Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Power

Solar Modules Deployed In France In 1992 Still Provide 79.5% of Original Output (pv-magazine.com) 70

French photovoltaics group Hespul tested solar panels installed in 1992, reports PV Magazine: The testing showed that the modules still produce on average 79.5% of their initial power after 31 years of operation. In a previous testing carried out 11 years ago, the panels were found to produce 91.7% of their initial power. "This result exceeds the performance promised by the manufacturers who said the panels would have maintained 80% of their output after 25 years," said Hespul.

The drop in performance is on average 20.5%, or 0.66% per year over 31 years, and 1.11% per year over the last 11 years... Another more recent study carried out by the US Department of Energy's National Renewable Energy Laboratory (NREL) on 1,700 American sites totaling 7.2 GW of power, showed a median degradation of around -0.75%/year. Moveover, another research study focused on 4,300 residential installations in operation in Europe and used different data processing methodologies. Depending on the methods, a median loss of -0.36% to -0.67%/year was obtained.

Thanks to long-time Slashdot reader storkus for sharing the news.
This discussion has been archived. No new comments can be posted.

Solar Modules Deployed In France In 1992 Still Provide 79.5% of Original Output

Comments Filter:
  • by gweihir ( 88907 ) on Saturday June 15, 2024 @12:41PM (#64551617)

    No surprise.

    • Re: (Score:3, Insightful)

      The results from France won't be the same everywhere.

      PV panels degrade faster in high heat, high humidity, and high UV.

      France has none of those.

      • by coopertempleclause ( 7262286 ) on Saturday June 15, 2024 @12:58PM (#64551649)

        Another more recent study carried out by the US Department of Energy's National Renewable Energy Laboratory (NREL) on 1,700 American sites totaling 7.2 GW of power, showed a median degradation of around -0.75%/year

        The difference between mean degradation in the US to France is 0.09% per year.

      • You obviously have never been in France in the summer. It can get very hot and very humid.

        • Just for information, the paper mentions the panels were installed in the Ain district. Assuming main city of Bourg-en-Bresse, the temperature record (from Wikipedia) is: Daily min-max: January 0-6 C , July 14.2-26.6 C / Absolute records: -17.6 to +39.3 C. Rain 80-100 mm every month.

      • by caseih ( 160668 )

        Putting bifacial panels in a vertifcal orientation is something that can help in places with high heat. Vertical panels won't produce as optimally as perfectly inclined panels will, but they do run cooler which can increase the lifespan of a panel and actually produce more power on the hottest days. And in climates like we have in Canada, winter-tme generation is not as bad as it could be thanks to light bouncing off the snow and hitting the back side of the panels.

        It's very interesting to watch this tech

        • ...Vertical panels won't produce as optimally as perfectly inclined panels will, but they do run cooler which can increase the lifespan of a panel and actually produce more power on the hottest days.

          According to the video I recently watched, the vertical panels will actually produce MORE average power over a number of days than sky-facing panels. The explanation was that the extra heat endured by sky-facing arrays increased the resistance and reduced the efficiency more than the extra light increased the power output.

          I think tracking arrays probably outperform bi-facial vertical installations, but they have their own problems. And when you consider the decreased downtime caused by hailstorms, bi-facial

      • > France has none of those

        Parts of France routinely top ~40C/100F. The current record is 46C/115F from just a few years ago. On multiple occasions they had to throttle their nuclear powerplants because it was too hot to run them at full capacity.

        Of course, in any other context you'd be complaining about France being too far north for solar to be viable, because the southernmost part of France is at the same latitude as Boston, MA.

        It's very unlike you to post bullshit. I'm shocked. /s
        =Smidge=

      • France has a fuckton more humidity than California, having lived in both places for 18 and 30 years respectively. Paris has the same odds of rain every month of the year, essentially.

        Meanwhile I can't recall any rain between May and September for so many years. I actually have a roof leak I found in late April, and haven't had it taken care of yet. I know it probably isn't raining any time soon. Most likely not before that last week of September or first week of October, like clockwork.

        My Enlighten daily so

    • It depends (Score:4, Insightful)

      by Roger W Moore ( 538166 ) on Saturday June 15, 2024 @01:15PM (#64551681) Journal
      That depends - solar panels started off a lot less efficient 31 years ago vs. today so it is still not clear how today's panels will age after 31 years. However, it certainly looks promising.
      • That depends - solar panels started off a lot less efficient 31 years ago vs. today so it is still not clear how today's panels will age after 31 years. However, it certainly looks promising.

        I came to say pretty much this - especially given that today's panels have greater power density, and therefore probably suffer from more self-heating and greater temperature rise. But TFA mentions that tests on more recent panels give similar results. Obviously those tests don't have three decades of depth, but the fact that those panels' degradation is well within original projections is at least suggestive of longer-term good performance.

      • by gweihir ( 88907 )

        Sure. But it also means that the accelerated ageing models were somewhat pessimistic back when. And the models used today are based on the older ones.

  • 2nd life (Score:5, Informative)

    by dargaud ( 518470 ) <slashdot2@nOSpaM.gdargaud.net> on Saturday June 15, 2024 @12:44PM (#64551631) Homepage
    And one thing that is little known, is that old (15yo, 20yo...) solar panels that get replaced are being sold for nothing in Africa and used there to provide electricity in areas that never had it before.
  • How efficient were solar panels 30 years ago ?
    • 11-15%-ish. But also you can still buy crappy panels if you want to, they are cheaper.

      • The shittiest ones available when I installed ~2 years ago were about 16% but those were the bottom end no name old tech no warranty Chinese junk. Real panels from a real company were 18% bottom end. Tip top end was 23% under very specific conditions. I got 22.5% under wider range conditions. Price difference between 18 to 22 wasn't super dramatic and my math says was worth doing.

        • How do you measure the efficiency of your panels?
          What measurement do you do to get 22.5%?
          My 16 panels are 400W per panel, made in Chinese, installed 2y7m ago and in summer they peak at 400W per panel (6.4kW total).

          • The official manufacturer's spec sheet.

            Some googling should get you detailed specs of any real company's panel specs. If specs aren't available, move on. There are plenty more options.

      • Those were the top end that NASA had in their space craft. When I wrote my dissertation on the topic around that timeframe, it was 5-8% for commercial panels, only towards the end of the 90s did it go to about 10-11%. Moreover in 1992, these things were expensive, on the order of $20/W without installation, which back then was custom fabrication for the frames/mounts and micro inverters hadnâ(TM)t been invented yet so thick DC cables is what I had to use and they werenâ(TM)t cheap. I built the mic

  • by opakapaka ( 1965658 ) on Saturday June 15, 2024 @01:28PM (#64551697)
    Here is what we have experienced out in the field: Panels last forever as does ground mount racking Roof mount racking will eventually screw up the roof AGM batteries last 10-20 years Wiring lasts forever Inverters last at least 10 years but can get a little weird firmware wise Charge controllers get fried regularly Monitoring equipment gets fried or goes out of date regularly Which is all to say, do not forget that panels are not the only component of solar production. It is way fun though. Said it once, say it again: self install some grid tie solar panels on your house.
    • by dwywit ( 1109409 )

      " Roof mount racking will eventually screw up the roof "

      Do tell. I've had racks on my roof for almost 30 years, haven't noticed any problems. My racks are box or angle aluminium secured to a corrugated steel (zincalume) roof.

      Agree about the batteries - my current set were installed in 2009.

      My previous inverter lasted from 1996 to 2022.

      I had a charge controller remote monitor needing a new faceplate - the membrane over the actual button wore out with me pressing multiple times a day to see what was going on.

      • by AmiMoJo ( 196126 )

        The effect on the roof depends on the type of roof. Particularly if the roof is constructed to take solar PV from the start, it should be no problem at all.

        As for batteries, it depends how deep you cycle them and how hot the environment is (lofts aren't the best place to put them), but newer chemistries are both cheaper and should last a lot longer. Since weight and density are not big issues with stationary batteries, they can be optimized for longevity. Also it's easy to DIY them and swap individual packs

        • The batteries have gotten more complicated though. We have a 350 kWh bank which is 48 battery units, 192 lbs/unit. Each one has a mini-computer onboard. I hope they last a long time.
      • Standing seam roofing is great for panel mounts since it is non-penetrating mounting. The roof we have an install on with standing seam is due for replacement and its issues are due to weather. It sounds like you're describing the "wavy" steel though which I don't have experience with.

        With the asphalt/fiberglass composite shingles, we have seen some nasty water entry over long periods in a harsh environment. What a moldy mess. Technology is getting better. I am about to install some Aire Hug mount points
        • by dwywit ( 1109409 )

          Yes, it's "wavy". GIS for "corrugated zincalume"

          It's fastened to the rafters with self-tapping roofing screws, which come with a plastic or rubber grommet for a water-tight seal, and you *only* fasten it through the peaks, not the troughs. It's designed so that rainfall drains away from the peaks and runs down the troughs, so if it's installed by competent tradies, you don't get leaks.

          The panel racking means you remove the roofing screws and fasten the racks via new screws (with new grommets) into existing

  • by ffkom ( 3519199 ) on Saturday June 15, 2024 @01:35PM (#64551703)
    I would assume they still do, since electronics back in the 1990s were often built to last. But that changed, so an inverter bought today may not see half of that life span... if the "cloud service" it depends on had not already bricked it 2 years after installation...
    • My solar manufacturer can access my system remotely to update firmware, etc, but if they entirely go away, the phone app can directly talk to my system over local WiFi, no cloud required.

      • Which brand do you use? I have Enphase, 28 M215 and 42 IQ8+ .
        Had some D380 and M215 also, but they got swapped for various reasons.

        The original Envoy-R gateway had a local API. I forget for which reason, but I had to upgrade it to an Envoy-S last year at no cost. The S version no longer has a local API, unfortunately. It did at one point, before I got mine, but a forced firmware update did away with it. Same story with the 2 IQ Envoys/Gateways.

        I use Home Assistant to fetch the production data. It now requir

        • Enphase IQ7 from just before the IQ8 became available. I thought about waiting but didn't see anything I cared about in the IQ8 and they said it would cost a little more.

          I'm not accessing the API through my own code. The phone app has a "connect locally function" but now that I think about it, it does require a login so my original post is wrong. I think if cloud went away I'd be unable to login to the phone app and I'd be fucked. Hmmmm... bad.

          I was going to say I'm only interested in the newest data s

          • by madbrain ( 11432 )

            The IQ8 series introduced microgrid - the ability to power your house even when the utility is down. I didn't set up my 42 IQ8+ that way. There were some technical reasons.

            The main one is that my house has so many loads that it's not realistic for the PV to power all the loads at once. You want to have a separate electrical panel for vital loads, and leave the rest unpowered. I have loads on the main panel, and 3 different subpanels, 2 of which are remote and quite far from the main - in on the opposite sid

            • I have power from my batteries when main is out. What do you mean?

              My house has 2 mains. 400 AMP each. Each main has its own controller and set of batteries so I have 2 separate systems I have to switch between in the app that are on different profiles/accounts under my master login. Each has a mix of vital and non vital loads. I have a sub panel for the pool pumps, heater, lights, etc. Last month we generated 2.6MWh and consumed 2.0MWh. Obviously we produce less in winter. It averages out a fair amo

      • So, in 10 years you'll be keeping a 15 year old phone with android 10 and the app installed on it, somewhere in a safe, because the manufacturer retired the app, and you need to access your inverter...

        My parents just got a solar installation, but the guy explaining it needed "the cloud" got a hopefully polite enough earful from me, and then a bunch of links to homeassistant integrations.

        • No, as I explained to the other guy, I realized that even though local access is Wi-Fi, I still need a cloud login to open the app.

          However, if it went that way, then why not? I have remotes for ancient audio equipment. How would an old phone be any different?
          I actually do have an ancient android sitting in a kitchen drawer that's perfectly functional on WiFi even though it can't talk to a modern cell tower anywhere in the world,

    • by dwywit ( 1109409 ) on Saturday June 15, 2024 @08:19PM (#64552293)

      My previous inverter lasted from 1996 to 2022. When it failed, it let out a LOT of magic smoke, but that was all, no fire.

      So I bought the same brand to replace it.
      https://www.latronics.com.au/ [latronics.com.au]

      And here's something surprising in this day and age - they're shutting down manufacturing at the end of this month as the owner wants a change (he's moving into electronics recycling), and they're releasing all their inverter IP to the public domain.

    • by AmiMoJo ( 196126 )

      A lot of early 90s electronics is dying now. Capacitors dry out, and a lot of kit from that era had soldered in batteries with a 10 year lifespan. Some time after they die they tend to start leaking, destroying the PCB in the process.

  • Solar cells lose efficiency cause by captured charge carriers. Would repeated application of current in the reverse direction mitigate against this?
    • Re: (Score:1, Offtopic)

      It depends on the state of the tachyon array and the phase on flow emitters. If you cross the beams, then ... well, just don't cross the beams whatever you do.

    • by Agripa ( 139780 )

      Solar cells lose efficiency cause by captured charge carriers. Would repeated application of current in the reverse direction mitigate against this?

      No, current in the reverse direction could make it worse by creating more trapped charge. To fix it, the silicon junction would need to be annealed at high temperature.

      I think the problem is the same as with bipolar transistors when their base-emitter junction is forced into reverse breakdown and hot carriers damage the junction creating spots where minority carriers can recombine, which reduces hfe at low currents. This is irrelevant for power transistors because at high currents, hfe is already reduced.

  • How does the efficiency of these 31 y/old PVs compare with modern PVs produced today ? Not quite the simple question as there will be different types at different prices.

    What were the reasons for the degradation in performance ?

    What was the overall financial benefit ? Ie cost of production, installation & maintainance compared to buying from the grid. Again not a simple question as prices vary by location, type of PV, etc.

    • And yes I can figure that out but, really, could they not have said so?

      • For that matter, it is not transparently obvious for any who gave up math after 8th grade whether all of the studies used the same compound vs mean computation.

    • by ceoyoyo ( 59147 )

      It's not compounded, at least not for the French ones. 0.66 * 31 = 20.46 = the 20.5% total.

      100% -> 79.5% over 31 years is a loss of 0.74% / year, which is close to the 0.75% / year from the US study.

      100% -> 91.7% over 20 years is 0.43% / year though, so it does seem like the French ones are accelerating.

  • by mspohr ( 589790 ) on Saturday June 15, 2024 @04:13PM (#64551927)

    My first solar array installed in 2014 now 10 years later has no measurable degradation.
    First year 6908 kWh, last year 6858 kWh ... less than 1% degradation

    • First year 6908 kWh, last year 6858 kWh ... less than 1% degradation

      Yes, but the sun is getting brighter as it migrates from Proton-Proton to CNO nucleosynthesis.

      Stellar nucleosynthesis [wikipedia.org]

      • by groebke ( 313135 )

        First year 6908 kWh, last year 6858 kWh ... less than 1% degradation

        Yes, but the sun is getting brighter as it migrates from Proton-Proton to CNO nucleosynthesis.

        Stellar nucleosynthesis [wikipedia.org]

        Uh-huh. In what time frame? I see your snark, now for some context...

        "The energy output of the Sun has not fluctuated by more than perhaps 0.1% to 0.2% in human history" (Source: Taylor, David; Northwestern University June, 2012)

        So, in a ~300,000 years, the sun has not changed luminosity or temperature more that +/- 0.2% of it current value. At most, a 0.000011% change in stellar output from 2014. Are you implying stellar evolution has sped up geometrically?

        The .07% in array power output shift in output is

    • My PV array installed in 2010 produced more in may 2024 than may 2011.
      I posted a more detailed thread called "anecdotal data".

      It appears weather difference more than counters any panel degradation, in my case, on a monthly basis.

      The peak days did show about a 10% reduction, though.

    • Comparing yearly production, or even monthly, is not a precise enough indicator of panel degradation, IMO. I had a production increase over a 13 year period. A deep dive showed that it was due to weather. The peak days showed 10% degradation. But weather more than countered it.

      The differences in winter can be massive, especially in El Nino years. Good luck measuring degradation then.

  • My Renesola Virtus II hybrid panels (10kW) installed 13 years ago are within 5% of their original performance. I don't have daily data from that far to tell exactly, but within the last 5 years for which I have detailed data, there is no measurable drop. I have never even cleaned them!

    • I actually witnessed an increase in production over 13 years for the same month of May, one of the peak production months for me.

      I do have daily production data, as well as per micro inverters data. Comparing peak days only, there was 10% degradation. But monthly, it was more than countered by weather.

  • I truly hate how Firefox for mobile reloaded the page and completely erased my lengthy post when my finger accidentally touched the screen and pulled down. Does Chrome do that too? I have been resisting it, but this is utterly ridiculous.

    Anyway, I compared my 2011 production vs 2024 for the month of May, which is one of the top production months. This is for a 6.58 kW DC system - 28 x 235W US made Sharp panels. The micro-inverters went from 190W to 215W between these 2 dates, but there wasn't any clipping

"I got everybody to pay up front...then I blew up their planet." "Now why didn't I think of that?" -- Post Bros. Comics

Working...