Modification of water treatment processes for chlorine stability and disinfection by-products control

Supply of drinking water is facing challenges in terms of quantity and quality of raw water and from the increasingly stringent water quality standards, the supplied water has to comply with. The goal is to produce water that is stable enough so that disinfectant, added to control bacteriological risks, can be maintained through a distribution system without producing an excessive amount of disinfection by-products (DBP). Trihalomethanes (THM) and haloacetic acids (HAA) are the most commonly regulated DBP in water supply systems and are the DBPs water utilities are struggling to comply with. The presence of NOM (natural organic matter) and bromide in raw water are mainly responsible for chlorine loss and the formation of undesirable DBP during chlorine disinfection. New DBP with potentially higher health risk concerns often containing bromine are increasingly reported. The DBP guidelines continue to be reviewed, introducing more stringent DBP limits and adding new compounds. Studies indicate that both NOM and bromide concentrations in raw water sources are increasing with time, putting pressure on the treatment and maintenance of disinfectant residual while complying with DBP guidelines. The majority of water treatment plants across the world adopt a conventional treatment train consisting of coagulation, sedimentation and filtration (CSF). The CSF configuration removes some NOM but is not sufficient to achieve the goal in many cases and hence the enhanced removal of NOM is needed. Even worse, bromide cannot be removed by the CSF. These plants are (or will soon be) under pressure to make incremental upgrades to improve water quality before the need to switch to new and more expensive technologies. This work focuses on options available for improvement of the treated water quality in CSF based water treatment plants and distribution systems. Since bromide cannot be removed by the CSF, improvements were first targeted on NOM removal but later bromide removal and additional processes are attempted. First, an approach is developed to evaluate available technological options to achieve the generic water quality goal – the last customer receives water containing sufficient disinfectant residual and lower DBP than the regulatory limit. The approach included understanding the target NOM removal for a given situation and an approach to select the best available technology to remove NOM. Traditionally, DBP are measured for regulatory purposes, but treatment process is selected and eventually modified (i.e. the target NOM removal adopted) based on either trihalomethane formation potential (THMFP) or HAA formation potential tests. These tests measure the maximum amount of THM/HAA which can form under the worst possible conditions (often elevated temperature and chlorine concentration). Such an approach fails to describe the optimal scenario treatment plants should operate. NOM, chlorine concentrations, retention time and temperature in the water supply system are important factors that determine chlorine reaction rates and concentrations of DBPs formed. The proposed methodology combines the DOC (dissolved organic carbon) measurements with the existing two organic component chlorine decay and THM formation models based on laboratory tests to evaluate chlorine and THM profiles in a given distribution system under specific conditions. DOC measurements are used to quickly optimize the water treatment process and the chlorine and THM models are used as the best available predictors of chlorine and THM profiles in a distribution system. This approach enables the assessment of the proposed changes and optimization of a water treatment plant – network operation for specific conditions. The evaluation of technologies identified enhanced coagulation, a treatment where coagulant salts are added beyond the requirement to remove turbidity, is the best available technology (BAT). More specifically the enhanced coagulation using ferric salts at low pH is the identified as BAT for reduction of NOM. While it serves as an extension rather than an addon process to CSF, it does introduce additional coagulant salts and thus increases the salinity and the sludge which needs disposal. Ferric sludge recycling could improve performance of enhanced coagulation under some conditions. The experiments with various regenerating chemicals (sodium hydroxide and others) have proven that NOM can be washed from the collected ferric hydroxide sludge. The regenerated coagulant could be reused for coagulation of colloidal particles and removal of NOM. The regenerated ferric hydroxide is less efficient in NOM removal than fresh ferric chloride (a larger dose of ferric hydroxide is needed to remove the same amount of NOM). However, since it is recycled within the process and it enables the reduction of sludge and chloride, it is viewed as a promising technology in specific situations such as water sources with low alkalinity and high NOM concentrations. The chlorine decay rates and the resulting concentrations and toxicity of the formed DBP are strongly influenced by bromide concentration in chlorinated water. A performed literature review on bromide removal could not identify a readily scalable bromide removal process suitable for a CSF-based plant. Analysis of the situation identified a couple of potentially promising research direction for bromide removal. The most promising out of these is derivatisation of bromide to bromamine and its removal. The two organic components chlorine decay model was in this work extended to include the impact of bromide concentration on chlorine decay rates. Application of low ozone dose (~2mg O3/L) before enhanced coagulation resulted in the lower required dose of ferric salt, lower treated water DOC, better chlorine stability and lower formation of THM. It was estimated that the use of pre-ozonation could achieve better water quality at lower chemical cost than the enhanced coagulation alone. The results of this work can assist in the modification of existing CSF based water treatment plants and improvement of the delivered water quality for a targeted distribution system