Unless the land is valuable, you might as well leave old solar farms there. Even when they're 50 years old they'll still output some power, even if half the inverters are dead and the panels are down to 50% efficiency.
Here in Brazil, we have already gotten to the point where the bottleneck to adding new grid-scale solar and wind is often not the land or the panels/turbines, but the amount of power the grid can carry, especially between the Northeast (which has lots of wind potential and plenty of solar potential) and the Southeast/South (where the power consumption is the highest). A miscalculation in this amount (the solar and wind power plants did not react as specified to a power disturbance, so the calculations were too optimistic) was a major factor in the recent country-wide blackout (https://en.wikipedia.org/wiki/2023_Brazil_blackout but that article seems to be missing the later updates from https://www.ons.org.br/Paginas/Noticias/Ocorr%c3%aancia-no-S...).
...What?
> Here in Brazil
Ah. When your country is destroying 8,000 square km/year in Amazon rain forest (and disappearing the indigenous people who live on it), yeah, I guess you would think land is 'cheap.'
Anyway - you're missing the point. The "cost" is lost opportunity. At a certain point, between increased new panel efficiencies and the currently installed panels losing efficiency, the operator is missing out on enough power generation that they can upgrade and make more money in the long run.
The question would seem to be, at what point is the marginal profit from the trickle of electricity from fully-depreciated old panels less than the cost of ripping them out & selling the scrap? Presumably there is such a point, as the old solar panels increasingly approach zero electrical output, if their materials aren't so worthless as to not even be worth the cost of scrapping.
I could totally imagine that most installations aren't allowed to top up a few old solar farms with new panels without also bringing the rest of the farm up to modern standards re: fire and electrical safety.
Also, I could totally imagine that grid connections are based on the manufacturers rated power output, and not the lowered output because the panels are old.
Reasons for repowering would be that newer modules are more efficient for reasons other than age (in the case of the Swiss array, new modules would be over twice as efficient) and that the modules are a fraction of the cost of the array - modules are very cheap now. In a very naive calculation, a module pays for itself in less than a year https://ourworldindata.org/grapher/solar-pv-prices (assumptions: $0.26 / kWh, more than 1000 full-power hours in a year). A new module (again naively, without installation or capital costs) replacing a 20% degraded one would then pay for itself in about five years.
So, I guess we are now in a situation where solar modules are almost free - there are now other problems to be solved and they seem difficult, storage especially...
Wholesale rates for non-dispatchable electricity are under $.02 / kWh.
Not to forget, costs per kWp go doen year after year. At the sake time Wp per panel go up year after year as well. This alone makes repowering much more attractive than just letting old plant running.
Just one thing, depreciation is all about cost accounting and has exactly zero to do with cash flow.
The back of napkin maths I do for household level purchases and investments do not scale to those levels. Nor do they to company level finances.
Following those practices are, obviously, no guarantee for success. Not following them on the other hand is a near guarantee for failure.
The reason I said the other comment was banal is because it amounts to "do what makes sense", which well, yeah.
Otoh I think people are forgetting there's a point where the maintenance gets to be more
Land. Property tax. inspections. That sort of thing...
If you re-power that plant with more efficient panels, your production will sometimes be capped by the grid connection. But it’s fine because you will have significantly higher production in the morning and evening hours.
This can make financial sense.
I hope they put the used panels on the second hand market, such that users with different economics ( home installs, countries with lower labour and land costs) can buy and use them.
Most situations aren't like that though.
If you have limited land, it makes sense to make it more productive, if you need to spend significant amounts to set up a new farm over and above the raw panels, it makes sense to reuse what you already have.
Similar thing to houses, where there is available land to build new ones.