Recently the International Council on Clean Transportation (ICCT) published a total cost of ownership report on trucking decarbonization which found that energy costs for battery electric trucks and green hydrogen fuel cell trucks would have almost identical energy costs per mile in 2050. Clearly there was a methodological issue, as at least three times as much electricity is required for the same forward motion as with batteries. Now they’ve asserted that their choice was deliberate and methodologically sound to the significant detriment to their credibility.
Energy price per kilometer should be at minimum three times more for the hydrogen pathway, but instead was only 70% more in 2030 and 10% more in 2050. Further, the report claimed that hydrogen fuel cell trucks would be cost competitive with diesel trucks in 2030. All results confound the most basic of comparison logic.
There are various ways in which this result might have been reached, and so I set out to assess what they had done to arrive at this completely inaccurate result. Possibilities included centralized green hydrogen production in high renewables areas in northern Africa with pipeline transmission and distribution with inaccurate production and transmission costs, a confusion over blue hydrogen versus green hydrogen prices or an inaccurate assumption of the energy requirements and cost of manufacturing hydrogen from water and electricity. I assessed it and published on this a few days ago with a call for a retraction of their erroneous report.
The last possibility turned out to be the case followed closely by not aligning assumptions across scenarios. They had relied on an earlier ICCT assessment report which assessed manufacturing hydrogen at truck refueling stations from green electricity and water and found it was the cheapest pathway. However, that report had a significant flaw in that it assumed that a small scale MW electrolyser at the station would be powered by wholesale costs of electricity with only minor adders from the power purchase agreements for wind and solar.
There are multiple flaws with this that are worth calling out.
Electrolyzers are expensive pieces of kit. Amortizing that cost across the lower cost kilograms of hydrogen requires high utilization with firmed electricity. That’s very true for major industrial scale electrolysis facilities and it’s if anything more true for less efficient tiny electrolysis facilities. The wholesale cost of electricity is not what anyone pays who wishes to run their facility 60% of more of the time.
Any facility wishing firmed electricity ends up paying the equivalent of grid commercial costs of electricity as they have to buy either pay electrical utilities for the firming services or buy storage themselves or reduce utilization and increase costs. Assuming a MW scale electrolyzer facility would be setting up significant additional contractual agreements and infrastructure for this is also problematic.
I consider that earlier report naive and that the findings aren’t supportable.
And then to the total cost of ownership report. They used the local manufacturing cost of hydrogen from the earlier report without disclaimers about its assumptions on the price of electricity. They assumed a scenario where this was put in place and electricity at the refueling station was indeed being paid for at wholesale rates.
Then for battery electric trucks at exactly the same location using electricity from the same wires, they assumed that the electricity would be at commercial rates, 3-4 times the rates for the hydrogen manufacturing.
To the credit of the analysts, they still found that battery electric trucking was cheaper than diesel or hydrogen in all scenarios, but the published cost of energy per mile for the hydrogen scenario was both a fraction of any realistic scenario and had very different assumptions for otherwise identical scenarios.
This is an easy error to make. The ICCT has historically, in my opinion, attempted to be a fair and honest analyst of transportation. I assumed that they would accept their mistake, correct it and move on quickly.
Sadly, that’s not what has transpired. First, they were silent for several days after the mistake was pointed out to them. Then, earlier today, six days after the first identification of the mistake in a communication with one of the lead authors, the other lead author informed me that they were standing by their report with its grossly inappropriate comparison and conclusion, and further that they wished to publish a rebuttal in CleanTechnica. I’ve connected them to CleanTechnica’s editorial staff so it may surface.
The full text of the communication from Felipe Rodríguez is:
“Thank you for your comments and patience while we thoroughly reviewed our work in response. We take this criticism seriously. It’s refreshing to see such thorough scrutiny rather than a flat-out dismissal of the findings. It’s this kind of engagement that enriches scientific discourse.
In brief, we stand firmly by our findings and methods. Your analogy points to a potential bias in our methodology. Rest assured, our choice of ‘apples’ for hydrogen and ‘oranges’ for batteries was deliberate and methodologically sound. To touch on a few key points:
➡ We chose PPAs for our 1 MW electrolyzer to ensure the use of renewable electricity, essential for genuinely green H2. We simply cannot model grid-mix here
➡ Our approach uses the same structure of grid fees and taxes for green H2 production and charging. However, there will be differences in demand fees due to the different load profiles (stable 1 MW for green H2 vs peaky 20 MW for the charging hub)
➡ The grid upgrades needed for a 20 MW charging hub will impact the electricity cost that the CPO sees
➡ The infrastructure cost of the MW charging hub, especially given its low utilization, impacts the electricity price that the trucker sees. This is the charging price we reported.
Our report encompasses various scenarios (pessimistic, central, optimistic) that differ in multiple parameters, with the cost of electricity at the meter being just one of them. We report here our central scenarios, which are always in the range of those in the scientific literature. Still, given the interest this has aroused, we plan to extend our analysis with an additional scenario in an addendum where a PPA for the MW charging hub is presented.”
This is, of course, errant nonsense that is so heavily weighted to try to justify hydrogen as an option that it’s hard to lend it credence. To paraphrase, what the ICCT is saying is:
There’s absolutely no way to make hydrogen at the refueling station without additive green electricity, so the only mechanism that would be supported is with PPAs which justifies the cost point. It’s a completely unrealistic and flawed scenario, but they are correct that it is the only one that is remotely credible as a decarbonization strategy.
On the other hand, they say that using grid electricity in batteries directly is completely viable and cost effective without PPAs as a decarbonization strategy and so that’s what people will do. That’s also true. But further they say that the same PPAs can’t be used for battery electric vehicles.
They ignore the cost of firming for hydrogen and then assert that the much lower cost of firming a third of the electricity with batteries isn’t completely viable. This burdens the electricity case and unburdens the hydrogen case inappropriately. Further, they ignore both locality and temporality. The three concepts are necessary for manufacturing actually green hydrogen as otherwise whatever carbon footprint the electricity has is multiplied by a factor three to four.
Briefly, additionality requires that a green electricity consumer buy as much electricity from specially built wind or solar farms that are adding low carbon electricity to the grid as they consume. Temporality requires that the electricity consumed by the end user be matched to the generation of the wind and solar farms (firming) or marginal new high carbon electricity will be demanded and hence be provided. Locality requires that the wind and solar farms be fairly close in geographical terms to the demand for hydrogen, otherwise the high carbon local electricity will be expanded with a major new demand source.
Of the trio of requirements, only the first is met. As a result, they could set up a hydrogen refueling station in Poland with its 80% coal grid and build a wind farm in Spain that manufactures low-carbon electricity without any temporal connection to their hydrogen manufacturing, and be increasing coal demand locally. That would mean that a kilogram of hydrogen put into a truck in Poland wouldn’t be green or low carbon, but pitch black.
At present, Poland’s grid intensity is 635 grams of CO2e per kWh, down 20%. Including compression, removal of water vapor, storage and pumping into trucks, roughly 60 kWh of electricty is required per kilogram of hydrogen. That means the carbon intensity of a Polish kilogram of hydrogen would have a carbon debt of 38 kilograms of CO2e. That’s well above the carbon intensity of an energy equivalent mass of diesel with it’s roughly 10 kilograms of CO2e.
They are indeed correct that using unabated grid electricity to manufacture hydrogen is not a viable solution, but don’t lean into what that means sufficiently with temporality and locality, as a result their cost figures are vastly below reality.
By comparison, Germany’s grid intensity of 333 grams of CO2e per kWh results in double the carbon debt for the same energy as diesel. France’s nuclear heavy grid has a low carbon intensity of around 70 grams of CO2e per kWh and so hydrogen manufactured there is only 40% of the carbon intensity of diesel.
Obviously, battery electric trucks have a third to a quarter of the carbon intensity per kilometer with grid electricity as hydrogen fuel cell trucks do, so they are clearly the sensible choice and that ratio will simply continue, not change.
The answer to this conundrum isn’t to build three times the wind and solar to massively extend hydrogen demand, but to reduce hydrogen demand as much as possible as quickly as possible and to build as much wind and solar to decarbonize all grid electricity rapidly. It’s unclear why hydrogen for energy advocates find this difficult to understand, and it’s unclear why they think that buying wind energy in Spain is particularly useful when they are adding demand in Poland.
The ICCT has significantly weakened its credibility. Discussions I’ve had with a global set of energy and decarbonization analysts over the past week have left us scratching our heads. While I had given the Council the benefit of the doubt earlier, assuming that they had made an honest mistake, now it’s clear that inside the ICCT there are strong pro-hydrogen factions and stakeholders who have warped the organizations’ discussions and analysis.
While the ICCT has rejected retraction, I call again for doing so. Further, I think that they should retract their earlier and naive green hydrogen and rework with it temporality and locality requirements that would cause the cost per kWh to rise to roughly the same as commercial rates of electricity. Finally, they clearly have an internal problem with institutional bias that is impacting their governance. I would recommend that they make a serious effort to redress that, which might include changes in senior leadership, mandates and standards.
