Science of the Total Environment 524–525 (2015) 40–51

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Methane emission from sewers Yiwen Liu, Bing-Jie Ni, Keshab R. Sharma, Zhiguo Yuan ⁎ Advanced Water Management Centre, The University of Queensland, QLD, Australia

H I G H L I G H T S • • • • •

Sources and sinks of methane in sewers are identified. Both offline and online methane quantification methods are reviewed and assessed. Comprehensive methane production/emission data is presented and synthesized. Models for predicting methane production in sewers are reviewed. Effects of sulfide-control chemicals on methane production in sewers are detailed.

a r t i c l e

i n f o

Article history: Received 24 February 2015 Received in revised form 3 April 2015 Accepted 3 April 2015 Available online xxxx Editor: Rolf Halden Keywords: Sewer Methane Greenhouse gas Emission Mitigation

a b s t r a c t Recent studies have shown that sewer systems produce and emit a significant amount of methane. Methanogens produce methane under anaerobic conditions in sewer biofilms and sediments, and the stratification of methanogens and sulfate-reducing bacteria may explain the simultaneous production of methane and sulfide in sewers. No significant methane sinks or methanotrophic activities have been identified in sewers to date. Therefore, most of the methane would be emitted at the interface between sewage and atmosphere in gravity sewers, pumping stations, and inlets of wastewater treatment plants, although oxidation of methane in the aeration basin of a wastewater treatment plant has been reported recently. Online measurements have also revealed highly dynamic temporal and spatial variations in methane production caused by factors such as hydraulic retention time, area-to-volume ratio, temperature, and concentration of organic matter in sewage. Both mechanistic and empirical models have been proposed to predict methane production in sewers. Due to the sensitivity of methanogens to environmental conditions, most of the chemicals effective in controlling sulfide in sewers also suppress or diminish methane production. In this paper, we review the recent studies on methane emission from sewers, including the production mechanisms, quantification, modeling, and mitigation. © 2015 Elsevier B.V. All rights reserved.

Contents 1. 2.

3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . Sources and potential sinks of methane in sewers . . . . . . . . . . 2.1. Methane production in anaerobic sewer biofilms . . . . . . . 2.2. Methane production in sewer sediments . . . . . . . . . . . 2.3. Knowledge gaps . . . . . . . . . . . . . . . . . . . . . . Methane measurement in sewers . . . . . . . . . . . . . . . . . 3.1. Off-line methane measurement and manual sampling methods 3.2. Online measurement . . . . . . . . . . . . . . . . . . . . 3.3. In-sewer dissolved- and gaseous methane data . . . . . . . . 3.3.1. Dissolved methane concentrations . . . . . . . . . 3.3.2. Gas phase methane concentrations . . . . . . . . . 3.4. Overall methane emission estimation . . . . . . . . . . . . Factors affecting methane production and emission in sewers . . . . 4.1. HRT . . . . . . . . . . . . . . . . . . . . . . . . . . .

⁎ Corresponding author. E-mail address: [email protected] (Z. Yuan).

http://dx.doi.org/10.1016/j.scitotenv.2015.04.029 0048-9697/© 2015 Elsevier B.V. All rights reserved.

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

41 41 41 42 43 43 43 44 44 44 45 45 47 47

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

4.2. A/V ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. COD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Modeling of methane production and emission in sewers . . . . . . . . . . . . . . . . 5.1. SeweX: a mechanistic model predicting methane production by anaerobic sewer biofilm 5.2. A model predicting methane production in sediments . . . . . . . . . . . . . . 5.3. Empirical models predicting methane production in sewers . . . . . . . . . . . . 5.4. Further model development . . . . . . . . . . . . . . . . . . . . . . . . . . 6. Effects of chemical dosing on methane formation in sewers . . . . . . . . . . . . . . . 6.1. Oxygen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. NO− 3 6.3. Fe3 + . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4. Elevated pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5. FNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. Conclusions and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Methane (CH4) is a highly potent fugitive greenhouse gas (GHG) that contributes significantly to climate change (IPCC, 2006). Over a 100-year horizon, 1 ton of CH4 will induce a warming effect equivalent to 21 tons of CO2 (IPCC, 2006). The global average atmospheric concentration of methane increased from approximately 0.7 ppm in 1750 to 1.8 ppm in 2013. It is estimated that up to 50% of methane emission is due to anthropogenic activities (IPCC, 2006). The plentiful carbon flow into wastewater systems creates potential for GHG emission (Keller and Hartley, 2003). If just a minor fraction of the carbonaceous compounds contained in wastewater was converted to CH4, it would result in significant GHG emission, and consequently, concern has grown significantly in recent decades (Scanlan et al., 2008). Wastewater systems comprise wastewater treatment plants (WWTPs) and sewer networks. However, studies to date have considered methane emission from WWTPs to be the major contributor (Czepiel et al., 1993; Daelman et al., 2012; Souza et al., 2012; Wang et al., 2011). Further, due to the lack of data, the Intergovernmental Panel on Climate Change (IPCC) concluded that, “…wastewater in closed underground sewers is not believed to be a significant source of methane” (IPCC, 2006). The IPCC conclusion has been challenged by the methane data reported by Guisasola et al. (2008). The authors measured liquid phase CH4 concentrations of up to 20–25 mg/L in rising main sewers in Australia. The authors highlighted that, with this level of methane production, CH4 emission from these sewers could contribute an additional GHG contribution of roughly 48–60% above that from a WWTP. Depending upon the ventilation conditions, methane could also accumulate to high concentrations in sewer headspace. Gas phase methane concentrations of up to 50,000 ppmv, i.e., 5% by volume (vol), have been detected in the air of a gravity sewer (GWRC, 2011). This is concerning because CH4 is a highly volatile gas and displays a Lower Explosive Limit (LEL) of approximately 5% vol. Uncontrolled release of methane could cause an explosion when in contact with air in confined spaces such as sewers, and thus poses a serious safety risk (Spencer et al., 2006). In addition, methane generation in sewers may consume a significant amount of soluble chemical oxygen demand (COD) (Guisasola et al., 2008), which is detrimental to nutrient removal in downstream WWTPs. The work undertaken by Guisasola et al. (2008) stimulated systematic studies on methane formation in sewers. The fundamental mechanisms underpinning methane formation in sewers have been illustrated by Guisasola et al. (2008) and Sun et al. (2014). Increased quantitative monitoring of methane in sewers has enabled significant progress to be made in terms of GHG accounting (Chaosakul et al., 2014; Foley et al., 2009; Guisasola et al., 2008; Liu et al., 2014, 2015b; Shah et al., 2011). Based

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

41

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

47 47 47 47 47 47 48 48 48 48 49 49 49 49 49 49 50

on current limited data, several models have been proposed for the prediction of methane production in sewers (Chaosakul et al., 2014; Foley et al., 2009; Guisasola et al., 2009). In addition, effects of different chemical dosing strategies on methane formation in sewers have been studied in detail (Ganigué and Yuan, 2014; Gutierrez et al., 2009; Jiang et al., 2011b, 2013b; Zhang et al., 2009). This body of work illustrates the recognition of the importance of understanding, quantifying and mitigating methane emission from sewers in recent years. In this paper, we present a review of the important outcomes and findings arising from the research on methane production and emission from sewer systems to date. Included is a description of sources and potential sinks of methane in sewers, methods for methane measurement, and rates for methane production and emission thus far measured. Also discussed are models available to predict methane production, and the effects of chemical dosing on methane production in sewers. Finally, current knowledge gaps are highlighted. 2. Sources and potential sinks of methane in sewers Sewer systems are an important and integral component of urban water infrastructure, which collects and transports wastewater from residential houses or industry to WWTPs. Operationally, sewer systems can be divided into two categories, i.e., fully-filled pressure sewers (rising main sewers), which are anaerobic, and partially-filled gravity sewers, where re-aeration takes place. In addition to transporting wastewater, sewers also act as biological reactors with various microbial processes. Generally, there are five major phases in a sewer pipe: namely the suspended wastewater phase, the wetted sewer biofilms, the sediments, the sewer air phase, and the biofilm on pipe surface exposed to sewer air, with the latter two being present in gravity sewers only. In-sewer microbial processes mainly take place in biofilms and sediments, with little contribution from the suspended biomass in the water phase or in the gas phase (Mohanakrishnan et al., 2009a). Wetted anaerobic biofilms with a thickness of a few hundred micrometers feature in rising main sewers. In gravity sewers, both biofilms and sediments below the water surface are in partially anaerobic or fully anaerobic conditions even when oxygen is present in the bulk wastewater, due to limited penetration of the oxygen (Gutierrez et al., 2008). Therefore, anaerobic fermentation and sulfate reduction using organic matter or sulfate as electron acceptors can occur in deeper layers of biofilms and sediments (HvitvedJacobsen, 2002). 2.1. Methane production in anaerobic sewer biofilms Utilizing the products of anaerobic fermentation, methanogenic archaea (MA) within sewer biofilms can produce CH4 from acetate or

42

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

100

Soluble biodegradable COD (mg/L)

60

40

20

10

B

Modelled COD

100 6

4 50 2

0

0 0

200

400

600

8

Modelled Sulfate

S-SO42- (mg/L)

Measured SRB Modelled SRB Measured MA Modelled MA

80

Fraction

150

A

800

0

200

400

0 800

600

Depth (µm)

Depth (µm)

Fig. 1. A: Relative abundance of SRB and MA; B: Depth profiles of sulfate and soluble biodegradable COD in a sewer biofilm, adapted from Sun et al. (2014). The surface of the biofilm was defined as depth 0 μm.

hydrogen under anaerobic conditions, through methanogenesis (Hvitved-Jacobsen et al., 1998). Sewers usually have limited sulfate (ca. 10–30 mg/L) and sufficient carbonaceous substrates (ca. 200–500 mg/L chemical oxygen demand, COD) to support both sulfate-reducing bacteria (SRB) and MA, thereby allowing simultaneous methane and sulfide production, despite their competition for substrate (Guisasola et al., 2008). SRB dominate the top layers of biofilms where the sulfate concentration is relatively high (Fig. 1). In comparison, MA require low-sulfate conditions. Therefore, the deeper layers of biofilms, where sulfate is limited or absent due to the diffusional limitation of sulfate while an unlimited supply of methanogenesis precursors are available, are ideal habitats for MA (Sun et al., 2014). The wastewater COD can be roughly divided into three categories, namely the fermented COD such as volatile fatty acids, readily fermentable COD such as carbohydrates, and the slowly biodegradable COD such as proteins, polysaccharides and lipids. While the slowly biodegradable may contribute to sulfate reduction and methane formation in sewers following the hydrolysis and fermentation processes, the hydrolysis process likely plays a relatively minor role due to the high abundance of fermented and fermentable COD in wastewater. All known MA are obligate anaerobes (Whitman et al., 1999) but are different from other Archaea in that MA are very sensitive to pH and temperature. MA grow optimally at temperatures N28 °C and pH range 5.5–9.0 (Whitman et al., 1999). MA grow slowly, with typical doubling times between 1 and 9 days. It has been reported that Methanosataceae, Methanomicrobiales, Methanosarcinaceae, Methanococcales and Methanocaldococcaceae exist in rising main sewer biofilms whereas Methanobacteriales is absent (Mohanakrishnan et al., 2009b). A recent study found 90% of the MA population in the biofilm

of a laboratory rising main sewer reactor belonged to the genus Methanosaeta, which is an obligate, acetoclastic methanogen (Sun et al., 2014). MA that use other substrates such as hydrogen accounted for less than 10% of the total MA population in the sewer biofilm, possibly because hydrogenotrophic MA were outcompeted by hydrogen-utilizing SRB (Kristjansson et al., 1982; Zhang et al., 2011).

2.2. Methane production in sewer sediments Previous studies suggested that activities which take place in the biofilm are the most important microbial transformations in sewers (Hvitved-Jacobsen, 2002). However, current studies have also revealed significant methane production from gravity sewer sediment which is biologically active (Liu et al., 2015a). The average methane production rate of 1.56 ± 0.14 g CH4/m2-d, is comparable to the areal rate of 1.26 g CH4/m2-d from biofilms in a rising main sewer pipe (Foley et al., 2009). Our further studies showed a large variability of methane production in sewer sediments, i.e., varying from 0.13 to 2.09 g CH4/ m2-d. Pore water sampling results (Fig. 2) indicated that the main methane production zone was located near the sediment surface (0–2 cm), but extended deeper than the sulfide production zone (0–0.5 cm). There was minimal methane production activity in the deeper layer of the sediment (2–3.5 cm) due to limited penetration of fermentable COD. A clear stratification of the microbial community occurred with sediment depth. SRB dominated at ca. 0–0.5 cm and coexisted with MA between ca. 0.5–1.0 cm. Below this depth, MA dominated the microbial populations (Liu et al., 2015a).

300

1.0

A 0.8

0.6

COD (mg/L)

Exp SRB/(SRB+MA) Model SRB/(SRB+MA) Exp MA/(SRB+MA) Model MA/(SRB+MA)

0.4

200

Modelled fermentable COD Measured sulfate Modelled sulfate Measured methane Modelled methane

100

10

5

S-SO42- / CH4 (mg/L)

Fraction

15

B

0.2

0.0 0.0

0.5

1.0

1.5

2.0 Depth (cm)

2.5

3.0

3.5

4.0

0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0 4.0

Sediment depth (cm)

Fig. 2. A: Fraction of SRB and MA; B: Depth profiles of sulfate, methane and fermentable COD in the sewer sediment, adapted from Liu et al. (2015a). The surface of the sediment was defined as depth 0 cm.

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

2.3. Knowledge gaps

43

sewer environment is not fully understood yet, and this may have implications for estimating methane emission from sewers.

It has been reported that CH4 may be oxidized by microorganisms under anaerobic, anoxic, and aerobic conditions in biofilms and sediments (Islas-Lima et al., 2004; Schreiber et al., 2010; Waki et al., 2005). These processes play an important part in regulating methane release in nature, e.g., in seas and lakes. Biofilms and sediments in gravity sewers typically have a shallow aerobic zone at the surface followed by an anaerobic zone deep in the profile (Gutierrez et al., 2008). The depth of the aerobic zone depends on the DO level in the bulk phase. Therefore, the CH4 produced in the anaerobic zone could be oxidized by methanotrophs under both anaerobic and aerobic conditions, with sulfate and oxygen being the respective terminal electron acceptors. In addition, nitrate-dependent methane oxidation (DAMO), which has been experimentally demonstrated recently (Haroon et al., 2013), may also happen if the wastewater contains nitrate due to, for example, industrial discharges, agricultural runoffs, or purposeful nitrate dosing for sewer sulfide control. However, the limited evidence has indicated that methane oxidizers grow very slowly (Valentine and Reeburgh, 2000). For example, a reported areal rate of ca. 0.02 g CH4/m2-d for aerobic or anaerobic methane oxidation in lake or ocean sediments (Bastviken et al., 2002; Iversen and Jørgensen, 1985), is substantially lower than the methane production rate of sewer sediments, i.e., 1.56 ± 0.14 g CH4/m2-d (Liu et al., 2015a). Damgaard et al. (2001) measured the methane profile in a biofilm grown at a sewage outfall, and found that a portion of the methane produced in the deep anaerobic layer was oxidized in the superficial aerobic layer. However, no additional research has been conducted since, and there is no direct evidence for the presence of methane oxidizers in sewer biofilms or sediments submerged in the water. Therefore, consumption of dissolved methane via methane oxidation in the sewer liquid phase is likely to represent a weak sink. Biofilm growing on the crown and sides of a sewer pipe, above the sewage level and exposed to air in gravity sewers, could potentially be another sink for methane. However, a recent study has shown that aerobic methanotrophs are totally absent in these biofilms due to the unfavorable acidic environment (Cayford et al., 2012), ruling them out as a major sink. Methane could be removed by various types of sewage odor treatment units (Burgess et al., 2001) employed to remove hydrogen sulfide from extracted sewer air. However, there is currently very little known about the potential for this to occur. In addition, methane can be effectively oxidized (50%) by activated sludge in aeration units at WWTPs (Daelman et al., 2014). Clearly, the potential for methane sinks in the

A

3. Methane measurement in sewers 3.1. Off-line methane measurement and manual sampling methods To date, the primary method for CH4 measurement in sewers has comprised manual sampling at regular intervals over several hours followed by off-line gas chromatography (GC) analysis (Guisasola et al., 2008; Foley et al., 2009; Shah et al., 2011). Particular sampling arrangements are required for measuring gas phase methane concentrations. Gas may be sampled from a ventilation point (Shah et al., 2011) or from a purpose-built sampling chamber connected to the sewer headspace (Liu et al., 2014), using gas bags or evacuated Exetainer® tubes (Fig. 3A). The gas samples are then analyzed for methane using a GC equipped with a flame ionization detector (FID). Dissolved methane sampling in rising main sewers is generally done using a small but flexible pipe connecting a sampling tap at ground level to the tapping arrangement of the underground pipe. Samples are collected from the pipe using a hypodermic needle and plastic syringe as shown in Fig. 3B. This procedure prevents exposure of sampled wastewater to the atmosphere and oxygen (Foley et al., 2009). For sampling dissolved methane in gravity sewers, manholes, wetwells and pumping stations, wastewater samples are usually collected with a sampling device consisting of an open-head cylindrical container. The container is lowered and filled below the water level, and then gently retrieved. Within the container, sample aliquots are extracted with a plastic syringe from ca. 5 cm below the water surface to avoid contact with air (GWRC, 2011). Alternatively, a submersible pump can be used to collect sample from below-ground at low speed in order to avoid turbulence. Sub-samples are subsequently extracted into an evacuated Exetainer® tube (Labco, Wycombe, UK) or a pre-treated serum bottle (Daelman et al., 2012). The contents of the tube or bottle are mixed overnight to reach gas–liquid equilibrium. Dissolved methane is then measured by GC, and the sample concentration is calculated using Henry's Law and mass balance. A more accurate method using evacuated Exetainer® tubes for both gas and liquid phase methane sampling and measurement has been proposed recently (Sturm et al., 2014). This method uses nitrogen gas to thoroughly flush the tubes before sampling, in order to minimize the residual methane present in the Exetainer tubes. Sewer systems are highly dynamic (Sharma et al., 2008) as sewage flows fluctuate substantially over time, leading to varied wastewater

B

Fig. 3. (A) A purpose-built device for gas sampling or infrared (IR) gas sensor application in sewer headspace: a gas pump continuously recycles the gas from the sewer headspace to the chamber and then back to the sewer. A chiller is used in the gas line feeding the chamber to maintain the desired level of 50–70% relative humidity (RH) for the IR sensor (Liu et al., 2014); (B). Collection of dissolved methane sample directly from the rising main into an airtight syringe, adapted from Foley et al. (2009).

44

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

hydraulic retention time (HRT). In addition, in rising main sewers, pumps are frequently turned on and off resulting in intermittent flow, which further adds to sewer dynamics. Liu et al. (2015b) observed a number of short-term fluctuations (e.g., within an hour) in dissolved or gas phase methane concentration, which was likely caused by the intermittent operation of pumps. Similar to the dynamics observed for hydrogen sulfide production in sewers (Sharma et al., 2008), CH4 concentrations in both the liquid and gas phases are also expected to fluctuate. It is difficult to capture the anticipated fluctuation in CH4 concentration with manual sampling. Therefore, continuous and extended monitoring of CH4 concentrations in both gas and liquid phase is important for the accurate quantification and thorough understanding of CH4 production and emission from sewers. In addition, methane production and emission data are expected to vary from site to site (Foley et al., 2009), and manual sampling is not feasible for long-term quantification of methane concentrations over a large number of sampling sites in extensive sewer networks. Therefore, online sensors for continuous CH4 measurement are optimal for the accurate quantification of methane emission. However, an online sensor may be calibrated using manual sampling. 3.2. Online measurement Although many online sensors for gas phase methane monitoring are available, most of them are not applicable in sewer conditions due to interference from hydrogen sulfide (Deng et al., 1993; Schierbaum et al., 1992). Infrared (IR) spectroscopy is the most promising method for online CH4 measurement in sewer conditions (GWRC, 2011). However, a key feature of sewer air is the high humidity, typically in the range 80–100% RH (Joseph et al., 2012), which could potentially interfere with IR CH4 measurement (Sun et al., 2011). Liu et al. (2014) evaluated the suitability of IR spectroscopy-based online sensors for measuring methane gas in humid and condensing sewer air. An IR sensor with external power supply was extremely robust in variable and high humidity. A battery-operated IR sensor was sensitive to changes in humidity, but the problem was resolved by maintaining the humidity on the sensor probe surface at 50–70% RH through increasing surface temperature or refrigeration (Fig. 3A). Both sensors exhibited excellent linearity, and can be applied with factory calibration. The detection limit of sensors i.e., ca. 0.023–0.110% vol, corresponding to a dissolved methane range of 0.005 to 0.026 mg/L under equilibrium conditions at 20 °C and 1 atm, was suitable for measuring methane gas in sewers. In-sewer application of the sensor with external power supply for nearly 1 month confirmed accuracy and longevity. In the future, infrared spectroscopy will be a powerful tool for accurate quantification of methane emission from sewers. Further studies should also evaluate the performance of gas flow meters in sewer conditions.

Furthermore, the concept of an innovative and fully automated sewer gas monitoring system based on a floating and drifting embedded sensor platform (SewerSnort) has been proposed (Kim et al., 2009). This sensor float can be introduced upstream and drift to the end of the network, collecting location-tagged gas measurements, thus providing a gas concentration profile along the sewer line. To date, the experiments have been based on a dry land emulator, and verification in actual sewers is needed before field application. A limited number of commercial sensors are available for online, dissolved methane measurement (Camilli and Hemond, 2004; Lamontagne et al., 2001; Tsunogai et al., 2007). These are mainly designed for measuring methane in clean water, using gas-permeable membranes to extract methane gas from water, and cannot be used in sewage containing a large amount of impurities as well as high sulfide concentrations (Boulart et al., 2010). Recently, a new online, dissolved methane sensor has been developed by Liu et al. (2015b). This device uses an online, gas phase methane sensor to measure methane under equilibrium conditions after stripping from the sewage. The data are then converted to liquid phase, dissolved methane concentrations according to Henry's Law. The detection limit (ca. 0.24 mg/L) and range (ca. 0–24.2 mg/L) are both suitable for sewer application, and can be adjusted by varying the ratio of liquid-to-gas phase volume settings according to specific applications, i.e., at a ratio of 4, a resolution of 0.09 mg/L can be achieved at the expense of a reduced measurement range of 0 to 9.3 mg/L. The sensor demonstrated good performance over a six-week period at the end of a rising main sewer network. 3.3. In-sewer dissolved- and gaseous methane data 3.3.1. Dissolved methane concentrations The significant contribution of sewers to methane production and emission was not recognized until recently (Table 1). Over a 4-h campaign with half-hourly manual sampling of sewage, dissolved methane with concentrations of 4.4 to 6.1 mg/L was detected at the end of a 828-m long rising main (UC9, Gold Coast, Australia). The average hydraulic retention time (HRT) was 2.5 h, yielding an average methane production of 1.1 kg CH4/day given an average flow of 200 m3/day in the pipe (GWRC, 2011). However, due to diurnal fluctuations in wastewater flow, the actual HRT of this sewer varied between 1.5 and 6 h, and the current estimation may misrepresent the actual methane production without considering the entire range of HRT. Dissolved methane concentrations of 11–33 mg/L (case I, GWRC, 2011) and 3.4–6.6 mg/L (case II, Foley et al., 2009) were measured at the end of a 1100-m long rising main (CO16, Gold Coast, Australia), during the respective 4- and 6-h sampling campaigns in the early morning with hourly, manual sampling. The average daily flow in this pipe was 707 m3/day, resulting in an average daily methane production of 9.8

Table 1 Dissolved methane concentrations and methane emission in rising mains. Name

Type

Length (m)

Diameter (mm)

A/V (m−1)

HRT (h) average (min–max)

Average wastewater temperature (°C)

Dissolved CH4 (mg/L) average (min–max)

Daily flow (m3/day)

Production (kg CH4/day)

Overall emission: kg CO2-e/m3

References

UC9 CO16 CO16 C27 C27 Perth B RV

Rising main Rising main Rising main Rising main Rising main Rising main Gravity

828 1100 1100 4400 4400 15,000 1000

150 300 300 525 525 900 1000

26.7 13.3 13.3 7.6 7.6 4.4 –

2.5 (3.1–4.6) 2.6 (3.9–11.0) 2.6 (1.5–7.3) 9.1 9.1 – 27.9 (22–31.4)

27.7 22.5 23.5 24.6 20.3 – 33.3

5.3 (4.4–6.1) 15.3 (11.0–33.0) 5.2 (3.4–6.6) 9.1 (5.0–15.0) 7.1 (3.5–12) 4.8 10.1 (8.0–13.7)

200 707 707 2840 2840 11,000 –

1.1 9.8 2.6 24.6 19.0 52.8 –

0.11 0.32 0.11 0.19 0.15 0.10 –

RV

Gravity

1000

1000

–

7.8 (0–12)

30.2

4.6 (0.1–11.4)

–

–

–

CO16 PS

Pumping station Pumping station

–

–

–

–

23.5

1.5 (1.0–1.92)

707

–

–

GWRC (2011) GWRC (2011) Foley et al. (2009) Liu et al. (2015b) Liu et al. (2015b) Liu et al. (2014) Chaosakul et al. (2014) Chaosakul et al. (2014) Foley et al. (2009)

–

–

–

–

–

0.51

2000

–

–

Liu et al. (2014)

OR3 PS

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

and 2.6 kg CH4/day, respectively. This substantial difference in the results could not be fully explained by the differences in HRT (3.9– 11.0 h for the 4-h sampling and 1.5–7.3 h for the 6-h sampling) and temperature (22.5 °C and 23.5 °C, respectively). A long-term continuous monitoring campaign would provide more insight. Recently, a newly developed online measuring device was used to monitor dissolved methane concentrations at the end of a 4.4-km long rising main sewer (C27, Gold Coast, Australia), over a period of three weeks in both summer and early winter (Liu et al., 2015b). This rising main sewer carries a flow of 2840 m3/day, leading to an average HRT of 9.1 h. The measured data showed a wide variation in dissolved methane concentration i.e., 5–15 mg/L in summer and 3.5–12 mg/L in winter. Based on these measurements, the average daily production of methane was estimated as 24.6 and 19.0 kg-CH4/day, respectively. Also, the large variation in methane concentration within a short time span (i.e., 15 min) clearly demonstrates that it is practically impossible to adequately capture the dynamics with manual sampling (Liu et al., 2015b). Chaosakul et al. (2014) measured dissolved methane levels in a 1-km long gravity sewer (Thailand, Table 1). The concentrations varied from 8.0 to 13.7 mg/L in dry weather (HRT of 22–31.4 h) and 0.1–11.4 mg/L in wet weather (HRT of 0–12 h). This observation demonstrates that while methane emission occurs in gravity sewers following sewage discharge from an upstream rising main, a significant proportion of methane still remains in the liquid phase, and eventually emits at the downstream gravity section or inlet head works of the downstream WWTP. Pumping station (PS) methane concentrations have also been reported in the literature (Table 1). The dissolved methane concentrations are close to 0 if the station receives ‘fresh’ sewage from nearby households (Liu et al., 2015b). If the PS receives aged sewage from other upstream pumping stations, then dissolved methane concentrations usually vary from 0.5 to 2 mg/L (see data for CO16 and OR3 in Table 1). 3.3.2. Gas phase methane concentrations In pressure sewers, methane can be produced and accumulated beyond saturated concentrations (Table 1) because in-sewer pressure is in excess of atmospheric pressure (Guisasola et al., 2008). When sewage flow from an enclosed anaerobic sewer pipe is discharged into a ventilated space, i.e., pumping station, wet-well, gravity sewer or even influent works of WWTPs, a large proportion of dissolved methane is stripped off to the atmosphere under turbulence, resulting in significant emissions (Table 2). Liu et al. (2014) reported that gas phase CH4 concentrations varied between 7000 and 12,000 ppmv with an average concentration of 9000 ppmv in a manhole receiving 17,000 m3/day of sewage discharge from a gravity sewer (Perth A, Western Australia, Table 2). In addition, Chaosakul et al. (2014) detected gas phase methane concentrations between 7164 and 17,183 ppmv at the end of a 1-km long gravity sewer (Thailand). Furthermore, in a study in Melbourne, Australia, almost

45

half of the 14 manholes along a gravity sewer line (GWRC, 2011) contained methane concentrations of up to 50,000 ppmv or 5%, i.e., equivalent to the methane Lower Explosive Limit (LEL) (Table 2). Another field study in Boston of the USA during autumn also found a large number of locations where methane concentrations in manholes exceeded the LEL (Phillips et al., 2013). These results indicate that insewer methane formation is not limited to warm countries such as Australia. However, more data should be collected in cold regions to determine the extent of methane formation in courtiers with a cold climate. Besides the issue of direct GHG emission, methane at high concentration poses a serious safety issue in sewers. These data demonstrate that gravity sewers can emit a substantial amount of methane. However, due to the lack of appropriate tools and methods, the data cannot be used to calculate the methane production and emission rates. By conducting gas phase methane measurements at discharge manholes, other studies attempted to illustrate direct methane emission rates after rising mains during sewage transport (Table 2). In a US study, gas phase methane concentrations of 500–900 ppmv were detected using off-line measurement at the discharge location of a 5.3 km rising main (HCPS). The wastewater flow was 1855 m3/day, yielding a direct CH4 emission of 7.44 kg/day (Shah et al., 2011). Gas phase methane concentrations of 2500–45,000 ppmv were also measured by an online methane sensor at the discharge manhole of a 1.3-km long rising main (CO16, Gold Coast, Australia) with a daily flow of 707 m3/day (GWRC, 2011). Furthermore, methane concentrations in the headspace of a discharge manhole of a 15-km long rising main (Perth B, Western Australia, Table 2) with a daily flow of 11,000 m3/day fluctuated from 15,000 ppmv to 29,000 ppmv, with an average concentration of 20,000 ppmv over 12 days, as measured by an online sensor (Liu et al., 2014). Considering the higher methane concentrations in the last two cases, substantial direct methane emission rates can be expected in discharge manholes. In addition, grab sampling at 65 different pumping stations in Georgia, USA revealed methane emission rates between 1.13 and 11.68 kg CH4/day (GWRC, 2011). For example, at SMPLS pumping station receiving a sewage discharge of 378.5 m3/day, gas phase CH4 varied between 65 and 275 ppmv, resulting in a methane emission rate of 1.18 kg CH4/day. Similarly, online measurement at OR3 pumping station (Gold Coast, Australia) with a daily flow of 2000 m3/day showed a gas phase methane concentration of 1400–2800 ppmv (Liu et al., 2014), which is one order higher than that measured at SMPLS. Both these cases illustrate that pumping stations are potential ‘hot spots’ for direct GHG emission. 3.4. Overall methane emission estimation For quantifying methane production in a rising main sewer, sewage flow data, and dissolved methane concentrations at the

Table 2 Gas phase methane concentrations in sewer air. Name

Type

Length Diameter A/V CH4 (ppmv) average Daily (min–max) (m) (m) (m−1) flow (m3/day)

Direct Comment emission (kg CH4/day)

Reference

Perth A RV RV SEW

Gravity Gravity Gravity Gravity

6000 1000 1000 –

1800 1000 1000 –

– – – –

17,000 – – –

9000 (7000–12,000) 17,183 (1,3500–23,000) 7164 (65–19,000) 1500–50,000

– – – –

Liu et al. (2014) Chaosakul et al. (2014) Chaosakul et al. (2014) GWRC (2011)

Perth B HCPS CO16 OR3 PS SMPLS PS US PS

Rising main Rising main Rising main Pumping station Pumping station Pumping station

15,000 5310 1300 – – –

900 406 300 – – –

4.4 9.9 13.3 – – –

11,000 1855 707 2000 378.5 –

20,000 (15,000–29,000) 600 (500–900) 10,000 (2500–45,000) 2200 (1400–2800) 116 (65–275) –

– 7.44 – – 1.18 1.13–11.68

– Dry weather Wet weather Data from several manholes receiving industrial wastewaters Measured at discharge manhole Measured at discharge manhole Measured at discharge manhole – – Data from 65 pumping stations across the USA

Liu et al. (2014b) Shah et al. (2011) GWRC (2011) Liu et al. (2014) Shah et al. (2011) GWRC (2011)

46

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

Fig. 4. Sampling locations for measuring methane production and emission in rising main (A) and gravity (B) sewers.

upstream pumping station and at the end of the pipe are required (Fig. 4). Because gravity sewers are partially filled, both liquid and gas phase methane levels in the upstream manhole and at the end of the sewer, and both water and gas flow rates are required for the evaluation of methane production. Considering that methane oxidation in sewers is expected to be a slow process (Valentine and Reeburgh, 2000), it is reasonable to assume that the majority of the methane formed would be eventually stripped to the atmosphere via ventilation in gravity sewers or at WWTPs. Therefore, these data can also be used to calculate potential overall emission rates from sewer systems. In some studies, the quantification of overall CH 4 emissions has been done by direct measurement of methane gas flux from a discharge manhole (Shah et al., 2011). However, this is expected to underestimate emissions as CH 4 could also be emitted at other locations in the network. To date, overall methane emission data are available for single-pipe rising main sewers e.g., UC09 and CO16, and a small rising main network i.e., C27. The overall methane emission potential of these rising main sewers varies substantially, ranging from 0.10 to 0.32 kg CO2-e/m3 with an average value of 0.18 kg CO2-e/m3 of wastewater transported (Table 1, assuming that 1 kg CH4 is equivalent to 21 kg CO2 in terms of global warming potential). In contrast, previous studies have identified that WWTPs are an important source of methane emission (Bousquet et al., 2006), with direct

methane emission factors ranging from 0.14 to 3.44 g CH4/influent m3, i.e., 0.003–0.072 kg e-CO2/influent m3 (Daelman et al., 2012). It should be noted that in calculating methane emission factors for WWTPs, the direct emission arising from stripping sewage at the influent works has been taken into account (Wang et al., 2011), and that this is in fact the methane production from sewers. Considering both direct and indirect emissions, de Haas et al. (2014) estimated that the total GHG emissions from WWTPs in Australia is 0.5–2.0 kg CO2-e/m3 with a mean value of 1.0 kg CO2-e/m3 of wastewater treated. Using this data, methane production in sewers UC09, CO16 and C27, contributes ca. 18% of the total GHG emission during wastewater handling and treatment. It should be noted that those particular sewers are only a small part of a much larger network, and hence more methane production is expected when the wastewater is transported through the remaining parts of the network before reaching the WWTP (Pikaar et al., 2014). Therefore, methane production from sewers plays an important role in contributing to overall methane emissions over the entire wastewater systems. In summary, due to the operational complexity of sewer systems and dynamic nature of methane emissions it is impractical to estimate overall CH4 emissions from large networks through either online or off-line measurements. Instead, a more appropriate use of field data would be for model calibration and validation when incorporated into system-wide mathematical modeling.

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

4. Factors affecting methane production and emission in sewers Some key factors regulating methane production and emission in sewers have been identified in recent studies (Foley et al., 2009; Guisasola et al., 2008). These factors include HRT, pipe area-to-volume (A/V) ratio, COD, and temperature, which also affect dissolved sulfide concentrations. 4.1. HRT Guisasola et al. (2009) found that dissolved methane concentration was positively correlated with HRT in sewers. Through observing that methane concentrations increased along the length of the sewer in the field studies, Foley et al. (2009) similarly concluded that increased retention time caused increased methane production. Liu et al. (2015b) observed a clear diurnal pattern, with higher dissolved CH4 concentrations overnight and lower concentrations during the day, likely caused by the diurnal fluctuation in HRT in the network (Sharma et al., 2013). In addition, Chaosakul et al. (2014) detected a higher methane concentration in both liquid and gas phase in a gravity sewer during periods of lengthy HRT (Table 2). 4.2. A/V ratio It has also been documented that the dissolved methane concentration is related to the A/V ratio of the sewer pipe. A higher A/V ratio enables more biofilm growth per unit volume of the wastewater and thus gives a higher methane production rate. Both Guisasola et al. (2009) and Foley et al. (2009) revealed that higher A/V ratio resulted in a higher methane production. 4.3. Temperature Liu et al. (2015b) demonstrated that temperature also plays an important role in methane production. A higher methane production rate was observed in summer as compared with winter (Table 1). The results from pumping stations in the USA (Table 2) showed that the concentration of CH4 in the gas phase was, in 80% of the cases, higher in summer than in winter (GWRC, 2011). 4.4. COD Liu et al. (2015a) demonstrated the dependency of methane production in sewer sediments was mainly related to fermentable COD concentrations. Also, trade waste containing high COD discharged into domestic sewers was found to significantly increase methane production (Sudarjanto et al., 2011). Since dissolved sulfide concentration has a positive correlation with HRT, A/V ratio, COD, and temperature (Sharma et al., 2008), it is also likely to be correlated with methane concentration. Liu et al. (2015b) reported that the dissolved CH4 and sulfide profiles in a rising main had a strong positive correlation, with an R2 value of 0.47. Guisasola et al. (2008) reported similar trends with methane and sulfide profiles, both displaying positive correlation with HRT. Thus, high methane concentrations are expected at sulfide ‘hot spots’ (severe odor or corrosion), and this could provide a convenient way of locating likely areas with high methane levels in a sewer network. 5. Modeling of methane production and emission in sewers It is impractical to quantify methane concentrations either by online or manual methods over a large number of sampling sites along extensive sewer networks, particularly on a long-term basis. However, mathematical modeling is a viable option for predicting and estimating methane production and emission in sewer networks, and therefore

47

can be a powerful tool for supporting industry in operational optimization and the development of mitigation strategies. 5.1. SeweX: a mechanistic model predicting methane production by anaerobic sewer biofilm Sharma et al. (2008) developed the SeweX model to predict the dynamics of sulfide and volatile fatty acid (VFA) production in sewers. SeweX is a dynamic sewer model, describing in-sewer biological, chemical and physical processes. It predicts both the temporal and spatial variations of wastewater composition, including sulfate and sulfide, using sewer network configuration, pipe geometry, sewage characteristics and hydraulic data as inputs. The most recent version of the SeweX model is now expanded to include methane generation by incorporating the model developed by Guisasola et al. (2009). SeweX is the first model to predict the spatial and temporal variation in sewer methane concentration. The following processes underpinning methane production in sewers are included in the model: 1. 2. 3. 4. 5. 6. 7.

Acidogenesis Acetogenesis Acetoclastic methanogenesis Hydrogenotrophic methanogenesis Hydrogenotrophic sulfidogenesis Acetate-based sulfidogenesis Propionate-based sulfidogenesis.

Fermentation has been modeled as two separate processes, acetogenesis and acidogenesis. Three fermentation products are considered, namely H2, acetate and propionate. The biofilm-catalyzed processes were modeled using Monod kinetics, and higher values for saturation constants were used to account for diffusion limitations in biofilms. This model can be utilized to study the effect of key sewer operational parameters on methane formation. The methane component of the SeweX model was initially calibrated using data collected from lab-scale experiments. Subsequently, the model was validated using manually sampled, off-line methane data. Therefore, more online field measurement data are needed for better calibration and validation of the methane related kinetics. Furthermore, the in-sediment biological reactions in gravity sewers are modeled as biofilm processes due to the knowledge gap related to biological transformations in sewer sediment. 5.2. A model predicting methane production in sediments Recently, Liu et al. (2015a) built a detailed, one-dimensional sediment model to predict methane and sulfide production and microbial distribution in a sewer sediment based on biological reactions proposed by Guisasola et al. (2009). For the soluble components involved in the biological reactions, the first step relates to diffusion into the sediment where the reactions take place, and only vertical diffusion in the sediment is considered. All model parameter values were obtained from literature and no model calibration was performed. However, the modelpredicted methane production profiles within the sediment matched very well with the experimental data. As with SeweX, more online field measurement data are needed for better calibration and validation of the model. The proposed model may be useful in practical applications to determine the contribution of sewer sediments to the overall sewer network emissions. A simple half-order kinetic equation (Eq. (1)) was derived from the detailed sediment model to predict methane production in sewer sediment (Liu et al., 2015a). The model is simple as it involves only one parameter to be calibrated i.e., k (see Eq. (1) below), and could potentially be incorporated into the dynamic SeweX model. However, more field data is required to examine the accuracy of this

48

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

sediment model and to understand the dependency of k on sewer conditions.

where, rCH4 = areal methane production rate (g CH4/m2 day); k = rate constant for methane production ((g CH4/m)0.5/day); SF = bulk fermentable COD concentration (mg/L).

processes. In addition, processes which serve as a sink for methane in sewers should be included in the model once an understanding is established. There is increasing interest in studying the effect of the interactions among urban water system components. This can be enabled through integrating the WWTP model and the SeweX model, resulting in better prediction of methane emission over the entire wastewater system (Guo et al., 2012). With further development of individual sewer models, the integrated model will be more reliable for standard applications.

5.3. Empirical models predicting methane production in sewers

6. Effects of chemical dosing on methane formation in sewers

Foley et al. (2009) has proposed a simple empirical model (Eq. (2)) for estimating methane production in a rising main sewer, based on the observation that it is related to the wastewater HRT and the A/V ratio of the pipe. This simple equation offers a tool for water authorities to predict methane emissions from a rising main sewer.

Recent studies have shown a range of chemical dosing approaches to mitigate sulfide emission from sewers (Zhang et al., 2008). Methane formation can also be suppressed by commonly used chemicals such as nitrate, oxygen, ferric salts, hydroxide (pH elevation) and free nitrous acid (FNA) (Table 3) which are used for treating sulfide-related problems (Ganigue et al., 2011). The reason for this phenomenon is that the methanogens are slow growers and are very sensitive to environmental conditions as compared with SRB (Whitman et al., 1999). In contrast to SRB, methanogens usually inhabit the deeper zone of sewer biofilms or sediments, and are usually protected due to limited penetration of the dosed chemical. For effective control of methanogens, higher dosage of chemicals may be needed to achieve full penetration during the initial dosing period, when overall bacterial activity is high. However, continuous dosing, as required for sulfide control with most chemicals, may not be necessary.

r CH4 ¼ k  S F

0:5

ð1Þ

−5

CCH4 ¼ 5:24  10

 ½A=V  HRT þ 0:0015

ð2Þ

where CCH4 is the concentration of dissolved methane (kg/m3); 5.24 × 10−5 kg/m2/h equals the rate of methanogenic activity of the pipeline biofilm; 0.0015 kg/m3 equals the average residual concentration of dissolved methane. However, it should be noted that the methane production rate (5.24 × 10−5) is expected to be affected by other factors such as wastewater composition (specifically COD) and temperature, and thus may vary from system to system. Therefore, more field data is required in order to develop and calibrate this empirical model for its application. Chaosakul et al. (2014) proposed a similar empirical model to predict methane formation in gravity sewers based on A/V ratio, HRT and wastewater temperature (Eq. (3)). The model has been calibrated with field methane data and partially validated using rising main sewer data.

CCH4 ¼ 6  10

−5

ðT−20Þ

 ½A=V  HRT  1:05

þ 0:0015

ð3Þ

where C CH 4 is concentration of dissolved methane (kg/m 3 ); 6 × 10 − 5 kg/m2 /h equals the rate of methanogenic activity of the pipeline biofilm; 0.0015 kg/m3 equals the average residual concentration of dissolved methane; 1.05(T − 20) is a function of temperature. 5.4. Further model development Currently, the potential for biological CH4 oxidation has not been factored into sewer models as there is a lack of understanding of those

6.1. Oxygen Ganigué and Yuan (2014) found that long-term oxygen injection at 15–25 mg/L reduced methane formation by 47% in laboratory-scale sewer reactors. The methane production rate also dropped to 15% during short-term oxygen injection over an exposure time of 6 h, however, methane production fully recovered after ca. 20 days. It was likely that full control over CH4 production was not achieved due to only partial oxygen penetration into sewer biofilm. A compounding factor is that oxygen injection into sewers may potentially lead to N2O production via the development of a nitrifying microbial community. However, negligible N2O production was observed in the study reported in Ganigué and Yuan (2014). It should be noted that the presence of oxygen also promotes heterotrophic activity, resulting in oxidation of a significant amount of organic matter in the sewage, which will in turn, impact nutrient removal processes at the WWTP. Gutierrez et al. (2008) suggested that oxygen should be dosed at a downstream sewer location for maximum effectiveness in sulfide control. The biological or chemical oxidation of methane using oxygen is

Table 3 Comparison of the effects of chemical addition on methane production in sewers. Chemical

Dosing levels

Condition

Dosing plan

CH4 reduction level (%)

H2S reduction level (%)

Methane production recoverya

Reference

Oxygen Oxygen Nitrate Nitrate Nitrate Nitrate Hydroxide Hydroxide Iron salts FNA FNA

15–25 kg O2/ML 15–25 kg O2/ML 30 kg N–NO2− 3 /ML 30 kg N–NO2− 3 /ML 17 kg N–NO2− 3 /ML 50 kg N–NO2− 3 /ML pH = 9 pH = 11.5 21 kg Fe/ML 0.26 kg N-FNA/ML 0.26 kg N-FNA/ML

Laboratory Laboratory Laboratory Laboratory Field Field Simulation Field Laboratory Field Laboratory

Continuous Continuous Continuous Continuous One shock One shock Continuous Shock for 6 h Continuous Shock for 8 h Shock for 12 h

47 – 42 94 13 27 98 97 43 – 99

– 35 – 66 – – 99 67b 99 N80c –

– – – – 100% in 2 days 100% in 2 days – 3% in 15 days – – 20% in 14 days

Ganigué and Yuan (2014) Gutierrez et al. (2008) Jiang et al. (2013b) Mohanakrishnan et al. (2009a) Shah et al. (2011) Shah et al. (2011) Gutierrez et al. (2009) Gutierrez et al. (2014) Zhang et al. (2009) Jiang et al. (2013a) Jiang et al. (2011a)

a b c

Percentage of methane production recovered over a certain time after dosing. 67% sulfide reduction in next 2 days following shock dosing. The SRB activity recovered gradually over a period of 7 days. Over 80% decrease in sulfide production for 10 days.

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

very slow as compared with that of sulfide, and oxygen should be dosed at multiple locations within the sewer system over an extended period to achieve network-wide control on methanogenic activity. This approach is likely to incur significant costs for chemical usage, which could hinder its application. 6.2. NO− 3 A recent study conducted by Jiang et al. (2013b) demonstrated that the long-term addition of 30 mg N/L of nitrate resulted in a 90% reduction in methanogenic activity in a rising main sewer reactor. The longterm dosing reduced methane concentrations in the effluent by 42%, while negligible nitrous oxide was produced. It was suggested that methanogenesis may persist in deeper biofilms due to the availability of soluble organic substrates, and the limited penetration of nitrate and sulfate (Liu et al., 2015c). Field trials (Table 3) further showed that following one nitrate shock dose of 50 mg-N/L, the intermittent addition of nitrate to sewers reduced methane production by 27%, but full recovery of methane production occurred after 2 days. Similar to oxygen, nitrate dosing also promotes heterotrophic activity, thereby affecting the nutrient removal processes at the WWTP.

49

pumping station for a period of 8 h. MA are more sensitive than SRB to FNA, and the laboratory dosing study demonstrated complete methane control, giving good evidence that FNA is effective for both methane mitigation and sulfide control in sewer systems. The current practice of selecting chemicals and design of dosing locations/rates is mainly based on an individual's experience. Instead, the approach should be based on specific features of the site in question. In this respect, the SeweX model functions as an empowering tool in supporting the decision-making. Also, constant, flow-paced and profiled dosing rates are currently applied during the chemical dosing, also based on experience. Sulfide and methane concentrations, and flows show significant temporal and spatial dynamics, and the current methods could result in over-dosing of chemical during periods with low sulfide and methane production, and conversely under-dosing during other periods. However, use of the newly developed, simple, online dissolved methane/sulfide sensor (Liu et al., 2015b), enables dynamic chemical dosing based on online measurement of flow rate and dissolved methane/sulfide concentrations, potentially achieving more cost-effective and responsive mitigation methods.

7. Conclusions and outlook 6.3. Fe3+ Zhang et al. (2009) investigated the impact of long-term ferric chloride dosing (ca. 21 mg/L) on methanogenic activity of sewer biofilms, and reported that effluent methane concentrations of a sewer reactor were reduced by 43%, together with almost complete control of sulfide production i.e., 99%. Iron salts have been the most common chemicals used to control sulfide, accounting for about 66% of the total sewage treated with chemical dosing (Ganigue et al., 2011). Further field trials are required to confirm the effectiveness of Fe3+ in mitigation of methane production. 6.4. Elevated pH Gutierrez et al. (2009) reported that a long-term elevated pH to 8.6– 9.0 suppressed the growth of methanogens in sewer reactors. The model simulation of long-term hydroxide dosing at the upstream of the network indicated that full control of both methane and sulfide production was achievable. Further field trials (Gutierrez et al., 2014) showed effective methane control could be attained with short-term, moderately-elevated pH. The exposure of the entire sewer line to pH 11.5 for 6 h was adequate for complete control of methane production for more than 2 weeks. A 67% decrease in sulfide production occurred in the 2 days following the alkali shock-dosing. The SRB activity recovered gradually over a period of 7 days, while the methanogenic activity took much longer to recover. Recently, a new caustic generation system (Pikaar et al., 2011, 2013) has been able to continuously produce, in-situ, ca. 3% hydroxide w/w from sewage at a lower cost than the scenario of caustic dosing, once a week at pH 11. This electrochemical method is a promising technology for methane control in small sewers, but field trials are needed to confirm effectiveness.

Although sewer systems contribute minimally to global GHG emissions, methane production and emission from sewer systems can be significant in terms of the overall carbon footprint of wastewater systems (a conservative value of ca. 18%), and therefore, should be accounted for. Current data suggest that methane is mainly produced by methanogens in the deeper layers of sewer biofilms and sediments in both rising main and gravity sewers. Methane emission mostly occurs under turbulence at structures exposed to the atmosphere such as pumping stations, wet-wells, and influent works of WWTPs. Methane production in sewers can be affected by factors such as HRT, A/V ratio, temperature, and COD loading. Furthermore, sewer methane production and emission display significant temporal and spatial variation, indicating that online measurement will be the most suitable method for quantification. In addition, an appropriate sampling protocol for both rising main and gravity sewers has been presented to enable more accurate estimation of sewer methane production and emission. Based on the current understanding and availability of data, several dynamic and empirical models have been developed. These models may prove to be very useful tools to support GHG accounting and mitigation strategy development. Future research is likely to focus on both quantification of in-situ methane production, and reduction of emissions from specific sewer sites. More detailed and comprehensive field data will be required for the development of reliable accounting guidelines and mathematical models for the prediction of methane emission from sewer networks. In addition, further field demonstration of the mitigation methods will be essential for trialing other cost-effective chemicals, and for the development of better control options such as online dynamic control of dosing rates. Future studies should also focus on identifying potential methane sinks in sewers.

6.5. FNA Acknowledgments Recent laboratory studies demonstrated the inhibitory effect of FNA on SRB and methanogenic activities in sewer biofilms (Jiang et al., 2010, 2011b). Jiang et al. (2011a) reported that with intermittent dosing, FNA exerts a strong biocidal effect on methanogens in sewer biofilms. A 12-h exposure with 0.26 mg-N/L of FNA suppressed methanogenesis in the sewer reactor, resulting in only 20% recovery of methane production in the following two weeks. This cost-effective dosing method has also been verified in an actual sewer (Jiang et al., 2013a). Sulfide production was reduced by more than 80% over a period of 10 days after FNA was dosed once at a

The authors acknowledge the Australian Research Council, the Council of the City of Gold Coast, the Melbourne Water Corporation, the South East Water, the Western Australia Water Corporation and the District of Columbia Water and Sewer Authority for the financial support through project LP110201095. Yiwen Liu acknowledges the support of an Endeavour International Postgraduate Research Scholarship (IPRS) and The University of Queensland Centennial Scholarship (UQCent). Bing-Jie Ni acknowledges the support of an Australian Research Council Discovery Early Career Researcher Award (DE130100451).

50

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51

References Bastviken, D., Ejlertsson, J., Tranvik, L., 2002. Measurement of methane oxidation in lakes: a comparison of methods. Environ. Sci. Technol. 36 (15), 3354–3361. Boulart, C., Connelly, D., Mowlem, M., 2010. Sensors and technologies for in situ dissolved methane measurements and their evaluation using technology readiness levels. TrAC Trends Anal. Chem. 29 (2), 186–195. Bousquet, P., Ciais, P., Miller, J.B., Dlugokencky, E.J., Hauglustaine, D.A., Prigent, C., et al., 2006. Contribution of anthropogenic and natural sources to atmospheric methane variability. Nature 443 (7110), 439–443. Burgess, J.E., Parsons, S.A., Stuetz, R.M., 2001. Developments in odour control and waste gas treatment biotechnology: a review. Biotechnol. Adv. 19 (1), 35–63. Camilli, R., Hemond, H., 2004. NEREUS/Kemonaut, a mobile autonomous underwater mass spectrometer. TrAC Trends Anal. Chem. 23 (4), 307–313. Cayford, B.I., Dennis, P.G., Keller, J., Tyson, G.W., Bond, P.L., 2012. High-throughput amplicon sequencing reveals distinct communities within a corroding concrete sewer system. Appl. Environ. Microbiol. 78 (19), 7160–7162. Chaosakul, T., Koottatep, T., Polprasert, C., 2014. A model for methane production in sewers. J. Environ. Sci. Health A 49 (11), 1316–1321. Czepiel, P.M., Crill, P.M., Harriss, R.C., 1993. Methane emissions from municipal wastewater treatment processes. Environ. Sci. Technol. 27 (12), 2472–2477. Daelman, M.R.J., van Voorthuizen, E.M., van Dongen, U.G.J.M., Volcke, E.I.P., van Loosdrecht, M., 2012. Methane emission during municipal wastewater treatment. Water Res. 46 (11), 3657–3670. Daelman, M.R.J., Van Eynde, T., van Loosdrecht, M.C.M., Volcke, E.I.P., 2014. Effect of process design and operating parameters on aerobic methane oxidation in municipal WWTPs. Water Res. 66, 308–319. Damgaard, L.R., Nielsen, L.P., Revsbech, N.P., 2001. Methane microprofiles in a sewage biofilm determined with a microscale biosensor. Water Res. 35 (6), 1379–1386. de Haas, D.W., Pepperell, C., Foley, J., 2014. Perspectives on greenhouse gas emission estimates based on Australian wastewater treatment plant operating data. Water Sci. Technol. 69 (3), 451–463. Deng, Y., Nevell, T.G., Ewen, R.J., Honeybourne, C.L., 1993. Sulfur poisoning, recovery and related phenomena over supported palladium, rhodium and iridium catalysts for methane oxidation. Appl. Catal. A Gen. 101 (1), 51–62. Foley, J., Yuan, Z.G., Lant, P., 2009. Dissolved methane in rising main sewer systems: field measurements and simple model development for estimating greenhouse gas emissions. Water Sci. Technol. 60 (11), 2963–2971. Ganigué, R., Yuan, Z., 2014. Impact of oxygen injection on CH4 and N2O emissions from rising main sewers. J. Environ. Manag. 144, 279–285. Ganigue, R., Gutierrez, O., Rootsey, R., Yuan, Z., 2011. Chemical dosing for sulfide control in Australia: an industry survey. Water Res. 45 (19), 6564–6574. Guisasola, A., de Haas, D., Keller, J., Yuan, Z., 2008. Methane formation in sewer systems. Water Res. 42 (6-7), 1421–1430. Guisasola, A., Sharma, K.R., Keller, J., Yuan, Z., 2009. Development of a model for assessing methane formation in rising main sewers. Water Res. 43 (11), 2874–2884. Guo, L., Porro, J., Sharma, K.R., Amerlinck, Y., Benedetti, L., Nopens, I., et al., 2012. Towards a benchmarking tool for minimizing wastewater utility greenhouse gas footprints. Water Sci. Technol. 66 (11), 2483. Gutierrez, O., Mohanakrishnan, J., Sharma, K.R., Meyer, R.L., Keller, J., Yuan, Z., 2008. Evaluation of oxygen injection as a means of controlling sulfide production in a sewer system. Water Res. 42 (17), 4549–4561. Gutierrez, O., Park, D., Sharma, K.R., Yuan, Z., 2009. Effects of long-term pH elevation on the sulfate-reducing and methanogenic activities of anaerobic sewer biofilms. Water Res. 43 (9), 2549–2557. Gutierrez, O., Sudarjanto, G., Ren, G., Ganigué, R., Jiang, G., Yuan, Z., 2014. Assessment of pH shock as a method for controlling sulfide and methane formation in pressure main sewer systems. Water Res. 48, 569–578. GWRC, 2011. Global Water Research Coalition Report, N2O and CH4 emission From Wastewater Collection and Treatment Systems: State of the Art Report. Haroon, M.F., Hu, S., Shi, Y., Imelfort, M., Keller, J., Hugenholtz, P., et al., 2013. Anaerobic oxidation of methane coupled to nitrate reduction in a novel archaeal lineage. Nature 500 (7464), 567–570. Hvitved-Jacobsen, T., 2002. Sewer Processes: Microbial and Chemical Process Engineering of Sewer Networks. CRC Press, USA. Hvitved-Jacobsen, T., Vollertsen, J., Nielsen, P.H., 1998. A process and model concept for microbial wastewater transformations in gravity sewers. Water Sci. Technol. 37 (1), 233–241. IPCC, 2006. In: Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T., Tanabe, K. (Eds.), Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse Gas Inventories Programme. IGES, Japan. Islas-Lima, S., Thalasso, F., Gomez-Hernandez, J., 2004. Evidence of anoxic methane oxidation coupled to denitrification. Water Res. 38 (1), 13–16. Iversen, N., Jørgensen, B.B., 1985. Anaerobic methane oxidation rates at the sulfate–methane transition in marine sediments from Kattegat and Skagerrak (Denmark). Limnol. Oceanogr. 30 (5), 944–955. Jiang, G., Gutierrez, O., Sharma, K.R., Yuan, Z., 2010. Effects of nitrite concentration and exposure time on sulfide and methane production in sewer systems. Water Res. 44 (14), 4241–4251. Jiang, G., Gutierrez, O., Sharma, K.R., Keller, J., Yuan, Z., 2011a. Optimization of intermittent, simultaneous dosage of nitrite and hydrochloric acid to control sulfide and methane productions in sewers. Water Res. 45 (18), 6163–6172. Jiang, G., Gutierrez, O., Yuan, Z., 2011b. The strong biocidal effect of free nitrous acid on anaerobic sewer biofilms. Water Res. 45 (12), 3735–3743.

Jiang, G., Keating, A., Corrie, S., O'Halloran, K., Nguyen, L., Yuan, Z., 2013a. Dosing free nitrous acid for sulfide control in sewers: results of field trials in Australia. Water Res. 47 (13), 4331–4339. Jiang, G., Sharma, K.R., Yuan, Z., 2013b. Effects of nitrate dosing on methanogenic activity in a sulfide-producing sewer biofilm reactor. Water Res. 47 (5), 1783–1792. Joseph, A.P., Keller, J., Bustamante, H., Bond, P.L., 2012. Surface neutralization and H2S oxidation at early stages of sewer corrosion: influence of temperature, relative humidity and H2S concentration. Water Res. 46 (13), 4235–4245. Keller, J., Hartley, K., 2003. Greenhouse Gas Production in Wastewater Treatment: Process Selection is the Major Factor. pp. 43–48. Kim, J., Lim, J.S., Friedman, J., Lee, U., Vieira, L., Rosso, D., et al., 2009. SewerSnort: A Drifting Sensor for In-situ Sewer Gas Monitoring. IEEE, pp. 1–9. Kristjansson, J.K., Schönheit, P., Thauer, R.K., 1982. Different Ks values for hydrogen of methanogenic bacteria and sulfate reducing bacteria: an explanation for the apparent inhibition of methanogenesis by sulfate. Arch. Microbiol. 131 (3), 278–282. Lamontagne, R.A., Rose-Pehrsson, S.L., Grabowski, K.E., Knies, D.L., 2001. Response of METS sensor to methane concentrations found on the Texas-Louisiana Shelf in the Gulf of Mexico. DTIC Document. Liu, Y., Sharma, K.R., Murthy, S., Johnson, I., Evans, T., Yuan, Z., 2014. On-line monitoring of methane in sewer air. Sci. Rep. 4. Liu, Y., Ni, B.-J., Ganigué, R., Werner, U., Sharma, K.R., Yuan, Z., 2015a. Sulfide and methane production in sewer sediments. Water Res. 70, 350–359. Liu, Y., Sharma, K.R., Fluggen, M., O'Halloran, K., Murthy, S., Yuan, Z., 2015b. Online dissolved methane and total dissolved sulfide measurement in sewers. Water Res. 68, 109–118. Liu, Y., Sharma, K.R., Ni, B.-J., Fan, L., Murthy, S., Tyson, G.Q., et al., 2015c. Effects of nitrate dosing on sulfidogenic and methanogenic activities in sewer sediment. Water Res. 74, 155–165. Mohanakrishnan, J., Gutierrez, O., Sharma, K.R., Guisasola, A., Werner, U., Meyer, R.L., et al., 2009a. Impact of nitrate addition on biofilm properties and activities in rising main sewers. Water Res. 43 (17), 4225–4237. Mohanakrishnan, J., Sharma, K.R., Meyer, R.L., Hamilton, G., Keller, J., Yuan, Z.G., 2009b. Variation in biofilm structure and activity along the length of a rising main sewer. Water Environ. Res. 81 (8), 800–808. Phillips, N.G., Ackley, R., Crosson, E.R., Down, A., Hutyra, L.R., Brondfield, M., et al., 2013. Mapping urban pipeline leaks: methane leaks across Boston. Environ. Pollut. 173, 1–4. Pikaar, I., Rozendal, R.A., Yuan, Z., Rabaey, K., 2011. Electrochemical caustic generation from sewage. Electrochem. Commun. 13 (11), 1202–1204. Pikaar, I., Rozendal, R.A., Rabaey, K., Yuan, Z., 2013. In-situ caustic generation from sewage: the impact of caustic strength and sewage composition. Water Res. 47 (15), 5828–5835. Pikaar, I., Sharma, K.R., Hu, S., Gernjak, W., Keller, J., Yuan, Z., 2014. Reducing sewer corrosion through integrated urban water management. Science 345 (6198), 812–814. Scanlan, P., Elmendorf, H., Santha, H., Rowan, J., 2008. How green is your footprint the impact of greenhouse gas emissions on strategic planning. Proc. Water Environ. Fed. 2008 (6), 783–792. Schierbaum, K.D., Weimar, U., Göpel, W., 1992. Comparison of ceramic, thick-film and thin-film chemical sensors based upon SnO2. Sensors Actuators B Chem. 7 (1), 709–716. Schreiber, L., Holler, T., Knittel, K., Meyerdierks, A., Amann, R., 2010. Identification of the dominant sulfate-reducing bacterial partner of anaerobic methanotrophs of the ANME-2 clade. Environ. Microbiol. 12 (8), 2327–2340. Shah, A., Willis, J., Fillmore, L., 2011. Quantifying methane evolution from sewers: results from WERF/Dekalb Phase 2 continuous monitoring at Honey Creek Pumping Station and force main. Proc. Water Environ. Fed. 2011 (6), 475–485. Sharma, K.R., Yuan, Z., de Haas, D., Hamilton, G., Corrie, S., Keller, J., 2008. Dynamics and dynamic modelling of H2S production in sewer systems. Water Res. 42 (10), 2527–2538. Sharma, K., Ganigue, R., Yuan, Z., 2013. pH dynamics in sewers and its modeling. Water Res. 47 (16), 6086–6096. Souza, C., Chernicharo, C., Melo, G., 2012. Methane and hydrogen sulfide emissions in UASB reactors treating domestic wastewater. Water Sci. Technol. 65 (7), 1229–1237. Spencer, A.U., Noland, S.S., Gottlieb, L.J., 2006. Bathtub fire: an extraordinary burn injury. J. Burn Care Res. 27 (1), 97–98. Sturm, K., Yuan, Z., Gibbes, B., Werner, U., Grinham, A., 2014. Methane and nitrous oxide sources and emissions in a subtropical freshwater reservoir, South East Queensland, Australia. Biogeosciences 11 (18), 5245–5258. Sudarjanto, G., Sharma, K.R., Gutierrez, O., Yuan, Z., 2011. A laboratory assessment of the impact of brewery wastewater discharge on sulfide and methane production in a sewer. Water Sci. Technol. 64 (8), 1614–1619. Sun, Y., Liu, W., Zeng, Y., Wang, S., Huang, S., Xie, P., et al., 2011. Water vapor interference correction in a non dispersive infrared multi-gas analyzer. Chin. Phys. Lett. 28 (7), 073302. Sun, J., Hu, S., Sharma, K.R., Ni, B.-J., Yuan, Z., 2014. Stratified microbial structure and activity in sulfide-and methane-producing anaerobic sewer biofilms. Appl. Environ. Microbiol. 80 (22), 7042–7052. Tsunogai, U., Kou, S., Kimoto, H., Kimoto, T., 2007. Development of in-situ methane sensor in seawater. 51th Annual Meeting of the Geochemical Society of Japan. Valentine, D.L., Reeburgh, W.S., 2000. New perspectives on anaerobic methane oxidation. Environ. Microbiol. 2 (5), 477–484. Waki, M., Suzuki, K., Osada, T., Tanaka, Y., 2005. Methane-dependent denitrification by a semi-partitioned reactor supplied separately with methane and oxygen. Bioresour. Technol. 96 (8), 921–927. Wang, J., Zhang, J., Xie, H., Qi, P., Ren, Y., Hu, Z., 2011. Methane emissions from a full-scale A/A/O wastewater treatment plant. Bioresour. Technol. 102 (9), 5479–5485.

Y. Liu et al. / Science of the Total Environment 524–525 (2015) 40–51 Whitman, W.B., Bowen, T.L., Boone, D.R., 1999. The methanogenic bacteria. In: Dworkin, et al. (Eds.), The Prokaryotes, an Evolving Electronic Resource for the Microbiological Community. Release 3.0. Springer-Verlag, New York. Zhang, L., De Schryver, P., De Gusseme, B., De Muynck, W., Boon, N., Verstraete, W., 2008. Chemical and biological technologies for hydrogen sulfide emission control in sewer systems: a review. Water Res. 42 (1–2), 1–12.

51

Zhang, L., Keller, J., Yuan, Z., 2009. Inhibition of sulfate-reducing and methanogenic activities of anaerobic sewer biofilms by ferric iron dosing. Water Res. 43 (17), 4123–4132. Zhang, J., Zhang, Y., Quan, X., Liu, Y., An, X., Chen, S., et al., 2011. Bioaugmentation and functional partitioning in a zero valent iron-anaerobic reactor for sulfate-containing wastewater treatment. Chem. Eng. J. 174 (1), 159–165.