Vaclav Smil is correct: never forecast

Posted on Updated on

A new piece at “Smart Planet” takes on Vaclav Smil. Rather delightfully it accuses Smil of criticising straw men:

What really aggravates Smil is “exaggerated, unwarranted claims regarding the pace and near-term exploits of new renewable conversions.” He rails at length against aspirational models put forth by transitionistas like Al Gore, Amory Lovins, and T. Boone Pickens, hurling epithets like “grandiose,” “a grand delusion,” and “pathetic prayer-like advertisements.” He beats Matthew Simmons and M. King Hubbert like rented mules. He declaims the application of Moore’s law to solar PV because the growth of PV isn’t about increasing efficiency.

But in so doing, he is essentially assailing straw men and missing the point.

And then the author rather rapidly morphs into one of these suppose straw men that Smil criticizes:

Energy analyst Gregor Macdonald’s analysis of BP data shows that solar power consumption has grown at a compound annual growth rate of 63.2 percent over the past five years. He believes it should be able to maintain that rate as its rapidly falling costs and simplicity of deployment undercut alternatives. If so, solar would provide more than 10 percent of global electricity demand by 2020, surpassing nuclear generation, even while global power consumption grows by 3.4 percent per year.

63.2% growth of solar power for the next seven years! This is not analysis, it is a Ray Kurzweil-esque fantasy. As is well known, exponential growth in the hands of some people is a blunt and inaccurate instrument.

Let’s do the basic sums. Seven years of growth at 63.2% will see total global solar capacity go from 100 GW today to 3,000 GW in 2020. Not only that, but it will require there to be enough global solar manufacturing capacity to produce 1,200 GW of new solar panels each year from 2020. How exactly do we go from 60 GW today to 1,200 GW in 2020 when there is currently an oversupply of solar manufacturing capacity?

And we must ask what we are we going to do with this 1,200 GW of solar manufacturing capacity from 2020 on? Solar power has a fundamental problem: the sun goes down. Adding 1,200 GW of new solar to electricity grids each year will quickly see solar power saturating, making investing in 1,200 GW of manufacturing capacity seem like a bad idea, and quickly. Or will we figure out how to store electricity on a multi gigawatt scale within a decade? Care to predict more exponential growth?

And as a final piece of entertainment. What would happen if we grew solar at a rate of 63.2% not simply up until 2020, but a few years after 2020? Well, you guessed it. We can get to 100% solar in no time at all, around 2025 to be precise:


(note: I’ve extrapolated global electricity production from 2011 (at 22,000 TWh, taken from BP’s Energy Outlook) at a 3.4% growth rate, as the piece used, and started solar of at 100 GW of solar in 2013 increasing it at 63.2% per year assuming a capacity factor of 15%.)

So, Vaclav Smil is correct: never forecast.


3 thoughts on “Vaclav Smil is correct: never forecast

    […] Vaclav Smil is correct: never forecast […]


    stephengn said:
    June 20, 2013 at 9:37 pm

    Wow, you’d better go tell Warren Buffet that he’ll never make anything of himself with all these financial forecasts he’s made over the years. All that financial planning is for the birds


    […] I wrote a month ago about the strange claims from Chris Nelder that solar power could easily reach 10% of global electricity supply by 2020. Here he was: […]


Comments are closed.