Friday, January 31, 2014

Review: Spain's Photovoltaic Revolution

Review of Spain's Photovoltaic Revolution - The Energy Return on Investment
Pedro S. Prieto, Charles A.S. Hall
Springer, 2013

No getting around it, I was disappointed.

First, I pay $47 bucks, and I get a 3/8" (1cm) thick x 6" x 9.25" paperback.
(was expecting something larger).

Well, OK, if it has real data, from a real owner of a real PV system, then maybe it's worth it.

Or maybe not.  Maybe part of my sense of disappointment is the let down in my expectations, I was looking for things I didn't know.  If you don't know anything about EROEI, a much better place is in Energy and the Wealth of Nations - Understanding the Biophysical Economy, Hall & Klitgaard, Springer 2012.

Disappointed to have dated information, the last info on world oil and gas production was 2010 in a book with a 2013 copyright. (pg. 2).

pg. 3 - "… where something is expensive in dollars, it will be expensive in energy".  This doesn't follow, look at any art auction, luxury autos, etc.  Or more to the point, home prices in the housing bubble that ended in 2008 with loss of monetary value, in some cases over 50%.  Did the embodied energy in those houses change when their price collapsed?  No.

Was disappointed to see Prieto and Hall ascribing the lack of monuments of hunter-gathers to low EROEI (n.b. they say EROI, which many find confusing: is it money or energy invested?).  Because hunter-gathers typically were transient, they didn't stay around long enough to do any building, nor why would they?  Without a king ruling over them to crack the whip, why would a small band even attempt a big pyramid?  People are lazy and social, way better to hang out with the tribe and talk story than haul big stones around.  Big irrigation and structural works have to wait until there's a king with a big enough population to boss them around.

They seem to buy into the notion of agriculture as the fruit of "progress", having either not read or not been impressed with the evidence that agriculture was not so much "invented" as was forced into practice by people under desperate conditions (climate,  Pleistocene Overkill, overpopulation).

see:
(short book)
Neanderthals, Bandits and Farmers: How Agriculture Really Began - Colin Tudge
(long details)
The Food Crisis in Prehistory: Overpopulation and the Origins of Agriculture - Mark Nathan Cohen

They ascribe (from others) the EROEI of 10:1 for hunter-gatherers, but here we get into the hard problem with EROEI - how to draw the boundaries.  Most hunter-gatherers, under good conditions, only work a few hours a day.  Do we say the hour they work gives them 10 hours of free time?  Or do we say they feed themselves, a young child, and elder and a mate - thus EROEI 4:1 ?  This begs the question: what is the minimal EROEI of a society needs for maintenance or for growth?  While it is clear that hunter-gatherers typically had more energy available than they needed, an actual number for EROEI is now less clear after pondering this book.

I'm puzzled over the reference on pg. 15 to "Al Gore's Kyoto Protocol".  He advocated for it, but the protocol was the work of many, many others, and is done as a UN initiative.  Is there a sub-text of disparaging climate change mitigation and/or environmentalism/renewable energy?

One of the premises of the book is that Spain is a great place to study PV due to the good sun resource, and stable PV generation.  But this is un-clear, because of the way the Spanish government created the boom and bust of PV in Spain.

Chapter 3 is a decent history of that government incentive, and how the market exploded due to overly generous incentives.  One thing missing in this book though, is that Spain's retail electrical rates are (still) partly regulated, and were/are kept artificially low (can you say "bread and circuses"?). But with the oil and other energy price shocks, that government subsidy exploded.  Then with the financial crisis beginning in 2008, the Spanish government was overwhelmed with debt, and had to take action, meaning cut spending, so very few people in Spain are happy.  My point is that not just the PV land rush is to blame for the austerity crisis.

I think Prieto & Hall place a bit too much blame on industry, whereas I see industry just responding to rather goofy government regulations.  First, the feed in tariff offered was far too large, about 15% bigger than Germany's.  This attracted the fraudsters that Prieto & Hall rightly complain about, as well as the inexperienced, both resulting in higher costs.  Second, there was no reservation system to limit the systems to what the government had budgeted.  The 2010 goal was about 400 MW as I recall, but that level was already surpassed in 2007.  In 3/4 of the next year, while the Spanish government was figuring out what to do, another 2,700 MW of PV was installed.  In that respect, they wildly succeeded, over-achieving by nearly 10 times (not percent, times!).  But obviously the government blew their budget.

Third, the Spanish government arbitrarily set a limit of 100 kW per system (in an attempt to encourage small businesses), which businessmen merely circumvented by setting up little shell companies, one per 100 kW system.  This not only made the systems less efficient generators, it added to cost and presented opportunities for fraud.

The only new thing I learned is that due to the system limits and tariff structure, typical systems were overbuilt by 8%, which again increased costs - but those costs/inefficiencies were more than covered by the too-high subsidies.

Now the Spanish government is trying to stave off default, and to pay creditors of the historical electricity subsides as well as make payments on the current feed in tariffs.  The honest thing to do would be to raise electricity rates, but they're already high (Spain has no fossil fuels but a bit a lignite, and limited hydroelectricity), so that is politically unpalatable and would likely further depress the economy.

Which brings us back to the issue raised by Prieto & Hall, the EROEI of PV, which they claim is the major problem with the solar program.

I see several problems with their details, though compared to the wild days of EROEI 50+ imported oil, they are correct.  Society will be poorer, whether they like it or not.  The only question is how much poorer (and how soon).

(1) it's based on dated information, particularly systems costs.  I can see why, that they wanted to look at a couple of years of production from those systems, though it's still disappointing to see (only) old information used.
(2) the Spanish installations are typically overbuilt by 8%, which gives cost and some power, but much of this power is thrown away during peak sun times.  This is an artifact of government policy and not intrinsic to the technology.
(3) the calculation of energy investment based on total national energy use and GDP is rather loose.

They use 5.5 million Euro / MWp system cost, plus 30% O&M (operations and maintenance) over a 25 year plant life (pg. 44).  But those numbers are old.
In Tracking the Sun IV,
http://emp.lbl.gov/sites/all/files/lbnl-6350e.pdf
it says > 100 kWp systems are $1.9 in Germany pg. 20.
O&M costs don't scale down much, so I'll use 30% of the old 5.5 million Euro/MWp = 1.65 million Euro/MWp O&M. US$1.9 is now 1.41 Euro, round this up to 1.45 for a total lifetime system cost of 3.1 million Euro instead of 7.15 million Euro for the old numbers.  So the rough EROEI from page 44 goes from 2.41 to 5.56.
While this is not great news in an absolute sense, EROEI of 5.5 means essentially twice the civilization of 2.4.

When they did their "how much energy out" calc, they down-rated for 8% due to the Spanish overbuilding (pg 49).  Nobody else that I know of did that, and they're not doing it anymore in Spain, so this is somewhat dubious.  Tossing this means their "losses not generally accounted for" go down from 23.5% to 15.5%, and the rough EROEI with current costs goes to 6.15.

The rough calculations were done on the basis of the total economy of Spain and how much energy they use.  They take the energy used by Spain in 2008 in million tons of oil equivalent and divide by the GDP, and end up with 7.16 MJoule/Euro.

Now, is this really true?
Notice that with the updated costs, the EROEI more than doubled, but did the actual embedded energy go down by half?  Certainly it went down, people are getting smarted about racking for example, and don't use so much concrete these days, but did the actual energy invested halve?

Another example of the questionable nature of their money-to-energy assumption - PV system prices in the U.S. are higher, due to lack of scale, more paperwork, and more cost in customer acquisition.  If a U.S. PV system is twice the price of a German system, does it have twice the embedded energy?  Seems bizarre to assume so.  Two neighboring states, one taxes PV systems, the other doesn't, otherwise the system cost is the same.  Is the system in the taxing state have a worse EROEI?

It would have been interesting to compare other forms of energy generation using these metrics, but they are not in evidence.

They admit to the imprecision of the total energy/total GDP assumption, and in fact show that the true range for energy intensity per monetary unit vary widely.  In Table 4.1, they show the results of a study from Carnegie Mellon University that shows a range from 21.4 MJ/$ down to 1.54 MJ/$, with the US average of 9.68 MJ/$.  So where does PV fit in?

So they go through the parts and process of building of a PV plant in some detail in Chapter 6.  While it's a comprehensive list, my problem and disappointment is all the "we assume this, then double that".

I paid $47 for an EROEI analysis based on assumptions?  And then once they get their mostly economic assumptions made, they use some rough factor of the 7.16 MJ/Euro (half for this, twice for that) to get energy.   They end up with an EROEI of 2.45 with their detailed calculations, which corrected for current system prices goes to 5.65.  They provide a sensitivity analysis case that corrects for the Spanish aberrations, resulting in 2.84, with current prices that's 6.55.

There's a bit of an unfortunate gulf between the Energy Payback Time and LifeCycle Analysis studies of modules, which show current EPBTs under a year for modules, but are based on exact measurements of the energy used per machine, per factory, etc., and the final system EPBT.  But there are a few recent studies out there that show whole systems, with increasing detail.

slide show from Fraunhofer refers to 1.5 years EPBT for systems.  Even if one doubled that to include O&M, one gets an EROEI of 8 over 25 years.
See slide 32 more more info:
http://www.ise.fraunhofer.de/de/downloads/pdf-files/aktuelles/photovoltaics-report.pdf

A short article of EPBT:
http://www.bnl.gov/pv/files/pdf/236_PE_Magazine_Fthenakis_2_10_12.pdf

A major nit - page 80 talks of ingots and cells mostly  "… imported from the world microchip technology sector …"  Microchips and silicon PV have about one commonality, they start from polysilicon.  Otherwise they're different industries, and the PV industry, even in 2007, used several times more polysilicon than the chip industry (thus the insane poly prices in those years).

So, I'm disappointed.  I was looking for hard numbers, and I got a ".. we are left with a somewhat ambiguous picture of solar energy in Spain."

They do raise a lot of good questions though, and I agree with their call that we need to do a lot of work on EROEI - for all energy sources.

As for the book, unless you're a complete EROEI geek and won't miss $47 at all, I can't really recommend it.

original 20140131
edit 20140131 US PV system costs vs. EROEI

No comments:

Post a Comment