SolarCity Pushing Industry To 40% Increase In Useful Lifetime of Solar Power Installations (electrek.co)

An anonymous reader writes: SolarCity released a new report that says solar power systems have a usable lifetime of at least 35 years, which is 40% longer than what the market expects. Electrek reports: "The key finding of the report is that power degradation (annual efficiency loss) of solar panels supplied to SolarCity is as much as 35% lower than for a comparable industry-wide selection of non-SolarCity panels, which are typically expected to last for 25 years. In the study here, SolarCity looked at greater than 11,000 panels to determine their data points and come to their conclusion that their solar panels are performing well beyond expected industry standards. Today, standard efficiency solar panels put out by Tier 1 suppliers are generally warranted to lose no more than 0.7% efficiency per year for the first 25 years -- this is the Power Production Warranty. The key finding in this study is that the annual 0.7% efficiency loss is too high an estimation -- and that the number ought be closer to 0.5%. While it might seem a small number -- a difference of 0.2% -- when applied over a multiple decades timeframe, it means that instead of the standard twenty five year assumed productive life, we can expect at least another ten years of production above 80% of the original system output. Large installers like SolarCity, able to do this type of wide-scale research -- and to also demand higher quality, are showing their ability to pull the manufacturers of the world upward. With SolarCity building their own solar panel Gigafactory we ought expect the quality levels to be even greater in the near future.

Comments

Actual repor link:

[Ecuador]

I get a 404 error on the first link. The actual report link is at : http://www.solarcity.com/sites/default/files/reports/SolarCity%20Photovoltaic%20Modules%20with%2035%20year%20Useful%20Life.pdf
As for the degradation of panels, I have a 10kW system made of relatively inexpensive Chinese Renesola Virtus II hybrid panels for over 3 years now and I have seen no measurable degradation in performance so far. And I even look at the peak days of month to avoid the issue of daily weather and still it seems the peak has not reduced (rather increased by 1% which should be within the margin of error) in these three years. So there goes two myths my installer told me "you need to pay me to service and wash them every year to maintain peak output" and "output will measurably drop yearly anyway though". No I haven't washed them. So, I don't know if they will continue this trend or will suddenly drop in efficiency, but at least for the first few years they seem stable.

Re:Actual repor link:

[Martin Blank]

Panels do better than asphalt roofs because they're tested to higher levels. Asphalt shingles get various ratings based on tests using various size balls of ice to see how they take it. Solar panels are generally tested using one-inch steel balls impacting perpendicular to the panel, usually at or near terminal velocity. The ice used in asphalt testing can break, absorbing some of the energy, but the steel ball gets no such benefit.

Extreme hail can still damage a solar installation, and hail might be able to damage non-panel components, but if your panels are suffering damage from hail, you stand a good chance of finding holes punched in the asphalt and wood the rest of the roof is made of.

Math

[Okian Warrior]

It's even simpler than that. How can any company, that has only existed for 9 years, have any proof their product lasts 35 years let alone even 10?

By using math.

Re:Math

[SydShamino]

One 35-year old working unit isn't proof, it's an anecdote. For proof you need statistics. And those are... math!?!?!

Interestingly enough, you might not even need one unit to have reached 35 years old for the statistics to yield that age. For example, do you not believe that the half life of Thorium 229 is 7340 years? We certainly haven't observed a sample that long, yet we understand how the decay process works and can thus use statistics to extrapolate a relevant average decay period. Similarly, if we humans understand the aging mechanisms of solar panels well enough, including knowledge that there are no threshold events that cause the decay to change to a newer, faster model, then it's reasonable to look at the start of the curve, fit it to the known methods, and advance it to extrapolate a useful life.

Data vs andecdotes

[sjbe]

So one witness in court claiming he saw the murderer is not proof but an anecdote?

Unless the testimony of that witness can be corroborated in some way then that is correct. By itself it is rarely sufficient because shockingly, people lie and are routinely wrong even when well intentioned. When it comes to science and engineering, eyewitness testimony is close to worthless and anecdotal evidence is generally considered dubious for good reason.

You americans are really quite dumb. Or your school system ... or what ever.

Really? You don't understand the difference between data and anecdotes and we are the dumb ones? Curious bit of logic you have there...

BTW: solar panels, as in photovoltaics, exist since over 50 years. We have a pretty good idea how long they last and degrade.

You do realize the panels we produced 20+ years ago are not identical to the panels we produce today, right? That's like claiming an 8086 chip from Intel should have the same failure modalities and rates as a modern Pentium chip. Your point is correct that we have a fairly good idea how they degrade but the reasons why are a lot more complicated than the fact that solar panels have been made for a long time.