The problems with Volkswagen's diesel cars came to the attention of regulators after the International Council on Clean Transportation and West Virginia University began researching "real-world" emissions performance of several vehicles in 2013 and 2014. WVU tested three models in the United States – VW Jetta, VW Passat and BMW X5 – and contributed its results to the ICCT's final report released in October, 2014. Together, European and U.S. testers rigged 15 vehicles from multiple manufacturers with portable emissions-modelling systems that include gas analyzer probes that snake up the tailpipe, and a connection to the car's onboard computer.
The report found, on average, the vehicles emitted seven times the regulated limit of nitrogen oxides (NOx), and some were 25 times or more over their emissions limits. It also showed that some vehicles passed the real-world tests (a BMW, for instance). NOx in the atmosphere contribute to smog and acid rain and have been linked to higher levels of lung disease.
The California Air Resources Board's (CARB) timeline of events state that after WVU reported its results at a conference in May, 2014, the U.S. Environmental Protection Agency began to push Volkswagen for an explanation. In December, 2014, VW offered to do a voluntary recall to recalibrate two of the main systems it uses to capture NOx. CARB resumed testing the recalled vehicles in May, 2015, and reported to VW on July 8 that the real-world tests remained significantly high. In joint meetings between July 8 and Sept. 3 – during which the EPA threatened actions that could block the sale of 2016 VWs – the company admitted its vehicles have software designed to kick in during vehicle-certification testing.
The EPA calls any physical or computer system designed to evade standards testing a defeat device. To sell a car in North America, the company must prove it is in compliance with the U.S. Clean Air Act. Cars must pass tests such as the EPA's Urban Dynamometer Driving Schedule (UDDS), also known as "the city test" – an 11.9 km test at an average speed of 30.6 km per hour, or the Highway Fuel Economy Test, which is a 16-km test at an average speed of 77.2 kph. The test measures pollutant in emissions and involves very specific calibrations to which car-makers and testers have agreed.
VW has yet to describe publicly how it engineered its defeat device, but the EPA says VW's onboard computer software recognized the test conditions – perhaps because the car steering wheel was not turning at high speed – and activated technology that would capture NOx and allow it to pass the test.
Why would VW fake the results?
VW has not explained why it would disable NOx capture technology, but the answer may lie in a mix of fuel economy and customer satisfaction issues.
The main technology for capturing and breaking down NOx is selective catalytic reduction, which uses a chemical catalyst, usually aqueous urea (a synthetic product sometimes called AdBlue), that is stored in a tank on the vehicle that needs periodic refilling. The more NOx captured, the faster the urea tank needs refilling, and compact cars have limited space for large tanks, meaning owners might have to do it themselves between service appointments.
Another is called Lean NOx Trap Technology, which is part of the vehicle exhaust system. Its main requirement is a lot of heat, which can reduce fuel efficiency, and may drain off a little engine power from the drive train, as much as 10 per cent.
Which models are affected?
VW has admitted about 11 million vehicles have the defeat device software. According to the EPA, the affected diesel models include:
Jetta (model year 2009-15), Jetta Sportwagen (2009-14), Beetle (2012-15), Beetle convertible (2012-15), Audi A3 (2010-15), Golf (2010-15), Golf Sportwagen (2015), Passat (2012-15)
One of the worst offenders in the ICCT tests was a VW Jetta equipped with a Lean-NOx trap. This is a vehicle that requires an EPA certificate called Tier 2 Bin 5/ULEV II, which means it is supposed to emit only 0.05 NOx (grams/mile). The ICCT test of this vehicle found 1.8 grams per km, which converts to 2.88 grams per mile.
VW scandal by the numbers
11 million: Number of Volkswagen diesel vehicles the company says could be affected worldwide
482,000: Number of VW diesel passenger cars sold in the United States since 2008 in which the company installed software that could skirt emissions tests
23 per cent: Proportion of all cars sold by Volkswagen in the United States in August with diesel engines
$26-billion: Loss in market value by end of day Tuesday as Volkswagen's ordinary shares fell to a four-year low
$7.3-billion (U.S.): Amount the company has set aside, initially, to cover the fallout
$37,500: Potential fines from the U.S. Environmental Protection Agency VW could face per vehicle
Editor's note: A Wednesday feature on Volkswagen's emissions deception said the ICCT (International Council on Clean Technology) test of a VW Jetta found 1.8 grams of nitrogen oxides per km. The story incorrectly converted that to 3.96 grams per mile, when in fact it is 2.88 grams per mile.