My last post on Tuesday reported on the Soho Forum climate change debate that had taken place the previous day. Debater Andrew Dessler, arguing in favor of rapid reductions in human greenhouse gas emissions by the method of vastly increasing electricity production from wind and solar generators, had heavily relied on the assertion that wind and solar are now the cheapest ways to generate electricity. An important slide in his presentation showed comparative costs of generation from various sources, with wind and solar clearly shown as least expensive. At the bottom of the slide, the acronym “LCOE” was legible.

LCOE stands for Levelized Cost of Electricity. I first encountered this term a couple of years ago, and thought that I should get on top of it to understand its significance. It took me about a half hour to figure out that this metric was completely inapplicable and invalid for purposes of comparing the costs of using dispatchable versus non-dispatchable generators as the predominant sources to power an electrical grid that works. The reasons are not complicated, but do take some minutes of thought if the matter has not previously been explained to you. In Tuesday’s post, I asked as to Dessler’s reliance on this LCOE metric:

[I]s he aware of this [inapplicability of LCOE] and therefore intentionally trying to deceive the audience? Or, alternatively, is he innumerate, and does not understand how this works quantitatively?

Some commenters on the post were quite harsh in their judgments of Dessler. They argued for the inference of intentional deception, on the basis that no one claiming expertise in this field could really be so obtuse as to think LCOE was a valid metric for the purpose for which Dessler was using it.

So today I thought to look at how others go about comparing the costs of generation of electricity from wind and solar versus dispatchable sources like fossil fuels or nuclear. I can’t say that I was surprised to learn that LCOE is everywhere as the metric of choice for the comparison. Moreover, it is almost impossible to find any discussion of why LCOE is completely misleading when comparing the cost of a grid powered predominantly by dispatchable sources to the cost of a grid powered predominantly by intermittent wind and solar sources backed up by storage.

Consider, for example, the International Renewable Energy Agency, going by the acronym IRENA. IRENA is a UN offshoot, launched in 2009 and based in Abu Dhabi, that currently has 168 member countries including all the big ones. IRENA’s mission is to advocate for and promote “renewables” as the way to go for the world’s energy system. Surely, with all the big countries (and most of the small ones) backing its efforts, IRENA’s utterances can be relied upon as definitive.

IRENA puts out annual reports on the costs of renewable power generation. The latest one, titled “Renewable Power Generation Costs in 2021,” just came out in July. Here is the press release, dated July 13, 2022. Excerpt from the press release:

New IRENA report shows almost two-thirds of renewable power added in 2021 had lower costs than the cheapest coal-fired options in G20 countries. . . . IRENA’s new report confirms the critical role that cost-competitive renewables play in addressing today’s energy and climate emergencies by accelerating the transition in line with the 1.5°C warming limit and the Paris Agreement goals. . . . “Renewables are by far the cheapest form of power today,” Francesco La Camera, Director-General of IRENA said. “2022 is a stark example of just how economically viable new renewable power generation has become.”

Amid the excited claims that renewables are “by far the cheapest” sources of power, the term LCOE does not appear anywhere in the press release. To find that that is the metric being used to make these “by far the cheapest” claims, you need to go to the main Report. Excerpt:

The global weighted average levelised cost of electricity (LCOE) of new utility-scale solar PV projects commissioned in 2021 fell by 13% year-on-year, from USD 0.055/kWh to USD 0.048/kWh. . . . The global weighted average LCOE of new onshore wind projects added in 2021 fell by 15%, year-on-year, from USD 0.039/kilowatt hour (kWh) in 2020 to USD 0.033/kWh.

Here is the featured chart, showing that costs of power from solar PV cells have now fallen well below the costs of power from natural gas:

You can see right there that here in 2022 power from natural gas is at least three times as expensive as power from solar PV cells. But the title of the chart gives away that the metric for comparison is LCOE.

Look around for others making cost comparisons of ways to produce electricity, and you will find more and more of same. From Bloomberg, June 30, 2022, “Renewable Power Costs Rise, Just Not as Much as Fossil Fuels”:

The costs for renewable plants plunged for a decade as production of solar and wind equipment surged and technologies improved, but the supply-chain chaos triggered by the pandemic ended those steady declines last year, according to BNEF’s biannual survey of the levelized cost of energy. . . . New onshore wind now costs about $46 per megawatt-hour, while large-scale solar plants cost $45 per megawatt-hour. In comparison, new coal-fired plants cost $74 per MWh, while gas plants are $81 per MWh.

From the Guardian, June 23, 2021 (citing last year’s report from IRENA — also based on LCOE):

Almost two-thirds of wind and solar projects built globally last year will be able to generate cheaper electricity than even the world’s cheapest new coal plants, according to a report from the International Renewable Energy Agency (Irena). . . . Francesco La Camera, Irena’s director general, said . . . ““Today renewables are the cheapest source of power.”

So it’s not just Dessler. Some big international agency of “experts” adopts LCOE for making these cost comparisons, and everybody just nods along without ever putting in the 30 or so minutes of critical thinking that would be needed to figure out that this is completely wrong.

To reiterate points previously made, the LCOE metric assumes that wind and solar generators are essentially the same kind of thing as dispatchable fossil fuel-powered generation plants. Just build about the same amount of nameplate capacity, and everything will work out just fine. But in fact a predominantly wind/solar system requires vastly more infrastructure to make a fully-functioning reliable grid: some combination of a 4x or 5x overbuild of generators, vastly more transmission lines, and 20 or 30 days of battery storage. These elements could easily multiply the cost of electricity to the consumer by a factor of 5 or 10 or more. Nobody knows, because there is no functioning demonstration project from which reasonably precise costs can be extrapolated. And frankly, there never will be such a demonstration project, because the costs are so enormous that it can never be done. Meanwhile, everyone just nods along as if LCOE comparisons are meaningful.