ISO 5667-14:1998 download.Water quality — Sampling — Part 14: Guidance on quality assurance of environmental water sampling and handling.
Introduction
Quality control procedures are required for the collection of environmental water samples for the following reasons:
a) to monitor the effectiveness of sampling methodology;
b) to demonstrate that the various stages of the sample collection process are adequately controlled and suited to the intended purpose, including adequate control over sources of error such as sample contamination, loss of determinand and sample instability. To achieve this quality control procedures should provide a means of detecting sampling error and hence a means of rejecting invalid or misleading data resulting from the sampling process;
c) to quantify and control the sources of error which arise in sampling. Quantification gives a guide to the significance that sampling plays in the overall accuracy of data;
d) to provide information on suitably abbreviated quality assurance procedures that may be used for rapid sampling operations such as pollution incidents or groundwater investigations.
This part of ISO 5667 is one of a group of International Standards dealing with the sampling of waters. It should be read in conjunction with the other parts of ISO 5667 and in particular with Parts 1, 2 and 3.
The general terminology is in accordance with that published in ISOITC 147, Water quality, and more particularly with the terminology on sampling given in ISO 6107-2.
NOTE Bias Is Ihe total systematic error as contrasted to random error There may be one or more systematic error components contruting to the bias, A lSrgew systematic catherence Irom the accepted reference value Is reflected by a larger bias value.
3.3
precision
closeness & agreement between independent results obtained under stipulated conditions
(ISO 3534-11
NOTE 1 The variation associated with test results from repealed sampling operations will be SuCect to variation from analytical sources as well as trom sources connecled with m sampling process A conarison ci random error Prom repealed sampling operations with that from repeated analysis & the sane sanipte can be used to dec. the contribution of sarrØrxg to overall random error.
NOTE 2 PrecIsion depends only on the distribution 04 random errors and does not relate to the true value ci the specified value, (ISO 3534-1). The measure of precision Is expressed in terms ci a standard deviation ve. Improved precision Is reflected ax a smet slaridard deviation value.
NOTE 3 The Independence ad test results rellecls the extent to which results are obtained in a manner not axfluenced by any previous res4it on the same test oblecl tISO 3834-11. Quantitative measures 01 preoslon depend critically on stipulated conditions. The well-known terms repeatability and reproduobility relate to speatic types of stipufated conditions. The former term corresponds to meaallements made under the sane controlled Isame method, strictly a&xered to in the same loratory) conditions; the letter term refers to the sane method used ax difterent laboratories.
3.4
rspresentaflv.ness
extent to which the condition of all the samples taken troni the body of waler retlects conditions in water of interest
3.5
comparabIlity
degree of agreement with respect to control over random and systematic errors
3.6
Certified Reference MaIerlal
CRIII
stable, homogeneous material, with a composition closely matching that of the sample to be analysed, for which the concentrations of the determiriands ci interest in that material are known with a known degree of uncertainty
NOTE In most chemical analyses the traceability of measuremeni can be obtawied by a series of caatrahone that demonstrates diet no tess ci determinand or contamination occurs during the sample treatment This traceatility can be based on the analysis of a CRkL
3.7
blank
observed value obtained when measurement Is made on a sample identical to the sample of interest, but in the absence of the determinand.
NOTE Field blank samples are laloratoiy blank samples which are taken into the held, treated as samples and analysed as a check on sampling procedoree.
(ISOTR 13530J
3.8
splice
known cantity 01 detenThnand which is added to a sample, usually for the purpose of estimating the systematic error of an analytical system by means of a recovery exercise
PSO1R 13530J
5 Sampling quality control techniques
5.1 General
Sampling is defined Ni ISO 5667-2 as the process at removing a portion, intended to be representative, of a body of water (or sludge or sediment) for the purpose of examination for various defined characteristics.
Guidance is given below with respect to quality control procedures which can be used to identify and quantify errors assoaated with sampling.
A complete overview of quality assurance applied ID sampling is outside the scope of this document. However, it is important to emphasize that the qualify control measures discussed below should idealy be applied in the context of a well organized approach to quality controL This would include a review of the whole approach to sasoplrng with respect to its fitness for the intended purpose. Within this, the choice of sampling techniques, sampling locations, numbers and types of sample taken, training of sampiing statt. sample transport, preservation and storage should be considered. The chosen approach should be adequately documented and a system at record-keeping established. A suitable qualey control programme could contain any or all of the techniques listed below. The effort expended on sarripling quality control is dependent on the obedlves of the programme, but It is recommended that at least 2% of analytical efforts should be devoted to quality control for sampling.
As noted earlier, quality control measures in sampling have three main objectives:
a) to provide a way of monitoring and detecting sampling errors and hence a means at rejecting invalid or misleading data;
b) to act as a demonstration that sampling errors have been controlled adequately; and
C) to indicate the variability of sampling and thereby to give a griide to this important aspect of error.
The tollowing quality control techniques are described below.
— the collection of replicate samples as a check on the precision of sampling;
— the use of field blank samples to monitor sources of sample contamination,
— the use of spiked samples as quality controls to assess sample stability during transport and storage.
5.2 Replicate quality control samples
This term can be used to cover a range of approaches to quality control which aim to assess the random error assoctated with different levels of the sampling process:
a) analytical variance: replicate analyses of the same sample prepared in the laboratory can be used to estimate short-term analytical error;
b) analytical + subsampling/transporl variance: analyses of replicate samples taken ii the field (B1 and 82) from the buli sample (B) (the sample obtained by a single application of the sampling procedure). The difference between such data gives an estimate at analytical plus sampling variance (includes storage but excludes the effect from sampling containers).
C) analytical + tolai sampling variance: analysis of bulk samples obtained by separate application of the sampling procedure. This provides an Indication of the variance at the whole process of sampling and analysis (A1 and A2).
The relationship between the different sampling variances in examples b) and c) lilustrated schematically in figure 1.