3 ms·
But how do you consider "full system costs" for something that needs people to manage the waste for 1000+ years even after you have closed the plant? We can ev
by Kon5ole 4y ago
But how do you consider "full system costs" for something that needs people to manage the waste for 1000+ years even after you have closed the plant?
We can evidently not even predict inflation correctly from one month to the next so it seems unlikely that we can know what a nuclear plant actually costs.
France has certainly not known, they're currently over 40 billion dollars in debt for their plants (to be paid by taxpayers when they nationalize the operator), and they have over a million cubic meters of waste that needs manpower to manage for longer than France has existed as a nation.
Instead of nuclear plants in 0.1% of the Sahara, I propose solar panels on 10% of existing parking lots. Much easier, much faster, much cheaper, and coming generations will thank us instead of curse us. :)
//crashoverrideCIA ;)
- kragen 4y agoIt's debatable, but at least everyone† agrees that solar panels don't impose major additional costs after initial construction. So I think we can come to less debatable conclusions by noting that the fairly knowable initial construction cost of nuclear power plants is higher than that of utility-scale solar plants, per watt, in most of the world. That way we don't need to argue about how long we have to hire guards for the million cubic meters of slightly contaminated gloves and concrete—even if that cost is zero, nuclear still loses, except in places like Norway. By the way, France has existed as a nation for about 1700–1800 years. Maybe you mean "longer than France has existed as a state." ______ † There are a few coal-industry shills raising alarms about cadmium, but of course the currently popular kinds of solar panel contain no cadmium, and even the ones that did contain cadmium didn't pose a real problem because of the minuscule quantities and their resistance to corrosion.