CARB's Fuel Cell Detour on the Road to Zero Emission Vehicles
By Alec Brooks
On April 20, 2004, California governor Arnold Schwarzenegger signed an executive order aimed at creating a network of 200 hydrogen refueling stations throughout the state by 2010. The goals for hydrogen powered transportation sound good on paper – zero emission vehicles propelled by hydrogen produced with clean renewable energy.
It sounds almost too good. Is the governor’s Hydrogen Highways program really the bold step that jump starts the future of transportation, or is it a monumental boondoggle that leads to a dead end? The conventional wisdom is that the former is true, but a growing body of analysis and evidence leads instead to the dead end conclusion.
The California Air Resources Board (CARB) has recently made a U-turn in their zero emission vehicle (ZEV) program just as everything was coming together. Battery electric vehicles had been improving rapidly and were almost universally loved by their drivers. Battery durability was shown to be better than expected and cost parity with conventional gasoline powered vehicles was clearly attainable.
By dithering on the ZEV mandate and putting all their chips on fuel cell vehicles, CARB has set back the development of practical zero emission vehicles by at least a decade. The current situation has much similarity to the California energy deregulation fiasco of the 1990s.
California is again being led down the primrose path by automakers, with the promise of a future vehicle vision that on the face of it, sounds great. In the 1990s, California was manipulated by energy companies; now it is the automotive companies’ turn. By convincing CARB that fuel cell vehicles will be the ultimate future solution, they have bought ten or more years to avoid bringing real zero emission vehicles to market.
Support for fuel cell vehicles derives primarily from conventional wisdom and commonly accepted facts about these vehicles. These “facts” have been repeated so often that they have become generally accepted as the truth. But it turns out that many of the facts are not supported by the data; much of what is universally accepted about hydrogen and fuel cell vehicles is, in fact, wrong.
This article documents my long, and so far unsuccessful, effort to get CARB to pay attention to the real and serious issues with hydrogen and fuel cell vehicles rather than simply jumping to a conclusion. I felt that there needed to be real discussion and analysis of whether hydrogen powered vehicles were what the state really should strive for in the near term or even in the long term. What started out as my brief reply to CARB’s reply to my original written comments has expanded into a very lengthy article. The length is needed to provide the relevant background, to give a detailed response to CARB’s comments, and finally, to make recommendations as to what should be done to fix the mess CARB has made.
In condensed form, here are some of the key highlights of this story:
- The commonly held belief is that fuel cell vehicles will have two to three times the fuel economy of gasoline powered vehicles. But so far, fuel cell vehicles are losing. The mid-sized gasoline powered Toyota Prius gets 13 percent better EPA fuel economy than the subcompact Honda FCX fuel cell vehicle.
- Fuel cell vehicles are energy pigs. Fuel cell vehicles that operate on hydrogen made with electrolysis consume four times as much electricity per mile as similarly-sized battery electric vehicles.
- Battery electric vehicles powered by today’s available battery technology can have twice as much driving range as current fuel cell vehicles.
- Fuel cell vehicles will inconvenience their drivers with more than ten times as much time devoted to refueling compared with battery electric or gasoline powered vehicles.
- CARB chairman Alan Lloyd hung battery electric vehicles out to dry in reaction to auto company threats to pull fuel cell vehicles out of California.
- As they were publicly trumpeting the arrival of the second generation EV1, General Motors privately vowed never to produce any more of them.
For the last several years, CARB has grappled with amending the ZEV regulations to address multiple lawsuits brought by automakers. On January 9, 2004, the final regulations were published, along with the FSOR, or Final Statement of Reasons. The FSOR is a document in which CARB is required to note all substantive comments and recommendations submitted to them during designated comment periods, and to explain how that information was considered in the rulemaking process.
On February 20, CARB issued a supplement to the FSOR to the California Office of Administrative Law, the agency that is charged with making sure that state agencies like CARB follow the proper rulemaking procedures.
The supplement covered two items: a request by Volkswagen for relief on a specific aspect of the regulation, and a response to comments that I had made at the December 2002 ZEV workshop (and subsequently submitted as written testimony prior to the March 2003 board meeting). CARB’s official explanation of how this last minute response to my comments came into being is misleading if not outright false. The FSOR supplement itself is a slipshod job, and should be an embarrassment to CARB. This is the story of how CARB came to issue the response to my comments and what is wrong with what they said in it.
I have been following the ZEV mandate since it was adopted at the November 1990 CARB meeting, at which I testified. In 1990 and 1991, I was a participant in the key design reviews for the EV1 as it was being developed at GM, and witnessed first-hand the trials and tribulations in getting that vehicle to market. The EV1 development team at GM was by and large incredibly dedicated to putting out a quality product. I leased an EV1 in 1996 on the first day they were available. The EV1 that I drove for more than three years had its share of glitches, but the overall experience was quite positive. I continued to closely follow GM’s efforts (or lack thereof) to bring EVs to market, and CARB’s seemingly continuous revisions to the ZEV program. I became very involved in the 2000 – 2002 review of the ZEV mandate. I testified at nearly every public workshop and board meeting and met several times with the chairman and senior staff.
By about 2000, it appeared that CARB had waffled so much on the mandate that car makers were emboldened to cancel their EV programs. Instead of EVs, they urged us to wait while they developed the silver bullet vehicle of the future, the fuel cell vehicle. It has been almost universally reported that fuel cell vehicles would be two, three, or even four times as efficient as conventional vehicles, and that making hydrogen with clean renewable electricity would finally end our oil imports and give us a wonderful and sustainable solution to our transportation needs.
In the 1990s, fuel cell vehicles started replacing battery EVs as the next big thing. Battery EVs were said to be a mature technology that wouldn’t improve much more, and in small volume, they cost much more to produce than a conventional car. Fuel cell vehicles were touted as the solution to all the ills of battery electric vehicles. There would be no more range problems. They could refuel in a jiffy, and they would emit only pure water from the tailpipe. They sounded almost too good to be true.
My first exposure to fuel cell vehicles was in 1990 when I was contacted by "Dr." Roger Billings, the (self proclaimed) guru of hydrogen. He claimed that he had developed the first fuel cell vehicle with financial support from the Pennsylvania Energy Office. But his claims sounded suspect. I spoke with an official at the Pennsylvania Energy Office about the project and found that they couldn’t actually say whether the vehicle was powered by a fuel cell or the batteries onboard. The ‘fuel cell’ in the vehicle was a shiny metal box with a few pipes and wires, but no evidence of a cooling system or other usual subsystems. This experience instilled in me a healthy skepticism about fuel cell vehicles that has served me well over time.
At an EV industry conference in Atlanta in 1999, I saw a fuel cell vehicle up close for the first time – the Ford P2000. It was on display outside of the conferences’ gala evening event. The engineers babysitting the car fired it up and had it ‘idling’. It sounded like a lot of commotion – pumps, fans, blowers, etc, for a car that was, after all, just sitting there doing nothing. I asked the engineer with the car how much electricity (being produced by the fuel cell) was being used to run all those devices. The answer surprised me – it was 3 kilowatts! By contrast, the standby power of an EV is about a tenth as much. I knew that the average power requirement to drive a similar-sized EV over the EPA urban driving cycle was just 4 kW. Here was the fuel cell vehicle just sitting there using 3 kW! I asked about driving range, and at first got the stock answer – something like 100 miles. But I asked again about the city cycle range, and didn’t get an answer, except an impression that it was a lot lower.
At the 2001 Environmental Vehicles Conference in Detroit, I test drove Ford’s Focus FCV fuel cell vehicle. Again, I was struck by the cacophony of noises when the car was sitting there turned on but otherwise doing nothing. The engineer hosting the test drives confirmed that 3 kW number again. The driving experience was poor by comparison to what an EV driver might have come to expect. Accelerator pedal response was slow as the air pumps whined up to speed, and there was also an annoying lag when releasing the accelerator pedal; power continued to flow to the wheels for about a half-second after it was no longer being requested. I also got the sense from the test drive host that the range was not very good.
I also started to look at the fuel side of the picture. At present, most hydrogen is produced from natural gas, resulting in CO2 emissions of about 12.5 kilograms per kilogram of hydrogen produced. For transportation applications it turns out that it will be more efficient and more practical to use natural gas directly in natural gas powered hybrid vehicles, rather than first converting the natural gas to hydrogen and then using the hydrogen in a fuel cell. The emissions of natural gas vehicles are so small that they are considered for practical purposes as clean as EVs, as evidenced by their allowance into California carpool lanes with a single occupant. The only other vehicles with this privilege are zero emission vehicles – battery and fuel cell powered. The long term goal often stated for hydrogen is to produce it by electrolysis, with renewable electricity, such as wind or solar. There are, of course, other uses for renewable electricity, such as offsetting coal or natural gas based electricity generation. So it is vitally important to understand how much electrical energy would be required to make hydrogen for fuel cell vehicles, and to compare that with the electricity requirements of battery electric vehicles.
I wanted to find out the efficiency of the electrolysis process that makes hydrogen by splitting water. At an industry conference, I went to the Stuart Energy Systems booth, where a prototype ‘personal fueller’ electrolysis device was on display. The idea was that you’d make hydrogen in your garage by connecting the personal fueller up to your 240V power and the water supply. I asked the company rep in the booth what the efficiency of the device was. I was assured that the personal fueller was extremely efficient, greater than 99 percent! I didn’t believe it and decided to try a different approach. To avoid any confusion about what people defined as efficiency, I’d instead try to find out how much electricity, in kWh, was needed to produce and compress one kg of hydrogen (which is the energy equivalent of one gallon of gasoline).
I sent an email to Stuart Energy Systems asking for this information. It took three tries and 5 months to get the answer: 55 kWh. Subsequent data from Stuart’s data sheets for their large "community fueler’ product places the number at 63 kWh. This is a pretty big number. Other companies’ electrolyzers are just as bad or even worse. Honda’s electrolyzer installed at their research center in Torrance California consumes 64 kWh/kg, and Proton Energy’s is up at 72 kWh/kg. If one kg of hydrogen is processed through a fuel cell at an average efficiency of 50 percent (a generous assumption given today’s FCVs), the electricity you get back out will be 16.6 kWh, for a round trip efficiency of only 23 to 26 percent from the electricity consumed by the electrolyzer. Or put another way, you have to put in about four times as much energy as you get back out; three fourths of the input energy is wasted before any of the input energy reaches vehicle’s drive motor! By contrast a lithium battery and good battery charger could return about 51 kWh of that same 63 kWh input, over three times as much energy back out for the same amount put in. So right away, an EV has a three fold energy advantage over an FCEV.
But there is more: a fuel cell is a one-way device; it produces electricity from hydrogen and oxygen. A battery is a two-way device, it can accept power as well as deliver it. This is a crucial difference. The power requirements for a vehicle are highly bi-directional. Positive power is needed to accelerate, go up hills, and maintain speed. Negative power is needed to slow down or go down hills. This negative power represents recoverable energy that can be used again to drive the vehicle. With conventional cars, the negative power is thrown away in the friction brakes or in engine braking. With battery electric vehicles, much of that energy is recaptured in the battery through regenerative braking. With fuel cell vehicles, since the fuel cell itself can’t accept that energy, a separate energy buffer is sometimes employed. Honda uses a supercapacitor bank. Toyota uses a Prius battery pack. GM has nothing to absorb this energy; it is wasted as heat in the friction brakes. But the Honda and Toyota approaches are limited in the energy and power that can be reabsorbed, reducing the recapture of energy compared to battery electric vehicles.
The combination of the electrolyzer energy consumption numbers, the big difference in standby power, and the inferior regeneration energy capture all suggested to me that fuel cell vehicles would use many times more electricity per mile than battery electric vehicles.
I started to search for data on the actual fuel economy and driving range of fuel cell vehicles. I discovered that the data was usually given by vague statements, such as "range of up to 150 miles". Data for standardized test cycles or real world driving was never quoted. It seemed like it was a big secret. At the 2001 Michelin Challenge Bibendum, a competition for environmentally friendly vehicles, the participation of the fuel cell vehicles from the California Fuel Cell partnership came with the proviso that actual hydrogen consumption numbers and driving range would not be disclosed as part of the results. This seemed strange for a competition that was all about energy efficiency.
There were scattered bits of information that hinted of a range problem. In their "highlights of 2001" report, the California Fuel Cell Partnership touted 754 refueling events during the year, with 34,000 vehicle miles covered – an average of only about 45 miles between refuelings. In the Fuel Cell Partnership’s "Rally Through the Valley" drive from Sacramento to Los Angeles, there were a lot of refueling stops. Some of these stops coincided with public events and test drives. But on the last day of the rally, the driving leg was from Bakersfield to Los Angeles with no events scheduled along the way. The vehicles were to travel directly to Griffith Park in Los Angeles for an afternoon test drive event. The distance from Bakersfield to Griffith park is just 105 miles, but there were two refueling stops along the way: one at the base of the Grapevine, and the other at Castaic. Driving range appeared to be a big problem but was being kept very quiet.
In late 2002, the EPA came through with official fuel economy test data for the Honda FCX fuel cell vehicle. The EPA fuel economy guide for the 2003 model year listed only two zero emission vehicles. One was the battery electric Toyota RAV4EV, and the other was the Honda FCX fuel cell vehicle. These vehicles are about the same size and utility, although the FCX has only two doors, and is about 200 pounds heavier than the RAV4EV. At last, here was the data for an apples-to-apples comparison of the energy consumption of fuel cell and battery electric vehicles tested over the same driving cycles. The published results confirmed what I had expected. The FCX was rated at 50 miles per kg of hydrogen (combined city / highway), and the RAV4 EV was rated at 30 kWh / 100 miles, or 0.30 kWh/mile. With the electrolyzer energy consumption number that I had at the time – 60 kWh/kg – the FCX would need 1.20 kWh of electricity (to produce hydrogen) per mile. This is fully four times as much as the RAV4EV’s 0.30 kWh/mile. To put this into perspective, this is the same ratio as between a Hummer and a Honda Civic! (But unlike the comparison of the Hummer to the Civic, in this case the vehicles are about the same size. The fuel cell vehicle was an energy pig!)
As I read more and more gushing articles about fuel cell vehicles, it became apparent to me that I had never seen these basic energy comparisons with battery electric vehicles expressed anywhere. Every time fuel cell vehicles were mentioned, it had always just been accepted as fact that they were going to be wonderfully efficient.
In 2000, the energy efficiency guru Amory Lovins came to AC Propulsion, where I was working at the time, for a short visit. At the end of the visit, I drove him to the Los Angeles airport in a prototype 4-door electric vehicle. It wasn’t a particularly lightweight vehicle, having a large liquid-cooled nickel-metal-hydride battery pack. The trip was on the freeway in heavy traffic with lots of speedups and slowdowns. Lovins was very interested in the digital displays in the vehicle that tracked battery Amp-hours and kiloWatt-hours (kWh). The displays showed both net values and cumulative energy captured by regenerative braking. In AC Propulsion’s electric vehicles, virtually all vehicle slowing in normal driving is accomplished with regeneration, leading to large amounts of energy returned back to the battery.
During the drive Lovins confidently told me all about his fuel-cell-powered hypercar concept, and rattled off a string of very optimistic efficiency numbers which he multiplied together to get an astounding overall efficiency. In addition to his efficiency numbers all being on the high side of reasonable, he had forgotten to include the difference between the higher and lower heating value of hydrogen, a common mistake. The higher heating value represents the energy of the hydrogen-oxygen reaction if the water byproduct of the hydrogen-oxygen reaction is accounted for as condensed liquid water. The lower heating value is the energy of the reaction if the water is in the vapor state. The difference between the two is the heat of vaporization of the product water. Electrolyzers often have efficiency quoted relative to the higher heating value of the hydrogen produced, while fuel cell efficiencies are usually referenced to the lower heating value. The difference between the two for hydrogen is about 15%.
After dropping Lovins off at the airport, I drove home and recharged the car overnight on my separately-metered EV charging circuit. The data displays in the car had indicated 201 Wh/mile DC energy consumption over the 80.8-mile trip. Energy recovered from regenerative braking tallied up at 5.27 kWh, or 65 Wh/mile; without regeneration, the energy consumption would have been 32 percent higher. The energy used to recharge the vehicle back up to full was 26.5 kWh at the AC meter, or 328 Wh/mile. The inefficient liquid-cooled battery resulted in quite a bit of extra energy consumption compared to the DC value, but still, 328 Wh/mile is a respectable number for a four door EV. A lithium-battery-powered vehicle of similar capability would have used only about 250 Wh/mile AC on the same trip. I sent Lovins a follow up email to note that with the amount of hydrogen that could be produced with that same 26.5 kWh of electricity, his futuristic hypercar concept, even if it was as efficient as he claimed it would be, would not have been able to travel as far as the here-and-now EV.
In the summer of 2003, I attended the Future Car Congress in Washington DC, sponsored by the Department of Energy and the Society of Automotive Engineers. In the keynote session, Toyota’s fuel cell vehicle executive engineer Norihiko Nakamura candidly summed up the cost problem facing fuel cell vehicles:
"If a certain level of mass production can be achieved the cost should be dropped drastically. But a great amount of effort is needed to bring the cost to even two to three times that of a standard vehicle." (emphasis added)
The Future Car Congress had a ride and drive event to provide the opportunity to drive a variety of vehicles, including hybrids, fuel cell vehicles, a battery electric, and a diesel. During the course of the day, only the fuel cell vehicles had to leave the ride and drive to go refuel. Not even the battery electric RAV4EV had to go recharge.
At the September 7, 2000 CARB meeting, the board strongly reaffirmed their commitment to the ZEV program. The board directed the staff to work out the details and bring back a new ZEV rule for a vote in January 2001. At the January meeting, the board actually boosted the number of ZEVs that would have been required under the staff proposal, and approved new rules that would require thousands of battery EVs starting in 2003. In these new rules, CARB started to show a favoritism to fuel cell vehicles. In the early years of the new proposed rules, battery electric vehicles were to be awarded with 8 or 12 ZEV credits, while fuel cell vehicles would receive a whopping 40 ZEV credits. This seemed backwards, given all the apparent disadvantages of fuel cell vehicles compared to battery electric vehicles. It seemed very premature to declare that fuel cell vehicles were preferable over battery electric vehicles. There should be analysis, discussion, and informed debate as to whether fuel cell vehicles were even a good thing to strive for, even in the long term.
A subsequent set of proposed changes also included an "alternate compliance path" which would allow automakers to meet all of their pure ZEV requirements by producing just a small number of fuel cell vehicles for demonstration programs. These vehicles would not even have to be sold in California. (In fact, they wouldn’t have to be sold at all – just placed for a limited amount of time in demonstration programs). This was a depressing turn. Here was CARB planning to put all their chips on fuel cell vehicles – vehicles that used four times the energy of battery electrics when the hydrogen is produced by electrolysis, had lower range than battery electrics with current technology, and didn’t appear to have any hope of ever being cheaper to manufacture than battery electric vehicles. It defied logic that they were going down this path. Maybe they didn’t know the real numbers.
In order to stimulate discussion of the relative merits of fuel cell and battery electric vehicles, I decided to make a detailed presentation comparing the two vehicle types at the CARB public workshop on December 5, 2002. The purpose of the workshop was for CARB to capture public feedback on their proposed new ZEV rule structure. I prepared a written document to go with my presentation. It was long – I expected it would take 25 minutes to present, which is much longer than is typically allotted. The workshop was well attended by CARB staff and representatives from the automakers and many EV drivers. Board Chairman Alan Lloyd was present. The organization of the workshop was a little informal, with no assigned order of speakers. I knew from attending past workshops that the attendance tails off as the day goes on. I secured a speaking slot in mid morning, and launched into my presentation hoping that I wouldn’t be cut off for going on too long. I stuck to reading from my prepared text and went along at a fast clip and got all the way through. The response from the audience was very positive, even from Tom Cackette, deputy executive officer of CARB, who happened to be sitting next to me in the hearing room. I later got several very complimentary emails from people that that seen the webcast or read the online printed copy.
A recording of the presentation is online at http://www.socalev.com/Sounds/CARB-AM08.mp3 and the written version titled "Perspectives on Fuel Cell and Battery Electric Vehicles" is at: http://www.arb.ca.gov/msprog/zevprog/2003rule/2003rule.htm
The main points of my presentation were:
- Fuel cell vehicles have significant range problems. Battery EVs with modern batteries can easily have twice the driving range of fuel cell vehicles.
- Fuel cell vehicles fuelled by hydrogen made by electrolysis will use four times as much electricity per mile as a similar-sized battery electric vehicle.
- Using natural gas to make hydrogen for fuel cell vehicles will consume as much or more natural gas per mile as just burning the natural gas directly in a ICE hybrid vehicle. The higher energy per unit volume of natural gas would provide more than twice the driving range as a fuel cell vehicle given equal tank volume and pressure.
- CARB’s electric vehicle cost assumptions were wrong. New cost data was presented for lithium ion and sodium nickel chloride batteries. The Prius was cited as an example of what a battery EV would cost to manufacture in volume.
- CARB was not being realistic about the cost challenges facing fuel cell vehicles. It was noted that Toyota said they’d have to work very hard to get the cost of a fuel cell vehicle down to double or triple that of a conventional car.
After the presentation I felt that maybe I had made a difference and that CARB would pay attention to what I had said. Previously, Alan Lloyd and some of the CARB executive staff had sought out my ideas and advice on changes they were considering for the mandate. But the only follow up I had from CARB was from their staff attorney who asked me why I thought their approach with multiple categories of ZEVS was an invitation to another lawsuit. But I heard nothing about any of the rest of what I had said. I had expected at least some follow up questions. But there was just nothing. At a subsequent fuel cell vehicle event, I asked a CARB staffer whether CARB was going to acknowledge my comments and respond. I was told that I’d have to wait for the final statement of reasons document at the end of the rulemaking process. I was told that CARB was required to respond in that document.
blog comments powered by Disqus