tl;dr: Even the most optimistic predictions of eTruck penetration are far too pessimistic.
Reading the article, I found it extremely pessimistic. Or perhaps “optimistic” given the general tenor of their business. Here’s “Exhibit 1” from their analysis:
The first thing I saw is how slowly they expect electric trucks to penetrate the market, especially heavy duty trucks (HDT) and in China. This may not be clear at first, given that they are actually assuming an exponential growth rate, but a quick look at the axis scales shows how slowly they expect it to grow.
Here’s the money quote:
Three critical assumptions most affect TCO breakeven points. The assumptions that drive TCO uncertainties include the development of fuel and electricity efficiencies for ICE or BECV technologies, the cost of batteries, and the cost of fuel and electricity. [Their emphasis]
Battery Costs: The Key error
The key error, in my view, involves “the cost of batteries,” where they are probably using a forecast like one of the ones below.
“Moore’s Law” for Li-Ion Batteries
The problem with those forecasts is that they are totally ignoring the real exponential growth of deployment, and the way it drives real manufacturing costs, as the big red arrow and star above demonstrate. So let’s try using “Moore’s Law”, which predicts an exponential decrease in cost over time. We’ll start with a graph of 2010–2016 costs from Bloomberg New Energy Finance (BNEF):
These average out to an annual “growth” rate of about 0.805 (the fact that the “growth” rate is smaller than one means the cost will decline). In this case, then, the cost will drop each year to roughly 80% of the previous year. Here’s a projection for through 2030:
Now, the most important thing about this cost decline is that costs become a fraction of what they were real quickly. By 2020 they’re less than half what they are this year.
Of course, we can’t be sure they’re going to get this cheap this fast. But most of the projections seem real sure they won’t, with no good reason that I can see.
Although there’s no proof in any way, the document I linked above suggests that Wright’s “Law”, also called “Learning Curve” is probably the best predictor of actual costs. This “Law” says that for every doubling of deployed technology, the cost will tend to decline by some fixed fraction, such as ~16% for Lithium-Ion storage for Electric Vehicles.
Don’t confuse the annual “growth” rate with learning rate, one depends on time (“Moore’s Law”) the other on deployed capacity (Wright’s “Law”). They do seem to predict very similar outcomes during the early growth phase of new technology, but as deployment fills its capacity, or competing technology reduces advantage, deployment can slow despite reducing costs.
Also, while learning curve can reduce manufacturing costs, shortage in materials (in this case probably Lithium and Cobalt) can stall the cost reduction and thus slow the entire process. In addition, governmental interference, such as tariffs, can distort otherwise natural growth curves.
Working together, these two “Laws” can roughly predict cost reduction for batteries.
Now, the next issue with the McKinsey report involves an unwarranted assumption regarding recharging:
In the example shown, the earliest breakeven point occurs at a distance travelled of about 200 kilometers a day. This sweet spot of operation means the battery is large enough to enable efficient operation without too many recharges, while ensuring sufficient annual distance to benefit from the lower cost per kilometer.
They appear to be assuming that the battery will be built into the actual truck, rather than swappable. When I went looking for info on this, I found a very interesting recent article: Can Electric Truck Battery Swap Really Work? Obviously the technology is very simple, easily matured in a few years. Here’s an example of a forklift with swappable battery packs:
Now, the immediate objection would be that a trucking firm would need to buy many redundant batteries. This would raise the cost, although it would allow much greater ROI for the trucks themselves (in full use). And note how, if batteries continue their exponential cost decline (see above) by 2022–2023 it would be possible to own three batteries for every truck for the same cost as one today, allowing most of them to sit in the charging stations.
Full-time truck usage will probably be feasible by about the same time, using self-driving trucks, so the drivers can nap during most of the drive, only waking up for critical stretches. In addition, swapping out drivers could allow trucks to effectively run most of the time, excluding only quick battery-pack swaps and trailer exchanges.
(I should mention that while researching this article, I found a story why “Standardized Electric-Car Battery Swapping Won’t Happen: Here’s Why”. I certainly do not agree that the problems they mention will necessarily block development of swappable batteries in private automobiles, but in any event they don’t really apply to use by large trucking companies or cooperatives.
(I also found this article on optimizing. It seems to be pointed to passenger transport rather than trucks, but its methods could probably be ported to modelling a truck system.)
Putting Together the Technologies
So here, we can imagine a much more rapid roll-out in all categories of electric truck, due to the ability to use swappable batteries to rapidly reduce the TCO per mile driven:
- Trucks don’t have to sit idle while charging
- Driver time can be optimized
The key to making this work, in addition to rapid decline in battery costs, is a sophisticated scheduling system; sophisticated enough to plan for charged batteries being available at the right charging stations when the right trucks arrive. As far as I know, all the problems in designing and building such a scheduling system are standard software engineering problems, so it shouldn’t be an issue.
An Extra Benefit
Another technology that’s getting cheaper fast is solar power. Which means that charging stations could be supplied with enough solar power to make them (at least partly) independent of the electrical grid.