THE UNCERTAINTY OF EQUIVALENCY

ABSTRACT
Throughout Europe, more and more legislation is being introduced, through European Directives, which either requires or refers to methods for the measurement of air pollution. Quite often these methods are not available as European standards (ie CEN methods), although member countries may have their own national standards. These may be used in the interim, but would need to be revoked once a CEN method was introduced, because CEN standards are mandatory in Europe. There may also be other standards available from ISO or the US EPA, but it is not a requirement that these be used by the European Union Member States. Standards exist in many forms, and often the accuracy (sic) is quoted. Unfortunately, this only provides a very rough estimation of the precision of the standard. Rarely does it explain the true accuracy (uncertainty). If they are to be used to demonstrate compliance with European legislation, then this is far from satisfactory.

Where member countries have in the past relied on their own National standards, and have put a lot of time, effort and money into their development, they may find it difficult to change their networks or instrumentation. To ensure future consistency throughout Europe, it is fundamental that a CEN Reference Method exists. Where the Reference method is based on a laboratory technique, is very precisely described or legislation is approved before a CEN method is available, it may not be possible for member countries to make the necessary changes, especially where extensive networks have been introduced. In these instances it would be necessary to be able to demonstrate equivalency with the CEN Reference Method, or the specific requirements of the EU Directive. However, no such methodology or guidance exists at present, for this to be done on a consistent basis.

This paper discusses the need to move to standard methods of known uncertainty, which are drafted more as performance standards, so that equivalency can be easily demonstrated, if required. In this way we not only encourage better standards, but we do not stifle innovation in measurement techniques.