17. CO2 specific emissions: There is good correlation with brake specific emissions but a possible problem exists due to the influence of H/C ratio. For LPG variations of CO2 is less than 10% and on O2 is less than 2%; For CH4 the influence on CO2 is less than 25% and for O2 the difference is less than 3%.

18. PEMS equipment provides a fairly accurate looking at the relation between power and CO2 (straight line) for diesel. For dual-fuel also the relationship between CO2 and power varies with the fuel ratios.

19. If O2 is measured instead of CO2 there also is a good correlation for diesel and dual-fuel.

20. Conclusions on CO2 and O2 measurements: 1) calculation of CO2 specific emissions for D-F has disadvantages; 2) the amount of O2 needed to release a certain amount of energy from an air/fuel mixture appears reasonably insensitive to the fuels used for a D-F engine. However, more validation work is necessary. Also, PEMS equipment does not always measure O2 emissions. So using O2 emissions seems to be a difficult approach. As such, back-to-back measurement was used and ‘shall be lower than or equal to the operation on diesel.’ A derogation is possible for CO; no CH4 requirement (no diesel value for CH4); and discussion is on-going about a possible NMHC derogation.

21. Back-to-back approach: there would be a choice of test bench measurement (brake specific emissions); on road test using PEMS (average emissions in g/second). Use of chassis dyno will be under further discussion depending on comments made from Germany.

22. Mr. Rijnders expressed confidence in the draft regulation yet there is concern about the next steps in validating the specific testing methods. Mr. Dekker said that a back-to-back approach was preferred over using O2 emissions because these have not yet been validated. But the on-road test and test bench measurement is validated but not the chassis dyno approach. More data is needed from the manufacturers in order to validate the method. But the chassis dyno is more popular with D-F suppliers yet it has not been an approved methodology, therefore, it is possible that this will be removed from the proposal. Also there is not much time to validate the approach. Mr. Rijnders asks, therefore, if chassis dyno approach can be removed for now and if, in the future it can be validated, it can be reinstated. The GFV will await a document from Germany stating their concerns and the next GFV (23rd February 2016) will consider whether or not to remove the chassis dyno testing option for testing.