Report No DWI0486

Sept 1988

Pressure for compliance with the EEC Bathing Water Quality Directive has prompted investigation into methods of disinfection for marine discharge of wastewater. Disinfection may be an alternative to discharge through long sea outfalls or may provide a temporary solution while outfalls are constructed. Chlorination is the established disinfectant for both water (eg. UK) and wastewater (eg. US), but increasing concern over its environmental impact has led to development of alternatives including UV; ozone; lime; peracetic acid; chlorine dioxide; bromine chloride and gamma irradiation.

The Department of the Environment commissioned CES to undertake a comparative evaluation of disinfection techniques. The objectives of the study were to review both current operational techniques and experimental techniques for the inactivation of bacteria, viruses and pathogens in sewage effluents, with particular reference to marine discharges. The inactivation efficiency of the various techniques was to be compared and the actual or potential capital and operating costs evaluated. In addition, promising techniques for future application and aspects of disinfection requiring further research were to be identified.

The report covers firstly the properties of individual operational or experimental disinfectants, reviewing their production and application; chemistry; mechanism of disinfection; inactivation efficiencies; factors affecting efficiency; and environmental impact. Operational and cost data from pilot and full scale trials, both reported in the literature and provided by UK Water Authorities are presented. Subsequent sections of the report comprise a cross comparison of disinfectant alternatives in terms of their inactivation efficiency, environmental impact and cost; and a brief summary of Water Authority attitudes to disinfection. Factors affecting the choice of disinfectant have been discussed and areas for future research identified.

1. Operational disinfection systems

• Chlorine is the most widely used wastewater disinfectant; in the US approximately 62% of total municipal wastewater is chlorinated. Chlorine is applied either as elemental chlorine (a dense corrosive gas), or as a hypochlorite compound (typically sodium hypochlorite solution) supplied as a 14-15% solution or generated on-site from brine or seawater using electrolysis (OSEC). Hazards associated with the handling of chlorine gas makes the use of sodium hypochlorite solution or OSEC preferable in densely-populated areas.
  
  Chlorine reacts rapidly with ammonia and organic compounds to form chloramines and chlorinated organics. High concentrations of ammonia in wastewater give combined chlorine residuals. Despite a much slower bactericidal activity than free chlorine, given sufficient contact time, chloramines are equally effective disinfectants. However, both free and combined chlorine show low inactivation efficiency for resistant microorganisms such as bacterial spores and cysts.
  
  Chlorine residuals (both free and combined) are acutely toxic to aquatic organisms at low concentrations and are persistent due to their stability. The US EPA has proposed stringent discharge requirements for total residual chlorine (mean 7.4µg/l for saltwater); dechlorination is necessary at many chlorination facilities. Certain chlorinated by-products, eg. trihalomethanes (THMs) are carcinogenic and there are significant adverse environmental effects associated with chlorinated organics formed during wastewater disinfection.
  
  Despite its disadvantages, chlorine remains one of the most cost-effective disinfectants available with low capital and operating costs of under 2p/m³. It is also suitable for application to raw sewage, unlike ozone and UV. Its ease of application makes it suitable for emergency disinfection or as a temporary measure.
• Chlorination/dechlorination was used at over 5000 wastewater disinfection plants in 1987. The two most common dechlorination processes use sulphur dioxide or granular activated carbon (GAC), though sodium bisulphite, sodium sulphite, sodium thiosulphate, sodium metabisulphite and biotin may also be used. Sulphur dioxide gas requires similar dosing equipment as chlorine and reacts with chlorine residuals on an equal molar basis (approximately 1mg/l is required to remove 1mg/l chlorine). A disadvantage is that excessive dosage of sulphur dioxide leads to depression in pH and dissolved oxygen, necessitating reaeration. The total cost of chlorination can be increased by up to 30-50% with the addition of a dechlorination step. Dechlorination has been shown to reduce both chlorine residuals and the mutagenic activity of water. GAC dechlorination can reduce mutagenicity in drinking water both by application before (removal of precursors) and after (removal of organochlorine compounds) chlorination.
• Ozone is being used increasingly for wastewater disinfection in the US and to a lesser extent in Europe. In 1987 there were 19 operational plants in the US and 2 small scale wastewater plants in France. Ozone is an unstable gas which is generated on site by electrical discharge through air or oxygen. It decomposes rapidly in aqueous solution and under alkaline conditions hydrolyses to form the OH radical, which is a powerful oxidant. Conditions for oxidation are improved in acidic pH, through slower direct oxidation. Ozone is both an efficient bactericide and virucide, characterised by its rapidity and the low concentrations required. Typical ozone doses in practice for secondary effluents are 10 mg/l for a contact time of 10 min. Ozonation is currently only practicable on effluents treated to at least secondary quality. Ozonation of lower quality effluents is limited by the high ozone demand associated with high COD and SS levels. One advantage over chlorine is that ammonia levels in wastewater have little effect on disinfection efficiency. Though ozone appears not to produce THMs and may even destroy a number of THM precursors, it oxidises a wide range of natural organics in wastewater and can lead to significant changes in the nature and concentrations of certain organic compounds. Ozone destroys most of the non-volatile organic constituents in wastewater but produces others; concentrations of mutagenic micropollutants can be increased by ozonation.
  
  Though highly toxic, the hazards from ozone are minimised because it is utilised immediately. However, operators need protection by adequate leak detectors as concentrations of ozone generated are in excess of occupational exposure limits. Accident risks at ozonation plants are low; in over 70 years of large scale disinfection of potable water in Canada and Mexico no fatalities have been reported. Costs are largely generating equipment costs and the energy costs involved in ozone production; approximately 30% of total costs arise from amortisation of capital and 17% from power consumption. Comparison of operational wastewater disinfection plants in 1984 found that capital and operating costs for ozonation were approximately 2.5 and 2 times those of chlorination/dechlorination respectively.
• Ultra-violet (UV) disinfection systems were operating at 53 wastewater plants in the US and Canada in 1984, with 64 in the design or construction phase. These facilities operate sucessfully on secondary effluents with flow capacities from 2.7m³/d - 2.2m³/s. Wastewater quality is a major limiting factor because high concentrations of solids can absorb UV and protect microorganisms by encapsulation. Secondary treatment is necessary for UV to be both technically and economically feasible. Factors governing the efficiency of disinfection are UV dose (a product of lamp intensity and contact time) and wastewater quality. Typical operational dosages for secondary effluents range from 30 to 50 mWs/cm². Inactivation of viruses requires 3-4 times the dose for bacteria, whilst bacterial spores and cysts are up to 9 and 15 times more resistant. Comparison of UV dose with chemical dosages is difficult but the range of UV doses for different pathogens appears narrower than that of chlorine. It is a more efficient virucide than chlorine.
  
  An advantage of UV irradiation is that it produces no harmful by-products and causes only slight chemical changes in non-volatile organics in wastewaters. UV does not produce a lasting residual thus there is no residual toxicity to marine organisms. There is no economy of scale for UV because the effluent flow rate is proportional to the number of lamps. The major capital outlay is for UV lamps, while operational costs arise from lamp replacement and electricity consumption. Capital costs tend to be higher than those of a dechlorination plant whilst operating costs are of the same order of magnitude.
• The Clariflow single-stage lime treatment process was developed in the UK by Blue Circle Industries plc in conjunction with Southern Water and Portsmouth Polytechnic. The first plant has been in operation at Sandown, Isle of Wight, since 1985 and successfully treats up to 21,000m³/d of raw, screened sewage. A patented lime-based slurry facilitates the production of an upflow sludge blanket. A minimum pH of 10.5-11.0 is required to achieve adequate coliform disinfection, although at lower pH improved removals of BOD, SS and metals are observed. A high pH regime has been shown elsewhere to be effective for inactivation of other pathogenic microorganisms including viruses.
  
  The effluent is necessarily of a high pH, which may have a localised ecological impact at the site of discharge. The process also generates a large volume of sludge which must be dewatered and disposed of either to agricultural land or, as at Sandown, to landfill. Capital costs are high, although these may decrease for subsequent plants. Operating costs are high, typically 5 times those of chlorine. Costs rise disproportionately with pH, therefore it has been suggested that a lower pH regime may be operated in winter when bacterial quality is less critical.

2. Experimental disinfection systems

• Chlorine dioxide has had extensive use as a water disinfectant in Europe and -the US but has yet to be used as a wastewater disinfectant. It is both a powerful bactericide and virucide even at high pH levels and has an important advantage over chlorine in that it does not appear to produce THMs.
  Chlorine dioxide is a yellow explosive gas produced in situ from the reaction of sodium chlorite with either chlorine gas or hydrochloric acid. Even when generated on site there are hazards involved in chemical handling, although development of a new method of production under vacuum eliminates the explosive hazard. Chlorine dioxide is a more powerful oxidant than chlorine and is also a more effective virucide. Use of indicator organisms may thus yield conservative performance data for its disinfection of secondary effluent. Chlorine dioxide dosages of 0.05-1.0 times and residuals of 0.3-0.1 times that of chlorine achieve equivalent bacterial reductions. Though THMs are not formed, chlorine dioxide can react with organics to yield other potentially hazardous chlorinated or unchlorinated by-products, some of which are known carcinogens. Potential environmental impacts on marine biota are not well documented. Chlorine dioxide is technically feasible as a wastewater disinfectant, as demonstrated by short term operation at a full scale wastewater plant, but operational costs are high because of the price of the feed chemical. Capital costs are comparable to chlorine, with only a slight modification of generation and feed equipment necessary. Total costs tend to be 2-5 times as high per m³ as chlorine.
• Peracetic acid (PAA) exists as an equilibrium mixture with hydrogen peroxide, acetic acid and water and is produced in the UK, solely by Interox, under the trade name Oxymaster (12% w/w PAA). Application of PAA is simple and direct and it is suitable for disinfection of all sewage types. Whilst it has been shown to be an efficient bactericide at concentrations of 15-20 mg/l PAA and 2 min contact time; it is less effective as a virucide and sporicide, with dosages of 100 mg/l and contact times of 30 min required for viral inactivation.
  
  The constituents of Oxymaster are known to cause toxic and mutagenic effects, but its rapid reaction and dissipation means that significant residuals are unlikely to occur. The highly oxidative nature of PAA makes it capable of reacting with organic compounds in wastewater; there is concern over the possible formation of epoxides and (through free chlorine formation by peroxide radicals) chlorinated organic compounds. Concentrations might be expected to be low but there are currently inadequate data to evaluate the significance of this aspect. By-products were not detected in seawater during disinfection trials with PAA. Although capital costs for PAA are low, the high feed chemical costs means that operating costs may be up to 6 times those of chlorine per m³ of sewage treated.
• Bromine chloride is an interhalogen compound formed from an equilibrium mixture of bromine and chlorine. It is hazardous at low concentrations; although the reliability of storage and dispensing methods have improved, this has been a major constraint in its widespread application. Hydrolysis products of bromine chloride react with nitrogenous compounds to form bromamines (analogous to chloramine formation by chlorine). The inorganic bromamines are more effective bactericides than inorganic chloramines while organic bromamines also display some disinfectant properties. Though chlorine and bromine chloride are equally effective as bactericides, bromine chloride is a more efficient virucide, particularly in the presence of organic or inorganic interfering substances.
  
  Bromine-chloride is less toxic to fish than chlorine because of the shorter half-life of its residuals. However, formation of brominated THMs may pose more serious problems than the chlorinated analogues. Brominated organic compounds can accumulate in fish exposed to effluents disinfected with bromine chloride. Bromine chloride requires much the same handling and dosing equipment as chlorine but due to the higher rate of reactivity of bromine chloride, smaller contact tanks may be used thus reducing capital costs. Operating costs are reportedly of the same magnitude as those for the chlorination/dechlorination process.
• Ionising radiation has been investigated as both a sludge and wastewater disinfectant; it is already used in over 30 countries to increase the shelf life of food. Radiation energy can be produced from an electron accelerator or a gamma-radiation source such as Co-60 or Cs-137. Energised electrons produce a much higher dose rate than Co-60 or Cs-137 sources but, unlike gamma radiation sources, have a relatively short penetration range. Although effective in the inactivation of most pathogens, the safety aspects and high costs involved are serious drawbacks. Costs arise from the handling and shielding equipment and source replenishment. Gamma-radiation tends to be cheaper than electron acceleration at small facilities due to the investment costs of a high voltage accelerator. Although there is considerable industrial experience in the US with the use of gamma-radiation sources, radiation is inherently hazardous and requires proper protection and shielding facilities.

3. Possible disinfection techniques
   
   Embryonic techniques considered included bromine; heat; sensitized photo-oxidation; quaternary ammonium compounds; activated carbon adsorption; protein precipitants; filtration; and ultrasound. Although several have been evaluated at pilot scale for the disinfection of wastewater, technical limitations (filtration, ultrafiltration) or cost considerations (activated carbon, heat) mitigate against their use at full scale. None were considered to provide a viable technique which could be developed for application within the next 5 years, although ultrasonication could be a useful pretreatment method for ozonation and UV irradiation.
4. Comparisons of alternative disinfectants
   
   Comparative summaries of both operational and experimental disinfection systems are given in Tables 1 and 2.

• In considering inactivation efficiency, it appears that the most effective disinfectants in terms of the range of doses required for bacterial and viral inactivation are chlorine dioxide followed by ozone (see Figure 1). UV and gamma-irradiation also appear effective, although data on viral inactivation in wastewaters are limited. The halogens, bromine chloride and bromine, are more effective than chlorine as virucides, though as bactericides their efficacies are comparable. PAA requires high doses to be effective as a virucide. Chlorine, both free and combined, is an efficient bactericide but its use as a virucide appears limited. The limited information available on lime treatment suggests that the Clariflow process is likely to be effective for virus inactivation.
• The potential for residual toxicity effects on the marine environment is dependent on the half-life of the residual. Chlorine has the greatest hazardous potential, although the use of dechlorination with sulphur dioxide can attenuate toxicity. This is particularly important in relation to discharges to small rivers (as in many cases in the US), where the capacity for dilution is much less than for marine discharges. Bromine residuals are generally shorter-lived and both bromine and bromine chloride have been shown to be less toxic than chlorine. Whilst PAA may have a toxicity equal to that of chlorine, it is unlikely to have a significant effect in practice because it is unstable and short-lived. Although demonstrably toxic, ozone appears to present few problems regarding residuals, due to its short half-life and rapid dissipation from water. There is little available information on UV, which cannot exhibit a direct toxic effect but may cause a physical alteration of certain compounds and hence an indirect toxic effect.
• By-product formation by chlorination is unequivocal and health hazards of eg. carcinogenic THMs, are widely documented. The environmental effects of wastewater chlorination have been demonstrated in trials by Welsh Water, where bioaccumulation of chl orinated organics in marine biota was observed. However, a decline following termination of disinfection may have implications for seasonal disinfection, suggesting that there may be potential for a degree of purification of the biota during non-disinfection periods.
  
  The effects of by-products formed by alternative disinfectants are still largely unquantified. Whilst the strongly oxidative nature of ozone, PAA and chlorine dioxide may be expected to alter the components of the effluent matrix, many of the oxidation products do not appear to be harmful. By-products of concern are aldehydes, hydroperoxides and mutagenic activity of ozone; peroxides (possibly mutagenic) and small amounts of chlorinated organic compounds by PAA through generation of free chlorine; and brominated organics by bromine and bromine chloride. Generation of halogenated organics may occur with chlorine dioxide if free chlorine is present, although chlorine dioxide itself does not form THMs. There is no evidence for mutagenic activity of chlorine dioxide. Gamma-irradiation would be anticipated to cause mutagenic activity, but there is insufficient operational experience to provide evidence for this in practice.
  
  Due to the lack of long-term operation of any of the disinfection processes other than chlorine, evaluation of their hazardous effects remains largely unresolved. However, whilst the absence of any deleterious effects seems unlikely, it would seem that the environmental impact of disinfectants such as ozone and UV will be of a lesser magnitude than that arising from the use of chlorine compounds.
• The relative costs of the different processes as reported in the literature are specific to a particular plant size. For example, chlorine shows distinct economies of scale and whilst there is no technical limit on the size of a UV plant, the costs tend to become prohibitive at larger facilities because of the high operating costs. On an economic basis, chlorine, hypochlorite and UV (at small-medium plants) are the most cost effective. However this does not take into account any remedial costs for precluding adverse environmental effects, eg. by dechlorination. A dechlorination system, necessary to prevent residual chlorine toxicity, adds 30-50% to the costs of chlorine and sodium hypochlorite systems, and makes them comparable with the costs of the halogens such as bromine chloride and chlorine dioxide. Also additional benefits of a process such as colour removal with ozone or phenol destruction with chlorine dioxide, are not accounted for in a simple cost analysis. Ozone is generally reported to be some 2-8 times more costly than chlorine, sodium hypochlorite approximately equal to chlorine, and chlorine dioxide and bromine/chlorine up to two times as expensive as chlorine.
  
  Costs obtained from UK manufacturers for a theoretical case study based on a population of 25,000 and a DWF of 4,500 m³/d are shown in Table 3. From this it can be seen that the costs of disinfecting crude sewage, for which not all systems are appropriate, increase according to the following series;
  
  Cl₂(gas) < Cl₂(hypo) < Cl₂(OSEC), PAA < Clariflow.
  
  As with all chemical disinfectant systems, there is greatest sensitivity of cost to dose. The costs given for chlorine are based on a low dosage of 20 mg/l typically used in pilot studies but estimates are also given for a higher dosage of 100 mg/l, as used at a full-scale UK plant. Chlorine gas and hypochlorite remain the cheapest options even at higher dose rates; whilst the cost of OSEC approaches that of PAA. OSEC is significantly cheaper than the Clariflow process, which is the most expensive option with a total cost approaching that of land-based treatment. However, the Clariflow system does have additional benefits such as removal of BOD, 55 and heavy metals. There are insuffient data to evaluate the costs of ozonation of crude sewage because it is generally only used for treated effluents. Assumed costs based on the latter are intermediate between PAA and Clariflow.
  For the disinfection of secondary effluents, costs increase as follows;
  
  Cl₂(gas) < Cl₂(hypo) < Cl₂(OSEC) < UV, PAA < 03
  
  There is a smaller range between alternatives for treated effluents; UV appears attractive particularly if compared with chlorination plus dechlorination, and is of a similar cost to PAA. Ozone remains one of the most expensive options. The comparative magnitude of process costs obtained in the case study is similar to that reported for operational plants absolute costs vary with plant size and effluent quality.

5. Water Authority attitudes to disinfection
   
   A summary of disinfection trials carried out by Water Authorities is given in Table 4. The general concensus among Water Authorities would appear to be to opt for long sea outfalls wherever possible. Even with adequate screening and disinfection, the public acceptability of swimming in sewage polluted waters is questionable. Although there is a great deal of interest in the potential for disinfection, there is a general reluctance for its adoption on a long-term basis because of the lack of proven efficacy and the high operating costs. The awareness of environmental impact observed by water authorities, both in-house and increasingly amongst consumers, also tends to mitigate against disinfection. It is however being seriously considered, as for example by South West Water, as a short term measure while outfalls are constructed.
6. Conclusions
   
   There are several processes available which can disinfect sewage to achieve coliform levels permitting discharge through comparatively short outfalls. Estimates of UK costs show prices ranging from approximately 1.6 p/m³ to 19 p/m³ for year-round disinfection or 2.0 p/m³ to 27 p/m³ for six monthly operation. These costs would reduce slightly for very large outfall schemes, and in some cases such as Clariflow, are associated with additional benefits such as increased BOD, 55 and heavy metal removal. In this respect (although not included in the present remit), it would be of interest to compare the costs of disinfection options with those for a long sea outfall. The figures presented should also be viewed in the context of the costs levied by Water Authorities for full biological treatment for subsequent discharge to an inland watercourse, typically approaching 25 p/m³. In relation to this, those processes at the lower end of the cost range indicated may be attractive if they can be shown to be environmentally acceptable.
   
   A major drawback with disinfectants is the formation of hazardous by-products. It is difficult to interpret existing information in the context of the marine environment rather than in relation to human consumption, since much of the toxicological data for chlorinated by-products relates to exposure via drinking water. The processes with the least by-products are either very costly or not presently effective for raw sewage. The only option without associated by-products is a long sea outfall. Until further evidence proves otherwise, it is considered that disinfection by chemical means should only be used on a short term basis, while-an outfall is being constructed. It is interesting to note the apparent recovery of marine organisms from the effects of chlorinated discharge in the Welsh Water study; this suggests that seasonal disinfection may be less environmentally detrimental than year-round treatment. Nevertheless, the possible establishment of reservoirs of pathogens in sediments during the winter months would have to be reviewed at individual locations.
   
   Where standards are imposed rigorously and beaches have to remain open for five years or so until an outfall scheme (or land-based treatment) comes on-stream, a choice of disinfectant may well have to be made. If economic considerations prevail, chlorine compounds (specifically hypochlorite) are attractive. Seasonal application and dechlorination to preclude residual toxicity could be used to reduce environmental impact. However, experience at Weston-Super-Mare, where dosages at the upper end of the chlorine range in Table 3 are applied and compliance is still not always achieved, has to be borne in mind. Although about three times more expensive than gaseous chlorine, the lower degree of by-product formation associated with chlorine dioxide suggests that it merits further investigation as a wastewater disinfectant.
   
   If economic aspects are not overriding and minimal environmmental impact is considered important, the Clariflow process could have application. However, construction time, land acquisition, sludge disposal and an overall cost approaching that of land-based treatment would have to be considered. Ozone is a potentially attractive option in economic comparison to Clariflow, but its efficacy on crude sewage is not well documented. The intermediate option in terms of both cost and likely impact is PAA, however it is not possible to assess the latter factor adequately on the basis of current data.
   An examination of embryonic technologies for sewage disinfection identified no systems which showed sufficient development potential for use within the next five years. Aspects identified for further research include the following:

• a detailed review of existing installations and operational experience in the US, particularly with regard to environmental effects;
• more comprehensive studies on the efficacy of alternative disinfectants for viral inactivation;
• an evaluation of effects of the use of viral indicators and of the phenomenon of viable, non-culturable cells on the apparent efficiency of disinfectants;
• field trials to assess the degree and significance of by-product formation by disinfectants, particularly chlorine, chlorine dioxide and PAA;
• a practical investigation of the applicability of chlorine dioxide for UK marine discharges.

Copies of this report may be available as an Acrobat pdf download under the 'Find Completed Research' heading on the DWI website.