Yesterday, the Green media was getting extremely excited about a new paper, which claimed that renewable energy was going to save society millions and billions and trillions of pound by 2050 (or something like that). Ten trillion pounds by 2050 said the BBC.
As readers here know, I keep a close eye on the cost of renewables, and have published papers on both offshore and onshore wind, showing that the financial accounts of operators in both sectors show no sign of significant cost reductions. It’s not just me either: my findings closely match those of the energy economist, Professor Gordon Hughes, the energy analyst Kathryn Porter, and an important paper in the peer reviewed literature.
So the idea that renewables are going to save us lots of money is, at first sight, pretty implausible. I decided to take a look at the underlying paper, which comes from the Martin School, at Oxford University.
The methodology is, in essence, extremely crude: it involves extrapolating historic cost trends out along an expected curve (Wright’s law, apparently), while jazzing it up a little with what they call a stochastic methodology, which seems to just generate an uncertainty window. The latter details are, for the purposes of this post, largely irrelevant, however – it’s all gazing at tea-leaves in my opinion. What interested me was that they were generating predictions of future cost reductions from historic falling cost trends. As already noted, lots of people find no such cost reductions in windfarm accounts (and in fact, I have some limited data on solar, which tells a similar story).
Where are the Oxford Martin team getting this data from? Buried deep in the supplementary information to the paper, I learned that it came from the International Renewable Energy Agency (IRENA). I have been trying to get to the bottom of IRENA’s claims of cost reductions for some time, and recently worked out what I think is the underlying reason.
It’s all to do with the way currency is handled. IRENA’s work is all demoninated in USD (and as a result, so is the Oxford Martin paper). But the problem of using a single currency is that the final cost figures will be affected by any currency fluctuations. And boy, have there been some big currency fluctuations in the last ten years. In particular, against the dollar, sterling has depreciated by 30%, the Euro by 25%, the Yen by nearly half, and the Brazilian Real by two thirds. When reporting in USD the costs of operators in any of these places, any reduction of less than these values represents an underlying increase in costs. Take the UK for example. According to IRENA, onshore wind costs fell from 0.086c/kWh in 2010 to 0.071c/kWh in 2019 (see figure below).
But during that time, the exchange rate went from roughly 1.60 to 1.28, so in Sterling terms the equivalent figures are 5.3p and 5.5p – a small increase! This is exactly the situation I reported for UK onshore windfarms in my paper at the start of this year (Their absolute figures are much lower than mine though, presumably because they are using different assumptions to me).
This is clearly going to completely undermine IRENA’s overall figures for renewables cost trends. The UK, the Eurozone, Japan and Brazil between them have a very significant proportion of the world’s renewables. The only major centre for renewables that doesn’t seem to have suffered a depreciating exchange rate against the dollar is China, for which, interestingly, IRENA doesn’t detect much of a reduction in wind costs and for which the solar trend seems to have bottomed out (see figure below).
It didn’t have to be like this, of course. IRENA could have created some sort of an index (1990 = 100, or something) which would eliminate the currency effect. But they didn’t, and now the Oxford Martin School guys have picked it up without understanding it, and have extrapolated what is, in essence a foreign exchange fluctuation out to 2050 and have concluded that renewables will save the world.