Recognising that this is a minor portion of a larger post - there are a lot of industrial companies that couldn't possibly cover the damage if something goes badly wrong (eg, explosives supplier flattens a major port, 3rd world mining accident, electrical utility causes wildfire, most mid-sized building companies if a skyscraper collapses, etc). Even many insurance companies won't pay out if a flood or fire hits a city, because they can't afford to cover large-scale disasters. Contamination in drinking water is also a pretty big risk. At that level of probability, the only way of dealing with these threats is government. No other organisation else has the manpower or resourcing to deal with rare catastrophes.
It is a fact that liability for a nuclear catastrophe would overwhelm a companies ability to pay. However, we do a lot of things that are more risky and more costly at that level of probability. The fact that there is risk is a very real consideration. That fact we can't insure it is not - there are too many things where, if a rare event happens, the liability cannot be shouldered by a corporation.
Fukushima cleanup, one of the worst disasters we've seen in 30 years, seems to have cost about $15 billion damages + $60 in compensation [0][1] vs $250 billion due to the actual disaster that caused the nuclear plant to fold (the tsunami). And this is as an unexpected, low probability, cost. If a corporation can't handle this, it is tiny bikkies to a government to cover the $15 billion in actual cleanup. And the probability of the government needing to act is tiny to start with. The risks here are so small we don't even talk about them outside the context of nuclear power.
[0] https://en.wikipedia.org/wiki/Fukushima_disaster_cleanup#Cos... [1] https://www.forbes.com/sites/jamesconca/2016/03/10/after-fiv...
Of course this is why in developed countries they are required to buy insurance. Insurance companies exist which can take on extremely large liabilities for industry, the point is that Nuclear has never been able to afford the size of premiums which would make the full risk of its liabilities profitable to cover.
Fukushima was by no stretch an example of the worst nuclear event liable to occur. A bad accident or attack on any one of the worlds hundreds of nuclear plants CAN seriously irradiate many thousands, even millions of people in some locations, poison water tables and make land unlivable for decades - continuing harm to many creatures even if people are able to avoid it.
After each disaster like Chernobyl or Fukushima we don't just have a new "rate of deaths so far" line to work with, that average line is a one dimensional extremely simplistic measurement. Scientists and engineers who have taken it and as a valid summary of hazard should really know better.
Calculating an actuarial probability was never the problem. You can't know the lower bound on the probability of something that has never actually happened, but you can calculate an upper bound based on the period of time it hasn't happened, which is all you really need. They can calculate what the premiums should be.
The issue is that you're asking them to carry a hundred billion dollars in insurance, but insurance companies have capitalization requirements. AIG's market cap is less than seventy five billion. They can't write an individual policy they don't have the resources to pay out on in the event of a claim, regardless of the probability that it actually happens.
So the problem isn't the risk calculation, it's the amount of insurance you want them to carry. The solution is to find someone who could actually pay out a claim that large in the unlikely event that it comes to that, which could only be the government, so that's what they did.
> Calculating an actuarial probability was never the problem.
Of course is wasn't - revealing and paying the resultant premiums is the problem, which is why governments make alternative arrangements for nuclear.
[1] - https://www.worldatlas.com/articles/largest-insurance-compan...
I think you mean two thousand billion. Which is problematic since there are more than twenty nuclear power stations in the US (each of which would need to be insured), even if that wasn't revenue.
Insurance companies are competitive. If they take in a hundred billion dollars, they pay out more than ninety in claims the same year. They can't pay a claim using the money they're already using to pay other types of claims.
> Of course is wasn't - revealing and paying the resultant premiums is the problem, which is why governments make alternative arrangements for nuclear.
The resultant premiums wouldn't be that bad if it was only the risk of a nuclear disaster they had to account for.
But a smart insurance company is going to take into account that US courts have a way of connecting sympathetic victims to deep-pocketed or well-insured corporations, even when there is only a tangential relationship with them, when it's the only way to save a devastated community. For example, if Fukushima had happened in the US, there is a decent chance that courts would have tried to tack a lot of the cost of the damage from the tsunami on the plant operators. Or for another example, PG&E and the recent wildfires in California.
That's a different kind of risk, but it's one an insurance company that doesn't want to go bankrupt has to account for, and it's also the one that makes buying that amount of insurance prohibitively expensive in the private market.
This statement > "which is problematic since there are more than twenty nuclear power stations in the US" shows a misunderstanding about the insurance business. It is not necessary to have assets to cover worst case events everywhere simultaneously. I expect you do understand that but have not given this subject your best attention.
Because they expected to. If you take in $145B in premiums and pay out $135B in claims, everything is fine. If one year they took in $145B in premiums and paid out $235B in claims, where does the rest of the money come from?
In theory they could collect the premiums and accumulate them over time so that by the time there is a claim there is the money to pay it, but that allows you to pay a single claim of that size in year 50, what is covering it in year one?
> Nuclear plants liability are capped in north america at 13 billion USD the nuclear industry is simply exempt from arranging full cover - unlike other industries.
What other industries? I don't see oil and coal companies each having to be insured against claims for billions in damages from climate change. PG&E isn't covering the full liability from the camp fire, they're filing for bankruptcy.
> This statement > "which is problematic since there are more than twenty nuclear power stations in the US" shows a misunderstanding about the insurance business. It is not necessary to have assets to cover worst case events everywhere simultaneously. I expect you do understand that but have not given this subject your best attention.
That's assuming you can consider them all to be independent. The problem with very low probability events like this is that if they do happen the cause is often another very low probability event. You get something like the Yellowstone Supervolcano or a coordinated terrorist attack and all the claims come at the same time.
Or even at the same time as other types of claims -- you get damage to a nuclear plant or three from a ten thousand year storm or an 8.5 magnitude earthquake and now you're paying that claim on top of the disaster claims from what caused it. Insurance companies want to pool risk, not assume 100% of the damages from a massive disaster with a single dependent cause.
The idea of holding oil companies liable for climate change has a ring to it, but insurance wont cover it because there is no legal precedent. There is legal precedent to claiming damages on industrial accidents and nuclear industry accidents. If one nuclear plant fails badly for any number of reasons and fallout is blown towards a city or across valuable farmland, the claims could easily exceed 13 billion USD.
Rarely in the amount of the theoretical maximum amount of damage they could possibly do. In practice when a company has more liability than insurance coverage, they end up in bankruptcy and it's the insurance policies of the victims that end up paying out, or some kind of government assistance. This is especially true for very rare and very large claims, because nobody really carries that much insurance.
For example, Boeing no doubt has insurance, but their worst case failure mode is something like the hypothesized worst-case scenarios for the Y2K or Y2038 bug where some integer rolls over and every plane they've ever made crashes into the ground at the same time. There is no way they have enough insurance to cover that -- it would be a trillion dollars. But neither do we expect them to, because it's very unlikely even though it could theoretically actually happen.
The same kind of thing could happen to a major auto company, or really any company that runs the same software on a million different pieces of industrial hardware. A company could sell a million pacemakers and have them all fail at once, or elevators, or emergency dispatch equipment. None of them are insured for that amount of liability.
The oil companies are rare in the sense of having extremely deep pockets, enough to cover a major claim like that on their own, but they're about the only instance of that. And even they would be completely bankrupt if the courts ever start allowing claims against them for the damage from climate change.
What does ”uninsurable” mean? If nuclear power plants are uninsurable Silicon Valley is uninsurable.
Silicon Valley is insurance because the liability exposure is substantially less.
It is easy to bring up problem after problem with alternatives. It is not easy to show why the status quo is somehow better, because it is not.
A death is a death. Someone dieing because they fell off a roof is equally as bad as dieing due to radiation.
I do support an alternative solution: making rooftop installs safer, and more utility scale ground mount solar. Both are infinitely more tractable as problems than properly managed nuclear power. The US is already on track to install more utility scale solar this year (~150GW) than total existing nuclear generating capacity, even accounting for capacity factor. One year!
It should be obvious when you think about it; if "it'll never happen" (because it can't or hasn't happened) is true, then insurance actuaries could establish that and there would be no problem with insuring the "safest source of power", but governments have to wave the responsibility for full liability instead. There is a real probabilistic risk of great and persistent tragedies from nuclear plants, the more there are the greater that risk.
Fortunately wind, solar and storage are cheaper already than nuclear [1] and their prices are continuing to drop fast. There may be a good case for improving safeguards on these industries too - it should not be as expensive as trying to fully safeguard nuclear has been.
[1] https://www.lazard.com/perspective/levelized-cost-of-energy-...