Taste and Odour

(endorsed 2011)

Guideline

The taste and odour of drinking water should not be offensive to most consumers.

General description

Taste and odour are two of the primary criteria consumers use to judge the quality and acceptability of drinking water. People’s sense of taste and smell tends to vary, and so the acceptability of the same water can vary from person to person and from day to day for the same person. Similarly, one individual within a group may be more or less sensitive to a particular substance than the group as a whole. Whilst taste and odour present in water does not generally have a health impact, the presence of tastes and odours may raise consumer concern with regard to water quality.

Sources of taste and odour

Taste and odour in drinking water can result from naturally occurring inorganic chemicals; from biological activity, either in the source, treatment process or distribution system; as a by-product of water treatment processes; or from chemical contamination at any point from source to tap.

Inorganic compounds are generally present in water in substantially higher concentrations than organic compounds. Taste thresholds for some commonly occurring inorganic ions are about 0.1 mg/L for manganese, 0.3 mg/L for iron, 3 mg/L for copper, 3 mg/L for zinc, 250 mg/L for chloride, and 250-500 mg/L for sulfate. Most of these ions have health guidelines at concentrations higher than their taste thresholds (except copper at 2 mg/L). In most cases the customer would reject the water for aesthetic reasons before it would be of health concern.

Contamination of source water from spills, discharges or leaks of organic compounds can result in unpleasant taste and odours. Diesel fuel, for example, has a taste and odour threshold of 0.0005 mg/L. Methyl tert-butyl ether (MTBE) is the most commonly used fuel oxygenate added to reduce atmospheric concentrations of carbon monoxide and other aromatics. MTBE has frequently been detected in samples of shallow groundwaters, particularly in the United States. It affects the taste/odour of water at concentrations below 0.030 mg/L (Young et al. 1996).

One of the most common odours in water is described as “earthy”, “musty” or “woody”. Compounds most often linked to these odours are geosmin and 2 methyl isoborneol (MIB), which have similar low odour threshold concentrations of 0.00001 mg/L (10 ng/L) (Young et al. 1996). Cyanobacteria that produce these compounds include taxa representing the genera Anabaena, Aphanizomenon, Planktothrix, Oscillatoria and Phormidium in either planktonic and benthic habitats. Actinomycetes grow preferably in terrestrial habitats such as exposed sediments and vegetative debris, and are considered to enter aquatic habitats mainly in run-off from the shoreline.

Production of odorous compounds has been reported for most of the major algal classes and other odours produced by particular algae have been described as sweet, aromatic, cucumber, flowery, geranium, nasturtium, violets, fishy, peaty, grassy, mouldy, and vegetable. These odours originate from a variety of odorous compounds produced by the algae including aldehydes, ketones, alkenes, alcohols, terpenes, sulfides, amines, hydrocarbons, fatty acids, esters, carbonyl and aromatics. Cell concentrations as low as 500 cells/mL for some cyanobacteria and for a range of other algae are sufficient to taint a water supply.

Disinfection chemicals can contribute taste or odour to water. The odour threshold for free chlorine varies with pH, but is generally considered to be between 0.1 and 0.4 mg/L, whilst monochloramine and dichloramine odour thresholds are considered to be 0.5 mg/L and 0.15 mg/L respectively. A study by Piriou et al. (2004) has determined taste thresholds of 0.05 mg/L for free chlorine, 0.1 mg/L for monochloramine and 0.2 mg/L for chlorine dioxide using trained French panellists with flavour profile analysis. Untrained panellists were around 2-4 times less sensitive and the US consumer panel was 5-10 times less sensitive than the French consumer panel. This result can be linked to the different chlorination practices in the two countries (residuals are around 0.1-0.2 mg/L in France compared with 1.0-3.0 mg/L in the USA).

A number of organic compounds produced as by-products of disinfection, particularly chlorination, can cause tastes and odours. Some chlorinated phenols, for example, have an antiseptic smell and a very low taste and odour threshold, varying from 0.002 to 0.0001 mg/L, whilst some brominated phenols have a threshold as low as 0.0000005 mg/L (0.5 ng/L) (Mackey et al. 2004).

A range of chloroanisoles can result in earthy/musty odours (Young et al. 1996). For example 2,4,6-trichloroanisole (TCA) is produced from the action of biofilms in distribution systems on the disinfection by-product 2,4,6 trichlorophenol. TCA is detected at lower concentrations (typically <0.000001 mg/L [<1 ng/L]) than MIB or geosmin, but it is less frequently responsible for odour incidents.

Dimethyl di- and tri-sulfides (DMDS and DMTS) are responsible for septic/swampy odours. They have odour threshold concentrations at low ng/L concentrations. It has been suggested that these compounds may be produced by microorganisms in distribution systems (Franzmann et al. 2001, Heitz et al. 2000).

Taste and odour can also arise from impacts on the supplied water within the customer’s property, such as contaminants in direct or indirect contact with water (e.g. contaminants from kettles, refrigerators, dishwashers or washing machine hoses). The compound 2,6-dibromophenol, identified as probably responsible for a “plastic” or “chemical” taste in water after it is boiled, has a taste threshold concentration of 0.0005 mg/L (Whitfield et al. 1992, Adams et al. 1999). Odours resembling kerosene and cat urine were found to be more intense and more diverse when chlorine dioxide ($\text{ClO}_2$) was used, and increased numbers of complaints about odours in domestic water supplies were associated with the presence of new carpets in customers’ homes (Dietrich et al. 1992).

Measurement

Sensory analysis

Sensory methods provide a qualitative classification and a semi-quantitative determination of taste and odour intensity. Sensory techniques include flavour profile analysis and sensory gas chromatography.

A panel experienced in flavour profile analysis (Krasner 1995, McGuire 1995, Suffet et al. 1999) often represents the first step in coping efficiently with taste and odour episodes. The twenty-first edition of Standard Methods for the Examination of Water and Wastewater (APHA AWWA WEF Method 2170 - 2005) presents flavour profile analysis (FPA) as a technique for identifying taste and odour samples. FPA uses a group of four or five trained panellists to examine the sensory characteristics of samples. Flavour attributes are determined by tasting, odour attributes by sniffing the sample. Panellists must be able to detect and recognise various flavours present and quantify them according to standards. FPA requires well trained panellists and data interpreters, and reproducibility of results depends on the training and experience of panellists. These panels are useful for assessing complaints by consumers, potentially identifying the source of an adverse flavour; and for assessing the impact of a new or improved treatment process on taste and odour.

Sensory gas chromatography (GC) (Bruchet 1999, Suffet et al. 1999) is often used to complement chemical analysis when identifying odorous compounds. The effluent from the GC capillary column is directed to the nose of an operator through tubing and a sniffing funnel. This technique has been used for plastic odours detected from polyethylene pipes.

Chemical analysis

Chemical analysis during a taste and odour episode can identify the compound(s) responsible for the organoleptic characteristics of the sample and thus potentially ensure that the episode is not linked to a possible health threat. However, as the human nose is very sensitive, to obtain similar sensitivity from chemical analysis it is first necessary to concentrate the samples.

Closed loop stripping analysis (CLSA) (Bruchet 1999, Crozes et al. 1999, Khiari et al. 1999) and liquid-liquid extraction using methylene chloride (Khiari et al. 1999, Ventura et al. 1995) represent two methods of choice for the concentration of the most common odorous compounds. Solid phase micro extraction is an alternative technique to CLSA for the extraction of odorous compounds from water (Bao et al. 1999). Purge-and-trap and headspace methods can possibly be used as complementary sample preparation techniques for volatile compounds but these methods have limits of detection at or greater than 0.001 mg/L. Stir bar sorptive extraction (Twister extraction) is the most recent development in extraction techniques for taste and odour compounds (Benanou et al. 2004, Baltussen et al. 1999). A stir bar coated with polydimethyl siloxane is placed in the sample, which is stirred for 30-120 minutes. The compounds are extracted onto the stir bar surface through this procedure, and the bar is then placed into a thermal desorption unit, which is connected online to the gas chromatograph-mass spectrometer. The limit of detection for this method is reported to be <0.000001 mg/L (<1 ng/L) for both geosmin and MIB.

A gas chromatograph-mass spectrometer, for separation and detection of the off-flavours compounds, is a prerequisite for any laboratory wishing to identify taste and odour compounds.

Treatment of drinking water

Inorganic compounds, which can cause taste and odour, can be removed by the use of appropriate treatment processes. For example iron and manganese can be removed during conventional treatment with pre-oxidation. High salinity water may require the use of reverse osmosis to make the water palatable.

Organic substances producing taste and odour are generally more common in source waters. Volatile compounds can sometimes be removed by aeration. However, the application of activated carbon, in powdered (PAC) or granular form (GAC), is often the most effective treatment for the removal of a range of odour compounds. In Australia, PAC is the chosen method for the treatment of cyanobacterial metabolites due to its ease of application and because it can be applied only when required (Newcombe 2006). However, this method is quite costly if continuously applied. In addition, high levels of PAC will increase the load on the sludge treatment facilities, and may result in additional wear and tear on mechanical equipment (pumps etc) resulting in higher maintenance costs. GAC has many other advantages over PAC, such as its long life, higher adsorptive capacity, the ease of process control, more efficient use of the carbon, and the ability to regenerate the carbon for reuse (Herzing et al. 1977). Unfortunately at present there are no GAC regeneration facilities in Australia, so the waste adsorbent must be disposed and replaced.

Oxidation can also reduce tastes and odours. The effectiveness of the process is dependent upon factors such as the type of oxidant used, the type of reaction (addition or substitution), the structure of the compound, contact time, and environmental factors (e.g. pH, temperature, the presence of interfering compounds). Chlorine, chloramine and chlorine dioxide are common oxidants/disinfectants used in drinking water treatment. They have been effective in treating a variety of tastes and odours. However, MIB and geosmin have been found to be virtually resistant to this form of oxidation due to their tertiary structures (Lalezary et al. 1986, Anselme et al. 1988, Glaze et al. 1990). Ozonation is more effective for MIB and geosmin; approximately 50% removal of these compounds has been recorded in the laboratory, and at pilot and full scale (Ho et al. 2002, Ho 2004). The combination of ozone and GAC (sometimes called biological activated carbon or BAC) can be very effective in removing a range of odour compounds.

When disinfection results in tastes and odours being formed within distribution systems, the removal of precursors such as natural organic matter during the treatment process, increased oxidation, or change of oxidant can sometimes solve the problem.

Tastes and odours arising out of reaction with biofilms can be reduced by preventing or reducing biofilm growth. This is best accomplished by introducing treatment processes that reduce the food source for bacteria, the assimilable organic carbon, and hence reduce biofilm growth. The presence of a disinfectant residual will also reduce biofilm growth but care must be taken not to introduce disinfectant-related tastes and odours.

If materials used in distribution systems result in discernable tastes and odours, then replacement of the materials or in situ lining of the pipework is recommended. The use of non-return devices or back-flow prevention devices can eliminate taste and odour issues associated with the back-siphoning of water, including hoses attached to dishwashers and washing machines installed within close proximity to a draw-off point. The introduction of the Australian and New Zealand Standard AS/NZS 4020 (2005) to ensure materials used in contact with water comply with a range of tests, including the formation of taste under standard test conditions, should eventually eliminate taste and odour problems associated with materials in contact with drinking water. However the interaction of disinfected water with components such as plastics in customers’ kettles or new plumbing can be more difficult to control.

Health considerations

Taste and/or odour in potable water may indicate pollution of the water, insufficient water treatment and/or inadequate maintenance of the distribution system. Odours of a biological origin can indicate increased biological activity, for example by algae. Some algae can produce toxins and the detection of these algae by odour provides a useful early warning of potential problems, although odour does not necessarily indicate the presence of toxins.

Derivation of guideline

It is clearly unsatisfactory for a water authority to be supplying water that is objectionable in taste and odour to consumers. This will undermine consumer confidence and may lead to the use of water from sources that are aesthetically more acceptable, but potentially less safe. Such sources may include untreated private household supplies, bore water or water treated through poorly maintained domestic filters. It is also unrealistic to expect complete consumer satisfaction with aesthetic characteristics of a water supply; therefore an appropriate guideline should be that the taste and odour of drinking water not be offensive to most people. Due to the subjective nature and range of causes of taste and odour, it is not possible to set a quantitative guideline.

Operational guideline for MIB/geosmin

The major cause of taste and odour episodes in Australian water supplies is the presence of MIB and/or geosmin. Based on experience in water utilities, action based on concentrations of MIB/geosmin suggested below will help to minimise customer complaints.

At treatment plant inlet

Total MIB and/or geosmin >10 ng/L

Increase sampling to every 2 days at treatment plant inlet

Start MIB/geosmin analysis on treatment plant outlet

At treatment plant outlet

Total MIB and/or geosmin >10 ng/L

Introduce powdered activated carbon dosing to treatment plant

Regular measurement and identification of algae should also be undertaken to complement MIB/geosmin analysis at the inlet to the water treatment plant. Depending on the species-specific relationship between cell numbers and MIB/geosmin concentrations, additional monitoring may be necessary at the treatment plant inlet when algal organisms known to be producers of these compounds exceed approximately 1000 cells/mL.

Guidelines in other countries

The 2004 World Health Organization (WHO) Guidelines require that taste and odour be acceptable to the consumer.

The 1998 European Economic Community Standard (Council Directive 98/83/EC), requires that taste and odour be acceptable to consumers and that there be no abnormal change.

The 2005 Canadian Guidelines stipulate that drinking water shall have an inoffensive taste and odour.

Since 1979 the United States Environmental Protection Agency has listed odour in the secondary drinking water contaminant standards and has listed a secondary maximum contaminant level for odour of 3 expressed as a Threshold Odour Number (TON). They also recommend a variety of reports for further information on identification and control of taste and odours.

References

Adams JB, Lock SJ, Toward MR, Williams BM (1999). Bromophenol formation as a potential cause of ‘disinfectant’ taint in foods. Food Chemistry, 64(3):377-381.

Anselme C, Suffet IH, Mallevialle J (1988). Effects of ozonation on tastes and odors. Journal of the American Water Works Association, 80(10):45-51.

AS/NZS (2005). Testing of products for use in contact with drinking water, AS/NZS 4020:2005. Standards Australia, Standards New Zealand.

APHA AWWA WEF Method 2170 - Flavor Profile Analysis (2005). Standard Methods for the Examination of Water and Wastewater, 21st Edition. American Public Health Association, American Water Works Association, Water Environment Federation.

Baltussen E, Sandra P, David F, Cramers CJ (1999). Stir bar sorptive extraction (SBSE), a novel extraction technique for aqueous samples: Theory and principles. Journal of Microcolumn Separations, 11(12):471-478.

Bao M, Griffini O, Burrini D, Santianni D, Barbieri K, Mascini M (1999). Headspace solid-phase microextraction for the determination of trace levels of taste and odor compounds in water samples. Analyst, 124:459-466.

Benanou D, Acobas F, de Roubin MR (2004). Optimization of stir bar sorptive extraction applied to the determination of odorous compounds in drinking water. Water Science & Technology, 49(9):171–178.

Bruchet A (1999). Solved and unsolved cases of taste and odor episodes in the files of Inspector Cluzeau. Water Science & Technology, 40(6):15-21.

Crozes G, Hagstrom J, Suffet I, Young C (1999). Bench-scale evaluation of adsorptive processes for taste and odors control using rapid small-scale column tests and flavor profile analysis. Water Science & Technology, 40(6):39-44.

Dietrich AM, Orr MP, Gallagher DL, Hoehn RC (1992). Tastes and odors associated with chlorine dioxide. Journal of the American Water Works Association, 84(6):82-88.

Franzmann PD, Heitz A, Zappia LR, Wajon JE, Xanthis K (2001). The formation of malodorous dimethyl oligosulfides in treated groundwater: the role of biofilms and potential precursors. Water Research, 35(7):1730-1738.

Glaze WH, Schep R, Chauncey W, Ruth EC, Zarnoch JJ, Aieta EM, Tate CH, McGuire MJ (1990). Evaluating oxidants for the removal of model taste and odor compounds from a municipal water supply. Journal of the American Water Works Association, 82(5):79-84.

Heitz A, Kagi RI, Alexander R (2000). Polysulfide sulfur in pipewall biofilms: its role in the formation of swampy odour in distribution systems. Water Science and Technology, 41(4-5):271-278.

Herzing DR, Snoeyink VL, Wood NF (1977). Activated carbon adsorption of the odorous compounds 2-methylisoborneol and geosmin. Journal of the American Water Works Association, 69(4):223-228.

Ho L (2004). Removal of Cyanobacterial Metabolites from Drinking Water using Ozone and GAC. PhD Thesis. University of South Australia, Mawson Lakes Campus.

Ho L, Newcombe G, Croue J (2002). The reactivity of ozone with well characterised NOM fractions and its effect on the destruction of MIB and Geosmin. Water Research 36(3)511-518.

Khiari D, Bruchet A, Gittelman T, Matia L, Barrett S, Suffet I, Hund R (1999). Distribution-generated taste-and-odor phenomena. Water Science & Technology, 40(6)129-133.

Krasner S (1995). The use of reference materials in sensory analysis. Water Science & Technology, 31(11):265-272.

Lalezary S, Pirbazari M, McGuire MJ (1986). Oxidation of five earthy-musty taste and odor compounds. Journal of the American Water Works Association, 78(3):62-69.

Mackey ED, Baribeau H, Fonseca AC, Davis J, Brown J., Boulos L, Crozes GF, Piriou P, Rodrigues JM, Fouret M, Bruchet A, Hiltebrand DJ (2004). Public perception of tap water chlorinous flavour. AWWA Research Foundation, Denver, CO.

McGuire M (1995). Off-flavor as the consumers measure of drinking water safety. Water Science & Technology, 31(11):1-8.

Newcombe G (2006). Removal of natural organic material and algal metabolites using activated carbon. In: G Newcombe, D Dixon (eds), Interface Science in Drinking Water Treatment: Theory and Applications, Interface Science and Technology Series, 10, Elsevier, Amsterdam, the Netherlands.

Piriou P, Mackey ED, Suffet IH, Bruchet A (2004). Chlorinous flavour perception in drinking water. Water Science & Technology. 49(9):321-328.

Suffet I, Khiari D, Bruchet A (1999). The drinking water taste and odor wheel for the millennium: beyond geosmin and 2-methylisoborneol. Water Science & Technology, 40(6):1-13.

Sugiura N, Iwami N, Inamori Y, Nishimura O, Sudo R (1998). Significance of attached cyanobacteria relevant to the occurrence of musty odor in Lake Kasumigaura. Water Research, 32(12):3549-3554.

Ventura F, Matia L, Romero J, Boleda M, Marti I (1995). Taste and odor events in Barcelona’s water supply. Water Science & Technology, 31(11):63-68.

Whitfield FB, Shaw KJ, Walker DI (1992). The source of 2,6-dibromophenol: cause of an iodoform taint in Australian prawns. Water Science and Technology, 25(22):131-138.

Young WF, Horth H, Crane R, Ogden T, Arnott M (1996). Taste and odour threshold concentrations of potential potable water contaminants. Water Research, 30(2):331-340.