Wastewater treatment plants can be significant sources of nitrous oxide (N2O), a potent greenhouse gas. While our understanding of N2O emissions from suspended-growth processes has advanced significantly, less is known about emissions from biofilm processes. Biofilms may behave differently due to their substrate gradients and microbial stratification. In this study, we used mathematical modeling to explore the mechanisms of N2O emissions from nitrifying and denitrifying biofilms. Our ammonia-oxidizing bacteria biofilm model suggests that N2O emissions from biofilm can be significantly greater than from suspended-growth systems. The driving factor is the diffusion of hydroxylamine, a nitrification intermediate, from the aerobic to the anoxic regions of the biofilm. The presence of nitrite-oxidizing bacteria further increased emissions. For denitrifying biofilms, our results suggest that emissions are generally greater than for suspended-growth systems. However, the magnitude of the difference depends on the bulk dissolved oxygen, chemical oxygen demand, and nitrate concentrations, as well as the biofilm thickness. Overall, the accumulation and diffusion of key intermediates, i.e. hydroxylamine and nitrite, distinguish biofilms from suspended-growth systems. Our research suggests that the mechanisms of N2O emissions from biofilms are much more complex than suspended-growth systems, and that emissions may be higher in many cases.

INTRODUCTION

Nitrous oxide (N2O) is a potent greenhouse gas, with a global warming potential 300 times greater than that of CO2 (IPCC 2013). Wastewater treatment plants can be an important source of N2O (Ahn et al. 2010). Emissions from wastewater may result from incomplete denitrification (Kampschreur et al. 2009; Lu & Chandran 2010; Pan et al. 2012), but also from nitrification (Tallec et al. 2006; Kampschreur et al. 2008, Kampschreur et al. 2009; Lu & Chandran 2010; Wunderlin et al. 2012; Ye et al. 2014; Daelman et al. 2015). During nitrification, N2O emissions may result from nitrifier denitrification as well as chemical degradation of hydroxylamine (NH2OH) (Schreiber et al. 2009, 2012; Harper et al. 2015; Soler-Jofra et al. 2016).

While extensive research has investigated N2O formation in suspended-growth systems (Kimochi et al. 1998; Colliver & Stephenson 2000; Kampschreur et al. 2008; Ahn et al. 2010; Lu & Chandran 2010; Aboobakar et al. 2013), few studies have explored its formation in biofilm-based processes, such as the moving bed biofilm reactor, biological aerated filter, and granular sludge. Such processes have been gaining in popularity, and therefore it is important to understand their potential for N2O emissions.

Bacteria in suspended-growth systems are directly exposed to the bulk liquid. Thus, the formation of N2O depends exclusively on the conditions in the bulk environment. For example, little or no N2O formation would be expected from denitrifying bacteria if the bulk is fully aerobic, as denitrification would be inhibited by O2. In a biofilm, however, O2 gradients exist. Even if the bulk liquid were aerobic, bacteria in the deeper biofilm could experience anoxic conditions, allowing nitrate reduction and the formation of N2O.

A number of models have been developed to predict N2O formation by nitrifying (Kampschreur et al. 2007, 2008; Ni et al. 2011; Mampaey et al. 2013; Ni et al. 2013, 2014) and denitrifying microorganisms (Hiatt & Grady 2008; Ni et al. 2011; Kampschreur et al. 2012; Pan et al. 2013). In particular, recent models have improved the prediction of N2O by explicitly considering formation and consumption of nitrification and denitrification intermediates, and modeling the competition of key enzymes for intracellular electron mediators (Pan et al. 2013; Ni et al. 2014; Ni & Yuan 2015). A recent study used such a model to predict the mechanisms of N2O formation from nitrifying biofilms containing ammonia-oxidizing bacteria (AOB) (Sabba et al. 2015). However, this study did not consider the effects of nitrite-oxidizing bacteria (NOB) in nitrifying biofilms, nor did it assess the effects of gas stripping, e.g. via aeration. More importantly, there are no studies addressing the mechanisms of N2O formation from denitrifying biofilms.

The objective of this study was to use mathematical modeling to systematically explore N2O production and emissions from nitrifying and denitrifying biofilms. Note that we use the words N2O production (the net formation of N2O from the biofilm) and emissions (loss of N2O from the reactor in liquid and gas phases) interchangeably in this paper. Also, the intent was not to accurately predict N2O emissions, but to explain how the mechanisms of N2O formation in biofilm differ from those from in suspended growth systems. We assessed N2O emissions from nitrifying biofilms consisting solely of AOB, or AOB plus NOB. Denitrifying biofilms were studied separately, to clearly establish the mechanisms of N2O formation by each population.

METHODS

The biofilm models used to predict N2O production in biofilms were based on traditional diffusion-reaction mass balances for the relevant chemical species in both nitrifying and denitrifying biofilms.

The nitrifying model considered N2O formation by AOB via two pathways: the hydroxylamine (NH2OH) pathway and the nitrifier denitrification pathways. This approach is based on a recently published model (Ni et al. 2014; Sabba et al. 2015). While in our previous work we focused on the mechanisms of N2O formation in biofilms consisting exclusively of AOB (Sabba et al. 2015), in this work we expanded on the previous work and explored the effects of gas stripping and of the combined presence of AOB and NOB within the biofilm. Furthermore, we studied the mechanisms of N2O emissions in denitrifying biofilms, with and without gas stripping.

Parameters for the nitrification model are reported in Table 1. While a base AOB density, assuming uniform distribution of AOB, was used in most studies, different biomass densities were also tested. For additional tests with AOB along with a constant and uniformly distributed population of NOB, the NOB were simulated using the conventional ASM model, i.e. without electron mediators (Picioreanu et al. 1997). The denitrifying model included N2O formation by heterotrophic bacteria, and was adapted from Pan et al. (2013).

Table 1

Parameters used for the nitrification model

Parameter Symbol Value Units Source 
Concentrations in influent 
 Oxygen Cin,O2 from 0.001 to 5 (varied) mg L−1 Typical range 
 Ammonia Cin,NH3 80 mgN L−1 Chosen 
 Hydroxylamine, nitrous oxide, nitric oxide, nitrite, nitrate Cin,i mgN L−1 Chosen 
Initial concentrations C0,i Cin,i mgO2 L−1 Chosen 
Biomass concentration in the biofilm 
 Ammonia oxidizers, AOB CF,XAOB 50 (base case) 50 or 35 (with NOB) g L−1 Typical value, Wanner et al. (2006)  
 Nitrite oxidizers, NOB CF,XNOB 0 (base case) 15 (with NOB) g L−1 Picioreanu et al. (1997)  
Concentration total redox mediators CT,Med 0.01 mol kg−1 Ni et al. (2014)  
Biofilm thickness LF 100 (base case) 2, 50, 100 (varied) μm Typical values 
Liquid flow rate Q 11 mL min−1 Chosen 
Liquid volume in the reactor VB Chosen 
Biofilm surface area AF 0.5 m2 Chosen 
Gas volume VG 3.5 Chosen 
Gas flow rate QG L min−1 Chosen 
Gas-liquid mass transfer coeff. kLa 100 h−1 Chosen 
Henry gas-liquid coefficient N2O (25 °C) HN2O 0.611 mol mol−1 CRC Handbook (2014)  
Parameter Symbol Value Units Source 
Concentrations in influent 
 Oxygen Cin,O2 from 0.001 to 5 (varied) mg L−1 Typical range 
 Ammonia Cin,NH3 80 mgN L−1 Chosen 
 Hydroxylamine, nitrous oxide, nitric oxide, nitrite, nitrate Cin,i mgN L−1 Chosen 
Initial concentrations C0,i Cin,i mgO2 L−1 Chosen 
Biomass concentration in the biofilm 
 Ammonia oxidizers, AOB CF,XAOB 50 (base case) 50 or 35 (with NOB) g L−1 Typical value, Wanner et al. (2006)  
 Nitrite oxidizers, NOB CF,XNOB 0 (base case) 15 (with NOB) g L−1 Picioreanu et al. (1997)  
Concentration total redox mediators CT,Med 0.01 mol kg−1 Ni et al. (2014)  
Biofilm thickness LF 100 (base case) 2, 50, 100 (varied) μm Typical values 
Liquid flow rate Q 11 mL min−1 Chosen 
Liquid volume in the reactor VB Chosen 
Biofilm surface area AF 0.5 m2 Chosen 
Gas volume VG 3.5 Chosen 
Gas flow rate QG L min−1 Chosen 
Gas-liquid mass transfer coeff. kLa 100 h−1 Chosen 
Henry gas-liquid coefficient N2O (25 °C) HN2O 0.611 mol mol−1 CRC Handbook (2014)  

A continuous, ideally-mixed biofilm reactor incorporating nitrifying or heterotrophic bacteria was modeled. The two separate models evaluated one-dimensional, planar biofilms. A hydraulic retention time of 6 hours for the nitrifying and 1.5 hours for the denitrifying condition was used. As initial values, all concentrations in biofilm and bulk liquid were taken as equal to the corresponding influent concentrations. As base condition, a biofilm specific surface area of 125 m2 m−3 was used. Biomass growth, decay, attachment and detachment were not considered. Biofilms of different thicknesses were modeled. Thicknesses of 2 μm for the nitrifying and 5 μm for the denitrifying biofilm were assumed to represent ‘suspended growth’. While these thicknesses were chosen arbitrarily, they both had essentially had no substrate gradients within the depth and therefore behaved as suspended growth.

For the denitrification process, O2, nitrite (NO2), nitric oxide (NO), N2O, nitrate (NO3) and chemical oxygen demand (COD) were included as state variables. Note that the COD is assumed to be readily biodegradable. For the nitrification process, the model additionally considered ammonia (NH3) and NH2OH as state variables, but did not include COD. The conditions tested in both models are listed in Tables 1 and 2. All model equations and process matrices are provided in the Supporting Information in Tables S1–S3 for the nitrifying model and Tables S4–S6 for the denitrifying model (the Supporting Information is available with the online version of this paper). The denitrification model was used to predict the effects of bulk O2 and NO3 concentrations on N2O production in denitrifying biofilms, assuming an influent NO3 concentration of 14 mgN L−1. The bulk COD concentration was 720 mgCOD L−1, such that COD was not rate limiting within the biofilm. The assessed biofilm thicknesses were 5, 50 and 400 μm. In the denitrification model, we added an O2 reduction process with a high maximum reduction rate, qmax, and a very high relative affinity for Mred such that O2 reduction was prioritized over denitrification. This novel approach to modeling O2 inhibition guarantees that, as long as O2 is present, it will keep Mred at very low levels and inhibit the reduction of nitrogen oxides. This approach is a more fundamental alternative to the conventional ‘oxygen switch’ used in the ASM models, and allows the distinct inhibitory effect of O2 on each enzyme to be included via the Mred concentration.

Most modeling runs were without N2O stripping. However, for some runs we explored the effects of stripping on N2O emissions. As NO usually does not accumulate and acts as a transient compound, its stripping was not included in the model. To simulate N2O stripping during aeration, an additional transfer term was included in the liquid N2O mass balance as , and a further equation was solved for the gas phase concentration, CG,N2O (mol/m3 gas), as dCG,N2O/dt = QG/VG(0−CG,N2O)−kLa(CG,N2OHN2OCB,N2O).

The model was implemented on the COMSOL Multiphysics platform. Equations for one-dimensional diffusion and reaction, for a fixed biofilm density and thickness, were solved with variable time step on a biofilm domain discretized with a mesh size of 1 μm. Steady state was assumed to be reached when effluent concentrations were stable. Steady state for all conditions was obtained after maximum simulation time of three days for nitrifying biofilms and one day for heterotrophic biofilms.

A summary of the nitrifying and denitrifying conditions used for the model can be found in Tables 1 and 2, respectively. A complete list of stoichiometric matrices, reaction rates and other model parameters for both models can be found in the Supporting Information (Tables S1–S6).

Table 2

Parameters used for the denitrification model

Parameter Symbol Value Units Source 
Concentrations in influent 
 Methanol (as COD) C0,COD 720 (non-limiting) mg L−1 Chosen 
 Nitrate C0,NO3 Range 0.0001–50 mg L−1 Varied 
 Nitrite C0,NO2 mgN L−1 Chosen 
 Nitric oxide C0,NO mgN L−1 Chosen 
 Nitrous oxide C0,N2O mgN L−1 Chosen 
 Oxygen C0,O2 Range 0.00001–4 mgO2 L−1 Varied 
Biomass concentration in the biofilm 
 Concentration total redox mediators CT,Med 0.01 mol kg−1 Pan et al. (2013)  
 Biofilm thickness LF 400 (base case) 5, 50, 400 (varied) μm Typical values 
 Liquid flow rate Q 44 mL min−1 Chosen 
 Liquid volume in the reactor Vb Chosen 
 Biofilm surface area AF 0.5 m2 Chosen 
 Gas volume VG 3.5 Chosen 
 Gas flow rate QG L min−1 Chosen 
 Gas-liquid mass transfer coeff. kLa 100 h−1 Chosen 
 Henry gas-liquid coefficient N2O (25 °C) HN2O 0.611 mol mol−1 CRC Handbook (2014)  
Parameter Symbol Value Units Source 
Concentrations in influent 
 Methanol (as COD) C0,COD 720 (non-limiting) mg L−1 Chosen 
 Nitrate C0,NO3 Range 0.0001–50 mg L−1 Varied 
 Nitrite C0,NO2 mgN L−1 Chosen 
 Nitric oxide C0,NO mgN L−1 Chosen 
 Nitrous oxide C0,N2O mgN L−1 Chosen 
 Oxygen C0,O2 Range 0.00001–4 mgO2 L−1 Varied 
Biomass concentration in the biofilm 
 Concentration total redox mediators CT,Med 0.01 mol kg−1 Pan et al. (2013)  
 Biofilm thickness LF 400 (base case) 5, 50, 400 (varied) μm Typical values 
 Liquid flow rate Q 44 mL min−1 Chosen 
 Liquid volume in the reactor Vb Chosen 
 Biofilm surface area AF 0.5 m2 Chosen 
 Gas volume VG 3.5 Chosen 
 Gas flow rate QG L min−1 Chosen 
 Gas-liquid mass transfer coeff. kLa 100 h−1 Chosen 
 Henry gas-liquid coefficient N2O (25 °C) HN2O 0.611 mol mol−1 CRC Handbook (2014)  

RESULTS AND DISCUSSION

Nitrifying biofilms

Effect of O2 and thickness on N2O emissions

We first explored N2O emissions from an AOB biofilm as a function of bulk O2 for biofilm thicknesses of 2, 50, and 100 μm. We then selected one biofilm thickness, 100 μm, and analyzed its behavior in more detail. Finally we evaluated the effects of NOB on the overall N2O emissions.

The nitrifying model suggests that thicker biofilms have greater N2O emissions than thin biofilms, which represent suspended-growth systems (Figure 1(a)). A range of thicknesses was simulated. The emission rates increased with increasing in O2. However, this behavior was different for thinner biofilms and suspended growth systems, where N2O reached its maximum at much lower O2 levels than for thicker biofilms. Biofilms with greater thicknesses (e.g. 50 and 100 μm) followed similar general trends with regards to N2O emissions (Figure 1(a)). This trend confirmed that thicker biofilms not only had higher emissions, but also had N2O emissions for a much wider range of O2 values. The main cause is the diffusion of NH2OH, an AOB nitrification intermediate. Diffusion of reaction intermediates has been previously shown (De Beer et al. 1997; Stewart 2003; Sabba et al. 2015). NH2OH forms in the outer, aerobic regions of the biofilm and diffuses to the inner, anoxic regions of the biofilm (Figure 1(b)). The higher emissions for thicker biofilms occurred on a basis of a higher biomass content for biofilms, respectively.
Figure 1

(a) N2O production rates for AOB biofilms of different thicknesses with a constant biofilm surface area, per unit reactor volume, as a function of bulk O2, and (b) net component rates over the biofilm depth (x) for the 100-μm biofilm. Results are for the base case conditions at a bulk O2 of 0.9 mg L−1.

Figure 1

(a) N2O production rates for AOB biofilms of different thicknesses with a constant biofilm surface area, per unit reactor volume, as a function of bulk O2, and (b) net component rates over the biofilm depth (x) for the 100-μm biofilm. Results are for the base case conditions at a bulk O2 of 0.9 mg L−1.

To better understand and explore the mechanisms that lead to N2O formation, a base case of a 100-μm biofilm was considered (Figure 1(b)). Figure 1(b) shows the net rates of formation or consumption of nitrifying key species and O2. In a suspended-growth system at steady state, the rate of NH3 oxidation should equal the rate of NH2OH oxidation. In biofilms, however, some NH2OH may diffuse into the deeper portions of the biofilm, resulting in a net formation of NH2OH in the outer biofilm and net consumption in the interior (Figure 1(b)).

The external portion of the biofilm (right side of Figure 1(b)) has high nitrification rates due to the high concentrations of NH3 and O2. This can be seen in Figure 1(b), where there is net consumption of both compounds and net formation of NH2OH and NO2 as products. Essentially, all of the electrons from NH2OH oxidation are utilized for O2 reduction in this zone, allowing little NO2 reduction. However, at greater depths, around 60 μm, O2 becomes limiting and the rate of NH3 consumption approaches zero. In this zone, little NH3 reduction takes place, but electrons produced from NH2OH that diffuses from the outer layers are used for NO2 reduction, leading to a spike in N2O formation. Below 30 μm, NH2OH is no longer available and the rate of N2O formation decreases to zero. This is also true for larger thicknesses where different biomass concentrations might be present. With increasing thicknesses, the inner portions become inactive, while the amount of active biomass close to the bulk liquid remains similar. The NH2OH pathway contributed only to a small extent to the N2O overall production, while the nitrifier denitrification pathway was the main contributor for most of the N2O produced. These results are similar to those found by Sabba et al. (2015).

Effect of NOB on N2O emissions

Sabba et al. (2015) studied a biofilm consisting solely of AOB. However, nitrifying biofilms commonly typically include both AOB and NOB. While NOB do not directly produce N2O, they may affect N2O formation by AOB by modifying the surrounding environment.

If the total density of AOB only is 50 g L−1 (base case), and NOB provide an additional 15 g L−1 density for a total of 65 g L−1, the N2O emissions increase with respect to the base case (Figure 2). This is because a higher overall biomass density of AOB and NOB leads to higher O2 gradients in the biofilm. This promotes a higher gradient of NH3 oxidation rates and O2 concentrations, leading to greater diffusion of NH2OH into the deeper biofilm. It also contributes to the formation of an anoxic zone within the biofilm depth.
Figure 2

Emissions from base case 100 μm nitrifying biofilm with AOB and NOB. The presence of nitrite oxidizers (NOB) in the biofilm may enhance the N2O production. N2O production rate without NOB (base case, AOB 50 g L−1) and with 15 g L–1 NOB (AOB 50 or 35 g L−1).

Figure 2

Emissions from base case 100 μm nitrifying biofilm with AOB and NOB. The presence of nitrite oxidizers (NOB) in the biofilm may enhance the N2O production. N2O production rate without NOB (base case, AOB 50 g L−1) and with 15 g L–1 NOB (AOB 50 or 35 g L−1).

Interestingly, even if the AOB density drops to 35 g L−1, maintaining a total biofilm density of 50 g L−1, the N2O formation rates with NOB are higher than if the biofilm is exclusively composed of AOB. This is because NOB have a higher specific rate of oxygen consumption in our model. Based on these considerations, the presence of NOB in a nitrifying biofilm may actually increase N2O emissions.

Effects of gas stripping on N2O emissions from nitrifying biofilms

In this section, we evaluated the effects of gas flow on both suspended growth (modeled as thin biofilms) and biofilm systems. Results are shown in Figure 3.
Figure 3

N2O bulk liquid concentration for nitrifying conditions as a function of bulk O2 for a 2 μm suspended growth and 100 μm biofilm system, with stripping (solid line) and without stripping (dashed line).

Figure 3

N2O bulk liquid concentration for nitrifying conditions as a function of bulk O2 for a 2 μm suspended growth and 100 μm biofilm system, with stripping (solid line) and without stripping (dashed line).

Note that our model did not include NO stripping due to aeration. NO stripping can reduce N2O formation, as NO is a precursor to N2O. However, our preliminary simulations show that NO is mostly converted to N2O within the biofilm. NO stripping is more significant for thin biofilms, but these produce little NO due to the lack of substrate gradients.

Research has shown that N2O can be stripped from suspended growth systems (Rassamee et al. 2011; Law et al. 2012; Wu et al. 2014). Including stripping simply shifts N2O emissions from the liquid phase to the gas phase (Figure 3). When stripping was included, the N2O concentration in the liquid phase, for both suspended and biofilm systems, decreased to near-zero levels (Figure 3). But this did not impact N2O formation rates, since stripping was not linked to aeration and the O2 concentration remained constant (data not shown). This situation is different for denitrifying bacteria (below), as stripping and biological reduction are competing processes, i.e. more stripping leads to less biological reduction in the biofilm.

Denitrifying biofilms

Effect of O2 and NO3 on N2O emissions

At low NO3 concentrations, emissions from the 5-μm biofilm were higher than the thicker biofilms, but they quickly reached a low maximum rate of N2O production (Figure 4(a)). This is due to full penetration of NO3. The lower emissions at lower NO3 concentrations in the thicker biofilms are explained by the partial NO3 penetration within the biofilm depth. Thicker biofilms require higher NO3 concentrations to reach maximum denitrification rates throughout the biofilms. For greater biofilm thicknesses, the higher biomass concentration accounted for a higher rate of N2O formation.
Figure 4

N2O production rates for denitrifying biofilms of different thicknesses with a constant biofilm surface area, per unit reactor volume and time, as a function of bulk NO3 (a) and bulk O2 (b). Net component rates over the biofilm depth (x). Results are for the base case conditions with anoxic bulk conditions (net rates of component formation or consumption) (c) and biofilm N-species concentration within biofilm depth (d). Results in (c) and (d) are for the base case conditions with a 400 μm biofilm in the presence of anoxic bulk.

Figure 4

N2O production rates for denitrifying biofilms of different thicknesses with a constant biofilm surface area, per unit reactor volume and time, as a function of bulk NO3 (a) and bulk O2 (b). Net component rates over the biofilm depth (x). Results are for the base case conditions with anoxic bulk conditions (net rates of component formation or consumption) (c) and biofilm N-species concentration within biofilm depth (d). Results in (c) and (d) are for the base case conditions with a 400 μm biofilm in the presence of anoxic bulk.

N2O emissions from ‘suspended growth’ and biofilm systems were assessed for different O2 bulk concentrations (Figure 4(b)). At low bulk O2 concentrations, the amount of N2O produced per unit reactor volume for the 5-μm ‘suspended growth’ scenario was higher than the thicker biofilms. Specifically, the 5-μm scenario reached its maximum N2O emissions at near-zero O2 concentrations. The N2O emissions then dropped steeply around 0.1 mgO2 L−1 and approached zero at around 0.3 mg O2 L−1. For the 400-μm biofilm, the emissions of N2O were slightly higher at low bulk O2 concentrations and then had minimal decrease with increasing O2. This is due to the O2 consumption of oxygen in the outer layers, allowing denitrification in the inner layers. In a similar fashion to Figure 4(a) greater biofilm thicknesses, the higher volumetric biomass concentration accounted for higher rate of N2O formation.

Finally, to better evaluate the mechanisms of N2O formation in denitrifying biofilms, a base biofilm thickness of 400 μm was considered in more detail (Figure 4(c)). For the 400-μm biofilm with an anoxic bulk, the bulk NO2 and N2O concentrations were 0.12, and 0.07 mgN L−1 (Figure 4(d)), respectively (data not shown). When both the NO3 concentration and the rate of NO3 reduction start to decrease, at around 300 μm, the NO2 reduction rate starts to increase and there is a net production of N2O around 280 μm (data not shown). The inner portion of the biofilm, around 250 μm, has a low concentration of NO3 and a higher net NO2 consumption rate (Figure 4(c)). In the last region of the biofilm, where NO3 is mainly depleted and NO2 is at low concentration, the N2O consumption rate leads the process rates and uses the available electron mediators to reduce N2O to N2. Thus, this region is a net sink for N2O produced in other regions of the biofilm or the bulk.

Effects of stripping on N2O emissions

The effects of gas stripping were evaluated for a 5-μm system, representing suspended growth, and a 400-μm biofilm system (Figure 5(a)5(d)). Stripping may occur due to low levels of aeration or due to N2 gas production. In Figure 5(a)5(d), we compared N2O production rates with and without stripping as a function of bulk NO3 for both suspended growth and biofilm systems. All these scenarios were tested for non-limiting COD conditions.
Figure 5

N2O production rates in denitrifying biofilms with a constant biofilm surface area, with and without stripping, for a 5-μm biofilm (‘suspended growth’) (a), (c) and 400-μm biofilm (b), (d) as a function of bulk NO3 (a), (b) and (c), (d) as a function of bulk O2.

Figure 5

N2O production rates in denitrifying biofilms with a constant biofilm surface area, with and without stripping, for a 5-μm biofilm (‘suspended growth’) (a), (c) and 400-μm biofilm (b), (d) as a function of bulk NO3 (a), (b) and (c), (d) as a function of bulk O2.

For the 5-μm ‘suspended growth’ biofilm, the maximum N2O production rate was reached at very low NO3 concentrations (e.g. 0.1 mgN L−1) (Figure 5(a)). This is because the thin biofilm becomes fully penetrated and saturated by NO3 at relatively low bulk concentrations. Stripping increases the N2O emission by maintaining low bulk N2O concentrations, which allows the bulk liquid to become a sink for N2O. Thicker biofilms require a higher NO3 concentration to reach the maximum bulk N2O concentration (Figure 5(b)). Stripping has a greater effect on thicker biofilms (Figure 5(b)).

The effects of gas stripping on N2O emissions as a function of bulk O2 were also evaluated (Figure 5(c) and 5(d)). For suspended growth systems (5 μm biofilm) without stripping, no gaseous N2O emissions occurred throughout the sweep of O2 concentration for both biofilm and suspended growth. Higher emissions were found when stripping was included for both biofilm and suspended growth (Figure 5(c) and 5(d)). Emissions in this case represent the net emissions; comparing emissions by the formation rate in denitrifying systems would not allow a fair comparison, since denitrifying systems might both produce and consume N2O. For suspended growth systems, emissions were higher at low O2. However, they dropped suddenly after the O2 concentration reached 0.1 mg L−1 where a full inhibition of the suspended growth occurred (Figure 5(c)). No emissions occurred afterwards. Biofilms performed differently (Figure 5(d)). They tended to have higher emissions throughout the O2 sweep. Emissions were highest at low O2, and decreased when O2 increased.

CONCLUSIONS

Our model suggests that N2O emissions from both nitrifying and denitrifying biofilms behave differently from suspended growth systems. In suspended growth systems, all bacteria are exposed to the same bulk concentrations of substrates and intermediates.

As found previously, NH2OH formed in an aerobic zone of a nitrifying biofilm diffuses to an anoxic zone, resulting in a spike in N2O formation rates and higher N2O emissions. However, a novel aspect of this study is that the presence of NOB can also enhance emissions. This was due to the high rate of O2 reduction by NOB leading to an increase in the O2 gradient within the biofilm.

Diffusion of intermediates was also important for the denitrifying biofilm, where NO3 and NO2 reduction govern the activity in the outer portion of the biofilm. Thus, the inner portion of the biofilm has lower concentrations of both compounds, and these can diffuse and be consumed elsewhere. This same aspect applies to N2O, which can both be exported to the bulk and diffuse towards the deeper regions of a denitrifying biofilm and be reduced.

For denitrifying systems, gas stripping increased emissions by decreasing the amount of N2O available for reduction in the deeper, anoxic regions of the biofilm. An increase in influent flow rate mimics stripping effects, creating a more pronounced gradient between biofilms and the bulk environment, leading to higher emissions.

These results identify important mechanisms that affect N2O emissions in nitrifying and denitrifying biofilms. Future research should address the behavior of biofilms containing both nitrifying and denitrifying bacteria, mimicking simultaneous nitrification and denitrification systems.

ACKNOWLEDGEMENTS

F.S. and R.N. were supported by NSF project CBET0954918 (Nerenberg CAREER award) and WERF project U2R10. F.S received additional support from the Bayer Corporation Fellowship. C.P. work was supported by a Melchor Visiting Professor grant from University of Notre Dame. These results were presented at the 5th IWA/WEF Wastewater Treatment Modelling Seminar 2016, Annecy, France (WWTmod2016) seminar and the fruitful discussions are kindly acknowledged.

REFERENCES

REFERENCES
CRC Handbook of Chemistry and Physics
2014
95th edn
.
Section 5: Solubility of selected gases in water. Online at http://www.hbcpnetbase.com/.
Daelman
M. R.
van Voorthuizen
E. M.
van Dongen
U. G.
Volcke
E. I.
van Loosdrecht
M. C.
2015
Seasonal and diurnal variability of N2O emissions from a full-scale municipal wastewater treatment plant
.
Sci. Total Environ.
1
(
536
):
1
11
.
De Beer
D.
Schramm
A.
Santegoeds
C. M.
Kuhl
M.
1997
A nitrite microsensor for profiling environmental biofilms
.
Appl. Environ. Microbiol.
63
(
3
),
973
977
.
IPCC
2013
Climate Change 2013: the Physical Science Basis
.
Working Group I Contribution to the IPCC 5th Assessment Report
.
IPCC
,
Cambridge
,
UK
and New York, NY, USA
.
Kampschreur
M. J.
Tan
N. C.
Kleerebezem
R.
Picioreanu
C.
Jetten
M. S.
van Loosdrecht
M. C. M.
2007
Effect of dynamic process conditions on nitrogen oxides emission from a nitrifying culture
.
Environ. Sci. Technol.
42
,
429
435
.
Kampschreur
M. J.
van der Star
W. R. L.
Wielders
H. A.
Mulder
J. W.
Jetten
M. S. M.
van Loosdrecht
M. C. M.
2008
Dynamics of nitric oxide and nitrous oxide emission during full-scale reject water treatment
.
Water Res.
42
(
3
),
812
826
.
Kampschreur
M. J.
Temmink
H.
Kleerebezem
R.
Jetten
M. S. M.
van Loosdrecht
M. C. M.
2009
Nitrous oxide emission during wastewater treatment
.
Water Res.
43
,
4093
4103
.
Kampschreur
M. J.
Kleerebezem
R.
Picioreanu
C.
Bakken
L.
Bergaust
L.
de Vries
S.
Jetten
M. S. M.
van Loosdrecht
M. C. M.
2012
Metabolic modeling of denitrification in Agrobacterium tumefaciens a tool to study inhibiting and activating compounds for the denitrification pathway
.
Frontiers in Microbiol
3
,
370
.
Kimochi
Y.
Inamori
Y.
Mizuochi
M.
Xu
K.
Matsumura
M.
1998
Nitrogen removal and N2O emission in a full-scale domestic wastewater treatment plant with intermittent aeration
.
J. Ferment. Bioeng.
86
(
2
),
202
206
.
Mampaey
K. E.
Beuckels
B.
Kampschreur
M. J.
Kleerebezem
R.
van Loosdrecht
M. C. M.
Volcke
E. I. P.
2013
Modelling nitrous and nitric oxide emissions by autotrophic ammonia-oxidizing bacteria
.
Environ. Technol.
34
,
1555
1566
.
Ni
B. J.
Yuan
Z.
Chandran
K.
Vanrolleghem
P. A.
Murthy
S.
2013
Evaluating four mathematical models for nitrous oxide production by autotrophic ammonia-oxidizing bacteria
.
Biotechnol. Bioeng.
110
,
153
163
.
Picioreanu
C.
Van Loosdrecht
M. C. M.
Heijnen
J. J.
1997
Modeling the effect of oxygen concentration on nitrite accumulation in a biofilm airlift suspension reactor
.
Water Sci. Technol.
36
(
1
),
147
156
.
Sabba
F.
Picioreanu
C.
Perez
J.
Nerenberg
R.
2015
Hydroxylamine diffusion can enhance N2O emissions in nitrifying biofilms: a modeling study
.
Environ. Sci. Technol.
49
(
3
),
1486
1494
.
Schreiber
F.
Loeffler
B.
Polerecky
L.
Kuypers
M. M. M.
de Beer
D.
2009
Mechanisms of transient nitric oxide and nitrous oxide production in a complex biofilm
.
ISME J.
3
,
1301
1313
.
Soler-Jofra
A.
Stevensa
B.
Hoekstraa
M.
Picioreanu
C.
Sorokin
D.
van Loosdrecht
M. C. M.
Pérez
J.
2016
Importance of abiotic hydroxylamine conversion on nitrous oxide emissions during nitritation of reject water
.
Chem. Eng. J.
287
,
720
726
.
Stewart
P. S.
2003
Diffusion in biofilms
.
J. Bacteriol.
185
(
5
),
1485
1491
.
Wanner
O.
Eberl
H. J.
Morgenroth
E.
Noguera
D.
Picioreanu
C.
Rittmann
B. E.
Van Loosdrecht
M. C. M.
2006
IWA Task Group on Biofilm Modeling. Mathematical Modeling of Biofilms.
IWA Scientific and Technical Report No. 18
,
IWA Publishing
.

Supplementary data