THE DETERMINATION OF TASTE AND ODOUR IN DRINKING WATER IN RELATION TO WATER QUALITY REGULATIONS
Report No FR0254

R I Crane

Dec 1991

SUMMARY

I OBJECTIVES

To review the methods of analysis for the assessment of taste and odour with respect to UK Regulations concerning the quality of drinking water and to recommend laboratory studies to develop reliable procedures.

II REASONS

The assessment of drinking water for odour and taste is one of the more basic tests carried out by water undertakers and it is one of the few water quality assessments that consumers can readily carry out themselves. The Water Supply (Water Quality) Regulations 1989 (and their Scottish equivalents) have set mandatory standards for odour and taste in drinking water. Such standards for what is a subjective assessment lead to many methodological and interpretational problems. The methodology available and associated problems need to be reviewed in order that more reliable techniques can be identified and developed.

III CONCLUSIONS

The taste and odour standards in the EC Directive do not reflect that the determinations required are very subjective and the results will have considerable uncertainty attached to them. None of the currently-used methods are likely to produce results with the UK target precision of +/-1 dilution number.

The meaning of the numbers specified in the Directive appears to have been interpreted differently in the UK and some other Member States. The UK has introduced a scale which appears to be one point more lenient than used elsewhere. However, any real differences would probably be masked by the inherent lack of precision of the analytical methods.

Odour and taste are not precisely definable entities and any numerical value obtained is very dependent on the method used. Therefore in order to produce results comparable between different water undertakers a standard method must be used. At present a variety of methods are used in the UK and Europe and there would probably be differences in the results produced by these methods. A CEN task group is attempting to produce a harmonised (standard) method for the assessment of taste and odour in drinking water. This CEN method could become the standard method for all in the future in relation to testing compliance with the EC Drinking Water Directive and national regulations.

More specific conclusions related to the methodology of taste and odour compliance testing are given. In addition other conclusions on the assessment of taste and odour in drinking water are drawn which are not directly related to compliance with regulations.

IV RECOMMENDATIONS

• Pressure should be maintained for the revision of the taste and odour parameters in the EC Directive to provide more sensible and fully-documented standards. This documentation should include the purpose of compliance with the standard and the accuracy of associated analytical methods required to assess compliance.

• The analytical method should be standardised to guarantee comparable results.

• Any further development and testing of methodology undertaken in the UK should be compatible with the development of the "standard" method by CEN.

• Whatever method is used, its performance needs to be determined. Some further work is needed on determining which are the best statistical procedures for assessing the performance of the method and relating this to how compliance with a standard will then be judged from the results. It may be that the apparently inevitably low levels of precision of this type of method will override the apparent justifications for some of the method refinements suggested below.

• The stability of samples needs to be studied, especially in relation to odour.

• Some investigation on how any standard methods cope in practice with a wide range of substances, particularly very volatile chemicals, is required. The work should involve testing how a range of substances with differing volatilities achieve equilibrium vapour concentrations under test conditions.

• The standard method should include a screening process whereby samples which have virtually no taste or odour can be excluded from the expensive, full-scale test procedure.

• The panel size should be in the region of eight for full-scale testing and no more than three for screening. The pool of available panellists should be large enough to ensure that the same people do not always form a panel.

• Panels members should be trained and have known sensitivities to a range of specific substances and also, if possible, to real samples. The question of whether a panel should consist of "average" or "sensitive" people still needs to be resolved.

• The performance of panellists should be monitored.

• When presented with a set of samples containing a dilution of the test sample and reference samples panellists may be required to identify which sample they regard as not being the reference ("forced-choice" method) or they may be allowed to say that they cannot tell the difference. The statistical correction required in the forced-choice method is known but the statistical implications of the alternative method should be investigated.

• The choice of reference water is an area of uncertainty. The best option appears to be a groundwater of consistent quality which is reasonably close to the mineral content of the range of waters to be tested. A more detailed assessment of how the reference water affects the result of a test is needed.

• Procedures for running control samples should be included in the design of a method.

• A room and apparatus should be reserved for taste and odour testing.

• Qualitative testing for taste and odour has a useful role. There is room for greater standardisation in qualitative testing. There are some good reasons (sample instability, showing good customer relations etc) for training samplers to perform this type of testing. A standard list of descriptors of tastes and odours should be used.

V RESUME

The report outlines the regulations that cover the assessment of taste and odour in drinking water, discusses the underlying principles of such assessments and describes the various approaches used in the UK, Europe and elsewhere. The review is based on available literature and discussion with laboratories carrying out such assessments in the UK and Europe.

Copies of the Report are available from FWR, price £35.00 less 20% to FWR Members