Evaluation and Modification of Constant Volume Sampler Based Procedure for Plug-in Hybrid Electric Vehicle Testing 2011-01-1750

Plug-in hybrid electric vehicles (PHVs) consume both fossil fuel and grid electricity, which imposes emission testing challenges on the current constant volume sampler (CVS) test method. One reason is that in the charge-depleting cycle, PHVs having all-electric range operate the engine for a small portion of the traction energy need, causing the CVS to overdilute the exhaust gas. The other reason is that the dilution factor (DF) in the EPA calculation has an error caused by ignoring the CO₂ concentration in ambient air. This paper evaluates these challenges by testing a Toyota PHV on the industry standard CVS system combined with additional continuous sampling methodology for continuous diluents, smooth approach orifice (SAO) measurement for ambient air flow, and fuel flow meter (FFM) measurement for fuel consumption. The current EPA DF can produce an error resulting in higher mass calculation. For the test car, the error for THC is in the range 0-20% on the urban dynamometer driving schedule (UDDS), and 0-75% on the highway fuel economy test (HWFET), though it is estimated that this error amounts to less than 4.7% error for non-methane organic gas (NMOG) compared to the current SULEV emission standard. For the test car, the error for CO, NO_x, and CO₂ is in the range of 0-8%, 0-3.6%, and 0-0.5%, respectively, which are negligible compared to the current SULEV emission standard. An alternative procedure is proposed for the charge depleting cycle using the continuous sampling method to ignore the time period before the engine first starts, thereby eliminating overdilution. This technique can reduce the impact of potential gas analyzer and DF error to one third compared to the current CVS method. The FFM and CVS showed matching measurements during most portions of testing and only demonstrated discrepancy of fuel consumption measurement during the first engine start cycle. The detailed reason why the FFM measured 13 grams more CO₂ than the CVS on average is likely associated with exhaust left in the vehicle tailpipe and CVS sampling line, though more investigation is needed with future testing.