In affluent nations, there are two principal approaches to control the microbial water quality in the distribution network. The first approach is to maintain a disinfectant residual in the water during distribution to provide a barrier against ingress of microbial contaminants and limit microbial growth. The second approach is to control the risk of ingress through strict maintenance of the physical and hydraulic integrity of the network and to control the growth of microbes by distributing biologically stable water and using materials that do not leach nutrients. In the latter approach, drinking water is distributed without the presence of a disinfectant residual. This approach is found in some European countries, the former in most of the other affluent nations.
The different approaches have been the subject of debate that amalgamated in the late nineties (Trussel, 1999, van der Kooij et al 1999a, 1999b, Haas, 1999, LeChevallier, 1999). In these debates, arguments pro and con distribution with a disinfectant residual are highlighted. The most important arguments for the use of a disinfectant residual are:
– The presence of a residual disinfectant reduces the risk of microbial contaminants that may enter the distribution network. With the increasing complexity of the distribution network, the costs of the network and the open nature and aging of the infrastructure a residual disinfectant is necessary to inactivate microbial pathogens that may enter the network through cross-connections, mains breaks, repairs and leaks. Also for smaller systems with limited resources, a disinfectant residual is a relatively simple and cheap solution to improve the microbial safety.
– The presence of a residual disinfectant controls the growth of microorganisms in the network. It limits non-compliance with microbial water quality standards such as total coliforms and heterotrophic plate counts. It is also argued that limiting the amount of nutrients in the water and materials is difficult and would require substantial investments in additional treatment (LeChevallier, 1999) and focus on carbon-sources alone may not be enough (Haas, 1999). Also, most networks now in use have been installed in the past decades with the materials of that time and a control strategy should also consider this situation.
– The presence of a residual disinfectant may serve as sentinel for a breach of integrity of the system. When a distribution network is monitored with a consistent sampling strategy (or even on-line sensors) a reduced disinfectant residual concentration may signal that a contamination event has occurred. The most important arguments against the use of a disinfectant residual are:
– Disinfectants react with organic and inorganic compounds in the water and this creates disinfection by-products (DBP) in small quantities. DBP formation depends on many factors, but DBP such as trihalomethanes (THM) are found in the vast majority of cases of chlorination. More than 600 DBP have been identified (Richardson et al., 2008). Several (groups of) by-products have been associated with illnesses in humans. Some of the DBP (bromate, NDMA, benzaldehyde) show carcinogenic activity in longterm animal studies and several others are classified as possible carcinogens. A large body of epidemiological literature is accumulated over the years and meta-analyses have been conducted to assess the health risk of DBP. Lifetime exposure to chlorinated drinking water is associated with bladder cancer with a risk level of approximately 1 in 1000 (Hrudey & Charrois, 2012). For colon and rectal cancer and also for reproductive health outcomes results are less univocal. A recent review (Nieuwenhuijsen, 2009) indicated that there appears to be some evidence for an association between exposure to DBPs, specifically THMs, and congenital health, particularly in relation to the health effects little for gestational age/intrauterine growth retardation and, to a lesser extent, pre-term delivery, but evidence for relationships with other outcomes such as low birth weight, stillbirth, congenital anomalies and semen quality is inconclusive and inconsistent.
– The reaction of disinfectants with organic compounds in the water may also yield compounds, such as halogenated phenols and anisoles that give rise to taste and odour complaints by consumers. Taste and odour complaints are the most frequent cause of consumer complaints and consumers have a negative opinion about chlorinous taste and odour, both in terms of aesthetics and safety (Crozes et al., 2007).
– The sensitivity of pathogens to the disinfectants used in the network differs. The disinfectants used are effective against bacterial pathogens, less so against viruses and even less so against parasitic protozoa (Payment, 1999 EPA, 2012). Chlorine and chloramine are not effective against Cryptosporidium. (LeChevallier & Au, 2004).
– The use of a disinfectant residual may mask the failure of the integrity of the system and ingress of microbial contamination (Craun & Calderon, 2001). Water quality testing for coliforms or E. coli that are very sensitive to chlorine may indicate that the water is not contaminated, while infectious pathogens that are more resistant may be present in the water. This is of particular importance in samples taken after repairs and maintenance.
– Disinfectant residuals are not very effective against microbes in biofilms on pipe walls or sediment. The disinfectants react with the biofilm matrix but do not reach the microbes (LeChevallier et al., 1988). Also microbes in biofilm particles that slough of the wall and enter the bulk water again are more difficult to reach and inactivate (Behnke et al., 2011).
– Disinfection is targeting the symptoms rather than the cause of the microbiological issues in the network. The cause of ingress is insufficient hydraulic and structural integrity and hygiene. The cause of biofilm formation is the quality of the treated water and the materials used in the network. Targeting the cause is more effective and is also not sensitive to failures in the disinfection.
– In many settings, disinfectant residuals are not maintained throughout the entire network. That means that only a part of the network and consumers is protected by the presence of a residual and part is not (Payment, 1999, Gauthier et al., 2001). In such settings, disinfectant residuals may even enhance regrowth as they react with the organic compounds in the water and produce compounds that are more readily biodegradable (Skadsen, 1993; Zang and DiGiano, 2002).
– The use of toxic chemicals such as chlorine and chlorine dioxide requires production and may require transportation of these chemicals with a risk of accidents and spills.
5.1.3 Microbial safety in water legislation in the Netherlands Water legislation in the Netherlands has adopted Quantitative Microbial Risk Assessment as the central approach to demonstrate the microbial safety of (surface) water supply systems (Anon, 2001, 2011). Water utilities have to monitor the presence of reference pathogens (Campylobacter, Cryptosporidium, Giardia, culturable Enteroviruses) in the source water and demonstrate the adequacy of the treatment processes to remove these pathogens to produce drinking water that has a quality that corresponds with a probability of infection of less than 1 in 10,000 persons per year. Although not identified specifically in the legislation, maintaining this high quality through the distribution network implies the same safety level applies to the water delivered to the consumer.
5.2 Good engineering practice Good practices in construction, operation and maintenance of distribution networks is of paramount importance in protecting the water quality, both with and without a disinfectant residual in the network. When the network is in good condition and actively managed, this creates the proper conditions to consider distributing water without disinfectant residual.
Several enteric pathogens, particularly viruses and protozoa are (much) more resistant to chlorine and related disinfectants than indicator bacteria such as E. coli. This means that infectious pathogens may still be present in contaminated water, where the indicator bacteria for faecal contamination have been inactivated. Several outbreaks of viral and protozoal illness have been reported from water in which no E. coli was detected (Craun and Calderon 2001; Anderson and Bohan 2001).