Calculating The Full Costs Of Electrifying Everything Using Only Wind, Solar And Batteries

For several years now, advocates of “decarbonizing” our energy system, along with promoters of wind and solar energy, have claimed that the cost of electricity from the wind and sun was dropping rapidly and either already was, or soon would be, less than the cost of generating the same electricity from fossil fuels. These claims are generally based on a metric called the “Levelized Cost of Energy,” which is designed to seem sophisticated to the uninitiated, but in the real world is completely misleading because it omits the largest costs of a system where most generation comes from intermittent sources. The large omitted costs are those for storage (batteries) and transmission. But as we now careen recklessly down the road to zero emissions, how much will these omitted costs really amount to?

A guy named Ken Gregory has recently (December 20, 2021, updated January 10, 2022) come out with a Report at a Canadian website called Friends of Science with the title “The Cost of Net Zero Electrification of the U.S.A.” A somewhat abbreviated version of Gregory’s Report has also appeared at Watts Up With That here. Gregory provides a tentative number for the additional storage costs that could be necessary for full electrification of the United States system, with all current fossil fuel generation replaced by wind and solar. That number is $433 trillion. Since the current U.S. annual GDP is about $21 trillion, you will recognize that the $433 trillion represents more than 20 times full U.S. annual GDP. In the post I will give some reasons why Gregory may even be underestimating what the cost would ultimately prove to be.

First, some background. A huge and, I would submit, obvious engineering issue that permeates the question of powering an electrical grid with only intermittent sources is how to assure that there is always sufficient electricity available to meet demand at every minute throughout the year. Somebody needs to carefully study actual generation from wind and sun hour by hour (maybe even minute by minute) throughout a year, and then calculate exactly how much storage it’s going to take to get through all the sun and wind droughts that occur during what could be long periods of calm, overcast, hot, cold and nights. This kind of work does not involve sophisticated mathematics, but it does involve detailed effort to create an appropriate spreadsheet and find and carefully input reams of publicly available data. To their incredible shame, none of the genius planet saviors who are imposing decarbonization on us ever undertake this task, and for that matter they have never even admitted it is an issue, as far as I have been able to determine. Take a look, for example, at New York’s Climate Action Scoping Plan, or California’s AB32 Climate Change Scoping Plan, or for that matter any of the regulations and Executive Orders coming out of the federal government, and you just never find any mention of this subject.

And so it has fallen almost entirely to a small number of amateurs to be the ones to address what is in fact the critical, multi-tens-of-trillions-of-dollars issue of “decarbonization.” On November 22, 2018 I discovered on a website called Energy Matters a detailed calculation by a guy named Roger Andrews, looking at two cases (California and Germany) and considering supply from wind and solar and demand on an hour-by-hour basis to figure out how much storage it would take to get through a year using only wind and solar. Andrews, a retired guy working by himself at his home in Mexico, had taken the time to do the critical detailed work that none of the world saviors could be bothered with. I first reported on his results in a post on November 29, 2018. The conclusion of Andrews’s work was that, for those two cases, going to an electricity system of all wind, solar and batteries could drive up the cost of electricity by a factor of approximately 14 to 22.

Depending on assumptions used, you might conclude that those numbers were too high, or maybe too low; but the great thing about Andrews’s work was that he laid out his methodology sufficiently that you could tweak different assumptions and see how that affected the result. But it also was very clear that no amount of tweaking of assumptions would avoid the key driver of the vast costs, which was shown to be that generation from wind and solar is seasonal — basically, for wind much higher in the Spring and Fall than in the Summer and Winter, and for solar much higher in Summer than Winter — and therefore a storage system to get through a year would need to store vast amounts of power for months on end. For Andrews’s cases of Germany and California, the answer both times was that there would need to be some 30 full days worth of power consumption stored for six months and more.

And then, in February 2019, Andrews died. I posted this “Eulogy For Roger Andrews” on February 5, 2019. Nobody immediately stepped up to take Andrews’s place.

But in the last few days I have found that two different guys have now stepped into the breach left by Andrews, one of them being a much more direct successor than the other. The first of the two, but the one who is the less direct successor to Andrews, is Thomas Tanton. Tanton is a professional energy consultant, with a firm called T2 and Associates. In October 2020 Tanton produced a detailed report for something called the Energy & Environment Legal Institute with the title “Cost of Electrification:
A State-by-State Analysis and Results.”
I am somewhat embarrassed that I had not come across this Report previously, because I have represented E&E Legal in some matters (although by the time of this Report they had changed leadership and I was no longer representing them).

Tanton’s Report has a somewhat different focus than did Andrews’s work. Where Andrews sought to calculate the cost of using just wind, solar and batteries to supply the currently existing electricity demand for his chosen jurisdictions (Germany and California), Tanton’s main focus is on the incremental costs of electrifying all the currently non-electrified parts of the economy in his chosen jurisdictions (which consist of all of the 50 U.S. states). Currently non-electrified parts of the economy include things like transportation, home heating and cooking, agriculture, and so forth. But Tanton also analyzes, or appears to analyze, the incremental additions to supply as being all wind, solar and batteries, as well as the replacement of existing fossil fuel generation.

Tanton comes up with a total cost for the entire U.S. of his full electrification model of some $18 trillion to $29 trillion, before even getting to additional major things like what to do with stranded assets, which would bring the total to more like $36 trillion. From the Executive Summary:

Electrifying the entire nation, with a goal of eliminating the direct consumption of fuel would cost between $18 trillion and $29 trillion in first costs. In addition, constructing and implementing an “all-electric” nation will require consideration of two other significant costs: stranded assets and deadweight losses .

Those costs — essentially, well more than a full year’s GDP for the entire U.S. — would certainly be ruinous, and far more than enough to sink the supposed energy transition before it gets very far. But I would say that Tanton’s treatment of the costs of incremental battery storage that would be needed to get to a full wind/solar/battery system is very superficial and inadequate. He nowhere does the kind hour-by-hour calculations that Andrews did as to how much storage would be needed and at what cost.

Which brings us to Ken Gregory. As you will see, Gregory’s work is very much in the methodology of Andrews. Gregory is a retired engineer who lives in Calgary, Alberta, Canada. He joined something called the Friends of Science Society in Calgary in about 2003, and has been writing articles and newsletters for them on the subject of climate science since 2008. Gregory’s Report is basically framed as a critique of Tanton’s work, most particularly focused on the question of needed storage. From the Executive Summary:

A report by Thomas Tanton estimates a capital cost of US$36.4 trillion for the U.S.A. economy to meet net zero emissions using wind and solar power. This study identifies several errors in the Tanton report and provides new capital cost estimates using 2019 and 2020 hourly electricity generation data rather than using annual average conditions as was done in the Tanton report.

So, with the correction of that Gregory calls errors, what figure does he come up with for the full electrification of the U.S. economy with a wind/solar/battery system:

This study finds that the battery costs for replacing all current fossil fuel fired electricity with wind and solar generated electricity, using 2020 electricity data, is 109 times that estimated by the Tanton report. The total capital cost of electrification is herein estimated, using 2020 data, at US$433 trillion, or 20 times the U.S.A. 2019 gross domestic product.

Off by a factor of 109, amounting to some $433 trillion. Holy sh*t! And yes, that is about 20 times U.S. annual GDP.

So who is right? Gregory, like Andrews, provides detailed information as to his methodology and assumptions. Like Andrews, he got hour-by-hour data on electricity generation from existing wind and solar facilities, extrapolated the data to the point that assumed wind and solar facilities produced sufficient power to meet the country’s demand for a year, and then calculated how much storage would be needed to cover the intermittencies. Here is Gregory’s chart of the pattern of combined wind and solar generation in the U.S. in 2020:

The horizontal axis gives hours through a year, starting in January. There are 8760 hours in a 365 day year. So, for example, hour 4400 is some time in early July. And now for the key calculation: how much power will you need to store up at the maximum to make it through the year, and how long will you need to keep it stored until full discharge? Here is Gregory’s chart for that:


The red line represents the 2020 case and is derived from the 2020 chart immediately above. The blue line is a 2019 case based on actual U.S. power usage and wind/solar production for that year. You can see that the results are very similar, and the seasonal pattern is the same.

The maximum battery storage required is about 250,000 GWH in the 2020 case, coming some time in about early July. It turns out that the 250,000 GWH is just about 30 days usage for the U.S. So Gregory gets for the U.S. a result for maximum storage requirement that is very comparable to what Andrews got for California. This is somewhat noteworthy because California’s existing renewables generation portfolio is heavily weighted to solar rather than wind, while the full U.S. portfolio is heavily weighted to wind rather than solar. This shifts the seasonal pattern somewhat, but does not eliminate or reduce the pattern significantly.

Gregory uses a cost of battery storage of $437/kwh, which he gets from this EIA 2020 Report (also used by Tanton). Multiply $437/kwh times the 250,000 maximum storage requirement and you get a mere $87 trillion or so of storage costs to just get the existing U.S. electricity system to run all year at current demand.

Oh, but the plans of the Green New Dealers call for electrifying the whole rest of the currently-unelectrified part of the economy. It’s the extrapolation to the rest of the economy that gets Gregory’s total storage costs up to the $433 trillion.

To his credit, Gregory does not end there. He has considered many, many other cases to see how much the costs can be reduced. Those cases include things like (a) way overbuilding wind and solar facilities to reduce the need for storage even though it means wasting vast amounts of electricity at times of overproduction, (b) keeping some or much of the fossil fuel capacity around, (c) carbon capture and storage, and so forth. If you have time, go through Gregory’s report and consider whether any of these scenarios actually make any sense.

Before concluding, I want to particularly mention three of Gregory’s assumptions that are critical to the final cost calculations, and indeed to the potential feasibility of the whole endeavor:

  • The assumed cost of $437/kwh for battery storage could be criticized. Although it is a number from a current EIA report, many are predicting that battery storage costs will soon be dropping rapidly. Some predict costs within a few years of as little as $100/kwh. Andrews used $200/kwh. Obviously the number you pick could swing the bottom line wildly. But even $100/kwh would only drive the $433 trillion down to about $100 trillion — five times U.S. GDP.

  • Gregory explicitly states that he assumes throughout his calculations that you get back 90% from a battery of whatever amount you charged into it. “I assumed that the battery efficiency is 90% [8] , meaning that charging the battery with 100 MWh and discharging 90 MWh leaves the battery storage unchanged.” That assumption may be reasonable when you charge one day and discharge the next, but here we are talking about holding the thousands of GWHs of charge in batteries for six months and more. I don’t even think that battery technology exists that can do this. If it did, you could easily lose half or more of the charge in six months, or be forced to spend a big percentage of the energy in the batteries to maintain their environment to prevent dissipation. This is a potential huge cost driver of yet additional costs that Gregory does not consider.

  • Also not considered by Gregory are costs of additions to the transmission system to get power from where it happens to be produced on any given day to where it is needed, which could be all the way across the country. Gregory: “The use of S+W generation from 48 contiguous states implicitly assumes that there is sufficient unconstrained transmission capacity to share any excess or to cover any shortfall among the states.” Again, there could be huge costs here not yet considered.

Overall, I want to congratulate Mr. Gregory on what I think is a highly competent and useful piece of work — as opposed to almost everything else that exists on this subject. Will anybody pay attention? The answer is, no for now, but eventually reality will win out, and attention will be paid.