Reduction of microbial risk associated with greywater by disinfection processes for irrigation

Greywater is defined as untreated household wastewater generated from showers, bathtubs, hand basins, floor wastes, laundry tubs and washing machines. In some definitions, the wastewaters generated from kitchens and dishwashers, in homes and restaurants, are included within the greywater term, while others have prohibited the inclusion of kitchen wastewater with greywater due to its lower quality and classified it with the black water from toilets (WHO 2006; Schäfer et al. 2006; Al-Mashaqbeh et al. 2012). Maimon et al. (2014) demonstrated that when kitchen wastewater is included with greywater, higher concentrations of Escherichia coli are detected, while Friedler (2004) reported that kitchen wastewater does not typically have significant faecal contamination. Indeed, high counts of E. coli in greywater, including kitchen water, might be due to the available organic matter and nutrients necessary for bacterial regrowth (Winward et al. 2008c; Zapater et al. 2011; Maimon et al. 2014).

Nevertheless, the greywater generated from kitchens and restaurants might have microbial loads, which result from the washing processes for vegetables and meat. These wastewaters are more relevant to the greywater than to the sewage due to their lower microbial loads and chemical content, such as nitrogen levels, which are 90% less than those in sewage (Jong et al. 2010; Oron et al. 2014). In general, although there is no consensus about the classification of kitchen wastewater as greywater or black water, in this review, the term greywater will include all non-toilet wastewater resulting from households and restaurants.

Greywater represents an important alternative source of water in arid and semi-arid countries, especially those that experience water scarcity (Oron et al. 2014; Maimon et al. 2014). The quality of greywater is lower than potable water, but it is of higher quality than sewage (Prathapar et al. 2005; Jamrah et al. 2006). Greywater contains different pathogens, such as bacteria, viruses and parasites. Bacteria are the most common among these pathogens and represent the biggest challenge because they have the ability to multiply without the need for an intermediate host. Bacteria have the ability to survive for a long time under stressful environmental conditions, and have the potential to regrow after the disinfection processes for greywater. Therefore, improvement in the collection and treatment of greywater is very important for public and environmental health (Katukiza et al. 2014a; Bani-Melhem et al. 2015). Today, a wide range of greywater treatment technologies have been investigated to restore and maintain the physical, chemical and biological integrity of greywater pollution (Wurochekke et al. 2014c). Some of those technologies have exhibited an efficiency for the reduction of pathogenic bacteria and are being used on land for treatment processes, while other technologies are still under investigation. Moreover, the efficacy of the treatment technology to be applied for treating greywater depends on the ability of this technology to meet the requirements of the standard limits. The main criterion for the reuse of greywater in agriculture is pathogenic bacteria. In this review, pathogenic bacteria in the greywater, the treatment technologies used for disinfection processes and the potential of greywater as an alternative resource for irrigation are discussed based on microbiological concepts.

Although greywater does not contain faeces, unlike sewage, the concentrations of pathogens are not negligible. Greywater contains several opportunistic pathogens, such as Staphylococcus aureus and Pseudomonas aeruginosa, in addition to enteric organisms (Ottoson & Stenstrom 2003a; Gross et al. 2007; Winward et al. 2008a). In this section, the bacterial loads of greywater are reviewed to understand the health risk associated with greywater to humans and the environment. The concentrations of pathogenic bacteria in greywater have been reported to be in the range of 5.4–8.9 log₁₀ CFU/100 mL for total coliforms (TC), 2.8–7 log₁₀ CFU/100 mL for E. coli, 2.4–4 log₁₀ CFU/100 mL for enterococci, 3.2–4.4 log₁₀ CFU/100 mL for P. aeruginosa. S. aureus ranged from 3.4 to 5 log₁₀ CFU/100 mL, while Salmonella spp. average 4 log₁₀ CFU/100 mL (Casanova et al. 2001; Winward et al. 2008c; Santasmasas et al. 2013; Katukiza et al. 2014b; Bani-Melhem et al. 2015).

Casanova et al. (2001) investigated the microbial quality of greywater from sinks, showers, washing machines and dishwashers over a seven-month period. They revealed that greywater contained a higher concentration of TC, faecal coliforms (FC), enterococci and P. aeruginosa than the standard limits recommended by the WHO guidelines (WHO 1989). Ottoson & Stenstrom (2003a) studied the prevalence of faecal contamination of greywater and associated microbial risks. The study found high concentrations of faecal indicator bacteria (coliforms, E. coli and enterococci) in greywater. The results of quantitative microbial risk assessment (QMRA) that was performed on Salmonella typhimurium and Campylobacter jejuni indicated that the growth conditions for Salmonella spp. in greywater sediments increased the probability of infection up to 1,000-fold. Recently, the QMRA wiki page, which is a community portal for current quantitative information and knowledge developed for the QMRA field, was set up as a collated database/resource relating to QMRA (2015). It contains hazard infectious doses (ID), dose response, exposure assessment, risk characteristics, and risk management of pathogens. QMRA has been used by several authors for estimating the risk infection of target pathogens in drinking water (Hunter et al. 2003, 2009). Petterson & Stenström (2015) have used QMRA for quantifying the log₁₀ reduction across a free chlorine disinfection drinking water. The study demonstrated the importance of accounting for variable residence time in QMRA, where the log₁₀ reduction of pathogens appeared high and small parcels of water with short residence time can compromise the overall performance of the disinfection system.

Birks et al. (2004) compared the microbial loads of raw greywater, rainwater and groundwater. The study noted that greywater exhibited a higher faecal bacterial load than the rainwater and groundwater, which may be due to the nutrient content in greywater. E. coli, FC, S. aureus and S. typhimurium were detected in washing, kitchen, cooking, bathroom and shower wastewater generated from apartments located at Gyeonggi-do in Korea (Jong et al. 2010). The concentrations of E. coli (4 × 10³ CFU/100 mL) were more than the standard limits recommended by the Enforcement Decree and Regulation of the Water Supply and Waterworks Installation Act in Korea. Some pathogenic bacteria survived in the greywater even after the treatment processes – P. aeruginosa and S. aureus were not affected by the greywater treatment process and remained with concentrations ranging from 1 to 3.9 × 10³ and 2.4 × 10⁴ CFU/100 mL, respectively (Maimon et al. 2014).

The hazard risks associated with the presence of pathogenic bacteria depend on the ID required to cause disease. Therefore, the bacterial contents of greywater are not necessarily an indication that the pathogen will cause infection to humans or animals. S. aureus and P. aeruginosa have the potential to cause skin infection and intestinal colonization if their concentrations exceed 10⁵ and 10⁶ CFU/100 mL (Maimon et al. 2014).

The ID for Shigella sp. is 10–200 cells, 10⁵ cells for E. coli. For non-typhoidal salmonellosis, the ID is approximately 10³ cells, while it is 10⁵ cells for enteric fever by ingestion (Kothary & Babu 2001; Ryan & Ray 2004). Again, the ID of the bacteria is not the main factor that detects the pathogenicity of the bacteria. Moreover, the ability of bacteria to multiply and grow in greywater may increase their number such that it reaches or exceeds an infectious dose. The potential of bacteria to reproduce in greywater relies on the available nutrients and the environmental factors, such as pH, temperature and presence or absence of competition processes with endogenous microorganisms. Therefore, Shigella sp., which has a low ID, cannot persist for a long time in the environment. Conversely, E. coli and Salmonella spp., with a high ID, can increase in the environment due to the ability of both bacteria to grow in greywater as a secondary habitat (Gordon et al. 2002).

However, the health risks associated with pathogenic bacteria in greywater are not limited to their abundance and/or concentrations, but also in terms of the quality of their pathogenicity, such as antibiotic resistance. Greywater contains many antibiotic-resistant bacteria. Nuñez et al. (2012) examined the prevalence of antibiotic-resistant bacteria in raw greywater obtained from a channel located in the area of Ingeniero Budge, Buenos Aires Province. They showed that coliform bacteria resisted ampicillin by 34% and cephalothin by 17%. About 38% of enterococci exhibited resistance to vancomycin, while E. coli, Enterococcus faecalis and Stenotrophomonas maltophilia exhibited multi-resistance to antibiotics. Al-Gheethi et al. (2015a) studied the prevalence of antimicrobial resistance among E. coli, Klebsiella pneumoniae, E. faecalis and S. aureus in greywater generated from resultants at different locations in Penang, Malaysia. They showed that these pathogens exhibited multi-resistance to the investigated antibiotics (ampicillin, amoxicillin, ciprofloxacin, tetracycline and erythromycin). The presence of antimicrobial resistance among the bacterial population in greywater exacerbates the distribution of antibiotic resistance among bacterial communities. This is due to the horizontal transmission of resistant plasmid genes between bacteria. Therefore, greywater is considered to be a reservoir of antibiotic resistant bacteria, which increases the potential health risk to humans.

Regarding the health risk associated with parasites, Cryptosporidium parvum and Giardia lamblia, as well as helminth ova, are the important opportunistic intestinal pathogens of humans (Rangel-Martıneza et al. 2015). The main concerns about these pathogens lie in their ability to form cysts which can survive for a long time in the environment followed by transmission to humans via contaminated drinking water and vegetables irrigated with contaminated greywater. It has been reported previously that C. parvum and G. lamblia are more predominant in either warmer or wetter seasons. However, Rangel-Martıneza et al. (2015) found no association between dry-cold or dry-warm periods and the presence or absence of these parasites. Further, the climate changes during different seasons might affect the number of parasites available; for example, C. parvum is more prevalent than G. lamblia (Kurniawan et al. 2009; Ajonina et al. 2012; Julio et al. 2012). Cryptosporidium spp. are capable of completing a life cycle within the same host and have the ability to cause reinfection to the host. Many reports in England, the USA and Japan have revealed the presence of correlation between the disease transmitted by polluted water and the presence of C. parvum, due to the potential of this parasite to resist chlorine disinfection. Cryptosporidiosis is more prevalent among immune-depressed people (Snelling et al. 2007).

Helminth parasites, which demonstrate a significant health risk in reused waters, include Ascaris lumbricoides and Ancylostoma duodenale. These helminths have a simple life cycle with no intermediate hosts, and are capable of causing infection via the faecal–oral route as well as having a high potential to survive and resist environmental stresses (Toze 2006). Ascariasis is the most common helminthiasis which infects children under 15 years of age (Silva et al. 1997; Jiménez 2003). The infection from parasites and helminth in greywater occurs after the consumption of food or water contaminated with oocysts (Carey et al. 2004). C. parvum, G. lamblia, Entamoeba histolytica, as well as A. lumbricoides are significantly more infectious to humans than most enteric bacterial pathogens because these pathogens have a low infective dose (<10 oocysts) (Rose & Gerba 1991; Dillingham et al. 2002).

Many viruses, such as rotaviruses, adenoviruses, reoviruses, caliciviruses and astroviruses have been found in wastewater. However, enteric viruses are the most common and important viruses. Enteric viruses are obligate intracellular parasites and are specialized to infect humans (Haas et al. 1999; Toze 2006). These viruses are more resistant to treatment processes and environment than bacteria. Therefore, the presence of these viruses in greywater represents an important risk to humans and is rarely a problem for other animals. The main concern about the presence of enteric viruses in greywater is their low dose infectivity (<10 viral particles) and their long-term survival in the environment, besides the limited extent of reduction or inactivation during treatment processes (Asano 2005). As mentioned before, the health hazards associated with greywater may turn into health risks depending on the environmental conditions and types of pathogens.

The biological indicators are microorganisms used as a model to investigate the presence and response of pathogenic bacteria to treatment processes in environmental systems. These indicators have similar characteristics to pathogenic bacteria, although the enumeration and the detection procedure is easier than for pathogens. Biological indicators include specific bacteria, fungi, viruses and parasites. The criteria for the selection of indicator bacteria include their ability to survive in the environment without an increase in their number, and a similar response to environmental factors as in the case of pathogenic bacteria. Therefore, they must be present in the contaminated environment and absent in the non-polluted environment (Cooper & Olivieri 1998; Dumontet et al. 1999; Lepeuple et al. 2004; Bitton 2005; Myers et al. 2007).

Among several bacteria, faecal bacteria (TC, FC and enterococci) are the historical indicators. These bacteria have been used for over 100 years (Ashbolt et al. 2001; US EPA 2007). The correlation between these indicators and pathogenic bacteria such as Salmonella spp. and S. arureus have been reported in the literature (Efstratiou et al. 1998; Efstratiou 2004). However, other reports in the literature revealed that these indicators do not correlate to the pathogens, especially in non-domestic wastewater. In some cases, the faecal indicators increased in number in the environment (Polo et al. 1998; Byappanahalli & Fujioka 1998). According to the US EPA, FC is used to determine whether the water meets the state water quality standards, while E. coli or enterococci are considered to be a health risk from recreational water. Moreover, Enterococcus sp. has been suggested to be useful for indicating the presence of viruses, particularly in sludge, seawater and biosolids. This is because these organisms are relatively easy to enumerate and survive longer than FC in these environments (Bitton 2005; Mote et al. 2012). Efstratiou et al. (2009) suggested that TC or FC alone might provide an adequate indicator for Salmonella presence in seawater. However, a statistically significant improvement demonstrated that enterococci alone might be more useful for the detection of Salmonella presence.

Indeed, the efficiency of the bacteria to be used as a biological indicator depends on the type of waste. Several studies have suggested using S. aureus and P. aeruginosa as indicators in greywater, because they have been detected even in the absence of E. coli (Yoshpe-Purer & Goldman 1987; Mates 1992; Tosti & Volterra 1998; Garland et al. 2000). Kenner & Clark (1974) reported that P. aeruginosa may be a better indicator than FC of pollution in potable, direct reuse, bathing and recreational water. Gross et al. (2007) noted that S. aureus and P. aeruginosa survived in greywater in which E. coli was reduced to non-detectable levels (<1 CFU/100 mL). In medical waste, S. aureus and P. aeruginosa are the biological indicators for testing the efficiency of treatment processes for those wastes (Efaq et al. 2015). With respect to biosolids, the US EPA (2003) demonstrated that Salmonella spp. are bacteria of great concern as well as good representatives of the reduction of other bacterial pathogens because they are typically present in higher densities than other bacterial pathogens and are at least as hardy.

Wheater et al. (1980) reported that the number of P. aeruginosa in freshwater streams reflected the level of domestic pollution. Coronel-Olivares et al. (2011) suggested the use of P. aeruginosa as a new indicator for determining the quality of treated sewage effluent. This was due to its importance as an opportunistic pathogen. According to Warrington (2001), P. aeruginosa is the best indicator for recreational use, especially in marine water, where E. coli does not survive as well. Although E. coli is a good indicator in sewage-contaminated water supplies and chlorinated water high in organics, this bacterium is not as useful in water with high organic waste from vegetable processing. S. aureus has been proposed as being a useful indicator in hospital hygiene for microbiological standards (Dancer 2004). Crone & Tee (1974) found that Staphylococcus is more frequent than FC in swimming pools and considered staphylococci as an indicator for human contamination. Jin et al. (2013) used S. aureus as an indicator to evaluate the hydrothermal treatment process in terms of achievement of the hygienic safety of food waste.

It can be concluded that the selection of indicator bacteria depends on the type of wastewater. However, P. aeruginosa and S. aureus are the most appropriate for greywater from laundry and shower resources while E. coli and E. faecalis are the best indicators for the presence of faecal contamination.

Greywater consists of all non-toilet wastewater that contains oils, fats, detergents, soaps, nutrients, salts and particles of hair, food and lint as well as different types of microorganism. Greywater can be reused and recycled for several applications; it has been recycled as a source of biodegrading bacteria and a medium for the production of enzymes. Alrumman et al. (2014) used potato wastewater as a simple and cheap medium for the production of α-amylase by Bacillus axarquiensis compared to starch broth medium. They found that the amount of α-amylase in potato wastewater was less than in the starch medium by 13.8%. However, after the simple addition of starch, nitrogen, phosphate and calcium into the potato wastewater medium, the production of α-amylase increased four times more than production in the starch broth medium. Phong et al. (2014) isolated 102 bacterial isolates from the wastewater of food processing plants and restaurants in Can Tho city, Vietnam. Among these bacteria, 11 bacterial isolates exhibited good lipase production. Acinetobacteria soli strain has been proposed for wastewater treatment because of its high ability for lipid degradation.

Moreover, in arid and semi-arid countries, greywater is used extensively for irrigation. This is due to severe water scarcity, rainfall fluctuation and the rise in water pollution (Bani-Melhem et al. 2015). The reuse of greywater for irrigation in those countries was intended to recover nutrients and for artificial aquifer recharge, as well as limit the resupply costs (Garland et al. 2000; Abu Ghunmi et al. 2008; Dalahmeh et al. 2011). Greywater is rich in nutrients that might be used for high plant growth. Alongside its benefits, greywater contents, such as soap and detergents, might have a negative effect on plants. The primary health hazard concern associated with the use of greywater is excreta-related pathogens. The presence of faecal pathogen bacteria in greywater, such as FC, E. coli, S. enterica, S. aureus and P. aeruginosa, might represent a high risk for humans and animals (Dixon et al. 1999; Chen et al. 2013; Benami et al. 2013; Maimon et al. 2014). The improper reuse of greywater for garden irrigation may pose a major hazard to the environment in the long term (Anda et al. 2010; Mohamed et al. 2012, 2013, 2014a, 2014b).

According to Al-Hamaiedeh & Bino (2010), the criteria required to reuse greywater include hygienic safety, aesthetics, environmental tolerance, and technical and economic feasibility. In this section, the potential reuse of greywater for irrigation will be discussed based on hygienic safety.

Garland et al. (2000) studied the effect of Igepon TC-42 (sodium N-coconut acid-N-methyl taurate, a common ingredient of soaps and detergents) on wheat and lettuce plants as well as the survival of pathogenic bacteria. They investigated three different modes: 875 mg m⁻² on growing area once a day, continuous addition of the same concentration d⁻¹ and the variable addition of 0 ± 3,000 mg m⁻² d⁻¹. They noted that the growth rate of wheat was not affected, while the lettuce yield significantly reduced with the pulse and continuous treatments. Igepon was degraded rapidly by microorganisms within 2 days, with a half-life of <1 hour. The maximum survival was recorded for P. aeruginosa, which persisted in the environment surrounding the wheat for 70 days and the lettuce for 28 days. E. coli and S. aureus were reduced to below the detection limits within 35 days. According to this study, the detergent might not pose any potential growth risk to certain plants. However, the survival of some pathogenic bacteria on the irrigated plants would represent an important source and cause severe disease for plants, animals and humans.

It has been reported previously that bacteria can survive in stressful conditions by entering a viable but nonculturable (VBNC) state (Winfield & Groisman 2003). Salmonella spp. have high survival rates in aquatic environments (Chao et al. 1987), and Salmonella spp. spread with manure have survived for >10 months in soil (Jones 1986; Baloda et al. 2001). The contamination phenomenon of poultry farms has been revealed due to the ability of Salmonella spp. to persist for >1 year even with the use of disinfectants (Davies & Wray 1995). Jacobsen & Bech (2012) reported that Salmonella spp. could survive in soil for prolonged periods and could be transported to freshwater resources in high numbers.

The survival of pathogenic bacteria in greywater used for irrigation depends on the type of microorganism, source of pathogens (humans, plants), irrigated plants, frequency of irrigation, characteristics of the irrigated soil, temperature, pH and competition with protozoa (Guan & Holley 2003; García et al. 2010). Franz et al. (2011) reported very important findings, in that they found that E. coli strains isolated from humans survived longer than those isolated from animals (on average 211 vs. 70 days). Wang et al. (2014) investigated the survival time of E. coli O157:H7 in 14 soils collected in eastern China from the warm-temperate and subtropical zone. They showed that the E. coli O157:H7 survival times were between 1.4 and 25.8 days. E. coli O157:H7 has survived in neutral or alkaline soils and warm-temperate zones for longer than in acidic soils and subtropical zones. The survival rate also depends on the soil carbon and total nitrogen, with the minimum survival rate being recorded in the presence of amorphous Al₂O₃ and the relative abundance of Chloroflexi.

The environmental risks associated with greywater reuse are represented in the adverse effects on plant growth due to microbial contamination. However, Benami et al. (2013) compared the presence and abundance of E. coli, K. pneumoniae, S. enterica, P. aeruginosa, E. faecalis and Shigella spp. in soil irrigated with treated greywater and that irrigated with freshwater (control). The study was conducted for 6 months and revealed that the abundance of these bacteria in the greywater and freshwater irrigated soils did not differ. Based on these results, they suggested that greywater irrigation has no effect on the diversity and abundance of the tested pathogens and indicators in the soil. In addition, Winward et al. (2008c) stated that no incidences of illness have been linked to greywater reuse and the health risks might appear to be low compared to sewage. It depends on the quality of the greywater, disinfection practices and management practices. The occurrence of faecal indicator bacteria in greywater, as reported by Casanova et al. (2001), Ottoson & Stenstrom (2003a) and Friedler (2004), has demonstrated the potential presence of faecally transmitted pathogens. Moreover, the WHO (2006) has reported that the disease outbreaks caused by Vibrio cholerae, Salmonella spp. and Shigella sp. are associated with the reuse of wastewater, excreta or greywater for irrigation of vegetables. Therefore, pathogenic bacteria associated with greywater still present a health risk even in the absence of reports in the literature; this point should be taken into consideration. The potential of pathogenic bacteria in greywater to cause human disease is discussed below.

The absence of evidence for the diseases caused by pathogenic bacteria, which have been transferred by vegetables irrigated with contaminated greywater, is due to the absence of accurate techniques that could detect the bacterial infection source. In the infections caused by parasites, it is easy to detect the source of the infection because parasites have a clear life cycle (reservoir, intermediate and vector host). In the bacterial infection as a typhoid disease, it is difficult to detect how S. typhi is transferred to humans – whether it is from drinking contaminated water or from contaminated food – because Salmonella spp. are able to grow and multiply in the environment (Lopez-Torres et al. 1987; Jimenez et al. 1989; Brettar & Hofle 1992; Winfield & Eduardo 2003). In addition, Salmonella spp. do not have a specific reservoir and are capable of colonizing a wide variety of microorganisms (Foltz 1969). This feature supports the Salmonella life cycle, as Salmonella spp. have the ability to pass through different hosts into the environment and back into a new host (Thomason et al. 1977). Some authors reported that bacteriophages have the ability to identify and distinguish different strains of bacterial species isolated from different origins (disease, food, water, environment) or geographical locations (Schofield et al. 2012). Wangkahad et al. (2015) have used bacteriophages to detect the source of pathogen contamination. The study investigated the association between bacteriophages of Aeromonas caviae, Enterobacter sp. and K. pneumoniae in water and the contamination sources (human vs. animals) in Thailand. The study found that bacteriophages were detected in polluted samples from human faecal sources but not in non-polluted samples. The presence of A. caviae was associated with human faecal sources, whereas Enterobacter sp. and K. pneumoniae were prevalent in human and animal faecal sources.

In general, the concerns in terms of health risk related to the reuse of greywater in irrigation lie in the potential of these pathogens for regrowth or persistence and transference into the food chain (Rose et al. 1991). For instance, the ability of P. aeruginosa to regrow in greywater makes it an opportunistic pathogen of concern for greywater reuse (Winward et al. 2008a). Hence, the strategy of greywater reuse for irrigation purposes is subject to strict regulation in developed countries (Table 1). In contrast, although developing countries also have strict regulations for the reuse of greywater in agriculture, these countries lack the developed techniques and the power to enforce such regulations. Therefore, surface irrigation, which is not permissible in developed countries, is the most common use in developing countries.

In comparison, black water with high concentrations of pathogenic bacteria would appear to have a higher risk to humans than greywater. It has been demonstrated that the irrigation of fresh vegetables by black water represents the main cause of diarrhoeal disease due to the high microbial load in the vegetables even after the washing process (Ronner & Wong 1993). The movement of pathogenic bacteria through the soil and the contamination of groundwater, drinking water, runoff and erosion containing pathogens and contaminating surface water increases the possibility of human infection (Sahlstrom et al. 2004).

In order to ensure the safe reuse of greywater, these wastes should meet certain limits. In some countries, there is a quality safety plan and alternative systems of treatment of greywater (Wurochekke et al. 2014c). Currently, many countries have regulated the microbiological standards for greywater reuse (Table 2). Moreover, most countries focus on the health or environmental risk from pathogenic bacteria, and very little is mentioned regarding the potential presence of heavy metals or disinfection by-products (DBPs) and pharmaceutically active compounds (Toze 2006). This differs from black water or biosolids where the standards focus on pathogenic bacteria as well as viruses, parasites and helminths. As well, the standards for black water include heavy metals and toxic substances. These differences are due to the nature of black water, which is more complicated than greywater. Therefore, the discussion of standards presented here focuses only on the presence of pathogenic bacteria as a criteria for the quality of greywater. These standards range from the stringent guidelines of the US EPA and Korea, where FC should be non-detectable in water intended for urban reuse, to the less demanding guidelines of Germany and the WHO (<1,000 CFU/100 mL). In the USA, California requires that FC should not exceed 2.2 CFU/100 mL as an average over a 7-day period or 23 CFU/100 mL in 30 days. In the UK, since 1988, Defra and its predecessors have described in detail the microbiological standards of different types of wastewater. In Jordan, the standards for greywater are divided into two categories based on their utilization. In Category A, where the greywater is used for the irrigation of cooked vegetables, the FC should be <100 MPN/100 mL, while in category B, where the greywater is used for the irrigation of tree crops, the FC should be <1,000 MPN/100 mL.

The disinfection technologies aim to eliminate the microbial loads of greywater that might pose a potential risk for humans and plants, and thus provide safe and aesthetically acceptable greywater that is appropriate for the purpose of irrigation. These technologies include chemical (chlorination and ozonation), physical or mechanical (filtration processes) and radiation disinfection (ultraviolet (UV) irradiation). The degree of disinfection process proposed must take into account the type of reuse and the risk to the population of exposure (Matos et al. 2014). In this section, the disinfection techniques for greywater are reviewed and discussed based on the efficiency in eliminating pathogenic bacteria and toxic by-products, and to reuse greywater for agricultural purposes. The advantages and disadvantages of disinfection processes for greywater are illustrated in Table 3.

Chlorination was the most common method for disinfection in the twentieth century because it was cheap and simple (Dunnick & Melnick 1993; Winward et al. 2008a). The efficiency of chlorination to kill pathogenic bacteria in greywater relies on the greywater quality, such as the turbidity, presence of organics and particles that might protect pathogenic bacteria from the action of chlorine (Winward et al. 2008a). Mohamed et al. (2015) suggested double doses of chlorination for water with high turbidity (100 NTU). A high dose of chlorine and the removal of impurities from greywater by filtration improve the efficiency of the chlorination process (Tal et al. 2011). However, the free and combined chlorine residue is toxic; therefore, there is a requirement to de-chlorinate or to remove the chlorine before disposal to the environment. Recently, many developed countries such as the USA have prevented the use of chlorination due to its toxic by-products and the occurrence of resistance and regrowth among pathogenic bacteria (Morris et al. 1992; Nieuwen-Huijsen et al. 2000). The reason for the occurrence of bacterial regrowth in chlorinated greywater might be due to its highly reactive characteristics, which accelerate the chlorine decay process (Tal et al. 2011).

Most DBP of chlorination are carcinogenic; one of them is nitrosodimethylamine, which has been reported to be a probable human carcinogen (Pehlivanoglu-Mantas et al. 2006). Cantor et al. (1987) reported the first indication of the association between bladder cancer risk and the consumption of chlorinated water. The bacterial resistance of chlorination has been reported since the 1980s. Ridgway & Olson (1982) indicated that bacteria isolated from a chlorinated system have more resistance to the combined and free forms of chlorine than those isolated from an un-chlorinated system. Recently, some authors have revealed several pathogenic bacteria that are resistant to chlorination (Dungeni et al. 2010; Cherchi & Gu 2011). The formation of DBPs such as total trihalomethane and haloacetic acid were also reported in the disinfection process of water by titanium dioxide photocatalysis.

The problem of bacterial resistance to chlorine has been extended to include resistance to antibiotics. The authors reported that there is a correlation between the resistance to chlorine and antibiotics, as bacterial cells exposed to chlorine may acquire resistance to antibiotics (Murray et al. 1984; Shi et al. 2013). In greywater, the ability of pathogenic bacteria to resist chlorine is due to the bacteria having been in contact with high concentrations of chlorine during the washing processes, thus, the efficiency of chlorination to eliminate these pathogens would be weak.

In order to increase the efficiency of chlorination, Santasmasas et al. (2013) suggested that the disinfection of greywater by chlorination might be more effective if the greywater were treated using a biological oxidation process, as this would lead to degradation of organic matter and thus reduce the toxic by-products. In their study, the authors treated greywater in four stages: a screening process, in which the debris was separated; then biological oxidation to decompose the organics; a filtration process using ultrafiltration membrane technology; finally, the greywater generated was disinfected with chlorine. They showed that chlorination reduced the TC and FC by 6 log₁₀ CFU/100 mL and that the disinfected greywater met the standard limits recommended for reuse in irrigation.

Benami et al. (2015) monitored the presence of FC, E. coli, S. enterica, S. aureus and P. aeruginosa during the greywater treatment and disinfection process (raw, biologically treated, and disinfection by chlorine). The evaluation study was conducted bi-monthly over the course of a year. The samples were analysed using both culture and molecular methods based on DNA-based quantitative polymerase chain reaction (qPCR). The culture-based method revealed that the pathogenic bacteria were inactivated by chlorine. However, the DNA-based analyses indicated that there was no significant difference between the pathogenic bacteria in the raw and disinfected greywater. To understand the discrepancies between the results, the authors repeated the analysis of E. faecalis as a model bacterium, and obtained similar results. Based on these results, the authors concluded that the inactivation efficiency cannot be estimated by DNA-based qPCR. In fact, the explanation of these results is the viable but nonculturable (VBNC) state of the inactivated bacteria. It has been reported that under unfavourable conditions, bacteria may transfer into a dormant state called VBNC, which means that there are slow metabolic processes and no cell division (Oliver 2005); in such a case, the culture-based technique is not able to detect the bacterial cells. Therefore, the molecular-based technique will be more accurate to detect the presence or absence of VBNC bacterial cells (Chen et al. 2012). However, some authors reported that this technique might give false positive results. For this reason, a decade ago the molecular-based technique using qPCR was developed for greater accuracy. The development included using propidium monoazide or ethidium monoazide dyes (Fijalkowski et al. 2014). In addition, new technologies have been introduced to detect the viability of bacterial cells, such as the detection of metabolic activity, the positive energy status, responsiveness and the reverse transcription quantitative PCR assay (Fittipaldi et al. 2012; Jiang et al. 2013).

Helmi et al. (2015) have investigated techniques to detect the viability of a bacterial population in chlorinated and non-chlorinated water. The techniques used included flow cytometry (FCM), solid-phase cytometry (SPC), epifluorescence microscopy (MSP) and culture method performances. The total bacterial counts were quantified by using 4′,6′-diamino-2-phenylindole (DAPI) (for MSP) and SYBR Green II (for FCM), and the viable and non-viable bacterial cells were distinguished by using SYBR Green II and propidium iodide dual staining (for FCM). In FCM and SPC, the Chemchrome V6 was used to distinguish the active bacterial cells, while 5-cyano-2,3-ditolyl tetrazolium chloride (CTC) was used in the MSP method. The study recorded significant correlation between total bacteria and active cells as determined by microscopy and FCM. In contrast, the correlation between bacterial counts and active cells was not significant, as determined using solid-phase and FCM. Moreover, the culture-based method using R2A medium showed bacterial recovery after chlorination. Therefore, FCM might be a useful and powerful technique for drinking water production monitoring.

In general, chlorination has become an unacceptable process for disinfection of water due to the toxic by-products, bacterial resistance and bacterial regrowth after the disinfection process. Therefore, the alternative technology used is ozonation. However, the responses of microorganisms to ozonation have been reported to be similar to those of chlorination (Xu et al. 2002; Vital et al. 2008, 2010).

A mechanical disinfection process by filtration would be more efficient to remove pathogenic bacteria but not for the inactivation of bacterial cells, as the cells removed by the filtration systems are still active. This technique could not increase the microbial resistance to antibiotics or produce toxic by-products. Numerous systems of membrane filtration have been developed for the removal of pathogenic bacteria, the efficiency of which depends on the pore size of the membrane filter. However, the physiochemical characteristics of greywater significantly affect the filtration process.

Winward et al. (2008c) evaluated the efficiency of three configurations of constructed wetland: horizontal flow reed bed, vertical flow reed bed (VFRB) and a green roof water recycling system (GROW). A membrane bioreactor (MBR) and a membrane chemical reactor (MCR) were used for the removal of heterotrophic bacteria, TC, E. coli, enterococci, sulphite reducing Clostridia, S. aureus and P. aeruginosa from the greywater. The MBR was conducted in continuous flow and under aerobic conditions to provide a mixing of the biomass. The MCR consisted of an advanced oxidation process using UV light and titanium dioxide in a closed reactor, and membrane filtration was used for the disinfection process. They reported that the three constructed wetlands met the German reuse standard for TC and P. aeruginosa (4.0 and 2.0 log₁₀ CFU/100 mL, respectively) but not the California State standards (2.2 log₁₀ CFU/100 mL). Among the constructed wetlands, VFRB provided the best overall removal of faecal indicators, in which 44% of the treated samples met the California State standards. MBR provided the highest quality treated greywater and had the most robust treatment efficiency. This system was not affected by an increase in the influent greywater strength.

Jong et al. (2010) investigated the efficiency of an MBR consisting of anaerobic-anoxic-oxic reactors, and a membrane filter (pore size 0.45 μm) to remove E. coli, S. aureus and S. typhimurium from greywater. The maximum removal of these bacteria after the anaerobic-anoxic-oxic reactors was 20%, 5.5% and 5.6%, respectively. After the membrane filter, E. coli reduced by 67.5 compared to 27.7 and 20.4% for S. aureus and S. typhimurium, respectively. However, the E. coli count in the treated effluent was still more than the standard limit recommended for irrigation reuse (1.3 × 10³ CFU/100 mL).

Katukiza et al. (2014a) studied the treatment of greywater in Uganda using a two-step crushed lava rock filter. The filter unit consisted of two filters in series. The filters (R1 and R2) were composed of 10 cm of crushed gravel under-drain (for the sedimentation process) and graded crushed lava rock (2.56–5 mm for the first filter and 1.18–2.56 mm for the second filter). Crushed lava rock was chosen in the study as it has a higher specific surface area and porosity compared to sand and gravel (Kalibbala et al. 2012; Sekomo et al. 2012). The hypothesis for treatment was based on a two-step design; the primary treatment of greywater was performed by R1, while the secondary treatment was performed by R2. Katukiza and co-workers found that the sedimentation process did not reduce the bacterial loads. The TC, E. coli and Salmonella spp. counts did not differ significantly before and after the sedimentation process. The log reduction was 3.9, 3.8 and 3.2 after R1 and R2 for TC, E. coli and Salmonella spp. respectively. However, the bacterial loads of the treated effluents did not meet the WHO guidelines for greywater. The concentration of E. coli in the treated effluents was 3.7 log₁₀ CFU/100 mL compared to 3 log₁₀ CFU/100 mL required by the WHO; in addition, the Salmonella spp. was 2.73 log₁₀ CFU/100 mL. Based on these results, the designed filter system was insufficient to reduce the pathogenic bacteria in greywater and to meet the international standards for reuse for agricultural purposes.

Molaei (2014) studied four types of filtration system for the removal of Salmonella spp. and E. faecalis from artificial greywater in Sweden. These filters were biochar, bark, activated carbon and a mixture of bark and activated carbon. The filtration systems were designed in column experiments (height 65 cm, diameter 4.3 cm) for 60 days. The biochar filter was the most effective in removing Salmonella spp. (3 log reductions). The efficiency of the bark and filter mixture ranged from 1 to 2 log reductions. The activated carbon filters exhibited the highest efficiency for removing Salmonella spp. and E. faecalis (7 vs. 5 logs, respectively). The high removal of active carbon appeared to be efficient for the disinfection of greywater. However, this efficiency was recorded during the first week and decreased significantly after 15 days to 1 log reduction. After 30 days of the filtration process, the concentration of Salmonella spp. increased to >3 log₁₀ CFU/100 mL. For all the filter types investigated, the efficiency percentage reduced after 1 week of the disinfection process. In order to achieve greater removal of pathogenic bacteria, the filter should be replaced every week at least, which means that the current disinfection process would be more expensive to apply to land.

Bani-Melhem et al. (2015) evaluated a submerged membrane bioreactor (SMBR) system (pore size of 0.04 mm) for removing TC and FC from greywater in Jordan. The SMBR was operated for 42 days at constant transmembrane pressure (13 kPa). They noted that 99.99% of the TC and FC concentration was removed and that the greywater met the standard limits recommended for reuse in irrigation.

According to the above literature, the disinfection of greywater using membrane filtration has some efficiency for the removal of pathogenic bacteria. However, these technologies have many disadvantages. Membrane filtration is less effective in highly polluted greywater and requires constant maintenance. This technology is not able to remove dissolved organic matter, and does not cause any physical or chemical damage to bacterial cells. Hence, the remaining bacteria in the disinfected greywater might grow and multiply again.

The disinfection process, which does not lead to complete damage of the bacterial cells to prevent regrowth after the disinfection process, would be inadequate. For instance, although the disinfection process by UV irradiation is safer in terms of toxic by-products, the ability of bacteria to resist UV disinfection and regrow after the disinfection process has been recognized (Alonso et al. 2004; Munir et al. 2010; Ting et al. 2011; Wang et al. 2012). Huang et al. (2013) reported that the population of antibiotic-resistant bacteria in water increases when exposed to low doses of UV light or chlorine. Therefore, the selection of the disinfection process for greywater should be in terms of being free from toxic by-products and having the ability to completely destroy bacterial cells.

The potential for bacteria to regrow depends on the inactivation mechanism caused by the disinfection process. An understanding of the factors affecting bacterial growth would help in the selection of an appropriate process that has the ability to cause complete damage to the bacterial cells. Among the many factors that affect bacterial survival and growth in the environment, some are chemical, such as pH, water activity, moisture content, nutrients and disinfectants, while others are physical, such as temperature, pressure and irradiation (Johnson et al. 2005; Hungaro et al. 2014).

Among the various factors, temperature is the most critical in terms of the survival of the bacteria (Tang 2009). Hence, temperature is a basic consideration in the development of several disinfection technologies, such as pasteurization, which has been used for over 200 years. Most pathogenic bacteria are mesophilic, with an optimal temperature of 37 °C. Therefore, thermal treatment above the optimal temperature would lead to denaturation of the proteins, enzymes and nucleic acids of bacterial cells. However, the thermal disinfection of greywater has become increasingly unacceptable in recent years due to the high energy and increase in the Earth's temperature.

Thermal disinfection at moderate temperature (45 °C) has been reported as being effective for the reduction of E. coli. Wastewater treated in such a manner has met the WHO guidelines (<3 log₁₀ CFU/100 mL) after 48 hours and US EPA (<14 CFU/100 mL) after 1 week of treatment (45 °C) (Al-Gheethi et al. 2013a). These results indicate that a moderate temperature is effective as a disinfection process for wastewater. However, the treatment periods required to achieve the standards are lengthy. Thus, this process might be inappropriate for a high volume of greywater or even for a private house because the greywater is generated daily, whereas this process requires 1 week to meet the greywater standards for reuse. Therefore, Al-Gheethi et al. (2013b) used solar disinfection (SODIS) for the elimination of FC, E. faecalis, Salmonella spp. and S. aureus in secondary effluent and lake water. The treatment process was conducted inside transparent polyethylene terephthalate bottles for 1, 2, 3, 4, 5, 6, 7 and 8 hours. It was revealed that the temperature of the samples increased to >45 °C during the treatment process. FC, Salmonella spp. and S.aureus were eliminated from the secondary effluent after 8 hours and after 7 hours from the lake water. No regrowth of these pathogens was observed, as determined by the absence and presence technique using an enrichment medium. E. faecalis reduced to less than the detection limits in the lake water but was still detectable in the secondary effluent. In order to eliminate E. faecalis, Al-Gheethi et al. (2013b) stored the treated samples inside closed PPT bottles at room temperature for 2 weeks; the bacteria were eliminated during the storage period. This was because E. faecalis is aerobic and died off in the closed bottles due to the absence of oxygen and competition from indigenous microorganisms.

SODIS is a technique used extensively for drinking water; this technique improves the microbiological quality of drinking water due to a combination of the temperature and UV, which effectively destroys pathogenic bacteria (WHO 2002; Gomez-Couso et al. 2009). In the study conducted by Al-Gheethi et al. (2013b), the authors combined three factors: temperature, UV and anaerobic condition. This treatment exhibited efficiency in the reduction of faecal indicators and the elimination of pathogenic bacteria.

Pansonato et al. (2011) demonstrated the efficiency of SODIS in the inactivation of pathogenic bacteria in greywater. They investigated SODIS for the inactivation of E. coli in greywater. The samples were exposed to an average solar radiation intensity of 518 W/m². The results indicated that the SODIS system has the potential for TC and E. coli inactivation in pre-treated greywater by >2 log₁₀.

Currently, the US EPA (2015) uses the SODIS method as a model for the design of residential systems for the sanitation of greywater for the irrigation of lawns and gardens. This project started in August 2014. It is expected that a system will be invented that can increase the efficiency of the solar radiation and heat for the inactivation of microbes.

Gross et al. (2007) studied the recycled vertical flow bioreactor system for removing pathogenic bacteria (E. coli, S. aureus and P. aeruginosa) from greywater generated from the kitchen. In this system, the greywater was subject to preliminary sedimentation and then pumped into plastic containers consisting of three layers: the upper container contained organic soil with lime pebbles and heterotrophic bacteria as decomposing organisms; the lower container was used as a lower ‘reservoir container’; and the middle layer consisted of plastic filter media. The greywater generated from the reservoir container was pumped again into the upper container. The treatment process was conducted for 2–3 days, and no disinfection process was performed in this treatment. Samples from the input and output points were collected and subjected to microbial analysis, which revealed that at the end of the treatment process, E. coli was reduced to less than the detection limits (<1 CFU/100 mL). However, the concentrations of P. aeruginosa did not change in the samples collected before and after the treatment process. S. aureus declined in the treated effluent but still averaged 100 CFU/100 mL. In this study, although no disinfection process was implemented, E. coli was completely removed. However, S. aureus and P. aeruginosa survived during the treatment process. This might be because the kitchen greywater represents a secondary environment for E. coli, while S. aureus and P. aeruginosa are indigenous bacteria in these wastes. Thus, the survival of these bacteria was greater than for E. coli. In addition, the biodegradation of organic matter by decomposing bacteria during the treatment process would prevent the regrowth of E. coli. Friedler (2004) reported that kitchen wastewater does not contain faecal contamination. Therefore, such waste might only contain a small concentration of E. coli, whereas, in the presence of the organic matter and nutrients in kitchen wastewater, E. coli might regrow to high concentrations (Winward et al. 2008c; Zapater et al. 2011; Maimon et al. 2014).

The disinfection of greywater through the use of plant products that have antimicrobial properties was studied by Winward et al. (2008b). They assessed the antimicrobial properties of essential oils (EOs) from origanum oil (Thymus capitatus) and carvacrol as a disinfection process for greywater in the UK. They found that 468 mg/L of origanum oil reduced TC to less than the detectable levels after 30 min of contact time. The inactivation rate of TC increased with an increase in the contact time. No regrowth of TC was recorded in the greywater over 14 weeks, despite the organic concentrations and particulate size in greywater reducing the efficacy of the disinfection process with origanum EOs. The study opened up a new direction for the disinfection of greywater. EOs are aromatic compounds extracted from plants (Sangwan et al. 2001). EOs have the ability to inactivate bacterial cells by disruption of the cell membrane, causing leakage of the cell contents and eventual cell lysis (Burt 2004). As yet, toxic by-products generated from the disinfection process using EOs have not been reported. However, in this study, the authors reported that disinfection with origanum EO was limited by organics. Thus, they indicated that disinfection by origanum EO would be more effective for preventing bacterial regrowth at a low concentration if the greywater was treated by a MBR (Jefferson et al. 2001). Otherwise, high concentrations of origanum EO should be used. Finally, the authors concluded that the disinfection of greywater with EOs appears to be impractical due to the land area required to produce sufficient origanum EO and the anticipated costs.

The storage of greywater is an important element in all greywater recycling systems. Clearly, water needs to be stored until its utilization. However, the microbial load of the stored water needs to be evaluated carefully. Rose et al. (1991) reported that coliform bacteria increased to 1 and 2 log₁₀ CFU/100 mL during the storage period over 48 hours. Furthermore, Dixon et al. (1999) indicated that TC increased by 2 log₁₀ CFU/100 mL after 24 hours of storage. In contrast, Gerba et al. (1995) found that TC and FC reduced by 3 logs in stored greywater. Previous studies for Salmonella sp. and Campylobacter sp. seeded into greywater indicated that no regrowth occurs (Rose et al. 1991; Ottoson & Stenstrom 2003b). Al-Gheethi et al. (2013a) noted that the concentrations of E. coli and E. faecalis reduced in the stored effluents to less than the detection limits after 2 weeks of storage at room temperature. The effluent met the standard requirements for irrigation. It has been suggested that the presence of competing microorganisms prevents the regrowth of pathogenic bacteria or induces them to enter into a viable but non-culturable state (Ottoson & Stenstrom 2003b). Moreover, Al-Gheethi et al. (2013a) reported that the storage system for secondary effluents should not be more than 2 weeks due to the increase in antimicrobial resistance among the surviving bacteria. They noted that after 4 weeks of storage, the concentrations of TC and Salmonella spp. reduced. However, the resistance to ampicillin, amoxicillin, cephalexin, cefuroxime and ciprofloxacin increased significantly among these bacteria.

The microbiological qualities of greywater during the storage period depend on the nature of these waters. The potential of pathogenic bacteria to multiply and grow in the treated water is less than that in untreated waters. Nevertheless, this potential will rely on the nature of the treatment processes and storage conditions of the treated waters. The storage conditions for the disinfected greywater affect the ability of bacteria to regrow and multiply. The bacterial cells inactivated using UV irradiation could grow again under suitable storage conditions. Al-Gheethi et al. (2013b) found that Salmonella spp. S. aureus and E. faecalis grew again in the effluents treated by SODIS for 6 hours, when the samples were stored at 37 °C for 4 days. Generally, bacterial regrowth may occur if the disinfection process does not lead to physical damage of the bacterial cell, as in the case of ozonation and different bio-filter systems (Berney et al. 2006; Vital et al. 2010).

Greywater is considered to be an alternative resource for irrigation. These waters have lower microbial loads than sewage. However, there is still a potential health risk. Therefore, effective disinfection processes rely on the ability to divert microbial cells rather than inactivation only. Advanced technical analysis to evaluate the viability of microbial cells in greywater is required before the reuse of disinfected greywater for agriculture proposes.

The authors gratefully acknowledge the Ministry of Higher Education of Malaysia for the research project financial support under fundamental research grant scheme (FRGS) vot No. 1453.