Hacker News Reader

Comment by zizee | Hacker News Reader

"A small modular reactor should last a minimum of 60 years. Probably more, up to 100, frankly, if maintained properly. Wind and solar, after about 20 years you have to replace everything.”

I'm a big fan of the SMR concept, but this line about having to throw everything away for solar after 20 years is just wrong.

View on HN

Can you elaborate? Is there a solar farm in prod right now older than 20 years that you know of? Im all for solar, and honestly didn't consider this angle..

The main reason why solar farms are replaced before 20 years are up is because modern panels are much more efficient than they were 20 years ago. By replacing the panels you can get 2-4x as much power in the same footprint and using the same infrastructure.

Wouldn't it be more cost-effective to put up the new panels somewhere else? Are panel costs such a negligible part of solar farm costs that expending doesn't make sense?

So you don't _have_ to replace everything. But there's a compelling economic argument in upgrading the components in the Solar array as better ones become available.

I guess you don't need permits for upgrading to newer components.

Eventually this pressure will go away though, as panels can only get so efficient. Future generations of panels are unlikely to make such big improvements.

And because thus is an economic pressure/incentive, its really the opposite of what the original quote was suggesting. In 20 years we throw it all away because there has been amazing improvements doesn't gel well with their "this will last 60 years" statement.

There's a panel in Germany that was in use for 36 years: https://www.presse.uni-oldenburg.de/einblicke/54/files/asset...

Efficiency went from 8.55% to 8.2%

The panels are often warrantied for 20-25 years. Most things outlast their warranty period. Inverters are less reliable, warranties are half that for residential stuff but commercial stuff will come with a service contract. Wires and racks can last quite a long time.

You don’t need anything else (e.g., a river to dump waste heat).

the older inverters tended to die after 10 years. But more recent ones are supposedly better, but they don't exist long enough to really produce statistics.

Is it 12 years? 15? 20? no one knows. But since most warranties are 10 years, its safe to assume they easily get there, or it would become too expensive for the manufacturer

If nuclear technology improves in efficiency as much as solar has, we'll want to replace the SMRs also in 20 years. Check out this amazing graph [1] of solar efficiency improvements from 1976 to the present. I wonder which kind of cells are on typical roofs.

1. https://www.nrel.gov/pv/cell-efficiency.html - high resolution at https://www.nrel.gov/pv/assets/pdfs/best-research-cell-effic...

Well, maybe 30 years rather than 20, but they do degrade.

See: https://www.nrel.gov/news/features/2022/aging-gracefully-how...

"A major question in the solar energy industry is exactly how much we should expect solar modules to degrade each year...and when they will eventually degrade so much that they no longer produce adequate power...For modules built today, it is probably 30 years."

I wonder if those guys have ever met the people from the other side of the NREL office who are running the PV Lifetime project. They have commercial, non-research panels in the field that are aging much less than 0.5% per year.

That stat is when they think it'll drop below 80% of original production, not when it needs scrapped.

..same way as the steam generator, turbines, bearings, etc. etc. in a thermal generator need replacing.

Except a quarter of the cost.

That does seem exaggerated, though people throw around 25yr as a standard lifetime for PV[1] (with an approx. degradation of 1% output per year). 20-25yr for a wind turbine also looks believable (pretty good given that's not solid state like the PV).

1: https://energy.mit.edu/news/study-even-short-lived-solar-pan...

PV is now being sold with a 25-year warranty - it'll still have at least 80% capacity after that time.

As the article rightly points out, it often just makes more economic sense to replace them earlier due to improvements in panel technology. There isn't really a technical reason to replace them.

Modern panels, such as the ones I just purchased, are warrantied for 92% original efficiency after 25 years

There are panels in NREL's PV Lifetime Project that are on pace to have 80% rated output after 200 years. https://www.nrel.gov/docs/fy22osti/81172.pdf

PV "lifetime" is overblown fossil industry propaganda. It does not factor into any economic decision.

It's also irrelevant. Cost of maintenance and service lifetime are built into the price/MWh.

Windmill technician IS the largest growing job in the US this year

Reply on news.ycombinator.com