Natural organic matter as precursor to disinfection byproducts and its removal using conventional and advanced processes: state of the art review

Providing safe drinking water is essential for sustaining human life on earth. With the growing demand for water, it is becoming difficult for drinking water industries to meet the quality needs, both chemically and microbiologically. The chemical aspect refers to chemical contaminants in water sources that are a direct threat to human life. One such contaminant is disinfection byproducts (DBPs) which are formed as a result of disinfection of the water in the treatment process itself. Disinfection is crucial for maintaining the microbiological safety of water, i.e. it aids in inactivating microbial pathogens (bacteria, virus, protozoa etc.) that can cause various water-borne diseases (Gomez-Alvarez et al. 2016). One such disinfectant is chlorine and it is the most widely used across the globe. DBPs are generally formed by the reaction of disinfectants such as chlorine with organic precursors present in source water; these organic precursors are mainly called natural organic matter (NOM) and NOM acts as a forerunner to DBPs. Some of the chlorination disinfection byproducts are shown in Table 1. Trihalomethanes (THMs) are the major class of DBPs formed. Though THMs is not a regular water quality parameter, various studies have reported their occurrence in water systems across the globe and stringent guidelines have been imposed for controlling THM levels in water supply systems (Golfinopoulos 2000; Rodriguez et al. 2003; Ivahnenko & Zogorski 2006; Wang et al. 2007; Kumari et al. 2015). THMs constitute four main volatile organic compounds (VOCs): trichloromethane (chloroform), bromodichloromethane (BDCM), dibromochloromethane (DBCM) and tribromomethane (bromoform). Total trihalomethanes (TTHMs) are the sum of the mass concentrations of chloroform, BDCM, DBCM and bromoform in μg L⁻¹ (Frimmel & Jahnel 2003). THMs have been classified as a probable and possible human carcinogen in group 2B and C (IARC 1999). The removal of THMs after their formation is difficult and involves resource-intensive processes such as activated carbon adsorption or air stripping. Therefore, efforts should be directed towards optimizing the operation of existing water treatment plants to minimize THM formation or developing treatment techniques to degrade natural organic matter (NOM), which are the DBP precursors (Rook 1974).

NOM is a complex mixture of heterogeneous chemical fractions with different polarity, chemical composition, charge and molecular weights (Nebbioso & Piccolo 2013). The chemical characteristics of NOM can be a useful tool to study their correlation with DBP formation. The complete removal of NOM by conventional water treatment processes is relatively inefficient, leading to the formation of DBPs, during either post- or pre-chlorination (Murray & Parsons 2004). NOMs are usually quantified in terms of dissolved organic carbon (DOC). DOC is the part of total organic carbon (TOC) which can be filtered through a 0.45 μm membrane filter. The absence of regulatory guidelines for DOC strengthens the need for more focus on NOM removal as it can be a source of various water quality issues such as color, taste, odor and DBPs. Advanced oxidation processes (AOPs) are highly efficient water treatment processes that can degrade natural and recalcitrant organic matter. AOPs oxidize the highly complex organic matter into simpler compounds, therefore increasing the biodegradability of the NOM. Biological activated carbon (BAC) or biofiltration is also known to remove biological organic matter from source water (Chien et al. 2008; Korotta-Gamage & Sathasivan 2017). AOPs can completely mineralize the NOM, but at higher energy and thus higher cost inputs. Therefore, combined with biofiltration process BAC can be an economical solution for the removal of NOM.

The organic matter in surface and ground water is predominantly natural organic matter (NOM) and this NOM is a complex mixture of organic compounds with different molecular size and properties (Lamsal et al. 2011; Sillanpää & Matilainen 2014). NOM is derived from plants, animals, microorganisms and their waste and metabolic products. Therefore NOM is omnipresent in all natural water sources and even in soil and sediments (Aiken 1985; Suffet & MacCarthy 1988). It is present in particulate, dissolved and colloidal forms. The amount and characteristics of NOM are site-specific, i.e. they are climate, topography and geology dependent (Fabris et al. 2008; Wei et al. 2008). The characteristics of NOM vary both regionally and with time (Wei et al. 2008). Aquatic NOM is a heterogeneous mixture consisting of both hydrophobic and hydrophilic compounds. The hydrophobic fraction, which accounts for more than half of total dissolved organic carbon (DOC), is predominated by humic substances (Sillanpää & Matilainen 2014), primarily humic acids and fulvic acids and other phenolic compounds and carbon with conjugated double bonds. Aquatic fulvic acid is considered to be the major portion of hydrophobic fraction (Aiken et al. 1992). Hydrophobic NOM is rich in aromatic content and is composed of primarily humic material. Humic material is formed by decaying vegetative matter, such as lignin. Lignin is found in plants and is quite resistant to biodegradation, yet is reactive to oxidants, such as chlorine. These characteristics of the aromatic hydrophobic humic material tend to form higher THM levels. The hydrophilic fraction of NOM primarily consists of aliphatic carbon and nitrogen bearing compounds such as carbohydrates, proteins and amino-acids (Sillanpää & Matilainen 2014). The composition of NOM is presented in Table 2.

NOM is predominantly responsible for water quality issues such as color, odor and taste. Colorless water may also have NOM present in significant levels. NOM can act as a major source of microbial re-growth in the water distribution system, if present in treated water. The major issue with NOM is formation of unwanted products such as DBPs upon reaction with chemicals like chlorine. Therefore, it can be concluded that NOM can pose serious water quality issues in any drinking water treatment industry if not well treated. The majority of studies suggest that hydrophobic compounds are the major contributors of DBPs (Fabris et al. 2008; Wei et al. 2008) while few consider the hydrophilic part to be the contributor of DBPs. The source of origin of NOM also plays a very important role in deciding the nature of NOM. Autochtonous and allochtonous are two classes based on the source of origin of NOM. Allochtonous NOM is derived from sources on land that are external to the aquatic system and autochthonous NOM is generated within the water column having mainly algae as a source (algal NOM) (Wershaw et al. 2004; Leenheer et al. 2004; Winter et al. 2007; Berggren et al. 2015). Allochtonous NOM is dominated by hydrophobic content whereas autochtonous NOM is dominated by lower molecular weight hydrophilic molecules (Wershaw et al. 2005). Therefore defining NOM into operationally defined chemical fractions, i.e. hydrophobic, transphilic and hydrophilic, will also help in determining the source of NOM origin.

The presence of NOM in water is well acknowledged but it was in the late 1970s that NOM was identified as a precursor to disinfection byproducts, mainly THMs. THMs have been shown to cause severe health impacts in various epidemiological studies and health risk assessments (Dodds et al. 1999; Richardson 2003; Villanueva et al. 2006, 2015; Wang et al. 2013). Since then the reaction of NOM with disinfectants and other chemicals used in water treatment, and the influence it exerts on virtually every aspect of water treatment, has begun to be appreciated. The idea of haloform formation in water treatment plants stated by Rook (1974) is still a problem in many developing nations like India. The history of THM recognition is shown in Table 3. The United States Environmental Protection Agency (USEPA) was the first agency to set THM standards in 1979. Later on, USEPA also implemented first and second disinfectants/disinfection by-products rules (D/DBP) in 2000 and 2006 respectively. Later on, the European Union (EU) and the World Health Organization (WHO) also set guidelines and standards for THMs. Since then, THMs have been regulated in different nations across the world, primarily developed nations. A few guidelines set for THMs across different parts of the world are depicted in Figure 1. The identification and monitoring of THMs and their sources in developing nations like India is non-existent. In 2012, in its new draft the Bureau of Indian Standards (BIS) included standards for THM levels in BIS code: IS 10500: 2012. Very few studies to date have reported the THM identified in some states of the country (Rajan et al. 1990; Thacker et al. 2002; Sharma & Goel 2007; Hasan et al. 2010a, 2010b; Basu et al. 2011; Mishra et al. 2012).

The reaction of NOM with chlorine is dependent on the chemical characteristics of NOM itself, i.e. hydrophobicity, polarity, nature of functional groups present, aromaticity etc. Chlorinated DBPs such as THMs are generally formed by the reaction of naturally derived organic matter with chlorine (Farkas et al. 1949; Gallard & von Guntem 2002; Westerhoff et al. 2004), but THMs or DBPs can be formed from anthropogenic sources like wastewater treatment plants (Yang et al. 2014). The properties of wastewater effluent derived organic matter (EfOM) are completely different from NOM; EfOM from biological wastewater treatment plants consists of biodegradation and soluble microbial products. The characteristics of NOM and EfOM converge but aromatic moieties in both are of completely different origin (Yang et al. 2014). THMs are mainly studied in drinking water treatment plants where the main source of influent is surface water (lesser anthropogenic influence), but sometimes there can be incidental introduction of treated wastewater into the drinking water treatment plant (in developing nations). NOMs have complex chemical composition; different chemical fractions contribute differently to THM formation. Humic and fulvic acids (hydrophobic fractions) are the most important precursors to DBPs. NOMs can be fractionated on the basis of polarity into hydrophobic, transphilic and hydrophilic using XAD resin fractionation. The different NOM fractions react differently according to coagulant, amount of coagulant, chlorine, ozone and in terms of DBP formation potential (DBPFP) (Fabris et al. 2008). Table 4 describes the role of different chemical groups on THM formation. Various studies have researched NOM surrogates instead of the source NOM itself, with major compounds being aniline, resorcinol etc. in aromatic moieties and L-aspartic acid, 3-oxopentanedioic acid, 2.4-pentanedione etc. in aliphatic properties (Bond et al. 2012). NOM characterization provides a useful insight into NOM composition, reactivity towards chlorine and removal options. Among all the techniques well established in literature, this review will focus on techniques such as UV absorbance at 254 nm (cm⁻¹), specific UV absorbance (SUVA) (SUVA = UV₂₅₄/DOC*100), Fourier transform infrared spectroscopy (FTIR), XAD resin fractionation, and fluorescence excitation − emission matrix (FEEM). SUVA is widely used for estimating the chemical characteristics of DOC in the source water, its amenability to coagulation, and the reactivity with chlorine toward DBP formation (Kitis et al. 2001; Weishaar et al. 2003; Fearing et al. 2004; Van Verseveld et al. 2007; Hua et al. 2015). The significance of SUVA in characterizing NOM in terms of THM forming potential (THMFP) is shown in Table 5 and Table 6 describes the efficiency and use of different NOM characterization techniques.

The detailed mechanism of THM formation and the effect of different process parameters on THM formation is well explained by Rook (1974). Chlorine is usually applied in the form of sodium hypochlorite or in gaseous form. Chlorine readily reacts with water and forms hypochlorous acid (HOCl) which in turn dissociates into hypochlorite ion (OCl⁻). The structure of NOM, especially humic acid, is very complex. The compounds such as resorcinol bear a close resemblance to aromatic NOM with chlorine, with most of the research focused on surrogates such as resorcinol (Frimmel & Jahnel 2003; Bond 2009). Aromatic or phenolic compounds such as resorcinol are considered as the main precursors of THMs. One such reaction mechanism is shown in Figure 2. The formation of THM is affected by various process parameters like chlorine dose, residual chlorine, reaction time, temperature, pH, NOM source, concentration and inorganic sources like bromide ion etc.

The most common technique employed for the removal of NOM from water treatment systems is coagulation-flocculation followed by clarification (sedimentation or flotation), filtration and disinfection. Coagulation is mainly employed for the removal of turbidity and with that NOM and some of the hydrophobic compounds are also removed. This is the most conventional method employed by drinking water treatment systems across the globe. There is no specific technique employed for the removal of TOC or DOC, but various treatment techniques employed remove TOC/DOC along the way. Non-conventional or advanced methods include adsorption, membrane filtration, ion exchange, biofiltration, AOPs or integrated methods such as AOP followed by biologically activated carbon (BAC).

Coagulation is the destabilizing of solid colloidal matter which will result in the formation of micro flocs. The micro flocs thus formed start to agglomerate, leading to the formation of larger flocs due to Brownian motion and this process is called flocculation. Chemical coagulation is generally achieved by the addition of iron or aluminum salts. Coagulating aids such as PACl (poly aluminum chloride) can also be used to enhance coagulation. The possible mechanism of NOM removal by coagulation is shown in Figure 3(a). Most of the NOM is believed to be removed by coagulation. However there is still ambiguity regarding reduction in the hydrophobic or hydrophilic part by coagulation. Most of the studies suggest the removal of the hydrophobic higher molecular weight (HMW) part is more efficient as compared to the hydrophilic lower molecular weight (LMW) part. Such a phenomenon may be due to higher aromaticity of the HMW fraction which carries high charge density and a higher level of negative charge due to the presence of ionic groups such as carboxylic and phenolic groups. Therefore, the HMW fraction tends to dominate the colloidal charge nature of water and is more amenable to removal by coagulation. Enhanced coagulation was introduced as a new regulatory requirement in the USA, primarily aimed at removing DOC and thereby DBP precursor. The main aim is also comparable with optimized coagulation, i.e. maximum removal efficiencies in terms of turbidity, particulate TOC, DBP precursor, least residual coagulant, sludge production and operating cost (Edzwald & Tobiason 1999). All this is mainly achieved by increasing coagulant dose and adjusting pH (Yan et al. 2006).

Collins et al. (1992) studied the effect of slow sand filtration for NOM and subsequently THM precursor removal. The possible mechanism of removal was physical straining, adsorption and biodegradation. The organic matter degradation was found to be dependent on filter biomass which in turn was found to be dependent on the cleaning and maintenance procedures adopted (the filter harrowing technique is more effective than surface scraping cleaning) (Collins et al. 1992). Although slow sand filtration is inexpensive and is the most widely employed water treatment process, it is unable to decrease the DOC to those levels that will prevent DBP formation below the set standard limit (Moncayo-Lasso et al. 2008).

Rapid sand filtration can also contribute to DOC reduction and the main mechanism involved here is adsorption on the flocs and biodegradation. The net DOC removal is dependent on the operating conditions or bioprocesses occurring inside the filtration system. Biological processes occurring inside the filtration system may lead to biodegradable DOC (BDOC) elimination whereas assimilable organic carbon (AOC) removal is dependent on the oxygen concentration in the filter bed (Korth et al. 2001). In both RSF and SSF, biodegradation is among the major pathways for NOM reduction.

Membrane filtration alone is not effective in DBP precursor removal. It is able to achieve only less than 10% DOC removal. However, with pre-treatment or in combination with conventional techniques such as coagulation, comparatively better NOM removal can be achieved.

The DBP removal through microfiltration along with coagulant is a site specific system and should be optimized for a particular site because of spatial variation of NOM. One more advantage of using membrane filtration along with coagulation is that the membrane filtration uses a physical barrier for achieving microbial and particle removal, therefore coagulation is not required to achieve the filtration objective and the process chemistry is specifically optimized for the removal of NOM (Vickers et al. 1995).

GAC is a highly porous, effective adsorbent widely used for drinking water treatment for color, odor, taste and organic contaminants including NOM removal. GAC has macroporous, rough surfaces with widely distributed fissures and ridges in contrast with the non-porous smooth surfaces of sand filters. The mechanism of DOC removal by GAC is represented in Figure 3(b). The total organic carbon (TOC) concentration was found to be lower in GAC-filtered water than in sand-filtered water (Hyde et al. 1987). The adsorption process of the GAC mainly depends on the surface area, pore structure and surface chemistry (Moreno-Castilla 2004). The rough surfaces of GAC also provide an excellent site for microbial attachment and provide shelter to newly attached bacteria, protecting them from shear forces that are a major hindrance during biofilm development. The type of GAC also plays a very important role in DOC removal and biodegradation occurring in the biofilters or biologically activated carbon filters (BAC) (Karanfil et al. 1999). GACs are either chemically or steam activated and are prepared from different sources such as coconut husk, wood or coal. Various studies reported steam activated coal based carbon to be the best adsorbent for DOC removal (Yapsakli & Çeçen 2010). The major advantage of GAC in terms of adsorption is that it can adsorb both readily and slowly biodegradable organics although higher molecular weight compounds are not easily removed by GAC due to their larger size (sieving effect), whereas intermediate and lower molecular weight compounds are easy to remove (Yan et al. 2006; Xing et al. 2008). GAC filtration is suggested to be most effective in the removal of intermediate molecular weight compounds (IMWs). The GAC, which has bioactivity on its surface and removes a significant amount of DOC by biodegradation, is called biological activated carbon (BAC) (Nishijima & Speitel 2004). Pre-treatment of water supplied to GAC may lead to increased BAC performance. Therefore, in the case of pre-treatment before GAC, such as oxidation where all the non-biodegradable DOC is converted to biodegradable DOC (BDOC), it becomes easier to remove the BDOC fraction by BAC (Yapsakli & Çeçen 2010).

The effect of each unit process on organic matter removal needs to be assessed to obtain a better perspective while choosing the appropriate water treatment process for enhanced NOM removal. One such study has been conducted by Chen et al. (2007), who reported the effect of conventional processes (along with different modifications) on organic matter removal. The effect of different process parameters in terms of TOC, UV₂₅₄, TTHMFP (total trihalomethane formation potential), THAAFP (total haloacetic acid formation potential) reduction was reported. Maximum removal was 30, 36, 41 and 55% for TOC, UV₂₅₄, TTHMFP and THAAFP respectively (Chen et al. 2007). The effect of different treatment processes in drinking water treatment plants on DOC and THMFP reduction is described in Table 7.

NOMs act as a forerunner to DBPs and conventional removal processes such as coagulation, filtration etc. do not guarantee the total NOM removal or reduction in DBP formation potential (DBPFP) (Moncayo-Lasso et al. 2008). A well-explained review on removal of NOM from drinking water by AOPs is given by Matilainen & Sillanpää (2010). The next section will cover the advancements in AOP for NOM removal in the last 10 years.

AOPs involve the generation of highly reactive radical intermediates, especially the OH^. radical (Glaze et al. 1987). The advantage of AOPs is the conversion of high molecular weight (hydrophobic) organic compounds (HMWs) into low molecular weight (hydrophilic) organic compounds (LMWs) with a system operating at ambient pressure and temperature and sometimes complete mineralization. Most studies suggest HMWs to be the root precursor of DBPs (Zhang & Jian 2006; Liu et al. 2010). The factors that make OH^. radical advantageous over other oxidants are its higher oxidizing capacity (Table 8), non-selective nature and the fact that the reaction rate constant of OH^. radical with organic species is usually several orders of magnitude higher than oxidation processes, as shown in Table 9. Westerhoff et al. (2007) studied the reaction of several dissolved organic matter surrogates and have established their reaction rate constants, as demonstrated in Table 10.

The most commonly used combination for AOPs is O₃/H₂O₂, UV/H₂O₂, UV/O₃, Fe^2+/H₂O₂, Fe²⁺/H₂O₂ + hv, vacuum UV (VUV), and UV/TiO2. The first free radicals are generated followed by a chain of reactions as shown in Table 11. The reaction of NOM with OH^. radical occurs in three ways (Matilainen & Sillanpää 2010):

1. addition of OH^. radical to double bond;

2. H-atom abstraction yielding a carbon centered double bond which can react rapidly with oxygen to form organic peroxy radicals which can subsequently lead to the production of aldehyde, ketone or CO₂;

3. OH^. radical gaining electron from organic species.

Ozonation can also be considered as an AOP in the case of higher pH values and in combination with peroxide. Ozone is unstable in water and tends to decompose rapidly and another major oxidant that is formed from ozone decomposition in water is OH^. radical. Table 11 shows the reaction mechanism of ozone with NOM.

The extent of NOM mineralization depends on various factors, i.e. radiation dose, oxidant dose and reaction time. Factors that affect radical formation are mainly pH, temperature, presence of ions, pollutant type as well as presence of scavenging agents such as bicarbonate ion. The rate of oxidation is dependent on radical, NOM and oxygen concentration. This section also combines the literature from 2000–2018 on various studies of AOP for NOM removal, as explained in Table 12.

The GAC which has bioactivity on its surface and removes a significant amount of DOC by biodegradation is called biological activated carbon (BAC) (Nishijima & Speitel 2004). BAC is one of the most promising, eco-friendly and economically feasible processes for enhancing water treatment performance. BAC is advantageous compared to GAC as eventually adsorption sites become saturated with organics leading to GAC exhaustion. The biofilm formation starts over the rough porous surface of GAC and bacterial colonization starts utilizing organics on the surface as the food source. The biofilm developed has the potential to degrade organic pollutants, including biodegradable organic matter by biodegradation, therefore prolonging the life of the carbon bed and without requiring regeneration like GAC (Dong et al. 2015; Korotta-Gamage & Sathasivan 2017). Since microbes are attached to the surface, the supply of organics or substrate to microbes in biofilm is mainly controlled by a bulk and surface transport phenomenon. The substrate must be transported from the liquid phase to the biofilm outer surface and then to microbes inside the biofilm by diffusion. The factors that affect the rate of substrate utilization within a biofilm are: (a) substrate transfer to biofilm; (b) diffusion of substrate into the biofilm; (c) substrate utilization within the biofilm; (d) substrate growth yield; and (e) biofilm detachment.

Other factors affecting BAC performance are as follows:

• (a)

  Filter media and characteristics: Media characteristics play a huge role in pollutant removal and the choice of media depends on type of pollutant to be removed. In the case of drinking water pollutants, GAC is the best media type, though GAC particle size has been shown to have little to no effect on NOM removal (Velten et al. 2011).

• (b)

  EBCT: This is a key design and operational parameter of any biofilter. The removal of organics usually increase with increase in EBCT up to optimum value (Laurent et al. 1999). Han et al. (2013) compared up-flow (UBACF) and down-flow (DBACF) BACs and showed that the retention time for UBACF (10 min) was slightly higher than DBACF (8 min) due to 25% bed expansion in the case of UBACF. The NOM removal by BAC took place in two phases, the first phase being the start-up phase in which non-ozonated feed water was directly supplied to initiate bacterial colonization; and the second phase is steady state of biodegradation in which bacterial respiration and biomass assimilation accounted for most of the NOM removal. Most of the studies suggest that it takes around 60–90 days to achieve the steady state (Velten et al. 2011; Han et al. 2013). In the start-up phase, NOM removal efficiency of DBACF was slightly higher than UBACF whereas in the steady phase removal efficiency of UBACF was 10% higher than DBACF.

• (c)

  Backwashing: It is important to use an appropriate backwashing technique for filter backwashing to maintain the microbial attachment on the BAC surface and to restore head losses (Miltner et al. 1995; Ahmad & Amirtharajah 1998; Putz et al. 2005). Though backwashing is an important aspect to be considered while running a BAC system, a few studies neglected backwashing for the sake of an in-depth depth study of the overall distribution of biomass over the BAC bed (Gibert et al. 2013a, 2013b).

• (d)

  Temperature: Bacterial activity tends to increase with temperature increase within a range of 10–30 °C because this temperature range is favorable to bioactivity of many bacterial communities (Billen et al. 1992; Chaudhary et al. 2003).

• (e)

  Pre-oxidation: BACs mainly target the lesser complex or LMW organic compounds and have been shown to reduce the chemical dose, i.e. chlorine, coagulant etc. (Seredyńska-Sobecka et al. 2006), thereby increasing the performance of the BAC filter.

• (f)

  Depth of the filter: The apparent BET surface area of GAC particles tends to decrease faster in the top portion of the DBACF over time, indicating a higher level of adsorption and thus NOM removal in the upper portion of the filter (Moore et al. 2001). Also, bacterial growth varies along the depth of the filter; in UBACF, the highest attached biomass concentration is mainly found in the middle because of the presence of oxidant at the entry preventing microbial growth at the bottom of the filter (Urfer & Huck 2001; Han et al. 2013; Fu et al. 2017) whereas in DBACF the highest biomass concentration is found at the top and decreases along the depth due to nutrient limitation (Servais et al. 1994).

• (g)

  Biomass concentration: The performance of BAC is dependent on the attached biomass concentration, which varies separately in UBACF and DBACF filters. UBACF tend to show better NOM removal performance because of the more diverse microbial environment and more even distribution of species in UBACF than DBACF. This higher NOM removal efficiency may be attributed to the easy wash out of deposited extracellular metabolites in UBACF leading to enhanced biological activity and thus biodegradation of NOM. The characterization of biofilm, i.e. predicting their occurrence and behavior especially in a full-scale drinking water treatment plant, is still not a much studied topic and requires an in-depth study from a research point of view (Gibert et al. 2013a, 2013b).

NOM removal by AOP can lead to mineralization of organic matter but with higher energy requirements and thus cost inputs. Oxidation is generally achieved at higher doses, as lower doses are proven to be insufficient for DOC reduction (Toor & Mohseni 2007; Sarathy et al. 2011). Therefore, for achieving economic viability of the system, integration of AOP with a biological system like BAC is the best possible alternative. The AOP in conjunction with BAC takes advantage of the partial oxidation products formed by AOP that are further utilized by microbes in BAC as substrate, thereby minimizing the DOC to the best possible concentration. NOM is partially oxidized and HMW compounds are transformed into smaller and more biodegradable compounds such as carboxylic acids and aldehydes (which are byproducts in the case of ozonation). Hydrogen peroxide is the most commonly used oxidant in AOPs, which remains unchanged in treated water and needs to be removed; BAC aids in that also. Sarathy et al. (2011) showed that raw water spiked with 10–12 mg/L of H₂O₂ after passing through the BAC column for 10 days (EBCT = 4.2 min) showed a 93% reduction in H₂O₂ and a 100% reduction with EBCT of 20 min. This H₂O₂ degradation in BAC can be attributed to the presence of catalase produced by bacterial species to protect themselves from external H₂O₂. A few studies that utilized AOP in conjunction with BAC for DOC removal are shown in Table 13.

Growing water demands combating heightening emerging contaminants in the water matrices calls for new advancements in the drinking water treatment sector. One such contaminant is DBP, especially THMs, which are formed upon reaction of chlorine (the most commonly used disinfectant) with NOM. THMs came into the limelight in 1970 and within the space of a year became a significant public health parameter in the USA with its first disinfection byproduct rule. Stringent guidelines are available across the globe for THMs but mostly for developed nations. The cognizance of THMs in developing nations like India is still lacking. THMs have proven to have a carcinogenic nature and thus need major focus from a public health point of view. For targeting THMs generally instead of targeting them directly, their precursors, i.e. NOM, are targeted as they are the root cause of other water quality issues in drinking water treatment industries. NOMs occur ubiquitously in surface water regimes and are site-specific too; their complexity in terms of their chemical nature makes it more difficult to treat them, especially from a THM point of view. The conventional treatment processes like coagulation, flocculation, sedimentation, filtration etc. are not able to remove NOM, especially in terms of its THMFP, and thus require a more advanced form of treatment. AOPs are promising a technology that can completely mineralize NOM but its high performance efficiency is compensated by its high cost (higher electrical energy per order), therefore there is a need to rely on a more techno-economically feasible option. AOP in conjunction with biofiltration or BAC is one such viable option. AOP with lower energy inputs can partially oxidize NOM into simpler or more biodegradable products which can be further removed by BAC column with no or minimum cost inputs, thereby making it overall a more techno-economically feasible way of treating NOM.

This work was financially supported by NBCC, India and FIG grant, Indian Institute of Technology, Roorkee, Uttrakhand, India.