A couple of commenters on yesterday’s post raised interesting issues that I thought called for another post on the same subject.
Commenter Arthur proposes to solve the battery expense problem by having the 100% renewable system be in effect only from mid-March through July, which are the peak months for renewable generation in California as shown on the charts in yesterday’s post. Arthur concludes, “Battery expense solved!”, and presents a formula suggesting that mid-March to July is 37.5% of the year, so emissions would be reduced 37.5% without the battery expense.
Where to start? Looking at the chart, you can see that in the mid-March to July period in California, demand ranges between about 20,000 MW at the lowest point, and up to about 39,000 MW at the highest, while the supply from the hypothesized 100% wind/solar system ranges between a low of only 17,000 MW on July 31 and a high of about 48,000 MW on several days in May and June. There are plenty of times when supply is less than demand. And this is only giving you supply and demand on a daily basis, which is sufficient for Andrews’s purpose. But demand must match supply on a second-by-second basis, and without doubt would fail to do so most every evening, when the sun goes down and people are home running their air conditioners and using their TVs and computers. Without fossil fuel backup, what keeps the power running during those times? It still must be batteries. A good rough estimate would be that it would take about one-third as much battery capacity to keep this system up and running for the four-and-a-half month period than the full year proposal modeled by Andrews.
The right way of looking at it is that as the percent of generation desired from renewables increases, the need for battery capacity increases at an accelerating rate. But if we got the need for batteries down to about one-third, that would mean that instead of multiplying the cost of electricity by a factor of about 15 - 20, instead it would be a factor of about 5 - 7.
If we could get the need for batteries down to one-third that of a full-year system, would that be a worthwhile thing to do? Arthur suggests that his mid-March to July proposal would lead to a 37.5% reduction in fossil fuel emissions, which “would certainly satisfy the demands of even the most grim of GW predictions.” Unfortunately, these proposals only address emissions from the electricity-producing sector, which are only about 40% of all emissions. (Emissions from things like heating and transportation would be almost entirely unaffected.) So the overall reduction in emissions would be only .4 x .375 = 15%. Sadly, Arthur, I am sure that no climate crusader will be satisfied with that. And your electricity bill will still go up by a factor of at least 5.
Then there are other considerations I did not mention that undoubtedly will make the cost increases still larger. Commenters Denis Rushworth and Kevin Kilty mention two of these: (1) the need to replace the batteries from time to time, and (2) the need in the future world to produce the batteries with much more expensive energy from the 100% renewable system. And here’s a third: the system that Andrews models requires that batteries store large amounts of charge for periods of time up to 9 months. Batteries don’t store charge for that kind of time period without losing some substantial part of it. Technical data from this post at Battery University suggest that a fully-charged lithium ion battery stored at room temperature will only retain about 80% of charge after a year. My own experience with computers is that charge dissipates far faster than that, but maybe the 80% can be achieved with ideal conditions. Anyway, that would add another 25% or so to Andrews’s cost increases.
The main point is that even Andrews’s model considers an idealized system where many future engineering problems have not yet been addressed. As each one gets addressed, the costs will only go up further. I just wonder how far anyone will get down this road before the ridiculousness becomes obvious to everyone.