4 ms·

It’s always seemed so obvious to me: electronics depreciate at a rapid rate, solar panels are electronics, do not finance electronics over 20 years and think it

It’s always seemed so obvious to me: electronics depreciate at a rapid rate, solar panels are electronics, do not finance electronics over 20 years and think it’s going to work. For a 20 year lease, do you really want to be using a 19 year old system stuck on your roof ? Big clunky, weathered with spiderwebs. What does that look like?

Agathos 8y ago
I'm very excited to be making the last payment on my 500 MHz Athlon this year!
throwaway5752 8y ago
Solar panels are not electronics. Solar panels usually are warrantied to at least 80% of nameplace capacity out to 25 years. And there are hard caps on 1) PV efficiency (Shockley–Queisser) 2) insolation that are non-negotiable physics. If that system is clean and reasonably maintained, it will be 80% as good as they day you bought it (and probably 80% as good as a brand new system) in a quarter of a century.
- jpeg_hero 8y ago
  True, you’ll still have your 80% but that will be compared to the state of the art at the time as a trade off. The production of contemporary panels should (hopefully) be much much better. Imagine the advances in the next 2 decades: higher efficiency, lighter weight, more module, standards we can’t conceive of. Like the other poster in this thread said: sure your Athlon 500mhz is still runninnfine, even at 80% efficiency
  - cf498 8y ago
    So you also assume an exponential increase in solar panel efficiency ? Or in your energy consumption? Its a rather simple calculation, where do you think the cost of energy will be in 20 years and how will the amount of energy you consume develop in that time.
  - throwaway5752 8y ago
    I'm a bit disappointed in your response. Did you look up Shockley–Queisser (https://en.wikipedia.org/wiki/Shockley%E2%80%93Queisser_limit https://en.wikipedia.org/wiki/Shockley%E2%80%93Queisser_limi...)? The maximum theoretical single p-n junction efficiency is low 30%s. Sunpower is the best consumer grade product and they are in the low 20%s right now (https://us.sunpower.com/products/solar-panels/ https://us.sunpower.com/products/solar-panels/). There is not that room for improvement. There are higher efficiency pv cells, but they are stacked layers with different bandgaps and/or involve concentrators. And the hard cap is insolation (the amount of sunlight/area) and I am personally very pessimistic that solar tracking will make it in a reliable way to consumer/residential application. Your Athlon example only works because of where we were in the transistor density s-curve (aka "Moore's law" when it was in the earlier phase and looked exponential). Solar is a lot closer to the analogy at 5 nm scale where quantum tunneling becomes a concern and incremental progress is much slower and more expensive.
    - caf 8y ago
      Residential-scale solar tracking doesn't really look that different from electric garage door openers, does it? Those seem to work pretty reliably for years with ~daily cycling.
      - throwaway5752 8y ago
        It's tougher than that. They are outdoors and have to stand up to weathering and storms. In the case of dual axis trackers they have an additional degree of freedom. Finally, an the toughest, in my opinion, is how little net energy budget they have to work with to be economical: they have to require less energy to build, install, and operate than the incremental solar energy they generate over their operation life. They give an approximately 40% production improvement (that's roughly the single/dual axis midpoint. towards the high side), so if you say a 10 yr life with zero maintenance and controlling a single 250w panel you're talking 9 mWh, or $1080 retail electricity value generated. A 40% improvement on that is only $440 worth of energy budget per tracker over a 10 year life. That glosses over a ton of complexity like if area is amenable to ground vs rooftop, if it's higher latitude, high density panels, stationary panel/tracking mirror, and most importantly having multiple panels per tracker. Zero maintenance/replacement is not realistic, either, since it would cycling dozens of times a day and could be in hurricane/tornado areas. But given a lower end garage door opener is $150 before installation costs, it gives an idea how tough that would in that price range. Consider that panels are generally engineered to withstand 100+ mph gusts, and this would be a relatively finely tuned machine attached to one or more 18 square foot kites that would be expected to withstand the same conditions. https://news.energysage.com/solar-trackers-everything-need-know/ https://news.energysage.com/solar-trackers-everything-need-k... was a pretty good article I found on it.
randombit 8y ago
Solar panels don't depreciate in the same way as most silicon. 20 years later, panels will still be producing power just fine, albeit with some degradation. Source: my house has 23 year old solar panels, my garage has 10 year old panels, they all still produce plenty of electricity. There are plenty of reasons to avoid dealing with these companies, but this isn't one of them.
- antisthenes 8y ago
  Out of curiosity, how much do your 23 year old solar panels produce compared to when they were new? 90%? 85%?

Agathos 8y ago

I'm very excited to be making the last payment on my 500 MHz Athlon this year!

throwaway5752 8y ago

Solar panels are not electronics. Solar panels usually are warrantied to at least 80% of nameplace capacity out to 25 years. And there are hard caps on 1) PV efficiency (Shockley–Queisser) 2) insolation that are non-negotiable physics. If that system is clean and reasonably maintained, it will be 80% as good as they day you bought it (and probably 80% as good as a brand new system) in a quarter of a century.

jpeg_hero 8y ago
True, you’ll still have your 80% but that will be compared to the state of the art at the time as a trade off. The production of contemporary panels should (hopefully) be much much better. Imagine the advances in the next 2 decades: higher efficiency, lighter weight, more module, standards we can’t conceive of. Like the other poster in this thread said: sure your Athlon 500mhz is still runninnfine, even at 80% efficiency
- cf498 8y ago
  So you also assume an exponential increase in solar panel efficiency ? Or in your energy consumption? Its a rather simple calculation, where do you think the cost of energy will be in 20 years and how will the amount of energy you consume develop in that time.
- throwaway5752 8y ago
  I'm a bit disappointed in your response. Did you look up Shockley–Queisser (https://en.wikipedia.org/wiki/Shockley%E2%80%93Queisser_limit https://en.wikipedia.org/wiki/Shockley%E2%80%93Queisser_limi...)? The maximum theoretical single p-n junction efficiency is low 30%s. Sunpower is the best consumer grade product and they are in the low 20%s right now (https://us.sunpower.com/products/solar-panels/ https://us.sunpower.com/products/solar-panels/). There is not that room for improvement. There are higher efficiency pv cells, but they are stacked layers with different bandgaps and/or involve concentrators. And the hard cap is insolation (the amount of sunlight/area) and I am personally very pessimistic that solar tracking will make it in a reliable way to consumer/residential application. Your Athlon example only works because of where we were in the transistor density s-curve (aka "Moore's law" when it was in the earlier phase and looked exponential). Solar is a lot closer to the analogy at 5 nm scale where quantum tunneling becomes a concern and incremental progress is much slower and more expensive.
  - caf 8y ago
    Residential-scale solar tracking doesn't really look that different from electric garage door openers, does it? Those seem to work pretty reliably for years with ~daily cycling.
    - throwaway5752 8y ago
      It's tougher than that. They are outdoors and have to stand up to weathering and storms. In the case of dual axis trackers they have an additional degree of freedom. Finally, an the toughest, in my opinion, is how little net energy budget they have to work with to be economical: they have to require less energy to build, install, and operate than the incremental solar energy they generate over their operation life. They give an approximately 40% production improvement (that's roughly the single/dual axis midpoint. towards the high side), so if you say a 10 yr life with zero maintenance and controlling a single 250w panel you're talking 9 mWh, or $1080 retail electricity value generated. A 40% improvement on that is only $440 worth of energy budget per tracker over a 10 year life. That glosses over a ton of complexity like if area is amenable to ground vs rooftop, if it's higher latitude, high density panels, stationary panel/tracking mirror, and most importantly having multiple panels per tracker. Zero maintenance/replacement is not realistic, either, since it would cycling dozens of times a day and could be in hurricane/tornado areas. But given a lower end garage door opener is $150 before installation costs, it gives an idea how tough that would in that price range. Consider that panels are generally engineered to withstand 100+ mph gusts, and this would be a relatively finely tuned machine attached to one or more 18 square foot kites that would be expected to withstand the same conditions. https://news.energysage.com/solar-trackers-everything-need-know/ https://news.energysage.com/solar-trackers-everything-need-k... was a pretty good article I found on it.

randombit 8y ago

Solar panels don't depreciate in the same way as most silicon. 20 years later, panels will still be producing power just fine, albeit with some degradation. Source: my house has 23 year old solar panels, my garage has 10 year old panels, they all still produce plenty of electricity. There are plenty of reasons to avoid dealing with these companies, but this isn't one of them.

antisthenes 8y ago
Out of curiosity, how much do your 23 year old solar panels produce compared to when they were new? 90%? 85%?