Learning and cost curves, and why a dollar spent on coal, oil, gas, or nuclear infrastructure represents an investment that will be stranded, and probably very soon

Excerpt from George Harvey on Clean Technica, June 2019

…we have a tool allowing us to predict not only that the price of PVs, concentrated solar power (CSP), and hydrogen will fall, but, given their past history and with a little careful analysis, the approximate rate at which they may continue to fall. This is equally true for the other new technologies, including wind power, batteries, geothermal power, geothermal heating, electric vehicles, and so on. We can do this because of a law of economics, Wright’s Law, which is popularly known as the “learning curve” or the “experience curve.”

Wright’s Law has been used by Tesla to model cost declines, according to an article that appeared in CleanTechnica in April. I tracked a reference in that article to another in IEEE Spectrum, and from that to a Santa Fe Institute working paper (you can download the pdf HERE) showing Wright’s Law at work in 62 technologies for making goods ranging from computer memory to various chemicals, PVs, beer, wind turbines, televisions, crude oil, and more. It worked for all of them.

Wright’s Law relates the cost of a technology to the number of units that have been produced. As experience is gained with a technology, knowledge and efficiency in producing products with that technology increases. It becomes both better and less expensive.

“Experience Curve Effects,” an article in Wikipediasays, “The Learning Curve model posits that for each doubling of the total quantity of items produced, costs decrease by a fixed proportion.” By observing the declines in cost over increasingly long times, we can get an increasingly accurate understanding of what that proportion is.

My observation of this is that the cost of the technology is only a part of the final cost of a product. This means that eventually, given the cost declines, the cost of technology becomes a minor part of the cost of a product. For example, in the earliest days of automobiles, they were expensive toys for rich people. As more were made, imaginative people saw ways to make them both better and less expensive. The declines in prices continued for three decades, into the depression, when other factors were responsible for the greatest changes.

In some cases, the declines seem simply to have gone on and on. The first computer I worked with had core memory that sold for $1 per bit when it was first released, or so I was told. At that price, a 16-gigabyte thumb drive would have cost $128 billion. But now, I can buy such a drive for the cost of a cheap lunch. And I expect the price will keep falling because technology is still such a large part of the price.

Large reductions that we have now for various types of renewable energy indicate we are not even close to a point where declines calm down yet. A fall of 46% over eight years for CSP is pretty steep. Perhaps we could do careful modeling to find just how costs are likely to decline, but when the curve is as steep as it is, we can make one prediction easily. Wright’s Law says the price will probably continue fall for quite a while.

The news here is that while solar PVs and wind power are now at parity with fossil fuels, their costs can be predicted to continue to fall, each year, for some time to come. And the costs of batteries and CSP will continue to fall, going below those of fossil fuels in the near future. With each year that passes, the costs of renewable energy will fall farther below parity.

At the same time, old technologies, such as coal-burning, nuclear, or gas-burning power plants, will have costs that do not decline, so they will need to be able to charge high prices or be subsidized to keep operating. An exception is fracked gas, produced by using a new technology, but I believe it will not save the day for gas-powered plants, because the cost of gas is not a sufficiently great part of the overall cost of running the plant.

The lesson is that a dollar spent on coal, oil, gas, or nuclear infrastructure represents an investment that will be stranded, and probably very soon.

And that brings me to speculation. I know there will be comments saying I am wrong, but I think it is something to consider.

Though I have not seen any careful investigation of the subject, I believe a case could be made that Wright’s Law could go into reverse, with costs for a technology increasing as it becomes obsolescent and used less often. I am suggesting that there may be cases in which decreasing demand will increase costs to customers. And yes, I am aware of the fact that I am suggesting that the economic theory called the “law of supply and demand” does not always work.

As increasing numbers of EVs are used for transportation, companies selling diesel oil and gasoline will feel hit by competition from electric energy. In a situation ruled by supply and demand, those companies would cut prices to keep their customers. But they cannot do that indefinitely. The oil companies have fixed costs which they will still have to cover, despite their reduced cash flows.

Fossil fuel companies might not be able to recover by aggressively reducing prices in a market where their competition has continually reducing costs, under the influence of Wright’s Law. As competition increases from ever-cheaper electricity taking customers away, oil companies will have to charge higher prices to cover fixed costs. But higher prices will drive away more customers. I am suggesting that the equilibrium set up by adjustments in supply and demand might be impossible to establish in such a case, and that the “law” of supply and demand could prove inoperable in the terminal situation.

In the end, I suspect, the price of oil will go up in the face of decreasing demand. When the next-to-last gas station in a town closes, only one will be left, and it will set its own price. When the last gas station in town closes because it is losing customers, the gasmobile driver will have to drive out of town buy gas. The last customers, possibly poor people who can barely afford to buy any sort of car at all, will have to pay increasingly high prices for fuel.

By the way, another article in CleanTechnica appeared on the same day as those I mentioned in the first paragraph. It was called, “Wall Street Issues ‘Peak Car’ Warning.”