Section 1065.920 - PEMS calibrations and verifications(a)Subsystem calibrations and verifications. Use all the applicable calibrations and verifications in subpart D of this part, including the linearity verifications in § 1065.307 , to calibrate and verify PEMS. Note that a PEMS does not have to meet the system-response and updating-recording verifications of § 1065.308 and § 1065.309 if it meets the overall verification described in paragraph (b) of this section or if it measures PM using any method other than that described in § 1065.170(c)(1) . This section does not apply to ECM signals. Note that because the regulations of this part require you to use good engineering judgment, it may be necessary to perform additional verifications and analysis. It may also be necessary to limit the range of conditions under which the PEMS can be used or to include specific additional maintenance to ensure that it functions properly under the test conditions. As provided in 40 CFR 1068.5 , we will deem your system to not meet the requirements of this section if we determine that you did not use good engineering judgment to verify the measurement equipment. We may also deem your system to meet these requirements only under certain test conditions. If we ask for it, you must send us a summary of your verifications. We may also ask you to provide additional information or analysis to support your conclusions.(b)Overall verification. This paragraph (b) specifies methods and criteria for verifying the overall performance of systems not fully compliant with requirements that apply for laboratory testing. Maintain records to show that the particular make, model, and configuration of your PEMS meets this verification. You may rely on data and other information from the PEMS manufacturer. However, we recommend that you generate your own records to show that your specific PEMS meets this verification. If you upgrade or change the configuration of your PEMS, your record must show that your new configuration meets this verification. The verification required by this section consists of operating an engine over a duty cycle in the laboratory and statistically comparing data generated and recorded by the PEMS with data simultaneously generated and recorded by laboratory equipment as follows: (1) Mount an engine on a dynamometer for laboratory testing. Prepare the laboratory and PEMS for emission testing, as described in this part, to get simultaneous measurements. We recommend selecting an engine with emission levels close to the applicable duty-cycle standards, if possible.(2) Select or create a duty cycle that has all the following characteristics: (i) Engine operation that represents normal in-use speeds, loads, and degree of transient activity. Consider using data from previous field tests to generate a cycle.(ii) A duration of (6 to 9) hours.(3) Starting with a warmed-up engine, run a valid emission test with the duty cycle from paragraph (b)(2) of this section. The laboratory and PEMS must both meet applicable validation requirements, such as drift validation, hydrocarbon contamination validation, and proportional validation.(4) Determine the brake-specific emissions and mass rate emissions, as applicable, for each test interval for both laboratory and the PEMS measurements, as follows: (i) For both laboratory and PEMS measurements, use identical values to determine the beginning and end of each test interval.(ii) For both laboratory and PEMS measurements, use identical values to determine total work over each test interval.(iii) If the standard-setting part specifies the use of a measurement allowance for field testing, also apply the measurement allowance during calibration using good engineering judgment. If the measurement allowance is normally added to the standard, this means you must subtract the measurement allowance from measured PEMS emission results.(iv) Round results to the same number of significant digits as the standard.(5) For each test interval and emission, subtract the lab result from the PEMS result.(6) The PEMS passes the verification of this paragraph (b) if any one of the following are true for each constituent:(i) 91% or more of the differences are zero or less than zero.(ii) The entire set of test-interval results passes the 95% confidence alternate-procedure statistics for field testing (t-test and F-test) specified in § 1065.12 .70 FR 40516, July 13, 2005, as amended at 73 FR 37345, June 30, 2008; 75 FR 68467, Nov. 8, 2010; 79 FR 23814, Apr. 28, 2014; 88 FR 4688, Jan. 24, 2023