Abstract:
Chlorine, added as a disinfectant in most surface water treatment plants, reacts with natural
organic materials (NOM) present in water and produces several disinfection byproducts
(DBPs) including trihalomethanes (THMs). The factors that promote the formation of THMs
include pH, temperature, NOM, chlorine dose, reaction time, bromide, etc. On the other
hand, presence of ammonia acts to suppress formation of THMs by reacting with free
chlorine, and thus reducing the amount of chlorine available for THMs formation.
In this study, trihalomethane formation potential (THMFP) at six surface water treatment
plants (WTPs) was evaluated; the plants included the Saidabad WTP, Chandnighat WTP,
Pirojpur WTP, Madaripur WTP, Narsingdi WTP and Moulovibazar WTP. The water quality
of these WTPs deteriorate progressively from January to May as dry season progresses,
though the raw water quality at the WTPs located outside Dhaka are far better than those
located within Dhaka (i.e., Saidabad and Chandnighat WTPs). THMFP at the selected WTPs
were estimated using three different models. THMFP at SWTP and Chandighat WTP have
been found to be much greater than those at the four WTPs outside Dhaka, because of higher
concentration of organic matter in raw water. However, none of the models incorporate effect
of ammonia in the predictive equations.
In the present research, effect of ammonia on THMs formation was studied through carefully
controlled laboratory experiments. Three sets of experiments were carried out, with five
samples in each set. For any particular set, DOC/NOM (added in the form of humic acid),
chlorine dose, and reaction time were kept constant, while ammonia was varied from 0 to 40
mg/l. The results of the laboratory experiments showed very distinct and significant effect of
ammonia on formation of THMs. The effect of ammonia was particularly pronounced at low
doses and diminished gradually as ammonia concentration increased. Presence of ammonia at
concentrations up to 5 mg/l led to a drastic drop in THMs formation. Available models,
particularly the widely used model of Amy et al. (1998), significantly over-estimated the
THMFP in the presence of ammonia. Since none of the existing predictive models for THMs
formation considers ammonia, an effort was made to incorporate the suppressive effect of
ammonia by modifying the model by Amy et al. (1998). Multiple regressions were used to
estimate the model parameter. The developed model was able to describe the experimental
data quite well. The proposed model could be considered as a significant improvement over
the existing model of Amy et al. (1998), and could be used for estimation of THMs formation
in the context of Bangladesh, where high ammonia is often encountered in the raw water at
surface water treatment plants.
