Why Not a 40-MPG SUV?
Technology exists to double gas guzzlers’ fuel efficiency. So what’s the holdup?
To get a sense of the auto industry’s progress in fuel efficiency, look no further than the 2002 Chevy Blazer. The model with automatic transmission, six cylinders, and four-wheel drive gets 18 miles per gallon (mpg), two miles less than a comparably equipped Blazer did in 1985. Indeed, in those 17 years the average fuel economy of the entire fleet of U.S. cars and light trucks declined from 26 mpg to 24 mpg-in part because of the rising proportion of gas-guzzling sport-utility vehicles (SUVs). Yet in March, when auto industry lobbyists claimed that building more fuel-efficient cars would be “too difficult,” the U.S. Senate once again killed legislation that would raise the country’s Corporate Average Fuel Economy standards. It was a familiar dance; Congress has not raised the standards even once during those same 17 years.
It’s not that automotive technologies haven’t improved; it’s that the improvements have been geared toward delivering power, not efficiency. Since 1981 the auto industry has hiked horsepower 84 percent, allowing vehicles to accelerate faster even though they have gotten heavier, according to the U.S. Environmental Protection Agency. “That’s what consumers want,” says Fritz Indra, executive director of advanced engineering for General Motors’ Powertrain division. “Each year Americans want a little more space inside, a little more power.”
But is it really too difficult to build a reasonably priced SUV that can get 40 mpg and still provide the performance, comfort, and reduced emissions consumers expect? The surprising fact is that an assortment of fuel efficiency technologies exist in industry and university labs. Even more startling is that many of these technologies are based on the conventional internal-combustion engine. They don’t require complex electric-gas hybrid drive trains like those under the hoods of the Toyota Prius and Honda Insight (see Visualize). Nor are they based on anything as exotic as fuel cells. If the automotive industry put some corporate horsepower behind moving these technologies into production-and that’s a big if, given the lack of U.S. regulations and consumer demand-the gas-saving technologies could start hitting showrooms within five years. Indeed, if it chose to, Detroit could manufacture a 40-mpg SUV by the end of the decade.
The gains would come largely from emerging technologies such as improved control systems that minimize energy losses in the engine and transmission, as well as efficient electrical components-from water pumps to engine valves-that could replace belt-driven mechanical systems. Existing technologies, such as advanced transmissions and fuel injection systems, could also play key roles if they were adopted more widely.
Indeed, if all new cars and light trucks adopted available and emerging gas-saving technologies, the average fuel economy of U.S. cars would surge to 46 mpg, up from today’s 27 mpg. And SUVs could average 40 mpg, up from today’s 21 mpg, according to a recent study prepared in part by John DeCicco, a senior fellow at Environmental Defense, a New York Citybased environmental group. (The study was coauthored by Feng An, a modeling expert at Argonne National Laboratory, and Marc H. Ross, a physicist and automotive-policy expert at the University of Michigan.) Two-thirds of the benefit would come from improving the power train, and the rest would come from cutting weight and reducing aerodynamic drag and rolling resistance. And even though retail prices of vehicles would rise some $1,000 to $2,000, depending on the model, consumers would save that much at the gas pump within five years. “The industry doesn’t lack the technology, it lacks the priority,” DeCicco says.
Such improvements in gas mileage would have a huge impact on U.S. oil dependence and the environment. According to the Union of Concerned Scientists, if the U.S. fleet’s fuel use improved to 40 mpg, the nation would save two million barrels of oil a day-75 percent of all the oil the United States imports from the Middle East. And it could mean a 30 percent decrease in greenhouse gases, chiefly carbon dioxide.
Automakers-while not debating the essential truth of such numbers-say that reliable, affordable versions of these new components and software controls are harder to implement than it may seem. But the manufacturers, while characteristically tight-lipped about production plans, have created advanced prototypes of these technologies and even installed early units in some vehicles. Because many of these technologies are readily available and based on the internal-combustion engine, they could have tremendous impact in the next several years. “There’s a lot of potential here,” says John Heywood, director of the Sloan Automotive Laboratory at MIT. “It’s our best hope for continuing to reduce emissions and fuel consumption of our ever growing vehicle fleet.”
Greater fuel efficiency starts with some seemingly simple ideas, for example, shutting down the engine to eliminate wasteful idling whenever the car is not moving. But for an engine to shut down at every red traffic light, the vehicle would need a high-power device able to restart the engine immediately-far faster than a traditional starter motor-when the driver taps the gas. Sounds straightforward. But early versions were generally too slow or too noisy to keep drivers happy.
That’s changing. Several auto suppliers have built prototypes of integrated starter-generators, which replace both the starter motor and the alternator, that are fast enough to crank the engine in less than half a second (see “An End to Idling,” below). “The big issue is reducing cost,” says Thomas Keim, an MIT electrical engineer who directs an MIT-industry consortium on advanced automotive electronics.
Illustration by John MacNeil
Most designs use induction motors, which must be controlled by expensive electronics that can rapidly switch between the starter-generator’s dual roles of starting the engine and acting as the alternator to generate electricity. But the MIT consortium has developed a starter-generator with a simpler version of the electronics, which could reduce the cost by 20 percent. “In the automotive world, a technology that is 20 percent cheaper tends to drive the more expensive choice out of the market,” Keim says. Ford has built a prototype of the consortium’s design and has completed initial tests. But there’s another roadblock: the difficulty of generating enough power to keep such devices humming. Rapid cranking takes lots of power. Like most other experimental starter-generators, the MIT device, built in anticipation of a future where cars employ a higher voltage standard, operates at 42 volts. Trouble is, virtually all of today’s cars still use a 12-volt system. (Exceptions include gas-electric hybrids and a 42-volt luxury Toyota sedan sold only in Japan.) Tethering a starter-generator to a 12-volt system is theoretically possible, but the starter-generator is just one of a growing number of advanced automotive electrical components on industry drawing boards. Taken together they exceed the system’s limits. Some industry experts say the voltage shift will take years: replacing an entire electrical system-which in a new car generally costs as much as the engine and transmission combined-is an expensive and complex task. But even without a complete replacement, a stopgap approach could provide a voltage boost. Manufacturers are crafting ways to install 42-volt systems beside existing 12-volt systems. The existing low-power system would remain in place, continuing to provide electricity to familiar light-duty equipment such as lights, radios, power seats, and window motors. The new high-power system would serve only heavy-duty equipment such as starter-generators and electric compressors. The dual system “would obviously add cost,” says Xingyi Xu, an engineer at Ford Research Laboratory in Dearborn, MI. Automakers would need to see considerable gains in fuel economy and consumer conveniences, he says, so “at the present time it is difficult to make that justification.” Even if justifiable, such systems would take at least five years to reach market, Xu says. But because they can provide the power needed to transform the average car from a mechanical into a more efficient electromechanical machine, 42-volt systems represent an enabling technology.
Souped Up Software
New equipment such as powerful starter-generators would raise fuel efficiency. Yet a vehicle could achieve even greater gains if the entire power train were controlled electronically.
Each component could be adjusted continuously to consume the least power necessary as driving conditions change and-equally important-could be controlled in an integrated way for systemwide savings. “Fuel efficiency would be even greater than the sum of the components,” says Frank Lohrenz, an electrical engineer at Siemens VDO Automotive in Regensburg, Germany.
The savings could be significant. Integrated software-control systems could provide a 10 percent fuel-efficiency boost (see “The Networked Car,” TR September 2002). Siemens’ software, for example, optimizes the delivery of torque-the turning force of the power train. To do this, the system electronically registers how far and how fast the driver pushes the gas pedal. Then through electronic control of such basic mechanical components as the engine, transmission, and a future starter-generator, it delivers the requested torque. The Siemens’ technology takes into account 20 parameters-including vehicle speed, engine rotation speed, and transmission gear-before deciding how best to deliver torque from the combined efforts of engine throttling, transmission gear ratio, and starter-generator activation.
Engineers have traditionally considered each component as a stand-alone unit, but there is plenty of interest in integrated control because it’s cheap: integrated control depends largely on software, and thus, its 10 percent fuel-efficiency boost comes at relatively low cost. “If a manufacturer sells one million cars,” Lohrenz says, “the cost could be less than $5 per vehicle.”
The most radical advance for improving gas mileage would come about by remaking the engine itself. This means rethinking the century-old mechanics that open and close the engine valves, which let a fuel-and-air mixture into combustion chambers and release exhaust gases. For decades, a camshaft has performed this job. A spinning shaft, it moves levers that open and close valves approximately 100 times per second in a fixed pattern.
The camshaft technology works well, but it wastes fuel. The traditional configuration provides no way to change patterns in order to deliver, for example, lots of power for accelerating onto a highway and to cut back on unneeded power-saving fuel-at highway cruising speeds. In recent years, though, engineers have added mechanical equipment to the camshaft, allowing some enhanced valve control. This control includes, for example, the ability to open valves only partway when little power is needed. Honda and BMW have developed and installed such “variable valve” systems in many production cars, improving fuel economy by 5 to 10 percent.
But the ultimate move toward optimization throws the camshaft away. Instead, electromechanical actuators would provide software-driven control for each valve (see “The Camless Engine,” below). By providing full control over the timing, lift, and duration of each valve motion, such a camless engine optimizes power delivery with the least possible fuel at every engine-rotation speed. The payoff is huge: a camless engine could improve fuel economy by 10 to 18 percent while also increasing engine torque by 15 to 20 percent at low speeds for faster acceleration.
Illustration by John MacNeil
The problem is that to prevent excessive wear and minimize engine noise and vibration, valves must decelerate before landing. A camshaft, though relatively inefficient, does this quite well, thanks to its ovoid shape, which produces a corresponding acceleration and deceleration in the valve motion. Actuators are different; they slam up and down, on and off. The way to make actuators as gentle as camshafts involves a combination of hardware and software, and many companies are working on the problem. Anna Stefanopoulou, a mechanical engineer at the University of Michigan, has already designed several promising software schemes. Within the past year Stefanopoulou’s team has optimized several algorithms and is now testing ways to use feedback from the valves to achieve high-speed motion with gentle landings. Meanwhile, to aid in this soft landing, Mohammad Haghgooie, a physicist with Ford Motor, is testing springs and pneumatic and hydraulic dampers to lessen the impact of valves without slowing them down. If successful, these improvements in electromechanical valve software and hardware could bring a camless engine to market in 2008, Haghgooie says.
Brake for Progress
Not all of the advanced fuel-efficiency technologies are still in the emerging stage. Even without camless engines and sophisticated software, assorted technologies for achieving better fuel efficiency are available. The list includes the “continuously variable transmission.” Unlike today’s automatic transmissions, which generally have four fixed gear ratios that clunk into place once engine rotation speed increases to a certain level, a continuously variable transmission delivers any of an infinite range of gear ratios on the fly. A Dutch company patented the technology decades ago; now the patents are expiring, and the transmission is already being installed in some models in the United States. The payoff can be big: in the 2002 Saturn VUE, the continuously variable transmission boosts fuel economy 7 to 11 percent, according to General Motors.
Improvements in fuel injection are also on the shelf, thanks to a recent advance known as “gasoline direct injection.” By replacing the traditional indirect-injection engine with this technology, the 2002 Volkswagen Polo has improved fuel economy for city driving by 13 percent. The benefit comes from exploiting the dynamics of how fuel and air mix. In the traditional indirect injection setup, gas and air are mixed outside the cylinder and then injected. With direct injection, fuel and air begin mixing only when they are inside the cylinder, enabling the engine to use an ultralean fuel mixture during steady, low-power driving.
All in all, there is no shortage of technology available and almost ready for the auto industry to adopt. And yet, SUVs still get an average of only 21 mpg. Asked why, General Motors’ Indra cites familiar industry arguments: innovations are too expensive; new components add weight, negating benefits. He says also that weight reduction-which, according to the DiCicco study, accounts for nearly one-third of the formula for boosting mileage-cuts into safety. That’s the argument the industry used as part of its lobbying blitz to kill tougher fuel-efficiency legislation last March.
Roland Hwang, a vehicles expert at the National Resources Defense Council, an environmental group based in New York City, says that argument is “irresponsible.” He claims the auto makers are fueling consumers’ fears about safety only to persuade them to buy bigger vehicles, which, he says, yield the highest profits. He notes that federal and insurance industry tests show that the safety record of SUVs is about the same as that for other cars. Even Honda America’s manager of environmental and energy analyses, John German, agrees that “if all vehicles weighed 100 pounds less, there would be no impact on safety.”
The larger point is simply that with no mandate from Washington or the public, the auto industry has little motivation to change. Doug Patton, a senior vice president at Denso International America in Southfield, MI, puts the subject into perspective: “What is the customer demanding? What is the government requiring? That’s how we look at it.”
To researchers, this is discouraging talk. MIT’s Heywood says most of these technologies have been in development for years. If the automakers wanted to, he says, they could readily make them inexpensive and reliable. “The car companies don’t give their engineers enough credit for being able to solve practical problems,” Heywood says. “Until management says, Okay, let’s really go for it,’ the technology doesn’t get past an advanced development prototype.” Giving such orders, he adds, “won’t happen until management thinks it has evidence that the technology will make the product sell in the marketplace or will create a new marketplace.”
Even if tough new efficiency laws are passed, others note, recent history suggests the auto industry won’t accede without a fight. “Industry leaders fought catalytic converters. They fought seat belts. They said air bags would bankrupt the industry. But once the requirements are passed they find a way,” Hwang notes.
For now, the auto industry is still content to fill showrooms with perennial gas hogs. But more efficient technologies-and the software to control them-are waiting for that final push into mass production.
Become an MIT Technology Review Insider for in-depth analysis and unparalleled perspective.Subscribe today