Method to account for measurement uncertainties when performing metrological tests within the scope of the German X-ray Ordinance (RoeV) and the German Radiation Protection Ordinance (StrlSchV)

Abstract

The requirements placed upon metrological tests conducted within the scope of radiation protection are often defined by tolerance ranges that are stipulated by legal limits or minimum technical requirements set out in guidelines and standards. A metrological test is designed to determine whether a tolerance range, and hence its pertinent requirements, have been complied with.
In doing so, the measured quantity value may well be within the tolerance range, yet the true value of the measurand may in fact be outside of the tolerance range. The full measurement result is therefore needed in order to assess this situation. This requires a measurement uncertainty along with the measured quantity value as an estimate for the true value underlying the measurement.
The measurement uncertainties to be taken into account must first be determined in order to be able to assess whether the measurement results comply with the requirements. This in turn requires a mathematical and statistical approach.
The stipulation of a standard measurement uncertainty must be based on the ‘Guide to the expression of uncertainty in measurement’ (GUM) and/or GUM Supplement 1. Alternatively, transparent conservative estimates of measurement uncertainties can be provided which are based on information from other sources (guidelines, standards, recommendations from specialist societies, etc.).
This recommendation serves to enable a uniform procedure to be used to test conformity with requirements. A requirement may be deemed complied with, i.e. the measured quantity value complies with the requirements, if, in the case of a stipulated tolerance range, the true value of the measurand lies within the tolerance range at a level of probability to be determined. The SSK considers it appropriate that the probability level for a correct decision in favour of conformity should be at least 95%, while the probability level for an incorrect decision should not exceed 5%.
As there is a great deal of demand for recommendations relating to this topic, the recommendation has been laid out such that it can be universally applied. The enclosure contains detailed examples on how to use the recommendations in practice.

The provisions of the RöV (2003), which expired on December 31, 2018, have been transferred to a revised StrlSchV, which came into force on December 31, 2018.