meet the emission standards for the legally mandated vehicle lifetime. Following a satisfactory agency review of these data, the manufacturer is issued a certificate of compliance, which allows the production of vehicles to proceed. Produced vehicles are then tested to ensure that the new vehicles meet the applicable emission standards. In addition, in-use vehicles are tested to ensure that the vehicles maintain their performance over time. A failure to maintain in-use performance can result in a recall. Manufacturers are also required to provide a warranty on the emission controls in the vehicle.
The initial exhaust-emission test used to certify cars, referred to as the federal test procedure (FTP), was based on measurements that sought to replicate driving during a typical vehicle trip in downtown Los Angeles during the 1960s. In this procedure (known as a test cycle), the vehicle exhaust is continuously collected and quantified as the vehicle is started and run through a series of operating modes (for example, cold start, accelerations, and decelerations). However, the FTP had a number of shortcomings when it was used to forecast in-use mobile emissions: (1) driving behavior and conditions vary so widely that it is difficult to represent all; (2) because of limitations in the dynamometers used at the time the test procedure was developed, the maximum acceleration during the test had to be limited unrealistically to 3.3 miles per hour per second (mph/sec); and (3) most of the vehicle speeds in the FTP are below 30 mph, a speed that is not reflective of the actual speeds used by most motorists in the United States. Thus, “off-cycle” emissions occurred during vehicle-operating modes that were not tested in the FTP. Although vehicles were certified to appropriate standards, their real-world emissions were higher than would be expected by the standards.
Congress sought to remedy the off-cycle emissions problem in the 1990 CAA Amendments by requiring EPA to control off-cycle emissions. To do so, EPA developed the supplemental federal test procedure (SFTP), which tests cars over an extended range of speeds and accelerations. Certification to the SFTP will be fully phased in by the 2004 model year. However, questions persist whether the SFTP fully captures the large increases in emissions that occur due to rapid accelerations and other changes in the vehicle load.
In addition, EPA took steps to address the issue of evaporative emissions. The test cycle for evaporative emissions was originally developed around 1970, and initial controls on evaporative emissions used a test procedure to measure emissions that would occur from a parked car (called diurnal emissions) and from a car that had just ceased operation (called hot-soak emissions). Like FTP for exhaust emission, these tests proved to be poor predictors of actual mobile evaporative emissions. Between 1987 and 1993, EPA developed a new test procedure for evaporative emissions