Mr. Kurokawa presented an analysis by JARI on the influence of the different methods under discussion within the HDH work program on the NOx emission. Basis for the comparison was the NOx emission on the WHTC. The 30 sec moving average method came closest to the WHTC emission level (90 ). If a second-by-second slope is used the ratio is 77 mainly caused by a shift in the engine map towards lower loads. In conclusion, the 30 sec moving average slope, which shows a load frequency second closest to WHTC after the second by second slope, is preferable.

Mr. Sanchez suggested that additional emission testing should be done.

As regards the HILS models, it was agreed that it is not acceptable to only allow for Matlab to be used but it should at least in principle be allowed to use other software packages. As a consequence, the models have to be described in detail in the gtr.