TOTAL VIABLE COUNT OF TREATED WATER IN THE KANDY MUNICIPAL DISTRIBUTION SYSTEM

Abatma r The general bacterial population of treated water in the Kandx distribution system was studied. The mean total viable counts at 3 5 O ~ and 20 C were found to be 80/ml and 1291rnl respectively. The mean total viable counts at 3 5 O ~ and 2 0 O ~ were not significantly different. The frequency of detection of coliforrns was increased to levels up to a total viable count of 500lrnl but decreased thereafter.


Introduction
Potable water of good bacteriological quality is generally associated with attainment of less than one total coliform/100 ml of water.Yet there are many other micro-organisms common to the flora of finished water whose numbers far exceed those of the coliform group.Some of these organisms may be a factor in creating health problems in very young and the debilitated persons in a c ~r n r n u n i t ~. ~~' In addition high non-coliform population in finished water have been implicated in suppressing coliform growth in test media.'t 3 f 5 Even though water treatment is adequate and Clp disinfection is provided, quality could deteriorate in the distribution system as a result of growth of organisms other than detectable coliforms.
Unless there is considerable collection of data within a country the total microbial count is difficult to use as a water quality standard.This study provides data with regard to the levels of non-fastidious heterotrophic bacteria in treated water in one of the main water distribution networks in Sri Lanka.

Experimental
Eighty-four samples of water from the Kandy water distribution system'was examined for the total viable count and the total coliforms.The samples were collected from street standposts which are directIy connected to the main and trmsported in accordance with the methods recommended by the WHO.O

Total viable count
One ml quantities of the sample and of the 10-I dilution were added to petridishes and 5 ml of glucose extract agar was poured over and mixed well.Duplicate plates for each dilution was used.After the plates were set, one set of plates were incubated at 35OC for 24 hours and the other set at 20°C for 48 hours.All colonies on selected plates (plates having colony numbers between 30 and 300) were counted after the incubation time.

Total coliforms
Most probable number technique using MacConkey broth was used in the estimation of total coliforms according t o the WHO re~ornmendations.~~

Mean total viable count in the Kandy Water Distribution System
The mean total viable counts, the standard deviation and the range of counts at the two incubation temperatures 35OC and 20°c are shown in Table 1.The mean total viable counts at 35OC and 20°c are 80 and 129 respectively.A significant difference of viable counts was not observed by the "t" test for paired observations.2 and 3. 23 and 25% of the samples have total viable counts < 10 at 35OC and 2 0 ' ~ respectively.

C. P. Kodikara
The data presented in Tables 2 and 3 show the relationship between the total viable count and the detection of coliforms at the two temperatures 3 5 O ~ and 25OC.At 3 5 O ~, the frequency of detecting coliforms does not show a significant increase after a total viable count of 500/ml (Table 2) and at 20°C the frequency of detecting coliforms increases upto a total viable count of 500/ml and decreases thereafter (Table 3).

Percentage of samples of water below a total viable count of 500/ml
Of the samples of the Kandy distribution system 77 to 78% were below 500/ml at 20°C and 3 5 O ~ respectively.In 13-14% of the total samples analysed it was unable to estimate an accurate count due to the plates being covered with proteus.

Discussion
The total viable count for drinking water as described in standard methods of American Public Health Association 1975, is the plating of small quantities (usually 1.0 ml or 0.1 ml) of a properly collected water sample in a nutrient agar medium and incubating aerobically for a fixed period at a prescribed temperature (35OC for 24 hours in 20°C for 48 hours).
Although no constant relationship appears to exist between standard plate count and the number of potentially pathogenic invaders that might be present, it is logical to assume that change occurences are proportionally greater as the general bacterial population increases.In addition high non coliform populations in finished water have been implicated in suppressingcoliform growth in test media.'f3i5 Therefore water having a high noncoliform population would indicate a negative coliforin test giving the consumer a false sense of security.The initial level of suppression of coliforms has been shown to be a general bacterial population exceeding 1000/m1.~Therefore an establishment of a standard plate count limit of 500/ml in the distribution water has been considered as a useful m e a s ~r e .~ The results of the present study shows that the mean total viable count of Kandy distribution water is below 500/ml and therefore the adverse influence of excessive bacterial population interfering on routine monitoring procedures of the Kandy water distribution system could be considered very low.The present study too shows that the frequency of detection of coliforms increasing as the standard plate count increased to levels upto 500/ml but decreasing thereafter (Table 3).
The total viable count could be considered as a quality control of water treatment processes and the efficacy of sanitation measures us d on distribution lines.~i s c h e r ~ ,indicates that a total viable count &low 500/ml could be maintained in the distribution lines by maintaining a residual C12 level of approximately 0.3 mg/l throughout &e-_distribution lines.Of the samples of the Kandy distributiori water 77 to 78% showed counts below 500/ml at 20°C and 3 5 O ~ respectively, indicating that the deterioration of water quality in the distribution system is very low.
Water used in food, beverage cosmetic and drug industries should be of a higher standard with regard to the total viable count becau'se these organisms create taste, odor and spoilage problems in the finished products.2It has been suggested that the water used in food processing should have a plate count of less than 50 bacteria/ml.'With regard to the water in the Kandy distribution system supplemental treatment procedures are required if the water is used for food, beverage, cosmetic or drug industries to maintain good keeping properties of the products.
The importance of using two temperatures of incubation namely 20°C and 35OC in determining the total viable count is that 20°C covers the growth requirements af most natural soil and water bacteria and S5OC the organisms of animal human reservoir.In the present study the total viable counts at 2 0 ' ~ and 3 5 ' ~ were not significantly different.Theoretically, an .increased count at 35OC is more critical from a hygienic point of view than an increased count at 20°C.

Table 2 .
Bacterial Plate Counts at 35Oc Versus Colifonn detection in water in the Kandy water distribution system

Table . 3
. Bacterial Plate count' at 20°c Versus Coliform detection in water in the Total viable count versus coliform detection