Survey of Swimming Pool/Spa Sanitizers and Sanitation Systems€¦ · and sanitation systems used...

John A. Wojtowicz – Chapter 7.1 121

Table 1 – U.S. Sanitizer Consumption in Swimming Pools, Spas, andHot Tubs (1992–5)

A) 1992, B) 1995, C) 1993

Sanitizer Thousands of Metric TonsCalcium Hypochlorite (65% av Cl)ChloroisocyanuratesSodium Hypochlorite (Cl2 Equivalent)Chlorine GasBromochlorodimethylhydantoinLithium Hypochlorite

50.048.8A

37.313.6B

7.7C

2.9

Basic information on the various sanitizersand sanitation systems used in swimming pools,spas, and hot tubs is discussed from the standpointof mode of action, disinfection, algae control, oxida-tion of contaminants, compatibility with ancillarychemicals, and cost. The main chemical sanitizersused in swimming pools and spas are calcium,sodium, and lithium hypochlorite, chlorine gas,chloroisocyanurates, and bromochlorodimethyl–hydantoin. Chlorine is the lowest cost and by far themost widely used sanitizer because it performs allthree sanitizer functions effectively, i.e., disinfec-tion, algae control, and oxidation. Use of bromine islimited primarily to indoor applications because itcannot be effectively stabilized. Systems employingozone, polyhexamethyl biguanide, metallic ions (cop-per, silver, or zinc), persulfate–type oxidizers, UV–hydrogen peroxide, and electrolyzers are used to asmall extent. In addition they do not offer an effec-tive alternative, significant improvement in perfor-mance and/or cost effective advantage to chlorine.

Sanitizer Consumption

The US sanitizer consumption is summarized inTable 1 (Wojtowicz 1993). Calcium hypochlorite,chloroisocyanurates, and sodium hypochlorite rep-resent the major portion of the sanitizer market,whereas chlorine gas, bromochlorodimethyl–hydan-toin, and lithium hypochlorite are a minor segment ofthe market.

Sanitizer/Oxidizer Cost

The cost of sanitizers/oxidizers are listed inTable 2. The data show that bromine is much moreexpensive than chlorine and that potassiummonopersulfate is very much more expensive thanchlorine as an oxidizer.

Chlorine

Chlorine Sources

Chlorine is marketed in various forms as shown

Originally appeared in theJournal of the Swimming Pool and Spa IndustryVolume 4, Number 1, pages 9–29Copyright © 2001 by JSPSIAll rights of reproduction in any form reserved.

Survey of Swimming Pool/SpaSanitizers and Sanitation Systems

John A. WojtowiczChemcon

122 The Chemistry and Treatment of Swimming Pool and Spa Water

Sanitizer/Oxidizer Weight (lb) Cost ($) $/lb Equiv. Av. ClCalcium Hypochlorite (65% av Cl) 25 45 2.77Sodium Dichloroisocyanurate 25 50 3.21Trichloroisocyanuric Acid 25 50 2.22Bromochlorodimethylhydantoin 25 110 7.59Potassium Monopersulfate 25 55 11.22

Table 2 – Sanitizer/Oxidizer Cost (at time of writing)

Table 3 – Chlorine Sources

Compound Form % Av. ChlorineChlorine Gas Liquefied gasCalcium Hypochlorite Granules, TabletsLithium Hypochlorite GranulesSodium Hypochlorite, Bleach SolutionSodium Dichloroisocyanurate, Dichlor* GranulesTrichloroisocyanuric Acid, Trichlor Tablets

*Marketed in hydrated and anhydrous forms.

10065 and 75

3510–15

56 and 6290

in Table 3. Sanitizer usage varies geographically, e.g.,in the Sun Belt, consumption varies in the order:chloroisocyanurates > calcium hypochlorite > sodiumhypochlorite, whereas in the snow belt, consumptiongenerally varies in the order: calcium hypochlorite >chloroisocyanurates > sodium hypochlorite. Granu-lar calcium or lithium hypochlorites can be added tothe pool by broadcasting the solid or by feeding asolution by metering pump. Chlorine gas is added tothe pool water via a vacuum injector or a porousdiffuser. Calcium hypochlorite tablets are dispensedthrough feeders or the skimmer. Sodium hypochlo-rite is added to the pool either manually or bymetering pump. Dichlor is broadcast and is usedprimarily in spas or hot tubs. Trichlor tablets aredispensed in feeders, skimmers, or floaters.

Hypochlorite ions can also be generated electro-chemically from chloride ions by dosing the pool withsalt (i.e., sodium chloride). Chlorine, formed at theanode by oxidation of chloride ions, reacts with waterforming hypochlorous and hydrochloric acids.

2Cl– → Cl2 + 2e–

Cl2 + H2O HOCl + HCl

Sodium atoms, formed by reduction of sodium ions atthe cathode, react with water forming sodium and

hydroxyl ions (i.e., sodium hydroxide), and hydrogengas.

2Na+ + 2e– → 2Na

2Na + 2H2O → 2Na+ + 2OH– + H2

In an unseparated cell, the products of the anode andcathode compartments react with each other; theoverall reaction is:

Cl– + H2O → ClO– + H2

Hypochlorite ions react with hydrogen ions to formequilibrium amounts of hypochlorous acid:

ClO– + H+ HOCl

Some chlorine generators add salt to the gen-erator rather than to the pool or spa. They can employa diaphragm that separates the anode and cathodecompartments. The chlorine formed in the anodecompartment is dissolved in the recycled pool water.The sodium hydroxide solution in the cathode com-partment can be added to the pool for pH adjustment.


HOCl H+ + ClO–

By contrast, hypochlorites such as calcium hy-pochlorite, ionize in water to calcium and hypochlo-rite ions. Hypochlorite ions partially react with hy-drogen ions to form equilibrium amounts of hypochlo-rous acid (ClO– + H+ HOCl).

Ca(OCl)2 + x H+ → Ca2+ + (2–x) ClO– + x HOCl

Literature Data – Chlorine is an effectivebroad–spectrum disinfectant (Block 1991). It is veryeffective against enteric bacteria such as E. coli (anindicator of fecal contamination) and viruses such asPolio 1 and Rotavirus, although it is less effectiveagainst protozoa such as Giardia (Hoff 1986); seeTable 4. The inactivation mechanism by hypochlorousacid varies with the organism. With bacteria respira-tory, transport, and nucleic acid activity are ad-versely affected (Hass and Engelbrecht 1980).

Effect of pH – The rate of disinfection bychlorine varies with pH because of the change in theratio HOCl/ClO–. However, it should be rememberedthat ClO– is a reservoir of HOCl. As HOCl is consumedit is replenished by the equilibrium reaction: ClO– +H+ HOCl.

Effect of Temperature – Although highertemperatures increase the ionization of hypochlor-ous acid to a small extent, the disinfection rateincreases. For example, in an unstabilized pool at pH7.5, raising the temperature from 77 to 86°F willcause an increase in HOCl ionization of ~4%, however,the disinfection rate increases by a factor of 1.4(Wojtowicz 1996a).

By contrast, the concentration of HOCl in sta-bilized pools actually increases significantly withincreased temperature due to increased hydrolysisof chloroisocyanurates. Indeed, the extent of hy-drolysis of monochloroisocyanurate ion (the majorchloroisocyanurate in pool water) increases by 200%over the above temperature range (Wojtowicz 1996a).

Effect of Stabilization – Although cyanuricacid stabilizes chlorine against decomposition bysunlight, it does so by reducing the concentrations ofHOCl and ClO– via formation of chloroisocyanurates(Wojtowicz 1996a). However, adequate disinfectionin pools is obtained by maintaining the NSPI recom-mended ideal free av. Cl (2–4 ppm) and maintainingcyanuric acid in the ideal range, i.e., 30–50 ppm.Stabilization does not increase much beyond 50–75ppm, but higher concentrations of cyanuric acid canreduce the disinfection rate (Table 5). To offset this

Photochemical Decomposition

Chlorine is decomposed by the UV rays (>290nm) in sunlight, e.g., about 90% decomposition occursin three hours. The decomposition is due to thephotoinstability of hypochlorite ion, which has anabsorption maximum at 290 nm and absorbs UV outto about 350 nm.

Stabilization of Chlorine

Chlorine can be stabilized against photochemi-cal decomposition by cyanuric acid (CA) via formationof chloroisocyanurates (monochloroisocyanurate ionis the main species at pool pH), which are relativelystable to UV light because they absorb well below 290nm. As little as 25 ppm CA reduces the extent ofdecomposition to about 30% after 3 hours (Nelson1967).

Disinfection

Active Agent – Disinfection by chlorine is dueprimarily to formation of hypochlorous acid, e.g., atpool pH chlorine gas reacts with water to produce amixture of hypochlorous acid and hypochlorite ion.

Cl2 + H2O → (1–x) HOCl + x ClO– + (1+x) H+ + Cl–

The proportions of hypochlorous acid and hypochlo-rite ion vary with pH due to the equilibrium shownbelow (at pH 7.54, it is 50:50):

Table 4 – Antimicrobial Activityof Chlorine

Ct values for 99% inactivation at 5°Cand pH 6–7 (Hoff 1986)

* Ct is the product of the free chlorineconcentration in ppm and the contact time inminutes. Higher temperatures will result inlower Ct values.

Microorganism Ct (ppm•min)*E. coli 0.034 – 0.05Polio 1 1.1 – 2.5Rotavirus 0.01 – 0.05


tendency, the av. Cl concentration can be maintainedat a higher level or the cyanuric acid concentrationcan be limited by frequent back–washing of the filter.

Effect of Contaminants – Nitrogen com-pounds introduced into pool and spa water reactwith chlorine forming combined chlorine com-pounds that are relatively biocidally ineffective(Wojtowicz, JSPSI 4(1)2001:30-40). These com-pounds can be oxidized by chlorine. NSPI recom-mends superchlorination in pools (i.e., at 10 timesCAC) when combined chlorine exceeds 0.2 ppm.

Algae Control

Chlorine at 2 ppm is effective in controllingmany strains of algae (Palmer and Maloney 1955).Green algae are free floating and are easy to control,whereas, the firmly attached black (i.e., blue–green)algae are the most difficult to control. Mustard(yellow) algae are loosely attached to pool surfacesand are easier to control than black algae. The key toalgae control is maintaining appropriate chlorine andcyanuric acid levels combined with periodic shocktreatment and brushing/vacuuming. Even a heavyinfestation of black algae can be eradicated by mul-tiple shock treatment (up to 30 ppm av. Cl) andbrushing and vacuuming.

Oxidation of Contaminants

Chlorine is an effective oxidant for most swim-ming pool and spa contaminants (Wojtowicz 1998a,2001a). Oxidation of nitrogen compounds (both inor-ganic and organic) produces small amounts of nitro-gen trichloride that may be irritating to the eyes ofsome bathers at sufficiently high concentrations(Wojtowicz 1998b). However, due to its volatility and

decomposition by free chlorine, heat, and sunlight, itis not a significant problem in outdoor pools. Periodicovernight superchlorination or shock treatment withchlorine or hypochlorite oxidizes ammonia and or-ganic nitrogen compounds preventing their build–upand minimizing nitrogen trichloride formation dur-ing normal chlorination of the water. Regular use ofshock products containing Trichlor and Dichlor arenot recommended since they greatly increase thecyanuric acid concentration.

Reaction With Ancillary Chemicals

Because chlorine is a strong oxidant, it reactswith organic–based ancillary chemicals added to pooland spa water such as quaternary ammonium com-pounds, copper chelates (copper citrate and coppertriethanolamine), scale and stain inhibitors(polyacrylates and hydroxyethylidene diphosphonicacid), clarifiers (polyelectrolytes), defoamers (sili-cones), and enzymes (Fitzgerald 1968). Consequently,these chemicals will have only a transient existencein the pool/spa water. Users should take into accountthe fact that ancillary chemicals have chlorine de-mands and will lower the free chlorine concentrationof the water.

BromineElemental bromine is a member of the halogen

family and thus its chemistry is similar to that ofelemental chlorine.

Bromine Sources

Figure 1 – BCDMH

Bromine Compounds – Commercially avail-able bromine compounds are listed in Table 6. Theprimary source of bromine for swimming pool and spasanitation is 3–bromo–1–chloro–5,5–dimethyl–hydan-toin (BCDMH, see Figure 1), which has a total av.halogen content equivalent to ~56% av. Cl. Commer-cial product is often referred to as 1–bromo–3–

Table 5 – Minimum av. Cl for30–sec 99% Kill Time*

*Estimated values for E. coli at 85°F(Wojtowicz 2001c).

Cyanuric Acid, Available Chlorine,ppm ppm 25 0.75 50 1.5100 3.0150 4.5200 6.0

N

NO Br

OH3C

H3C

Cl


BrCl + H2O → HOBr + HCl

Bromine is a liquid (boiling point 59°C, 138°F) and ismore difficult to meter than bromine chloride whichis a gas (bp –5°C, 23°F). In addition, bromine is moreexpensive and only half of the bromine is usable fordisinfection. Neither Br2 nor BrCl are employed insanitation of pools or spas.

Although bromine and bromochloro derivativesof Dichlor and Trichlor are known, they are notproduced commercially and therefore are not avail-able for swimming pool and spa sanitation.

In–Situ Generation of Av. Bromine – Hypo-bromous acid and hypobromite ion can also be gener-ated in situ by oxidation of bromide ion (typicallysupplied by calcium or sodium bromide) with chlorinecompounds such as chlorine gas, hypochlorites (Ca,Li, or Na), or chloroisocyanurates (Dichlor or Trichlor).For example, reaction of bromide ion with hypochlo-rite ion or hypochlorous acid, formed on addition ofa hypochlorite sanitizer to swimming pool or spawater, will form hypobromite ion and hypobromousacid.

HOCl/ClO– + Br– → HOBr/BrO– + Cl–

As with hypochlorite ion and hypochlorous acid,hypobromite ion and hypobromous acid are in equi-librium with each other at swimming pool pH.

HOBr H+ + BrO–

Other oxidizing agents such as potassiummonopersulfate or ozone can also be used.

Br– + H+ + KHSO5 → HOBr + KHSO4

chloro–5,5–dimethyl hydantoin. However, this is in-consistent with the fact that the more stable N–Clmoiety prefers the electronic environment providedby the adjacent electron donating methyl groups.BCDMH is marketed in tablet form for use in feeders.Since BCDMH has a lower equivalent av. Cl contentand is less soluble than Trichlor, it requires largerfeeders for equivalent feed rates. BCDMH is alsomarketed in granular form for use in spas. In water,the bromine substituent hydrolyzes forming an equi-librium concentration of hypobromous acid. All of thebromine substituent analyzes as free bromine. Bycontrast, the chlorine substituent is very tightlybound, hydrolyzing to only a very slight extent, andanalyzes as combined chlorine. In the presence ofbromide ion and excess dimethylhydantoin (H2DMH),the chlorine substituent can form the monobromoderivative. However, outdoor sunlight exposure testsunder simulated swimming pool conditions show thatreaction 1 does not go to completion even after 75%decomposition of total av. halogen.

BrClDMH + Br– + H2DMH 2BrHDMH + Cl–

BrHDMH H2DMH + HOBr

The hydantoin ring can be cleaved by hypochlo-rite ion (e.g., during shock treatment). The dichloroderivitive forms N–chloroisopropylamine, NCl3, andCO2 (Patterson and Grzeskowiak 1959).

Some instances of bather skin irritation havebeen reported in spas sanitized with BCDMH (Pennyand Rycroft 1983).

Both bromine and bromine chloride producehypobromous acid on reaction with water.

Br2 + H2O → HOBr + HBr

% TheoreticalCompound Form Av. Br Equiv. Av. ClBromochlorodimethylhydantoinA Tablets 66.2B 58.7Bromine Liquid 100 44.4Bromine Chloride Liquefied Gas 138.5 61.5

Table 6 – Bromine Compounds

A) A new product is now available that is dibromodimethylhydantoin, 49.6% equivalent av. ClB) Also contains 29.4% av. Cl

H2O


Stabilizer Conc. NaBr Av. Halogen % DecompositionC

ppm ppm as ppm av. ClB

Cyanuric Acid 50 0 3.4 35“ 50 100 4 100“ 150 0 3.6 19“ 150 100 4.6 96“ 300 100 4.6 74

BROMIshield 50 100 5.0 76DMH 50 100 4.6 74

A) Studies carried–out outdoors on a one–gallon scale in shallow glass containers in bright sunlight attemperatures of 70–85°F.

B) Initial concentration. Analysis performed using the FAS–DPD method.C) Exposure time 4 hours.

Table 7 – Comparative Stabilizer StudiesA (Wojtowicz 2000a)

Br– + H+ + O3 → HOBr + O2

Bromide ion can also be oxidized electrochemically toproduce available bromine. This was demonstrated ina swimming pool and a spa dosed with 2200 ppmsodium bromide (Olin Corp. 1983).

Br– + H2O → BrO– + H2

Hypobromite ions reacts with hydrogen ions to formequilibrium amounts of hypobromous acid: BrO– +H+ HOBr.

Disinfection

As with chlorine, the disinfecting properties ofbromine are due to hypobromous acid. Disinfectionby bromine is less sensitive to pH because HOBr isless ionized than HOCl over the swimming pool pHrange. Nevertheless, bromine is in general less effec-tive than chlorine (at normal pool pH) against bacte-rial spores (Marks and Strandskov 1950), bacteria(Zhang 1988, Gerba and Naranjo 1999), and viruses(Taylor and Johnson 1972) on a ppm basis. Althoughbromamines are better disinfectants than chloram-ines, they are readily decomposed by free bromine.Use of BCDMH results in build–up ofdimethylhydantoin that can reduce the concentra-tion of hypobromous acid, resulting in a decreaseddisinfection rate.

Algae Control

Swimming pool tests with electrochemicallygenerated bromine showed that it was effective incontrolling green, blue–green, and mustard algae.

Stabilization

Although bromine reacts with cyanuric acidforming bromoisocyanurates analogous to chloroiso-cyanurates, it cannot be stabilized as effectively aschlorine. Indeed, very high concentrations of cyanu-ric acid are required to obtain significant stabilizationas shown in Table 7. At 300 ppm of cyanuric acid, theextent of stabilization is similar to 50 ppmdimethylhydantoin (DMH, the parent compound ofBCDMH). Some research on bromine stabilizers hasrecently been published (Nalepa 1999).

A product called BROMIshield is currently onthe market that claims to reduce decomposition ofavailable bromine caused by intense sunlight (Dumas1999). This product appears to be dimethylhydantoinbased on a similar degree of stabilization (see Table7). Additionally, BROMIshield behaves likedimethylhydantoin in that it reacts with sodiumhypochlorite in the presence of bromide ion to formsignificant amounts of combined chlorine a portion ofwhich is still present even after 4 hours exposure tosunlight. However, if the sodium hypochlorite is firstreacted with sodium bromide to form hypobromite,then little or no combined chlorine is formed on


reaction with either BROMIshield or DMH. How-ever, the stability is somewhat lower under theseconditions.


With the exception of creatinine, the mainswimming pool contaminants are oxidized more rap-idly by bromine than by chlorine as shown in Table 8.The mechanism involves conversion of the chlorinecompounds to bromo derivatives that are less stable.Excess chlorine above the stoichiometric amount alsoincreases the oxidation rate as in the case of ammo-nia.


Like chlorine, bromine is a strong oxidant and itcan react with ancillary chemicals added to pool andspa water such as quaternary ammonium compounds,copper chelates, scale and stain inhibitors, defoam-ers, and enzymes. Consequently, these chemicals willhave only a transient existence in the pool/spa water.

Ozone

Properties

Ozone is an allotropic form of oxygen thatcontains three oxygen atoms. It is a pale blue gas atordinary temperatures, slightly soluble in water, andhas a pungent odor. Because of its instability (boththermal and explosive), it must be generated on site.

The half–life of ozone in tap water at 20°C is less than30 minutes. In addition to this thermal decomposi-tion, aqueous ozone is also decomposed by sunlight.Because ozone is a gas, it tends to escape fromaqueous solution. This tendency and the fact thatozone is toxic (the maximum allowable concentrationin air is only 0.1 ppm for an 8–hour exposure), is thereason that ozone cannot be used as a primarysanitizer.

Although ozone is a stronger oxidant than chlo-rine from a thermodynamic standpoint, it is notalways kinetically superior. For example, experimen-tal data show that chlorine is a much better oxidantfor bather contaminants such as ammonia, urea, andcreatinine that are oxidized only slowly by ozone.Thus, one of the advertised benefits of ozone inswimming pool/spa treatment, that it is a strongeroxidant than chlorine, is not supported by actualdata. Ozone technology has been comprehensivelyreviewed (Wojtowicz 1996b, 2001b).

Ozone Generation

Ozone can be generated from the oxygen in airby ultraviolet (UV) light or electric discharge (alsocalled corona discharge).

3O2 + UV light or electrical energy → 2O3

Corona discharge (CD) ozone generators (ozonators)produce much higher concentrations than UVozonators, i.e., 1–2 wt. % vs. <0.1 wt. %).

UV Ozone Generators – Low–pressure mer-cury lamps produce low concentrations because they

Compound Av. Cl/N Br– Ion Reaction Time % Oxidation Mol Ratio ppm Min. of Compound

Table 8 – Effect of Bromide Ion on Oxidation of Nitrogen Compoundsby Chlorine (Wojtowicz 1998a, 2001a)

Ammonia“

Urea“

Creatinine“

Glycine“

040040040040

1010606060606010

38 (87)79

7 (8)143.42.620

>26

1.8 (3.6)1.8

1.8 (3.6)1.81.81.81.81.8


also emit 254 nm radiation that decomposes ozone inaddition to the 185 nm radiation that’s responsible forformation of ozone. Lamps optimized for 185 nmradiation produce higher concentrations, typicallyabout 300 ppm. Some UV ozonators have air filtersand dryers but do not employ lamp cooling. UV lampshave very low energy efficiency compared to CDozonators. Over the past two decades numerous UVozonator manufacturers have entered the marketonly to go out of business a few years later.

CD Ozone Generators – By contrast to UVozonators, properly designed CD ozone generatorsproduce much higher concentrations, typically 1–2 %covering a wide range of production rates. Typical CDozonators employ dryers to lower the moisture con-tent of the inlet gas; a dew point of at least 60°F isrequired for optimum output. They may also utilizecooling to reduce the temperature of the CD cells.Some CD ozonators on the market are only margin-ally better than UV ozonators with ozone concentra-tions of only 0.06–0.2 wt. %.

Ozone Transfer into Water

The ozone produced by ozonators is transferredinto water by devices such as porous diffusers orventuris that disperse the gas into very small bubblesfor more intimate contact with water. Venturis gen-erate a vacuum that draws air through the ozonatorand injects the resultant ozone–air mixture into thewater circulation system. Compressors are used withporous diffusers and can also be employed to improvethe effectiveness of venturis. The transfer efficiencyincreases with the ratio of the water to gas flow rates.For a given transfer efficiency, the aqueous ozoneconcentration increases with the gas phase ozoneconcentration. The calculated ozone absorption forUV ozonators is generally in the 70–90% range. SinceUV ozonators do not decompose the unabsorbedozone in the off gas, this can pose a potential healthrisk to bathers in indoor spas.

Disinfection

Ozone at appropriate concentrations is an effec-tive broad–spectrum disinfectant, but its biocidalproperties can be affected by presence of readilyoxidizable matter in the water. Ozone cannot be usedas a primary sanitizer because it is both volatile andtoxic. If the ozone residual is sufficient for effectivecontrol of microorganisms, the concentration of ozoneabove the water will exceed the permissable exposure

limit (for an 8–hour exposure) of 0.1 ppm (OSHA1975). Thus, ozone requires a primary sanitizer andis typically used in conjunction with chlorine.

Algae Control

Ozone at appropriate concentrations is toxic tomany types of algae. However, swimming pools can-not benefit from this because ozone cannot be used asa primary sanitizer. UV ozone was shown to beineffective in algae control in two swimming pool tests(see Table 9).


Kinetic Data – The reactivity of ozone variesgreatly and depends on the functionality of thesubstrate. Reaction rate constants determined in thelaboratory show that ozone reacts very slowly withbather contaminants such as ammonia, urea (themain swimming pool contaminant), and creatinine(Hoigne, et al 1983–1985). Although ozone reactsrapidly with amines and amino acids, it reacts slowlywith many other organic compounds, e.g., aliphaticalcohols, aldehydes, and acids. Many organic nitro-gen compounds yield ammonia as an intermediateproduct. For example, in the oxidation of the aminoacid glycine, the ammonia formed is only slowlyoxidized at pool or spa pH because it is primarily inthe form of ammonium ion that does not react withozone.

Laboratory Data – Laboratory tests showedvery slow oxidation of bather contaminants such asammonia, urea, and creatinine even at relatively highozone and substrate concentrations (Eichelsdorferand Jandik 1985, Wojtowicz 1989a).


As in the case of chlorine and bromine, ozone willreact with ancillary chemicals added to the pool or spawater.

Evaluation of UV Ozonators

Disinfection – Two brands of commerciallyavailable UV ozonators were evaluated under swim-ming pool and spa conditions. The results, summa-rized in Table 9 indicate that typical UV ozonators bythemselves are not effective for either pool or spatreatment (Wojtowicz 1985). The bactericidal effec-


tiveness of ozone, evaluated under spa conditions,initially showed no inactivation of bacteria at spatemperature. In a second test starting at roomtemperature, a very slow kill rate was observed. Afterone hour the temperature increased from 77 to 88°Fand the inactivation was only 55%.

Oxidation – The oxidation of urea by ozone wasevaluated in a 300–gallon spa (100°F, pH 7.5, 0.3 gozone/hour). A total of 27 g of urea was added duringthe 36–hours of the test. Essentially no oxidation ofurea occurred based on the fact that there was littleor no change in the urea concentration and nobyproduct nitrate formation was observed. Similarresults were obtained in another study (Adams et al.1999).

Bromine Generation – A UV ozonator rated at7.5 g ozone/day was evaluated for generation ofavailable bromine from sodium bromide.

O3 + Br– → O2 + BrO–

Based on the amount of available bromine formed,the ozone generation rates at 25 and 35°C corre-sponded to 1.6 and 0.8 g/day or efficiencies of only 21and 8%, respectively.

Use of UV Ozone in Pools and Spas

Pools – UV ozone is applied directly into poolwater without use of a contact chamber as with CDozone, which is typically applied at a concentration of1 ppm and maintained for at least 2 minutes. Bycontrast, UV ozonators generate an ozone concen-tration of only 0.02 ppm in the external recycle loopand the contact time prior to entry into the pool is lessthan 1 second. The average ozone concentration inthe pool after one turnover (i.e., 6 hours) provided bythe ozonators in Table 1 is about 40-fold lower andamounts to only 0.5 ppb at a water temperature of

85°F, assuming only ozone decomposition. Assumingno reaction with bather contaminants, about 99% ofthe applied ozone will decompose over the course of6 hours. The total ozone dose, assuming no decompo-sition, is equivalent to only 0.03 ppm av. Cl in termsof potential oxidizing capacity. The low ozone concen-tration and dosage provided by UV ozonators pre-cludes significant contribution to disinfection or oxi-dation of bather contaminants.

Spas – Ozonating the water while the spa is inuse is not recommended because the unabsorbedozone in the ozonator vent gas can amount to 60 ppm.If the spa is treated after use with ozone at theaverage feed rate of 0.184 g/h (see Table 2) over a 6–hour period, the calculated average ozone concentra-tion, considering only decomposition, will be 1.6 ppb.The total ozone dose, assuming no decomposition, isequivalent to 0.4 ppm av. Cl in terms of potentialoxidizing capacity. This represents only 5% of therecommended shock dose of 8 ppm av. Cl. Further-more, most (~99%) of the applied ozone will simplydecompose resulting in negligible oxidation of bathercontaminants.

Claims – UV ozonator manufacturers typicallyclaim lower chlorine consumption (typically 60 to80%) and the ability to operate pools and spas at lowerchlorine concentrations (0.5–1.0 ppm). However, thereis a lack of published data on disinfection, algaecontrol and oxidation of bather contaminants ob-tained by independent researchers under actual pooland spa conditions in support of these claims.

As discussed above, UV ozone is too dilute tosignificantly contribute to disinfection and oxidationof bather contaminants, consequently a reduction inchlorine concentration and usage is not possible.Accordingly, NSPI recommended chlorine levels (1–3 ppm in pools and 3–5 ppm in spas) supplemented byperiodic shock treatment are necessary, (ANSI–NSPI 1995 and 1999).

Safety – As discussed earlier, ozone absorptionis incomplete, and since UV ozonators do not provide

Table 9 – Evaluation of UV Ozonators

Ozonator Test Ozone g/h ResultsA1 250–gal spa 0.25 Poor bactericidal performance.A2 6800–gal pool 0.5 Green algae bloom after 3 days of continuous operation*B1 250–gal spa 0.3 Poor oxidation of urea in synthetic bather insult.B2 6800–gal pool 1.0 Green algae bloom after 4 days of continuous operation*

*Water shock treated with calcium hypochlorite prior to test. pH 7.2–7.8, 80–85°F, 80 ppm alkalinity, 300ppm calcium hardness.


for off gas ozone destruction, the ozone concentrationabove the water at the point of entry of the ozonatedair into the pool or spa can be quite high (55 to 60 ppm).In indoor spas, this can cause the average ozoneconcentration above the spa to exceed the OSHAlimit of 0.1 ppm, creating a potential health hazard tobathers.

NSF Approval – UV ozonators with ratings upto 1 gram per hour ozone were tested by NSF andrequire the use of NSF certified brominators orchlorinators delivering 4 ppm bromine or 2 ppmchlorine (NSF 1985). Even larger output CD ozona-tors are subject to this requirement.

Cost – UV ozone generators, with productionrates of 0.25 to 0.44 g/h for pools of 18,000 to 50,000gals., retail for $500 to $700. These units come withventuri type injectors but do not have air filters,dryers, or off gas ozone destruction. The cost is afunction of the ozonator output and whether the unithas an air filter, dryer, or compressor.

Deficiencies of UV Ozonators – The followingsummarizes deficiencies of UV ozonators:

• Build up of toxic concentrations of ozone in indoorinstallations due to lack of off gas ozonedestruction

• Raises water pH by removing carbon dioxide• No separate contact vessels• Ozone inlet concentration too low and contact

time too short for significant disinfection oroxidation of bather contaminants

• Ozone output much too low to satisfy oxidizerdemand of pool or spa water in a practical time

• No method to measure the low ozone concentration• No way to tell if unit is functioning properly• Ozone output decreases with lamp age• No way to tell if lamps need replacement• Some units do not have air filters or dryers• No independent substantiation of effectiveness

in disinfection, oxidation, or reduced chlorineusage

Use of CD Ozone in Pools

European Practice – CD ozonators producemuch higher concentrations than UV ozonators andare commonly used in Europe, primarily in largecommercial or public pools. CD ozonators also re-quire additional equipment for a complete system

including: compressors, dryers, contact chambersand deozonators for treating vent gases and fortreating ozone–containing water before returning itto the pool. The most widely used ozonation techniquein Europe is the ozone–granular activated carbon(GAC) system that is covered by the German–devel-oped standard (DIN 1984). It involves treating allwater by coagulation, filtration, ozonation, GAC fil-tration, and chlorination. Ozone (about 1 ppm) isintroduced into the water in the external recycle loopafter a sand filter, through a porous diffuser in acontact chamber. After a reaction time of at least 2minutes, the ozonated water is filtered through GACto destroy unreacted ozone. This also destroys avail-able chlorine. The dechlorinated and deozonatedwater (with <0.05 ppm ozone) is then dosed with 0.5ppm av. Cl and returned to the pool. Althoughoxidation of contaminants is the primary purpose ofozonation, some destruction of microorganisms mayalso occur. In order to limit buildup of dissolvedsolids, a specified amount of water (~30 liters perbather) is purged from the pool.

Treating water by this process is cost effectiveonly for large, heavily used pools (e.g., public, com-mercial, or private). Data from European pools em-ploying the ozone–GAC process show that ozone canreduce operating costs by about 20%. Since ozone isnot effective in oxidizing bather impurities such asammonia, urea, and creatinine, removal of thesecontaminants will depend on processes that mayoccur in the GAC filter. For example, monochloram-ine can be partially converted to nitrogen and chlo-ride ion. Although the GAC filter may become biologi-cally active, potentially providing biodegradation ofsome contaminants, no data are available on theextent, if any.

North American Practice – Ozone–GAC sys-tems have been installed in several US cities (Rice1995). Modified systems are also being offered inorder to reduce costs. In retrofit installations, post–filter ozone injection is employed in conjunction witha combination contact chamber/GAC filter (Hartwig1996). For new installations, pre–filter ozonation isemployed which utilizes the filter as a combinationcontact chamber/GAC filter/sand filter. Still thissystem is cost effective only for large, heavily usedpools. Although DIN requires full flow ozonation,some systems employ only partial or slipstreamozonation (in some cases as low ~10%). Since ozoneonly increases the non–urea and ammonia CODreduction by about 20% and also requires a waterpurge and an effective GAC filter (i.e., biologicallyactive), any significant departure from DIN designwill be at the expense of water quality. CD ozonators


are not cost–effective for residential pools becausethe bather load is too low. This is probably also thecase for many intermediate sized public or privatepools.

Generation of Bromine – Ozone is sometimesused to generate av. Br from sodium bromide in smallEuropean public and semi–public pools and whirl-pools (i.e., spas). A spa test using a CD ozonator (7.5g ozone/hour) at room temperature showed only a50% bromine generating efficiency that would prob-ably be lower at typical spa temperatures.

Cost – The suggested retail price of residentialpool CD ozonators for 25,000 to 100,00–gal pools withozone production rates of 0.3 to 1.0 g/h ranges fromabout $600 to $1800. CD ozone generators employingair feed with ozone production rates of 1.2 to 7.4 g/hretail in the $800 to $3700 range and do not come withany peripheral equipment. Commercial CD ozonatorsemploying oxygen feed with ozone production ratesof 2 to 7 g/h retail in the $4,000 to $10,000 range anddo not come with off gas ozone destruction, contactchambers, or GAC filters. The cost of CD ozonatorswith higher production rates are as follows: 2–200 g/h (air) and 20–320 g/h (oxygen) – $10,000–$25,000;750–1800 g/h (oxygen) – $35,000–60,000.

Potassium Monopersulfate(PMPS)

Properties

Composition – Potassium monopersulfate(PMPS, KHSO5) is marketed in the form of a triplesalt (2KHSO5•KHSO4•K2SO4) of 85% purity. It con-tains ~4.5% active oxygen and is employed primarilyas a non–chlorine oxidizer.

Reaction with Halide Ions – Potassiummonopersulfate readily oxidizes bromide ion to bro-mine. It also oxidizes chloride ion to chlorine but ata much slower rate.

Stability – PMPS is decomposed by sunlightmuch faster than stabilized chlorine, e.g., an 8.6 ppmsolution of PMPS in tap water was 75% decomposedafter 6 hours (decomposition rate ~20%/hour) insunlight (Wojtowicz 2000b). Even in the absence ofsunlight PMPS decomposes at ~4%/hour at roomtemperature. The rate is even higher at spa tempera-ture: ~12%/hour.

Disinfection/Algae Control

While PMPS alone (or in the presence of car-

tridge derived copper and silver ions) is ineffective attypical swimming pool temperatures (~75–80°F), it ismore effective at spa temperatures. For example, at77°F PMPS provided only 16.8% inactivation of E.coli bacteria in 2 minutes (Gerba and Naranjo 1999),but at 104°F it provided >99.9999% inactivation. Nodata are available on disinfection of viruses andprotozoa, or on the algicidal properties of PMPS.


By contrast with chlorine, there is no publisheddata on the effectiveness of monopersulfate as anoxidant for typical pool and spa contaminants such asammonia, urea, amino acids, creatinine, etc. PMPSappears to be less reactive than chlorine since itbehaves like combined chlorine during DPD analysis.By contrast with chlorine, which oxidizes ammoniaand urea nitrogen primarily to elemental nitrogen,the oxidation product with PMPS is nitrate ion thatis a nutrient for algae. The recommended use ofPMPS at 1 lb/10,000 gal as an alternative shocktreatment for pools using the copper–silver or zinc–silver cartridges will be much less effective than achlorine shock, providing an oxidation capacity equiva-lent to only 2.3 ppm chlorine. Indeed, 3.3 lb of PMPSwould be required to equal the oxidizing capacity ofa 1 lb calcium hypochlorite shock.

Effect on pH

PMPS lowers pH due to formation of bisulfateions:

2KHSO5•KHSO4•K2SO4 → 3KHSO4 + K2SO4 + 2O

Uses

PMPS is used primarily as a non–chlorine shockin pools and spas and as a disinfectant in spas.

Cost

PMPS is much more expensive than chlorine,e.g., the cost of a 1–lb calcium hypochlorite shock is$1.80 while the cost of the equivalent amount of PMPS(3.3 lb) is $7.26.


Potassium Persulfate (PPS)

Properties

Composition – Potassium persulfate orperoxydisulfate (K2S2O8) is marketed as a whitepowder of ~95% purity.

Stability – The decomposition rate in stabilizedswimming pool water is about 5% per day. Decompo-sition proceeds as follows:

S2O82– + H2O → 2HSO4

– + 0.5O2

In the presence of oxidizable matter, the oxygen willbe consumed.


No data are available on the effect of PPS ondisinfection or algae control, however, by comparisonwith potassium monopersulfate it is not expected tobe significant.


Oxidation reactions of persulfate ion are nor-mally slow but can be catalyzed by sunlight and bymetal ions (eg, silver, copper, etc.) via formation ofsulfate ion radicals (SO4

– .) (Minisci et al 1983):

S2O82– + photon → 2SO4

– .

S2O82– + Ag+ → SO4

2– + SO4– . + Ag2+

The intermediate divalent silver (Ag2+) ions canoxidize organic matter faster than persulfate itself.Sulfate ion radicals are also more effective oxidantsthan persulfate. The Ag+ ions formed after oxidationof bather contaminants can be reoxidized to Ag2+ bysulfate ion radicals (or persulfate as above):

SO4– . + Ag+ → SO4

2– + Ag2+

No data are available on silver or copper catalyzedoxidations of bather contaminants by persulfate atswimming pool concentrations.

Effect on Water Chemistry

The end products of decomposition of persul-fate are hydrogen and sulfate ions. The hydrogen ionswill reduce pH and alkalinity.

Uses

Persulfate is used to a very small extent as anon–chlorine oxidizer in combination with coppersulfate.

Polyhexamethylene Biguanide(PHMB)

System Description

Polyhexamethylene biguanide (PHMB) is a poly-meric nitrogen compound [–(CH2)6 [NHC(=NH)]2NH–]N that is marketed as a 20% solution. It functions asa bacteriostat and is part of a three–componentsystem that includes a quat, i.e., a quaternary nitro-gen compound (dimethylalkylbenzylammonium chlo-ride) and hydrogen peroxide, which function as algistatand oxidizing agent, respectively. The system alsoincludes an enzyme–based filter cleaner and a metalchelator. After an initial dose of 10 ppm active, theconcentration of PHMB is maintained between 6 and10 ppm while the quat is maintained at 2–2.5 ppm viaweekly measurement/adjustment. Peroxide is addedevery 3–4 weeks at a dosage of ~27 ppm. This multi–component system is more expensive than chlorine.

Incompatibilities/Problems

Product literature indicates that PHMB is notcompatible with chlorine or bromine sanitizers, cop-per and silver–based algicides, ozone, persulfateoxidizers, most clarifiers and cleaners, and somestain and scale inhibitors. Excessive use of PHMB,quat, and enzyme can cause foaming and impart anodor and off taste to the water. Other problemsinclude persistent haziness or cloudy water anddevelopment of biological growths. These problemsmay necessitate partial drainage of the pool waterand replacement with fresh water and/or temporarilyincreasing the pH combined with vacuuming to waste.Since quats tend to be removed by filter media suchas diatomaceous earth (Fitzgerald 1960), increaseddosing frequency may be necessary.


Test Duration No. of Incubation Bacteria Countsdays Samples Period > 200 CFU/mL

Control PHMB 90 26 one week 1 7100 35 2 days 0 20

Table 10 – Bactericidal Comparison of Chlorine and PHMB


PHMB is a bacteriostat; the minimum inhibitoryconcentration (MIC) for E. coli is 20 ppm (4 ppmactive) and for Pseudomonas aeruginosa it is 100 ppm(20 ppm active) (Block 1994). Since the recommendedPHMB concentration is 30–50 ppm (6–10 ppm active),Pseudomonas aeruginosa may not be controlled.However, the combined effect of PHMB and the quatmay provide control. Because hydrogen peroxide is apoor oxidant, build–up of organic matter can promotebiological growths such as water mold and pink slime.

Swimming Pool Evaluation – The PHMBsystem was evaluated in a 6800–gal swimming pool(Sandel 1996). The data are summarized in Table 10.In the first year of the test, 27% of the pool samplesshowed bacterial counts above 200 CFU/mL. In thesecond year, 57% of the pool samples showed bacte-rial counts above 200 CFU/mL. By contrast with thefirst year results, which were obtained using a 7–dayincubation period, the second year results were basedon only a 2–day incubation period. This is indicativeof development of PHMB resistant bacteria. Bycontrast with the PHMB results, a calcium hypochlo-rite treated control pool showed negligible test re-sults above 200 CFU/mL. In addition, the PHMB testpool showed evidence of bacterial slimes in the poolskimmer. Furthermore, total organic carbon (TOC)increased with time and is associated with the pooroxidizing properties of hydrogen peroxide. The pres-ence of organic matter may in fact be responsible forthe development of PHMB resistant bacteria.


Hydrogen peroxide is a very poor oxidant forammonia and urea (the main pool contaminant), andother organic matter. Thus, build–up of organicmatter will occur and may cause cloudiness, develop-ment of biological growths, and inadequate disinfec-tion.

Copper, Silver, and Zinc Devices

System Descriptions – Sanitizer systems basedon metal ions employ copper, silver, or zinc. The ionscan be generated electrochemically (i.e., copper–silver or silver ionizers), or by dissolution (copper–silver and zinc–silver cartridges). Ionizers delivermuch higher concentrations than cartridges (Table11). The metal ion concentrations delivered by zinc–silver cartridges have not been disclosed. Thesedevices are installed in the external recycle loop ofthe pool or spa. Silver is not compatible with bromineor biguanide–based sanitizing systems.

Disinfection

Literature Data – The antimicrobial activity ofcopper and silver is thought to be due to binding tosulfhydryl groups in cellular proteins and enzymespreventing their participation in enzymatic reactions(Kutz et al 1988). On a ppm basis, the activity of metalions varies in the following order: silver > copper >zinc. Compared to chlorine, silver is a poor bacteri-cide, e.g., 86 minutes were required for 99.9% inacti-vation of E. coli in the presence of 0.03 ppm silver at25°C and pH 7.5 (Wuhrman and Zobrist 1958). Thepresence of chloride or phosphate ions significantlyincreased kill time. For example, 10 and 100 ppmchloride ion increased the 99.9% kill time of 60 ppbsilver by 25 and 70%, respectively. The 99.9% kill timewas also increased by 3 minutes for each 10 ppm ofhardness.

Ceramic cartridges containing finely dividedmetallic silver embedded in a ceramic substrate havebeen known since the 1930’s. One problem encoun-tered in their use was a gradual decrease in bacteri-cidal effectiveness with time due to build–up of anorganic slime that coated the silver particles neces-sitated periodic cleaning (White 1972). Evaluation inswimming pools showed that silver was unsatisfac-


tory as a bactericidal agent (Shapiro and Hale 1937).The data in Table 12 show that copper and silver,

either individually or together, are poor disinfec-tants (Kutz et al 1988). With chlorine alone, a high killrate was observed. With chlorine, copper, and silverpresent, disinfection improved to a small extent.Indeed, the data show that chlorine killed 99.9% ofthe bacteria and the copper and silver killed only0.09%. However, the concentration of copper wasabove 0.3 ppm and would most probably cause stain-ing. The data show that chlorine is necessary foradequate disinfection. It is important to mention thatthe tests were carried–out in well water with only 0.02ppm of chloride ion. Since typical swimming poolwater contains significant concentrations of chlorideion, disinfection rates will be lower than those shownin Table 12.

The disinfection data in Table 12 and providedby some manufacturers was obtained in the absenceof chloride ion and cyanuric acid. Chloride ion de-creases the bactericidal effectiveness of silver asdiscussed above. In addition, cyanuric acid is knownto reduce the effectiveness of chlorine. Thus, the lowrecommended av. Cl levels shown in Table 11 mightnot provide satisfactory control of bacteria in stabi-lized pools.

Testing of Ionizers – In a one–month test of a

copper–silver ionizer in a 16,000–gal outdoor aboveground residential pool, extremely high bacteriacounts (14,000 to 62,000 CFU/mL) were observed onthree successive periodic samples. The pool wasshocked with PMPS every two weeks. In a test of a spaionizer, high bacteria counts (>3,000/mL) of fecalcoliforms and streptococci were present in the waterwhile bathers were in the spa despite satisfactorycopper and silver levels (Sandel 1996).

Testing of Cartridges – Copper silver car-tridges provide even lower metal ion concentrationsthan ionizers, consequently the small enhancementin % kill shown in Table 12 would be further reduced.A cartridge designed for pools up to 16,000 gals. wasevaluated in a 6800–gal experimental abovegroundswimming pool (Sandel 1992). The recommendedchlorine level was 0.2 ppm. The bacteriological effec-tiveness of water from this pool was compared witha control pool treated only with chlorine using amodified AOAC protocol (AOAC 1981). The data,summarized in Table 13, show that water from thepool with the attached cartridge did not kill bacteriaat rates sufficient to pass the disinfection standardof the AOAC even with the presence of 0.24 ppmchlorine. In addition, the effect of the copper andsilver on the kill rate in the presence of chlorine wasvery small. Another study also reports poor bacteri-

Table 11 – Copper, Silver, Zinc Devices

Application

pool/spapoolA

SpaPoolD

PoolSpaG

Copperppm0.3

0.02–0.06“–––

Silverppm0.03

0.01–0.06“F

F

F

Zincppm

–––F

F

F

Chlorineppm~0.2

0.4–0.6C

0.5–1.0–

0.5–1.0

PMPSppm

––B

–E

–

Device

Cu–Ag IonizerCu–Ag Cartridge

“Zn–Ag Cartridge

““

A) Maintain at least 50 ppm cyanuric acid. Shock with 1 lb calcium hypochlorite or 1 lb of PMPS per 10,000gallons when pool becomes hazy.

B) For chlorine treated 250-gal spa: before each use add 1 tsp Dichlor (~4 ppm av. Cl) and once a week shockwith 1 tbs Dichlor.

C) For PMPS treated 250-gal spa: before each use add 1 tbs 85% PMPS (equiv. to ~4 ppm av. Cl) and oncea week shock with 3 tbs PMPS.

D) Shock once a week with 1 lb calcium hypochlorite or 1 lb PMPS/10,000 gal.E) Shock three times a week with 1 lb PMPS/10,000 gal.F) No data available.G) Shock once a week with Dichlor or PMPS according to manufacturer’s recommendations.


A) Provided by cartridge: copper ~25 ppb, silver ~30 ppb.B) Pool chlorinated with Trichlor tablets.C) Laboratory solution without cyanuric acid.

SolutionCu–Ag from poolA

Cu–Ag from poolA,B

Trichlor control pool“

Phosphate BufferC

Av. Clppm

00.240.232.080.95

0.5 min.100,00073,00070,000

850<1

E. coli CFU/mL1.0 min.94,00049,00052,000

<1<1

10 min.77,000

<1230<1<1

Table 13 – Comparative Bactericidal Performance ofCopper, Silver, and Chlorine

CopperB, ppm0.39

00.48

00.47

A) For E. coli bacteria (Kutz, Landeen, Yahya, and Gerba 1988).B) Provided by ionizer.

Table 12 – Comparative Bactericidal Performance ofCopper, Silver, and ChlorineA

SilverB, ppm0

0.060.04

00.04

Av. Cl, ppm000

0.200.20

One–min % Kill127

99.999.99

cidal performance for copper and silver ions providedby a cartridge. At 77°F, the copper and silver ionssupplied by the cartridge itself provided an averageof only 53% kill after 45 minutes against three testorganisms (S. faecalis, E. hirae, and P. aerogenosa)and had virtually no effect on the kill rate when PMPSwas present (Gerba and Naranjo 1999).

A zinc–silver cartridge attached to a 500–galtank circulated at 40 gpm required 30 minutes for99% kill (equivalent to 2.4 turnovers; Legend Labs).It’s main competitor (copper–silver) provided only13% kill under the same test conditions.

Limitations of Cartridges – Manufacturers ofcartridges claim that bacteria can be removed fromthe water and inactivated on the surface of thecartridge packing. However, the rate of disinfectionby this process will be very low because a typical poolrequires about 6 hours for one turnover of the waterand only a portion of the water (average ~30% for the

copper–silver cartridge) passes through the car-tridge. Furthermore, when the swimming pool pumpis off (typically about 18 hours) no filtration ofbacteria occurs. Even in spas, it typically takes about30 minutes for one turnover of the water. Effectivedisinfection in pools occurs on a minute time scale.The amount of water passed through the pool car-tridge in one minute is less than 0.1%. Thus, it is thechlorine in the water that will be doing virtually all ofthe disinfection, e.g., in one minute chlorine treatsthe entire contents of a 25,00–gal pool while thecartridge treats less than 69 gallons. Removal ofbacteria by the cartridge results in buildup of bacte-rial residues/organic slimes that reduces this disin-fection process and also decreases the release ofsilver into the water.

Potassium monopersulfate (PMPS), which isused as a non–chlorine shock with cartridges, is apoor disinfectant at pool temperatures but is moreeffective at spa temperatures.


Copper Concentration, ppmB

Algae % Control AlgistaticC AlgicidalD

Chlorella py. (green) 0 0.12 – 0.15100 0.21 – 0.44 >0.6

Phormidium in. (blue–green) 0 0.14 – 0.21100 0.59 >0.6

Pleurochloris py. (mustard) 0100 0.07 – 0.14E >0.6F

A) Tests in Allen’s medium employing 300,000 cells/mL.B) As copper triethanolamine.C) Contact time (days): C) 14, D) 9–10, E) 7, F) 1.

Table 14 – Control of Algae by Copper A

Algae Control

Literature Data – Silver, at 0.064 ppm, wasshown to be effective against blue green algae(Phormidium minn. and Plectonema sp.) but noteffective against green (Oocystis) and yellow–greenalgae (Pleurochloris sp.) (Adamson and Sommerfeld1980).

It is thought that copper inhibits the growth ofalgae by reacting with protein sulfhydryl groupsconsequently affecting cell membrane permeability(Kuwabara and Leland 1986). Studies using Allen’smedium have shown that low levels of copper (<0.6ppm) are only algistatic toward common swimmingpool algae (Fitzgerald and Jackson 1979) as shown inTable 14. The data show that copper is most effectiveagainst mustard (yellow) algae and least effectiveagainst blue–green (black) algae. Other data areconsistent with these results (Adamson andSommerfeld 1980). Algicidal concentrations are toohigh to be employed in pools because of the increasedpotential for staining.

Studies with green algae (Chlorella vulgaris)showed that using simulated swimming pool waterinstead of Allen’s medium resulted in a lower MIC(minimum inhibitory concentration), i.e., 0.042 ppmvs. 0.45 ppm (Grenier and Denkewicz 1997).

Zinc is less effective than copper as an algistatby more than an order of magnitude. No data areavailable on the algistatic/algicidal effectiveness ofPMPS, which is used as a non–chlorine shock.

Ionizer Testing – An ionizer with a 97:3 cop-per–silver electrode was evaluated in a heated 6800–gal outdoor aboveground test pool: temperature 80–85°F, pH 7.2–7.8, alkalinity 80 ppm and calcium

hardness 300 ppm (Wojtowicz 1988). The pool wasshock treated with calcium hypochlorite and the av.Cl allowed to dissipate prior to starting the test. Theinitial copper concentration was 0.3 ppm. This droppedto 0.2 – 0.25 ppm after three weeks. The silverconcentration was very low, typically < 2 ppb. Duringthe 52 days of the test, yellow–green algae developedon four different occasions, necessitating shock treat-ment with calcium hypochlorite. A control pool,stabilized with cyanuric acid and sanitized with cal-cium hypochlorite (free chlorine 1–3 ppm) over asimilar time frame did not develop algae growthdespite being treated three times per week withgreen, blue–green, and mustard algae.

Cartridges – The low concentrations of copperand silver provided by cartridges in conjunction withrelatively low chlorine levels would probably result inquestionable control of algae in stabilized pools.Indeed, regular use of a quat–type algicide is recom-mended by the manufacturer of copper–silver car-tridges in cases of persistent algae infestation.


Effective oxidation of contaminants is neces-sary for proper disinfection and algae control. Due tothe low recommended average chlorine levels usedwith copper, silver, and zinc devices, bather contami-nants will rise to higher levels, e.g., the urea concen-tration will be more than 10, 4, and 2.7 times higherin pools using ionizers, copper–silver cartridges, andsilver–zinc cartridges, respectively. This will en-hance the growth of bacteria and algae. Chlorine is agood oxidant for swimming pool contaminants whenused at appropriate maintenance concentrations (1–


3 ppm FAC) supplemented with weekly or biweeklyshock treatment (~8 ppm av. Cl).

The recommended use of PMPS at 1 lb/10,000gal as an alternative shock treatment for pools usingcartridges will be much less effective than a chlorineshock, providing an oxidation capacity equivalent toonly 2.3 ppm chlorine. Furthermore, PMPS is decom-posed by sunlight much faster than chlorine.

Claims

Chlorine Levels - It will be very difficult tomaintain the low chlorine levels (0.2 for ionizers, 0.4-0.6 for copper-silver cartridges, and 0.5-1.0 for zinc-silver cartridges) in water subjected to the twindemands of decomposition by sunlight and bathercontaminant oxidation. It will also be difficult tomeasure these low recommended chlorine levels bytest kit because they are at or near the bottom of thescale. Furthermore, these low chlorine levels will notprovide adequate disinfection in stabilized pools.

Chlorine Usage - It is claimed that ionizersand cartridges can reduce chlorine usage by up to90%. However, no actual pool or spa test data areprovided in support of these claims. What littlebactericidal data that is presented was obtained inthe absence of cyanuric acid, which is known todecrease disinfection rates. In view, of the minimaleffects of ionizers and cartridges (and the ions theyprovide) on disinfection, a significant reduction inchlorine usage does not appear to be feasible.

Staining/Cloudy Water

Staining will occur over an extended period oftime since all of the added copper and silver eventu-ally precipitate from solution and are deposited onpool surfaces. Indeed, about 30% of the silver and 10%of the copper added to the water by an ionizer wasfound to be lost each day. In short term tests, bluestaining was observed at ≥ 0.3 ppm copper. Coppercan also cause gray or black staining or discolorationwhile silver can cause brown or black staining. Mostmanufacturers recommend a maximum copper con-centration of 0.2 to 0.3 ppm to avoid staining. Someionizer manufacturers include a stain remover and astain inhibitor with their start up kits. Zinc can causecloudy water by precipitation of basic zinc carbonateat concentrations of a few ppm.

Cost

The cost of copper, silver, and zinc devices issummarized in Table 15. Ionizers are very expensive.Even the cartridge–based units are relatively expen-sive.

Non–Chlorine Silver–CopperFormulations

Two non–chlorine granular formulations fortreatment of pools and spas are available. Informa-tion on these is summarized in Table 16.

Product 1 – This product for pools consistsof a mixture of copper sulfate and potassium persul-fate (PPS). It does not contain a disinfectant. It israther expensive at $66 for 10 lbs. Its claims are:controls algae, oxidizes, adjusts pH, clarifies water,and maintains alkalinity and hardness. However, thisproduct does not contain any ingredients that willcontrol alkalinity or hardness. Indeed, the productcontains sodium bisulfate, an acidic compound thatlowers pH and alkalinity. Additional bisulfate isformed by decomposition of PPS.

The recommended concentrations are: cop-per 0.2 – 0.8 ppm (preferably 0.2–0.4 ppm) andpotassium persulfate 1.5 ppm minimum. Copperconcentrations of ≥ 0.3 ppm can cause staining ofplaster surfaces. The minimum concentration of 1.5ppm potassium persulfate is equivalent to 1.0 ppm av.Cl in terms of oxidizing capacity. Without taking intoaccount its oxidizing effectiveness, this concentra-tion will be insufficient to effectively oxidize bathercontaminants in the absence of an effective shocktreatment. Since this system lacks an effective dis-infectant, control of bacteria and other microbes willbe compromised.

Product 2 – This system consists of twoseparate products, a disinfectant (silver oxide) andan oxidizer (i.e., PMPS). Dosing with PMPS beforeeach use, and once a month treatment with silveroxide is recommended.

Ultraviolet Light – Hydrogen PeroxideEquipment Description

This system utilizes a cell that contains anultraviolet (UV) lamp. Swimming pool or spa water iscirculated through the cell. Only that portion of thewater within the cell is subjected to the combinedeffects of UV light and hydrogen peroxide. The UVcells are relatively expensive.


Disinfection

Hydrogen peroxide by itself is a very poordisinfectant (Block 1991). For example, it required500 ppm at 37°C (99°F) to inactivate E. coli in 10–30minutes and 15,000 ppm at 20°C (68°F) to inactivatePoliovirus type 1 in 75 minutes. Ultraviolet light caninactivate bacteria and viruses. For example, thereported 99.9% inactivation time for E. coli is 60seconds whereas S. facecalis required 165 seconds(White 1972). Although ultraviolet light can inacti-vate microorganisms, many bacteria can repair dam-age to their DNA. A major problem with this systemis that it does not provide a residual sanitizer concen-tration in the main pool or spa water. Also turbiditycan reduce the effectiveness of UV light.

Algae Control

No data are available on the effectiveness ofhydrogen peroxide as an algicide.


Hydrogen peroxide itself is a poor oxidant forswimming pool contaminants. However, ultravioletlight decomposes hydrogen peroxide into hydroxylradicals.

H2O2 + hν → 2OH

Due to their extreme reactivity, hydroxyl radicalshave a very short lifetime of about 10–6 seconds(Dorfman and Adams 1972). They react non–specifi-cally with swimming pool contaminants. The mecha-nism of the oxidation involves hydrogen abstractionby OH followed by rapid reaction with oxygen. Theefficiency of oxidizing swimming pool contaminantsby OH is reduced by reaction with radical traps suchas bicarbonate and carbonate ions.

Table 17 shows data on oxidation of swimmingpool contaminants by UV– H2O2 (Wojtowicz 1989).Whereas ammonium ion and urea showed little or no

Table 16 – Non–Chlorine Formulations

Product Application Disinfectant Algicide Oxidizer1A Pool/Spa 1.6% Copper Sulfate Potassium Persulfate2 Spa 1% Silver Oxide 1% Silver Oxide PMPS

A) Contains: oxidizer, algicide, clarifier, and 28% sodium bisulfate.

Table 15 – Cost of Copper, Silver, and Zinc Devices

A) Replacement electrodes: $100–150.B) Cartridge cost $90; requires replacement every 6 months.C) Fits inside cartridge filter; requires replacement every 4 months.D) 6–month cartridges: $89 and $151, respectively.E) Requires replacement every four months.

DeviceCu–Ag Ionizer

“Cu–Ag Cartridge

“Zn–Ag Cartridge

““

Application10,000–25,000 gal Pool

200–1,000 gal Spa5,000–25,000 gal Pool

250–1000 Spa20,000–gal Pool40,000–gal Pool

250–1000–gal Spa

Cost$1000–1500A

$400–900A

Flow Controller $130B

Cartridge $30C

Flow Controller $100D

Flow Controller $329D

CartridgeE


reaction, amino acids underwent significant oxida-tion. On average the ratio of the ammonia yield toTOC reduction was 0.97 indicating that the oxidationyields ammonia, carbon dioxide, and water.

H2NCH2COOH + [OH/O2] → NH3 + 2CO2 + H2O

Although creatinine underwent significant TOC re-duction, very little ammonia and no nitrate formationoccurred. The extent of TOC reduction will be lowerat the lower concentrations of these compoundsfound in pools.

A UV-hydrogen peroxide system (15 gal/min.)was evaluated over a 3-week period in a 250 gal spaat 100°F using a 4-6 hour duty cycle and a syntheticbather insult. Analysis showed no oxidation of ureaafter 107 hours of operation.

Electrolyzers

Equipment Description

Electrolyzers are electrolytic cells containingtwo electrodes, one negative (cathode) and one posi-tive (anode). A direct current from a power supply ispassed through the cell, from one electrode through

the water to the other electrode. Electrolyzers arerelatively expensive.

Electrolysis of Water

Electrolyzers essentially carry out electroly-sis of water, i.e., they dissociate water into its com-ponents: hydrogen and oxygen. Water ionizes to aslight extent to hydrogen and hydroxyl ions.

4H2O → 4H+ + 4OH–

The hydrogen ions are attracted to the negativeelectrode where they are reduced to neutral hydro-gen ions that rapidly recombine to form a molecule ofhydrogen gas.

4H+ + 4e– → 4H → 2H2

The hydroxyl ions are attracted to the positiveelectrode where they are oxidized to hydroxyl radi-cals that rapidly react to form water and oxygenatoms which themselves recombine to molecularoxygen.

4OH– → 4OH + 4e– → 2H2O + 2O → 2H2O + O2

The overall reaction is:

2H2O → 2H2 + O2

To avoid an explosion between the hydrogen andoxygen, air may have to be purged through the cell.

Swimming Pool Evaluation

This system was evaluated on a 20,000–gal poolin Las Vegas, NV from September 1994 to June 1995(Hafer 1995). During the first phase of the test(September 14 to December 6), the pool was treatedperiodically with a combination of sodium hypochlo-rite, Dichlor, and an algicide. The av. Cl ranged from0.09 to 0.25 ppm. Data for only 7 out of 83 days waspresented. Despite the use of chlorine and algicide,black algae were observed. In addition, high bacteriacounts > 200 CFU/mL were observed on numerousoccasions: heterotrophic 4 out of 7 and coliforms 3 out6. The test has been critiqued (Wojtowicz 1998b).

Table 17 – Oxidation ofSwimming Pool Contaminants

by UV/H2O2A

A) ~23°C, time 4 hours, pH 7.4, alkalinity 80 ppm,calcium hardness 250 ppm.

B) Containing 2.26 ppm nitrogen.C) Calculated.

CompoundB % Yield % Yield % TOC of NH3 of NO3

– ReductionAmmonia – 0 –Urea 2 0 2C

Creatinine 1.5 1 57Glycine 65 0 70α–Alanine 59 0 46Valine 63 2 59Lysine 35 0 47Glutamic Acid 58 0 69



Since electrolyzers do not provide a sanitizerresidual, they will be ineffective in disinfecting thepool and in control of algae.


Oxidation of water contaminants will not be verysignificant, because in contrast to the UV–H2O2system, which generates hydroxyl radicals in the bulkof the water flowing through the UV cell, electrolyzersgenerate hydroxyl radicals in a very thin film at thesurface of the positive electrode. The transient OHradicals and O atoms that are formed react extremelyrapidly in this film before they have a chance todiffuse into the bulk water. Thus, most of the waterpassing through the electrolytic cell is unaffected bythese reactive species.

References

Adams, V. Dean, Ph.D., Mike Silveri, and RobertMoore. “An Evaluation of an Electrolytic Processfor the Removal of Ammonia and Urea fromSimulated Spa Water.” NSPI Symposium SeriesVolume IV, 1999, pp. 44–52.

AOAC, 190, Official Methods of Analysis of theAmerican Association of Official AnalyticalChemists. 15th Edition, Arlington, VA.

Block, S.S. Disinfection, Sterilization, andPreservation. Fourth edition, Philadeliphia, PA:Lea & Febiger, 1991.

DIN 19643 (English): Treatment and Disinfection ofSwimming Pool and Bathing Water (Berlin: BeuthVerlag, 1984).

Dorfman, L.M. and G.E. Adams. “Reactivity of theHydroxyl Radical in Aqueous Solution.” NationalBureau of Standards, NSRDS–NBS 46(Washington, DC, US Government Printing Office,1972).

Dumas, Bob. “Bromine Stabilizer Passes Early Test.”Pool & Spa News, January 13, 1999, p 1– 2.

Eichelsdorfer, D. and J. Jandik. “Long Contact TimeOzonation for Swimming Pool Water Treatment.”Ozone Science and Engineering 7(2)(1985):93–106.

Fitzgerald, G. P. “Loss of Algicidal Chemicals fromSwimming Pools.” Applied Microbiology 8(1960):269–274.

Fitzgerald, G. P. “Compatibility of Swimming PoolAlgicides and Bactericides.” Water and SewageWorks, 115(1968): 65–71.

Fitzgerald, G. P. and D. F. Jackson. “ComparativeAlgicidal Evaluation Using Laboratory and FieldAlgae.” Journal of Aquatic and Plant Management17(1979):66–.

Gerba, Charles P. and Jaime N. Naranjo. “Comparisonof the Anti–bacterial Properties of Hot TubAdditives and Disinfectants.” Seminar at NSPIMeeting in Las Vegas, November 1999.

Grenier, J. and R. Denkewicz. “Improved Test Methodfor the Evaluation of Algicides and Algistats forSwimming Pools.” NSPI Symposium SeriesVolume II, 1997, pp. 63–68.

Hartwig, W. “To DIN or not to DIN: Ozonation of PoolWater in public and Commercial Pools.” Journalof the Swimming Pool and Spa Industry,2(1)(1996):25–33.

Hafer, D. “A Field Evaluation of the Bi–Polar OxygenSanitation System/Mineral Purification System.”Journal of the Swimming Pool and Spa Industry,1(3)(1995):39–51.

Hass, C. N. and R. S. Engelbrecht “Chlorine Dynamicsduring Inactivation of Coliforms, Acid–FastBacteria, and Yeasts”, Water Research14(1980):1749–

Hoff, J. C. “Inactivation of Microbial Agents byChemical Disinfectants.” U.S. EnvironmentalProtection Agency, Report No. EPA/600/2–86/067, 1986.

Hoigne, J. and H. Bader. “Rate Constants forReaction of Ozone with Organic and InorganicCompounds – I: Non–dissociating OrganicCompounds.” Water Research, 17(1983):173–183.

Hoigne, J. and H. Bader. “Rate Constants forReaction of Ozone with Organic and InorganicCompounds – II: Dissociating OrganicCompounds.” Water Research, 17(1983):184–185.

Hoigne, J., H. Bader, and W. R. Haag. “RateConstants for Reaction of Ozone with Organicand Inorganic Compounds – III: InorganicCompounds and Radicals.” Water Research,19(1985):993–1004.

Kutz, S. M., L. K. Landeen, M. T. Yahya, and C. P.Gerba. “Microbiological Evaluation of Copper–Silver Disinfection Units.” 4th Conference onChemical Disinfection, Binghamton, NY April10–13, 1988.

Kuwabara, J. S. and H. V. Leland. “Adsorption of


Selanastrum capricornutum (Chlorophyceae) byCopper.” Environmental Science and Technology5(1986):197–.

Lange’s Handbook of Chemistry. 13th Edition, J.A.Dean, ed. New York:McGraw–Hill Book Co., 1985.

Legend Labs, St. Paul, MNMarks, H. C. and F. B. Strandskov. “Halogens and

their of action.” Annals of the New York Academyof Sciences 53(1950):163–.

Minisci, F., A. Citterio, and C. Giordano “ElectronTransfer Processes: Peroxydisulfate, a Usefuland Versatile Reagent in Organic Chemistry.”Accounts of Chemical Research 1983, 16, 27-32.

Nalepa, C.J. et al. “Development of a BromineStabilizer for Outdoor Pools.” 4th Annual NSPIWater Chemistry Symposium, Las Vegas, NV,November 30, 1999.

Olin Corporation 1983.Palmer, C. M. and T. E. Maloney. “Preliminary

Screening for Potential Algicides.” 55(1955):1– .Shapiro, R. and F.E. Hale. “An Investigation of the

Katadyn Treatment of Water with ParticularReference to Swimming Pools.” Journal of theNew England Water Works Association, 51, 113(1937).

Taylor, D. S. and J. D. Johnson. “Kinetics of ViralInactivation by Bromine.” in Chemistry of WaterSupply and Treatment, A. J. Rubin, ed., AnnArbor, MI:Ann Arbor Publishers, 1975, pp 369–407.

White, C.G. Handbook of Chlorination. New York:VanNostrand Rheinhold, 1972.

Wojtowicz, J. A., Unpublished data 1985–2000.Wojtowicz, J.A. “Chlorine Monoxide, Hypochlorous

Acid, and Hypochlorites.” Kirk–OthmerEncyclopedia of Chemical Technology, FourthEdition, Vol. 5, New York, NY:John Wiley &Sons, Inc., pp 932–968, 1993.

Wojtowicz, J. A. (1996a) “Relative BactericidalEffectiveness of Hypochlorous Acid andChloroisocyanurates.” Journal of the SwimmingPool and Spa Industry, 2(1)(1996):34–41.

Wojtowicz, J. A. (1996b) “Ozone.” Kirk–OthmerEncyclopedia of Chemical Technology, FourthEdition, Vol. 17, New York:John Wiley & Sons,Inc., NY.

Wojtowicz, J. A. (1998a) “Fate of NitrogenCompounds in Swimming Pool Water.” NSPISymposium Series Volume III, pp. 40–44.

Wojtowicz, J. A. (1998b) Letter to the Editor, Journalof the Swimming Pool and Spa Industry,3(1)(1998):41–43.

Wojtowicz, J. A. (1999a) “Cyanuric Acid Technology.”Journal of the Swimming Pool and Spa Industry,to be published.

Wojtowicz, J. A. (1999b) “Oxidation of NitrogenCompounds in Swimming Pool Water.” Journalof the Swimming Pool and Spa Industry,4(1)(2001): pp. 30–40.

Wuhrman, K. and F. Zobrist. “Investigation of theBactericidal Effect of Silver in Water.” Schweiz.Z. Hydrol. 20(1958):218– .

Zhang, Z. “Disinfection Efficiency and Mechanismsof 1–Bromo–3–chloro–5,5–dimethylhydantoin.”Ph. D. Thesis, Univ. of Houston, 1988.

Survey of Swimming Pool/Spa Sanitizers and Sanitation Systems€¦ · and sanitation systems used...

Documents

Transcript of Survey of Swimming Pool/Spa Sanitizers and Sanitation Systems€¦ · and sanitation systems used...