In our October issue, HDT Equipment Editor Jim Park explored the differences between battery-electric and hydrogen fuel cell powered trucks. Here's an extended interview with Rick Mihelic, director of future technology studies for the North American Council of Freight Efficiency (NACFE) on the economics of batteries.
HDT: When talking about batteries themselves, Tesla seems to be the benchmark. Are all batteries created equally? Will the other suppliers deliver batteries with similar performance at a competitive cost?
Mihelic: One of the myths about batteries is that they're all the same; they're not. Batteries are optimized for duty cycles and individual OEM tradeoffs on many design factors. Regarding cost, Panasonic makes Tesla's batteries. They make batteries for other people as well. But there are differences as far as costs go. As I understand it, cost is based on production volumes. If you build a production line that can make 1,000 batteries, the batteries are more expensive than if it makes 1,000,000. Since Tesla is the highest volume user of batteries right now, I believe they've got the lowest possible pricing.
HDT: Are all batteries, the physical cell itself that actually stores the energy basically the same?
Mihelic: They can be similar, although there are many chemistries and many trade offs for battery designers. Think of your typical AA battery; the packs we use in cars and trucks are nothing more than a whole bunch of little batteries wired in parallel. Whether cylindrical cells, pouches or other designs, that small cell is really the core technology. Part of the discussion about setting the factory up is that once you get the production line going for a battery pack, it's going to be harder to change the design. At some point, you hit a limit on what the battery can do because you're building them in volume, and it becomes less and less likely that you're going to make radical changes to the battery because you set up a production line expecting to make a million batteries without altering the process. You can't, five weeks in, decide to change the production line to do a new battery. Achieving volume so that you can have the economy of scale becomes a stabilizing factor for the design.
HDT: As a battery supplier, can Tesla provide the number of batteries that we're going to need to populate a sizeable electric truck population?
Mihelic: One of the risks when it comes to batteries is that, as demand for automotive [and truck] batteries ramps up, demand for the raw materials could exceed the supply. Some are questioning whether they'll be able to ramp up the supply, particularly with cobalt. Lithium is fairly abundant and available in a lot of places. So, it's just a matter of throwing money at mining and production. But cobalt comes from a very small part of the world and there's not a lot of opportunity to ramp that up. People wonder what will happen as the number of battery-powered vehicles continues to grow. Where will all the raw materials come from? It's an unknown, and that means it's a risk.
The Department of Energy and other researchers have been working to find alternatives for things like the cobalt-based batteries and rare-earth magnets. Getting the raw materials for the rotors and any part that must be able to create a powerful magnetic field is a concern. China has the majority of the production of that raw material. And so, supply remains a concern, especially with the current trade war.
HDT: It's a bit like how we used to discuss oil as a finite resource and much of it had to come from overseas. Pricing was very volatile. But new technology was developed, like fracking, and now the U.S. is a leading exporter and pricing has been less volatile more recently.
Mihelic: Still, that's where the speculators come in. Right now, we have projections that say raw materials pricing will drop as battery capacity goes up. But all those assumptions are based on stability in the market. The minute there's a demand for something, you get speculators going in and driving up prices. You've got these illogical forces out there that drive the market in really strange ways. Two years ago, we had that $100 oil. Now it's down in the $50s. All that was speculation. Supply and demand really didn’t change much in that period
HDT: Exactly. It doesn't take much to set the market on its ear.
Mihelic: I think some of the projections on pricing are going to be tempered. I think as volume seriously takes off, for example if Freightliner gets into building 10,000- 20,000 electric trucks a year, that's a sudden demand for those raw materials that wasn't there before. So, where does the pricing go in a high demand, low supply world? The price always goes up.
HDT: It normally does, yes. We’re watching all these charts that show prices going down, but how can that be? If demand starts to ramp up and the speculators get involved, the prices for the raw materials needed for batteries won’t be stable for long.
Mihelic: Yes, that's my viewpoint. NACFE wrote about that in its recent report, "Medium-Duty Electric Trucks: Cost of Ownership" . We identified 20 generally unknown factors and that was one of the ones I spent a lot of time on. You don't really know where those prices are going to go. Projecting the innovation curve is challenging in the first place. But, where that pricing is coming from is by looking at the last 10 years, and the progression of energy density, and the costs. And things are heading the right way for now, but the demand's still pretty low. You're talking about well over half a million Teslas to date. And maybe close to that made by everybody else, so likely there are over a million electric vehicles in the US. It's very small numbers when you think about overall automotive use combined with commercial truck use. What happens if manufacturing suddenly spikes and you have tens of thousands of Class 8 trucks? Last year we made about 300,000 class 7 and 8 trucks in the US. If you start talking about making 50,000 trucks a year with batteries, the price almost has to go up. I just can't see how price goes down in that marketplace. But that's my personal opinion. I have no facts to prove it, just judgment based on what we see in other markets.
HDT: Is there a realistic estimate for what a 250 or a 500 kilowatt-hour pack will weigh? Does anybody know?
Mihelic: I dug into a lot of numbers, and companies like AVL have actually torn these vehicles down and done that analysis, actually putting battery packs on a scale. And, at the parts level, the best number I got was 14 pounds per kilowatt hour and the worst number was about 25. If you scour all the commercial advertising for vehicles that are being made and for prototypes that are being made, you can back into a list of numbers. And I pulled all of those about a year ago and put them into NACFE's list. So, 14-25 pounds /kWhr is the range right now. A good number is going to be somewhere around 14, and that should probably improve. As volume goes up, the designs get better, the cost accountants work less material into the battery, and weight goes down.
HDT: That makes sense. It’s possible to visualize that curve. So ultimately it's likely that the batteries are going to get lighter as they get, depending on which way the economy goes, less expensive or more costly.
Mihelic: I think there's a lot more opportunity on the weight side than there is on the cost side. And really, on the weight side, as you increase volume, you can be smarter about how you design the batteries. Audi did a battery pack for one of their cars where they used the pack as structure to help stiffen the vehicle. So instead of thinking of the battery as just something that you add onto the car, you design the battery as part of its structure. And they were able to stiffen it and lighten the final product by doing that.
So, that kind of stuff isn’t happening yet with trucks, because you've only got a handful of vehicles in prototype mode. When they get into production and the smart guys in manufacturing engineering start looking at things, they'll come up with ways to get better performance out of the structure, and maybe combine it with other things to get a better, lighter-weight chassis.
The lack of significant progress in weight doesn't bother me too much. I would think that it's a natural progression of the engineers’ desire to be more efficient, and the cost accountants that they work for wanting them to spend less on raw materials. Those two forces tend to make weight reduction happen fairly quickly.
HDT: Well that's a bit like what Nikola is doing with their hydrogen tanks. They claim that they're built into the structure of the floor of the cab. Can you put a finger on a cost per fuel cell stack, per kilowatt hour, and compare it to batteries? Is that even doable at this stage of the game or is it still not through prototyping?
Mihelic: One of the experts at AVL did a slide presentation a few months ago where he showed cost estimates for fuel cells. That presentation and what little I've been able to dig up on costs from other places is that they're still quite expensive compared to other things.
There aren’t a lot of fuel cell cars out there. There are a few in the field, but the volume just isn't there yet. The Mirai, which is the foundation fuel cell for the Kenworth truck, has not been through a lot of iterations or modifications. At this point, people are just trying to make things work. Cost is something that any manufacturer needs to think about, but you always have to just make it work before you can worry about making it more cost effective. So, I think they're just trying to walk before they can run, and the cost-reduction part of it comes after development is at a run, so reducing the cost and knowing how cost will end up is a distance downstream.
I believe that you shouldn't be focused on list price at this point anyway, because, as the old joke goes, nobody pays list, right? Price is determined by the particulars of the agreement between the seller and the buyer.
The state of California wants truck makers to go to a zero-emission vehicle, and they're setting up the infrastructure to fund grants, and to provide incentives and tax breaks for the long term because they want to migrate a huge population of people to zero emission vehicles. So, I expect grants and incentives are going to be around for a long while. If you want to get the state of California to be zero emission by 2045, you're going to have to convince people to buy these things, and if they're more expensive than what we have now, they're not going to be adopted. They're using dollars that will be constantly replenished to fund these grants, so I expect they're going to be around for a while.
When Tesla started, you had to build something like 200,000 vehicles before the tax breaks disappeared, if I remember correctly. So, they did that. I think they got to that point in about 2015 or 2016. So that's six or seven years of production before they hit that wall. Kenworth, Peterbilt, Volvo, Daimler, Nikola, Tesla, BYD and others each have fielded just a few Class 8 vehicles. So, at this point, we're quite a distance from hitting any such wall and seeing the incentives going away.
People get hung up on the list price of a battery electric truck, which is $300,000 right now, and they ask, “How can people buy that? Well, it may only cost you $150,000 or even $100,000 net, to buy a truck because the government is giving you money to buy it. That will be the reality. It's the out-the-door price that you need to think about, not the list price.
HDT: I guess that's where some of the fear comes in. When people say $250,000, everybody gasps and says, "No, it's not going to work in my business." Fleet managers need to get past that.
Mihelic: Yes, these people are hung up on the list price. And in fact, quite frankly, the list price has a lot of stuff in it that isn't necessarily based on hard numbers. So, you have profit margins all along the supply chain. If you buy a part through a car dealership, there's a massive markup because there is profit at each stage. So the list price is unrealistically high. The actual price that the manufacturer paid for the thing was much lower. Workhorse last year and UPS said that they could come out with a class six that was the same cost as the class six they were replacing without any incentives. So, it's probably economically feasible to make these vehicles a lot more cost effectively.
HDT: I can see that in the smaller classes, Class 5 and 6, simply because you need fewer batteries. But is that typical on the Class 8 side?
Mihelic: On the Class 8 side, the OEMs are taking a diesel truck, and not putting the diesel stuff into it, but putting the electric stuff into it instead. And so, they're taking something that was designed for a completely different powertrain and adapting it to an electric powertrain. There's got to be a lot of extra cost built into such a vehicle that that would go away if they did a purpose-built design, engineered right around the right powertrain. So Nikola gets to start more or less from scratch. Tesla gets to start from scratch. And so, they're going to be able to optimize their costs probably better than Freightliner, or Volvo, or PACCAR, because the present OEMs are starting from something else and adapting it, and that kind of design just generally has more cost built into it.
Unfortunately, apples-to-oranges comparisons like that, which leave the electric vehicle at a major disadvantage, are pretty common. Even the professional people that I work with at OEMs will occasionally use the most conservative value for the weight of batteries or the energy density of the battery. They'll use numbers that are far out-of-date. And it biases the result, leading to the conclusion that these trucks are just going to be too heavy and that the customer just won't be able to profitably carry freight with them.
So, the fact is that they're choosing bad numbers to start with. Then they do things like comparing duty cycles inaccurately. I look at data, official test data from smart people, and they'll compare one versus the other. But, when you examine the two duty cycles, you often find that they are different, and that has biased the information because they're not doing the exact same duty cycle. That’s bothersome. Anyone examining such comparisons needs to read the details to find these unfortunate facts out, and too few people read all the details, and so you just see the headline saying the electric truck won’t be economic to run. Credibility is everything and nobody seems to care enough about that.