BY BETH HAMM fieldreport

DBP reduction using mixed oxidants generated on site

MIXED OXIDANTS GENERATED ON SITE HAVE A DEMONSTRATED TRACK RECORD OF REDUCING BOTH TOTAL TRIHALOMETHANES AND HALOACETIC ACIDS WITHOUT THE ATTENDANT PROBLEMS ASSOCIATED WITH BROMATES AND CHLORITES OR ISSUES REGARDING SAFE HANDLING. he Safe Drinking Water Act (SDWA) moacetic acids. Chlorite and bromate are was established by the US Environ- limited to 1.0 mg/L and 10 µg/L, respec- mental Protection Agency (USEPA) tively. These limits affect all public water T in 1974 to protect the quality of the systems that add a , regardless nation’s drinking water. Although the use of their size. The rule went into effect in of in drinking water has dras- January 2002 for large surface water sys- tically reduced disease, these disinfectants tems and will go into effect in January can react with material in the raw water 2004 for groundwater and small surface and form disinfection by-products (DBPs) water systems (USEPA, 1998). Table 1 that are also haz- shows the maximum contaminant levels ardous to health. set by the D/DBPR for TTHMs, HAA5, Initially, the gov- chlorite, and bromate. ernment focused on reducing formation MIXED OXIDANTS of chloroform, bro- Onsite generation of a mixed-oxidant modichloromethane, solution, which consists primarily of dibromochlorometh- , uses a process similar ane, and bromo- to onsite generation of sodium hypochlo- form—all compo- rite. A brine solution is fed through an nents of a category electrolytic cell, power is applied, and the designated as total resulting oxidant solution is created. The trihalomethanes style of electrolytic cell and the operating (TTHMs). TTHMs parameters determine the formation of have been found to straight versus formation of cause liver, kidney, mixed oxidants. Mixed oxidants, produced by and central nervous system problems, as generators such as this one, have well as an increased risk of cancer. In 1979, the disinfection power of USEPA set a TTHM limit of 100 µg/L for and have also been documented to TABLE 1 MCLs for Stage 1 D/DBPR* “large surface water systems,” defined as inactivate even chlorine-resistant systems serving more than 10,000 people. . DBP† MCL—mg/L The 1996 amendments to the SDWA attempted to further balance the risks TTHMs 0.080 between microbial pathogens and DBPs. HAA5 0.060 The Stage 1 Disinfectants/DBP Rule Chlorite 1.0 (D/DBPR) lowered the permissible limit for Bromate 0.010

TTHMs to 80 µg/L. It also established a *MCLs—maximum contaminant levels, D/DBPR— Disinfectants/Disinfection By-products Rule new limit of 60 µg/L for the five haloacetic †TTHMs—total trihalomethanes, HAA5—the sum of acids (HAA5), i.e., mono-, di-, and five haloacetic acids trichloroacetic acids and mono- and dibro- field report

Mixed oxidants have the disinfec- the mixed-oxidant solution to inacti- tion power of chlorine, but they have vate chlorine-resistant organisms (e.g., also been documented by numerous the Cryptosporidium parvum oocyst) agencies to inactivate even chlorine- or achieve substantially higher inacti- resistant microorganisms, much like vation levels of other organisms at a or does. In lower concentration × time (C × T) addition, because only salt and water than is required with hypochlorite. are used as feedstocks, the use of Superior inactivation capability of a mixed oxidants does not raise concerns multitude of organisms using the about hazardous chemicals or safety. mixed-oxidant solution has been Furthermore, the solution also leaves a demonstrated at a number of institu- The -encrusted pipe sample (left) was durable chlorine residual in the distrib- tions, as described in the following disinfected by chlorine gas; the nearly spotless ution system. section. pipe (right) was exposed to mixed-oxidant Because of the presence of chlorine Inactivation studies. C. parvum treatment. in the mixed-oxidant solution, the for- oocysts. Initial testing in 1996 was mation of TTHMs and HAA5 will still of the limitations of current analytical conducted at the University of Arizona occur. However, anecdotal evidence techniques and the complexity and in Tucson using the excitation method. from a number of installations shows interferences of multioxidant chem- Researchers achieved an approximate DBP reductions ranging from 30 to istry. Classical laboratory methods for 2-log10 inactivation, whereas 50%. Neither chlorite nor bromate has detection of other oxidants do not hypochlorite had no effect (Sterling, been detected in mixed oxidant–treated function properly in the presence of a 1993). water. Depending on the level of pre- large chlorine matrix; thus, when these In 1997, the University of North cursors in the raw water, mixed oxi- methods are used, studies on mixed Carolina at Chapel Hill and the Cen- dants may be a viable alternative for oxidants have detected only chlorine ters for Disease Control and Prevention DBP reduction that also provides a (Gordon, 1998). compared mixed oxidants and sodium chlorine residual without the safety Although the chemistry of the hypochlorite for inactivation of the C. and handling issues associated with mixed-oxidant solution is problematic, parvum oocyst via infectivity assays on traditional chlorination. the biology is compelling. The mixed- neonatal mice. Mixed oxidants inacti- Mixed-oxidant chemistry. Manufac- oxidant solution exhibits several per- vated the organism by more than 3.6 turers’ claims of species other than formance characteristics that show it log10 (>99.9%) at a concentration of 5 chlorine in the mixed-oxidant solution contains more than just chlorine. The mg/L × 4 h. Hypochlorite at the same have not been directly verified because most dramatic evidence is the ability of dosage and C × T achieved no inactiva- tion whatsoever (Venczel et al, 1997). In 2000, the University of Colorado in Colorado Springs demonstrated a FIGURE 1 Comparison of DBP formation with use of chlorine versus mixed oxidants 2–3-log10 inactivation using poly- merase chain reaction amplification of fluorescent-labeled heat shock genes. Chlorine Mixed oxidants The results in the report unequivocally indicated that treatment of live C. Las Vegas, N.M. parvum oocysts with mixed oxidants

Crossville, Tenn. for different periods inhibited expres- sion of the heat-inducible Hsp70 gene, Greenfield, Iowa and the extent of the inhibition showed direct correlation with the duration Santa Fe, N.M. and nature of the disinfection treat- ment. Again, hypochlorite had no North Table Mountain, Colo. effect whatsoever. In 2000, researchers at the Univer- 0 1020304050607080 TTHMs—µg/L sity of North Carolina at Chapel Hill DBP—disinfection by-product, TTHMs—total trihalomethanes presented their findings at the AWWA Water Quality Technology Conference. Mixed oxidants produced extensive field report FIGURE 2 TTHM reduction at Greenfield Municipal Utilities, Iowa

100 10/29/96—mixed-oxidant disinfection in 90 place three months; TTHMs in distribution only 14% higher than at plant 80 (>4-log10) inactivation of test ,

—µg/L bacterial spores, and within 70 1–10 min. The level of inactivation of Distribution C. parvum oocysts in 10 min ranged 60 from 0 to >3 log10, depending on the 50 Plant cell design, experimental conditions, 40 and experiment (Sobsey et al, 2000). 7/30/96—first mixed-oxidant sample TTHM Concentration (one week after startup); TTHMs in Legionella pneumophila. At a dose 30 distribution 36% higher than at plant of 2 mg/L and exposure of 10 min, 20 mixed oxidants achieved total kill at 6/25/00 7/15/00 8/4/00 8/24/00 9/13/00 10/3/00 10/23/00 11/12/00

pH 8.0, compared with bacteria con- TTHMs—total trihalomethanes centrations > 2 cfu/mL with (Barton, 1996). Subse- quent studies showed that mixed oxi- dants achieved significantly higher FIGURE 3 Comparison of dosage and residual for chlorine versus mixed oxidants inactivation than sodium hypochlorite was able to achieve (Bradford et al, 1997). 2.0 . At a dose of 2 mg/L and exposure of 10 min, 1.8 mixed oxidants achieved total inactiva- 1.6 tion, compared with bacteria concen-

trations of 1,200 cfu/mL remaining mg/L 1.4 when sodium hypochlorite was used 1.2 under the same conditions (Barton, 1996). Subsequent studies showed that 1.0 mixed oxidants achieved 1.6–3.7-log 10 0.8 higher inactivation than sodium hypochlorite (Bradford et al, 1997). 0.6

Bacillus stearothermophilus. At a Concentration— Chlorine 0.4 dose of 4 mg/L, exposure of 5 min, and Chlorine Dose 0.2 pH of 8.0, mixed oxidants achieved total Mixed-oxidant Dose kill, whereas with sodium hypochlorite 0 Mixed-oxidant Residual bacteria concentrations > 12 cfu/mL Chlorine Residual remained (Barton, 1996). Subsequent studies indicated that mixed oxidants Bloomfield, N.M. Greenfield, Iowa achieved 1.5–2.5-log higher inactiva- Las Vegas, N.M. 10 Santa Fe, N.M. tion than sodium hypochlorite (Bradford Lamar County, Texas et al, 1997). . In 1996, the Uni- North Table Mountain, Colo. versity of Arkansas Center of Excel- lence for Poultry Science at Fayetteville conducted a study at a major poultry plant approved by the US Department limit, with an average of 222 cfu/mL, Salmonella typhimurium. Studies of Agriculture (USDA). The research more than twice the permissible level by the University of Georgia’s Depart- explored the ability of mixed oxidants (Waldroup et al, 1996). Subsequent ment of Poultry Science in 2001 versus sodium hypochlorite to inacti- studies conducted in 2001 by the Uni- showed mixed oxidants achieved a 2.1- vate E. coli below the USDA’s regula- versity of Georgia’s Department of log10 greater inactivation at a 2-mg/L tory limit of 100 cfu/mL. Of the Poultry Science in Athens showed that dose than hypochlorite was able to mixed-oxidant samples, 100% were mixed oxidants achieved a 2.4-log10 achieve at the same dosage (Russell, below the limit, with an average value greater inactivation of E. coli than 2001). of 22 cfu/mL. By contrast, half of the hypochlorite did at the same dosage Staphylococcus auereus. The Uni- hypochlorite samples were over the (Russell, 2001). versity of Georgia studies indicated field report

TABLE 2 Microflocculation effects using mixed oxidants

Coagulant Consumption Effluent Turbidity—ntu TTHM* Levels

Previous Current Percent Percent Installation Site Previous Doses Current Doses Percent Reduction Levels Levels Reduction Reduction

Crossville, Tenn. 90 gpd (0.34 m3/d) 70 gpd (0.26 m3/d) 22 NA† NA NA 47 Greenfield, Iowa 14.7 mg/L 8.9 mg/L 40 0.107 0.065 39 >20 Las Vegas, N.M. 10.5 mg/L 7.5 mg/L 29 0.07 0.03 57 44 Midwestern site Data not available Data not available Data not available 2.0 0.4 80 NA Santa Fe, N.M. 90 mg/L 54 mg/L 40 0.60 0.18 70 >45

*TTHM—total trihalomethane †NA—not applicable

that mixed oxidants achieved a 1.4- phy, 1998; Mangold Environmental TTHM and HAA5 concentrations for log10 greater inactivation than Testing, 1997). Neither chlorite nor the first quarter of 2001 to 40 µg/L, a hypochlorite at a 2-mg/L dose (Rus- bromate, both of which are regulated decrease of 47 and 49%, respectively sell, 2001). under the D/DBPR, has been detected (Brownfield, 2001). Listeria monocytogenes. Results in mixed-oxidant treated water. The • Greenfield (Iowa) Municipal Util- from the University of Georgia showed only by-product that has been ities reported that switching to mixed that compared with hypochlorite, detected is chlorate, which is not a oxidants resulted in a reduction in mixed oxidants at a 2-mg/L dose regulated substance and has no TTHM formation ranging from 20 to achieved a 2.3-log10 greater inactiva- known adverse health effects. Chlo- 40%. Initially TTHMs in the distribu- tion (Russell, 2001). rate formation is minimal and is pro- tion system were ~36% higher than Bacillus anthracis. Studies at the US duced in the range of 15 to 33 µg/L TTHM levels at the plant. After two Army Dugway Proving Grounds in per 1-mg/L dose as free available chlo- months of treatment with mixed oxi- Dugway, Utah, are in progress. Prelimi- rine (FAC), which is comparable to or dants, TTHMs in the distribution sys- nary results indicate inactivation of B. less than the levels found in commer- tem were only 14% higher than levels anthracis, Yersinia pestis (plague ), cial (Environmental Health at the plant and were reduced even fur- Klebsiella terrigena, Francisella Laboratories, 1995). ther as organic matter in the distribu- tularensis, and Vaccinia (smallpox) • Las Vegas, N.M., replaced chlo- tion system was eliminated (Duben, (Wright et al, 2001). rine gas with mixed oxidants for final 1996). DBP reduction. Another key indica- disinfection and also added mixed oxi- • The Sangre de Cristo Water Com- tor for the presence of oxidants other dants in pretreatment where previ- pany (Santa Fe, N.M.) also uses mixed than chlorine in the mixed-oxidant ously no disinfectant had been used. oxidants in pretreatment and final dis- solution is reduction in TTHMs and With chlorine gas, TTHM concentra- infection. The facility was unable to HAA5, as has been demonstrated at a tions averaged 80 µg/L. Within three pretreat with chlorine gas because of number of water utilities that have months after conversion to mixed oxi- excessive TTHM formation. Despite replaced conventional chlorination dants, TTHM formation was reduced the fact that mixed oxidants are with mixed-oxidant technology. to 45 µg/L, a 44% reduction. This injected preclarifier and in the clear- Although this phenomenon is not seen reduction was seen despite the fact well, whereas chlorine was only at every site, no utility using mixed that mixed oxidants were also being injected at a single point, TTHMs have oxidants has observed DBP concentra- added to the clarifier, something that been cut from an average of 60 µg/L tions above the levels produced by the utility could not do with chlorine with spikes >80 µg/L to an average of chlorine. In fact, in the great majority gas because of excessive TTHM for- only 33 µg/L, a reduction of ~50% of mixed-oxidant installations, both mation (Armijo, 2000). (Herrington et al, 1999). TTHMs and HAA5 are typically • When disinfecting with chlorine, • North Table Mountain Water and reduced by 30–50%, in contrast to lev- the Holiday Hills water treatment Sanitation District (WSD) in unincor- els formed with chlorine. plant in Crossville, Tenn., averaged 75 porated Jefferson County near Golden, A number of mixed-oxidant sites µg/L of TTHMs and 79 µg/L of HAA5 Colo., converted its filtration and dis- have also tested for production of in the first quarter of 2000. After con- infection systems in two stages. The fil- chlorites, chlorates, and bromates version to mixed oxidants late in the tration system achieved significant (Herrington et al, 1999; Yu & Mur- year, the facility reduced both its reduction in the formation of TTHMs field report

(37%) and HAA5 (30%). Subsequent and clarifiers. Biogrowth that was more contaminated line. However, the conversion to mixed-oxidant disinfec- able to survive with chlorine as a dis- pipe exposed to mixed-oxidant treat- tion from chlorine gas achieved an infectant is typically eliminated ment was spotless, as though it had additional 44% reduction in formation shortly after conversion to mixed oxi- never been used. of both TTHMs and HAA5. The com- dants. When a mixed-oxidant genera- • The KOA Kampground in Great bined effect of the new filtration and tor is first installed, flushing of the Falls, Mont., originally used hypochlo- mixed-oxidant disinfection was a total distribution system may be required to rite for disinfection of both its swim- decrease of 64% for TTHMs and 61% remove detached biosolids until the ming pool and potable water supply. for HAA5 (Jeschke, 2000a). system has stabilized. The stabiliza- The facility reported that conversion to Figure 1 compares DBP formation tion process usually takes one to two mixed oxidants led to the removal of with chlorine and with mixed oxidants months, and then the lines remain free previously accumulated biofouling in at these five sites. of biofilm or algae as long as mixed the distribution system. Black slime oxidants are used. that previously had formed in the THEORIES FOR BEHAVIORAL DIFFERENCES The ultimate effect is that reduced showers when the facility was using Laboratory testing and field installa- biogrowth decreases organic material hypochlorite disappeared. With tions have found that mixed oxidants that can react with chlorine, thus hypochlorite, loss of distribution sys- and chlorine exhibit significant behav- reducing DBP formation. Elimination tem pressure during frequent power ioral differences, although only chlo- of in distribution lines would outages resulted in biofouling in the rine can be measured in the mixed-oxi- mean that DBP levels in the distribu- pipelines, and the distribution system dant solution. Several explanations tion system would not be substantially always needed to be flushed when pres- have been proposed for the greater higher than levels seen at the treatment sure was regained. In contrast, since reduction in DBP formation achieved plant. Data from a study conducted by the campground switched to mixed by mixed oxidants in contrast to H.R. Green Engineers demonstrate this oxidants, the distribution system hypochlorite. These theories include effect (Duben, 1996). Over the three requires no flushing, and no discol- preferential action of nonchlorine oxi- months of the study period using oration is evident after a power failure. dants, elimination of biofilm, improved mixed-oxidant disinfection, the levels With mixed oxidants, algal growth has chlorine residual at a reduced dosage, of TTHMs in distribution water not formed on the pool area surface, and microflocculation to remove dropped more rapidly than the levels at thus eliminating the need for an algae- precursors. the plant. When the site first converted cide, a common requirement for out- Preferential-acting nonchlorine oxi- to mixed oxidants, TTHMs in the dis- door pools disinfecting with hypochlo- dants. Chlor- species other than tribution system were 36% higher than rite (Crayton et al, 1997). chlorine in the mixed-oxidant solution levels at the plant. After three months • The city of Las Vegas, N.M., also may be reacting with organic material of continuous operation, distribution reported substantial improvements in the water, reducing oxidant demand system TTHMs were only 14% higher with mixed-oxidant chemistry. When and eliminating some of the material than at the plant. Overall, TTHM the utility used chlorine gas, a 1.5–2 in. that causes TTHM and HAA5 forma- reduction in the distribution system (38–51 mm) thick algae mat would tion. These other species would be was twice as great as at the plant. Fig- grow in the clarifying basins. Staff had faster-acting and preferentially react ure 2 summarizes TTHM reduction at to drain the tanks every few weeks to with components in the water before the utility. scrub off the algae, which would peel the presence of chlorine in the mixed- Biofilm reduction with mixed oxi- off in sheets. The cleaning process took oxidant solution could cause an dants has been observed at a number ~20 h. Since conversion to mixed oxi- adverse effect. Because these other of other installations. dants, the facility has not seen any evi- species are short-lived and minute in • At the Diana Water Supply Corp. dence of algae growth (Armijo, 2000). comparison to the amount of chlorine in Diana, Texas, two line breaks • Hazlet, Sask., converted from present, they are not sufficient to occurred simultaneously. Both lines sodium hypochlorite on its well site to achieve total removal. Thus, the drew water from the same aquifer, but mixed-oxidant generation. With remaining organic material reacts with at that time the utility used chlorine hypochlorite, the facility experienced the remaining chlorine, producing gas to disinfect one line and mixed oxi- severe water quality issues and black TTHMs and HAA5, although at dants to disinfect the other line. The film growth on the cistern ladder and reduced levels. This theory is specula- line disinfected with chlorine contained walls. Within one month of conversion tive and has not been proven. biofilm only 200 ft (61 m) from the to mixed-oxidant disinfection, the Biofilm elimination. There is signifi- disinfection station. The line break in black film on the cistern ladder and cant anecdotal evidence of the ability the mixed-oxidant line occurred ~0.5 walls was nearly gone (Sletten, 2000). of mixed oxidants to clean biofilm mi (0.8 km) from the disinfection sta- • At a pilot-test site in the Mid- and algae from distribution systems tion, so the expectation was an even west, mixed oxidants were fed to one of the four standard clarifiers1 at the ments have decreased, there is less enabled the utility to reduce the dose plant. Within 8 h, biogrowth removal chlorine in the system that can react by 15 to 30% (between 1.4 and 1.7 on tubes was observed, and within a with organic matter. The end result is a mg/L) and still maintain a higher chlo- week, solids on the tubes had turned correlated decrease in DBPs. The fol- rine residual of 0.8 mg/L in the distrib- dark brown and dropped off (Gould, lowing sites have observed an ution system. In addition, the facility 2000). improved residual at a reduced dosage, reports a 44% reduction in TTHMs • Orange County Water District in and the majority of them have also (Armijo, 2000). Los Angeles, Calif., has developed reported reductions in TTHM, HAA5, • North Table Mountain WSD in novel techniques for evaluating biofilm or both. Colorado previously dosed at 1.2 mg/L formation and removal. Tests using • Bloomfield, N.M., used chlorine with chlorine gas in order to maintain a mixed-oxidant solution and chlorine gas in the clearwell at a dose of 2 mg/L 0.2-mg/L residual in distribution. With on selected biofilms were funded by the to maintain a 1.2-mg/L residual at the mixed oxidants, the utility has reduced Defense Advanced Research Projects end of the distribution system. With the dose 33% to 0.8 mg/L, which main- Agency. Initial results showed that mixed oxidants, the facility has been tains a steady residual of 0.3–0.4 mg/L mixed oxidants required a higher able to reduce the dose ~30% to only throughout distribution. The site also dosage than that required for chlorine 1.4 mg/L and still maintain the desired reports a 44% reduction in TTHM and to remove the biofilms. On further 1.2-mg/L residual at all points in distri- HAA5 levels (Jeschke, 2000b). inspection, however, the mixed oxidant bution. In addition, the plant is imme- • After switching to mixed oxi- solution also removed the protein sub- diately operational in the morning with dants, Santa Fe’s Sangre de Cristo strate that biofilms develop in order to no loss of residual overnight, whereas Water Company was able to decrease attach to material surfaces; chlorine, when chlorine gas was used, staff had the chlorine dose by 31% (from 1.6 to on the other hand, did not remove the to run the plant for hours each morn- 1.1 mg/L) and still maintain the same substrate. The initial conclusion was ing in order to get the residual up to residual at all points in distribution. that it would be much more difficult the necessary level (Ruybalid, 1999). The facility also reports ~50% reduc- and require more time for biofilms to • Greenfield Municipal Utilities in tion in TTHM formation (Herrington reestablish themselves on mixed oxi- Iowa provides water to the neighboring et al, 1999). dant–treated membranes than on chlo- community of Orient, 9 mi (14.5 km) Figure 3 compares the dosages and rine-treated membranes (Phipps & away. When the utility used chlorine gas residuals at the six sites before and Rodriguez, 2001). Additional research as a disinfectant, Orient had to boost the after conversion to mixed oxidants. is under way. incoming water with additional chlorine Microflocculation. Microflocculation Improved residual at a reduced in order to maintain the residual is defined as enhanced flocculation dosage. The FAC residual from mixed throughout its distribution system. producing either a reduction in coagu- oxidants is much more durable than When Greenfield converted to mixed- lant demand for the same final (filtered the FAC residual from traditional oxidant generation, the FAC residual water) turbidity or a reduction in final chlorination. The mixed-oxidant FAC continued throughout Orient’s entire dis- turbidity at the same coagulant residual can endure for long distances tribution system, enabling the commu- demand. Processes have been devel- and stays in the lines for a much nity to discontinue boosting. In addition, oped first to determine whether longer period of time. This is likely within a few months of conversion, microflocculation will occur and then related to the removal of substances Greenfield reported a TTHM reduction to optimize the microflocculation such as biofilm, which creates oxidant of 20 to 40% (Herrington et al, 1997). effect. An improved clarification demand within the distribution sys- • When using chlorine gas, Lamar process will demonstrate reduced tur- tem. In addition to a more durable County Water Supply District in bidity, reduced DBP formation, and chlorine residual, treatment plants typ- Brookstone, Texas, was unable to reduced chemical addition for such ically notice a reduction in the maintain the desired 0.2-mg/L residual chemicals as alum and polymer. required dosage at the plant to main- at the end of its 25 mi (40 km) long Enhancing the coagulation process tain the same residual at the end of the distribution system. In contrast, mixed essentially removes several of the pre- line. After the distribution system has oxidants at the same dosage maintain a cursors for DBP formation. A variety stabilized, most mixed-oxidant users 1-mg/L residual at the end of the line of installations using mixed oxidants in report a 30% final reduction in dosage (American City & County, 1997). pretreatment report a microfloccula- at the clearwell. • Las Vegas, N.M., used chlorine tion effect and a reduction in TTHMs A secondary effect of the reduced gas at a dose of 2 mg/L to maintain a and HAA5 not seen with traditional dosage is a correlated reduction in DBP 0.3-mg/L residual in the distribution chlorination. The fact that these sites formation. Because chlorine require- system. Use of mixed oxidants has were unable to chlorinate in pretreat- fieldreport

ment because of excessive TTHM for- treatment at its 8 mgd (30 ML/d) sur- operational evidence from numerous mation further substantiates this the- face water treatment plant in 1998. field installations. ory. Although they are adding more Before the change in treatment, turbid- In implementation guidelines for the oxidant overall because of dual injec- ity levels were 0.6 ntu, which was Stage 1 D/DBPR, USEPA describes a tion points, they see an overall decrease above the 0.5-ntu limit. Plant flows best available technology for users of in DBP formation. Table 2 shows had to be maintained at only 4 mgd chlorine, chloramines, and chlorine microflocculation effects at five facili- (15 ML/d) during spring in order to dioxide as “control of treatment ties using mixed oxidants. control turbidity even to those levels. processes to reduce disinfectant • Crossville, Tenn., uses mixed oxi- Upon conversion to mixed-oxidant demand and control of disinfection dants in pretreatment and final disin- generation, staff noticed the alum treatment processes to reduce disinfec- fection. Finished water turbidity con- demand dropping. They initially tant levels” (USEPA, 2001). Mixed oxi- centrations were always quite low reported a 39% reduction in alum dants alone, without the use of any (~0.1 ntu) even with chlorine gas, and demand, as well as a turbidity decrease other chemicals or treatment tech- use of mixed oxidants did not further to 0.25 ntu in summer and 0.01 ntu in niques, have been shown to reduce reduce turbidity. However, since the winter. They are also able to maximize both disinfectant demand and overall treatment plant converted to mixed plant flows to 10 mgd (38 ML/d), even chemical usage. Although current ana- oxidants, alum consumption has during springtime, with no adverse lytical techniques can detect only chlo- dropped by 22%, and TTHM and effect on turbidity concentrations rine in the mixed-oxidant solution, HAA5 concentrations for comparative because of a more rapid rate of floc chlorine alone cannot account for the quarters have dropped 47–49% formation (Herrington et al, 1999). reduced dosages, higher residuals, and (Brownfield, 2001). Recent reports from the operating staff lower levels of DBPs that mixed-oxi- • Iowa’s Greenfield Municipal Utili- indicate a 60% reduction in alum dant installations report. ties pretreats with mixed oxidants for demand from original dosages and Because mixed-oxidant solution is removal of manganese. When condi- even lower turbidity ratings. produced from sodium salt tions are optimal, the facility can elimi- (NaCl) and water (H2O), it is clear that nate the use of potassium perman- CONCLUSION the other constituents in the solution ganate, use 40% less alum, and Out of a number of disinfection are chlor-oxygen compounds. In hun- decrease effluent turbidity by 39%. alternatives, mixed-oxidant use is the dreds of installations utilizing mixed- The site also reports TTHM reductions only technology that offers safe opera- oxidant technology, not one has ranging from 20 to 40% (Herrington tion, a chlorine residual as required by reported DBP formation higher than et al, 1997). USEPA, and the potential for DBP levels found with the use of chlorine. In • In Las Vegas, N.M., use of mixed reduction. As USEPA enforces imple- fact, tests for by-products associated oxidants in pretreatment has resulted mentation of new DBP limits, facilities with the possible compounds in the in a 29% reduction in alum demand will need to investigate the best tech- mixed-oxidant solution usually showed and a 57% reduction in finished water nology for their sites and put manufac- lower TTHM and HAA5 formation turbidity. TTHM levels have dropped turer claims to the test. than with chlorine or onsite hypochlo- by 44% (Armijo, 2000). Compared with other chlorine tech- rite generators. In addition, no chlorite • At the Midwest pilot-test site, nologies, mixed oxidants provide sig- or bromate has been found in mixed- mixed oxidants were used in pretreat- nificant advantages, including superior oxidant treated water, and chlorate ment on a sidestream flow of water inactivation capability and reduced for- production is far below levels of con- and in a full-scale test on one of the mation of DBPs, which indicate that cern and common to all hypochlorite four existing upflow clarifiers. The the mixed-oxidant solution is more solutions. engineer reported elimination of pin- than just hypochlorite generated on In summary, drinking water facili- floc from the clarifier effluent, notice- site. There are a number of explana- ties, as well as wastewater facilities able improvement in effluent quality tions for decreased TTHM and HAA5 that follow California’s water recycling relative to the other clarifiers, and a formation, including preferential action criteria requirements for secondary positive effect on algae growth in the of nonchlorine oxidants, elimination of reuse of their water (California Code clarifier. In addition, total organic car- biofilm, improved chlorine residual at of Regulations, 2001), can benefit from bon dropped by 12%, providing fur- a reduced dosage, and microfloccula- a treatment practice that provides ther evidence for removal of precursors tion to remove precursors. Only one of superior disinfection, stable chlorine (Gould, 2000). these explanations—preferential action residual, no safety concerns, and lower • The Sangre de Cristo Water Plant of nonchlorine oxidants—is unproven, DBP formation. The operational began using mixed oxidants for pre- whereas the others are supported by improvements of treatment with mixed oxidants are well-documented and 40503, (859) 223-8000, e-mail levels and is a member of the AWWA should be taken into consideration by [email protected]. She has BS National Disinfection Committee, the any site designing or redesigning its and MS degrees in chemistry from Coagulation and Filtration Commit- disinfection process. Cumberland College and Eastern Ken- tee, and the Operations Standards tucky University, respectively. Hamm Committee. —Beth Hamm is a water has 14 years of experience in the utility consultant with Tetra Tech Inc., drinking water field. She is active in FOOTNOTES 800 Corporate Dr., Lexington, KY AWWA at the national and section 1Infilco Accelator, Richmond, Va.

REFERENCES Mangold Environmental Testing Inc., 1997. Analysis of Sample From American City & County, 1997. Supplier Using Mixed Oxidants for Greenfield Municipal Water Utilities. Greenfield, Iowa. Better Quality. Phipps, D. & Rodriguez, G., 2001. Comparison of the Efficiency of Armijo, F., 2000. Personal communication. Bacterial Removal and Reduction of Bacterial Viability by Barton, L.L., 1996. Disinfection of Simulated Cooling Water. University Mixed-oxidant Solution (MOS) and Chlorine (as NaOCl) on a of New Mexico, Albuquerque. Reverse Osmosis Membrane. Water Resources & Technol. Dept., Orange County Water District, Los Angeles. Bradford, W. et al, 1997. Results of Tests Comparing the Disinfection Effectiveness of Mixed-oxidant Solution and Sodium Hypochlo- Russell, S.M., 2001. A Comparison of MIOX Versus Bleach for Eliminat- rite in Simulated Cooling Water. Los Alamos Technical Associ- ing Pathogenic Bacteria Associated With Poultry Products. Uni- ates, Los Alamos, N.M. versity of Georgia, Department of Poultry Science, Athens, Ga. Brownfield, J., 2001. Memorandum Regarding Crossville Data. S.O.S. Ruybalid, C., 1999. Personal communication. Inc., Cleveland, Tenn. Sletten, L., 2000. Personal communication. California Code of Regulations, 2001. California Health Laws Related Sobsey, M.D. et al, 2000. Inactivation of Cryptosporidium parvum to Recycled Water: “The Purple Book.” Excerpts From the Oocysts and Other Waterborne Microbes by Oxidants Generat- Health and Safety Code, Water Code, and Titles 22 and 17 of the ed Electrochemically From From Portable Pen California Code of Regulations, Berkeley, Calif. and Bench-scale Systems. University of North Carolina, Chapel Crayton, C. et al, 1997. Final Report on the Validation of Mixed Oxi- Hill, N.C. dants for the Disinfection and Removal of Biofilms From Distrib- Sterling, C., 1993. Testing of the MIOX Electrolytic Disinfecting Sys- ution Systems. Montana Water Resources Center, Montana tem for Inactivation of Cryptosporidium Oocysts. University of State University, Bozeman. Arizona, Tucson, Ariz. Duben, M., 1996. Pilot Study Final Report: Mixed-oxidant Disinfection USEPA (US Environmental Protection Agency), 2001. Implementation System at Greenfield, Iowa. Howard R. Green Co. Consulting Guidance for the Stage 1 Disinfectants/Disinfection Byproducts Engineers, Des Moines, Iowa. Rule. EPA 816-R-01-012, Ofce. of Water (4606), Washington. Environmental Health Laboratories, 1995. Analysis of Sample From a USEPA, 1998. Stage 1 Disinfectants and Disinfection Byproducts MIOX SAL-80 Cell. South Bend, Ind. Rule, EPA 815-F-98-010, Ofce. of Water (4607), Washington. Gordon, G., 1998. Electrochemical Mixed-oxidant Treatment: Chemi- Venczel, L. et al, 1997. Inactivation of Cryptosporidium parvum cal Detail of Electrolyzed Salt Brine Technology. Miami Univer- Oocysts and Spores by a Mixed-oxidant sity, Oxford, Ohio. Disinfectant and by Free Chlorine. Applied & Envir. Microbiol., Gould, B., 2000. Midwestern Water System Claricone and MIOX Pilot 63:4:1598. Study Results. Lawrence A. Lipe & Assoc., Benton, Ill. Waldroup, A.; Doyle, M.; & Scantling, M., 1996. Onsite Generation of Herrington, R. et al, 1999. Performance of a Conventional Surface Chlorine Solutions for Poultry Processing Applications. Univer- Water Plant Using Mixed Oxidants for Microflocculation and sity of Arkansas, Center of Excellence for Poultry Science, Final Disinfection. Proc. 1999 AWWA Ann. Conf., Chicago. Fayetteville, Ark. Herrington, R. et al, 1997. Pilot Study Report: Mixed-oxidant Disinfec- Wright, J.D.; Dvorak, T.J.; & Harper, B.G., 2001. Interim Report: MIOX Dis- tion System at Greenfield, Iowa. MIOX Corp., Albuquerque, infection Pen and the Mesosystems Mesostill Agents of Biological N.M. Origin (ABO) Disinfection Test. Life Sciences Div., W. Desert Test Jeschke, R., 2000a. Memorandum to the North Table Mountain Water Ctr., US Army Dugway Proving Ground, Dugway, Utah. and Sanitation District Board of Directors. Golden, Colo. Yu, C.H. & Murphy, J.K., 1998. Demonstration Testing of Onsite Elec- Jeschke, R., 2000b. Personal communication. trolytic Generation of Disinfectants. EPRI-CEC, St. Louis, Mo.

COPYRIGHT © 2002 AMERICAN WATER WORKS ASSOCIATION.