Over the past few years, I seem to have developed something of a specialty in tracking the costs of trying to run an electrical grid using more and more power sourced from the intermittent renewables, wind and solar. I started looking into this subject because it appeared to me to be obviously a huge issue that must be addressed if one is going to attempt to replace fossil fuels with these renewables; but hardly anyone else seemed to be paying attention to this. On the side of the climate crusaders, you just get one unbelievably naive piece after another. For example, here is one of my favorites from the New York Times of February 6, 2018, “Why a Big Utility Is Embracing Wind and Solar,” by Justin Gillis and Hal Harvey, taking the position that a utility “will be able to build and operate the new [wind and solar] plants for less money than it would have to pay just to keep running its old, coal-burning power plants.” So, just knock down the old coal plants and put up some new, cheaper wind turbines and solar panels! Your electricity will come from clean and green sources, and the planet will be saved. What could be easier?
The missing piece is that the wind and sun provide power only intermittently, and cannot on their own keep a grid for millions of people up and running and functioning continuously. They need either some kind of 100% backup which is idle much of the time but ready to go at all times, or alternatively storage for what could be many days — or even a month or more — of power. Does that add a little or a lot to the costs? How would you know? Believe me, Gillis and Harvey — and many others of their ilk — do not and will not address this issue.
For some background on this subject, here are a few of my previous pieces:
How Self-Delusional Can We Be On The Cost Of Electricity From “Renewables,” February 10, 2018.
I return to this subject today because a Minnesota-based think tank called the Center of the American Experiment (CAE) is just out with a big Report titled “Doubling Down on Failure: How a 50 Percent by 2030 Renewable Energy Standard Would Cost Minnesota $80.2 Billion.” Center of the American Experiment is currently headed by John Hinderaker of PowerLine Blog. Hinderaker covers the issuance of the Report in his blog post Tuesday titled “Exposing the Costs of ‘Green’ Energy.”
The guys at CAE have done some detailed modeling and calculations, and back up their work with lots of spreadsheets that are like catnip to math geeks like me. In terms of cost of electricity to consumers, Minnesota starts from a current base of about 10.4 cents per kWh price of electricity to consumers, and an average annual consumer electricity bill of under $1000 per year. Here are CAE’s headline cost conclusions for a scenario of 50% of energy from wind and sun by 2030:
The Renewable scenario would cause electricity prices to increase an average of 4.18 cents per kilowatt hour (kWh), increasing the total retail price of electricity by 40.2 percent relative to November 2018 prices.
[T]he average Minnesota household using 748 kilowatt hours of electricity every month will see its monthly bill increase by $31.24 per month, or $375 per year.
That sounds rather bad. Maybe bad enough to wake up some people in Minnesota. But I’ve got some news for CAE: In my opinion, they have way, way underestimated the cost and price increases that will inevitably accompany an attempt to get generation from intermittent renewables up to 50% of electricity production. A much more reasonable estimate would be that getting to 50% renewable generation by 2030 would lead to price increases for electricity to at least 20 cents per kWh, and more likely 30+ cents per kWh; and an increase in electricity bills from under $1000 per year on average to at least $2000 and more likely $3000 or more.
On what do I base my opinion? Two things: (1) the experience of jurisdictions that have attempted to increase generation from renewables up toward the 50% goal, and (2) analysis of the CAE spreadsheets to see what they are leaving out. Let’s take these one at a time.
Multiple jurisdictions have attempted to increase their percent of electricity generation from renewables through the device of vastly increasing the number and capacity of wind turbines and solar panels connected to the grid. A few of these have managed to push the percent of generation from renewables up to the range of 30-45%, at least for some periods of time. Notable jurisdictions that have done this are Germany, Denmark, and South Australia. Here is a chart that I have reproduced here several times in the past, originally created by Willis Eschenbach of the site WattsUpWithThat, showing the relationship between per capita installed capacity of wind and solar generators and the cost of electricity:
According to RTE, Denmark got 43.4% of its electricity from “renewables” in 2017. According to Reuters, Germany slightly surpassed 40% of electricity from renewables in 2018. They have been rewarded for their virtue with consumer electricity prices in the range of 30 cents per kWh, which is about triple the current Minnesota price.
The current world champion of getting electricity from renewables is the Australian state of South Australia, which is not separately broken out on Eschenbach’s chart. From ABC (that’s Australian Broadcasting Corporation), October 2018:
The state's [South Australia’s] renewable energy generation fell in 2017 to 43.4 per cent from 46.6 per cent the previous year, however it is on track for 73 per cent renewable energy generation by 2020-21.
They’re still not up to 50% (although claiming they will get there shortly). And their reward? This chart gives the latest South Australia consumer electricity price as 43.45 Australian cents per kWh (slightly less for the first 100 kWh per month), which would be about 31 US cents. Once again, we are at about triple the current Minnesota price. The coming push to 50% from renewables will surely push that figure closer to 40 cents.
Analysis of the CAE spreadsheets
So how then have the CAE guys calculated that Minnesota can get to 50% generation from renewables by 2030 for a mere 40% increase in electricity prices, from around 10 cents to 14 cents per kWh, when every place else that tries this finds that the price goes to 30 cents or more?
The key spreadsheets in the CAE Report appear at pages 20 and 24 of the Appendix. Those spreadsheets show some key assumptions — principally that all coal capacity will be shuttered, that all nuclear capacity will be retained, and that wind and solar capacity will be dramatically ramped up. Another key assumption, reflected implicitly in these spreadsheets and stated explicitly elsewhere, is that the wind turbines will achieve 44% production of their capacity over the course of the year, all of which will be utilized to get to the 50%; and that the solar panels will achieve 18% [corrected — was “24%”], all of which will again be utilized.
Here are the two key flaws that I find:
The 44% and 18% utilization factors for wind and solar are way too high. You can actually look at the existing years reflected in the spreadsheets (2016-2017) [corrected — was “(2016-2018)”] and easily calculate that the existing [utility scale] solar facilities achieved a big 13.4% in 2017 [corrected — was “2018”], while the wind facilities achieved close to 40% in 2018, but only 36% in 2017. The authors justify using the higher figures on the grounds that they expect “efficiency” improvements. But this is not a question of “efficiency,” so much as a question that power from wind and sun cannot be generated, no matter how sophisticated the device, when the wind is not blowing and the sun not shining.
There is an implicit assumption that all of the electricity generated from the wind and solar facilities can and will be used. Not even close. For that assumption to be true, the 44% and 18% [corrected — was “24%”] capacity utilization factors would need to be met smoothly and continuously. Instead, production from the wind and sun swings wildly and randomly. Moreover, the nuclear plants cannot be turned off when the wind and sun suddenly come at 100% on a windy summer day; nor is it obvious that other facilities, such as hydro, can be so easily turned off. The result is that vast amounts of the renewable power need to be thrown away or pushed onto the neighboring states — just like Germany often needs not just to give away, but to pay Poland to take power off its hands. And then, of course, the renewable power that gets thrown away can’t count toward your 50% goal, so you need to add more wind and solar facilities to try to get there; but no matter how many of them you buy, you can’t get them to produce any power on a calm night, and you have more and more power to throw away on a windy summer day.
From the spreadsheet on page 24, you can see that in 2030, to get to the 50% of electricity from renewables, they envision 1871 MW capacity from nuclear, 2269 MW from “utility scale solar,” 1200 MW from “community solar,” and 6973 MW from wind. When all of those go at 100%, you will get 12,313 MWH every hour from those sources alone. But average usage in Minnesota is only 6849 MWH in an hour. Almost half of the generation needs to be thrown away. And by the way, there’s another 11,000+ MW of capacity in their hypothetical 2030 system, all of which needs to be completely turned off at these times. To the extent that it can’t be turned off, you will have even more to throw away.
The issues that I have identified ultimately are the reasons why nobody has actually yet succeeded in getting the percent of electricity production from renewables up as high as 50% in a system of intermittent renewables backed up by fossil fuels and nuclear. At some point in the 40s somewhere, you hit a wall where you need to add more and more wind turbines to generate tiny amounts of electricity on nearly calm days, while the amount of power thrown away on windy days explodes.
To correct the CAE work to account for the flaws I have identified and come up with real numbers, you need to do what my hero Roger Andrews has done for some of his calculations: get day by day wind and solar generation figures for the jurisdiction in question from existing facilities, and extrapolate those numbers to find out how much capacity you will need to get to your desired 50%, after taking account of that which must be thrown away. As of now, the CAE guys have only begun the hard work. On the other hand, we kind of know where the end result will come out, from the experience of Germany, Denmark, and South Australia: not a 40% increase in electric bills, but at least a tripling.
UPDATE, March 15: I got a communication from Isaac Orr, the lead author of the CAE Report. As a result, I have gone back and checked my numbers, and made a few corrections as indicated. I have indicated the change in each case. The changes do not affect my ultimate conclusions.