In January I posted a piece on coal under Trump, concluding
My prediction is that the pace of closures, and the loss of mining jobs, will roughly triple.
How’s it going? The coal plant closures continue, but it’s far too early to test my prediction. I do however have some new evidence on my side.
I’m not talking about the long-awaited announcement by Trump’s EPA of its replacement for Obama’s Clean Power Plan regulation in the form of the Affordable Clean Energy rule (ACE). The CPP was moribund once SCOTUS loyally suspended it, on the laughable pretext that the Trump Administration would shortly produce a workable alternative form of regulation of greenhouse gas emissions. ACE will be immediately caught up in litigation so it won’t have any effect either: it’s no longer regulations killing coal but economics. For instance, Republican Florida and Texas are both in the middle of a solar boom.
ACE is a stunt to retain the fraying support of Rust Belt voters along the lines of “At least we tried, but we had no answer to AOC’s superhero powers and superior Chinese and Danish technology, backed by the machinations of the Deep State”. Less an ace than a desperate lob from behind the baseline, inviting an easy smash. Fake news.
No, it’s something else.
My argument was based on a regularly updated FERC staff document on energy infrastructure, published here (click on the “Energy infrastructure” tab). The December table, which is what I went on, looked like this:
The latest April version:
The difference is that from January FERC added a less scary column of “high probability” capacity additions. Since the capacity factors of wind and solar are roughly half that of coal, the overhang that was so evident before – and the basis of my prediction – almost disappears. Coal isn’t doing well, but it doesn’t look like collapse. Trump’s favoured miners can breathe a sigh of relief.
Or can they?
1. Christian Roselund’s piece at PV Magazine that pointed out the contradiction between modest announced coal closures and much larger reported plans for new wind and solar was published on January 2. FERC added the cheerier column soon after, not giving their methodology, but citing the same consultants as before. If this had been planned for some time, surely they would have said so. As it is, and given the general untrustworthiness of the Trump Administration, it’s a natural suspicion that the addition is a political Band-Aid.
2. “Not highly probable” does not mean “unlikely”. The sources of the “All additions” column are utilities and big developers, and the consultants assembling the count are also professionals. A fair number of the projects they are looking at will no doubt eventually be cancelled for one reason or another: financing falls through, they can’t get a viable offtake PPA, permitting issues arise. But these are surely real projects not vapourware. A substantial number will still go ahead. The overhang is real.
As it happens, a different and equally professional team has come up with an independent estimate for new US utility solar capacity. This is the SEIA – the main American solar trade association – and Wood Mackenzie, a well-established consulting house who publish the GTM blog among other things. A chart with their estimate:
The total prospective additions here come to 78.8 GW, quite close to FERC’s 84.7 GW, confirming both. But SEIA/Wood Mac find 27.7 GW of solar with signed PPAs. A PPA, power purchase agreement, is a commercial contract for electricity supply over a long period (15 to 25 years), with the usual penalty clauses. Using their looser category, FERC only find 14.8 GW of “highly probable” solar projects. That’s only 53% of the SEIA total, far outside normal measurement error.
The SEIA criterion is more stringent than FERC’s, as well as much less subjective. Consider a developer with a defined and permitted project facing two competing and satisfactory prospective customers ready to sign profitable PPAs. The project does not meet the “signed PPA” test but it’s still a pretty sure thing.
The FERC estimate has a three-year horizon, while the SEIA doesn’t give one at all. But solar farms are routinely put up within two years of the final investment decision, and it’s hard to see why a willing customer (say Facebook) and willing developer (say Engie) would want to delay much longer. This can’t possibly account for much of the difference.
The obvious reading of the discrepancy is this:
- FERC’s new “highly probable” category is politicised rubbish, for solar anyway and probably for wind too.
- The true volume of planned solar and wind capacity in the USA still fits my scenario of an accelerating decline in coal consumption.
Footnote 1: the displacement proposition does not depend on the overall profitability of the solar and wind farms. Once they are built, their negligible operating costs automatically put them at the top of the despatch merit order in any reasonably functional market. Alternatively, if the offtake is through a long-term PPA with a corporate customer, it becomes their priority supply. New renewables displace coal, pretty much kwh for kwh.
Footnote 2: a nice if only marginally relevant photo of a wind technician doing girly work in the North Sea. The leading edges of the huge blades get worn and need regular taping to restore their aerodynamic smoothness. The photo is from the website of a company that recruits strong-nerved workers for this and other jobs in the industry. This is not a staged once-off image.