Pedro Prieto: many solar panels won’t last 25-30 years, EROI may be negative

Preface. Pedro Prieto and Charles Hall wrote the definitive book on the EROI of solar power, “Spain’s Photovoltaic Revolution. The Energy Return on Investment” and has built many commercial facilities himself and witnessed the failure of solar panels long before the supposed 25-30 years they were guaranteed to last.

This is being seen in England where there’s been a loss of 25% of power in the UK due to imperfections known as hot spots on solar panels:

Photovoltaics hot spots are areas of elevated temperature which can affect only part of the solar panel. They are a result of a localised decrease in efficiency and the main cause of accelerated PV ageing, often causing permanent damage to the solar panel’s lifetime performance. Dr. Dhimish discovered that of the 2580 panels he looked at, those that had hot spots generated a power output notably less than those that didn’t. He also discovered that location was a primary contributor in the distribution of hot spots (Solar power – largest study to date discovers 25% power loss across UK October 29, 2018

You may want to read my review of “Spain’s Solar Revolution” for background on what this post discusses, since what Pedro Prieto wrote assumes you’re familiar with the book.

Alice Friedemann  author of “When Trucks Stop Running: Energy and the Future of Transportation”, 2015, Springer and “Crunch! Whole Grain Artisan Chips and Crackers”. Podcasts: Practical Prepping, KunstlerCast 253, KunstlerCast278, Peak Prosperity , XX2 report

Pedro writes:

“Our study concluded that, when analyzed what we called “extended boundaries energy inputs”, about 2/3 of the total energy inputs were other than those of the modules+inverters+metallic infrastructure to tilt and orient the modules.

So even if the cost of solar PV modules (including inverters and metallic infrastructure) were ZERO, our resulting EROI (2.4:1) would increase about just 1/3.

Without including the financial energy inputs (you can easily calculate them if most of the credits/leasing, were requested in contracts at 10 years term with interests of between 2 and 6%, even if you consider as energy input derived from the financial costs, only the interests (returning the capital, in theory would only return, in my opinion, the previous PREEXISTING financial (and therefore, energy) surplus, minus amortization of the principal, if any (when principal is tied to a physical preexisting good, which is not the case, I understand in most of the circulating money of today, but you know much better than me about this).

We also excluded most of the labor energy inputs, to avoid duplications with factors that were included and could eventually have some labor embedded on it. And that was another big bunch of energy input excluded from our analysis.

As I mentioned before, if we added only these two factors that were intentionally excluded, not to open up old wounds and trying to be conservative, plus the fact that we include only a small, well-known portion of the energy inputs required to stabilize the electric networks, if modern renewables had a much higher or even a 100% penetration,  it is more than probable that the solar PV EROI would have resulted in <1:1.

And I do not believe any society can make solar modules even with 25 to 30 years lifetime. There are certainly working modules that have lasted 30 years+ and still work. Usually in well cared and maintained facilities in research labs or factories of the developed world. But this far away from expected results when generalized to a wide or global solar PV installed plant. Dreaming of having them 100 or 500 years is absolutely unthinkable.

Modules have, by definition, to be exposed more than any other thing, to solar rays (to be more efficient). You just look even at stones exposed to sun rays from sunrise to sunset and to wind, rain, moisture, corrosion, dust, animal dung (yes, animal dung, a lot of it from birds or bee or wasp nests on modules) and see how they erode. Now think in sophisticated modules  exposed to hail, with glass getting brittle, with their tedlar, EVA and/or other synthetic components sealing the junctures between glass and metallic frames eroding or degrading with UV rays and breaking the sealed package protecting the cells inside, back panels with connection boxes, subject to vibration with wind forces and disconnecting the joints and finally provoking the burning of the connectors; fans in the inverter housings with their gears or moving parts exhausted or tired, that if not maintained regularly, end failing and perhaps, if in summer, elevating the temperature of the inverter in the housing and provoking the fuse or blown of some vital components, etc.

I have seen many examples of different manufacturers of all types of modules (single/mono, multi/poli, amorphous, thin film high concentration with lenses, titanium dioxide, etc.) in the test chambers, after claims of the promoters to the manufacturers. I have attended to some test fields of auditing companies contracted by promoters, detecting hot spots in internal solderings just from the factory to the customs.

I have seen a whole plant of the so marketed as a promising first US brand specialized in thin film (confidentiality does not permit me to name, as yet) having to return it because it did not comply specs. Now, as I mentioned, I am in contact with a desperate promoter, seeking for more new modules to be paid (the manufacturer is broke and has disappeared) that will last a little bit more than those contracted (not Chinese) about 6 years ago and having failed about 2/7 of the total, without a sensible replacement, because present modules in the market have more nominal output power than those originally contracted for and with different voltage and currents that do not permit unitary replacements in arrays or strings, being forced to a complex and costly manipulation to reconfigure arrays in whole with old modules and creating new arrays with new modules and adapting inverters to the new currents and voltages delivered (Maximum Power Point Tracking or MPPT)

We mentioned many other examples of real life affecting functionality of solar PV systems in our book. The reality, 2 years after the publication of the book, proved us very optimistic. And we have many of the PBAs or circuits or connectors, etc. in our own country. Imagine when you install a solar village in a remote area of Morocco, or Nigeria or Atacama in Northern Chile and the nearest replacement of a single broken power thysristor or IGBT that is stopping a whole inverter and the plant behind (not manufactured in the country) and about 2,000 Km -or more- from the plant and need to pass customs like the one in Santos (Brazil), where tens of thousands of containers are blocked since more than one week (plus the usual 6 to 10 weeks custom procedures) for a fire in a refinery close to the only motorway leaving the Santos port to Sao Paulo.

I even contacted some German University (Saarland) designers of a very simple and superb device, and even they came to Spain to test it in my plant in a common attempt  to commercialize it in a joint venture. The device was a flat sensor kinetic platform of about 30×30 cm., able to measure the number of hits of hail, per square meter, the size and the speed of them.

The reason was double: in one side, it could help to prevent double axis tracking plants to order from the control room of the plant to move the towers to flag position against the prevailing wind and hail fall to avoid breaking of the module glasses. On the other hand, it would be a good device, for instance to fixed plants, to be used as hail measure pattern, a sort of standard accepted device by all interested parties, to help insurance companies and manufacturers to see if the damaged modules were caused by hail below or above manufacturer specifications.

It happened that we had to abandon the project, for lack of interest of both the insurance companies and manufacturers. The first, now have a good alibi, when a promoter raises a claim of its destroyed modules, to state that the hail was below size and speed of the the manufacturer specs and that should be responsibility of the manufacturer. The manufacturer, in its turn, when claimed by the promoter, would also claim that the hail was much higher in size and speed than the specified one. The promoter, with his modules destroyed and a fully fooled face, is so caught in the middle of nowhere, with the hail already melted and the plant destroyed. This is real life, ladies and gentlemen.

100 or 500 years lifetime? ha, ha, ha.”


This entry was posted in Pedro Prieto, Photovoltaic Solar and tagged , , , . Bookmark the permalink.

7 Responses to Pedro Prieto: many solar panels won’t last 25-30 years, EROI may be negative

  1. Sorry, your blog is really good but we have some 30 yrs old PVs, on the Looser in Alt Aussee in Austria, on harsch mountain condition and this from Tirol after 27 yrs still 90% of nominale:

    • energyskeptic says:

      That’s great, but one persons luck is not what matters, it is how tens of thousands of panels perform over time. I’m concerned about the use of solar to make our fossil fueled civilization survive, not individual home use, so how panels do overall is what matters to me. I wish you continued good luck

  2. Hugh Spencer says:

    Living as we do with solar (as our only renewable component) – and in a very trying environment for things electronic and mechanical, I agree with Pedro. Our first batch of 80 watt panels (BP) died after about 8 years – at least they got replaced – we were lucky and just got under the line when BP quit making panels. The current ones have survived about 10 years now. A friend in Cairns had amorphous panels installed – 50% failed within a year. We now have the 250 Watt panels as well – and it will be interesting to see their longevity. Our original panels were mounted on trackers (single axis – quite well made) but these have all failed (electronics). We do tend to live in hope. Unlike most users – we install, maintain and repair our own system – but in this coastal tropical community (which has no grid connection) provision of solar power is a highly divisive issue, and the environment – wet and fungussy) doesn’t help. However in grid-connected Australia the uptake of roof mounted solar is growing rapidly – despite the best efforts of energy suppliers and Federal Gov’t to slow it. It would be interesting to see the longevity of panels (mostly 250 watt) in this situation.
    Hot-spots .. interesting – and having just invested in a thermal camera – I’ll go and look at ours! Also – look at

    • energyskeptic says:

      And it’s not just how long panels last, there are dozens of reasons for why their performance degrades over time, even if they’re still working. It must save you a lot of money to be able to fix problems yourself. What a beautiful place you live in, we were there a few years ago and loved the climate, biodiversity, and so on, but it would drive me nuts to have to stay out of the ocean because of the saltwater crocodiles.

  3. John Matheson says:

    I have been buying 10 year old Sharp PV NEQ* panels manufactured in Japan. These panels were popular in Australia during the first round of government subsidised domestic solar systems. The panels come from people upgrading to larger systems and have a market price of AU$0.25 per watt (about US 18 cents per watt). The panels still meet original specifications for output and mechanically they are as good as new – only the anodising is a little dulled due to weathering. I don’t see why in another 10 or even 20 years time there would be any significant difference in condition or performance.

    • energyskeptic says:

      It’s the overall, at scale, stats that matter. For example, our house burned down in 1991 and the appliances we bought after that over 25 years later, are mostly all working, far longer than what consumer reports says is the average longevity for the particular products and makers we own. But glad your panels still work! Let’s hope the inverters and other associated solar components last 30 years as well…