Changes Implemented in Version 6 of the RDE Test Protocol Following the Comments on Version

Changes implemented in Version 6 of the RDE test protocol following the comments on Version 5 and the discussion during the meeting of the RDE drafting groupon 30 June 2014 (Ispra, Italy)

1.Introduction

This document details key changes introduced into the Appendices 1-4 of the draft RDE test protocol. The changes in the protocol reflect the stakeholder comments provided on draft Version 5 and the discussions of the RDE drafting group on 30 June 2014. Text marked red signifies numbers that shall be verified. Comments on the present draft are welcome.

2. Key changes

• General: after internal discussions it was agreed to remove from the entire document any reference to “good engineering practice.” JRC checked carefully and did not identify cases in which the omission would cause ambiguity.
• 1.2.2 Test parameters: We now suggest to measure parameters at a frequency of 10 ± 0.001 Hz. 10 Hz sampling can substantially increase the accuracy of time alignment. Are the new requirements feasible? Are there reasons to revert to the 1 Hz recording?
• 1.2.2 Test parameters: Footnote 6 – based on internal discussions and additional stakeholder feedback we removed the reference to the ECU as the preferred signal but suggest instead that a sensor (e.g., an optical sensor) shall be the preferred reference. This choice ensures independence of PEMS testing from communication with the ECU.
• Table 1: We now introduced also altitude and a new footnote 10 on the calculation of NOX emissions from NO and NO2.
• 1.2.5 Emissions sampling: We kept the provision of a maximum residence time of 3s for particle sampling. More detailed calculations should be provided to justify an extension of the residence time. Moreover, we specify here a penetration efficiency of 95% for low-pressure sampling systems.
• 1.3.4 Checking and calibrating the analysers for measuring gaseous emissions: We removed the 90% span criteria and refer to this requirement only in paragraph 1.5.3.
• 1.3.5 Checking the analyser for measuring particle emissions: We increased the threshold for zeroing to 5000 particles and removed completely all requirements for the response check.
• 1.4.2 Test run: We have now reintroduced the possibility to auto-zero the analysers. Auto-zeroing should be discussed further in the meeting on 4 September.
• 1.4.3 Test end: We now recommend switching off the engine at least 30 seconds before the sampling and data recording is stopped. Is this provision problematic?
• Appendix 2: We have rearranged the provisions of Appendix 2. The whole appendix should be checked in detail. The most important changes are detailed below.
• 2.2 Time alignment of parameters: We now include provisions for a two-step alignment process. The method shall be checked for correctness and practicability.
• 2.3 Cold start: Based on internal discussions, we changed the maximum allowable time to 2 min. This provision shall be discussed as well as the possibilities for simplifying requirements further by presenting only a 2 min time limit.
• DPF regeneration is now completely excluded from Appendix 2 and only dealt with in the main text of Annex IIIA.
• 2.4 Consistency check of vehicle speed and driving distance (and other correlation analyses): Based on 63 PEMS test conducted at JRC, we define linearity requirements and permissible deviations of total trip distances as determined by GPS and ECU vehicle speed signals. Comments on the requirements are most welcome.
• 2.5.1 Dry-wet correction: We now removed the equations making use of air mass flow rate and fuel mass flow rate and limit the dry-wet correction to one simple equation. This choice shortens and simplifies the text but should be discussed further in the next meeting.

2.5.2 NOx correction for humidity and temperature: We do not allow for humidity correction here – this choice shall be discussed further and potentially referred to the RDE group if disagreement cannot be resolved.

• 2.7.2 and 2.7.3 on the determination of exhaust gas mass flow: Provisions may be skipped as the linearity of air flow and fuel flow is not specified any more (only exhaust flow has to meet requirements of App.3). The equations should be discussed during the meeting.
• 2.8 Calculating the instantaneous mass emissions: We keep the originals provisions but would like to discuss whether the detailed calculations under 2.8.3 should be removed.
• Table 3: We removed the linearity requirements for air flow and fuel flow and require now only to verify and validate the resulting exhaust mass flow.
• 3.2.2 Frequency of linearity verification: We now specify that the linearity of analysers shall be verified within six months before the actual emissions tests. It should be discussed whether this requirement is necessary and feasible.
• 3.2.3 Procedure of linearity verification: After internal discussions, it was decided to remove the option of calibrating “out-of-calibration” EFMs on the chassis dynamometer. In the text, we now specify that EFM shall be verified every six month.
• 3.3 and 3.6: we now ask to verify accuracy and other requirements only for analyser and EFM families, not for individual instruments anymore.
• 3.3.1.2 Alternative analysers: We now include additional provisions for alternative analysers. If alternative analysers meet the provisions of paragraph 3.3.1.2 no further validation is mandatory (see changes in paragraph 4.2.1).
• 3.3.2.7 Rise time: We kept the requirement at less than 3s upon request; this choice is still debated and may be lowered again to 2.5s, if equipment manufacturers can agree.
• 3.3.3.2 Efficiency test for NOX converters: We now specify that the efficiency of the NOXconverter shall be checked at least within one week before the emissions test.
• 3.4.2 Gas dividers: We decided to keep the paragraph because it is slightly more explanatory than the text in R83. Comments are welcome.
• 3.6.1.1 General: Following the comment of ACEA we now require non-standard exhaust flow measurements to be verified only with respect to their accuracy.
• 3.6.1.5 Noise: We leave the requirement unchanged; they are currently a factor of 2 higher than the accuracy requirement.
• 3.6.1.6 and 7. Zero and span drift: We now specify that zero and span drift can be checked based on the primary pressure measurements.
• 4.2.1 Frequency of PEMS validation: Following concerns of DG-ENTR, we added the possibility of ex-post validation. The paragraph should be re-discussed during the meeting.
• 4.2.2 PEMS validation procedure: We now divided the paragraph into sub-paragraphs and specify that the effect of the CVS on the PEMS mass flow measurements shall be minimized.
• 4.2.3 Permissible tolerances for PEMS:We decided to keep the provisions for the ECU vs. chassis distances. JRC tests suggest deviations of 1.7±0.2% and around 350±100m.For all parameters, we now present only absolute accuracy requirements. Are the requirements in Table 7 appropriate?
• 4.3.2. Validation procedure: We kept the provisions but should discuss whether the validation of simplified EFMs is necessary or whether it can be absorbed in the validation of PEMS.We increased the threshold for excluding data points from the validation from 5% to 10% of the maximum value.

1