Validation method for fire gas analysis — Part 1: Limits of detection and quantification
|Publication Date:||15 November 2011|
|ICS Code (Protection against fire in general):||13.220.01|
In this part of ISO 12828, limits of detection (LD) and limits of quantification (LQ) are defined and calculated. It provides methods for determining suitable values for these two parameters for a specific analytical procedure and for a specific chemical species. It does not provide detailed guidance on methods of sampling and analysis of specific species which might be present in fire effluents. This guidance is contained in ISO 19701 and ISO 19702. The use of this part of ISO 12828 fulfils the requirement in ISO/IEC 17025 that a laboratory carrying out chemical analysis (e.g. of fire effluents) is able to characterize and evaluate a method by such parameters as LD, LQ and uncertainty. Examples of where the information contained in this part of ISO 12828 can be applied are:
a) Method validation: The parameters LD and LQ are required for all chemical analytical methods; they are as important as measurements of accuracy and precision.
b) Classifications based on toxicity indexes: Methods selected for analysis of effluents must have a minimum limit of quantification, consistent with the critical concentration used to calculate the contribution of each effluent to toxicity index. Furthermore, a toxicity index is not considered as zero when concentrations of toxic species are detected but not quantified (as they are below the limit of quantification). In this case, a contribution at least equal to the limit of detection for each measured species can be registered.
c) Round-robin comparison between two analytical methods: For a given working range, two methods can be compared only if the limits of these methods (calculated by using this part of ISO 12828) are similar for the lower range of concentrations to be measured. For example, if one laboratory provides values near its own limit of detection, and another laboratory gives results well above its own limit of detection, the reproducibility R assessment of the round robin can be artificially overestimated. In many round-robin tests, bad reproducibility R values can be found if some values are close to the limit of quantification and/or limit of detection. In such cases, no conclusion on the round robin can be given without an assessment of the limit of quantification value and the expression of results as described in this part of ISO 12828.
This part of ISO 12828 is intended for use by operatives familiar with chemical and physical analysis of fire effluents.
Examples of existing standards where the information contained in this part of ISO 12828 can be used are the analytical chemical methods in ISO 19701, ISO 19702, ISO 5660-1, and the chemical measurements in the methods discussed in ISO/TR 16312-2.