Hell and High Water in Buffalo News

On the problem with hydrogen cars:

A June 30 letter stating that we should consider using hydrogen fuel as an alternative to ethanol is misleading. My opinion on hydrogen fuel car is based on the book “Hell and High Water” by Joseph Romm, an MIT-trained physicist who managed energy-efficiency programs in the U.S. Department of Energy during President Clinton’s administration.

According to Romm’s analysis, the math for hydrogen cars simply doesn’t work out. There are two ways to create hydrogen, extract it directly from fossil fuels or split water molecules using electricity. For cars, the latter is the better approach.

A vehicle needs about one megawatthour of electricity to produce enough hydrogen fuel to travel about 1,000 miles. In order to produce the electricity needed, it will burn coal and as a result, will produce about 2,100 pounds of carbon dioxide. Driving the most efficient car (40 mpg) for 1,000 miles will produce an additional 485 pounds of carbon dioxide.

A vehicle powered by hydrogen fuel cells would indirectly create four times the carbon dioxide emissions of today’s most efficient gasoline cars. We need to invest in cleaner energy technologies and energy-efficient programs such as hybrid cars, wind and solar plants.

7 Responses to Hell and High Water in Buffalo News

  1. Lou Grinzo says:

    Let me second the opinion expressed above. As I said on my own site, even if you’ve already concluded hydrogen won’t have an impact on transportation in the near or mid-term, you should still read the book, as Romm does an exceptional job of discussing the market dynamics and tradeoffs in our energy and environmental decisions.

    Too bad that Romm fella doesn’t have his own, you know, blog, or something…

  2. Deane Little says:

    Romm’s argument assumes that you are using dirty hydrogen… generated from fossil fuels… to drive your hydrogen car 1000 miles. Of course such an approach would be foolish, burnign much more carbon than simply using a gasoling powered car. On the other hand, using renewable energy to generate hydrogen by electrolysis would allow you to drive your car as far as you want to without adding a single molecule of CO2 to the air. Romm’s is a typical straw man argument against hydrogen… there are clean and dirty ways of making hydrogen, and the anti-hydrogen crowd ALWAYS makes their arguments against the dirty form.

  3. Joe says:

    Actually, the article misstated my argument. My argument was actually that using renewable power to make hydrogen to fight global warming saves 485 pounds of carbon dioxide, but using that renewable power directly to displace coal power saves 2100 pounds — plus you avoid buying the expensive electrolyzer, hydrogen infrastructure, and fuel cell. Thus hydrogen fuel cell cars are a cost-inefficient way to use renewable resources to fight global warming — at least until we have eliminated carbon dioxide emissions from coal combustion.

  4. Earl Killian says:

    To Deane Little:

    Let’s say you decide to eliminate fossil fuels from personal transportation. You know how to generate electricity from wind or solar. What do you do with it? If you’re given a choice of (1) using the electricity to produce hydrogen, or (2) using the electricity to charge the batteries of plug-in vehicles, which do you choose?

    The plug-in vehicle case is easy to analyze, since production vehicles exist. The U.S. drove 2.7 trillion miles in 2004 in passenger cars, other 2-axle 4-tire vehicles, and motorcycles. Using 260 Wh/mi, 370 Wh/mi, and 150 Wh/mi for these three vehicle classes, you come up with 821 TWh needed per year (300 Wh/mi on average). The grid (power plant to wall plug) is 92%, so you need to generate about 900 TWh at the power plants. For wind, this will take about 410 peakGW of wind turbines (102 GW average, assuming 25% of peak). Wind turbines used to cost $1000/peakkW (anyone have a newer number?), so the cost of the wind farm is $410 billion. (Solar thermal in the desert (e.g. Stirling Energy or Ausra), is alternative to wind.)

    The hydrogen fuel cell vehicle (HFCV) case is a bit trickier, because HFCVs are not production-ready (this an issue all by itself). Since a HFCV is essentially a battery electric vehicle (BEV) with some (but not all) of the batteries replaced by a hydrogen tank and PEM fuel cell, we can compare them by just looking at the two paths: (1) electricity to plug, plug to battery, battery output; (2) electricity to H2, H2 compression, H2 shipment, H2 storage, H2 decompression, and H2 to electricity in the PEM fuel cell. For (1), the figures are 92% and 86% efficiency, giving 79% overall. For (2), Lovins proposes to extract some of the energy from H2 compression during decompression; I don’t know what is reasonable here, so let’s be very generous and assume the product achieves 90% efficiency. I don’t know what to assume for H2 shipment, so again be generous and assume 95%. H2 storage is another issue (there is leakage, and H2 does affect the stratosphere), but let’s ignore it here. Wikipedia gives the efficiency of electrolysis of water as 50-70% (the rest of the energy goes into heat). PEM fuel cells are at best 50% efficient (again, heat gets the rest of the energy). Multiplying these figures gives 21-30% overall efficiency. That is a factor of 2.6 to 3.7 reduction from the 79% efficiency of the BEV pathway (more detailed estimates often come up with a 4 times). Therefore, instead of $410 billion of wind turbines, one would need $1 trillion to $1.5 trillion.

    Given the choice between spending $410 billion, and $1 to $1.5 trillion, which do you choose?

  5. Earl Killian says:

    A postscript to the above reply to Deane Little:

    The above analysis ignored the cost of the vehicles themselves. Since BEVs are much cheaper than FCVs, it would be an even more skewed comparison if vehicle costs were included. Let’s say the cost of the H2 storage tank and PEM fuel cell could be brought down to a mere $10,000 premium over the batteries saved (the premium is currently much more than this). There were 136 million passenger cars on the road in 2004, 6 million motorcycles, and 92 million non-car 2-axle vehicles (mostly light trucks and SUVs), for a total of 234 million vehicles. To replace that fleet with FCVs instead of BEVs costs an additional $2 trillion. Moreover, this cost recurs with higher frequency than the 20 year lifetime of wind turbine farms, or the 30 year lifetime of solar thermal farms.

  6. Deane Little says:

    Again, I protest the numbers in the original post and the way they have been skewed to give every advantage to the petroleum powered vehicle and every disadvantage to the hydrogen car. So an efficient petroleum powered car (40 mpg) generates only 485 pounds of CO2 during a 1000 mile trip? That low number assumes that gasoline magically appears inside the gas pump, having apparently reproduced within the underground tank every night when the lights go out. In fact my understanding of the process is that petroleum is extracted from deep within the earth, pumped to the surface, transported thousands of miles, refined in an energetically wasteful process, transported to gasoline stations and then pumped into your car… where it is finally burned to create 480 pounds of CO2 (19 pounds per gallon). My humble guess is that the very complicated path the gasoline takes to reach your car might add just a few pounds of CO2 to the air. I am sure someone here knows the ratio, what is it? Not pretty I am sure, which is perhaps why it always gets ignored in the equation.

    Re Joe’s argument about using renewable energy to make hydrogen vs just using the electricity, point taken. But your numbers are MUCH higher than those projected in the Socolow Wedge model. If you crunch their numbers for producing hydrogen from wind vs just producing electricity from wind there is about a two fold difference in carbon saving efficiency. But their model (wedge 10 I think it is) suggests 1 megawatt hour generated by wind saves about 1267 pounds of carbon dioxide… not 2100. So depending on how much CO2 is generated drilling, pumping, transporting, refining, transporting again and finally burning gasoline, the numbers sure look a lot closer than your 485 vs 2100 argument. And who’s driving a 40 mpg car anyway?

  7. Joe says:

    The upstream CO2 numbers for gasoline add about 20% — I usually include them.

    As for Socolow Wedge model, I hope this is not come as a big shock to you but it’s analytically weak. In any case, 100% coal has about double the emissions of the grid. I use coal in the analysis because I’m assuming this is done under a carbon cap, where new renewables would displace coal.