DOE releases final report from 6-year national fuel cell vehicle demo; key targets met, with twice the efficiency of today’s gasoline vehicles
18 July 2012
The US Department of Energy (DOE) released the final report from its National Renewable Energy Laboratory (NREL) for a technology validation project that collected data from more than 180 fuel cell electric vehicles over six years (early 2005 through September 2011). These vehicles made more than 500,000 trips and traveled 3.6 million miles, completing more than 33,000 fill-ups at hydrogen fueling stations across the country.
The project—the Controlled Hydrogen Fleet and Infrastructure Validation and Demonstration Project, also referred to as the National Fuel Cell Electric Vehicle (FCEV) Learning Demonstration—met its key technical targets, the NREL report said. NREL also found that these early fuel cell vehicles achieved more than twice the efficiency of today’s gasoline vehicles.
The NREL report said that baseline data from the demonstration teams in 2006 showed a range of net system efficiency from 51% to 58% for first-generation (Gen 1) systems, which was very close to the target. Second-generation (Gen 2) vehicles showed an efficiency of 53% to 59% at quarter-power, within one percentage point of the target.
In 2009, NREL expanded this CDP to include a comparison of the efficiency at full power, for which DOE’s target was 50% net system efficiency. The data show first-generation systems as having 30% to 54% efficiency at full power while second-generation systems have 42% to 53% efficiency, exceeding the 50% target.
...this project has exceeded the expectations established in 2003 by DOE, with all of the key targets being achieved except for on-site hydrogen production cost, which would have been difficult to demonstrate through this project because the hydrogen stations were not designed, constructed, and utilized as full scale, commercial stations.
—Final Report
The Learning Demonstration project began in 2004 when DOE competitively selected four automaker and energy partner teams. (Daimler and BP; GM and Shell; Chevron, UTC Power and Hyundai-Kia; and Ford and BP.) Two of the four original OEM and energy partner teams concluded their projects on schedule (based on the original five-year planned project duration) and provided their last data at the end of 2009, while two of the vehicle OEMs and Air Products extended their projects and provided data to NREL for another two years.
The final report is the first comprehensive report to include all new or updated results (40 composite data products (CDPs)) published in the last two years; it recaps the highlights from the first five years and summarizes new results from the final two years of the project.
The three primary objectives of the project were to evaluate fuel cell durability, vehicle driving range, and on-site hydrogen production cost and compare to DOE’s targets. The three high-level DOE technical targets for hydrogen FCEVs and infrastructure were:
250-mile driving range. Second-generation vehicle driving range from the four teams was 196–254 miles, which met DOE’s 250-mile range target. In June 2009, an on-road driving range evaluation was performed in collaboration with Toyota and Savannah River National Laboratory. The results indicated a 431-mile on-road range was possible in southern California using Toyota’s FCHV-adv fuel cell vehicle.
More recently, the significant on-road data that have been obtained from second- and first-generation vehicles allowed a comparison of the real-world driving ranges of all the vehicles in the project. The data show that there has been a 45% improvement in the median distance between fueling events of second-generation vehicles (81 miles) as compared to first-generation vehicles (56 miles), based on actual distances driven between more than 25,000 fueling events.
Over the last two years, NREL saw a continuation of this trend, with a median distance between fuelings of 98 miles, which is a 75% improvement over the first-generation vehicles. The vehicles are capable of two to three times greater range than this, but the median distance travelled between fuelings is one way to measure the improvement in the vehicles’ capability, driver comfort with station location and availability, and how they are actually being driven, NREL said.
2,000-hour fuel cell durability. The maximum number of hours a first-generation fuel cell stack (2003–2005 stack technology) accumulated without repair is 2,375—the longest stack durability from a light-duty FCEV in normal use published to date of which the NREL team was aware. On average, the rate of the initial power degradation is higher in the first 200 hours and becomes much lower after that, similar to the fuel cell voltage degradation.
NREL also found that stack operation of around 1,000 hours is required to reliably determine the rate of the more gradual secondary degradation. Finally, significant drops in power were observed at 1,900–2,000 hours, providing a solid upper bound on first-generation stack durability. The maximum and average projected times until 10% voltage degradation for first-generation systems were 1,807 hours (best of the four teams) and 821 hours (average of all teams).
For second-generation fuel cell stacks (2005–2007 stack technology), the range of maximum hours accumulated by the four teams was approximately 800 to more than 1,200 hours, and the range of average hours accumulated by the four teams was approximately 300 to 1,100 hours. Relative to projected durability, the Spring 2010 results indicate that the highest single-team average projected time to 10% voltage degradation for second-generation systems was 2,521 hours, with a multi-team average projection of 1,062 hours. Therefore, DOE’s 2,000-hour target for durability has been validated.
Over the past two years, additional fuel cell durability data were acquired from updated GM and Daimler vehicles (2007–2009 stack technology) during their extended projects. Because there are only two teams, it is not possible to provide both the maximum and the average without revealing the individual results of both teams, NREL said, but reported that the average projected time to 10% voltage degradation of the two teams is 1,748 hours, a significant increase over first-generation and second-generation results.
$3/gallon gasoline equivalent (gge) hydrogen production cost (based on volume production). Cost estimates from the Learning Demonstration energy company partners were used as inputs to an H2A analysis to project the hydrogen cost for 1,500 kg/day early market fueling stations. H2A is DOE’s suite of hydrogen analysis tools, with the H2A Production model focused on calculating the costs of producing hydrogen.
Results from version 2.1 of the H2A Production model indicated that on-site natural gas reformation could lead to a cost range of roughly $8–$10/kg and on-site electrolysis could lead to a hydrogen cost of $10–$13/kg. 1 kg hydrogen is approximately equal to the energy contained in a gallon of gasoline, or gallon gasoline equivalent (gge). While these project results do not achieve the $3/gge cost target, two external independent review panels commissioned by DOE concluded that distributed natural gas reformation could lead to a cost range of $2.75–$3.50/kg and distributed electrolysis could lead to $4.90–$5.70/kg.
Therefore, NREL concluded, this objective was met outside of the Learning Demonstration project using distributed natural gas reforming.
From all of the project results that NREL has generated, it is our conclusion that FCEVs have advanced rapidly in the last seven years. As the automotive OEMs and other researchers worldwide continue to focus on the remaining challenges of balancing durability, cost, and high-throughput manufacturability, we are optimistic that improvements will result in a manageable incremental cost for fuel cell technology. We therefore expect continued progress to lead to several vehicle manufacturers introducing thousands of vehicles to the market in the 2014–2016 timeframe, at which time the hydrogen community will have its first true test of whether the technology will be embraced by the public.
—Final Report
Resources
K. Wipke, S. Sprik, J. Kurtz, T. Ramsden, C. Ainscough, and G. Saur (2012) National Fuel Cell Electric Vehicle Learning Demonstration Final Report NREL/TP-5600-54860
So, H2 from CH4 reformation still spews out C unless it is captured somehow. It seems likely someone will buy a a home H2 reformer/chiller/compressor unit, given the bounty of natural gas from fracking, so then the question becomes, "What is the well-to-wheels efficiency of such a system?"
On the plus side, if there ever is a zombie apacalypse, then these vehicles could still be supplied by Mad-Max-Beyond-Thunderdome-style methane production...so there's that at least.
Posted by: HealthyBreeze | 18 July 2012 at 11:44 AM
"1 kg hydrogen is approximately equal to the energy contained in a gallon of gasoline, or gallon gasoline equivalent (gge)." But FCs use fuel twice as efficiently as ICEs. So $4.90–$5.70/kg electrolyzed hydrogen would be equivalent to $2.45-$2.85 gasoline when considering energy delivered to the wheels? Am I understanding this correctly?
Posted by: David Snydacker | 18 July 2012 at 11:45 AM
Im interrested to buy a fuelcell car but not with a tank of compressed hydrogen that you have to fill at an hydrogen station. Put a water electrolyzed into the car and we make and stock the hydrogen into the car, no need for an external pump of hydrogen and a bothersome hydrogen infrastructure.
This study forgot to take into account the latest electrolyzers with catalysts that are very efficient at doing water electrolisis.
Posted by: A D | 18 July 2012 at 11:47 AM
"Therefore, NREL concluded, this objective was met outside of the Learning Demonstration project.."
Is this the 10th or 11th year of the highly decidered Bush Hydrogen Initiative..
Posted by: kelly | 18 July 2012 at 12:55 PM
Electrolysing urine instead of water will bring down the cost of hydrogen production by at least 50% if not more. It will also get rid of a lot of liquid waste.
Posted by: Mannstein | 18 July 2012 at 01:03 PM
Five heavy beer drinkers/passengers (1 L each/30 minutes) could produce almost all the urine required to keep a small FC powered car going at 100 Km/hr? Home made brew may not cost much more than gasoline.
Posted by: HarveyD | 18 July 2012 at 02:25 PM
@ David Snydacker,
I don't think $ and Kilowatt Hours are interchangeable.
Besides, if you look at $, then you're looking at total cost per mile, which has to include FCV cost amortized over vehicle life. I'd think the ~2,400 hours they forecast from the later generation of FC in their study would do maybe an average of 35 miles per hour lifetime, so maybe that's 84,000 miles. Does anyone see an FCV costing less than $42,000 and adding 50 cents per mile, plus maintenance any time soon?
Posted by: HealthyBreeze | 18 July 2012 at 06:29 PM
This strains credibility.
Is it April 1st? . . NO ! !
So after:
• more than 180 fuel cell electric vehicles
• more than six years
• more than 500,000 trips
• 3.6 million miles
• more than 33,000 fill-ups
it is their conclusion that they:
• believe FCEVs have advanced rapidly in the last seven years
• are optimistic that improvements will result in a manageable incremental cost for fuel cell technology
• expect continued progress to lead to several vehicle manufacturers introducing thousands of vehicles to the market at which time the hydrogen community will have its first true test of whether the technology will be embraced by the public. [Clearly this bears little resemblance to a true test.]
Then they proudly state; ” ...this project has exceeded the expectations . . . with all of the key targets being achieved except for on-site hydrogen production cost, which would have been difficult to demonstrate through this project because the hydrogen stations were not designed, constructed, and utilized as full scale, commercial stations.”
But that’s the only thing not already WELL known.
But they are right as to why this “validation project” and report are a HUGE waste of money.
And so how well did they do on the three primary (key) objectives of the project:
• To evaluate fuel cell durability
o The multi-team average projection of 1,062 hours validated DOE’s 2,000-hour target for durability? ?
• vehicle driving range,
o driving range from the four teams was 196–254 miles, which met DOE’s 250-mile range target.?
• on-site hydrogen production cost
o “project results ‘could lead to’ a cost range of $8–$10/kg vs. the $3/gge cost target”. Umm WHAT?
Posted by: ToppaTom | 18 July 2012 at 08:06 PM
@TT,
A HUGE waste of money, Eh?
If you are an engineer involved in product development and testing, you would quickly understand that during those thousands of hours and hundreds of thousands of miles of actual FCV's vehicular testing, they must have discovered thousands of problems and potential flaws and weakness of their design and manufacturing, as with any products under development, so that subsequent products would be much better, and would have avoided a lot of costly recalls later on. There is no substitution for these kind of extensive testing for any new kinds of product about to enter the market.
1) As to the durability of these fuel cell stacks,
"Relative to projected durability, the Spring 2010 results indicate that the highest single-team average projected time to 10% voltage degradation for second-generation systems was 2,521 hours.... Therefore, DOE’s 2,000-hour target for durability has been validated."
I think that the English is clear enough here. 10% degradation of voltage does not mean the end of life for the stack, just means that power will be reduced a little bit.
2) Driving range: 196-254 miles...clearly arrived at the 250-mi goal. Toyota's Advanced FCV achieved 431-mile range.
3) On-site H2 production cost: "two external independent review panels commissioned by DOE concluded that distributed natural gas reformation could lead to a cost range of $2.75–$3.50/kg and distributed electrolysis could lead to $4.90–$5.70/kg."
Count on my analysis an the third external independent source: 33kWh/kg of H2 divided by 0.78 as efficiency of the latest high-pressure electrolyzer = 42 kWh/kg of H2 x $0.05/kWh of off-peak electricity = $2.11 per kg of H2 for raw electricity cost. Now, feel free to add the amotized cost of the electrolyzer, the container and profit to bring the final cost per kg of H2 to $3-4/kg...depending on volume production of electrolyzers, containers and dispensers...the cost of these H2 dispensers will undergo 5-10 folds reduction in prices from early introduction to final mass production and higher market penetration.
Posted by: Roger Pham | 18 July 2012 at 10:33 PM
@Roger,
Do you have a link confirming the 42 kWh/kg number of a high pressure electrolyser?
Posted by: Arne | 18 July 2012 at 10:54 PM
No, that's just my estimation.
Posted by: Roger Pham | 18 July 2012 at 10:59 PM
Roger,
The best number I have seen for real-world non-paper technology is 60 kWh/kg for the Avalence high pressure electrolyser. (see page 9 in this document).
Until better technology is out there, that's the number that I use. Combined with the 60 miles per kg consumption of the Honda FCX Clarity, that results in 1 mile per kWh, about 3x less than an EV. Hydrogen cars have still a lot of catching up to do before being able to compete with EV's.
Range and refuelling times are important aspects and the H2 car seems to have the upperhand there (at least for now). But efficiency is king imo.
Posted by: Arne | 18 July 2012 at 11:43 PM
Good point, Anne.
I erroneously used the Low Heating Value of H2 at 33kWh/kg in the calculation, instead of using the High Heating Value of 39kWh/kg. So redo the calculation: 39kWh/kg divided by 0.78= 50 kWh/kg of input electricity. The 60kWh/kg stated in your reference dated back to 2004. Technology has improved since then.
50kWh/kg input electricity x $0.05/kWh = $2.5/kWh for raw electricity cost. This article is not about comparative efficiencies between BEV and FCV, but rather about practicality of H2-FCV and affordability of H2 as fuel, which have been demonstrated.
Talking about efficiency, H2-FCV can achieve almost 60% system efficiency. This, in comparison to about 20% efficiency of conventional ICEV, or 30-33% for HEV's, means that when H2 is at cost parity to current gasoline prices, the fuel cost per mile will be about 1/3 to 1/2 that of current ICE technology. That will make H2-FCV quite an attractive way to run on renewable energy, which is getting cheaper as the cost of solar PV's and wind turbines are continously declining. In fact, the cost of solar PV electricity is at near parity with coal-fired electricity. A solar farm dedicated to feed H2 electrolyzers does not need to be grid-tied and can be much cheaper and more efficient because the DC output does not have to be inverted into AC-grid-compatible output.
Posted by: Roger Pham | 19 July 2012 at 01:10 AM
I know that it is from 2004. But if newer, better technology exists, you should be able to show it. So, do you have a link to confirm this? You saying so is not evidence. I need independently verified numbers. Sorry.
Based on methane reformation or electrolysis? Or other technology that I need to be educated about?
You know that efficiency calculations based on a methane source of the H2 are meaningless, since it is a fossil fuel, iow a limited resource. But it is temporarily acceptable if carbon capture is implemented. If so, does your 60% figure account for the energy losses from carbon capture?
Posted by: Arne | 19 July 2012 at 02:20 AM
Roger, one more thing:
Large grid inverters are >98% efficient. The world record stands at 99%
Even the newest consumer level inverters are above 98% average efficiency. For example: Stecagrid 3000, 3600 and 4200
I don't see enough room for improvement to make your argument a valid one.
Posted by: Arne | 19 July 2012 at 02:28 AM
Roger:
I've been hunting around, but I can't find current efficiencies as high as 78% in current practise - perhaps you have a link?
We also need to include the energy for compression too, so perhaps the discrepancy lies there.
I think though that we should use best practise, as production fuel cell cars will presumably hit or exceed that, and the current leader is the Hyundai Tucson, which does 72 miles/kg, up from Anne's 60 miles/kg for the Clarity.
It is also really all a bit premature to worry too much about current efficiency from electrolysis, both because it is improving fast and because in practise for the time being most hydrogen is likely to come from reforming natural gas, just as most of the electricity for BEVs does.
On those terms we can hit around a 70% efficiency for reforming and compressing hydrogen from NG, which compares very favourably with generating electricity.
We simply do not need to solve all problems right away before making any progress.
If we must talk about theory though, then hydrogen production compares very favourably with producing electricity directly, as resources such as stranded wind can be made use of, which otherwise cannot be employed.
Posted by: Davemart | 19 July 2012 at 02:39 AM
The DOE seem to use the LHV of hydrogen in their calculations:
'Electrolyzer efficiency is 74.6% based on hydrogen’s lower heating value (LHV); it uses 44.7 kWh per kg of hydrogen produced.'
http://hydrogen.energy.gov/pdfs/10001_well_to_wheels_gge_petroleum_use.pdf
(Last page)
Posted by: Davemart | 19 July 2012 at 04:15 AM
Davemart,
The 72 mi/kg, is that according to Hyundai or the official EPA rating, which I could not google. Maybe you have a link for independent verification of that?
I did find this document. The conclusion says:
According to the Wikipedia page the process is 65% to 75% efficient. But these numbers are well-to-h2 not well-to-wheel. When doing a direct comparison, remember that modern CCGT plants are very efficient, around or slightly above the 60% mark. And they have the added advantage of being very usable for balancing fluctuating renewable energy generators, something that a hydrogen plant can not do.
You do not care about efficiency since this is only a short term worry. My worry is climate change and we have to get off fossil fuels asap. Introducing a new technology based on continued use of fossil fuels is not a smart way forward. Especially knowing that these investments have a tendency to persist for a very long time. Investors want to earn their money back and as soon as the hydrogen factories are built and H2 vehicles are on the road, inertia will keep it going for longer than we want. Just look at how difficult it is to get rid of our oil addiction.
I see you still cling to your fantasy of using 'stranded wind' for hydrogen production. 'Summer PV' would be a much better source, but never mind. Either way, I can not see the economics working out. Having a plant on standby 365/24 for switching on only under stormy conditions does not seem at all an attractive proposition, economically speaking.
Posted by: Arne | 19 July 2012 at 04:46 AM
Here is an efficiency calculation by Norsk Hydro, rather old now:
'Norsk Hydro Electrolysers (NHE) is today a leading
producer of alkaline electrolysers. Some of NHE’s
electrolysers have an efficiency of over 80% (high heating
value). Efficiency is an important factor in electrolysis
because the use of energy (~4.5 kWh/NM3H2) makes
up a significant portion of the costs at an electrolysis plant.'
http://www.bellona.org/filearchive/fil_Hydrogen_6-2002.pdf
Posted by: Davemart | 19 July 2012 at 04:50 AM
That is uncompressed hydrogen. You must add the energy for compression.
Posted by: Arne | 19 July 2012 at 04:51 AM
@RP,
Yes, A HUGE waste of money.
If you were a development engineer in the private sector you would understand that those thousands of hours and “hundreds of thousands of miles” (no 3.6 million miles) of actual FCV's vehicular testing, is of limited use because the problems, flaws and weakness relate to the design and manufacturing of development class FCV's, has very limited relevance to subsequent production vehicles.
They would differ substantially and have different flaws and weakness. Are these even prototypes or development level vehicles? Even thousands of miles is more than this stage warrants.
There is no substitution for extensive testing of production configuration vehicles.
1) As to the durability of these fuel cell stacks; with production (and final configuration) years away, lab testing is comparable to road testing. I admit a few miles of road testing is warranted. But only the gov would just do it the $imple way - 3.6 million miles.
3.6 million miles – that’s obscene.
A HUGE waste of money.
Maybe lab facilities were inconvenient or limited but money was not.
2) Driving range: You say “196-254 miles...clearly arrived at the 250-mi goal.” But since you say "Toyota's Advanced FCV achieved 431-mile range" why drive 3.6 million miles and get 196-254 miles?
A HUGE waste of money to conclude there are too few refueling stations today.
3) On-site H2 production cost: you say "two external independent review panels commissioned by DOE concluded that distributed natural gas reformation could lead to a cost range of $2.75–$3.50/kg ”
Now THAT study probably took a matter of hours or days and cost next to nothing. And even you say “Count on my analysis an the third external independent source”.
So what was the cost of:
• more than 180 fuel cell electric vehicles
• more than six years
• more than 500,000 trips
• 3.6 million miles
• more than 33,000 fill-ups?
Apparently it does not matter.
To be fair, it actually does NOT matter to the government, and relative to other gov programs maybe this was not
A HUGE waste of money.
Just more of the same. It’s not their money.
Posted by: ToppaTom | 19 July 2012 at 05:20 AM
Electrochemical compression to 12,000 psi takes around 3kwh/kg:
http://www.hydrogen.energy.gov/pdfs/review12/pd048_lipp_2012_o.pdf
For the most efficient other compression technology, you use around 7kwh.
In these cases on an LHV basis the total energy is including production comes to 47.7kwh/kg to 51.7kwh/kg.
Posted by: Davemart | 19 July 2012 at 07:05 AM
You are not going to get an EPA rating on a car which is not in production yet, so of course the 72mpg figure is Hyundai's.
Link:
http://www.greencarcongress.com/2010/03/hyundai-fcev-20100304.html#more
No doubt in practise mileage will be somewhat less than that, like every other car, ICE or battery.
It still boils down to moving a big, heavy, not very aerodynamic cars quite a way on the 5.6kgs of hydrogen it holds.
It has been tested by journalists, and none of them are saying they ran out:
http://www.caradvice.com.au/124489/hyundai-ix35-fcev-review-first-drive/
Posted by: Davemart | 19 July 2012 at 07:09 AM
@Anne:
A fantacist is someone who sticks solar panels on their roof in northern Europe, putting much of the cost on the bills of other people including the poor, and erroneously claims to run their car on this, when in fact no such thing happens as the car is used in the day, and the power provided to the grid in exchange is more or less worthless as it is provided when it is of least use!
I mention stranded wind in deference to eco-loons who engage in this sort of behaviour, as in comparison stranded wind to hydrogen is sane.
What I would do is simply use nuclear reactors to provide both electricity and hydrogen, and providing hysterics are not listened to it is safe.
Estimates put the death toll from evacuations at Fukushima at 600, all of which rational risk assessment could have avoided.
Even under the absurd notion of linear no threshold for radiation risks they are going to be indetectably low.
Posted by: Davemart | 19 July 2012 at 07:17 AM
Here is an analysis of hydrogen storage for wind:
http://205.254.148.40/hydrogenandfuelcells/tech_validation/pdfs/34364.pdf
Perfectly batty in my view, but not nearly as batty as solar pv in northern Europe.
BTW, in this document they use a value of 1kg of hydrogen is 11.1NM3H2, which on the Norsk Hydro figures I quoted earlier comes out to almost exactly 50kwh/kg before compression.
Posted by: Davemart | 19 July 2012 at 07:40 AM