51997PC0077

Amended proposal for a European Parliament and Council Directive relating to measures to be taken against air pollution by emissions from motor vehicles and amending Council Directives 70/156/EEC and 70/220/EEC (1) (97/C 106/05) (Text with EEA relevance) COM(97) 77 final - 96/0164(COD)

(Submitted by the Commission pursuant to Article 189a (2) of the EC Treaty on 26 February 1997)

Appendix 1 of Annex X reads as follows:

Appendix 1

1. This appendix describes the procedure to be used to verify the conformity of compliance of vehicles in service for the type I tail-pipe emissions test. The procedure deals with the case where there are gross emitters in the sample due to an emissions-related defect.

2. A vehicle is classified as a gross emitter when, for any regulated pollutant, the appropriate limit value given in section 5.3.1.4. of Annex I to this Directive is exceeded by more than 50 %.

3. Starting with a minimum sample of three vehicles, the sample is examined for the occurrence of gross emitters (according to the definition given in section 2 of this appendix).

3.1. If no gross emitter is found, the statistical procedure of section 4 of this appendix applies.

3.2. If a vehicle is found to be a gross emitter, the reason for the excess emissions shall be determined by the approval authority.

3.3. If more than one vehicle is found to be a gross emitter due to the same reason, a fail result for the vehicle sample shall be recorded.

3.4. When only one gross emitter is found, or more than one gross emitter is found but resulting from different reasons, the sample is increased by one vehicle, unless a cumulative sample size greater than the maximum sample size of 15 would be reached.

3.4.1. When, in the increased sample, more than one vehicle is found to be a gross emitter, for the same reason, a fail result for the sample shall be recorded.

3.4.2. If the maximum sample size of 15 is reached, provided no more than one gross emitter has been found in the sample, a pass or fail decision will be reached according to the statistical procedure of section 4 of this appendix.

3.5. Whenever the sample is increased due to the requirements of section 3.4, the statistical procedure of section 4 to this appendix shall be applied to the increased sample size.

4. Statistical procedure

4.1. With a minimum sample size of three and a maximum sample size of 15, the sampling procedure is set so that the probability of a lot passing a test with 40 % of the production defective is 0,89 (producer's risk = 11 %) while the probability of a lot being accepted with 65 % of the production defective is 0,15 (consumer's risk = 15 %).

4.2. The measurements of the pollutants given in section 5.3.1.4. of Annex I to this Directive are considered to be log-normally distributed and must first be transformed by taking their natural logarithms. I and n denote the current sample number.

4.3. If the natural logarithms of the measurements in the series are x1, x2, . . . xj and L is the natural logarithm of the limit value for the pollutant, then define:

dj = xj - L

>START OF GRAPHIC>

>END OF GRAPHIC>

>START OF GRAPHIC>

>END OF GRAPHIC>

4.4. Table 1 shows the values of the pass (An) and fail (Bn) decision numbers against current sample number. The test statistic is the ratio >START OF GRAPHIC>

>END OF GRAPHIC>

and must be used to determine whether the series has passed or failed as follows:

For 3 ≤ n START OF GRAPHIC>

>END OF GRAPHIC>

,

- fail the series if >START OF GRAPHIC>

>END OF GRAPHIC>

,

- sample another vehicle and test for a gross emitter if >START OF GRAPHIC>

>END OF GRAPHIC>

For n = 15,

- pass the series if >START OF GRAPHIC>

>END OF GRAPHIC>

,

- fail the series if >START OF GRAPHIC>

>END OF GRAPHIC>

.

>TABLE>

(1) OJ No C 77, 11. 3. 1997, p. 8.