Professor John McCarthy
Father of AI

Hydrogen

7 December 2008

Hydrogen is often advocated as an energy medium. Here are some relevant facts.

  1. Hydrogen is the lightest of the elements with an atomic weight of 1.0. Liquid hydrogen has a density of 0.07 grams per cubic centimeter, whereas water has a density of 1.0 g/cc and gasoline about 0.75 g/cc. These facts give hydrogen both advantages and disadvantages. The advantage is that it stores approximately 2.6 times the energy per unit mass as gasoline, and the disadvantage is that it needs about 4 times the volume for a given amount of energy. A 15 gallon automobile gasoline tank contains 90 pounds of gasoline. The corresponding hydrogen tank would be 60 gallons, but the hydrogen would weigh only 34 pounds.

  2. When hydrogen is burned in air the main product is water. Some nitrogen compounds may also be produced and may have to be controlled. Should greenhouse warming turn out to be an important problem, the key advantage of hydrogen is that carbon dioxide (CO2) is not produced when hydrogen is burned.

  3. Since hydrogen is not available in significant quantities in nature in pure form, the main present way of getting hydrogen is steam methane reforming, and this will probably remain the most economical way as long as methane (natural gas) is available cheaply and in large quantities, and hydrogen is required only in small quantities. When the price of methane goes up to more than three times its present price because of scarcity, hydrogen will be obtained by splitting water H2O into hydrogen H2 and oxygen O2. The chemical reaction is written

    2H2O + energy => 2H2 + O2.

    The well developed way of splitting water is by electrolysis. If fossil fuels, e.g. coal, oil or natural gas, are used to generate the electricity, there is no advantage over using the fossil fuels directly. Indeed you still get all the CO2, and there is a considerable loss of energy. Therefore, the large scale use of hydrogen depends on using either nuclear or solar electricity. In both the nuclear and solar cases, there are possible but undeveloped technologies that don't use electricity as an intermediate form of energy. [2003 September: There is a thermochemical process for splitting water that is claimed to be twice as efficient as elactrolysis. Here's an article on the sulfur-iodine cycle by Ken Schultz of General Atomic given at a Stanford University Global Climate and Energy sponsored meeting in 2003 April. To get good efficiency it requires reactors operating at a higher termperature (950 C) than present power reactors. Schultz's article also discusses solar thermal production of hydrogen. (There are also proposals to combine getting heat for houses and chemical processes and electrolysis with some saving of energy.)

  4. In either case, the law of conservation of energy tells us that all the energy to be obtained by burning the hydrogen must be supplied by the primary source, e.g. nuclear or solar. Of course, since these processes aren't 100 percent efficient, there is some loss of energy. Therefore, the use of hydrogen as an intermediate is justified only when there is some reason not to use the primary source directly. For vehicles the reason is that both nuclear nor solar power plants are too big to carry around, except that nuclear power is suitable for large ships.

  5. If there is large scale use of solar energy, the energy is likely to be generated far from where it is used and at a different time. Hydrogen has been proposed as both a storage and transmission medium. It should work for these purposes. I don't know how hydrogen pipelines compare with high voltage electric transmission as means of long distance transmission of energy.

    Hydrogen can be transported by pipelines similar to those used to transport natural gas. There are some addtional problems, because hydrogen tends to leak more and can embrittle some metals used for pipelines. The existence of a 240 km hydrogen pipeline in Germany operated by the company Air Liquide provides evidence that these difficulties can be overcome.

    In 2004 December I was informed that there is an 879km hydrogen pipeline network in Belgium, France, and the Netherlands operated by Air Liquide.

    However, the technology of efficient long distance transport of electric energy may be improved enough to obviate the advantages of hydrogen except for vehicles.

Hydrogen as a motor fuel

Hydrogen can be used as a motor fuel, whereas neither nuclear nor solar energy can be used directly.

Nuclear power requires heavy shielding to keep the neutrons away from people - too heavy for cars. It can be used in ships, and is used in American, British and Russian warships, especially submarines and aircraft carriers. The U.S. and Japan built commercial nuclear powered ships, one each (Savannah and Mutsu). (There were even proposals to use it in locomotives.) However, initial difficulties combined with anti-nuclear politics caused these projects to be abandoned and the ships mothballed. The Soviets built nuclear powered icebreakers, and these are in use. I think nuclear power will be revived for commercial ships when its political problems are overcome and the technology is further debugged.

Solar energy can't be used directly in cars except as a stunt. The current solar-powered cars are just religious exercises in the solar religion. The problem is that a solar array of a size that can be mounted on a car produces too little energy to give useful performance, and even that little isn't available at night or when it is very cloudy.

Hydrogen can be used as a fuel directly in an internal combustion engine not much different from the engines used with gasoline. The problem is that while hydrogen supplies three times the energy per pound of gasoline it has only one tenth the density when the hydrogen is in a liquid form and very much less when it is stored as a compressed gas. This means that hydrogen fuel tanks must be large.

Demonstrations of hydrogen powered vehicles have usually used compressed hydrogen gas. However, because of the low density, compressed hydrogen will not give a car as useful a range as gasoline. It may be even worse than using lead-acid batteries. Hydrogen can achieve a reasonable density adsorbed in metal hydrides, but then the weight of the metals makes the system very heavy.

The most practical way I know of using hydrogen as a motor fuel is to accept the difficulties of handling liquid hydrogen and solve them. There are three main problems.

  1. The low density. A hydrogen fuel tank will have three times the size of a gasoline tank. Also it must be insulated, and this will add to its bulk. This seems entirely bearable.

  2. Safety problems. Liquid hydrogen is cold enough to freeze air, and accidents have occured from pressure build-up following plugged valves. Some say these problems can't be overcome, but I side with those who think they can be overcome. In a collision the hydrogen tank may rupture, as can a gasoline tank. Limited accident experience suggests that the danger is somewhat less with hydrogen than with gasoline, because the hydrogen dissipates rapidly. The release of hydrogen into a confined space like a garage risks an explosion.

  3. Since the insulation can't be perfect, the hydrogen will gradually evaporate, typically 1.7 percent per day. This is too fast for a car to sit for months between uses. A tank of compressed hydrogen holding enough to get to a hydrogen station would solve this. If the engine is flexible enough to burn gasoline as well as hydrogen, a half gallon gasoline tank would suffice. Some automobile companies, e.g. BMW, have experimented with vehicles powered by liquid hydrogen. However, hydrogen cannot come into common use until the political obstacles to nuclear expansion are overcome or the technological obstacles to large scale solar energy are overcome. It is unlikely to be used as long as gasoline remains so cheap, i.e. as long as oil remains cheap and fear of global warming does not prevent its use. We hydrogen enthusiasts will just have to wait.

    October 2000 note: The German automobile company BMW has just announced a car to be powered by liquid hydrogen and a plan to build a network of hydrogen refuelling stations in Germany. The car is named the 745 hdi and it is based on the BMW 745, a top-of-the-line model. A range of 350 km is promised. This is a bit small but may be tolerable. The car can also burn gasoline. BMW built a small fleet of these cars and has demonstrated them in various countries including the US, both in L.A. and Detroit.

    Here's an article by BMW engineers. Oops, the link doesn't work directly as I have copied it, but the article and others can be found by Googling bmw+hydrogen.

    2003 November note: Jay Keller of Sandia Laboratory presented Why hydrogen? - Building an infrastructure at the Stanford University Global Climate and Energy workshop on Hydrogen Production, Storage, and Utilization: Technical Barriers and Research Opportunities in 2003 April.

    The large scale use of hydrogen for cars requires a very large investment in infrastructure. I suppose present gasoline stations can have hydrogen tanks and hydrogen pumps added, just as many gasoline stations also sell diesel fuel. Facilities for delivering hydrogen from the electrolysis plants to the users will be expensive. The transition to hydrogen will be triggered by one of two events. (1) The world will eventually run out of conventional motor fuel at reasonable prices. Opinions differ about when this will happen. Some pessimists predict a peak in world oil production in 2005. Other people think we won't run out in this century. I tend to the latter view considering that the oil sands in Alberta, Canada are being rapidly developed, produce oil profitably sellable at $12 per barrel and have resources greater than Saudi Arabia. (2) It may happen that increased CO2 causes real harm to the world and a crash program to stop emitting so much is required. There is not sufficient evidence of harm to convince governments to take drastic action in spite of the enthusiasm of the environmental community and its scientific supporters. [I can't take these proposals from these people seriously as long as they don't even mention nuclear energy. They are just playing with us.]

    2003 September: A Republican draft of the energy bill includes $1.1 billion to build a nuclear reactor to produce hydrogen. That would be a big step in the right direction.

    Energy and the hydrogen economy by Ulf Bossel and Baldur Eliasson analyzes the energy cost of a "hydrogen economy". The problem is that transporting energy as hydrogen uses much more energy than transporting it as oil or natural gas because of the low density of hydrogen, either as a liquid or as a compressed gas.

    Bossel and Eliasson's rhetorical object seems partly to knock off the idea of a hydrogen economy in favor of their preferred methanol economy. However, they propose to get the carbon for their methanol economy from biomass. That would satisfy greenhouse objectives, whereas getting the carbon from coal would not. I'm doubtful about the biomass part, and they don't offer specific arguments for its feasibility.

    New to me in their paper is the analysis of hydrogen transportation systems and their energy costs. The conclusion most interesting to me is that 2.12 times as much energy goes into generating and transporting liquid hydrogen by truck than you get into the fuel tank of the car. If their calculations are correct, the cost for the automobile user is regrettable but entirely bearable, because only a small part of the costs of operating a car are fuel costs, and a big part of the fuel price is federal and state tax.

    An alternative to transporting liquid hydrogen long distances is to electrolyze water locally, perhaps in the fuel station itself.

    Here's what Pimentel (1996, p. 211-212) has to say.

    In terms of energy contained, 9.5 kg of hydrogen is equivalent to 25kg of gasoline ( Peschka 1987). Storing 25 kg of gasoline requires a tank with a mass of 17 kg, whereas the storage of 9.5 kg of hydrogen requires 55kg, (Peschka 1987). Part of the reason for this difference is that the volume of hydrogen fuel is about 4 times greater for the same energy content of gasoline. Although the hydrogen storage vessel is large, hydrogen burns 1.33 times more efficiently than gasoline in automobiles ( Bockris and Wass 1988). In tests a BMW 745h liquid-hydrogen test vehicle with a 75 kg tank and the energy equivalent of 40 liters of gasoline had a cruising range in traffic of 400 km, or a fuel efficiency of 10 km per liter ( Winter 1986).

    At present, commercial hydrogen is more expensive than gasoline. Assuming $0.05 per kwh of electricity from a nuclear power plant during low demand, hydrogen would cost $0.09 per kwh ( Bockris and Wass 1988). This is the equivalent of $0.67 per liter of gasoline. Gasoline sells at the pump in the United States for about $0.30 per liter. However, estimates of the real cost of burning a liter of gasoline range from $1.06 to $1.32 when production, pollution, and other external costs are included (Worldwatch Institute 1989). Therefore, based on these calculations hydrogen fuel may eventually become competitive.

    The references above are copied from Pimentel (1996, p. 211). I plan to look them up, and this may change what I say.

    The above comparison between current costs of gasoline and hydrogen power for cars seems to be somewhat biased in favor of hydrogen. Taxes seem to be included in gasoline cost and not in hydrogen estimates, but roads will still have to be maintained when hydrogen is used as a fuel. Howover, I suspect the Worldwatch estimate of the "real cost" of burning a liter of gasoline is exaggerated.

    For me the decisive point is that the costs of a automobile transportation system using hydrogen produced from water using nuclear energy are low enough so that people worldwide who use automobiles will not give up the freedom they provide, regardless of efforts to get people to settle for public transportation or low range and low performance cars of one kind or another. This doesn't say that adequate batteries won't be developed to make electric cars better than liquid-hydrogen internal combustion powered cars. Maybe they will, but we won't settle for less mobility than hydrogen can provide. See also my scheme for swapping batteries frequently along highways.

    Mazda has experimented with hydrogen powered Wankel engine cars. They are trying to get California to declare them zero emission with some prospect of success. However, if California chickens out of the zero emission demand (as it should and probably will), I'll bet the Mazda will not be offered for sale any time soon.

Many people, including car companies, are being persuaded that cars of greatly lower performance, e.g. in size, in range and in acceleration, are acceptable and will be required in the future. For example, Daimler Benz (Mercedes) is now (1996) experimenting with a compressed hydrogen car. However, if just one manufacturer in the world, e.g. with a liquid hydrogen powered car, succeeds in maintaining present performance, then all the fine words about living with lower performance cars will amount to nothing.

West Virginia University has a Hydrogen Review page. It mentions several ways of using hydrogen for motor vehicles. It would seem to me from their numbers that liquid hydrogen is the winner - as stated above. However, the page makes no comparisons at all.

The Department of Energy supports research on hydrogen. The value of this page seems to be somewhat limited. First, it only mentions what is currently being supported by DoE, and doesn't mention what other organizations are doing. Second, it is entirely politically correct as is apparent in the first paragraph. Nuclear energy isn't even mentioned as a possible source, not even to be dismissed. Nevertheless, there is useful information.

A check on the DoE site in 2002 February shows that there is still no mention of nuclear energy. The site still has the Clinton (or perhaps it's Gore) religion. During the California energy crisis Vice-President Cheney made a speech or two favoring nuclear energy, but the push evaporated when there turned out to be enough natural gas for the time being. I guess the Bush Administration has enough problems to occupy all its attention and has no energy left to make its Department of Energy stop ignoring a source of energy favored by the Administration. Of course, maybe it was just the Vice-President who favored it.

Some companies that supply hydrogen and related technology: Air Liquide(sells gases including hydrogen), Air Products and Chemicals, Inc., The BOC Group, Praxair.

Air Products and Chemicals, Inc. has a page on a hydrogen fuel celled bus in Chicago that has a range of 250 miles - enough for a whole day. This may be good enough performance for cars.

Let me reiterate that the point of this page is not to show that one particular way of powering automobiles is best but to show that there are enough ways of keeping individual mobility in the advanced countries and for the backward countries to achieve it. Cars powered by liquid hydrogen fueled internal combustion engines with the hydrogen produced by nuclear power plants electrolyzing water will suffice to preserve our mobility.

1999 July Scientific American survey article

The 1999 July Scientific American has an article entitled "The electrochemical engine for vehicles" by A. John Appleby of Texas A & M University. The article concentrates on projects by the major manufacturers for powering cars. Most of them use a hydrogen-oxygen fuel cell with a platinum catalyst. They all work, but the main problems are the expense of the catalyst, the expense of the electrolyte and storing the hydrogen or making it from methane in a reformer. None are close to market. Liquid hydrogen, my favorite, is still in the running.

It is hard to know how serious these projects are even though a billion dollars will have been spent by 2000 July. It is prudent for the car companies to keep the Government appeased, and they also must hedge their bets against the possibility that governments world-wide will really force people to accept very low performance cars. None of them shows much sign of competing with the gasoline powered internal combustion engine at less than a very substantial multiple of present fuel costs.

Some schemes plan to get their hydrogen by reforming gasoline on board the vehicle. There is a chance of a slight increase in efficiency.

The article ends with

In time, we are likely to see an infrastructure for delivering hydrogen for fuel-cell powered vehicles. The result will be a more efficient, clean transportation sector, a reduction in imported oil and lower carbon dioxide emissions. A hydrogen delivery system will be built when it is technically feasible, affordable and necessary. But that point is still some years in the future.

The big taboo is on mentioning nuclear energy for splitting water as the source of hydrogen. There is no taboo about solar energy in various forms, but I doubt that these companies expect solar to succeed in producing enough energy at a reasonable cost. Most environmentalist scientists would like them to say that nuclear energy is unacceptable, but they don't want to say that - so they keep silent.

My guess is that all these companies are just hedging their bets. Also the author of the article does not feel it prudent to say all that is on his mind - and not just about nuclear energy.

2006 January note: The Hawaii Natural Energy Institute has a report on hydrogen. Useful appendices compare technologies and their costs for various ways of producing and using hydrogen. As is still common in the energy studies business, nuclear energy is ignored. Nevertheless there is much useful information. I'm sure many of the authors of the study know better, but they also know mentioning nuclear is still not useful for getting funding from agencies beholden to the environmental movement. However, at long last, the environmentalists are beginning to wake up and argue about nuclear energy among themselves.

2007 March note: The 2007 April Scientific American has an article entitled "Hydrogen" by Sunita Satyapal, John Petrovic, and George Thomas of the U.S. DoE that seems the most realistic of any I have seen. It recognizes the main problem as the storage of enough hydrogen to give a car an adequate range. Their goal is the 300 mile range of present cars. They discuss compressed hydrogen, liquid hydrogen, and several new chemical systems. As for liquid, they report BMW's intention to sell the Hydrogen7 in Germany and in the U.S. in places where hydrogen stations can be built. The car will have 300 mile range on gasoline but a disappointing 125 miles on hydrogen.

The article cites goals for 2010 and 2015 that seem more adequate. These are based on chemical adsorbents.

Boron

Graham Cowan makes a strong case for boron as a better automobile fuel than hydrogen. The idea is to use nuclear energy to split boron oxide, and to use the boron as fuel in suitable automobile engines. The boron oxide resulting from the combustion is condensed to a solid and stored at the boron station for later conversion to boron again. The main question in my mind is how to design an engine with horsepower per pound close to that of a gasoline engine.

Here's a link to a proposal to combine boron with water to get a hydrogen fuel. I don't know how long the link will last.

My goal for this page is to put in more numbers, e.g. the number of joules per kilogram of hydrogen required by electrolysis and also references to the literature on the production and use of hydrogen. I solicit information and references.

Here's a link to Magna Steyr, the manufacturer of the liquid hydrogen tank for the BMW 745h.

Here's a start on the numbers.

Table A.59 (of _Handbook_of_Physical_Calculations_, J. J. Tuma, McGraw-Hill New York, 1967 ISBN 0-07-065438-7) titled "Average Heat of Combustion" lists

"Gasoline(0.71) 11.50; Gasoline(0.77) 11.90; Hydrogen 33.90" with units 1E3_kcal/kg in each case.

The Scientific American article cited above lists 50 kw as the peak performance required for a car. That's less than what good performance cars give, but probably about right for economy cars.

This article doesn't claim to have shown that powering cars with liquid hydrogen produced by nuclear reactors is the best solution. I do claim that it is an adequate solution, and our descendants 100 years or 10,000 years from now will have automobiles if they want them.