The social problem with nuclear power is this. If some fool blows up the economy (or if a perverse structure of incentives drive a whole herd of bankers to blow up the economy) we all suffer for a few years, or even decades. But our grandchildren will be ok. If the same dynamic leads to a full reactor meltdown somewhere, then you could have 1% of the US uninhabitable for the next ten millennium (to say nothing of premature deaths and birth defects). That is longer than all of recorded history, for those who want to keep track. There is literally no punishment severe enough, no disincentive sufficiently large, to balance that out. You have to have a pretty high estimation of peoples abilities to not only intend the right thing, but to do the right thing 100% of the time (not 99.999999% of the time), to argue that of course we can handle it. Otherwise I would argue that the risk/reward calculation does not work out for society as a whole.
The direct cost of any such meltdown, were it to happen in the US, far exceeds any utility’s ability to pay. Or their insurance company’s. Or the re-insurance company. Assuming utilities can even get meltdown insurance. And even individual properties and businesses probably wouldn’t get their insurance claims paid out, since most policies have an “act of god” clause that excludes things like this. So we have a situation where the profits go to the owners and investors as long as it all works, and the costs get shoved onto other people – individuals and taxpayers, should things go wrong. Society has the wrong model for something as dangerous as this. For any given corporation (where by definition the directors are immune from liability and managers are unlikely to ever be prosecuted) the math may work out, since they can exclude most of the down side. And for people who want cheap electricity it likewise looks good. But that is only because they can externalizes the entirety of the risk. The rest of us should not accept that offer.
But can’t we build safe reactors? Maybe. Pebble bed designs sound like an interesting approach. But is that a good course with the cost of renewables continuing their decline? Right now there is not enough battery storage to build a reliable all-renewable grid. So something has to fill the gaps. It would be nice to avoid anything that outputs CO2. And so nuclear looks like the logical bet. But the problem in this country is that it takes $15 billion and 20 years to build a reactor. In the last four decades it is simply not cost effective, even just to rebuild the ones we have. Decommissioning a reactor takes a couple of decades and several billion dollars as well, and you can bet that, having profited off them for decades the electricity companies are doing everything they can to push that cost onto the public. Most of the reactors in this country are operating at twice the lifespan for which they were originally built, and are operated by for profit companies under a model whereby the officers of the company can personally profit by pushing past safety limits, but will not personally lose any prior-year bonuses if the whole thing blows up. This is why I just don’t trust our society to run utility-scale reactors. (I worry less about the reactors on military submarines and aircraft carriers; the people operating those are generally more responsible).
With a 20-year time horizon to bring new nuclear power on line, it may be an effective bet to gamble that batteries will be that far along by then. And most utilities appear to be taking this course. Meanwhile there seems to be plenty of natural gas during the transition.