mechanical engineering power 2003



Gauging Efficiency,
Well to Wheel

While the idea of using hydrogen to replace fossil fuels may seem like a dream come true, reality is more complex.

By Frank Kreith and R.E. West

George W. Bush has not been hailed as a "green" leader, but in his State of the Union Address, the president offered to support the dream of many environmentalists. "A simple chemical reaction between hydrogen and oxygen generates energy which can be used to power a car, producing only water, not exhaust fumes," Bush said. "The first car driven by a child born today could be powered by hydrogen and pollution-free."

Although it's far from new, the idea of using hydrogen to replace fossil fuels seems, at first blush, like a dream come true. The reality is more complex. Hydrogen does not exist in nature in a form that can be used as a fuel. It occurs naturally in water or in carbon compounds such as methane, and must be converted into the combustible form, H₂, by chemical processing. This processing requires energy from a primary source, such as coal, natural gas, uranium, wind, or the sun, and is thus at least somewhat inefficient and polluting.

Any analysis of the hydrogen-vehicle concept must take into account the steps necessary to make the hydrogen and then get it into the fuel tank. And any comparison between hydrogen vehicles and other existing and potential technologies must include all steps in the technology, a process called well-to-wheel analysis.

We have studied this issue for many years, and have made a well-to-wheel analysis of 12 significant technologies that could power the U.S. ground transportation system. Some are available today, while the others will be practical within the next quarter-century with no need for any major technological breakthroughs. And our results, which have been borne out by other researchers in the field, may surprise the president and the environmentalist backers of hydrogen vehicles. Compared with other fuel technologies, hydrogen vehicles are quite inefficient. On a realistic, well-to-wheel basis, a hydrogen-powered vehicle would be no friend to the environment if natural gas or any other fossil fuel were used as the primary energy source to generate the hydrogen.

REAL COSTS OF FUEL

How essential is an efficient and economically viable transportation system for the United States? Most American families spend more on driving their cars than on health care, education, or food. Moreover, the demand for transportation fuel is growing, not only in the United States but also all over the world, and its cost is escalating as a result. And because of the overwhelming dominance of petroleum as the fuel of choice, transportation will become more expensive in the coming decades.

In a seminal analysis of the growth, peaking, and subsequent decline of oil production in the United States, the oil geologist M.K. Hubbard predicted in 1956 that oil production in the U.S. would peak about 1973. His predictions have been borne out, and today more than 60 percent of the oil used in the U.S. is imported.

Subsequently, the Hubbard analysis was applied to worldwide oil production. Based upon the best estimates of total worldwide oil reserves, it is expected that the total oil production will peak somewhere within the next 20 to 40 years. Before we hit this peak, some other transportation fuel must be available to supplement oil.

Several alternative fuel options have been investigated to replace gasoline and diesel fuels. The main alternatives are propane, compressed natural gas, hydrogen, Fisher-Tropsch (F-T) diesel, methanol, ethanol, and electricity. But these fuels are derived from other, more basic feedstocks: biomass, coal, natural gas, nuclear fission, petroleum, or solar power.

Ethanol, derived primarily from corn, requires a different analysis than that done here to evaluate its efficiency; but unless the cost of production is significantly reduced, or its use is mandated, it's not likely to compete with natural gas.

Because of its environmental advantages and its relative abundance in stable regions, natural gas is considered to be the most likely near-term supplement for gasoline and diesel fuels. Natural gas is already used in both liquefied and compressed form to power the internal combustion engines of cars and trucks all over the world. But, it can also be used as the feedstock to produce other fuels, including hydrogen, methanol, and F-T diesel, or to generate electricity for electrolysis or charging batteries.

In conducting our analysis of alternative transportation technologies, then, we have used natural gas as the feedstock for each of these technologies, as well as the source of supply for all process heat and electricity requirements. In that way, all of these alternatives are compared on the basis of their efficiency with respect to the same energy source.

We also looked at a variety of engine and drivetrain configurations. The internal combustion engine is, of course, the current standard. And all-electric battery power has a long automotive history. Fuel cells, another technology, are essentially batteries wherein hydrogen and oxygen combine to generate electricity, and this electricity drives an electric motor that powers the vehicle. Because the electricity is generated by a chemical reaction rather than thermally, the efficiency of the fuel cell is not limited by the Carnot principle. The fuel-cell systems we analyzed use methanol or hydrogen as the fuel; the one has an onboard chemical reformer to generate hydrogen from the methanol, and the other uses hydrogen generated directly at a central plant.

A hybrid-electric vehicle, or HEV, uses a small internal combustion engine in combination with an electric motor and battery storage. (The engine and motor can be arranged in series or parallel. This analysis is based on a parallel configuration because this is the pattern chosen by automakers for commercialization.) In parallel, both the IC engine and the electrical propulsion system are directly connected to the drivetrain of the car. Either the engine or the electric motor can power the vehicle separately, but their outputs can also be combined when additional power is needed.

Since the IC engine of the hybrid-electric vehicle does not have to supply peak power alone, it can be significantly smaller than the engine of an equivalent-size conventional vehicle, and can operate at all times at or near its maximum efficiency. Excess energy from the engine is stored in the battery during periods of low power demand, whereas for acceleration or hill-climbing energy from the battery can be used to supplement the engine.

These and other features serve to make an HEV more efficient and less polluting than a conventional vehicle. Models such as the Toyota Prius and the Honda Insight are on the market today.

If sometime in the future automotive fuel cells were to be commercialized, they too could be placed into HEVs. But no such system is currently available and its development and commercialization are likely to be far in the future.

In all, we looked at six kinds of power trains and six fuels combined in 12 different fuel-engine combinations. Natural gas can be used directly in a conventional spark ignition engine, alone or combined with an electric motor in a hybrid gas-electric configuration, similar to the Prius. Natural gas can also be the feedstock to generate diesel fuel by a Fisher-Tropsch process—in which carbon monoxide and hydrogen react to produce liquid hydrocarbons—for conventional and high-compression diesel engines, as well as in diesel-electric hybrid configurations.

Hydrogen generated by steam reforming natural gas can be used directly in a fuel cell or as the fuel in a spark ignition engine, and a hybrid. Electricity generated by natural gas combined-cycle power plants can charge an all-electric battery-powered vehicle or can electrolyze water to make hydrogen for a fuel cell.

This chart, and the ones that follow, show the efficiency of six kinds of power trains—conventional and hybrid spark ignition, conventional and hybrid diesel, fuel cell, and battery-electric—and six fuels—natural gas, Fisher-Tropsch diesel, an F-T diesel-gas mix, methanol, and both steam-reformed and electrolized hydrogen—in a dozen different combinations.

When many people think about automobile efficiency, they treat the car as the entire system. In terms of tank-to-wheel efficiency, three of the technologies listed above stand out: the all-battery electric vehicle, the hydrogen fuel cell, and the diesel-electric hybrid. In fact, a conventional spark ignition engine is only half as efficient as a battery electric vehicle. Because of this presumed high efficiency, many companies support development of these three options. Battery electrics have long had vocal advocates, and they are often depicted as the "cars of the future."

But this sort of efficiency is illusory. One clear example of this was the effort by California to mandate the use of "pollution-free" electric vehicles. The 1990 California Low-Emission Vehicle Program required that, in order to reduce air pollution in metropolitan areas, at least 2 percent of all vehicles sold by each automaker be zero emission vehicles by 1998. The only technology available at that time to meet this goal was the electric battery vehicle. This mandate, however, ignored the lack of an adequate infrastructure, the state of zero emission technologies, and the pollution created in producing the electricity necessary to recharge the batteries.

And though the efficiency of battery-powered vehicles might be high, the production of electricity necessary to recharge the batteries was quite inefficient. Had it been implemented on a large scale, California's LEV program would have merely shifted pollution from tailpipes in the metropolitan areas to smokestacks of distant power plants.

California could have avoided an eventual debacle had it only done a well-to-wheel analysis before setting its emission standards. It's perhaps surprising that well-to-wheel analysis is not more widely performed since the process is rather straightforward. First, the individual efficiencies of the various steps necessary to bring the feedstock—in our case, natural gas—from the well and convert it to fuel in the tank (or electricity in the battery) are determined and multiplied together to get the overall well-to-tank efficiency. Then, the individual efficiencies of the steps involved in powering the vehicle wheels from the fuel in the tank are determined and multiplied together to give the overall tank-to-wheel efficiency. Finally, the well-to-tank and tank-to-wheel efficiencies are multiplied together to obtain the overall well-to-wheel efficiency of each technology.

The well-to-tank efficiencies of producing electricity or hydrogen are much lower than the efficiency of any of the other fuels investigated. Drawing methane from the ground, delivering it via pipeline and compressing it to final pressure retains nearly 90 percent of the gas's original energy content. But adding the steps of burning it in a combined-cycle power plant, sending the electricity generated across transmission lines, and using the electricity to electrolyze water to make hydrogen reduces efficiency by a factor of three.

Calculating over the full energy cycle, the high tank-to-wheel efficiencies of the hydrogen fuel cell and battery electric options are offset by the low efficiencies of producing the fuel in these forms. As a result, the overall efficiency of the hydrogen vehicle option looks much less promising than it is often portrayed in the commercial literature. At 13 percent, the well-to-wheel efficiency of the fuel-cell option with hydrogen produced by electrolysis is lowest of the options we studied. In fact, it is no better than that of a conventional gasoline-powered Chevrolet Silverado, which according to an analysis by General Motors, has a well-to-wheel efficiency of about 13 percent.

Since global warming gases are generated at approximately the inverse proportion to the well-to-wheel efficiency, hydrogen-fuel cell vehicles are far from "pollution-free" as long as the primary energy source is a fossil fuel.

The best performers we found were the hybrid vehicle technologies running on natural gas, F-T diesel, or a mixture of natural gas and F-T diesel. They have well-to-wheel efficiencies of 30 to 32 percent. Fuel-cell vehicles fueled by hydrogen generated by steam reforming natural gas at a central plant come in a close second at 27 percent.

Our analysis is based on the assumption that the energy or fuel necessary for vehicle propulsion is generated from natural gas. If electricity were produced from non-fossil sources, such as wind, solar energy, fusion, or fission nuclear energy, then it is not efficiency per se, but primarily cost and secondarily environmental effects that would determine their feasibility. Renewable sources are environmentally the most benign. Wind power is close to being price competitive for electric power generation in favorable locations, but the available wind-generation capacity is only a tiny fraction of what would be needed for an electrically powered transportation system.

At the present time, photovoltaic and solar thermal electric power are more expensive than electricity from combined-cycle, natural-gas power plants. Fusion power is nowhere near a reality, and fission power is not only more expensive than other electricity sources, but to be politically acceptable, a safe way of disposing of the nuclear waste will have to be demonstrated. Given these shortcomings of alternative, non-fossil power sources, it's unlikely that any of them would be ready to meet the 25-year time frame of our analysis.

CONSIDERING ALTERNATIVES

Other groups have made technical assessments similar to this well-to-wheel analysis, and some of the most prestigious ones have reached conclusions similar to those presented here.

Germany's Federal Environmental Agency concluded in 1999, "Hydrogen is an inappropriate choice (for transportation application) because of the high losses incurred in the production and processing stages." A 2001 study, "Well-to-Wheel Energy Use and Greenhouse Gas Emissions of Advanced Fuel/Vehicle Systems," performed by General Motors Corp. in cooperation with Argonne National Laboratory, BP, ExxonMobil, and Shell, concluded that, in terms of energy use, the performance of diesel-electric hybrids was nearly identical to that of fuel cell hybrids run on gaseous hydrogen or reformed gasoline. A report released in March 2003 by MIT's Laboratory for Energy and the Environment concluded, "Even with aggressive research, the hydrogen vehicle will not be better than the diesel hybrid in terms of total energy use and greenhouse gas emission by 2020."

All things being equal, hydrogen cars wouldn't make sense, given that a hydrogen fuel cell vehicle has a lower well-to-wheel efficiency than hybrid-diesel technologies. But the use of hydrogen as a transportation fuel has additional shortcomings and economic problems. One of the most difficult to overcome is the cost of developing an infrastructure to support the distribution of hydrogen fuel. Estimates of the costs of building such a hydrogen infrastructure vary widely. One Argonne National Laboratory study calculated the cost would range from $60 billion (for a 2 percent market penetration by 2020) to $500 billion (for a 40 percent market penetration by 2030).

Another challenge faced by the hydrogen vehicle is energy storage. To store a reasonable amount of hydrogen gas in a container, it has to be pressurized, typically to about 5,000 psi. Alternatively, to store hydrogen in liquid form requires cooling it to about -250°C, and then placing it in a large Thermos-bottle type of structure. As a cryogenic liquid, hydrogen occupies nearly four times as much space as gasoline for the same amount of energy stored. In addition, hydrogen storage and transportation systems require special alloys for pipes, valves, and tanks, because hydrogen embrittles steels. Consequently, the existing natural gas infrastructure cannot be adapted to hydrogen.

Although it is possible that future research may some day provide solutions to these and other problems for using hydrogen as a transportation fuel, the technology is not ready for commercialization and funding for an infrastructure isn't available today.

President Bush's vision of pollution-free vehicles no doubt helped his standing among voters who are concerned about the environment. But our results and others like them should be a warning to decision makers not to expect commercialization of hydrogen vehicles any time soon.

From an engineering and economic perspective, it is unlikely that hydrogen fuel cell vehicles will achieve market penetration in the coming decades. To protect the U.S. transportation system, policy makers should consider promoting other, more viable technologies and fuels for ground transport vehicles in preparation for the day when global oil production begins to decline. Starting today would not be too soon.

The original research cited in this article was first published in Transportation Quarterly, Vol. 56, No. 1, Winter 2002 (pp. 51-73).