How to Make Fewer Trips to the Pump

With gas nearing $2 a gallon, smart driving can save big bucks

America's devil-may-care attitude about burning gasoline may finally be changing. With gasoline prices pushing $2 a gallon in many regions, consumers are paying closer attention to the fuel-economy ratings on the window stickers of new cars. That's an about-face from the 1990s, when gas cost around $1.20 a gallon and people nonchalantly bought millions of thirsty pickups and sport-utility vehicles.

There's one problem: Driving conditions have to be near-perfect to get the mileage that manufacturers promise. Oh, the Environmental Protection Agency works hard to get those certified ratings close to real-world driving performance. But they don't account for several factors, such as the actual speed at which people drive, the frequency of air-conditioner use, the cleanliness of gasoline at most filling stations, and the growing traffic congestion in major cities--all of which hurt fuel economy.

The test works like this: Auto manufacturers put a vehicle on a dynamometer, which simulates driving conditions while the car remains stationary. The vehicle runs at an average of 48 miles per hour to represent highway performance and 20 mph for city driving. But the test assumes optimum driving conditions, such as a passive driver who doesn't speed and the purest gasoline. The agency takes the results of the auto-company tests and verifies them using dynamometers in its own laboratories. The EPA then subtracts 10% from the city miles-per-gallon rating, 22% from the highway results, and 15% from the average, to take into account road conditions that can hurt fuel efficiency.

Agency officials acknowledge that they are pleased if certified mileage is within 2 or 3 mpg of real-world performance rates. And auto makers concede that they use the EPA-approved numbers as more of a rough measure of fuel economy than a precise one. "Most people don't drive to replicate the numbers that are on the sticker," explains Mark Kemmer, a senior Washington representative for General Motors (GM ). "Our people think the EPA's numbers are relatively representative."

No one tracks how close the certified numbers get to real driving performance. But using a random sampling of vehicles, I found the EPA mileage, indeed, to be 2 to 3 mpg higher than actual performance. Sounds small. But take the 1999 Jeep Grand Cherokee owned by one of my colleagues. His SUV, which is certified at 16 mpg, has consistently gotten 13 mpg--a 19% discrepancy. I also tested an Acura MDX sport-utility vehicle. It is certified at 19 mpg, but I got 16.7 mpg over 280 miles of driving. Meanwhile, a Dodge Durango 4x4, certified at 14 mpg, averaged 11.7 mpg when I tooled around in it.

HYBRID TRIUMPH. When I drove the Honda Insight and Toyota Prius--both eco-friendly cars that run on a combination of gasoline and electric power--I got great fuel economy. Still, the cars' real-world fuel-economy results on the dashboard display were about 8 mpg short of the 64 mpg that Honda boasts and 7 mpg short in the 48-mpg Prius. In fairness, a Lincoln LS sedan I tested was right on the nose at 19 mpg.

The reasons for the discrepancies aren't hard to fathom. Vehicles are tested as if people spend 55% of their time driving in the city and the rest on the highway. But the Sierra Club estimates that people drive in cities about 75% of the time. "The test is flawed," says Sierra Club fuel-economy guru Dan Becker, who points out that people drive at about 13 mph in cities--indicating more fuel-guzzling, stop-and-go driving--not the 20 mph the EPA assumes. Also, all those luxury gadgets and options add weight to a vehicle, reducing fuel economy, notes GM's Kemmer. Yet both carmakers and the EPA rarely test loaded models.

On the highway, the EPA assumes a top speed of 60--which most drivers routinely exceed unless there's heavy traffic. In fact, just 21 states still have a speed limit as low as 55 mph on urban interstates, and every state except Hawaii allows cars to travel above 55 on rural interstates. Fuel economy drops precipitously when a car is driven faster than 55 mph.

Another flawed EPA assumption: the car's average startup temperature. The EPA assumes that temperature is 65 degrees when it's really in the more gas-guzzling 50-degree range. Fuel efficiency also falls fast when you turn on the air conditioning. Yet both car companies and the EPA test cars with the AC off. Air conditioning used to be a luxury option, but today, 98% of new cars are equipped with it. "I'm afraid people are using air conditioning far more often than we think," says Jeff Alson, senior engineer at the EPA's National Vehicle & Fuel Emissions Laboratory in Ann Arbor, Mich. Using the AC during stop-and-go city driving can cut fuel economy by 20%.

Other conditions just can't be mimicked in test labs. Wind is a big drag on fuel economy, especially on the highway, says the EPA's Alson. And while carmakers and the agency use the cleanest, best-performing fuel, water gets into gas at most filling stations, further cutting performance. If you live in a region where specially treated gasoline is mandated as a clean-air measure, that takes its toll as well.

EPA officials say they want to look into several factors that can hurt fuel economy, including how often people use the air-cooling systems in their cars. The agency also plans to study traffic congestion and how often people drive in the city vs. highways. If driving habits continue to change in ways that eat into fuel economy, the EPA could change its procedures to reflect lower real mileage, an agency official says.

What are the lessons from all of this? To meet or beat your car's fuel-economy ratings, you have to simulate the testing standards. For example, don't floor the accelerator when the traffic light turns green. Don't go much faster than 55 mph on highways. Avoid driving in congested cities. And hope for a cool summer, because if you really want good gas mileage, you'll have to do without the AC.

By David Welch

    Before it's here, it's on the Bloomberg Terminal.