Photovoltaic growth: reality versus projections of the International Energy Agency – the 2017 update

A fascinating picture, and blog post, comes my way, ht CR. There are various ways of looking at the same numbers; I've picked solar PV additions on a log scale, but you can also look on a linear scale, or look at total installations; see the post for more.

What we're seeing is that "official" forecasts of solar PV have lagged waay behind actual installations, and have done so with remarkable consistency. Despite repeated failure they have learnt nothing year on year. There's some discussion of just why the forecasts have been so bad, up to and including capture by Evil Fossil Fuel Barons, even though it isn't clear how that would make sense. Greenpeace also don't do a very good job, as the post notes. I tried to trawl back through GP's reports. But I got stuck because the 2005 report has ~70,000 PJ/a total energy baseline for 2000, whereas the 2010 one has 400,000; and that's illogical, captain. Also unpresciently, 2005 lumps solar PV, hydra and wind together; and the 2010 report is lead by pix of shiny mirrors.

The blog post quotes the IEA as pointing out that its reports are not supposed to be forecasts; this is probably about as useful as the IPCC saying the same about its projections. The IEA claims not to take into account new policies or "major new technologies" and that second point gets closest to the problem. Which I take to be not, really, any major new technologies but just steady technological improvements. Wiki has a nice pic showing growth by region; you can see the overall exponential growth continues, but Europe has clearly tailed off.

Although this is somehow news to me - clearly I've been asleep - others have noticed. The linked blog post provides examples, one of which is David Roberts at Vox. And, delightfully, I find myself able once again to disagree with him. He quotes GP saying Everything beyond projections for the next 10 years is simply a political statement from us, indicating what we want to see happen. This also becomes a work plan for us. If we see a renewable energy market isn’t performing as we want it to, we’ll try to jump in with campaigns—against fossil and nuclear fuels and in favor of renewables. And he likes this; because, effectively, he's a campaigner; and campaigners need something to campaign for. And I disagree because I wonder...

What are the consequences of this mis-forecast? Off in the real world, as opposed to scenario-land, solar PV keeps getting cheaper and people keep installing more of it. We can assume this is likely to continue, regardless of who campaigns for what and, probably, by this stage, largely regardless of government policies. Carbon taxes would help it, of course, but carbon taxes (or anything vaguely equivalent) are moving so sloowly that it seems solar PV will likely leap straight past that hurdle. I'm speculating here, of course. So a possible consequence of all this is that CO2 becomes less of a problem than we thought. Could it be that John McCarthy's semi-magical techno-optimism was actually right?


* US carbon emissions
How self driving cars will lead to small, shared, electric vehicles that will save our cities and climate (by Auke Hoekstra)


Phil said...

Also should link to "Sustainability of Human Progress"


JMC wasn't optimistic on solar power or electric cars. He was optimistic about nuclear and hydrogen fueled cars. He wasn't sure if humans could cause global warming, but thought it was a solvable problem if global warming happened.

Happy Heyoka said...

This is really interesting stuff.

One possibility is that most of the actual growth in installations is small scale and therefore off the radar of groups like the IEA?

I know here in Australia, "utility" scale solar is pretty slow to be realised - there are lots of not-quite-megawatt scale pilots - but the bigger installs are slow to happen for many reasons. I suspect the uptake is possibly worse than those bad predictions.

But in my hometown of 30k people, in aggregate, there would _easily_ be several megawatts worth of panels installed domestically.

Power prices here are quite volatile (there are shenanigans going on not unlike California in the 90's) plus the usual political nonsense in response to that.

I suspect one result of that is good number of individuals like myself are effectively going to opt out of the grid - not so much go "off grid" but choose to add storage and use the power locally rather than sell it back... so already overpriced power will become effectively more expensive to those who aren't in a position to do that.

That's not a great outcome socially and doesn't do much to address the need to supply "base load" to industry... but the need to charge my new EV and a payback period of around 6 years for the panels and storage at current energy prices means the spiders in my wallet win.

The rapidly evolving storage market, the fall in "retail" prices for panels - I can see how this is all "off the radar" of utility scale modelling by the likes of the IEA without invoking any conspiracy theories.

Prediction for Australia at least: in the name of "energy security" we'll plonk down a billion or so of tax payer funds for new coal infrastructure that won't even have finished paying for itself when it gets mothballed, the uptake of domestic storage will mutate into a new class of energy retailers and the existing ones will start to freak out and lobby for some regulatory lifeline that will just push folks like me totally off grid.

Tom said...

I used to get paid to forecast the growth of solar and other renewable forms of energy. My employers were happy with me because I was less wrong than other forecasters.

Future adoption of solar is tough to forecast accurately for the same reasons other sensitive materials are difficult. Plant and materials are subject to political considerations, demand is elastic and the technology is 'new' in the sense of reaching new markets. But in 2010 I achieved the remarkable feat of over-estimating the production and sale for the upcoming five year period.

I took comfort by reading old projections of internet adoption, which were much wronger, often by an order of magnitude.

Doing that work is difficult and not particularly well-paid.

William M. Connolley said...

> JMC wasn't optimistic on solar power or electric cars

Forecasting is hard, in detail. His main point, though, was that techno-optimisim was correct. And perhaps he was right.

> most of the actual growth in installations is small scale

I doubt that's true. I suspect the installed-base figures graphed are a mixture of known installations and known sales of panels; the latter doesn't have to track where they're put, only assumes they aren't wasted. If you read the crit of the IEA you'll see there is some crit that they aren't even up to date on current prices: they assume too high. That's a bit odd, because forecasting now shouldn't be too hard; the prices of panels aren't exactly secret.

> paid to forecast the growth of solar and other renewable forms of energy

What did your employers do with the information? Would it matter to them that the solar forecasts were too low?


Never understimate the synergy of subsidy and rent seeking.

OTOH, China has just launched the first commercial electric ship, powered by 240 Tesla-sized lithium batteries:

It's a coal barge

[wc: link this , SVP,- the Safari Blogger balk is still a probelm :]


William M. Connolley said...

When is Cupertino gonna replace that pile o' shite you're using as a web browser? Or are you just not very good at driving it ;-? Anyway: 'ere you go.

Phil said...

JMC's "main point, though, was that techno-optimisim was correct. And perhaps he was right."

If JMC is wrong, at least it is an interesting sort of wrong.

Even if there are techno-solutions to our problems, that doesn't mean that the politics will allow for them.

Steve Bloom said...

Cupertino's devotion to the circular economy now being manifest, maybe never.

I'm really happy to see RE going gangbusters, not just in solar but in (mainly offshore) wind, but sadly our climate emergency has not become self-solving thereby, even if storage proceeds at a similar pace. In particular, our mass social delusion re the carbon cycle feedbacks continues. Note e.g. that the fires of the last few months in CA far outbalance any recent emissions reductions (which were starting to look temporary for other reasons).

Sure, if the climate weren't shifting to a much dryer state that carbon would get sucked back up again within a few decades (assuming the soils didn't get too badly burnt, an unfortunate trend of late). I am not especially pleased that my forecast of West Coast ridging setting back in this winter has been accurate so far. The seasonal long-range forecast just came out and predicts the ridge isn't going anywhere for the rest of the winter. Yay me. (FWIW we'll still get a little rain, as was the case in the three strong drought years leading up to last winter's relative deluge.)

William M. Connolley said...

> the fires of the last few months in CA far outbalance any recent emissions reductions

Do they? I haven't seen any figures.

Steve Bloom said...

Yes, see links below. The southern ones should easily double the northern figures, even if Santa Barbara doesn't burn. Of course the comparison is as much a comment on the patheticness of the (temporary) emissions reductions as anything else.




We're seeing lots of strong signs of landscape type conversion like this, so hopes that the land sink can compensate seem unlikely to be borne out, especially as drying-induced soil emissions may be an even bigger issue..

William M. Connolley said...

> as cars do in those areas in an entire year

That's from 2007, and it isn't making any claims of global significance. Ditto the others. You should have made that clear; changes in *local* emissions are locally but not globally interesting.

Nick Barnes said...

That hilarious chart always reminds me of this hilarious chart. https://en.wikipedia.org/wiki/Itanium#/media/File:Itanium_Sales_Forecasts_edit.png

Steve Bloom said...

"in CA" seemed clear enough.

But if we want to speak about the global situation, recent OCO-2 results finding significant recent CO2 outgassing in the African rainforest even with decent rain (and without the drought, fires and land-clearing variously affecting the other two big ones). El Nino, sure, so it will be very interesting to see the subsequent few years of data, but not at all a good sign.

Steve Bloom said...

Also, you may wish to look up global fire trends. I'd call them significant. They're literally down due to grasslands due to human intervention (mainly fragmentation), but the underlying trend is up due to fires in forests, mainly tropical and boreal. We're now also seeing a sharp tundra fire increase, which boosts the permafrost feedback. Expect more!

Steve Bloom said...

Speaking of permafrost and soils in general: https://twitter.com/BenjaminSulman/status/940313901039677440. Ouch.

Hank Roberts said...

Russell, if you hilight the string:
-- 'ere you go.

and /view/source
you should see what you need to type to make the words function as a link
(you don't need to type: rel="nofollow" -- the blog software sticks that in)

Hank Roberts said...

(in Firefox on the Mac it's hilight (select) the line, then right-click on the selection, and View Selection Source, to see the raw html you need to be typing there.)

Steve Bloom said...

More about California, because there's no such thing as enough:

Dead trees, lots and lots of them. I've seen no estimate for the associated carbon footprint, although unlike the soil contribution it wouldn't be abrupt.

Hank Roberts said...


:,,, The company is no stranger to being accused of providing self-serving estimates. Earlier this year, Greenpeace and Oil Change International published a study that said Exxon's estimates for wind and solar power from 2005 came true nearly two decades early. The study also bashed the company for grossly underestimating how fast electronic vehicles may grow: Exxon reckons just 6 percent of the global car fleet will be powered by batteries by 2040 ..."

John Mashey said...

Humans are notoriously bad at forecasting exponentials (or more realistically, the high-growth part of S-curves) for new technologies. Silicon solar cells started at my old place, Bell Labs, in 1954, following century of research.

The Power EngineeringOf The World said...

JMC had not been upbeat about Solar power or perhaps electric powered automobiles. This individual has been upbeat concerning nuclear and also hydrogen supported automobiles. This individual had not been positive when human beings may cause international warming up, yet considered it absolutely was any solvable difficulty when international warming up took place.

John ONeill said...

'..solar PV keeps getting cheaper and people keep installing more of it.'
Only till its limitations become glaringly obvious. Greece, Italy and Germany have the highest percentage of their generation from solar - around 8%. It's now much cheaper to install than when most of the panels went in, but in Greece and Italy, though they get way more sun than Germany, new installations have slowed to a crawl, as they have in Spain. This despite the fact that in Italy and Spain, consumption of power peaks in July, whereas in the UK, France, and Germany, demand peaks in winter. Household demand is still mostly in the evening, and no plausible battery storage could cover a week's bad weather, never mind a whole winter of low production. Rome gets nearly three times as many sunshine hours in July as in January, but in Munich it's nearly six times as many. German subsidies were supposed to have been reduced for the latest round, but were kept up because of low uptake. Energy charts on the Fraunhofer website show a spike of electricity exports at noon on sunny days, as supply overwhelms demand. Who will they export to if all the neighbours install solar too ?