The internal-combustion engine was synonymous with the automobile throughout the 20th century. But its future is now at risk, since it faces competition from both the hybrid gasoline- and diesel-electric engine (that is, the hybrid) and from the fuel cell.¹ As befits any new technology, fuel cells (see sidebar "Challenge or opportunity?") and hybrids are attracting heavy investment and media attention. You would think that the internal-combustion engine had nowhere to go but out.

The internal-combustion engine will be the dominant power plant in new cars sold until at least 2025

But today’s internal-combustion engine is far more advanced and efficient than its predecessors. Over the past 20 years, automakers have significantly improved its power, its fuel efficiency, and its emissions, with more changes to come. Not that it will always outperform the alternatives; fuel cells—rapidly gaining market acceptance and slated to be in mass production for some premium markets by 2010—may become the leading technology of the late 21st century. However, given the current economics of the internal-combustion engine, we predict that it will still be installed in 90 percent of all new vehicles sold in developed economies in 2015 and remain dominant in new vehicles for at least another decade after that, both as a stand-alone technology and as an integral part of hybrids.

Only regulation could facilitate a quicker transition to the fuel cell. Indeed, environmental concerns about greenhouse gas emissions such as carbon dioxide and the geopolitical desire for energy independence may accelerate the demise of the internal-combustion engine, for governments could enact policy reforms to favor the development of alternative technologies² and their adoption by the consumer.³ But until the consumer is ready to embrace them, most governments are unlikely to accept the political risk of radical reform.

The lay of the land

Until the late 1960s, business economics and perceived consumer values shaped the automotive industry’s power plant choices. Carmakers could choose from a range of technologies.⁴ The value to the consumer of each of them was determined by its fuel and cost efficiency as well as its safety, durability, and ease of use. The convenience of the supporting infrastructure available for the internal-combustion engine was another very important consideration in the power plant choices of the automakers. In addition, increasingly strict emissions regulations have been influencing their priorities since about 1970.

A closer examination of these issues—technology, infrastructure, and emissions regulation—can help make it possible to forecast which technology will power cars in the coming decades. Other factors too may eventually be important. Fuel cells, for example, may have unique advantages for what auto engineers call "packaging": since they don’t need an engine bay, they offer greater freedom in styling and structural safety. For the time being, however, the three fundamental issues will prove decisive.

Here we focus on only two technologies: fuel cells and gasoline- or diesel-powered internal-combustion engines. Hybrids—the third contender—are clean and fuel efficient and have a valuable role to play in the near term, but they sacrifice performance and raise costs, since two separate technologies must be integrated and controlled.⁵ Although hybrids are in compliance with today’s lower emissions targets, only the fuel cell can power the zero-emission vehicles (ZEVs) that regulation will require in ever-increasing numbers.

Technology

At the beginning of the 1980s, the average horsepower per liter of cars in the US market had been drifting for 25 years—since the introduction of the high-compression engine, in the mid-1950s. A complacent industry was making few efforts to improve the underlying technology. But in the early 1970s, pressure for improved efficiency and emissions performance rose sharply. The US Clean Air Act as amended throughout the 1970s embodied in law the environmentalists’ demand for stricter emissions rules. Furthermore, the Arab oil embargoes of the 1970s squeezed the fuel supply and drove the need for more efficient engines.

Early efforts to meet new efficiency and emissions requirements succeeded, though at the cost of a huge erosion of power, drivability, and overall performance. But breakthroughs such as electronic engine-control systems and catalytic converters enabled the internal-combustion engine to more than double its average horsepower per liter, from 29 in 1980 to 64 in 2002, at a significantly lower cost, and to reduce its emissions sharply. In 1986, the engine of an entry-level car accounted for more than 15 percent of its total production cost. That figure has dropped to 8 percent today, even though engines now use more expensive materials (such as aluminum) and components. These developments represent a new and continuing S-curve in the internal-combustion engine’s evolution (Exhibit 1).

Indeed, the technology has come a long way, and automakers are committed to improving its power capacity, fuel efficiency, and emissions still further. Projections suggest that in these respects, internal-combustion engines will continue to gain at a rate of 1.5 percent annually—an impressive pace for a century-old technology and well in line with current R&D investment. (In general, on the contrary, returns to R&D investment fall over time.) During the next ten years, several other advances are expected, including continuously variable transmissions, infinitely variable engine-valve timing, direct fuel injection, cylinder deactivation (Exhibit 2), and drive-by-wire technologies (see sidebar "AUTOnomy raises the stakes").

In the past five years, the number of internal-combustion-related patents issued by the US Patent and Trademark Office has gone up 25 percent, a huge leap compared with the incremental increase in the number of such patents granted over the previous two decades. This upsurge suggests that innovation in the field isn’t in danger of slowing down. The fact that automakers continue to support such R&D should come as no surprise given their enormous investment in the technology.

How do fuel cells compare with the internal-combustion engine in raw performance? At the heart of a typical hydrogen fuel cell lies a proton-exchange-membrane⁶ (PEM) stack that electrochemically converts hydrogen and air into electricity and water (Exhibit 3). This elec-tricity directly powers the car’s electric motors and accessories. Depending on how efficiently the hydrogen is produced, fuel cells not only are clean "at the tailpipe" but also tend to use fewer resources along the whole chain, from the production of fuel to the turning of a car’s wheels (Exhibit 4). Fuel cells also have other potential advantages, such as instant-on torque response, less noise, and cheaper maintenance. In addition, fuel cells are more efficient because they generate electric power directly, so they will be well suited to cars that have increasing numbers of electrically powered features: the 2002 BMW 7 series, for example, has nine temperature-control fans just in the driver’s seat. The internal-combustion engine, by contrast, drives an alternator to meet a car’s electrical needs and incurs "parasitic" losses in efficiency by mechanically driving accessories such as power steering.

Nonetheless, internal-combustion engines are currently well positioned, technologically and economically, to outperform the fuel cell in powering vehicles. Although the fuel cell was commercialized at General Electric in the early 1960s for military and aerospace applications, current prototypes are still expensive producers of energy, and the reliability and durability of the PEM generate concerns, especially under real driving conditions. Despite the rapid development of fuel cells, they are still prohibitively expensive to produce if the goal is to match the range and performance of conventionally powered cars. Depending on the manufacturer, current estimates for the cost of PEM fuel cell prototypes range from $500 to $2,500 per kilowatt produced, which is still a figurative mile behind the internal-combustion engine’s $30 to $35 per kilowatt. But ten years ago, the cost of experimental PEM fuel cells probably exceeded $50,000 per kilowatt produced, and marked improvements in the underlying technology since then have captured the interest of the industry, not to mention an estimated $3 billion-plus in investments through 2004.

Another hurdle now being overcome is the amount of space needed for a fuel cell that can power a car, because the size and weight of the cell affects its performance and utility. The one in DaimlerChrysler’s 1994 "concept car" NECAR (New Electric Car) 1 filled the rear of a van, leaving room only for the driver and a single passenger. Six years later, the NECAR 5 power plant fit neatly within the Mercedes small A-Class engine bay and could power vehicles at speeds greater than 150 kilometers (90 miles) an hour.

Motoring infrastructure

A well-established infrastructure for fuel and repair services is vital for any driver. The internal-combustion engine clearly has the advantage here, for developed economies provide ready access to these services. The hydrogen fuel cell faces one of its greatest challenges in precisely this arena, since it lacks an infrastructure for its upkeep and maintenance. The creation of such facilities poses several potential problems. Building hydrogen storage facilities at filling stations (or the stainless-steel tanks needed for convertible methanol) and manufacturing tankers to supply those stations will require billions of investment dollars, for example. Experts predict that the infrastructure will develop gradually, beginning with large stations for centrally fueled fleets (of city buses, to give one example) and then moving to more dispersed and consumer-friendly locations, while existing gasoline stations are slowly converted to the fuel cell technology and new outlets are constructed to service it. Appropriately trained technicians and equipment must also be made available everywhere drivers might need them.

Because hydrogen doesn’t exist in a natural form that can be tapped, the generation of the vast quantities necessary to supply power to a large automobile market is also problematic. Energy- and emissions-efficient methods of extracting hydrogen from other compounds and of converting it for onboard use remain elusive. Solar-powered "farms" to extract hydrogen from water via electrolysis have been suggested but are not yet practical. Fuel cell vehicles also pose their own potential safety hazards: given the volatility of hydrogen gas, for example, stringent universal safety regulations must be imposed for storing, handling, and disposing of it.

The alternatives to a hydrogen gas infrastructure are equally troublesome. Onboard fuel reformation processes—which convert conventional hydrocarbon fuels such as natural gas or methanol into hydrogen—would require each car to contain all the essential elements of a small refinery. The increase in size, weight, complexity, emissions, and costs would further diminish the ability of fuel cells to compete with other technologies. Moreover, even if an onboard cryogenic tank could store liquid hydrogen at its vapor point (−253°C), the cost, the risk of accidents, and the problem of refueling would all present serious obstacles.

Indeed, the cost of deploying a reliable hydrogen infrastructure on par with current gasoline networks has been estimated at $100 billion and more. Unless governments subsidize the development of such an infrastructure, it is quite hard to imagine fuel cells competing economically with the internal-combustion engine in the foreseeable future.

Emissions and regulations

Emissions regulation is the Achilles’ heel of the internal-combustion engine. Carbon dioxide, the primary greenhouse gas, is an unavoidable by-product of fossil fuel combustion, whether the engine uses gasoline, natural gas, or diesel fuel or is an electric hybrid. If the public were convinced of the environmental dangers posed by air pollution and global warming, or of the geopolitical risks of an overreliance on fossil fuels, the technology could be regulated out of existence. The cleaner and quieter fuel cell is far better from an environmental point of view.

If pure hydrogen powers fuel cells, they emit almost no hydrocarbons, carbon monoxide, carbon dioxide, or nitrogen oxides. What carbon dioxide emissions there may be are by-products of the steam reformation of natural gas, currently the cheapest way to produce hydrogen.

By 2000, late-model automobiles emitted less pollution while running than automobiles did back in the 1970s while they were turned off

Although the internal-combustion engine generates far higher exhaust and evaporative emissions, the auto industry, despite considerable difficulties, has proved remarkably effective at reducing many of them (Exhibit 5): other than the greenhouse-enhancing carbon dioxide, they have fallen by 90 percent or more since 1968.⁷ In fact, by 2000, late-model cars emitted less pollution while running than 1970s-era cars did while turned off (large amounts of gasoline vapor leaked from old models).

Today’s safer, cleaner, and more efficient vehicles have been the result of the regulators’ willingness to impose restrictions and of the carmakers’ ability to respond to them. Reengineered catalyst technologies and new close-coupled high-flow exhaust-gas recirculation will further reduce emissions. What is more, BMW and Mazda are working to adapt internal-combustion engines to use hydrogen fuel, so they may eventually be able to piggyback on breakthroughs in techniques for storing it—a development that would reduce their emissions almost to fuel cell levels and further prolong their ascendancy, though with penalties in efficiency.

Nonetheless, this regulatory wave could be reaching its crest. Further restrictions may be beyond the automakers’ capacity to meet at a reasonable cost. Particularly in large metropolitan areas, the internal-combustion engine is facing a raft of proposed regulations to limit emissions; the possibilities include banning it from city centers and imposing special taxes for vehicles fueled by hydrocarbons. The California Air Resources Board (CARB), for example, has required high-volume automakers to sell a percentage of zero- or near-zero-emissions vehicles in the state by 2003. Because of the immaturity of pure-electric-vehicle technologies, CARB withdrew two previous ZEV phase-in milestones and has relaxed the 2003 targets. Yet carmakers might still miss the final deadline, thereby exposing themselves to millions of dollars in potential fines.

Most consumers support efforts to cut emissions and conserve fuel, but fewer would sacrifice cost, performance, or convenience

In April 2002, California became the first US state in which a bill restricting carbon dioxide emissions from automobiles was introduced. Will regulators go further and impose the ZEV standard on all automobiles or enforce carbon dioxide emissions limits that the internal-combustion engine can’t meet? Barring a dramatic shift in the level of consumer concern for the environment, these scenarios seem unlikely. Surveys show that while a majority of consumers support efforts to reduce emissions and conserve fuel in principle, fewer are willing to sacrifice cost, performance, or convenience.⁸ Any attempt to regulate the internal-combustion engine out of existence, it seems, would proceed very slowly.

Besides complacency, the major constraint on regulation is the potential loss to governments of revenues from fossil fuel taxes. This problem, in addition to the need to subsidize the hydrogen supply chain, may place an intolerable fiscal burden on those governments, in the developed world, that are thinking about using regulation to accelerate a switch to fuel cells before consumers have made that choice.⁹

Given the many advantages of the internal-combustion engine, it will remain the dominant power plant well into the present century, both as a stand-alone technology and in gasoline- and diesel-electric hybrids. Its tremendous capacity for improvement means that its competitors should take a long time to catch up or even to assume a strong position in the automobile market. Developing countries, which have less onerous greenhouse gas restrictions, will likely embrace the best available internal-combustion technology rather than confront the cost and infrastructure obstacles of alternative power.

Well into the middle of this century, more cars around the world will be propelled by the internal-combustion engine than by any other power source. These cars will require all the fossil fuel and maintenance support currently in place. While the fuel cell is an up-and-coming technology, its advantages are being realized more slowly than many had hoped. A brash leap into a fuel cell world is risky and, at present, unlikely. A well-planned transition will avoid a premature launch, a disappointed public, and a fallback to industry and environmental complacency.