BMW hydrogen-fueled  Mini
Experimental robotic liquid-hydrogen refueling station in Germany refuels hydrogen powered BMW Mini. Even at super-cooled densities, hydrogen has only a fraction of the caloric energy of gasoline.

The Hydrogen Emperor Has No Clothes

Executive summary of soon-to-be-published white paper on draw-backs of using hydrogen as an energy carrier

By John R. Wilson and Griffin Burgh

Messrs Wilson and Burgh are executives with TMG/The Management Group, a Detroit, Michigan-based consultancy.

Why (Not) Hydrogen?
Hydrogen, at about 18 atomic %, is the most common element in the earth's crust. However, due to its low density, this corresponds to less than 1 wt %, placing hydrogen 9th in order of occurrence by mass. Hydrogen does not occur to any significant extent on earth in its elemental form, although it is vastly more commonplace in outer space.

For commercial use, hydrogen must currently be extracted from those naturally-occurring resources in which it is abundant. These include water, natural gas, crude oil, coal and a few biological sources such as cellulosic biomass and sucrose. Hydrogen is not an energy source but, like electricity, is only an energy carrier. This has a profound impact on how and where we decide to make and use hydrogen.

Pure hydrogen contains no carbon and thus burns to form only water with no carbon dioxide or carbon monoxide (although oxides of nitrogen are formed when combustion takes place in air). Since it is a gas, hydrogen is, in principle, easily transported from the point of manufacture to the point of use; in fact the "packaging, transportation and delivery of the gas presents a number of difficulties. Hydrogen is currently the fuel of choice for fuel cells and, as BMW have shown, can also be used to replace gasoline in a suitably modified automobile combustion engine. Much earlier, coal gas (55% hydrogen) was commonly used in stationary single and multi-cylinder gas engines, mostly for on-site power generation. Such engines are still made (e.g., by Jenbacher) but mostly to run on landfill gas or refinery gas (impure methane).

Given its long history and these advantages, why is hydrogen not already in substantial use for transportation, home heating and other similar applications? There are several reasons which are discussed in more detail in our separate white paper:

Hydrogen does not occur naturally in the free, elemental state and is therefore not readily available. It is always present as water (H2O), as hydrocarbons (mostly alkanes, CNH2N+2) and a few other compounds. Dry coals contain about 4-6 wt% of hydrogen, compared to about 12.5 wt% in gasoline. Hydrogen is therefore a secondary energy source like electricity. It must first be extracted from those materials in which it occurs. In all cases, this requires a substantial investment of energy - energy that today must be obtained from another, typically non-renewable and 'dirty' source such as oil or coal.

The best-known method of manufacturing hydrogen is the electrolysis of water. This requires a large amount of electrical energy and the resulting hydrogen is prohibitively costly. The process also produces a large amount of oxygen (by weight, about 8 times the amount of hydrogen produced), which represents a waste of energy and has limited uses. Aside from the question of whether it is good policy to invest so much energy in converting a very attractive energy carrier (electricity) into one that is far less convenient and attractive (hydrogen), the source of the electrical power is key. Sources such as conventional coal, oil or gas-fired power stations generate pollutants and CO2 and thus appear to defeat the objectives of using hydrogen – energy independence and a cleaner environment. Clean, renewable electrical energy such as that from solar or wind generators would be ideal but these technologies have very far to go before they can offer either the size of energy resource or the low-cost power that will be needed to support a major hydrogen economy. Other "green" sources of power are not yet, and may never be, appropriate in scale, cost or reliability for a continuously-operating, large-scale electrolysis facility. Hydroelectric power is probably the most appealing, but, in 2003, that usually implies captive, dedicated hydroelectric power generated close to the electrolysis plant in a facility located in a remote area. Few new hydro plants are being built in populated areas due to public objections to new dams and associated flooding of farmlands and villages. The manufacture of hydrogen by electrolysis is likely to find a home only where there is no other choice of transportation fuel – for example, in Iceland, where the power can be generated inexpensively using geothermal power.

All other actual and proposed methods of hydrogen production involve thermochemical water-splitting – the reduction of water (as steam) by agents such as carbon or hydrocarbons at elevated temperatures. The established steam-hydrocarbon “thermochemical reforming" processes that are currently employed to produce the large quantities of hydrogen now used by the chemical, fertilizer and oil industries for ammonia manufacture, desulfurization of gasoline, diesel and jet fuels and for heavy oil hydrogenation are by no means 'clean'. Like conventional power stations, they generate large amounts of carbon dioxide and other pollutants, but in the near term they may be the only practical choice currently available for hydrogen production on the large scale required for even a 5-10% substitution of existing transportation fuels.

Reforming technology now exists (a) to convert inexpensive hydrocarbons of high hydrogen content (methane, even gasoline) into pure hydrogen by thermochemical reforming and (b) to remove and, if necessary, permanently sequester the carbon in minerals such as magnesium silicate. Present indications are that this is a far less energy-intensive clean route to hydrogen than electrolysis and that the resulting full-cycle carbon dioxide emissions are much lower.

Classical reforming technology uses natural gas as a feedstock. In North America, natural gas is now (2003) in very tight supply and is likely to remain that way. U.S. proven and immediately exploitable gas reserves are forecast to last less than ten years – i.e., they will be running out at about the time when they are needed for hydrogen production. There have been no large new discoveries for several years and none are expected. In any case, normal demand has been growing at a strong pace, especially for power generation, and is likely to absorb any growth on the supply side. Deep-sea offshore gas and imported liquefied natural gas (LNG) are likely to become important new sources, but only at a premium price. It appears that any hydrogen economy cannot depend on natural gas as a source of hydrogen. Nor should it, given the high priority of the many other uses for natural gas.

An alternative natural resource that is readily available and of which the U.S. has ample reserves – perhaps as much as 400 years worth – is coal. Coal is usually rejected out of hand by environmentalists as being “too dirty", but they are usually ignorant of the many clean coal developments that date from the 1970s. Hydrogen (or a synthetic natural gas) can be manufactured from coal by one of several gasification technologies that use oxygen (or air) and water injection. The best of these use carbon sequestration by reaction with dolomite or lime (typically the CO­2 Acceptor process) to capture any CO2 produced. The CO2 can be recovered and sold or permanently sequestered, for example in magnesium silicate minerals and the dolomite or lime recycled. Recent economic analyses funded by the USDOE indicate that zero emissions coal (ZEC) technology will become a very important hydrogen source if the hydrogen economy is able to overcome the other barriers that it faces (below).

Hydrogen has a very low volumetric energy density (energy content per unit volume), whether as liquid, gas or compressed gas or even when absorbed on or in any of the materials that have been proposed for its storage. The liquid must also be cooled (at very high energy cost), stored and transported at cryogenic temperatures. These factors raise major engineering issues in storage and transportation. The chemical and petroleum industries mostly use hydrogen close to where it is made. Thus, transportation issues seldom arise and the hydrogen is treated and evaluated economically as just another commodity chemical.

For pipelining, storage and delivery, the heating value of a gas in kJ/kg is not very meaningful. The volumetric heating value in kJ/m3 or kJ/liter is much more relevant. The volumetric heating value of various fuels in various states is tabulated below. All heating values are gross heating values (GHV).

H2 Fuels Comparison

Clearly, as energy carriers on a volumetric basis, the liquid fuels are better than the gases. Hydrogen can compete (but not very well) only when transported or used as a gas at very high pressure - or as a liquid - in very costly hardware.

Because of its low energy density, considerable energy must be used, not only to manufacture, but also to compress and/or store hydrogen and to move it through pipelines at high pressure. When all energy requirements are considered, much more energy is required to make and deliver hydrogen than can be recovered from it. A better alternative for transportation and delivery is a high hydrogen/low carbon content carrier like methanol, methane or even ammonia.

For a given rate of delivery of energy at the end of a pipeline, at least three times the volume of hydrogen must be moved compared to natural gas at the same pressure. This leads to much higher friction losses than for natural gas, notwithstanding the lower viscosity of hydrogen at ambient pressure.

Fuel cell and other hardware that currently exists to utilize hydrogen in conversion devices (other than simple external combustion devices such as stoves or furnaces) is prohibitively costly, and remains so despite very substantial development effort over many years. Only the internal combustion engine, modified for hydrogen, offers a reasonable cost in the near term, but, at 35% or so 'best case' thermodynamic efficiency, is only about half as efficient as current 'best case' hydrogen fuel cell technology with waste heat recovery. Fuel cell vehicular power systems, on the other hand, are still projected to cost about ten times as much as a typical auto engine, even in mass production. They are only about 65% efficient at high loads, 85% or so at low loads and currently offer a rather unimpressive vehicle performance compared to the gasoline engine of equivalent capacity. They also require expensive high-purity hydrogen.

The key issue addressed in this paper is not the cash cost of hydrogen, although that should obviously be as low as possible. It is the energy consumed to produce, compress, pump, liquefy (if appropriate), distribute and deliver the hydrogen relative to the energy that can be recovered from the hydrogen when it is consumed in, say, a fuel cell and used to power a vehicle. Since the recovered energy is, at best equal to the energy input to a hypothetical 100% efficient water electrolysis cell, everything else, from energy lost through generation inefficiencies to electrolysis losses to energy consumed for pipeline pumping and delivery, etc., is a contribution to the net energy loss of the total system. It is easy to come up with a realistic scenario in which the energy investment required for producing, compressing, transporting and delivering hydrogen is substantially greater than the combustion energy that can be recovered from the hydrogen. Even non-electrolytic production of hydrogen for energy transport and delivery is relatively inefficient when heating requirements and reforming inefficiencies are allowed for. Hydrogen is a poor choice of alternate fuel, particularly given the other choices available.


In our white paper we show that the technology required to support an economically and technically viable electrolysis-based hydrogen economy is not available today. In fact, it may never be available, since it requires large amounts of low-cost clean power. What is needed is fairly clear, but some of the barriers that must be overcome are set by the fundamental laws of chemistry, physics and thermodynamics. These WILL NOT be overcome. There are also huge economic barriers to be overcome, especially those involving capital costs. These are so high that the costs in both energy and dollar terms are likely to be so enormous that electrolytic hydrogen, for all its virtues, will never be a commercially viable option on a large scale. We believe, however, that there may be a number of niche opportunities that make sense.

Hydrogen manufacture by one of several alternative methods (all of them developed by the oil, coal and chemical industries that are today so out of favor with the environmental lobby) does offer some hope of at least much more efficient use of energy when making hydrogen. We examine these alternatives in much more detail in the white paper and conclude that it will always be energetically more attractive to deliver the hydrocarbon (such as methane, methanol or natural gas) directly to the end user. The penalty associated with producing hydrogen thermochemically (but not by electrolysis) may be acceptable only if society is willing to pay a high price in both cash and inconvenience for the cleaner environment and the better use of fossil fuels that can result. So far, there is little evidence that this is the case.

The apparent best process choice of the many available is coal gasification to hydrogen, methane, methanol or a substitute natural gas (SNG). If the fuel gas is to be used locally, and not for vehicles, hydrogen may be a good choice. Otherwise a methanol or SNG product is the preferred choice because it can be transported and delivered at much lower cost. These fuels contain some carbon and their use may require conversion to hydrogen and/or carbon sequestration at the point of use, depending upon whether the user’s objective is zero carbon emissions or merely reduced emissions.

The most important conclusion of our study is that although hydrogen may be technically feasible as a fuel in multiple uses (as it has been for almost two centuries), it is not, in most circumstances, economically feasible. Neither is it a “good fuel" despite the many claims made to the contrary. In most instances, the total energy cost of producing, compressing, liquefying, transporting and delivering it to the user will be far higher than the energy recovered from it. In addition, it is inconvenient and often dangerous to use. It makes no contribution whatever to energy independence - i.e., to weaning the U.S. off imported energy supplies - and almost no real contribution to eliminating or minimizing environmental issues such as global warming – that all has to be dealt with at the hydrogen or energy manufacturing plant and is independent of the choice of fuel.


What then should be our choice of a clean alternate energy strategy? We believe that, with the exception of a few special situations:

Hydrogen is not appropriate as a fuel, notwithstanding its zero carbon content - it is simply too energy-intensive, difficult and dangerous to produce and use.

If hydrogen is used, it must not be sourced from either electricity (itself a derived energy carrier) or natural gas (an energy source in short supply with many other priority users).

Coal, rather than oil or natural gas, should be the energy source for any alternate fuel strategy, especially one that is based on derived energy carriers.

Ammonia, despite difficulties in its use, may be the optimum choice of carbon-free energy carriers; however, we do not believe that it is necessary to use carbon-free fuels – reduced carbon fuels are sufficient.

Methanol, methane, synthetic natural gas (SNG), all represent excellent reduced carbon fuels (relative to direct coal, gasoline and other hydrocarbons) that can be derived from coal and used with much greater convenience and efficiency and at a much lower energy cost than hydrogen.

Since all of these low-carbon fuels can be used as direct fuels in alternative designs of fuel cell, we believe that emphasis should be placed on the further development of such cells, with appropriate capability included for additional carbon sequestration – if needed.

Direct methanol fuel cells may be the best choice for transportation use in view of their low temperature of operation, but much more work is required to improve their efficiency.

Hydrogen has been greatly oversold by "evangelists" in the USDOE and elsewhere and also by the environmental lobby, including some very persuasive writers who are adept at choosing half truths top fit their preconceived conclusions. In short, upon close and objective examination, we find that The Emperor Hydrogen has no Clothes.

Times Article Viewed: 8191
Published: 19-Jul-2003


blog comments powered by Disqus