40 CFR § 1066.501 - Overview.
Use the following procedures to test EVs and HEVs (including PHEVs):
(a) Correct the results for Net Energy Change of the RESS as follows:
(1) For all sizes of EV, follow SAE J1634 (incorporated by reference, see § 1066.1010).
(2) For HEV at or below 14,000 pounds GVWR, follow SAE J1711 (incorporated by reference, see § 1066.1010) except as described in this paragraph (a). Disregard provisions of SAE J1711 that differ from this part or the standard-setting part if they are not specific to HEV. Apply the following adjustments and clarifications to SAE J1711:
(i) If the procedure calls for charge-sustaining operation, start the drive with a State of Charge that is appropriate to ensure charge-sustaining operation for the duration of the drive. Take steps other than emission measurements to confirm that vehicles are in charge-sustaining mode for the duration of the drive.
(ii) You may use Appendix C of SAE J1711 for charge-sustaining tests to correct final fuel economy values, CO2 emissions, and carbon-related exhaust emissions, but not to correct measured values for criteria pollutant emissions.
(iii) You may test subject to a measurement accuracy of ±0.3% of full scale in place of the measurement accuracy specified in Section 4.4 of SAE J1711.
(3) For HEV above 14,000 pounds GVWR, follow SAE J2711 (incorporated by reference, see § 1066.1010) for requirements related to charge-sustaining operation.
(b) This paragraph (b) applies for vehicles that include an engine-powered generator or other auxiliary power unit that provides motive power. For example, this would include a vehicle that has a small gasoline engine that generates electricity to charge batteries. Unless we approve otherwise, measure emissions for all test cycles when such an engine is operating. For each test cycle for which emissions are not measured, you must validate that such engines are not operating at any time during the test cycle.
(c) You may stop emission sampling anytime the engine is turned off, consistent with good engineering judgment. This is intended to allow for higher concentrations of dilute exhaust gases and more accurate measurements. Take steps to account for exhaust transport delay in the sampling system, and be sure to integrate over the actual sampling duration when determining Vmix.