ramp down in actual effective power. Even a scrammed ‘hot’ reactor takes half a day to a couple days for daughter products to burn down to the point it doesn’t produce significant thermal power (hundred of megawatts to even half a gigawatt thermal).

That’s what got Fukushima btw - when they shut the reactor down and then the backup generators got destroyed, they lost their ability to pump water to cool the reactor (which requires significant electrical power), which proceeded to start to melt down the core, and causing massive hydrogen buildup, eventually blowing up the reactor building.

Some new designs allow more effective emergency passive cooling, but the issue remains - nuclear plants are great for baseline power, but they aren’t good for sub-day, hourly, or finer grained peaks. Both economically and technically. Think ‘fully loaded container ship’ or ‘multi-mile long train’.

Pumped storage, battery, or fossil fueled turbines are great for those faster reactions - and often can provide useful sub-second grid stability too. Think ‘speed boat’ or ‘passenger car’.

Wind power produces more at night. See eg here for an explanation: https://www.reddit.com/r/askscience/comments/10mvxm4/comment...
Hydrogen is a solution for seasonal leveling at high latitude (and for Dunkelflauten.)
I can’t understand why, with winters known to be how they are, there is no district heating. If that infrastructure had been in place, a transition to carbon-less sources of heat could be well underway.
Are there no rains in western Canada? Why is hydro not an alternative?
SMRs are 'fine' for smaller grids. For example, the province of New Brunswick has a CANDU 6 reactor, which is 1x660MW of capacity:

* https://en.wikipedia.org/wiki/Point_Lepreau_Nuclear_Generati...

CANDUs are pretty flexible in that there's a lot of maintenance (including refuelling) that can be done while it is running, but there's still some stuff that needs to be done when the system is powered down, which means taking down a large source of power for the grid.

If there were 2-3x300MW reactors, when there could be rotating maintenance without much impact to the grid.

Some people present Alaska as a good place for SMRs. But the largest grid there, the Railbelt Grid, has an average power flow of just 600 MW. A 300 MW reactor would be too large to integrate into that system.
CANDU are a great design; the calandria is a nice adaptation to the capabilities of Canadian maufacturing, and it is impressive to see the lines of reactors at Bruce, Pickering and Darlington. It is also a widely exported design, with China constructing reactors on a 4 year timeline. Granted a single large reactor would dominate the grid, but I presume there is grid connectivity to adjacent provinces. An alternative would be a large energy user (aluminium, heavy water production) to soak up the excess power and turn down when one unit comes offline.
One thing I was surprised to learn about nuclear costs is that there's a fixed component that is basically the same regardless of technology (SMR vs 'LMR'): the civil works.

There's a certain amount of concrete and such that needs to always be built, and if you go with a "cheaper" SMR, then the fixed cost becomes a large portion of the total project budget.

So unless there's a specific local need for ≤300MW, it might be better to go with a 600/900(+) MW design if you can tie into a large grid where all of those 'extra' MWs can be soaked up.

Does this prediction include the effect of cost and time overruns? These overruns are mentioned in the article, but then not dealt with at all -- but it is very important whether SMRs are more expensive according to made-up "planned" numbers or actual costs.

Flyvbjerg mentions SMRs as an example for modularity in his book, How Big Things Get Done, and predicts that they will be much less prone to overruns because experience can be accumulated along a series of reactors, whereas traditional reactors are one-off, bespoke projects which directly implies that they will be built with a lack of experience. Even if a nuclear power plant gets built that is "like" an existing one, it is never the same.

> traditional reactors are one-off, bespoke projects

TBF they don't have to be, obviously they don't number in the hundreds and site-specific concerns matter (especially as they have a large surface, they matter for SMRs too but the footprint means they matter a lot less). However you have to commit, hard.

During its buildup, France built 54 reactors of just two classes (34 C-class, with 3 variants, and 20 P-class, with 2 variants) in 22 years: https://fr.wikipedia.org/wiki/Liste_des_réacteurs_nucléaires...

If the overruns are due to salesmanship and lowballing, not unforeseeable problems, then one can expect SMRs to be subject to them just like traditional nuclear. This was the UAMPS/NuScale experience.
The cost over time generally tends to weigh heavily on the initial budget. Small reactors don't replace large ones for large needs, even if multiple are built (this hasn't even happened yet!).

What small ones can do is afford either government/public/private energy sources in localized areas. Infrastructure was built upon technology stacked on top of previous; dirt to stone, stone to asphalt, and on; etc.

The same is inherent with nuclear. It is easy to tie in to the existing grid, but the grids are extremely out of date for the growth of populations in general.

A large mix of SMR's could absolutely fuel energy needs in both the short and long term as technology continues to improve. The cost is a metric of current economics/interest. That's the problem right now - perspective states it's unaffordable because we've pivoted it that way.

It seems to me there's a lot more 'mass production' efficiency still to be found for small reactors. The whole article is just comparing historical costs.

Large reactors are highly bespoke, and therefore increasingly expensive, and the whole idea of small reactors is that they can be produced identically from an assembly line, and so you get the same kind of learning-curve price reductions as for solar panels or whatever.

I don't think that's the case, yet, though, so the argument might be compared to criticisms of solar in the 1980s and anyway this article seems to be a simple attack on nuclear in favor of 'renewables', which have already experience a learning curve.

You can produce significant fractions of a large modular reactor (e.g. AP1000) in a factory.

There are arguments that for a small modular reactor the civil works that happen onsite (e.g. foundations) could make up a larger proportion of the costs than for a large modular reactor!

Just how many small reactors do you think would have to be made to reach this promised land of "mass production"?
The United States Navy, which operates 83 nuclear-powered ships, demonstrates that small reactors are clearly a viable energy source.
Interestingly, if you said that “The Federal Government” should build and run small nuclear reactors, I bet you’d get a lot of pushback; I’m always fascinated by how “the military” is culturally considered very different from “the government” in the US. The military is perceived as able to do most anything while the government is considered lazy and wasteful.
Is energy independence and lower land use priced in?
There is a lot of empty practically free land in the world.

Also, it is estimated that it would take about ~7 million acres to power the US entirely with solar. The US currently uses about 40 million (!!!!) acres for corn for ethanol.

Land is just not a problem, at all.

https://elements.visualcapitalist.com/how-much-land-power-us....

https://frontiergroup.org/resources/ethanols-outsized-place-....

Land is a small fraction of the cost of a renewable energy source (save for biomass), especially in lesser populated regions.
While "common knowledge" says small can't compete with big when it comes to price/unit due to economies of scale, this article reads as very partial and anti-nuclear so I would trust the things that were said and the facts that were left out.

From the outside, this SMR situation looks a bit like monolith vs micro-services where there is a great deal of non-technical reasons why SMRs are a route being taken and it appears it's mostly political and organizational ie. approve and build the damn thing.

> Vogtle nuclear power plant being built in Georgia, US

It was finished on April 29 and has been online since that time.