Page 2: Guidelines for Canadian Drinking Water Quality: Guideline Technical Document – Colour

Definition and Measurement

The appearance of colour in drinking water is caused by the absorption of certain wavelengths of normal white light by dissolved or colloidally dispersed substances, by fluorescence in the visible wavelength region from substances that absorb white or ultraviolet light,Footnote 1 by the presence of coloured suspended solids, and by the preferential scattering of short wavelengths of light by the smallest suspended particles.Footnote 1,Footnote 2 Colour measured in water that contains suspended matter is defined as "apparent colour"; "true colour" is measured in water samples from which particulate matter has been removed by centrifugation.Footnote 3,Footnote 4 In general, the true colour of a given water sample is substantially less than its apparent colour.Footnote 4

The colour of a water sample is measured by visual comparison with a series of standard solutions containing known amounts of potassium chloroplatinate and added cobalt (II) chloride; the amounts of the latter are varied to match the normal yellow to brown hue of the test water at a given location.Footnote 4 As the platinum-cobalt standard method was designed to analyse naturally coloured water, other methods are employed to measure colours originating predominantly from industrial wastes or coloured minerals.Footnote 4 Difficulty in comparing the colour of a public water supply with standard colour solutions may therefore be indicative of pollution. Methods based on chemical oxidation,Footnote 5 absorption spectrophotometry,Footnote 6 for example, can also be employed to estimate the concentrations of organic substances that may cause colour in water. In general, however, very poor correlations are obtained between chemical results and colour measurements for water from different locations.Footnote 1,Footnote 10 This result should not be unexpected, as identical colours can be obtained by the standard method for waters containing different proportions of colouring agentsFootnote 10; that is, the values obtained by the standard platinum-cobalt method are not specific.

One true colour unit (TCU), or platinum-cobalt unit, corresponds to the amount of colour exhibited under the specified test conditions by a standard solution containing 1.0 mg of platinum per litre.Footnote 4 Thus, a sample exhibiting a colour of 15 TCU has a closer colour match with a standard containing 15 mg of platinum per litre than with standards containing 14 or 16 mg/L. A colour of 15 TCU can be detected in a glass of water by most consumers, and 5 TCU will be apparent in large volumes of water, such as in a bathtub; few people can detect a colour level of 3 TCU.Footnote 11

One major factor that affects the colour of natural surface water is pH. The increase in colour with increasing pH of the test sample is commonly referred to as the "indicator effect,"Footnote 1 and it is widely recommended that the pH of the sample be recorded together with the colour measurement to allow for this effect.Footnote 4 The alternative is that all colour measurements be done at a standard pH of 8.3.Footnote 12 Such standardization would be of value only in individual treatment plants, however, owing to the non-specific nature of the standard method.

Page details

Date modified: