We recently ran a perspective piece on Tesla and Electric Vehicles in which John Petersen suggested that "a grid-powered electric vehicle might make individual drivers feel warm and fuzzy about themselves, but from a public policy and resource-conservation perspective, it’s the most wasteful plan ever devised."
We had a rebuttal from NRDC attorney Max Baumhefner.
Today, we're publishing a guest piece from Nick Butcher which takes a close look at whether there is a battery material resource limitation to widespread electric vehicle penetration.
There's been much debate over the past days, weeks, months, years, and decades over the viability of electric vehicles (EVs).
But Tesla's Model S is a fast, luxurious, and frankly stunning sedan with a range of 300 miles and a recharge time of as little as 45 minutes. Case closed, right?
Hardly. The arguments rage, at least in some corners, loud as ever. I've recently decided to weigh in with my two cents, as I believe EVs are a part of the future mobility solution -- a big part -- and that economic solutions are readily achievable without any technical leaps.
These are three critical topics:
- EV batteries: Sustainable, or a waste of finite material reserves on toys for the rich?
- EV emissions & energy security: Clean, green, and sustainable, or no better than burning oil?
- EV economics: Will electric vehicles ever be directly commercially competitive with internal combustion vehicles?
In this post, I'll address the first point above: the battery production capacity of our planet given the known resource base, and the impact battery production will have on raw material demand. There is a lot of existing peer-reviewed work on this topic, some of which I'll reference.
Let's start with the myth. John Peterson recently stated it quite explicitly:
The bottom line is that grid-powered electric vehicles are unconscionable waste masquerading as conservation. There are enough batteries and battery materials to make electric vehicles for the few, the rich and the mathematically challenged, but there will never be enough batteries or materials to permit the implementation of grid-powered electric vehicles at a large enough scale to impact global, national or even local oil consumption. It's not an effective solution.
That's put pretty unambiguously. The general basis of the argument is that key materials for electric vehicles are supply-constrained and can't possibly scale to meet demand of more than a few hundred thousand units. If this were the case, we should then conserve those materials and try to use them in the most efficient manner possible -- which might not mean putting them in EVs at all.
But it's not the case. The resources are not scarce. "Peak battery" is not following hot on the heels of peak oil.
To understand why, you need to know that there are many different types of lithium-Ion battery, and they use different materials in their fabrication. Traditionally, cobalt was used as a major component (LiCoO2 cathode), but today, not only have cells been developed that use far less cobalt while delivering better performance, there are many cells which use no cobalt at all. I'll look at only a few of the most common for now -- research on numerous others is at an advanced stage, with many already in production. I'll present all the numbers at each stage so you can reassure yourself I'm not trying to pull a fast one; feel free to check my figures and let me know in the comments if you think I've made a mistake.
Below is a table of cell chemistries, detailing the compositions of each cell anode and cathode (negative/positive) by element. This is pretty dry, but it's a key input to what follows. No need to know it by heart.
To get 1 milliamp-hour of cathode capacity in a LiMnO4 cathode, you need 6 micrograms of manganese. If you want to check these calculations yourself, have a play around with the tool at WebQC. There's a good summary of the specific capacity per gram of each material on the wikipedia page, and in this recent presentation from TU Delft.
So far, so boring. Let's combine the milliamp-hour values with the voltages and get a table that tells us how many kilograms of a given material we need for 1 kilowatt-hour of battery. Below I give both the theoretical values and a 'real-world' value that reflects realistic figures today.
Want to know how many kilograms of cobalt there are in a 40 kilowatt-hour LiNiCoAlO2/C6 battery pack? Just multiply 0.278 by 40. For all further calculations, I'll use the 'real-world' values to be on the safe side.
Now that we know how much of each material is required to make a given battery, we can work out how many batteries we can make with what's available. Being strategic thinkers, we'll first look at reserves, and then drop down to production for the tactical level.
A few things to note:
- The units are 'Battery Packs,' which can be considered equivalent to 'Cars' -- except that you'll note I'm taking the extreme case (Tesla S Performance Edition) and assuming 85 kilowatt-hours per pack. If we made Nissan Leafs instead, we'd get more than three times as many. If we put it into Renault Twizys or GM EN-Vs, we'd get around fifteen times as many. Urban planners say that the mega-cities of the future will have many of these small EVs. I can't wait to see Tesla's take on this personal vehicle category.
- I'm not being greedy and suggesting that we use all of the resources for EVs. Quite the opposite. Except for lithium (which is in relatively limited use today, and EVs will be the main growth driver), I assume at least 80 percent of all other resources are used elsewhere. Overly conservative? Probably. Let's go with it.
- Because the resource numbers are critical, I didn't use a single source, but rather shopped around several. Please feel free to search yourself -- googling '<element> reserves/production' will get you there very quickly. I've just used fill-in values for carbon and oxygen. I hope I don't need to prove to anyone that they're abundant.
The results are in! As you can see, the alarmists' fears weren't entirely unfounded: if we stuck with lithium-cobalt-oxide, we'd have a real problem! Twelve million cars would be the 20 percent resource share limit.
Fortunately, that's not happening; LiCoO2 isn't even a good battery for EV applications. The next constraint comes with LiNiCoAlO2 (as Tesla uses in the Model S), but even that constraint doesn't start to bite until we've built 110 million vehicles, and that's assuming we only allocate 20 percent of the resource. Still, 110 million isn't much at all if we consider the total number of vehicles needed for the next hundred years. Are we stuck?
No. We don't need cobalt -- it's used today in some cases because it works well and it's abundant in comparison to current demand. But there are numerous substitutes under development and in production, and some are expected to perform even better. The two alternatives I show don't have any issues until we run into lithium at around 1.5 billion vehicles. LG Chem, BYD, and Samsung are heavily invested in LiFePO4, just to name three giants, and they're doing just fine (LG supplies GM, BYD builds their own cars, and Samsung is partnered with Bosch).
When you look at the details, it turns out that the only thing worth losing any sleep over is lithium. Here I was slightly concerned, as it's one material that's not already used in huge quantities. A USGS estimate puts lithium reserves at 10 million tons. That'd be a bit close for comfort! That was back in the 1970s, though -- a more recent study by Evans put the figure at 30 million tons. That's a bit better, but still tight (as you see above, 1.5 billion cars). But now, SQM estimates reserves may exceed 60 million tons! The evolution is outlined in this report, and the reason is clear. With USGS reserves already at 10 million tons, and annual demand currently only around 0.034 million tons, we have enough known reserves for 300 years at current extraction rates.
It's not that there's a shortage, it's that there's so much that until the last few years, no one has bothered to look for more. In fact, lithium exists at similar concentration in the earth's crust to lead and nickel. The question is only one of economic extraction and technology, and with current lithium prices only accounting for around 2 percent of the cost of a LiFePO4 battery in terms of $/kWh, that's not something we need to worry about anytime soon.
I'm only looking at mineral reserves, though. They're in the ground -- to get them into EVs, we need to have capacity in place to extract them and manufacture them into batteries. Fortunately, the major manufacturers are on the case.
Much as I'd like it to, EV production isn't going to increase to 30 million vehicles per year overnight. Let's take a quick look at how many EVs we could make today.
Even with today's levels of production and taking only a small share of the resource for vehicles, there's plenty in the pot for a half-million Tesla S class EVs per year. It's going to take a few years for EV production to ramp to these levels, though, and that gives time for the supply chain to adapt. Right now, the supply chain is adapting so fast that Roland Berger (among many others) is predicting significant overcapacity by 2015. IDC Energy Insights says production capacity will reach 26 gigawatt-hours that year.
What impact might this 26 gigawatt-hours -- remember, it's supposed to massively exceed demand -- have on raw material demand? Let's have a look.
Well, that was totally unexciting. In three years' time, the industry needs to ramp up lithium production a bit. No surprise there, and it's already underway. Aside from that -- and ignoring LiCoO2, which we've already established that no one seriously considers (and for good reason) -- the only minor red flag is the 8 percent of cobalt we might use if all the batteries were NCA, which they won't be. And even if they were, 8 percent is hardly a reach. Cobalt production increased by 27 percent in 2010 alone (Table 3 & 4). Batteries today already use 25 percent of cobalt! So much for EV growth monopolizing some precious slice of the resource pie.
But 26 gigawatt-hours? Too abstract - I want a number in cars. Let's take the bull by the horns and say we want 15 percent of total global light vehicle output, call it 10 million cars a year, to be Tesla Model S in 2020. The big version with all the bells and whistles and a huge battery. What then?
Finally, some numbers that look like at least a bit of a challenge. We have to significantly increase lithium production -- no debate there. But it's only an increase of 22 percent year-on-year for the next eight years. Trivial? No. Doable? Yes. We've seen that the resource base isn't a constraint, and that the production scale-up is already underway.
We'd be in stickier territory if everyone stayed with cobalt, but there are already perfectly good alternatives today, let alone a decade from now. Nickel would also need some attention, but, as with cobalt, we don't even need it at all. Manganese chemistries would drive a tiny bump in annual demand; 1 percent growth would make it irrelevant. Everything else doesn't even rate.
There are two things left to address, and I'm only addressing them because otherwise they'll be used as a refuge by naysayers. These two things are copper and neodymium, and it's very simple, so I'll make it quick.
Copper. At an absolute top-end estimate, an EV might require 100 kilograms of copper. Realistically, it'll be more like 50 kilograms; the Tesla Roadster uses about 30 kilograms. The average conventional car already uses about 20 kilograms. To avoid debate, let's use the figure 100 kilograms. Last year, global copper production was 20 million tons. Ten million EVs per year would consume at most 5 percent of this, worst case. So this is a non-issue.
Neodymium. This rare earth element isn't actually particularly rare. It's used in the magnets of permanent magnet motors. How much do we need to make our 10 million Teslas?
None! Tesla went with high-speed induction machines instead. The company doesn't use neodymium magnets, and the performance speaks for itself.
If Tesla, or any one else, wants to make 10 million long-range performance EVs per year in eight years' time, neither the raw material reserves nor the market production capacity will stop them.
We don't live on a strange island with only enough battery capacity/potential to meet 0.1 percent of our vehicle production capacity. We live on an island with copious battery reserves -- the thing we're running short of is oil, which can't be recycled. Hybrid drivetrains and light/efficient vehicles are a great choice if an EV doesn't suit you today, but they'll only ever reduce oil consumption, at most by 25 percent (hybrid alone) to 50 percent (hybrid and lightweight). It's a very worthwhile improvement, but once we double the vehicle fleet we're back where we are today - beset by supply constraints and global energy security tensions. They're no solution to the oil crisis, just a rearguard delaying action while we gather speed for the real shift.
I'm bullish on electric vehicles, and I've checked my sources. Have you?
Nick Butcher is an engineer from New Zealand, now living in Switzerland. He has almost a decade of experience in product development in the energy industry; most of it relating to grid interactive power electronics applied to a range of applications including electric vehicle fast charging and grid battery energystorage; both with engineering multinational ABB. He now splits his time between Ampard (an energy storage startup), SwissKitePower (Kite Energy), several other embryonic ideas, and freelance work.
Disclosure: The author has a long position in TSLA.