In reviewing anything I typically first by looking at the expertise of the source. So, in reviewing the Simon P. Michaux
paper the first thing I learned is his degree is in mining engineering. As the report is to address the mining challenges around phasing out fossil fuels it’s a good fit.
From my prior
review of the video the biggest issue I had was around the source data. He hadn’t blindly followed what was in the industry, he guestimated them with back-of-the-envelope calculations. Had he come up close to the accepted industry numbers that would have been fine, but he didn’t. That doesn’t mean he’s wrong – it means the discrepancies need to be explained. So, that’s why I’ve been eager to dig into his paper, are the mineral issues worse than most think as he does?
From the video I had a few basic questions:
- How did he calculate an additional 37.6PWH?
- Why 4 weeks of storage?
- Mineral replacement of cars
- H2 is cheap to get into, but sucks for storage and round-trip efficiency…why use it over more favorable combinations?
- What’s with the rare earth metals in solar panels?
- Why use historical mining rates for future predictions?
Except for the last, none of these items fall into his area of expertise, so it’s possible there are bad assumptions. Not in my area of expertise either, so take this all with a grain of salt. Michaux has said that his paper isn’t the end-all-be-all, rather it is proof that a more detailed and in-depth study is warranted.
How did he calculate an additional 37.6PWh?
Currently, the world uses 22.8 PWh in electricity and the EPA says that’s 25% and transportation is 29%, so 22.8 / .25 * .29 =
26.4 PWh.
For my back-of-the-envelope calculation, I’d take the amount of fuel currently consumed by vehicles as it’s a well-known non-controversial number. It should be less than the EPA number by about 15% as it wouldn’t account for planes, trains, or ships; so we’ll add that back on at the end. It also won't include EVs, but I don't think there are so many it would be significant.
There’s ~94MB/d in 2021 for gasoline and 2.9MB/d for diesel. A gallon of gasoline has 114,100 BTUs (33.44 kWh). Gasoline ICE ranges between 11–27 % and diesel ICE ranges from 25% to 37%. Using the highest efficiencies should give us the maximum amount of power. So, 94x42x33.44x.27x365 + 2.9x42x33x.37x365 = 13,010,681+ 542,820 = 13.6 PWh. EV’s have about a 77% efficiency, so that would be 18 PWh, and add in the 15% consumed by planes, trains, and ships and we get 18 / 0.85 =
21 PWh. So, not very far off from the EPA number, but still far enough away to make me wonder why the EPA number is that high.
From the paper, Simon estimated how many cars/trucks there are in the world and used the average annual mileage. The analysis is painstakingly 100s of pages. I understand why he went that way, he needs the number of vehicles to calculate the minerals/mining to replace with an EV fleet. But while his data sources are probably the best possible sources for actual number of vehicles, they are very low quality for calculating power requirements, at best 30% (which just happens to put them into range of the EPA numbers).
At this point, I’m convinced that the base number from his calculations are egregiously high, and that will ripple downstream to the quantity of minerals needed.
Why 4 weeks of Storage?
Most of the plans I’ve seen calculate around a day or less of storage is needed. The reason in those calculations are primarily economical, that is it is cheaper to overbuild renewables and share power to locations where renewables aren’t producing via an interstate grid.
On page 183 of his report, He cites a Steinke et al in a 2012
Grid vs. storage in a 100 % renewable Europe for the EU calling for 2 days storage plus 10% at 100% renewable, a book from 2015 Droste-Franke, chap 6
Electrochemical energy storage for renewable sources and grid balancing that calls for 4 weeks, and a 2020 book
Palmer, Energy storage and civilization: a systems approach proposing 7 weeks. To resolve this mystery, we’ll have to look more into it. There are a myriad of other papers showing much shorter timeframes, my guess is he was trying to be conservative and picked a number between them. It’s going to need more exploration, perhaps I’ll start with the NREL
report.
Mineral Calculations
A criticism was that he cherry-picked high-end EVs, in looking at the data I believe this is unfounded. It’s that he used existing data and pretty much all the existing EVs made are high-end. EVs for the rest of us are something manufacturers have only recently started to address. I couldn’t find the underlying assumptions for the mineral breakdowns in the report. For solar panels, I suspect he looked up all the materials used in the last decades as his baseline. Unfortunately, that’s exactly wrong as solar panels became cheaper over time as more exotic materials were removed from them. A modern roof-top solar panel is primarily Aluminum, Copper, Silicon, Silver, and Zinc with no exotics. Yes, specialty panels like CIGs are still manufactured with some exotic materials, but they're not what you'd put in a solar farm. Not in the paper, but from the video, his analysis is a 100% replacement for vehicles ignoring the recycling of ICE vehicles, which if included would greatly decrease the mineral needs.
Why H2?
High-density fuels are needed for planes and are preferable for long-distance transportation. He walks through the different biofuels and hydrogen was the only one left standing. Personally, I think his H2 analysis is a poster child for why the hydrogen economy isn’t as bright as hoped. In section 21 he rules out biofuels as being limited by quantity but believes there would be enough for aviation. On page 448, he has an interesting bit about biofuel from algae. I’ve got a softspot for algae as you could theoretically grow enough algae to power all transportation. The problem is it’s all wet, literally, and drying it to where it can be processed is so energy intensive it’s not currently worth it and he came to the same conclusion. The tech has promise though and a breakthrough there could be a game changer.
Why use historical mining rates?
Mining is his area of expertise, and I didn’t see anything about the rates in the paper like there was in the video. But I don’t see it, sure it’s not as easy saying if x is one guy with a shovel to get 5x in the same time you need 5 guys with shovels. But using historical mining rates seems even more wrong as mining is an endeavor that is based on supply/demand. That is historically we don't mine more as there's no need to mine more. I feel like I’m totally missing something here…any thoughts?