<THREAD>Why nuclear power plant accidents probably become significantly more likely in a war--the danger of a "common mode" failure at Ukraine's reactors.
Some wonk-ish thoughts to add to my analysis yesterday. (1/n)
carnegieendowment.org/2022/02/24/mos…
Some wonk-ish thoughts to add to my analysis yesterday. (1/n)
carnegieendowment.org/2022/02/24/mos…
Nuclear safety relies on defense in depth--multiple safety systems have to go wrong for an accident to occur.
The thinking here is that if a system fails only rarely, and its backup also fails only rarely, the the likelihood of both failing at the same time is very small. (2/n)
The thinking here is that if a system fails only rarely, and its backup also fails only rarely, the the likelihood of both failing at the same time is very small. (2/n)
The problem with this logic is that in a war, failure probabilities are unlikely to be independent, i.e. the conditions that lead to one system's failure could lead to the failure of its back-up too. This is called a common-mode failure.
Here's an example.
(3/n)
Here's an example.
(3/n)
Nuclear power plants use electrical power from the grid to cool a reactor if it's shut down. But power grids are often targeted in a war. Moreover, attacks on the electrical gird (not to mention other targets) can cause fires. (4/n)
Now, power plants also have emergency backup power sources (typically including but not limited to diesel generators). But no reactor operator wants to be without a grid connection with an approaching fire. (5/n)
Under normal circumstances, of course, the fire service would come and deal with the fire. But in a war? Are communications available to send out an alert? Are fire trucks tied up in relief efforts elsewhere? Would they be targeted in responding to the fire at the plant? (6/n)
It may not be possible to restore the grid connection quickly. So how much diesel is available on site (especially given that multiple reactors could be affected)? How much is available elsewhere (given the ongoing war effort)? Can it be transported to the site? (7/n)
In short, under normal circumstances, it's extraordinarily unlikely that a reactor would lose grid connections AND that it'd be faced with a fire AND that the fire service wouldn't be able to respond AND that diesel couldn't be transported to the site. (8/n)
In a war, however, this sequence of events, which could culminate in a major accident, becomes much more likely. Still unlikely in absolute terms, perhaps--but in relative terms, the unimaginable could become entirely conceivable. (9/n)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
