Disinfection By-Products in Drinking Water!
Chemical disinfection became an integral part of municipal drinking water treatment over 100 years ago as a vital tool in achieving its principal objective: protection of public health. Oxidation, while not as vital to achieving public health objectives as disinfection, has also been accepted as an important part of drinking water treatment.
ADVERTISEMENTS:
Typically, oxidation is used to address aesthetic concerns such as colour, taste and odour, which impact consumer perception and acceptance of a water as fit for human consumption; unfortunately, disinfection and oxidation produce a variety of by-products, which is the focus of this section.
The discussion of disinfection and oxidation by-products is presented in five sections. Following an introduction to the subject, separate sections are devoted to the by-products formed by chlorine, ozone, chlorine dioxide and chloramines.
Introduction to Disinfection or Oxidation By-Products:
The use of chemicals for disinfection and oxidation is very common, occurring at nearly every drinking water treatment plant in the industrialised world. Familiarity with the by-products of the chemicals used for disinfection and oxidation is important, as these by-products can significantly impact consumer health.
Some of these health effects have been identified in the recent past, and some of these health effects are not fully understood, making these by-products an ever-changing part of the drinking water treatment picture.
An overview of oxidation by-products, including a historical perspective, known by-products, regulatory requirements, practical considerations, the chemistry of chlorine, ozone, chlorine dioxide, and chloramine by-product formation, means of controlling (i.e. reducing) their formation at a water treatment plant and means of removing these by-products, if possible, after they form in a water treatment plant are presented.
Since the introduction of chlorine as a disinfectant in drinking water treatment at the turn of the twentieth century, chemical disinfection has been an integral part of municipal drinking water treatment.
ADVERTISEMENTS:
In addition to the use of chlorine as a microbial disinfectant, other benefits of chlorine—as well as other chemical disinfectants such as ozone and chlorine dioxide—were recognised, including the ability to eliminate colour and destroy many naturally occurring chemicals that cause objectionable taste and odour in the water. Consequently, water treatment plant operators commonly added as much disinfectant as necessary to achieve the desired aesthetic and microbial water quality.
In the early 1970s, researchers in the Netherlands and the United States were able to identify and quantify the formation of chloroform (CHCl3) and other trihalomethanes (THMs) in drinking water and relate this formation to the use of chlorine during treatment.
These early findings led to a large number of studies in the United States on the formation of these ‘by-products’ of chlorination. The studies included several monitoring surveys to assess the magnitude of the problem in drinking water treatment plants across the United States (i.e. occurrence studies), as well as studies to investigate how chlorination by-products were formed and what water quality and/or treatment conditions affected their formation.
It was also discovered that chloroform was not the only chemical formed as a result of the reaction of chlorine with natural organic matter (NOM) present in water and that in the presence of bromide ion (Br–) the reaction between chlorine, bromide, and NOM resulted in the formation of a mix of chlorinated and brominated chemical by-products.
ADVERTISEMENTS:
Using mass balance calculations it has been shown that the known chlorination by-products constitute between 30 and 60 per cent of the total organic halides (TOX) formed upon the reaction of chlorine with NOM and bromide. Therefore, much more work is needed to identify and characterise the remaining unknown by-products of chlorination.
The formation of Disinfection By-Products (DBPs) is not limited to chlorine disinfection. Haag and Hoigne reported the formation of bromate ion (BrO–3) when ozone was added to water containing bromide. Bromate was later identified as a suspected carcinogen.
Ozone addition to natural water was also implicated in the formation of numerous organic by-products, such as aldehydes. Thus far, the presence of these ozone organic by-products in drinking water has not been determined to be a public health concern at the typical levels at which they are formed.
Another disinfectant/oxidant used in water treatment is combined chlorine. In the 1970s, the US EPA identified combined chlorine as an alternative to chlorine because it was believed not to form THMs. Later research showed that while monochloramine, the principle component of combined chlorine, is less reactive than free chlorine, it does react to form DBPs, although at much lower concentrations than are formed with free chlorine.
ADVERTISEMENTS:
The prominent category of DBPs formed due to chloramination of potable water are haloacetic acids (HAAs), mostly the dihalogenated species. The presence of bromide ion will both increase the production of HAAs and shift the speciation to the more brominated species, which are thought to present a greater carcinogen risk.
In addition, Krasner and others reported the results of a 35 utility study in which they identified several new by-products, including cyanogen halides (e.g. CNCl), as by-products of chloramination.
However, no chloramine by-product was believed to be a significant public-health concern until Najm and Trussell reported that N-nitrosodimethylamine (NDMA) was a by-product of chloramine addition to drinking water and waste-water.
Chlorine dioxide forms by-products when added to water, such as chlorite (ClO–2) and chlorate (ClO–3) ions, both of which have been suspected to cause health effects. While the health effects of chlorate and chlorite continue to be a topic of debate, there was adequate concern that a limit on chlorite was adopted by the US EPA.
Chlorine By-Products:
Chlorine is by far the most widely used disinfectant in drinking water treatment. It is also the disinfectant that forms the greatest variety of known by-products. Of the known chlorination by-products, the primary by-products of concern in drinking water treatment are THMs and HAAs.
At an elementary level, chlorine reacts with NOM and bromide ions to form halogenated by-products, as shown in the following reaction:
NOM with or without bromide + chlorine by-products … (9.1)
According to Fuson and Bell the reaction between certain organic chemicals and chlorine to produce ‘haloforms’ has been known since 1822. However, it was only after the discovery of chloroform formation upon chlorination of drinking water in the early 1970s that the occurrence of haloform reactions in drinking water treatment was recognised.
Many researchers have explored the mechanisms of THM and other by-product formation upon chlorination of natural water. Because NOM is complex and has unidentifiable structures in natural water, it has not been possible to identify and verify the specific reaction mechanisms between NOM and chlorine.
Formation of THMs and HAAs occurs at the same time when chlorine reacts with NOM. Ideal conditions for THM formation are different from the ideal conditions for HAA formation. It is generally believed that the reaction mechanism leading to the formation of THMs is base catalysed, meaning the reaction is catalysed by hydroxide ions (OH ) present in the water and therefore proceeds faster at more alkaline pH values.
However, the formation of HAAs is enhanced under acidic conditions. Therefore, pH will directly influence whether THM formation is favoured or HAA formation is favoured and regardless of the exact mechanisms involved in the formation of chlorination by-products, higher concentrations of chlorination by-products are formed with higher concentrations of organic precursor material, bromide ions (inorganic precursor) or chlorine, as shown in Eq. 9.1.
The presence of bromide increases the amount of THMs that are formed when chlorine reacts with NOM. Bromide ions (Br) participate in the reaction between NOM and chlorine to form various by-products that have a mix of chlorine and bromine substitutions (e.g. bromodichloromethane, bromochloroacetic acid).
The pathway of bromide THM formation is: (i) chlorine oxidation of bromide ions to form hypobromous acid (HOBr), (ii) HOBr dissociates depending on the pH to form hypobromite ion (OBr~), and (iii) HOBr/OBr and chlorine reacts with NOM resulting in the formation of a mixture of chlorinated and brominated byproducts. The increase in THM formation when bromide is present occurs because bromide has a molecular weight of 79 g/mole compared to 35.5 g/mole for chloride.
Thus, an equal molar concentration of a highly brominated DBP (e.g. bromoform) compared to a purely chlorinated DBP (e.g. chloroform) will result in a significantly higher mass-based concentration of THMs.
The type of NOM present also influences the amount of DBP formation. Researchers have been able to fractionate NOM present in different water according to its molecular size and/or chemical characteristics. In natural water, NOM consists of humic (hydrophobic) and nonhumic (hydrophilic/polar) fractions.
In most, but not all, natural water, hydrophobic NOM contributes more to THM and HAA formation than hydrophilic NOM on a mass/mass basis. Many conventional and advanced treatment processes preferentially remove hydrophobic NOM over hydrophilic NOM, so it may be important to characterise the source water NOM to assist in selecting the appropriate treatment process for the source water.
Since the discovery of chlorination by-products and the implementation of the THM Rule in 1979, a large amount of work has been conducted to identify and evaluate alternatives for reducing DBP formation (primarily THMs and HAAs). The following alternatives are available to water utilities for reducing chlorination by-products:
1. Use an alternate disinfectant/oxidant.
2. Reduce the free-chlorine contact time.
3. Reduce the concentration of NOM before chlorine addition.
4. Remove bromide before chlorine addition.
5. Change the pH of the water during chlorination.
Removal of Chlorination By-products:
Chlorination by-products are organic chemicals, so they can be removed from water via several trace-chemical removal technologies. These include adsorption on activated carbon or for the more volatile by-products, removal with aeration and air stripping.
Several researchers have evaluated the removal of chloroform with adsorption on GAC, but the adsorption capacity of GAC for chloroform is quite low requiring a high GAC replacement or regeneration frequency, possibly every few weeks.
Similarly, air stripping is theoretically a viable approach for removing THMs from water because they are volatile chemicals. However, DBPs and THMs, in particular, continue to be formed after they are removed as long as there is a chlorine residual. For this reason, coupled with the fact that HAAs are not removed, air stripping is a non-viable alternative for DBP removal.
Chlorine Dioxide By-Products:
Chlorine dioxide (ClO2) is a unique oxidant because it has oxidation power at least equal to that of chlorine, yet, when added to water containing NOM, ClO2 does not break the C-C bond and therefore does not form halogenated organic molecules.
However, ClO2 does produce two inorganic by-products- chlorite (ClO–2) and chlorate (ClO3–). The presence of chlorite in drinking water is of concern because it is believed to have serious adverse health effects. In the United States, the Stage 1 D/DBP Rule of 1998 set the MCL for chlorite in water at 1 mg/l.
There are many sources of chlorite and chlorate in drinking water and their sources should be identified carefully before any mitigation measures are attempted. Chlorate can be present in the raw water entering a treatment plant due to agricultural use, as sodium chlorate has been used as an herbicide for almost a century.
In addition, chlorate salts are also used commercially as oxidising agents in special industries, which can result in significant levels of chlorate in raw water. Chlorate is also a degradation product of chlorine in liquid hypochlorite solutions. Therefore, the use of liquid hypochlorite solutions at water treatment plants may result in significant degradation to chlorate and add significant amounts of chlorate into the water.
Chlorite and Chlorate Formation during Chlorine Dioxide Generation:
In plants using chlorine dioxide, there are two sources of chlorite and chlorate. The first is the process of chlorine dioxide generation itself.
Typically, chlorine dioxide is generated by the controlled reaction between chlorine and sodium chlorite under very acidic conditions as follows:
Chlorite and Chlorate Formation as By-Products:
If excessively high concentrations of sodium chlorite are used in the generator, residual chlorite may remain in the product solution. The residual chlorite will then be injected into the water stream with the chlorine dioxide. New developments in chlorine dioxide generation technologies have improved the generator efficiency and greatly minimised the formation of chlorite and chlorate.
The second source of chlorite and chlorate in plants using chlorine dioxide is the by-products of the decay of chlorine dioxide after it is applied to the water. The decay of chlorine dioxide produces chlorite and chlorate in two ways- The first is through the oxidation of various water constituents such as reduced iron, manganese or NOM.
The reaction typically involves a one-electron transfer resulting in the formation of chlorite as follows:
Researchers have estimated that, in general, 50 to 70 per cent (by mass) of the chlorine dioxide applied during drinking water treatment is converted to chlorite. The formation of chlorite greatly limits the chlorine dioxide dose that can be applied during drinking water treatment knowing that the limit on chlorite is 1 mg/l, unless chlorite removal technologies are implemented downstream.
There is currently no measure that can be implemented at a water treatment plant to reliably reduce the formation of chlorite as a by-product of the decay of chlorine dioxide (Eq. 9.6), except through reducing the chlorine dioxide demand.
Reducing ClO2 demand is especially effective when chlorine dioxide is used to oxidise reduced substances such as iron, manganese and colour present in the source water. When chlorine dioxide is used for disinfection downstream of chemical precipitation, the chlorine dioxide demand can be reduced with improved NOM removal through coagulation or with improved iron and manganese pre-oxidation using other oxidants (e.g. permanganate). With a reduction in the chlorine dioxide demand, a lower chlorine dioxide dose will have to be added, thus reducing the amount of chlorite formed.
Removal of Chlorine Dioxide By-Products:
The uncontrolled production of chlorite is the primary obstacle facing widespread use of chlorine dioxide in drinking water treatment. Because implementing mitigation measures to reduce chlorite formation is not feasible, a significant amount of work has been conducted to evaluate options for the destruction of chlorite after it is formed.
While many options exist for chlorite destruction, the following discussion focuses on the most feasible options for a full-scale water treatment plant:
(i) Reduction with ferrous ion,
(ii) Reduction with activated carbon, and
(iii) Oxidation with ozone.
Reduction with ferrous ion and activated carbon are based on the idea of reducing chlorite to chloride by the following half-reaction-
When ferrous (Fe2+) is added to the water, it releases the two needed electrons to form ferric (Fe3+) with the combined reaction being as such-
The mass ratio of ferrous ion to chlorite is 3.3:1 mg Fe2+/mg ClO2– as shown in Eq. 9.9. The stoichiometry presented in Eq. 9.9 was validated by Latrou and Knocke who also showed that adding ferrous ions (as FeSO4-7H2O) at a mass ratio of 3:1 (mg Fe2+/mg ClO2–) resulted in virtually complete reduction of chlorite in less than 1 minute of reaction time (pH between 5 and 7; temperature between 5° and 25°C).
These researchers also verified that the added ferrous ion was oxidised to ferric (Fe3+) ion, which then enhanced downstream coagulation and flocculation by precipitating as Fe(OH)3(s). These findings were confirmed by Griese who, along with Latrou and Knocke showed that no chlorate was formed from the reaction between chlorite and ferrous ions.
Subsequently, Hurst and Knocke verified this approach under alkaline conditions (pH between 8 and 10) and studied the effect of dissolved oxygen on the Fe2+ dose required. Based on these results, a mass ratio between 3.5:1 and 4:1 is more appropriate under high-O2 conditions (> 5 mg/l) to satisfy the added demand for Fe2+.
Reduction with activated carbon is based on the fact that the surface of activated carbon is a good reducing agent and therefore, may be used to reduce chlorine to chloride. Chlorite is removed by GAC, but the removal efficiency increases rapidly over time and removal efficiency increases with increasing EBCT.
Oxidation with ozone is based on the chemistry of chlorite where chlorite is oxidised to chlorate ions with a strong oxidant such as ozone. Chlorine dioxide is a transient intermediate along the oxidation pathway from chlorite to chlorate.
This is reasonable because the oxidation state of chloride in chlorine dioxide is (+4), which is between that in chlorite (+3) and that in chlorate (+5). Increasing the ozone dose oxidises the intermediate chlorine dioxide to chlorate.
Water treatment plants that use chlorine dioxide as a raw-water pre-oxidant and utilise intermediate ozonation for disinfection can rely on the ozonation process to oxidise the chlorite by-product to chlorate. In plants that use chlorine dioxide followed by ozone, the chlorine dioxide dose can be increased in response to changing raw-water quality conditions without violating the chlorite standard in the finished water. This strategy is of limited value when chlorate is an issue.
Ozone By-Products:
Ozone is the second most frequently used primary drinking water disinfectant in the United States. It is also one of the most effective oxidants for the destruction of chemicals that cause colour, taste and odour in drinking water.
When applied to some raw water, it has been shown to improve the downstream chemical coagulation and clarification, as well as media filtration. As the US water industry searched for alternatives to chlorination in the wake of the discovery of the formation of THMs upon chlorine addition to drinking water, many utilities saw ozone as an excellent, but expensive, alternative.
Ozone addition to natural water forms two types of by-products. Organic by-products that are formed from the breakdown of large-molecular-weight NOM molecules by the added ozone are the first type of by-product.
The organic by-products are primarily aldehydes and ketoacids, which are not yet believed to have any significant adverse public health effects. The second type of ozone by-product is inorganic bromate (BrO–3).
In the presence of bromide, ozonation can result in the formation of bromate, which is classified by the US EPA as a ‘probable human carcinogen’ with a 10-6 cancer risk level of 0.05 μg/l.
When added to water, ozone reacts with NOM and bromide to form various by-products. The kinetics of bromate formation with ozone are far more rapid (i.e. minutes) than those of THM formation with chlorine (i.e. hours to days). A simplified schematic of the primary reaction pathways of ozone with NOM and bromide is shown in Fig. 9.14.
Reaction 1 in Fig. 9.14 represents the breakdown of NOM with ozone to form various organic by-products including aldehydes, ketoacids. While these by-products are believed to be benign at the levels formed during ozonation of drinking water, they are far more biodegradable than their ‘parent’ NOM molecules.
The increase in the biodegradability of the organic material present in the water is of concern because it can promote bacterial growth in the distribution system. The total concentration of the organic by-products is gauged by measuring the Biodegradable Dissolved Organic Carbon (BDOC) concentration present in the water or the assimilable organic carbon (AOC) concentration, which is believed to represent the more readily biodegradable fraction of the BDOC.
In reaction 2 in Fig. 9.14, ozone reacts with bromide to form hypobromite ions (OBr–) which in turn reacts with both ozone (reaction 3) and hydroxyl radicals (HO) (reaction 4) to form bromate (BrO–3). Hydroxyl radicals are intermediate products formed from the decay of ozone in natural water.
However, OBr– is also in equilibrium with its conjugate acid, hypobromous acid (HOBr) (reaction 5). If present in the water being ozonated, ammonia can react with HOBr (reaction 6) to form bromamine (NH2Br). HOBr can also react with NOM (reaction 7) to form brominated organic by-products, such as bromoform.
The reactions included in Fig. 9.14, especially those leading to bromate formation, are not the only reactions that take place between ozone, bromide, and NOM.
The reactions leading to the formation of bromate are shown in Fig. 9.14 and include two primary steps:
(i) The formation of OBr–, and
(ii) The subsequent formation of BrO–3 via two parallel pathways, one where OBr– reacts with molecular ozone (O3) and one where OBr– reacts with the hydroxyl radicals (HO.). Based on the work by Gillogly it appears that the HO- pathway contributed to bromate formation far more than the molecular ozone pathway.
The HO. pathway significance is supported by work that von Gunten and Hoigne reported, where the molecular ozone pathway contributed less than 30 per cent to the formation of bromate during ozonation, and work that Yates and Stenstrom performed, where the molecular ozone pathway generated less than 10 per cent of the bromate formed. In another study, Westerhoff confirmed that approximately 70 per cent of the bromate was formed through HO. mediated reactions.
The HOBr formed can react with NOM to form brominated organic by-products such as bromoform as shown in Fig. 9.14 (reaction 7). Reaction 7 is usually not significant for two reasons. First, bromide levels in natural water seldom exceed 0.5 mg/l.
If 50 per cent of the bromide (0.5 mg/l) is converted to OBr– and 50 per cent of the OBr– is converted to HOBr, only 0.15 mg/l HOBr will be formed (which is equivalent to about 0.11 mg/l as chlorine).
Therefore, the amount of bromoform that would form is expected to be quite small; Trussell and Umphres showed that the addition of 10 mg/l ozone to a natural water containing 1 mg/l bromide formed less than 1 μg/l of bromoform.
The second reason for the unlikely formation of brominated organics is, while decreasing water pH favours the conversion of OBr– to HOBr (pKa is 8.7 at 25°C) via the equilibrium reaction (reaction 5 in Fig. 9.14), it hinders the base-catalysed haloform reaction required to form bromoform.
It has been shown that bromoform formation upon ozonation decreased from 37 μg/l at pH 6 to approximately 15 μg/l at pH 8.5 (bromide = 1 mg/l; DOC = 3.4 mg/l; ozone dose = 10.2 mg/l).
There is little that can be done to reduce the formation of the organic ozonation by-products other than using less ozone and/or removing NOM before adding ozone. Using less ozone is usually not realistic because the ozone dose is typically determined by other factors such as taste and odour control or disinfection requirements.
Utilising less ozone and/or reducing the concentration of bromide before ozonation will result in the formation of lower levels of bromate.
Because using less ozone is usually not realistic, the following two measures have been used to reduce the formation of bromate in some natural water:
1. pH depression; and
2. ammonia addition.
Lowering the pH of water during ozonation is the most reliable and proven method for reducing bromate formation upon ozonation of bromide-containing water. The rationale for this control strategy can be deduced from Fig. 9.14. The pKa for the equilibrium reaction between HOBr and OBr– is 8.8 at 20°C.
Therefore, at lower pH values, a greater portion of the OBr– is converted to HOBr (reaction 5 in Fig. 9.14), thus making it unavailable for the bromate formation reactions (reactions 3 and 4). The published literature includes many examples of the effect of pH on bromate formation, one of which is shown in Fig. 9.15.
At an ozone dose necessary to achieve 0.5-log inactivation of Giardia cysts in the test water (3.4 mg/l), the bromate level formed at pH 8.0 was 2.6 times the level formed at pH 7.0 (13 μg/l compared to 5 μg/l). With an ozone dose necessary for 2-log inactivation of Giardia cysts (6.0 mg/l), the bormate level formed at pH 8.0 was 3.6 times greater than the bromate level formed at pH 7.0 (58 μg/l compared to only 16 μg/l).
In other studies it has been shown that further reduction in bromate formation can be achieved as the pH is decreased to 6.5 or lower, although acid addition for bromate minimisation must be followed by caustic addition for corrosion control in the distribution system.
Because reaction 5 in Fig. 9.14 is an equilibrium reaction between HOBr and OBr–, a fraction of the oxidised bromide will exist as OBr– even at low pH and still contribute to some bromate formation. To further minimise this fraction, ammonia can be added to serve as a ‘sink’ for HOBr by transforming it to NH2Br (reaction 6 in Fig. 9.14), thus creating a continuous driving force for the transformation of OBr– to HOBr.
The effect of ammonia addition on the formation of bromate in a natural water sample is shown in Fig. 9.16. Under ambient conditions (NH3-N < 0.03 mg/l as N), adding 5 mg/l of ozone resulted in the formation of 26 μg/l bromate.
Increasing the ammonia-nitrogen concentration to 0.7 mg/l (NH3-N = 0.7 mg/l as N) decreased the bromate level formed to less than the detection limit of 5 μg/l, which is greater than 500 per cent reduction in bromate formation. Combining pH depression and ammonia: addition to some water may be used to reduce bromate formation by more than 90 per cent.
Unfortunately, while pH depression has been demonstrated to reduce bromate formation, some studies do not show a significant impact of ammonia addition on bromate formation. The reason for the mixed ammonia results is not yet clear. For now, the impact of ammonia addition on bromate formation should be evaluated on a case-by-case basis.
Removal of Ozonation By-Products:
Removal of the two main types of ozonation by-products, inorganic bromate and organic by-products, are discussed below:
While the primary strategy for complying with a bromate standard is to minimise its formation in the first place, there are options for removing bromate from water after it is formed. The chemistry of bromate (BrO–3) is quite similar to that of nitrate (NO3–): both are monovalent anions, both are highly oxygenated, both are at the top of their respective oxidation scales and both anions are microbially reduced to benign forms under anoxic conditions.
Therefore, just like nitrate, bromate can be removed from water through the following water treatment technologies:
1. Ion exchange,
2. Membrane separation,
3. Biological reduction, and
4. Chemical reduction.
These four treatment technologies are not currently practical at full-scale for bromate removal, but a discussion of each of these treatment technologies is important for process understanding and possible future use.
1. Bromate Removal by Ion Exchange:
Anion exchange technologies can be used for bromate removal; however, similar to bromide, the low selectivity of most resins renders IX technology virtually impractical for bromate removal from drinking water. In addition, disposal of the spent high-concentration salt regeneration solution containing the eluted bromate is difficult because of the lack of viable handling and disposal options.
2. Bromate Removal by Membrane Separation:
Membrane separation processes using RO or NF membranes, can achieve greater than 90 per cent removal of a wide range of inorganic ions, including bromate. However, there are three primary drawbacks to the use of membranes for bromate removal, making membrane treatment not practical in this application.
The three primary drawbacks to bromate removal through membrane treatment are:
i. Membrane technology is more costly than any of the bromate formation control strategies, such as pH depression.
ii. Membranes produce a high-TDS residual stream that requires proper disposal (similar to IX brine disposal problems).
iii. Special pre-treatment is typically required or the membranes foul rapidly.
3. Bromate Removal by Biological Reduction:
Bromate can be reduced biologically to bromide by denitrifying bacteria when Dissolved Oxygen (DO) concentrations are low (< 2.5 mg/l) and Empty Bed Contact Times (EBCT) are high (> 25 minutes). Similar to nitrification, bromate reduction is greatly inhibited by increased DO concentration.
Unfortunately, this is a significant drawback to applying biological removal in a full-scale treatment plant because the DO concentration downstream of an ozonation process can be greater than 20 mg/l, especially when pure oxygen is used for ozone generation. The requirement for high EBCT is also a drawback as typical EBCT values in a water treatment plant filter are significantly lower than 25 min.
4. Bromate Removal by Chemical Reduction:
Chemical reduction of bromate to bromide can be achieved using reducing agents such as ferrous ions (Fe2+) or the surface of activated carbon.
Bromate can be reduced by Fe2+ under water treatment conditions according to the following reaction:
Typical bromate reduction through the reaction shown in Eq. 9.3 is on the order of 40 to 80 per cent, depending upon the dose of Fe2+. The unique aspect of using ferrous as a reducing agent is that it is oxidised to ferric (Fe3+), which is a typical coagulant in water treatment.
In plants that practice pre-ozonation, a ferrous salt can be added to the rapid mix chamber downstream of the pre-ozone contactor, which allows time for the reaction between ferrous ions and bromate to take place followed by the precipitation of the ferric coagulant formed.
Even though the DO levels are high, because the rate of oxidation of Fe2+ with oxygen is quite slow, not all of the ferrous ions are oxidised. The residual iron passing through the treatment plant can cause aesthetic water quality problems, rendering this bromate removal option impractical.
Ozonation of natural water forms various organic by-products such as aldehydes and ketoacids, all of which increase the concentration of the Biodegradable Organic Matter (BOM) in the water.
Due to the high biodegradation potential of these by-products, biological filtration downstream of ozonation has developed as the approach of choice for removing organic ozonation by-products. In typical dual-media filters (either anthracite-sand or GAC-sand), biological filtration will occur merely by allowing the water to pass through the filters without a disinfectant residual present.
Several factors affect the removal of BOM with biological filtration.
These include:
(i) BOM type and concentration;
(ii) Filter media type (i.e. GAC, anthracite, and/or sand),
(iii) Water-temperature, and
(iv) EBCT through the filter.
Depending on these factors, steady-state biological performance will be reached within a maximum period of 1 to 2 months. Various studies have concluded that either GAC or anthracite, compared to sand, is necessary to achieve good attachment of the bacterial cells.
Furthermore, research and full-scale operating experience have shown that anthracite performs as well as GAC in warm water for oxalate removal, as shown in Figs. 9.18 and 9.19. Data shown in Fig. 9.18 were gathered under relatively warm temperature conditions of 10 to 15°C and the anthracite-sand filter performed as well as the GAG-sand filter regardless of EBCT.
Similar performance in warm conditions was reported by Price and Krasner. On the other hand, under cold temperature conditions as shown in Fig. 9.19, the GAC-sand biofilter was still capable of removing a fraction of the oxalate (albeit with high EBCT values), while no removal was achieved with the anthracite-sand biofilter. The operational data confirm that under relatively warm temperature conditions, anthracite-sand biofilters perform as well as GAC-sand biofilters.
Biologically active filters develop head loss similar to conventional filters, requiring backwashing on a regular basis.
There are two types of concerns about backwashing of biological filters:
(i) Whether backwashing (especially with air scour) causes excessive detachment of the biomass off the filter media, and
(ii) Whether biofilters must be backwashed exclusively with non-chlorinated water to maintain their viability.
Miltner and Ahmad evaluated these two issues and concluded that rigorous backwashing did not adversely impact biofiltration performance. While AOC removal was impaired immediately after backwashing with chlorinated water, its removal was back to normal levels within a few hours of filter run.
Amritharajah concluded that air scour helped control the long-term buildup of head loss in a biofilter, which was confirmed by Teefy who noted that occasional backwashing with chlorinated water was necessary at a full-scale water treatment plant to prevent excessive biological growth, which otherwise results in shorter filter runs. The biofiltration performance achieved with filters backwashed with chlorinated water approximately every third wash.
Chloramine By-Products:
Converting free chlorine to chloramines via Eq. 9.2 has long been considered a cost-effective measure to reduce THM and HAA formation in chlorinated water. With tighter THM and H AA regulatory standards being implemented, an increasing number of water utilities are converting the secondary disinfectant in their distribution systems from free chlorine to combined chlorine.
Chlorinated by-products also form with combined chlorine, however, the THM and HAA levels formed with combined chlorine are typically less than 20 per cent of those formed with free chlorine. The formation of THMs and HAAs with combined chlorine is likely attributed to the low concentration of free chlorine that is always in equilibrium with monochloramine and also to formation that takes place during the process of mixing.
Based on recent research, it has been found that the reaction of combined chlorine with organic matter present in some waters may result in the formation of NDMA as a by-product. The US EPA currently classifies NDMA as a ‘probable human carcinogen’ and has estimated its 10-6 cancer-risk level at 0.7 mg/l, which is well below the levels measured in chloraminated drinking water. At the present time an action limit of 10 mg/l has been set in the State of California, but no federal MCLs have yet been established.
The chemical mechanism leading to the formation of NDMA is not yet known, but it is a direct byproduct of chloramine and not free chlorine as shown on Fig. 9.20. One proposed, but not confirmed, reaction mechanism is dimethylamine [(CH3)2NH], which reacts with chloramine to form dimethylhydrazine [(CH3)2,N2H2], which in turn is oxidised to NDMA by chloramine.
The research by Najm and Trussell found that the NDMA concentration increased with increasing chloramine dose, and the sequence of addition of chlorine and ammonia had no significant effect on the final levels of NDMA formed.
Further unpublished research by MWH engineers on untreated water from the West Branch of the California State Aqueduct has shown that NDMA formation peaks at a pH of about 7 and at a Cl2/NH3 ratio of approximately 5:1 on a weight basis (Figs. 9.21 a,b,c).
Although significant NDMA formation occurs in the first few hours, it appears that the process continues at a slow pace for several days (Fig. 9.21d). As a result of this prolonged formation time, the highest NDMA levels in the system are likely to be formed where the travel time is the longest.
In a survey of water systems conducted by the California Department of Health Services in 1999 it was found that 50 per cent of the system samples had no NDMA (MDL = 1 mg/l) and 5 per cent of the system samples had NDMA levels above 20 mg/l.
Based on measurements made since that time, higher NDMA levels are generally observed downstream of ion exchange and following the use of cationic polymers produced from the Mannich reaction.
Formation Control:
Alternatives for controlling the formation of NDMA can be gleaned from an examination of the circumstances that promote its formation. Generally, reducing the dose of chloramines, increasing the pH, and minimising the contact time in the distribution system are all practices that will reduce NDMA formation.
It also seems likely that either minimising the use of polyDADMAC cationic polymers or selecting carefully among them will also reduce the levels formed. Care in recycling filter waste wash water has also been shown to be of significance.
Removal of Chloramine By-Products:
The NDMA can be removed by reverse osmosis and photolysis. Although these alternatives may be practical in the treatment of contaminated groundwater or in the water ‘exiting ion exchange processes, it is not a practical solution when the NDMA present results from the by-product reaction with chloramines used principally for residual maintenance rather than primary disinfection.
NDMA forms slowly in the distribution system, reaching a maximum concentration in the consumer’s tap. As a result, any treatment for removal of NDMA would have to be applied at the consumer’s tap. Preventing the formation of NDMA, may be the most economical approach.