For the most part to my knowledge it’s the same as maintaining any large, complex piece of infrastructure. As it gets older spare parts get harder to find and have to be replaced with different similar parts requiring new engineering analysis, more and more big components like pipes and tanks get to the point where they need to be wholly replaced, etc…

Design lifespan is the point the designers expected a lot of annoying to replace things to wear out on paper for the cost of maintenance to rise, but now in the present we can inspect things to see how they actually did in practice.

This means that operations gets more expensive and you need to shut down for major work every now and then, but compared to the ever increasing cost of building an entire new plant just replacing the parts that have worn out in order to squeeze an extra fifteen or twenty years is probably going to be more cost effective to a point.

We just need them to hold in long enough for us to get enough renewables and storage capacity on the grid to replace all the fossil sources, at which point we can keep building renewables and replace the most most expensive to maintain nuclear and most fish limiting dams and the like.

Great points! Wind and solar are far easier to scale. Their main issue is land use, but when applied properly (with appropriate environmental impact assessments) that's not a major concern on its own, really it's transiting that power to use centers. Dealing with the individual property rights for a transmission line that doesn't benefit the person under it is and has been enough to kill energy projects.

In my opinion, nuclear's strength lies in its energy density. You could replace a coal or gas plant with a nuclear plant. This is an option being explored by a couple companies because it enabled them to use land no one wants that already has the cooling and transmission connections.

I support nuclear and it was a recurring theme in my environmental policy degree, but I am by no means against wind and solar. I think they are fantastic sources. They each have their trade offs. But we will need to make use of everything in the face of climate change.

One small note, nuclear is expensive, however be cautious when researching cost per Wh produced and look for the time scale. Wind and solar projects are often forecasted to run for just 20 years, they can certainly go longer though. Nuclear runs for 50+ years. Cost comparisons always use the lowest time scale. Nuclear obviously has a very high upfront cost that makes it stupid expensive for a 20 year plan, but over 50 it can reach parody or undercut renewables. Renewables are also done a disservice by these same reports by locking them to the low timescales when their leases are easily extended. But leases are also a large expense so renewal does bump the cost. Things get difficult to forecast with those known-unknows, so it's easier, and more accurate to take the lowest scale and say "this is the cost for 20 years" and let the reader decide if they want to math out the 2.5 multiplier. But then it wouldn't be accurate to the 20 years since renewal costs and...well, you see why we use the lower scale.