Pollution from Big Tech's data centre boom costs US public health $5.4bn
ft.com
ft.com
"The analysis does not account for the purchase of market-based instruments that are meant to represent investments in new renewable energy in the US and that tech companies buy to offset the pollution from their electricity consumption. ... 'Unlike carbon emissions, the health impacts caused by a data centre in one region cannot be offset by cleaner air elsewhere,' said Shaolei Ren, associate professor at UC Riverside."
Why not? Does clean air elsewhere not matter for the individuals elsewhere? Where is that "elsewhere?"
My admittedly ungenerous interpretation: factoring this context in would be really hard and would decrease the headline factor of the findings, so... publish without.
If your house had a carbon monoxide leak but your neighbor’s house did not, would you consider your carbon monoxide leak to be “offset” by that fact?
Not sure why we need an analogy, though. How about sticking to standard air pollution, which has direct impact of its own? CO poisoning seems a bit extreme as a comparison.
Likewise if you pay someone to plant a tree that'll reabsorb the carbon you've emitted over the course of 40 years, and they cut it down and burn it after 15 years - the carbon's still in the atmosphere.
So putting your data centre right next to a hydroelectric or nuclear power plant is the gold standard.
[1] https://www.theguardian.com/environment/2023/jan/18/revealed...
Where are you getting that from?
https://www.eia.gov/todayinenergy/detail.php?id=61364
According to the UN, China's share of mining dropped from 73% in 2020 to 21% in 2022. They also state
> The greenhouse gas emissions of Bitcoin mining alone could be sufficient to push global warming beyond the Paris Agreement's goal of holding anthropogenic climate warming below 2 degrees Celsius.
https://unu.edu/press-release/un-study-reveals-hidden-enviro...
In 2024 the percent of global energy usage for crypto mining has dropped to around 2% according to the IMF
https://www.imf.org/en/Blogs/Articles/2024/08/15/carbon-emis...
Someone less lazy than me could probably take some of these statistics to get an estimate of global energy consumed by crypto mining at its peak in 2020
That's not true for other pollutants though, which is one of the reasons that e.g EVs shifting emissions to powerplants is good because cars are used in dense urban environments where the pollution affects more people.
I've seen the work that big tech has published to justify the additionality of their clean energy purchases and offsets. They have the data to refute this if they can do so. I'd guess they can't due to their renewables and other offers being further away from population than these data centres.
Why do people keep adding "ality" to words? "Addition" works fine here:
"I've seen the work that big tech has published to justify the addition of their clean energy purchases and offsets"
Or more broadly "the extent to which something happens as a result of an intervention that would not have occurred in the absence of the intervention."
If you do something that causes air in one place to get 1 ppm dirtier and air in another place to get 1 ppm cleaner, and the populations are the same, the net impact on health cancels out.
But in general I agree that you should be able to look at tradeoffs of a set of actions and allow of the possibility that some negatives are offset by positives.
Not exactly. The other place might have people die from poverty, illnesses and godknowswhat much earlier so the effect of pollution shifting doesn't manifest itself or even gets masked off by rising life expectancies in general - that's the dirty secret behind the move of dirty and toxic productions to Asia.
The average ppm cancel out but the outcomes don't, because you have more sick people: Let's say 12 ppm is the maximum safe ratio for some pollution [edit: assuming it is a pollutant that has local impact, not like CO2 which has global impact]:
1) If place A has 10 ppm and place B has 10 ppm, then nobody is sick
2) If place A has 15 ppm and place B has 5 ppm, then people in place A are sick
Public policy generally doesn't work well with averages and similar analyses, because outcomes are usually discrete for each individual; it's not a stew where you can mix outcomes together and get something good.
As another example, if the economy results in one person making $1 billion and 999,999 making nothing, the average is $1 million per person - but what does that mean? 999,999 people are still in dire poverty. (It does have some significance, for things society does collectively - fund police and fire, and even help for poor people.)
It's actually slightly sub-linear, meaning that an increase for one group and an equal decrease for another group produces slightly less total mortality. See https://pmc.ncbi.nlm.nih.gov/articles/PMC9109601/#gh2329-bib... for some numerically fitted curves (search for GEMM).
Now if they offset by buying scrubbers that cleaned similar particulates elsewhere in the world you'd have a "average stayed the same" but the problem moved.
For instance, "Data centres cause pollution through high eletricity use" - high relative to what, residential housing? Aluminum smelting?
It's fine to be skeptical of the AI bubble, but this article (ironic for FT) reads like veiled de-development.
We also don't need any fizzy-drinks - shall we calculate the deaths due to pollution caused by that, let alone the primary effect on health of drink-victims? If you don't like what electricity is being used for, argue for specific regulation to make it more expensive.
Hard to tell if they're singling out AI because it's the current hot thing that everyone's talking about (and afraid of being replaced by), or if they're singling out "big tech" because they're rich and cool to hate on.
The journalists then learn what their audience responds to and creates narratives around that.
The result is the reality and the narrative often differ dramatically.
"News" is just "Reality TV" but in text format.
[1] https://www.theguardian.com/technology/2022/feb/18/bitcoin-m...
Nobody is writing "providing heat to elderly people in northern cities costs US public health $5.4bn" because most people feel that "keeping people from freezing to death" is worthwhile for society.
The problem is that's basically entirely based on vibes. For crypto opponents it's justified by crypto not doing anything that the traditional finance system can't do, and for AI opponents it's justified by AI only being able to rehash soulless copypasta while stealing jobs. As a result the whole argument just becomes a roundabout way of arguing "[technology] sucks".
Like, in what other context would you raise this concern about saving lives?
It's the same kind of dishonesty, picking out one hot-button source of pollution, apportioning out its health impact, and complaining about it alone.
[1] https://slate.com/technology/2016/12/self-driving-cars-will-...
https://www.irodat.org/?p=database#data
I worry about a public opinion chaining on switching to opt-out in the US so it has to evaluated carefully.
I'm guessing the US wait time is long because there are a lot of people with failing kidneys due to poorly managed diabetes and high blood pressure. But there are a whole host of auto immune and genetic diseases also.
Now we have better drugs for treating diabetes like Ozempic and SGLT so hopefully that improves the situation for kidney failures caused by diabetes.
I myself had an auto immune disease IgA nephropathy 10 years back but at that time there were hardly any drugs targeting it. I've since was on dialysis and recently got a kidney transplant. But the positive at the same time a half dozen drugs have been approved or under trial to slow down the progression of IgA nephropathy.
I mean, why not just say the organ shortage is bad, without pointing the finger at the latest hip, hot button issue, out of all the possible life-savers?
Of course, down this path too far and you start looking at prisoners and death sentences and organ harvesting, which I think most people would agree ain't right.
The article is part of a series about challenges with future technologies. They note that 1/5 of donations come from traffic accidents and there is already a major shortage. If self driving cars reduce most accidents that will mean nearly 1/5 of donations will need to be replaced
> If you don't like what electricity is being used for, argue for specific regulation to make it more expensive.
That reads like veiled de-development.
How do you reconcile saying both those things?
Same with this, I feel. It's not anti-pollution, it's anti-AI.
Originally cars ran on leaded gasoline, which was very bad. But we didn't ban cars, we got unleaded gasoline. Then they were emitting too much in terms of sulfur and nitrogen oxides, so we got low-sulfur fuels and catalytic converters. Then it was brake dust from friction brakes, but now hybrids have regenerative braking. Then it was CO2, so we got electric cars that can be charged purely from the sun -- they can emit less CO2 than most mass transit systems despite the lower passenger density because transit systems regularly use diesel buses or locomotives or run after dusk on electricity generated from coal and natural gas because they don't have internal batteries that can charge exclusively from the sun.
Somebody is going to invent some tires that last for 800,000 miles because they're made of something with such high internal cohesion they never get bald, or erode into something already commonly found in nature, and then people would find something else to complain about. And then someone would solve that too.
"Climatologists" are talking about CO2, which is the thing we already have everything we need in order to solve. You use electric ground vehicles, charge them with solar panels or nuclear reactors. Use synthetic or biofuels for aircraft. Switch to electric heat pumps to heat buildings. This isn't a matter of "we don't know how to do it" or "the only option is to reduce energy consumption", it's only a question of whether people are willing to adopt these things fast enough, because adopting them faster sometimes costs more in the short-term. So that's still a thing we actually have to do, but it isn't really a question of what to do, only a question of how to get people to do the known thing.
With how much life left on its battery before it needs a $10k+ replacement?
> In fact, many EV batteries may outlast the vehicles they are installed in, then enjoy a second life in a stationary storage application before finally being recycled, according to EVANNEX. “At the end of the vehicle’s life — 15 or 20 years down the road — you take the battery out of the car and it’s still healthy with perhaps 60 or 70% of usable charge,” Thomas said.
Meanwhile the battery replacement cost isn't $10k+. Nissan dealers charge $5500 for a new battery. Refurbished batteries are available for as little as $1000, with the obvious trade off that you'll have to replace it again sooner, but $1000 for another 3-5 years (with the option to do it again) could be a better choice for an older car than $5500 for another 10-15 years, depending on how much value you place on the inconvenience of disassembling your car to put it in again.
There is also an obvious market response if something like that was actually introduced. If burning carbon cost more, people would buy more fuel efficient cars. Then those cars enter the used market a few years later.
This is one of the reasons the only sensible carbon tax proposal is the one where all of the tax revenue is distributed back to the population. In the first year, there will be more demand for efficient cars than existing supply, so they'll cost more. People who drive a lot will buy them anyway, but can use the refund to offset the cost. People who don't drive as much would eat the cost of the carbon tax on a less efficient vehicle, which isn't so bad because they don't drive as much, and then gets offset by the refund until more efficient cars filter into the used market. When that happens the tax revenue goes down because people have more efficient cars, so the refund gets smaller, but by then people don't need it as much because they have the more efficient cars.
If all we have is opinions, let’s go mine. Fortunately, we have data.
Take Australia for example.
1. We had a carbon tax, sort of, briefly, but no one wanted it, and still doesn’t.
2. Fuel for vehicles here, petrol and diesel, has become wildly expensive, but cars have become huge to gigantic. Approximately no one drives a small car, approximately everyone owns at least two 4x4 and or SUVs, and those that don’t tend to drive a turbo charged Skyline or Subaru or similar.
Everyone drives relatively new cars compared to, say, 30 years ago when half the population drove a rusty floor pan around, you were luck to have anything resembling an actual car hanging off it.
No one here gives a shit about climate change, except maybe our Australia Broadcasting Corporation, but approximately no one consumes any of their media, and even if ya do 97% of the time their performing their mandate of proportional representation by publishing media aimed at the LGTBQI+ fictious community 100% of the time.
To the point where electric vehicle sales here in Australia have fallen off a cliff, and one hundred down ten to one the current buffoons get voted out and replaced by a potato promoting nuclear.
Good.
Australia briefly had something their politicians called a carbon tax, but it was really cap and trade, which is an entirely different system and is rubbish. In particular, you have to give ordinary people back the money the carbon tax collects or they'll rightfully hate it, which cap and trade does not.
And then, because the system they proposed was the miserable one, the opposition leader credibly threatened to repeal it (and then did), so hardly anyone bothered to try to save costs under a program they reasonably expected to quickly go away.
Actual carbon taxes have been implemented in parts of Europe and they do what they're expected to do.
> Fuel for vehicles here, petrol and diesel, has become wildly expensive, but cars have become huge to gigantic.
People used to say "buy a smaller car" as the way to save fuel, since that used to be about the only way to do it. Then carmakers learned you could design an SUV with the use of a wind tunnel instead of shaping it like a brick and then use a smaller displacement engine with a turbo, or a hybrid system, and then you get SUVs with better fuel economy than subcompacts got in the 90s.
Even the huge stonking V8s still get modern aerodynamics and get significantly better mileage than the old ones did.
> To the point where electric vehicle sales here in Australia have fallen off a cliff
Electric vehicles were never popular in Australia. Range is their Achilles heel and Australia is spread out. But then you do a plug-in hybrid that can operate as electric 90% of the time while completely solving the range issue.
Opinions / data, let's go with the data.
It's really not. Australia is mostly uninhabited, with most people living in very few high density places.
The average commute distance in Australia is 16 kilometres; 90% of Australians live on the East Coast; 73% of Australian's commute less than 20 kilometres.[1]
Even Western Australian has people living outside of Perth, which, by the way, is the most remote centre of commerce in the world, has an average commute distance of 20.4 kilometres.[1]
The average Australian vehicle travels about 12,100 kilometres annually, that's all vehicles including freight vehicles.[2] The average household in Australia owns 1.8 cars[3], so very many Australian's would be well served by owning at least one purely electric vehicle, but don't.
Plug-in hybrids are mostly not a thing here.[4]
1. https://www.abs.gov.au/ausstats/abs@.nsf/Lookup/by%20Subject...
2. https://www.abs.gov.au/statistics/industry/tourism-and-trans...
3. https://www.abs.gov.au/statistics/industry/tourism-and-trans...
Average commute or distance traveled per year isn't the interesting number. If you're considering an electric car with a range of 300km, it doesn't matter if your average commute is 16km or 100km, what matters is, do you take trips in excess of 300km?
https://finance.yahoo.com/news/more-aussies-holidaymakers-ta...:
> The number of survey respondents who have taken a road trip increased to 91.9% in 2024 from 88.4% in the same survey in 2022
And if anything, fewer total miles traveled per year would make EVs less attractive there because of less in terms of fuel savings.
> Plug-in hybrids are mostly not a thing here.
Plug-in hybrids are relatively new because they're basically normal hybrids with bigger batteries, so the batteries had to get cheap and light enough to make that work, and they got faster adoption in countries like the US because of tax incentives that favor PHEVs over traditional HEVs. Meanwhile from your own link, 25% of new cars in Australia are some type of hybrid of BEV, compared to 19% in the US. Consistent with "they are buying more fuel efficient vehicles", it's just that those vehicles are turbocharged low displacement or hybrid SUVs.
It's possible that the difference in tax incentives is causing carmakers to allocate PHEVs to the countries providing incentives specifically for those given a finite amount of battery production capacity, but according to your own numbers PHEVs are a good fit for Australians. Average commute is well within their all-electric range but then you still have no trouble taking the occasional road trip.
To reduce carbon emissions, one would put a tax on gasoline, natural gas, coal, and electricity from plants powered from those sources.
The thing is, that hits every single voter in the pocketbook.
And what's more, a voter on $20,000 drives about the same amount as a voter on $200,000 - so if the gas tax costs both drivers $1000 a year, the former will feel it a lot harder.
So our politicians are scared to do it.
They're afraid of the relentless propaganda campaign that will wash that reality away, funded by the concentrated forces that will lose business if forced to account for the damage their products cause.
Also, Americans think carbon fees are impossible, yet they're implemented all across the globe, including several US states. It's a political dysfunction that is particularly evident in the US, not a fact of life.
Quick googling found a stat that 1.6% of the whole world's electricity supply going on computing and also https://www.visualcapitalist.com/charted-big-tech-uses-more-... etc.
And we need more! On day 2 of his presidency Trump pledged to fast-track new power generation to drive data centres in the US. https://www.axios.com/2025/01/23/trump-ai-power-plants-data-...
Drill baby drill! Using electricity on compute is driving future dirty power generation.
I think we are back in 80s with current orange man defining direction like fashion, 'greed is good' at any cost mantra is certainly back
Greed pushes people out of the status quo.
Greed is why fusion reactors will be reduced in size, cost, and complexity.
That would imply that it isn't the major issue. Electricity generation as a whole is less than half of US energy consumption, US electricity generation is 35% nuclear or renewable, and computing is using 1.6% of electricity. This is a rounding error compared to e.g. transportation at 37% of all energy consumption, or HVAC at another double-digit percentage.
> And we need more!
New generation capacity in the US is disproportionately renewables because renewables will remain cheaper until they're closer to half of all generation and then have to really deal with the storage issues. Willingness to approve fossil fuel plants doesn't mean anybody really wants them. The large tech companies are buying nuclear:
https://www.nytimes.com/2024/10/16/business/energy-environme...
Not when that "dirty electricity generation" was planned to be shut down before the data centers arrived: https://www.theregister.com/2024/10/14/ai_datacenters_coal/, https://www.theguardian.com/technology/2022/feb/18/bitcoin-m..., https://www.theregister.com/2024/05/31/datacenter_power_crun..., https://www.theregister.com/2024/10/08/eric_schmidt_speech/.
https://westvirginiawatch.com/2025/01/16/west-virginia-isnt-...
West Virginia is not west Virginia.
Also West Virginia would more accurately be North Virginia.
So what do these degrowth people want exactly? Cant have resource extraction. Can’t have manufacturing, pollution. Can’t have tech, pollution. Just twiddling thumbs and money will come? Or keep all the benefits while offloading all the downsides to Mexico? And a few years later they will complain about dependence on foreign countries.
It feels like you're arguing against a strawman.
I can only speak for myself, but I would like CO2 emissions to be taxed appropriately, so that fossil fuel producers and consumers actually pay a somewhat proper price for the damage that they cause (negative externalities).
Sure, that might slow growth, but I see so evidence that this difference in growth has much of a negative effect on US citizens:
How much GDP growth do you think we actually need as a nation to stay satisfied? Double our GDP every 20 years? More? Less?
Because I know that I would trade a good fraction of my "growth boosted" past and present wage in exchange for climate change already being solved (and be happier as a result).
PS: I would also like to point out that any significant constant rate of growth is inherently unsustainable by definition; as a species, we need to arrange ourself with the fact that we'll be unable to continue growing our energy consumption exponentially-- the sooner, the better.
Yes. But air pollution is a negative externality that is not priced in for polluters/consumers, in exactly the same way as CO2.
You could argue that the case with air pollution is even more straightforward than with CO2, because those damages and costs are fully shouldered by the local population (unlike the consequences of climate change, which is probably gonna be felt more acutely in countries like Bangladesh, and thus less of a problem from a US citizens PoV-- note that this is not an argument I endorse).
> It also strongly implies “we don’t want you to pay, we simply don’t want you here”.
This feels like conjecture to me. The message of the article to me was "Tech giants operate data centers for hundreds of billions, but they are not paying for the consequences of their emissions which are mainly felt locally (costing billions)". That is not "we don't want you here" to me, but instead a "you should be responsible for the consequences of your actions".
I'm not exactly sure what your actual positions is. Are you arguing that industries/consumers should be exempt from paying the actual price for things (including external costs), because this (indirectly) subsidises local industry? And where do you draw the line? Because surely "the local gold extraction industry may freely pollute rivers with mercury" is across that line?
"The cost of treating illnesses has gone up due to X" rather than "more people are getting sicker due to X".
Makes it more of a detached conversation about pricing than about how something is hurting actual people, who perhaps have value beyond a purely economical one.
(I've always wondered how measurable the effect of subtly reducing the HFCS/sugar quantity in all sodas over decades would be.)
It’s easy to make laws about it:
- private jets/charters: only allowed if have enough passengers (calculate the CO2/person)
- old cars - illegal if don’t pass emissions check
"every bit" also costs political capital. Making any sort of change is going to cause a stir, and if you're going to cause a political shitshow, you might as well get good returns for it. Causing a political shitshow, but only making a modest impact is the worst outcome. Just look at the pasty tax in the UK[1]. The government proposed a change to the VAT rules that would have made a minimal fiscal impact, but generated so much flak that they were forced to back down.
[1] https://en.wikipedia.org/wiki/Pasty_tax#Aftermath
>start with the things with outsized effects with simple alternatives.
The examples you gave all cause disproportionate harm to a small group, which will raise a stink to their local representative, especially if they're being singled out.
An across the board carbon tax avoids both these problems.
Another thing to observe: The paper's claims are based on a McKinsey projection to 2030 that assumes 'a strong decrease of road transport emissions' and a 'surging demand for AI data centers' that 'outweighs power plants emission efficiency improvement'.
Contrast this to today where energy estimates are 2% for datacenters vs. 80% for road transport. Guess what is causing more outweighing of power plant emission efficiency improvement as that fleet of cars, trucks and lorries is electrifying (the reason for McKinsey's reduction projection)?
Now I'm not saying we should let data-centers get a free pass. Pollution is a serious problem. But in terms of priorities, curbing datacenters might not be the number one concern.
"The academics estimated that the cost of treating illnesses connected to this pollution was valued at $1.5bn in 2023, up 20 per cent from a year earlier. They found that the overall cost was $5.4bn since 2019."
Note that pollution is not defined in the article.
We wildly overengineered on safety regulations making them far more costly and risky to build than they actually need to be.
Even our older Gen II nuclear reactors from the 1990's are orders of magnitude safer than current fossil fuel power stations, in terms of both released radiation and lost human lifespan years.
Moreover, the higher electricity prices that resulted from using non-nuclear power lead to ICE vehicles remaining more prevalent than they should be, and non-heat-pump heating remaining more prevalent than it should be, leading to millions more years of human lifespan lost.
From a health perspective, blocking nuclear with overregulation was one of societies biggest health policy failures.
Didn’t we already learn better safe than sorry?
but I'm curious: what event do you think taught us that?
IMO, there's far mode damage being done to the world by, say, mercury in batteries than by nuclear accidents (of which chernobyl is the only serious one).
Nuclear Waste handling and disposal is a significant regulatory issue and includes many REACTIVE measures that have been hard-learned over the past century at sites like Pennsylvania, Fukushima, and Chernobyl.
This blind lashing out at any and all forms of regulation is ignorant. Each of these situations needs to be considered thoughtfully and contextually to find a balance between short term and long term objectives.
Coal generation has generally trended downward since the mid 2000s.
We don't need nuclear to replace coal.
Coal and nuclear are both reliable sources of energy, unlike solar or wind. I'm sure you're about to go into some battery related hand waving, but that doesn't help in two ways. First, there will be intervals where the lull in production exceeds storage capacity, which means fatalities in many cases. Second, battery is expensive, so the total system cost can exceed the cost of nuclear (especially with sane regulation and continually improving engineering).
Also, as far as "coal trending downward", that isn't the case outside the US.
We're going to need a vastly increased energy supply in order to meet the world's needs. Nuclear fission absolutely needs to be part of that mix, as we work to master fusion and other advanced, safe, and environmentally responsible forms of energy production.
Beyond that, high density energy sources are highly desirable off-planet. :-)
Relying on fusion power to clean up electricity generation is highly irresponsible, because in every remotely credible scenario, rollout is MUCH too slow to meaningfully affect climate change.
But I'm interested in trying to understand your view, and also which fusion power approach you put your faith in (and generally discussing this).
I strongly believe that nuclear power has no future, because it performs very poorly in an energy market where the marginal cost of producing is very frequently near-zero (thanks to renewables). Nuclear power is already quite expensive-- only running the plants half the time ruins cost competitiveness completely. You can see this happening already in countries like France and China, where nuclear and coal power plants are increasingly operating in load-following mode (i.e. not 100% all-the-time), which makes them even less cost competitive than in the past.
> I'm sure you're about to go into some battery related hand waving, but that doesn't help in two ways. First, there will be intervals where the lull in production exceeds storage capacity, which means fatalities in many cases. Second, battery is expensive, so the total system cost can exceed the cost of nuclear (especially with sane regulation and continually improving engineering).
Tow points here: Intermittency is a problem that actually shrinks at scale: The more spread out your wind parks/panels are, the smaller the periods where they provide unexpectedly little power (=> you need less buffering than you would naively assume).
Second point: a 100kWh battery is already affordable for a single household right now (thats basically big electric car battery). Price trends only go one direction there...
> Second, battery is expensive, so the total system cost can exceed the cost of nuclear (especially with sane regulation and continually improving engineering).
From this I assume you believe that nuclear reactors are not cost competitive mainly because of safety regulations, and "simply" fixing those regulations would make them able to compete on cost? This is likely incorrect. Consider:
Coal power plants are basically a minimally regulated, built-at-scale, super-simplified variant of a thermal power plant. They are the "ideal" that hyper-optimized nuclear reactor designs will never be able to reach (disregarding fuel costs here!). Even so: They struggle to compete with renewable on price already (disregarding fuel costs!). What are your thoughts on this?
My personal favorite: Tax carbon emission, use gas turbines as peaker plants, store energy long term via synthetic hydrogen, which is needed for carbon-free steel production anyway (and can also be used by gas peaker plants).
Then just let the market find out which fraction of batteries, wind, solar, carbon-taxed gas/coal, nuclear etc. works best.
I think you missed the point. Coal-fired plants leak radiation significantly. If this is not morally unacceptable in one case, why would it be in another?
They did not say that nuclear should have a free pass, just that the regulations should be consistent.
> And of course you could apply the same thing to other sources of energy and they will equally get cheaper.
Different technologies have different downsides and advantages. The parent was talking about 2 technologies having inconsistent expectations for the same metric. Personally, I do not want wind turbines everywhere or a nuclear plant in a city centre, but I think that both technologies (plus solar and hydro) need to be developed, because both have much smaller ecological and human costs than the alternative.
> Dam evey river.
That is counter-productive as there are diminishing returns, and not every river is worth daming.
The real question is if humans really can tweak the tradeoffs like this. Can you really change a design to be less safe on a controllable way.
Also, as you say different technologies have different advantages and downsides. That does mean different regulation.
But given current installation base, it's like 3x more deaths per unit energy than wind.
I am in favor of more nuclear power and I know very little about the field, but this argument never resonated with me.
If the Chernobyl series paints a reasonably realistic image of what could have happened had they not contained the melt down, we are talking "better part of a continent being uninhabitable for thousands of years". When that is the potential risk, comparing death counts from non-catastrophic scenarios isn't what we should be looking at.
It's not a realistic risk.
Chernobyl was absolute worse case. A massive reactor, with no containment structure (it was essentially in a shed), with a flammable solid state moderator, and no passive cooling. Yet, Europe is not uninhabitable today, infact the above-natural-background radiation in Europe is almost entirely coming from burning fossil fuels, not from Chernobyl.
In this case it's being used to mean that the regulations are overly restrictive or complicated, and impose higher costs than are reasonable for the benefits they provide. (Or potentially that being overcomplicated actually makes them less effective than if they were done properly.)
We know from natural disasters that insurance is usually insufficient to cover the true damage. Either governments have to step in, or the victims have to bear the costs. If the government has to pay for damages but someone else pays for safety features, there are clear incentives for overregulation.
The company operating the reactor obviously cannot pay. As long as shareholders are shielded from liability, it's easy enough to structure the company to ensure that it has no major assets beyond the power plant.
The problem you posed is just a market structure implementation detail. Government could operate the plants, or they could offer insurance to private sector operators.
If someone gets killed at work, how much do they get from workers' comp? What is a micromort and how much does one usually sell for? What did weregeld used to be back when it was still a thing?
I've not seen numbers updated with the latest data, the old data had them basically drawn on death per TWh and in the time between that data being gathered solar and wind have gone from a rounding error even when combined to each individual predcited to generate more than nuclear next year, so double nuclear generation per year and still accelerating away from it.
https://ourworldindata.org/grapher/death-rates-from-energy-p...
Fossil fuels also cause hundreds of billions of damage, simply when being used, no need for accidents. Your argument is useful, because it highlights that we are better equipped to deal with a predictable level of damage than rare events. But overall, having more nuclear plants and using less fossil fuels over a couple of decades would have left us better off by billions of dollars.
You now have 70% of budget to build battery/wind and still got enough time.
This is solved with current renewable prices without ideological preferences.
What is really needed is for humanity to say "enough" and stop using fossil fuels. Would it be a problem if training AI models (or running any other analytics) be lot more expensive in winter than summer? I doubt it. There is no reason not to use more renewables for this use case, then.
For example, the Gen II+ plants built in the 90's had an (inflation adjusted) price of about $500 million per GWe. Whereas, the cost of the recent Vogtl build was $15 billion per GWe!
Renewables proponents point to the cost of Vogtl (or similar) when they say that Nuclear has a higher cost than renewables, which is dishonest. The incremental cost of new reactors is far lower when you build many of them. The recent projects are mostly learning curve costs because of how long it's been.
The other part of this is that expensive silicon is not going to sit idle. It needs baseload power, and for solar that means adding batteries, and also downrating the nameplate capacity to it's wintertime capacity. These factors increase the real world cost of renewable power significantly.
Don't get me wrong, I think we should be investing in renewable as well, but it's still a substantially more expensive way to generate power than a serious roll out of nuclear would be.
Why not? Just have the excess of computational capacity in many different regions and compute where the sun is shining bright now.
It's cheaper to build a 1 GW baseload power plant to sit beside your existing GPUs, than to buy another 1 GW of GPUs to place on the other side of the planet. Also, latency.
Also, a large scale roll out of nuclear would be cheaper than solar regardless. Most of the cost of recent nuclear builds is learning cost. We build them so infrequently that people retire or move on before the next one is built, so we have to learn how to build them from scratch, and develop the supply chains from scratch each time.
[0] https://ru.wikipedia.org/wiki/Белорусская_АЭС#Стоимость_прое...
Unless, of course, it's the Vogtle plant or similar Western-built one, which cost about 17 billions per GW and took 10 years to construct. In this case, the price is comparable if you factor in the cost of money.
I fully agree that nuclear plants are cheaper in bulk (and when you can guarantee that there will be continuous future demand!).
But your numbers are extremely unrealistic. US GDP has basically quintupled since 1990-- we'll never be able to build nuclear plants for so cheap again.
Vogtle (and also current European and Korean reactors) give a much more realistic baseline for cost estimates than just inflation adjusting the price of old reactors that were built when everyone was poor (=> Baumol effect: https://en.wikipedia.org/wiki/Baumol_effect).
> It needs baseload power, and for solar that means adding batteries, and also downrating the nameplate capacity to it's wintertime capacity. These factors increase the real world cost of renewable power significantly.
Yes, but this is somewhat disingenuous framing; no consumer needs a baseload provider. Providing baseload is a privilege you get to enjoy when your marginal cost is much lower than anyone elses all the time. This is increasingly less the case for e.g. coal power in China, but even for nuclear plants in countries like France (mainly because renewables).
Nuclear power plants thus pair very poorly with renewables, because those eat into nuclear profits whenever they are available. It is highly doubtful to me that you could get nuclear plants price-competitive with renewables even if you committed massive amounts of money immediately, and the situation grows worse for nuclear basically every year.
Companies like Neoen show that with renewables/batteries, you can just install the things at scale while turning a profit, right now.
My conclusion is that any tax-funded investment into nuclear power is questionable/wasteful at best (and this also holds for fusion, but you can write that off as science, at least).