How do you compare engines according to fuel economy?
April 1, 2005
TAMPA, Fla. Comparisons of the fuel economy provided by different engines can require more than simply downloading reports from Electronic Control Modules (ECM). That's because each manufacturer track...
TAMPA, Fla. Comparisons of the fuel economy provided by different engines can require more than simply downloading reports from Electronic Control Modules (ECM). That’s because each manufacturer tracks related figures in a different way.
Herman Miller, fleet equipment manager for Shopko Stores, referred to his own ECM reports that suggested one truck’s fuel economy was 6.9 per cent better than it was in reality, and another model’s fuel economy was five per cent worse.
“You might not drive the right conclusion because that’s a 10, 11, 12-point spread,” he said of the figures that could encourage a fleet to choose one piece of equipment over another. “And using the ECM gallonage (tends to be) friendly to the engine manufacturer.”
At the very least, fleets need to be consistent in the approaches they use when comparing fuel economy, speakers said. And the work could begin with an understanding of the way each ECM tracks fuel-related figures.
Where one ECM will begin counting idle time after five minutes, another will count idle time after 15 seconds, said Caterpillar’s Bob Wessels. And the idling percentage can be calculated by dividing idle time by the truck’s running time or the engine’s entire running time.
The simple count of consumed fuel can vary five per cent between different engine designs, he added, noting that numbers can be plus or minus 0.1 kilometres per litre at the best of times.
Even counting the number of tire revolutions per kilometer can lead to a variable of plus or minus two per cent, because the diameter of a tire will wear down with use. (Some fleets set their ECMs to recognize a tire size that’s half way between a new tread depth and the point at which the tire is pulled for re-treading.)
“There are lots of variables to input correctly,” Wessels said, adding that a count of pump strokes and the number of kilometres recorded by a map will offer the best comparison.
“In the good old days, two-truck tests probably got the right answer,” added Chuck Blake of Detroit Diesel. But that could be a source of trouble with some of today’s designs. Some engines will improve fuel economy when they warm up, while others are better at slower speeds.
He recommended the Technology and Maintenance Council’s Type 4 test (Recommended Practice 1109-7) that requires test trucks to warm up for an hour, travel a minimum of 240 km and consume 114 litres of fuel. To remove the variables that can involve driving habits, two tractor-trailers chase each other until the test route’s half-way point, where trailers are switched and drivers change tractors.
“Repeatability with different vehicles is important,” Blake said. “Your first test is an indicator. Your second test is looking at a trend.”
Cruise controls should also be used to ensure speeds are constant.
“We like to see the same person taking the temperature reading (of fuel) at the same point in the tank,” he added.
Another important step is to shut off fuel crossover lines when conducting tests, to ensure the supply of diesel is coming from a single source.
In fact, there’s an argument for considering figures from a variety of sources.
Shopko, for example, compares driver trip sheets that record fuel purchases and distances against reports generated by fuel island computers.
FedEx drivers enter employee numbers, unit numbers and mileage at fuel islands before every fill up, said Dan Umphress, managing director of maintenance solutions at the LTL operation, which includes 10,000 tractors consuming 442,000 gallons of fuel per day. And that fleet’s system was also programmed to ensure the entered mileage can’t be lower than the last entry for the truck.
The data feeds fuel economy reports that can compare performance on a daily, weekly, monthly and an annual basis, as well as for the life of the vehicle.
“You can reach a point you drive yourself nuts,” Wessels said, referring to the array of variables outside an operator’s control. “The only approach I know to correct that is to run enough data (at least six months’ worth).
“If you don’t have a lot of information on a lot of routes and a lot of weather conditions, you’re at a real risk.”