Updating fuel economy testing is long overdue.
We’ve been sold a lie; a fable, a fraud, a fairytale of the turbocharged pot of gold at the end of the rainbow where fuel consumption is just about a non-existent factor in our lives. Manufacturers have sold us 1.4-liter Ecotec engines in Chevrolets producing the 153 hp that a 2.0-liter NA motor used to, and similarly sized engines from Fiat-Chrysler powering vehicles that traditionally required 2.0- or 2.4-liter powerplants. We even see larger engines like the 2.5-liter turbo 4 cylinder in the Mazda CX-9 replacing V6s with similar power.
But all of these return better claimed consumption figures than their larger, naturally aspirated forebears. But how? Is it magic, or maybe divine intervention? It’s neither – it’s simply a lie, a clever manipulation of the auto industry to circumvent stringent emissions laws in key sales markets like the US and Europe. But to understand why it’s a lie, we’ve got to look at how a turbocharged engine works, and how economy gets tested.
A normally aspirated (non-turbocharged) engine is a fairly simple thing. It delivers power in a simple, repetitive manner. The power curve generally starts off low, builds as the revolutions per minute (rpm) do, and then tapers off near redline. As engine speeds increase, so too does airflow into the engine – and as any pyromaniac will tell you, if you want more flame, you’ve gotta have more air. Turbochargers supply just that, juicing up the combustion process with compressed air. But this is a process that takes time – albeit fairly brief amounts – and isn’t always active if you’re in a low-load situation.
For these reasons, turbocharged motors have two distinctly different power curves – one on boost, and one off boost. The ‘off boost’ curve is how the engine behaves under low load – times when throttle inputs are so gentle that the engine’s management systems don’t feel the need for maximum power. Under these low load situations, the turbo isn’t active at all – the engine relies on natural flow of air into the intake and combustion chambers, and it supplies the minimum amount of fuel accordingly. In cars like the Chevrolet Cruze 1.4T, consumption is going to be that of a naturally aspirated 1.4-liter – pretty much next to nothing.
The ‘on boost’ power curve is the one that gives you maximum outputs – when a turbo is spinning at full tilt, providing as much air as possible for the combustion process. This yields maximum power and torque outputs – in the case of the Cruze, 153 hp and 177 lb-ft of torque – comparable to those of a traditional naturally aspirated 2.0- or 2.2-liter motor. But the more air you have for combustion, the more fuel you need too. Air flow sensors in the engine’s air intake detect how much air is coming in, and tell the fueling system to provide more or less accordingly. Under medium-to-high loads, when a turbocharger is active, the engine responds by sending in extra fuel for maximum torque and power production.
Extra fuel = worse fuel economy. It doesn’t matter how small, or big the engine gets – when it’s on boost and producing the power of a larger motor, it needs to drink like the larger motor does. It’s clever; an engine with the ability to behave like two different engines entirely. But manufacturers use that to their advantage, making the most out of flawed fuel economy tests.
In the USA, the tests have been updated more regularly than their European counterparts, though at their heart, they’re still archaic. These tests comprise three phases – all performed in a closed laboratory with no variables, and no supplementary items like air conditioning are used with the exception of one test. First up is the Urban Dynamometer Driving Schedule (UDDS) that evaluates city consumption. Also called the LA4 test, and originally derived in 1972, it's based around rush hour driving in Los Angeles; a time when there were far fewer vehicles, with better traffic flow, and very, very different vehicles and engines to what we have on the road today.
The UDDS test takes roughly 31 minutes, with a maximum speed of 56 mph. The test includes 23 stops and an average speed of just 20 mph. But the acceleration times allowed are far gentler than those we encounter in real life – like 12 seconds to accelerate from a standstill to 22.5 mph – and the stops allow for auto stop/start functionality to make a genuine difference – especially since the vehicles remain at a halt rather than creeping forward as they do in real life. Part two, known as the HWFET, is the Highway Fuel Economy Driving Schedule. It makes use of a pre-warmed engine and makes no stops. The average speed is 48 mph, with a top speed of 60 mph, and the test takes place over a simulated 10 mile journey.
The MPG figures are then adjusted down by predetermined percentages (10% for city and 22% for highway) in line with calculations made in 1984. These figures have remained unchanged for more than 30 years. From 2008 onwards, the EPA added a third test procedure to try and mimic real world situations better. The Supplemental Federal Test Procedure (SFTP) features three cycles. The first takes 10 minutes, covers 8 miles, averages 48 mph, and reaches a top speed of 80 mph. This test procedure features higher speeds, and importantly, more rapid acceleration. These crucial details mean the engine is placed under higher load – triggering the ‘on-boost’ power curve that requires more fuel.
The second, over 3.3 miles, tests air conditioning loads, and the third uses the city cycle in colder temperatures. Though the SFTP arguably produces more accurate real world figures, the results are skewed by antiquated tests used in the other two thirds of testing – tests that don’t account for real world behavior, and tests that don’t place real world loads on the engines. If these tests placed real-world appropriate loads on engines, they’d be tested in their on-boost state, and consumption figures we’d see would be far higher than claimed 27 mpg city value for the Chevrolet Cruze 1.4T, or the 25 mpg city value for the Fiat 500X 1.4T. But there’s also one other key issue…
The EPA has a limited testing capacity – only testing around 200-250 vehicles per year in 2010. At the time, that was roughly 15% of the new models on sale – considering each model variant has different factors including wheel and tire size, engine outputs, aerodynamic-affecting body cladding etc. For the most part, the EPA takes manufacturers at their word – using random testing in an attempt to keep manufacturers honest and testing specifically only when there are complaints or particular reasons that cause the EPA to doubt a manufacturer’s claims.
Ultimately, the key factors in determining fuel consumption are flawed. Trusting manufacturers who are doing their damnedest to avoid being caught out (we’re looking at your Dieselgate saga, Volkswagen) is risky business, but when test procedures are based on 30-40 year old data, and when the procedures themselves rely on engines operating in a low state of load, they ignore the single state almost all turbocharged cars will be operating in – the on-boost, high load state where the turbo is spooling, sending lots of air into the combustion chambers, along with plenty of fuel to keep the power outputs at their highest.
No one drives in the same way the EPA tests prescribe. No one drives their car in a laboratory without a radio, air conditioning, or without headlights. No one drives their car in 1972 downtown Los Angeles anymore – so why do our economy figures reflect people that do? Until such time as the tests take into account the state in which a turbocharged engine will actually be driven, turbocharged economy figures are destined to be one big scam.