If you properly maintain something -- a solar thermal power plant, a factory, anything -- it will last forever. You just need to replace each part when it's sufficiently worn down.
If someone builds a solar thermal power plant today and you ask me what that plant's efficiency will be in ten years, I would guess that the same plant would be more efficient, not less, because the company will have invested to not only maintain the plant in good working order, but also to improve its technology as it goes. For example, when the mirrors get scratched and dirty after a few years, the company might replace the mirrors with new mirrors that have a better coating technology than the original mirrors.
There is a big difference here between a utility-scale plant being maintained for long-term performance by professionals, and (say) a solar panel on my roof, which I do not have the time or expertise to maintain. A solar panel on my roof will only get more and more scratched and dirty and degraded over time. So its efficiency will go down and down, in a way that can be predicted and modelled.
It might be the case that it is not worth maintaining a solar thermal power plant in good working order. Maybe it's cheaper to let it get worse and worse, then throw everything out and rebuild it from scratch. I doubt that's the case! But if it is, the decay function is not something a physicist can predict from his armchair. You probably need to ask an engineer in the industry.