Traditionally engineers have considered treating water with 0.5 mg/L of free chlo-rine for 30 minutes as being adequate disinfection. Today, disinfection is defined in much broader terms, from physical and chemical removal of particles to chemical inactivation of microorganisms in the plant. With knowledge of the infectious dose for causing disease in humans and with an estimate of the concentration of parasites in the source water, the number of log-units of inactivation by the treat-ment processes can be determined to reduce the risk of waterborne disease to some level such as 1:10,000 or 1:100,000 cases per year. Consequently, source water quality will affect the required degree of treatment within the treatment facility.

With concern about raw water quality, water source protection is being promoted as part of the multiple barrier concept.

Whether the challenge is from bacteria or viruses or encysted protozoa, it is correct to consider these particles much as colloidal clay or macro molecules can be considered particles. Therefore, a major barrier in the disinfection of drinking water is in particle control - in other words, coagulation, flocculation, sedimentation and filtration. Understanding the chemistry of coagulation with good process control followed by well-designed (hydraulically speaking) flocculators and clarifiers will lead to excellent preparation of the particles for sedimentation and filtration removal.

A current American Water Works Association Research Foundation (AWWARF) project is travelling the United States collecting particle count, size and distribution data from a number of diverse surface water treatment plants. Their preliminary findings suggest that physical/chemical removal of particles in the Giardia and Cryptosporidium size ranges can be up to 4 log-units or more if the processes are well run, have tight operational control parameters, and have filter effluent turbidities less than 0.1 or 0.2 NTU. This is a remarkable observation given that the current Surface Water Treat-ment Rule in the United States gives a 2.5 log-unit credit for Giardia removal if there is coagulation, flocculation and sedimentation followed by filtration with less than 0.5 NTU of turbidity in the filtrate. The physical chemical removal processes can be considered to account for about 4 log-units of physical removal of parasites provided there is good process control and the processes are maintained in good operating order.

In small communities where a complex conventional surface water treatment plant is not practical, chlorination alone is not sufficient to protect public health when encysted protozoa are the issue. Particle control can be achieved with the simple yet effective slow sand filtration process or with high technology membrane processes such as ultrafiltration or microfiltration.

Figure 3. Giardia cysts after exposure to high concentrations of ozone.	Figure 4. Cryptosporidium oocysts after exposure to high concentrations of ozone.

While a long discourse on source water protection is possible, this article will focus on common sense. Sanitary surveys of the source watershed will document risks from sewage and from agricultural or other sources of contamination. An alert utility will document these potential risks and design their water intake and subsequent processes to account for "worst case" scenarios. In other words, a well-run utility will be proactive in anticipating deficiencies and taking corrective action before there is a serious problem. Regulation will not help this common sense approach.

While the water industry has struggled to redesign chlorination to be effective in controlling encysted parasites, there frequently has been inadequate protection against challenges from these pathogens. Many outbreaks of giardiasis have been attributed to surface water supplies that have only received chlorination. Cryptosporidium oocysts are unaffected by free chlorine or monochloramine under conditions found at most treatment facilities.

Simple concentration and time (CT) products for free chlorine and monochloramine for 2 log-unit inactivation of Giardia have been reported to be about 40 mg·min/L and 740 mg·min/L, respectively (pH 7, 20°C). Ozone has been shown to be effective for killing Giardia cysts and appears to act at the surface of the oocyst, as shown in Figure 3. In an AWWARF sponsored recontact time of about 5 minutes and an ozone residual of about 0.5 mg/L (apparent CT of 2.5 mgmin/L) greater than 2 log-units of inactivation could be obtained consistently. Chlorine dioxide has also been used to kill Giardia. Greater than 2 log-units inactivation has been obtained using a CT of about 5 mg min/L.

During the past five years there has been a significant amount of new information become available on the chemical inactivation of Cryptosporidium. Ozone is a very effective disinfectant of oocysts. Like Giardia, ozone appears to act on the surface of the oocyst as illustrated in Figure 4. Ozone conditions are about four times greater than those required for Giardia. Chlorine dioxide is second to ozone in effectiveness and has a number of advantages when compared with ozone. Low capital cost and ease of operation may make it an attractive alternative to free chlorine when only chlorine is used to treat the water. The design requirements for chlorine dioxide are about 10 times those required for Giardia.

One of the most interesting findings is that when disinfectants are used sequentially as happens in practice, there is greater inactivation than could be explained by the use of the single disinfectants. This synergism is the topic of another AWWARF research project conducted at the University of Alberta.

Another outcome of this research was the finding that when free chlorine was used first followed by monochloramine, detectable inactivation of C. parvum occurred. It has been found in the same study that Giardia muris, a useful surrogate for G. lamblia, is more readily inactivated when using more than one oxidant sequentially.

However, much is left to be discovered about the use of chemical disinfectants in the control of waterborne parasites. Questions related to the effect of temperature, pH, and water quality remain to be answered. Fundamental research into the mechanisms of infection and the effect that chemical oxidants have on this process may help to devise more effective methods of controlling the parasites in drinking water.

Using the global vision of disinfection as a series of barriers, future process design and operation will need to focus on quality control, using well designed and operated unit processes within a treatment facility. More potent chemical disinfectants such as ozone or chlorine dioxide will become more widely used in water treatment. While chlorine's days as a primary disinfectant of surface waters may be numbered, it will still serve an important role in protecting the quality of water in distribution systems.

The use of monochloramine following ozone treatment may result in savings in operating costs once the synergism between ozone and monochloramine has been well documented. Where feasible, source water protection could be promoted in some communities. However, this becomes a major political undertaking unless the watershed is owned or wholly controlled by the water supplier. Nonetheless, proactive utility managers will anticipate weaknesses in their systems and take corrective action to prevent the embarrassment of a major waterborne outbreak of parasitic disease.

Small systems will need to look towards more sophisticated technology than simple chlorination of surface waters such as slow sand filtration or membrane processes to simply and efficiently produce a drinking water that has low microbial risks.

And one final word: pathogenic bacteria were easily controlled until viruses became an issue in the post-war years. Adjustment in chlorination levels adequately protected public health. Then the waterborne parasites discussed in this paper became the issue. They are not so easily controlled and require new thinking, modern technology and quality control. Beware, the next "super bug" may be around the corner.

** See ES&E July 1996 for part one.

This article is condensed from original published in Environmental Science & Engineering, September 1996.

For further information, contact Gordon Finch at E-mail: Gordon.Finch@ualberta.ca