Hydrogen Internal Combustion Engine Vehicles: A Prudent

Hydrogen Internal Combustion Engine Vehicles: A Prudent Intermediate Step or a Step in the Wrong Direction?
Kenneth Gillingham Stanford University Department of Management Science & Engineering Global Climate and Energy Project Precourt Institute for Energy Efficiency Correspondence Address:
P.O. Box 16336 Stanford, CA 94309
January 2007

Abstract
Hydrogen internal combustion engine (ICE) vehicles present much of the same promise as hydrogen fuel cell vehicles (FCVs): reduced reliance on imported oil and reduced carbon dioxide emissions. Proponents envision hydrogen ICE as a bridging technology from gasoline vehicles to hydrogen FCVs. This paper examines the hydrogen ICE technology, focusing on relevant aspects such as power, fuel economy, tank size, and the state of the technology. An economic analysis is then performed to examine the potential implications of widespread adoption of hydrogen ICE vehicles in the United States. The case for hydrogen ICE depends most on key uncertainties in the evolution of vehicle and production technology, the cost of crude oil, and the valuation of carbon dioxide emission reductions. This analysis indicates that promoting hydrogen ICE vehicles may be a sensible policy goal as a transition strategy to hydrogen FCVs, but a more prudent policy would first promote gasoline-electric hybrids.
Key Words: climate change, carbon dioxide, hydrogen, technological change, internal combustion engines, fuel cells

Contents
1. Introduction..............................................................................................................................2 2. Hydrogen in Internal Combustion Engines...........................................................................3
2.1 Properties of Hydrogen .....................................................................................................4 2.2 Relevant Trade-offs ..........................................................................................................5 3. Comparison of Vehicle Technologies .....................................................................................6 4. Economics of a Hydrogen ICE Policy ....................................................................................9 4.1 Scenarios of Vehicle Technology Adoption ...................................................................10 4.2 Fuel Use ..........................................................................................................................11 4.3 Carbon Dioxide Emissions .............................................................................................11 4.4 Net Benefits ....................................................................................................................13 5. Conclusions.............................................................................................................................15 Figures...........................................................................................................................................20 References .....................................................................................................................................28

Hydrogen Internal Combustion Engine Vehicles: A Prudent Intermediate Step or a Step in the Wrong Direction?
Kenneth Gillingham∗
1. Introduction
At the center of on-going debates regarding energy security and global climate change issues lie the difficult issues inherent in the sizable light duty vehicle transportation sector. In contrast to most other sectors, in the light duty vehicle sector there are exceedingly few economically viable substitutes to the dominant energy source: gasoline. Concerns over reliance on gasoline imports from unstable regions of the world, as well as the potential negative consequences of global climate change from gasoline’s carbon dioxide emissions have motivated a vigorous policy debate on alternative pathways for the light duty vehicle transportation sector.
The advent of hybrid gasoline-electric vehicles leaves considerable opportunity for improving the fuel economy of the light duty vehicle fleet without a switch to a radical new technology. However, several technologies hold promise for powering vehicles with lowercarbon feedstocks. In particular, both hydrogen and electricity (e.g., in electric battery vehicles) can be used as energy carriers, in which energy can be generated from a variety of sources, including low-carbon sources, and stored as electricity or hydrogen for eventual use in powering the vehicle. For example, hydrogen can be produced through feedstocks as varied as coal gasification, natural gas steam reforming, electrolysis using solar or wind generated electricity, or direct dissociation in nuclear power production. Powering a vehicle using one of these energy carriers produces little or no tail-pipe carbon dioxide emissions (e.g., the product of hydrogen combustion with oxygen is water). This opens the possibility of running much of the transportation sector on energy derived from low-carbon sources, alleviating one of the major stumbling blocks in the way of reducing carbon dioxide emissions and oil imports.
∗ The author would like to gratefully acknowledge very useful conversations with James Sweeney and Chris Edwards, of Stanford University, and Dan Sperling of UC-Davis. Many thanks are also due to Amul Sathe of Stanford University for sharing his technical expertise. All errors are the full responsibility of the author.
2

Kenneth Gillingham

Hydrogen Internal Combustion Engines

In the 1990s, efforts to introduce battery-electric vehicles in California largely failed, mostly due to an extremely limited range. More recent efforts have shifted to promoting hydrogen. Since 2003, President Bush’s Hydrogen Fuel Initiative has received an appropriation of $150-250 million per year for hydrogen R&D (DOE 2007). In California, Governor Arnold Schwarzenegger signed an Executive Order that plans for a “Hydrogen Highways Network” to develop a hydrogen infrastructure in California (Schwarzenegger 2004). In July 2005, California Senate Bill 76 was signed, providing $6.5 million in initial funding to begin developing this infrastructure. These public policy actions underscore the importance many believe hydrogen has in the future of the transportation system.
But, there are many questions that remain unanswered concerning the economic feasibility and desirability of hydrogen in light duty vehicles. Moreover, hydrogen can be used in both fuel cell vehicles (FCVs) and hydrogen internal combustion engine (ICE) vehicles, and both technologies are currently being developed (Ford 2007). Most discussion and analysis of hydrogen has centered on the fledgling fuel cell technology due to sizeable potential fuel efficiency gains (e.g., NRC 2004). The advocates of hydrogen ICE vehicles see them as a crucial intermediate step to push the hydrogen production infrastructure forward, so it is ready for when FCVs are commercialized. However, there has been relatively little analysis of the merits of promoting hydrogen ICE vehicles as a transition step.
This paper aims to fill this gap through an analysis of the technical details and the economics of hydrogen ICE vehicles. Emphasis is placed on a comparison of hydrogen ICE light duty vehicles to the most prominent competing technologies of gasoline hybrids and hydrogen FCVs. The paper is organized as follows. Section 2 provides a brief overview of the history and technical specification of hydrogen ICE vehicles, Section 3 is a comparison of different vehicle technologies, Section 4 presents a scenario analysis of the economics of hydrogen ICEs, and Section 5 concludes.

2. Hydrogen in Internal Combustion Engines
Hydrogen-burning internal combustion engines trace their roots back to some of the very earliest developments in internal combustion engine development. Initially, gaseous fuels like hydrogen were preferred to liquid fuels like gasoline because they were considered safer to work
3

Kenneth Gillingham

Hydrogen Internal Combustion Engines

with, due to the low pressures used for the gaseous fuels and the quick dissipation of the gases in the event of a leak. In 1807 Issac de Rivas built the first hydrogen internal combustion engine, and although the design had serious flaws, it was a more than 50 years ahead of the development of gasoline internal combustion engines (Taylor 1985). Technological advances in gasoline engines, such as the development of the carburetor (which allowed air and gasoline to be consistently mixed), eventually led to other fuels being largely passed over in favor of gasoline. Until recently, hydrogen has been relegated to niche uses, such as in experimental vehicles or in the space program.
2.1 Properties of Hydrogen
There are several important characteristics of hydrogen that greatly influence the technological development of hydrogen ICE and FCVs.
Wide Range of Flammability. Compared to nearly all other fuels, hydrogen has a wide flammability range (4-74% versus 1.4-7.6% volume in air for gasoline). This first leads to obvious concerns over the safe handling of hydrogen. But, it also implies that a wide range of fuel-air mixtures, including a lean mix of fuel to air, or, in other words, a fuel-air mix in which the amount of fuel is less than the stoichiometric, or chemically ideal, amount. Running an engine on a lean mix generally allows for greater fuel economy due to a more complete combustion of the fuel. In addition, it also allows for a lower combustion temperature, lowering emissions of criteria pollutants such as nitrous oxides (NOX).1
Low Ignition Energy. The amount of energy needed to ignite hydrogen is on the order of a magnitude lower than that needed to ignite gasoline (0.02 MJ for hydrogen versus 0.2 MJ for gasoline). On the upside, this ensures ignition of lean mixtures and allows for prompt ignition. On the downside, it implies that there is the danger of hot gases or hot spots on the cylinder igniting the fuel, leading to issues with premature ignition and flashback (i.e., ignition after the vehicle is turned off).

1 The combustion of hydrogen and oxygen produces water as its only product, but the combustion of hydrogen with air also produces nitrous oxides (NOX), due to the high nitrogen content in air. Traces of carbon dioxide and carbon monoxide may also be present in emissions from seepage of engine oil.
4

Kenneth Gillingham

Hydrogen Internal Combustion Engines

Small Quenching Distance. Hydrogen has a small quenching distance (0.6mm for hydrogen versus 2.0mm for gasoline), which refers to the distance from the internal cylinder wall where the combustion flame extinguishes. This implies that it is more difficult to quench a hydrogen flame than the flame of most other fuels, which can increase backfire (i.e., ignition of the engine’s exhaust).
High Flame Speed. Hydrogen burns with a high flame speed, allowing for hydrogen engines to more closely approach the thermodynamically ideal engine cycle (most efficient fuelpower ratio) when the stoichiometric fuel mix is used. However, when the engine is running lean to improve fuel economy, flame speed slows significantly.
High Diffusivity. Hydrogen disperses quickly into air, allowing for a more uniform fuelair mixture, and a decreased likelihood of major safety issues from hydrogen leaks.
Low Density. The most important implication of hydrogen’s low density is that without significant compression or conversion of hydrogen to a liquid, a very large volume may be necessary to store enough hydrogen to provide an adequate driving range. Low density also implies that the fuel-air mixture has low energy density, which tends to reduce the power output of the engine. Thus when a hydrogen engine is run lean, issues with inadequate power may arise (College of the Desert 2001).
2.2 Relevant Trade-offs
Based on the above unique properties of hydrogen, there are several relevant tradeoffs pertinent to the use of hydrogen in ICEs.
The first relates to a decision that for the most part has already been made: whether to use a spark-ignition engine design (e.g., most gasoline vehicles), or a compression-ignition (CI) engine design (e.g., diesel vehicles). CI engines work by compressing air in the combustion chamber, increasing its temperature above the autoignition temperature of the fuel, such that injected fuel ignites immediately and burns rapidly. This small explosion causes the gas to expand and forces the piston down, creating mechanical energy that is be used to power the vehicle. Spark-ignited engines begin combustion at a much lower temperature and pressure through the use of an ignition system that sends a high-voltage spark through a sparkplug to ignite the fuel-air mixture.

5

Kenneth Gillingham

Hydrogen Internal Combustion Engines

Spark-ignition engines tend to be less expensive and have lower emissions of criteria pollutants (e.g., NOx and particular matter)2, but have lower power at low engine speeds and a lower theoretical efficiency than CI engines. Due to hydrogen’s wide range of flammability and low density, nearly all recent designs for hydrogen ICE vehicles call for CI engines (Ford 2007).3
A second relevant tradeoff is the type of transmission to use. Using hydrogen in a CI engine will most likely require the use of a continuous-variable transmission (CVT), as is commonly used in hybrid gasoline vehicles. The CVT may or may not be designed to be coupled with an electric battery and a separate electric motor that runs off recaptured energy from breaking. Here the tradeoff is between additional cost and improved fuel economy – although most recent hydrogen ICE designs include the battery and separate electric motor.
A third tradeoff is between power and fuel economy or emissions. Running a hydrogen engine lean reduces criteria pollutants and can improve fuel economy, but it comes at the cost of power due to the lower energy content of the fuel-air mixture. To ensure adequate power, turbocharging, super-charging, or not running the engine lean can all be used, but are likely to come at a cost of fuel economy and possibly criteria air pollutant emissions.
A final key tradeoff is between vehicle range and the hydrogen fuel tank size. Efforts are underway to improve storage of hydrogen in fuel tanks through compression or liquification of hydrogen, but the low density of hydrogen poses challenges to engineers attempting to decrease the tank size, yet ensure adequate range for hydrogen vehicles. Moreover, the hydrogen storage systems are likely to be heavier than standard gasoline tanks, increasing vehicle weight, which can decrease fuel economy.

3. Comparison of Vehicle Technologies
Table 1 presents estimates of some of the most important characteristics of the four most relevant types of vehicles: gasoline ICE, gasoline hybrids, hydrogen ICE, and hydrogen FCVs.
2 Recent technological advances have been successful in lowering criteria air pollutants for CI engines, albeit with higher manufacturing costs (Kliesch and Langer 2003). 3 Note that “diesel engine” is a general term applying to engines that work through compressed air ignition, so the CI engines described above could equally well be called diesel engines, and are often described as such. Diesel engines do not necessarily have to burn “diesel” fuel.

6

Kenneth Gillingham

Hydrogen Internal Combustion Engines

It must be emphasized at this point that many of these estimates, particularly on hydrogen FCVs are highly speculative due to the uncertainty in technology development, and the characteristics (e.g., size and weight) of vehicles that will be rolled out with each technology.
Hydrogen ICE vehicles tend to fall in a middle ground between the higher efficiency hydrogen fuel cell vehicles and the standard gasoline ICE vehicles. In many respects, hydrogen ICE vehicles can be thought of as diesel fuel hybrid vehicles that run off of hydrogen, rather than diesel fuel. Thus a critical difference between gasoline hybrids and hydrogen ICE vehicles is that the use of a CI engine design allows for greater engine efficiency: on the order of one third greater. Moreover, how engine efficiency varies with load and power differs between the engine types. Figure 1 provides a rough sketch of the relationship between engine efficiency and percent load for spark-ignition, compression-ignition (CI), and a single fuel cell (with equivalent output to the other engine types).
Spark-ignition engines have a maximum efficiency of 32.5% under normal conditions and at low loads have a much lower efficiency than this. Note that the additional electric engine in gasoline hybrid vehicles is highly efficient at very low percent loads, and is primarily used at low load levels, so gasoline hybrids do not suffer from this loss in efficiency at low loads as much. Compression-ignition engines tend to have a maximum efficiency rough in the range of 40%, and quickly reach efficiency levels close to the maximum efficiency at low percent loads. The greater maximum engine efficiency is in large part the reason why diesel vehicles have better fuel economy than conventional vehicles.4
A typical fuel cell stack can reach much higher efficiencies than either spark-ignition and CI engines, but it is important to note that as the fuel cell stack reaches maximum load, the efficiency drops precipitously, in contrast to the other engine types. The exact shape of this curve, and any quantitative estimates of fuel cell efficiency are highly speculative due to the many recent developments in fuel cell technology, but the general shape is robust (Edwards 2006).

4 An evaluation of 24 matched pairs of diesel to gasoline light duty vehicles in Europe and the United States found that indirect-injection diesel vehicles had 24% better fuel economy on average and turbocharged, direct-injection diesel vehicles averaged 50% better fuel economy, although much of that is due to the turbocharging (Schipper, Marie-Lilliu, and Fulton 2002)
7

Kenneth Gillingham

Hydrogen Internal Combustion Engines

This relationship has important implications for the power delivered to fuel cell vehicles, for additional fuel cells must be added to provide adequate power for some high-intensity uses – and the fuel cell stacks are one of the most expensive components of a fuel cell vehicle. Figure 2 indicates the relationship between power train efficiency and power in one particular study. As each of the fuel cell stacks incrementally reach 100% load, efficiency begins to drop.
This relationship may reduce the possibility of fuel cell heavy duty vehicles, which need to be able to provide sufficient power at high loads. Hydrogen ICE vehicles may be more economically attractive in these markets, since to the high cost of adding more fuel cells may make fuel cell vehicles prohibitively expensive. Of course, the exact relationship between power and efficiency depends on many factors relating to the specific application.
The rough estimates of the average and maximum engine efficiency in Table 1 follow from the discussion above. Equally important as engine efficiency is the efficiency of the transmission in converting the energy generated by the engine to propulsion. Gasoline hybrids, hydrogen ICE vehicles, and hydrogen fuel cell vehicles are all assumed to use CVT and hybrid transmission technology, which has approximately 60% efficiency, as opposed to a standard transmission, which has only around a 40% efficiency. Given these estimates and an estimate of the current average fleet-wide fuel economy of standard gasoline light duty vehicles, the fuel economy of each of the vehicle types is computed.5 These computed estimates for gasoline hybrids and hydrogen fuel cells match closely with those in NRC (2004).
Table 1 also highlights differences in engine sizeability, fuel tank size, cost of fuel, and emissions. All of these have either direct or indirect importance to the market feasibility of each vehicle type. The cost of hydrogen depends on the feedstock, as will be discussed in section 4, but there may even be a minor difference between the cost of hydrogen in ICE vehicles and fuel cell vehicles. Nearly all hydrogen fuel cells under development require very pure hydrogen to

5 Specifically, the total vehicle efficiency for each type is first computed by multiplying the engine efficiency by the transmission efficiency. Then, for gasoline hybrids, hydrogen ICE vehicles, and hydrogen fuel cell vehicles, the current gasoline ICE fuel economy is multiplied by the ratio of each vehicle type’s efficiency to the gasoline ICE vehicle efficiency. This methodology assumes that unobserved determinants of fuel economy change proportionally with vehicle efficiency.
8