Chevrolet and Ford cruise into Detroit Auto Show leading with truck introductions
Bee’ah to add 50 Tesla Semis to fleet; first in Middle East

DOE analysis suggests rapid convergence of FCEV and BEV TCOs; FCEVs less expensive for majority of LDV fleet by 2040; mass compounding

In 2020, battery-electric vehicles will be a cheaper vehicle option than fuel cell electric vehicles for the majority of the light duty fleet (79-97%), according to a new study by a team at the US Department of Energy (DOE) Fuel Cell Technologies Office (FCTO). However, the cost of the two powertrains will converge quickly, and by 2040, FCEVs will be less expensive than BEVs per mile in approximately 71-88% of the LDV fleet, according to the analysis. Additionally, FCEVs will offer notable cost advantages within larger vehicle size classes and for long distances.

Accordingly, the authors conclude in their paper, published in Transportation Reseach Part C, there will be a competitive market space for both FCEVs and BEVs to meet the different needs of light-duty vehicle consumers.

A common notion among automakers is that BEVs will compete among smaller vehicle size classes with shorter driving ranges, and that FCEVs will compete among larger vehicle size classes with longer daily ranges. A key factor that drives this assumed market segmentation is the difference in mass compounding.

For BEVs, as the capacity of the battery pack increases, an ever-greater fraction of that capacity is used to move the mass of the batteries rather than the mass of vehicle, passengers, and cargo. This results in a nonlinear relationship between vehicle purchase cost and vehicle range. For FCEVs, after adding the basic components of the powertrain—i.e., the compressed gaseous storage tank, fuel cell, balance of plant components, and small battery—an increase in vehicle range requires only slightly larger components, which has a relatively small impact on vehicle mass and cost. Differences in mass compounding between BEVs and FCEVs may also be visible across vehicle size classes as the ratios of mass, stored energy, and range change.

This paper advances the conceptual framework of mass compounding described above by examining costs of light-duty BEVs and FCEVs across a spectrum of vehicle driving ranges and size classes. Total cost of ownership (TCO)—including the time discounted vehicle purchase, operating, and maintenance cost—is estimated for FCEVs and BEVs for 77 market segments, defined by vehicle size class and vehicle effective range between refueling. Additionally, costs of range-related inconveniences are added to each vehicle segment. This segmentation helps elucidate the relative economic competitiveness of BEVs versus FCEVs into the future.

—Morrison et al.

Fraction of LDV fleet that is cost competitive for FCEVs (BEVs) as a function of days of inconvenience per year. Inconvenience penalty is assessed to vehicles that do not belong to a multi-vehicle household. Morrison et al. Click to enlarge.

The team used DOE’s Autonomie model to project vehicle component-level costs for FCEVs and BEV-50s through BEV-300s (50-mile to 300-mile range EVs at 50-mile increments) for the period from 2020-2040.

The paper assumes a 5-year lag between costs from Autonomie and real-world costs, given an assumed (and typical) 5-year lag time from initial vehicle R&D to the showroom.

The authors then performed 5 post hoc calculations on the Autonomie output:

  • Net present value of total cost of ownership (TCO) per mile for each vehicle range-size segment.

  • Linear scaling of Autonomie cost outputs for 5 generic vehicle classes for seven additional vehicle classes.
  • Translating the high-volume costs calculated by Autonomie to low volume costs—for the fuel cells, gaseous storage tanks and hydrogen production and delivery.

  • Interpolating costs of BEV-50 through BEV 300 assuming an exponential rate of cost increase, consistent with the mass compounding effect. Costs are adjusted to reflect real-world range.

  • Addition of an inconvenience penalty associated with vehicle range. BEV drivers are assumed to be inconvenienced when maximum daily mileage exceeds BEV range. FCEV drivers are assumed to be inconvenienced when daily range exceeds assumed refueling distance of 300 miles.

The TCO analysis did not include the time cost of refueling; vehicle performance; the capital cot of fuel infrastructure; or social costs.

Among their findings:

  • Both FCEVs and BEVs have market segments with sometimes substantial cost advantages over one another, especially in the early years. For example, a BEV-50 pickup truck in 2030 is more than $1.00 per mile cheaper than an equivalent FCEV.

  • The number of FCEV-competitive segments grow over time and the relative TCOs become more favorable for FCEVs and the fuel infrastructure is built up.

  • By 2040, for all vehicle classes except pickup trucks, all but the 50- and 100-mile range segments of the 77 range-size class segments are cheaper for FCEVs than BEVs.


  • Geoff Morrison, John Stevens, Fred Joseck (2018) “Relative economic competitiveness of light-duty battery electric and fuel cell electric vehicles” Transportation Research Part C doi: 10.1016/j.trc.2018.01.005



This study confirms that FCEVs will soon be competitive with BEVs, specially for all weather, extended range, larger vehicles.

A real breakthrough on 10X, much lower cost, higher performance future mass produced SS batteries could improve the competiveness of extended range 150+ KW BEVs?



The only thing that this report will probably confirm is how hard it is to make predictions more that 5-8 years into the future. One of my favorite quotes on this is from Buzz Aldrin (Apollo 11 astronaut): "You promised me moon colonies and all I got was FaceBook". There is also Yogi Bera comment: "It is hard to make predictions especially about the future."

Maybe by 2040, we will have almost free electric power from fusion which can be used for high temperature electrolysis although I would not bet on it. I agree with the comments about "Mass Compounding" but that ignores the possibility That we will have batteries other Lithium Ion. The Lithium Sulfur cell that Rice Univ and others developed promised about 4 times the energy density of Lithium Ion and extremely fast charging. Will this research or other similar research make it to the market. I do not know but I would rather bet on fast charging Lithium Sulfur than fuel cells. Maybe by 2040, there will be something completely different that we make us wonder why we ever looked at either batteries or fuel cells (Cold Fusion? :))



Almost every week there are news items on this forum about fuel cell and hydrogen technologies which would have at least as great an impact as lithium sulphur batteries.

I don't know what will pan out, but multiple possibilities across a broad front increase my confidence that we are on a technological wave which is going to get us places, although in no way the same as the overworked metaphor of Moore's Law.

I am also hopeful that at some time a host of improvements will occur in batteries, but I certainly don't know when.


I really can't evaluate the claims by the researchers, as it is behind a paywall and without the assumptions used such as the rate of improvements in battery energy and density and analogous stuff for fuel cells and hydrogen production, it is simply a claim, not an argument.


"Predictions are hard to make, especially when they are about the future" - Richard Feynman PhD


I am reasonably confident that this headline is obfcuse and misleading. Can anyone say from referencing the article that it refers in any way to the cost of operation per KLM?

My best reading is that it refers to the cost of powertrain including storage ank of vehicles with comparable mileage. So it simply states the obvious .

The only clever thing is the way it couches hints and misleading direction to confuse the main topic of interest for nearly everyone. That is the comparative cost of operation.

Why they couldn't just show that a 50klm bev is cheaper to both buy and operate than a 50klm F.C. and that there is a linear relationship to distance?

If you are stupid enough to pay the firewall but smart enough to be embarrassed for doing so assuming that will help you understand you will probably keep mum and the scam will achieve its objective.

It seems like the fake news dept has completely in control in its objective. Tell a lie once its a lie tell a lie a thousand times and it's becomes true.

It's very sad to see the methodical attempts of dumming down the conversation (esp0 in the USA.

" However, the cost of the two **powertrains*** will converge quickly, and by 2040, FCEVs will be less expensive than BEVs per mile in approximately 71-88% of the LDV fleet, according to the analysis. Additionally, FCEVs will offer notable cost advantages within larger vehicle size classes and for long distances.


The only solid thing in the extract is that FCEVs are more competitive as vehicle size increases.

But then anyone with any knowledge of the subject knew that anyway.


Note they are talking about single car households.
A much simpler and more immediate solution is to have a short range BEV and access to a second ICE or shared FCEV. Many households have 2 or more cars and there are car sharing schemes, or you could just rent one or borrow / swap one from a friend etc.
Governments could make this happen faster if they made it easier to share cars, especially in terms of insurance.

At the simplest level, all you need is one ICE driver who is prepared to swap cars every so often.
This could be an acquaintance or someone found by an app.



If this works out well, then one vehicle should do:

'An established and innovative vehicle company from China has placed an order for PowerCell S2 fuel cell stacks, worth approximately SEK 1 million. The customer, who develops passenger cars powered by fuel cells, will test the fuel cell stacks in the company’s electric cars.

“China has high ambitions in replacing today’s fossil-dependent vehicle fleet with electric vehicles powered by batteries and fuel cells, and has extensive support for the construction of these so-called New Energy Vehicles. This support has recently been adjusted to benefit fuel cell cars, and simultaneously we see cities and individual companies taking own initiative to further accelerate the conversion. We are proud to be able to contribute”, said Per Wassén, CEO of PowerCell Sweden AB.

For several years, the customer has used battery power in their vehicles, but has found that the limited capacity of the power grid, the lack of access to charging posts and the slow charging of the batteries limit the potential of progress.

“The fuel cell stacks, 20 kW each, ordered by the customer will be used as so-called range extenders in the cars. This makes it possible to drive a lot further than if you were forced to rely solely on battery power. It’s a smart way to combine the respective benefits of batteries and fuel cells, and it’s a background to the growing interest in fuel cells”, concluded Per Wassén.'

20KW should just about cover cruising, so the range becomes as big as the hydrogen tank is, whilst everyday driving is covered by the battery.

The questions are space, to fit all the stuff in, and of course cost.


@Dave, I am with you on the 20Kw range extender. IMO, it is all you need for a private car.
In my view, the power source doesn't matter that much as it won't be used that much. It could be a petrol or diesel generator - the problem here is petrol going "stale" and the emissions problems with diesel. Fuel cells are very nice in terms of emissions, but expensive and the H2 usually comes from natural gas (at present).
If the design is modular (in the medium term), then any 20 Kw generator should do.

(But I still like the BEV car swapping idea which is an administrative solution, rather than an engineering one. )


For a larger passenger car, would it make more sense to have ~10kwh battery and a 20kw stack? Or does it not scale well with downsizing?

Even a vehicle the size of a bus or truck could get by with a 100kwh pack and a 20-50kw stack and have tons of range. When driving depending on where you live, upto 20% of the time your car is idling while you drive.

If they could make a 20kw stack one fifth the size of a 100kw stack, they could probably fit most everything in under the hood.

other thoughts would be: Manufactures are trying to make a one size fits all solution now, and will differentiate in the future. There might be some voltage conversion going on, where its much more efficient to have a bigger stack and more potential to be more efficient. If they are making these for the full lineup, especially bigger cars and SUVs it probably would make sense to develop 100kw stacks. (Towing up grades might use every ounce of power)

Cost isnt an issue for fleets with autonomous vehicles. Well, not for the initial purchase really, as removing the driver can save $200k a year. So well see some crazy expensive tech become rather affordable.

Dave with idling and everything else, 20kw should be adequate for most everyone. SUVs and Trucks that tow might find it lacking when they do, or someone that lives high up in mountainous terrain.

Like in my other thoughts, I just think they were trying to make a one size fits all solution for the stack, and they'll differentiate later, with different size stacks, (after they perfect it), and use battery capacity to make up for deficits now in bigger movers.


I hope that H2 develops as a safe and affordable low-zero carbon alternative for the percentage of transport or other energy requirements that can't be met currently by known non fossil sources and with enough emphasis on the safety side H2's inevitable hiccups and disasters will be kept minimal.

For cities and air pollution the payoffs are substantial but not enough to future proof without efforts in many other sectors of civics and industry. At least we as a specie have some claim to a good record of identifying the risk areas.
Whether the lever operators get the right instructions is not so clear.

"We see New York Mayor de Blasio has announced that New York will divest its city pension funds from the fossil fuel industry. As if that wasn’t a big enough moment on its own, there’s more. He also announced a lawsuit against some of the biggest oil and gas corporations for the damage caused by climate change."

"For a city as iconic as New York" (among others)"to take a public stand in such a big way shows the tide is beginning to turn."


Rejection of subsidies for coal and nuclear power is a win for fact-based policymaking
January 10, 2018 10.39pm AEDT



Hydrogen does not go stale in the tank, and looking at the specifications for acceptable leakage from CF tanks, would take around 30 years to evaporate away!

But the big win is that a BEV with a fuel cell RE remains a ZEV, so that for instance city authorities would have no concern that air pollution might go up when the battery reaches its limits.

Also since it is ZEV, it is reasonable to have a smaller battery, which might not cover all the everyday needs some days.

If it is an FCEV, so what?

But if it uses something else, you have to be much more careful about optimising the battery pack to usually cover the daily driving distance.


CheeseEater said:

'For a larger passenger car, would it make more sense to have ~10kwh battery and a 20kw stack?'

Hopefully, but it depends on the battery, and it depends on the stack.

The battery has to cope with providing more power when the fuel cell stack is short, for instance in accelerating or short hills, so would need a good power output.

And the fuel stack has to run at something like its optimum efficiency.
I don't know what the performance curves are like for current stacks, let alone for whatever are still in development for the next generation.

Something like this:

would presumably radically reduce the size of the stack needed for a given power output


' I just think they were trying to make a one size fits all solution for the stack, and they'll differentiate later, with different size stacks'

Nissan have always developed a half size stack for use as an RE as well as a full sized one.

They have chosen not to release any FCEVs in their efforts to bring to market BEVs, right now, but have kept up a fair size development effort.

And they are also putting a 10KW SOFC into the NV 200 running on ethanol for delivery work.

And Toyota simply stuck two Mirai stacks together to power the Class 8 semi they are working on.

The modular nature of fuel cells means that they are inherently scalable.

The limits, as well as cost, have been bulk durability and power.

They have all been addressed adequately, but continue to improve fast.


Hi Arnold.

My own reservations about a high percentage of renewables in the grid have in recent years mainly centred on storage.

Hydrogen in my view can provide enough 'lubrication' in the system to make it far more practical, especially in places with a long dark winter.

I will be as pleased as anyone when far better batteries arrive.

The problem is that we don't know when that will be, and both resource limits in cobalt and the high embodied energy in big batteries represent considerable obstacles.

That plus the fact that it is relatively easy to provide somewhere for a lot of folk to plug in, and darned difficult to provide somewhere for everyone to plug in mean that there is and will be a place for fuel cell vehicles too.

I imagine we are in for an exciting time and will try a lot of things out in the process of optimising the solutions, both as a system and in vehicles, with different balances of fuel cells and batteries.


Well, I have managed to access the paper, and battery enthusiasts are not going to like the assumption that battery costs will fall from $360KWH in 2015 to $165KWH in 2040


With autonomous eCabs it could be a moot point, private car ownership could wane saving individuals lots of money.


I have little faith in anyone making technology predictions 25 years out. In re-reading some of this article, I see that the group making the prediction is the DOE Fuel Cell Technologies Office (FCTO) which has a built-in bias towards supporting fuel cells as that is their main function. Always interesting to go back and read old Popular Science articles predicting the future. At one time, you might think that we would all be driving cars powered by nuclear reactors by now. In 1955, I learned to drive an Allis Chamlmers tractor so I had an interest in the 1959 Allis Chamlmers Fuel Cell tractor. Well, the short of the story is that farmers still are not using fuel cell tractors.

Even if the cost (and size?) of the fuels become reasonable, there is still overall problem with generating hydrogen. Yes, every few days there is a new development but they are mostly lower cost catalysts but there is still the problem of supplying the energy to make the hydrogen and no we do not have surplus renewable energy. We currently generate about 6.5% of our electric power from wind and solar and about another 6.5% from hydro. Hydrogen and fuel cells are a bad storage solution as the overall efficiency is about 25%. Pumped storage (hydro) would be a better solution or just release less of the existing water used for hydro.

I tried to find the source of the quote: "Predictions are hard to make, especially when they are about the future". So take your choice, Niels Bohr? Samuel Goldwyn? K. K. Steincke? Robert Storm Petersen? Yogi Berra? Mark Twain? Nostradamus? Anonymous?. Maybe it comes from a Danish proverb. Anyway, it has a certain self-evident truth.


Davemart, you mention a 30 year evaporation timeframe for H2.I re read because my understanding is @ 30days willgoa long way towards emptying the tank for liquefied H if not cryogenically cooled.

I noticed (i think it was the KenworthFC truck) that the storage tank was twice the capacity required to store the capacity of liquid H. sothe gas was @ twice the volume or presumably 1/2 pressure.
That seems to be the requirement for storage at ambient temperatures.

My other inquiries into storage informed my belief that there is no real option for storage at any useful pressure/volume for transport applications over any useful time frame unless cooling is employed that is to say virtually the entire contents will discharge via over pressure valve or it will autoignite.
That is a no compromise situation as I understand it no amount of insulation or vacuum will extend the storage without cooling beyond days to ~a week..
The only way to realise the full energy content is to utilise the ullage on a continious basis.

This can be verified by comparing the autoignition temp and pressures that can be found on the internet.

I am interested to hear more on the storage tank you mention.


Hi Arnold:

I was referring specifically to the CF tanks in cars, which store compressed gas, not liquid hydrogen.
The figures I and another guy, a battery car enthusiast, dug out for the specs for compliant tanks are what I was talking about for the very low leakage rates.

I have no idea what the figures for liquid hydrogen are, as I have not looked at them at all.


sd said:

'I have little faith in anyone making technology predictions 25 years out.'

Yep. And when I looked at the methodologies employed my fears were confirmed, as in an effort to produce well grounded figures they extend much of current practise in some respects unduly far into the future.

For instance, they use the range of the Mirai as the range in 2040, when the Nexo already greatly exceeds that.

I typically have issues with folk making straight line projections, for instance the energy density of batteries has increased on average by x percent, so extrapolating the into the future (without regard to the underlying technologies) we arrive at such and such a density by 2025 or whatever.

In my view that is a problem for present batteries, as the chemistry is running out of gas, so projections looking at 2025 are often unduly optimistic in my view.

However, extend the time horizon to 2040 and all sorts of other, much higher density technologies may have come into widespread use, and in fact if my reservations about present chemistries prove right, then it is difficult to imagine that we won't manage to crack one or the other of them.

One of the areas they rule out as too complicated to model is all sorts of hybrids.

That includes, for instance, PHEV FCEVs, which I regard as offering the best of most of the component parts.

So if you are using a battery for everyday running around, concerns about the higher cost of hydrogen become moot, and if you are using a fuel cell for longer distances, then fast charging hassles are no more.

The earliest version coming, the Mercedes, is in my view a bit of an unhappy compromise, as the tech at this stage is not really up to cramming everything we need in, but if improvements in batteries are enough to enable fully competitive BEVs without subsidy or tax exemption, then that is likely in itself to have done much of the heavy lifting to enable a PHEV FCEV, as the battery component would reduce a lot in bulk.

I am not concerned about the fuel cell component, as the technology at this stage of the curve is advancing incredibly rapidly, and the bulk of the system as well as the cost will fall big time.

So I disregard this study, as it is projecting current(ish) technology further than is reasonable into the future.

I think both batteries and fuel cells will play a big part, but likely in combination rather than opposition.

I would however have no objection to driving a lithium air battery car using through the road charging if someone turns one out!


Hydrogen is the most efficient "escape artist" among all known gaseous elements. Put it under pressure and it'll disappear evermore quickly into the surroundings. The only way to keep it in bondage is to confine it within a cryogenic tank and that is heavy and awful expensive. The only reasonable solution is to combine it chemically with atmospheric CO2 as methane similar to NG; in this form it can be stored reasonably well. But then a reformer is needed to make it compliable to FCs (reduced to H2) or use a FC that can be fed with gas directly. Everything else is daydreaming and wishful fancies. An H2 infrastructure is abhorrently expensive.



Have a look at the specifications for carbon fibre tanks,

The hydrogen in an FCEV on the road right now would take literally years to leak away.



I think that you are confusing hydrogen with helium. From the Wikipedia article on hydrogen storage:

"Compressed hydrogen storage can exhibit very low permeation."

On the other hand, helium will leak thru anything. All you can hope to do is slow it down.

There are lots of problems with hydrogen storage and fuel cells in general but high permeation is probably not one of them.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name is required. Email address will not be displayed with the comment.)