Assessment of Activity of Bacteria in Integrated Fixed Film Activated Sludge

In this study, the feasibility of a manometric batch test method to measure biological activity of Integrated Fixed Film Activated Sludge (IFAS) microbial aerobic and anaerobic communities was investigated. Additionally, the substrate consumption ratio, the N2O emissions from the biological activity of the different microbial populations and the inhibitory effect of stormwater pollutants on the activity and N2O production were investigated as well. The obtained results from the aerobic tests showed qualitative correspondence with trends described in the literature, but differed greatly in quantitative terms (1 to 2 orders of magnitude). The anoxic test did not produce interpretable results, because values recorded with the manometric method could not be transformed using the method that had been destined for the transformation, and the results were contradictory to what was depicted in the literature. The stormwater toxicity test results were scattered so that an interpretation did not seem feasible, because the values for the experimental duplicates varied so largely that no larger pattern could be established. The trend of the results obtained for the N2O production agree with previous reports, however, because of the unreliability of the fluid analysis results (for example in terms of N2O production per nitrogen) mass balances to corroborate them were not possible to achieve. Overall the experiments did not provide the results that were expected and significant improvements to the methods and a further investigation of the influencing factors are necessary to ensure that the proposed method provide more accurately results.


INTRODUCTION
Wastewater treatment facilities remove pollutants and nutrients, before the contaminated stream is released into receiving water bodies, minimizing impacts in the environment. This is achieved through a chain of physical, biological and chemical treatments. The heart of most Wastewater Treatment Plants (WWTP's) treating municipal wastewater is the biological (or secondary) treatment stage (see Figure 1).
In the secondary treatment units, conditions are established to support biological processes in aerobic (aerated) and anoxic zones. In the biological treatment, most of the nutrients, such as carbon, nitrogen and phosphorus compounds are removed from the wastewater by microorganisms that use these compounds as source of energy and matter for their cell metabolism and growth. In the aerobic zones, the microorganisms use the oxygen that is supply through aeration, to oxidize the substrates (carbon to CO2 and ammonia to NO3) and in the anoxic zones facultative bacteria reduce NO2 and NO3 to N2 when using them as electron acceptors to respire organic carbons.
In coastal areas, nitrogen is the limiting nutrient for the growth of nuisance algae that can cause eutrophication. Because of eutrophication, low oxygen zones can occur, which have led to fish kills, closing of beaches and fishing grounds [1], [2]. In order to prevent these issues, the United State Environmental Protection Agency (USEPA) implemented programs with states to issue increasingly strict regulations for the nitrogen concentration of WWTP's effluent [3]. During the last decades, several alternatives to enhance nitrogen removal have been developed. One of these technologies is the Integrated Fixed film Activated Sludge (IFAS) system [4], which is a hybrid process that increases the nitrification capacity by providing support media for nitrifying bacteria to grow along with suspended biomass in the aeration tank of WWTP (see section IFAS for more details).
This study was conducted in cooperation with the Narragansett Bay Commission (NBC), which is especially interested in high performing nitrogen removal processes for their two WWTPs, which are the largest in the State of Rhode Island and are located on the northern end of the Narraganset Bay. Due to the upcoming repermitting of the plant, it is anticipated that stricter effluent standards for pollutants and nutrients will be set by the Rhode Island Department of Environmental Management (RI-DEM) as the permitting agency [5], [6]. To increase the performance of a wastewater treatment process it is important to adjust the process parameters (like aeration, solid retention time or hydraulic retention time) in a way that enables it to achieve the highest removal rates possible. In this case, the understanding how the components of the hybrid IFAS system (suspended and attached biomass) work and influence each other in the process of nutrient removal is needed. One way to characterize a process is measuring the biological activity of the microbial communities responsible for the different removal steps (carbon, nitrogen, and phosphorous removal, among other). The determination of the biological activity is important because the conventional biofilm describing parameters (like dry weight or biofilm thickness) do not always show linear correlation with its ability to consume substrates [7]. The biological activity can be measured via respirometric and molecular based methods, and by the measurements of substrate concentrations over time in continuous flow and batch experiments while manometric measurements of the gas phase in batch tests [7], [8], [9]. The molecular based methods assess the activity through the analysis of compounds produced by living cells. A prominent and accurate method is the analysis of the ATP content. ATP is produced by active cells and disappears instantly when cells die and is therefore a good indicator how active biomass is. Its main disadvantage is the complexity of its extraction process. An advantage of the method is that the values stay constant after samples are frozen. Another method described as very sensitive and simple is the INT-dehydrogenase, which measures the activity of the electron transport system (ETS) through the reduction of an added compound (INT) by electron diverted from the ETS. The dehydrogenase analysis works best for population in a stable state and is widely applicable (wide temp. range, anaerobic and aerobic activity) although it does not distinguish between biological and chemical reduction of the INT. It has been characterized as simple, sensitive and rapid and therefore suitable for wastewater treatment plants [7]. "The most conventional technique for microbial activity determination […] is the measurement of the substrate removal rate" [7].
This can be measured through influent and effluent concentrations in continuous flow experiments or start and end (and timed) measurements in batch tests. The disadvantage of these tests is that limitations by oxygen or substrate availability have to be prevented by the experimental design.
Respirometric methods use different means to measure respiration activity in terms of oxygen uptake rate (OUR). The OUR is a fundamental physiological characteristic of culture growth [10] and is a frequently used parameter, even though its sensibility and reproducibility are low and a distinction between primary and secondary metabolisms is not possible [7]. OUR measures the oxygen uptake of a microbial community (or a pure culture) and is directly tied to the substrate consumption of aerobic processes, because the oxygen is necessary as an electron acceptor for the substrate oxidizing bacteria. During the exponential growth phase of the bacteria the OUR increases, because of the higher substrate consumption, and it decreases again in the stationary phase, because of the lower metabolic activity [10]. The sensitivity and reproducibility of the measurements can be improved using sensors and microelectrodes. Respiration rate can be measured using DO-probes [11], gas flow analysis [12] or manometric techniques [8]. The manometric method measures the pressure drop in a closed system which in aerobic conditions can be correlated with oxygen consumption. This method has been also used to determine denitrification activity of biofilm from a post denitrification in Moving Bed Bio Reactor (MBBR) under anoxic conditions [8]. That study used the same principle, with the difference that the increase of pressure was allocated to the production of N2.
Main objective of this study was to assess the use of a manometric method for measuring the respiration activity of the heterotrophic, nitrifying and denitrifying bacteria. Furthermore, nitrous oxide (N2O) production was measured to determine the production of this gas associate with the different biological activities. Finally, the effect of stormwater pollutants on the different microbial populations was assessed in terms of activity and N2O production.  plant in several phases. In the early 1990s a planning process started to reduce the pollution from storm events, which lead to the construction of a three stage CSO abatement tunnel system, the last stage of which was finished in 2016 [14]. The tunnels capture the sewer overflow, to ensure that all stormwater gets stormwater treatment and none gets discharged untreated. In order to reduce the nitrogen discharge from the WWTPs effluent, enhanced aeration technology and the IFAS system were implemented in 2013 [15].
The biological nitrogen removal process consists of two phases: nitrification and denitrification (see Figure 2). Nitrogen enters the treatment plant mostly in the form of ammonia (NH3), which is transformed by biological ammonification from organic nitrogen (for example from fats and proteins) while the wastewater is transported in the sewer system to the wastewater treatment plant [16].
In the nitrification phase, the ammonia (NH3) is oxidized to nitrate in a two-step aerobic process. First Ammonia Oxidizing Bacteria (AOB) transform it to nitrite (NO2) followed by the transformation to nitrate by Nitrite Oxidizing Bacteria (NOB). The AOB first oxidize NH3 to hydroxylamine (NH2OH) using the enzyme ammonia monooxygenase (AMO) and then NH2OH to NO2 using hydroxylamine dehydrogenase (HAO). NOB use a complex enzymatic chain reaction to oxidize NO2 to NO3 [17].
Other microbes, which can oxidize ammonia are ammonia oxidizing archaea (AOA) and bacteria, which can oxidize NH4 under anaerobic conditions using NO2  [17]. There are also some autotroph bacteria capable of denitrification, among which some species are also nitrifiers (Nitrosomonas eutropha & N.europaea) [18]. If these species engage in nitrification under low DO levels, it is called nitrifier denitrification, which also brings some problems in terms of increased N2O production (see 1.1.3.Green House Gas production in Wastewater treatment plants).

The Integrated Fixed Film Activated Sludge (IFAS)
Heterotrophic and nitrifying bacteria compete for oxygen and space in the aerobic zone of WWTPs [19]. Heterotrophic bacteria grow faster than nitrifiers, so they win this competition [20]. Common measures to increase nitrification in an activated sludge process would be increased aeration and longer solids retention times (SRT) [21]. Since an increase of biomass concentration in the aeration tank is limited due to operational requirements (too high SRT decrease activity, growth rate and gas production from sludge treatment) [18], the SRT cannot be drastically increased, if good settlement qualities of the sludge are to be maintained [21]. Both increased aeration and increased reactor volume entail high cost, due to increasing energy requirements (aeration) and/or investment in new technology [21].
Integrated Fixed Film Activated Sludge (IFAS) system were developed to address these issues. The IFAS is a hybrid system, which consists of suspended sludge and biofilm (see Figure 3) that co-exist in the same tank. This separates the bacteria populations. In this case, slow growing nitrifying bacteria can thrive in the biofilm while the suspended biomass allows facultative aerobic bacteria cycle between the aerobic and anoxic tanks [22]. Previous studies have found that the IFAS system yields higher nitrogen removal than conventional systems [20], [22], [23]. The main advantages of the IFAS system are the enhanced nitrification capabilities in less space and the increased process stability in terms of its resilience to low temperatures and temporary disturbances like hydraulic stress, toxins or changes in their environmental conditions [7]. Also, it offers the possibility to add more media to increase treatment capacity [18] with reported values up to 70% of the volume of the aeration tank [22], and it can be used for simultaneous nitrification-denitrification at low DO conditions [18]. The disadvantages of the system are the need for higher DO levels due to the higher biomass content and possible transport of oxygen to the anoxic tank, the use of propriety products (the media and technology are sold by AnoxKaldness, Veolia), the higher difficulty of maintenance, due to the necessity to remove and store the media when maintenance in the tanks is necessary, and additional hydraulic head loss in the WWTP by the flow resistance of the plastic media [18].  wastewater treatment to anthropogenic N2O emission is about 3.2% [27], but N2O from these facilities might account for up to 26% of the GHG emissions of the water supply and sanitation sector combined [17], [27].
In the context of biological nitrogen removal, N2O is produced in both parts (nitrification and denitrification) of the biological nitrogen removal process ( Figure 2 and 4).
The two main microbial communities responsible for nitrification are the ammonia oxidizing bacteria (AOB) and the nitrite oxidizing bacteria (NOB). Of these the AOB are mostly associated with N2O production, mostly through nitrifier denitrification [28] or higher nitritation rates than nitrification ones, which lead to accumulation of NO2 and intermediates of the oxidation process. It has been suggested that during NH3 oxidation, highly reactive intermediates are released by AOB, which then are transformed to N2O through chemical processes [17] (see Figure 4). Nitrifying denitrification is a process where otherwise nitrifying bacteria (like Nitrosomonas europaea) reduce NO2 to NO, N2O and N2 under low oxygen conditions. The main production path of N2O through nitrifying denitrification is during hydroxylamine oxidation (HAO) [28]. Nitrifying denitrification is considered a survival metabolism at low O2 levels, and has been controversially discussed as a self-protection mechanism against NO2 levels [17]. Main drivers of N2O emissions from AOB have been identified to be: nitrite accumulation [8] [16], low DO concentrations [17], [27], excess inorganic carbon concentration [17], low pH conditions [17], [27]. NOB have only been connected to N2O production under anoxic conditions, but their metabolism has scarcely been studied [17]. The main contribution to the N2O production by NOB is indirect, through their respiration by which they control the NO2 accumulation, which causes increased N2O production by other bacteria. The accumulated higher concentrations of NO2 can then inhibit other bacteria and also lead to incomplete nitrifier denitrification. The main factors cited for NOB inhibition are high NH3 concentrations (although unspecific, because the inhibiting concentrations depend on the nitrite oxidizing species) and HNO2, which is correlated to NO2 accumulation at low pH [17].
In the denitrification process, NO3 and NO2 are used as electron acceptors in the absence of O2 and thereby are reduced to N2 through the intermediates NO and N2O.
When this process is not fully conducted, N2O is released. The crucial factor for this is the enzyme N2O reductase (N2OR), which accounts for the reduction of N2O to N2.
This enzyme is very sensitive to even very low concentrations of oxygen and is also inhibited by high NO2 concentrations, likely through stress caused by HNO2 and NO [17], [27]. Interestingly, the inhibitory effect caused by NO, unrelated to its origin, was found to be irreversible even if free NO only appeared temporary. Another factor observed to cause increase in N2O production from denitrification are low or very high COD:N ratios. At low COD:N (<3.5) ratios the N2O emissions increased when organic carbon became the limiting factor and the bacteria started to consume internal storage compounds. In other cases, the limited organic carbon can lead to an accumulation of NO2 which then caused an increased on N2O production. At high COD:N ratios an enrichment of aerobically denitrifying organisms can occur which could be connected to increased N2O production. [27] 1.

Main Objectives
The hypothesis developed for this study is that the Oxitop based manometric method can be used to assess the activity and greenhouse gas production of the different bacterial communities and the effects of inhibitory substances on the IFAS system. In other to probe this hypothesis, the main objectives of this work was the validation of the manometric method, quantification of the biological activity of the heterotrophic, nitrifying and denitrifying bacteria in the IFAS system using manometric measurement methods. Additionally, the response of the hybrid systems components to disturbance by synthetic stormwater and the production of nitrous oxide (N2O) emissions in the different processing steps were investigated as well. weeks, which is very long. In comparison the SRT at the Fields Point WWTP is around two weeks, which is also relatively long (compare [29]).

Reagents and Solutions
The substrate and nutrients concentrations in the liquid phase were analyzed using   The stormwater solution was mixed adapting a recipe that was used before by Kasareni et al. [30] (see Appendix I). The concentrations in the recipe were defined to correspond to 100% stormwater. Therefore, the maximum concentration of pollutants in the bottles was set to be similar to those found at the maximum stormwater input to the WWTP. The pollutant concentrations for the injection mixture were then calculated to reach those corresponding concentrations in the bottle with an injection of 1ml.

Biomass concentration
Total

Equation 1
For the biofilm total solid determination, the average amount of TS per support  The experiments were conducted in 250 ml bottles (see Figure 6). Additionally to measuring head opening, the bottles have two side sockets, which were closed with septi and screwcaps allowing fluids and gas sampling, while keeping the system closed. Below measuring head sodium hydroxide solution container is placed in order to absorbs the CO2 produced during respiration. This step is needed in order to only record the pressure reduction due to oxygen consumption (heterotrophic and nitrifying activity) or pressure increase due to nitrogen production (denitrification).
To analyze the rate of the pressure change, the periods with the highest, stable pressure change after the injection of substrate were selected and the slope of the pressure change in the selected time frame was calculated. Figure 7 shows the image of a representative graph of the change of pressure over time in an aerobic experiment. The pressure at time t=0 is determined to be 0 by the measuring system.  [18]. Values that can be found in the literature for the endogenous respiration are 0.037 d -1 for heterotrophic bacteria, 0.008 for AOB and 0.005 for NOB [32].

GHG production
The gas samples that were taken at the time of injection and the end of each experiment were analyzed in Professor Mozeman-Valtierras Lab in the CBLS Department of URI using a Shimadzu GC-2014 Gas Chromatograph, which was calibrated with three samples each of three different standards with concentrations of 0.508ppm, 2.125ppm and 10.02ppm of nitrous oxide. The gas samples were analyzed for their N2O concentration. An analysis for N2 was not possible, but the concentrations of CO2 and CH4 were also measured, although their calibrations were not as reliable as the one for N2O. Also, it should be noticed that the CO2 concentrations were not accurate, since NaOH was added to all bottles to bind CO2.

Sample Preparation
Immediately after a suspended sludge sample was drawn from the Tank 6, the pH was adjusted using NaOH or sulfuric acid to a value of pH 7 ± 0.3. Then the sludge was let to settle and a fraction of the supernatant was exchanged for PBS. After this, the sludge was either placed in the incubator to be aerated overnight (18-24h) for the aerobic tests or bubbled with argon gas for 30 minutes for the anaerobic tests.   Table 2. After substrate injection, the tests without stormwater injections ran for approximately four hours after before gas samples and fluid samples were drawn from all of the bottles and the experiment was ended.
For the storm water experiments the bottles, which were prepared in the same way as the others before, were run for one hour after the substrate injection and then 0.25ml of the storm water solution were injected every 45 minutes until 1ml was injected in each bottle. After the last injection, the bottles were run for another 45 minutes to 1 hour before gas and fluid samples were taken and the experiment was stopped, and pressure depletion rates were calculated in the same manner described before.
The gas samples that were drawn at the time of injection and the end of each experiment were analyzed for N2O, CH4 and CO2 and the fluid samples that were drawn at the begin and the end of each experiment were centrifuged and analyzed for COD, NH4-N, NO2-N and NO3-N for the aerobic tests and NO3-N and TKN for the anaerobic tests.

RESULTS AND DISCUSSION
In this chapter, the results will be presented in summarizing Tables (Table 3 to Table   6) and then discussed in two parts for the aerobic and anoxic experiments. Table 3 shows the results of the aerobic tests. The substrate assimilation rates in the suspended sludge were one to two orders of magnitude lower than the rates in the biofilm. Both materials showed low N2O production in the heterotrophic tests and higher production in the nitrification tests, although the highest peaks occurred in different tests, in NOB for the suspended sludge and the AOB+NOB test for the biofilm.  Table 4 shows the results of the manometric method and the gas sample analysis for the anoxic experiments. The values for the pressure change over time were not transformed into a substrate reduction rate, because the negative results do not comply with the theory on which the transformational calculations are based, after which the pressure was expected to increase due to the production of nitrogen gas.

Summary of the Results
The negative results in the first line indicate a decrease in pressure but contradicting the results in the second line also show a decrease in NO3-N, which should have produced an increase in pressure. Tables 5 and 6 show the results for the experiments in which a synthetic stormwater run-off solution was gradually injected into the bottles after they had been injected with a substrate (Glucose, Ammonia, Nitrite ore Nitrate), 45 min were left between the injections. Using the data from these measurements the assimilation rates after each injection were calculated. In the Tables 5 and 6 in the first column, it is first indicated which kind of process was tested and then following, the assimilation rates after the four stormwater injections (SW inj. 1-4).      Figure 9 illustrates how scattered the results were and making difficult its interpretation.
Within the suspended sludge results the highest assimilation rates can be found in the heterotrophic experiments and one order of magnitude lower rates for the ammonia oxidizing process while the nitrite oxidizing test shows about half the rate of the heterotrophic. These results qualitatively agreed with previous reports, that heterotrophic bacteria outcompete the AOB in the suspended phase [19], [20], [22].
The NOB show higher activity in the suspended phase than the AOB, which has been found before in suspended sludge, but not in ratios as high as the one found here (about one order of magnitude compared to 1:3 in other studies) [23], [33], [34].
Quantitatively the values reported in the literature are in the order of mgNOx/gMLSS  [22] and mgO2/gTVS [11], which is about two orders of magnitude larger than the results calculated from the pressure measurements.
The assimilation rates calculated from the pressure measurements in the biofilm experiments are all in the same order of magnitude ( ). The order of magnitude of the standard deviation variates, but they are in the same order of magnitude (ammonia ) or one order smaller (heterotroph and nitrite, ) as the one of the substrate assimilation rates. Within this close range, the nitrite oxidation rate is the highest compared to the heterotrophic and the ammonia oxidizing rates, which complies with the findings of Regmi et al. [22] and the premise that fewer heterotrophic bacteria are located in the biofilm [20].
Overall the rates found in the biofilm are one to two orders of magnitude higher than the suspended biomass phase (ratios larger than 10:1, p-values of 0.008 and 0.005 for the heterotrophic and ammonia test and 0.06 for the nitrite test). This does not agree to the ratios found by Regmi et al. [22], which are in the order of 5:1.7 for the AOB and 7.6:0.8 for the NOB between the biofilm:suspended phase, although they used MLSS instead of TS as normalization factor. The difference between the TS and the MLSS is that the TS additionally carries everything that is smaller than 45µm or dissolved in the fluid sample, which includes inorganic compounds which do not participate in the biologic processes. Therefore, the values calculated per MLSS will be higher than the ones calculated per TS. On the other hand, the results of this study agree with the results found by Plechna et al. [11] in qualitative terms (not in total values). Even though they found low OURs for biofilm compared to activated sludge, which was not the case in this study, when set in relation to the biomass, the and measured the DO concentration over a short period (less than 10 minutes) of time and found a difference of the factor 10 in the activities between biofilm and suspended biomass. Plechna also used low TS concentrations (2.5 g/l) in the activated sludge, because they had found the normal concentration to lead to a too fast decline of the DO, which could mean, that in our study as well, oxygen limitation occurred, against all efforts [11]. It might be that the combination of small test volumes and the relation to the biomass leads to a qualitative overestimation of the difference in activity between suspended sludge and biofilm, which could be amplified by the difficulty of the mass determination of the biofilm.
The results of the stormwater tests were very scattered and at times showed opposite behavior between duplicate bottles, which is reflected in the high standard deviations of the data set, however some information can be drawn from the results.
The calculated substrate assimilation rates from the pressure values recorded through the stormwater tests partly followed the anticipated pattern. They were expected to show the normal average assimilation rate after the substrate injection and after each injection the assimilation rate would decreasd, because of the inhibitory effect of the injected pollutants. At first, the assimilation rates increased in most cases after the substrate injection during the first and second SW injection, before the inhibitory effect could be detected, often after the third SW injection. This might have been due to the short time used, so that the bacteria were still increasing their assimilation rate because of the new food source (substrate injection) even after the first SW injection. This assumption is supported by Ren, who describes that in some toxicity studies, respirometric measurements methods took about an hour to show toxic effects [9]. Most bottles showed strong signs of inhibition after the third SW injection (equals to 27% SW, time frame from 1.5 to 2.25 h after 1 st SW injection).
A strong decline in pressure took place in the bottles at high SW pollutant concentrations. The change in pressure could not be accounted for by the expected patterns or patterns from the tests without SW. The change in pressure was not caused by the substrate assimilation, because this pressure drop was also clearly detectable in the control bottles. It is possible that the pressure decline was caused by the oxidation of the metals (Pb, Cd, Ni, Zi, Cu) in the stormwater solution or due to the increased nutrient supply caused by the dead biomass that could increase the metabolism of the active biomass. Another option could be a starting degradation of the poly aromatic hydrocarbons by bacteria, which are present as up to 1% in microbial communities and can in some instances react very fast when hydrocarbons are present [35]. It can have been contributing to this effect, that the concentration of stormwater run-off was increased successive, so that the bacteria had time to adapt, before toxic concentrations were reached.
The results for the N2O production show negligible increase or even decrease of the N2O concentration in the gas phase of both sets (suspended and biofilm) of the heterotrophic experiments, which could correspond with results found by Mannina, who found N2O consumption in the aerobic reactor [36], but opposing trends were found in the nitrogen transformation. In the suspended sludge, a lower production rate of the N2O can be seen with the ammonia oxidization and a higher production rate with the nitrite oxidization. For the biofilm, the opposite was observed. The same tendencies can be analyzed in the respective stormwater experiment, even though marginally inhibited (by 10-35%). The literature reports as causes for N2O emissions in the aerobic phase mainly low DO levels, NO2 accumulation and low pH.
The acidity as a cause can be ruled out because of the use of PBS to buffer changes in the pH [37]. DO could not be a cause, since there was an intensive aeration before the tests and the constant stirring. In the instances where the DO was measured at the end, it was at levels that were too high to suggest an anoxic environment in the samples (≈4mg/l), but considering the observation connected to DO by Plechna [11], it cannot be ruled out that regions of low DO in sludge flocks or the biofilm are due to possible limitations by the oxygen transfer rate in sludge flocs or biofilm [10]. The nitrite accumulation due to the direct injection of the nitrite could explain the high N2O production values in the nitrite oxidizing in the suspended sludge test. This might Figure 10: Extract of two exemplary pressure graphs from which gas samples had been extracted; Source H. Behrmann not have occurred in the ammonia test because of its better equilibrium between its ammonia oxidizing and nitrite oxidizing processes, which would result in a nitrite oxidization rate high enough to avoid nitrite accumulation, that would have resulted in a negative effect (increased N2O production). The pattern in the biofilm tests was the opposite, with high N2O production in the ammonia test and lower production rates in the nitrite test. This could mean that the concentration and activity of the NOB in the biofilm is high enough to oxidize the injected concentration of nitrite without inhibitory effects. The low N2O production in the nitrite test also indicates that the high production in the ammonia test is most likely not caused by nitrite accumulation. It is likely that the high N2O production rate could be caused by a higher oxygen utilization than oxygen transfer rate, which could lead to low oxygen concentration in the biofilm even though enough oxygen is dissolved in the fluid phase [10]. These areas of low oxygen in the biofilm can cause production of N2O due to nitrifier denitrification, aerobic denitrification or intermediates of the incomplete oxidization of ammonia [17].

Anoxic tests
It was expected that the that is produced in the anoxic zones is released in the aerobic zones when aeration lowers the transfer resistance [27], [36]. The decline of NO3 concentration in the fluid sample analysis and an increased production of N2O suggest that denitrification occurred but could not be detected by the manometric measurements. The observation that an increase in pressure could be detected in the bottles from which gas samples had been drawn at the begin of the experiment and which therefor started at low pressure levels (see Figure 10) suggests that the pressure in the bottles might have prohibited the release of the N2 and N2O into the gas phase. This is contradicted by the fact that Brådskär [8] found pressure increase with a similar but larger scaled experimental setup. It is possible, that the concentrations of biomass and substrate in the experiment were too low produce an observable pressure change. It also begs the question how much N2O was dissolved in the fluid phase and therefore did not get detected in the gas sampling.
The manometric values from the anoxic stormwater experiment are different with a high deviation, coming to inconclusive results. Some of the suspended sludge bottles showed patterns that also could be seen in the aerobic tests, but then also duplicate bottles produced opposite results in different timeframes, while switching their direction (positive/negative rates) in between timeframes. In the biofilm set, the control bottles showed patterns that were expected from the bottles with the substrate, while the bottles with the substrate showed high pressure depletion that increased with the successive SW injections.
The N2O production in both the suspended sludge and the biofilm anoxic experiments was higher than in the aerobic tests, which matches the literature that identified the anoxic zone as a main source of N2O, especially, when incomplete denitrification occurs [27], [36]. These results differ from continuous reactors where the dissolved N2O is transported to the aerated sections and stripped out [36]. In this study, the N2O production could be allocated to its process of origin, due to the batch tests with the different substrates, were dissolved N2O could not be transported out of the zone where the process took place. In the suspended sludge, the production rate was as high as the highest of the aerobic tests (NO2 to NO3), in the biofilm it even

Limitations
Limiting factors for this study were the small number of samples, which makes it difficult to identify outliers and larger trends. Also, the small volume of the samples, which was caused by the available equipment (bottles, stirrer plates, incubator) and easy handling, might have contributed to the high variation of the results, due to scaling effects and the normalization on the TS concentration. For the anoxic tests, it is very difficult to verify that they were actually anoxic, which could explain the negative results, even though all possible steps were conducted to produce anoxic conditions. Additionally, the results of the fluid sample analysis were not accurate, which make the verification of the manometric results impossible. Furthermore, the choice of a normalization factor is difficult, because of the difficulty to remove the biofilm from the support media and the identification of its components. The TS was a parameter, which was possible to determine, but it also entails distortion, because the composition of the biofilm and the suspended sludge are different from each other. Finally, variation on daily operation of the model WWTP, could affect the activity and concentration of the biomass as well.

CONCLUSION
The results of the manometric method are very variable; however, they agreed qualitatively with previous studies. The manometric method could be an option to measure aerobic activity using large sample volumes and repetitions that could produce better quality results, enable researchers to identify outliers and allow justified interpretations, but other methods like the substrate mass balancing or DO measurements would be a more efficient alternative, due to faster procedures and possibly lower sample volumes. For the stormwater test, longer time frames should be considered so that the influence of disturbance from the injection is reduced, otherwise automatized injections and gas sampling could be considered.
For the anoxic activity, the results from this study are contradictory and do not produce interpretable results. To determine the desired target concentration in the bottles these concentrations were multiplied by 0.4, which resulted in the concentrations below (column 2). These concentrations were then multiplied by 0.09 l/Bottle to calculate the total amount of each compound per bottle, which then also equals the concentration per ml in the solution, because it had to be added to the bottle in a 1ml injection. which can be transformed into Equation 5,

Equation 5
the reduction of air can be calculated from the pressure depletion. Since normal air was used, the depletion of oxygen is equivalent to 20.95% [31] of the determined n value. The resulting number (n*0.2095) can then be transformed into a mass [g] as shown in Equation 6.

Equation 6
Through stoichiometric calculations (see eq. 7, 8 and 9) the assimilation rate of nutrients can be calculated from the use of oxygen. The oxygen demands used for the calculations were: 4.57 g O2/g NH4 to NO3, 1.14 g O2/g NO2 to NO3 and 1.07 g O2/g C6H12O6 [18].