The Causes of Tastes and Odors
in Drinking Water
RICHARD D. HOAK
Senior Fellow
Mellon Institute
Pittsburgh, Pennsylvania
When people complain about a bad taste or odor in tap water their
dissatisfaction often arises from a difference in the flavor they have been
accustomed to. Actually, the taste of water varies considerably from city
to city. It is therefore a subjective measure of water quality, because
potability and tastelessness are not synonymous. These considerations
apply to moderate differences in flavor, for tastes and odors sometimes
occur which are unpleasant by any criterion.
Twenty years ago Baylis said, "Microscopical organisms and decaying
vegetation probably are responsible for more tastes and odors than any
other cause." Nevertheless, it has been customary to blame industrial
wastes for malflavors in water taken from streams subject to this kind
of pollution. It is not unreasonable for waterworks operators to take this
position, because there have been notable examples of taste and odor
problems resulting from excessive discharge of waste from manufacturing
plants. In recent years, however, there has been a trend back to Baylis'
opinion. This has come about as a result of increasing knowledge of the
contributions from algae, the Actinomycetes, decaying vegetation, and
biological activity in streams.
Among the various constituents of industrial wastes, none has been
as widely blamed as phenol for causing tastes and odors in water. This
resulted from the finding that partial chlorination of low concentrations
of phenol created an unpleasant medicinal odor. Since chlorination is
almost universally practiced at municipal waterworks, it was logical to
assume that occurrences of medicinal odors were caused by phenol.
Development of a sensitive analytical method for phenolic substances
was a significant factor in attributing medicinal tastes and odors to the
presence of phenol. The distilled 4-aminoantipyrine method measures all
substances which couple with the reagent to form a dye. Although it is
generally recognized that the method is non-specific, it has become customary to report results as phenol. This has led to the inference that
whatever is determined by the method is actually phenol, and that such
229