Mainstream Deammonification process monitoring by bacterial activity tests

Mainstream Deammonification process monitoring by

bacterial activity tests

ANDREA CARRANZA MUÑOZ

KTH ROYAL INSTITUTE OF TECHNOLOGY

SCHOOL OF ARCHITECTURE AND THE BUILT ENVIRONMENT DEGREE PROJECT IN ENVIRONMENTAL ENGINEERING,

SECOND CYCLE, 30 CREDITS

STOCKHOLM, SWEDEN 2020

TRITA-ABE-MBT- 2054

www.kth.se

Mainstream Deammonification process monitoring by bacterial activity tests

ANDREA CARRANZA MUÑOZ

Supervisor and Examiner Elzbieta Plaza

Degree Project in Environmental Engineering and Sustainable Infrastructure KTH Royal Institute of Technology

School of Architecture and Built Environment

Department of Sustainable Development, Environmental Science and Engineering

SE-100 44 Stockholm, Sweden

©Andrea Carranza 2020 Degree Project Master Level

Department of Sustainable Development, Environmental Science and Engineering School of Architecture and the Built Environment

Royal Institute of Technology (KTH) SE-100 44 STOCKHOLM, Sweden

Reference should be written as: Carranza, A., (2020) Mainstream Deammonification process monitoring by bacterial

activity tests.

Sammanfattning

Deammonifikation är en välanvänd teknik för rening av sidoströmmar med höga ammoniumkoncentrationer vid relativt hög temperatur, som till exempel rejektvatten från avvattning av rötslam eller industriellt avloppsvatten. Deammonifikationsprocessen har lägre driftkostnad än konventionella reningsprocesser, förbrukar mindre energi samt möjliggör högre biogasproduktion samtidigt som processen är enkel att implementera. Reningstekniken har dock ännu inte tillämpats i fullskala för rening av huvudströmmen på grund av den höga C/N-kvoten och de låga vattentemperaturerna i kommunalt avloppsvatten samt behovet av efterbehandling. Detta anses ha en negativ inverkan på anammoxbakteriernas tillväxthastighet och funktion vilket påverkar bakteriegruppens beteende i processen.

.

Syftet med detta examensarbete var att utvärdera om det är praktiskt genomförbart att använda deammonifikation för att rena kväve från kommunalt avloppsvatten, vilket följdes upp genom att studera bakterieaktiviteten i en pilotskalereaktor. De involverade bakteriegrupperna (AOB, NOB, heterotrofer och denitrifierare) övervakades genom att mäta den mikrobiella aktiviteten varje vecka med hjälp av batch-tester. Resultaten användes till att utvärdera olika driftstrategier och deras effekt genom att följa förändringarna i mikrobiell aktivitet hos de konkurrerande bakteriegrupperna.

Studien genomfördes i Stockholm under sex månader i en enstegs-IFAS-pilotskalereaktor (integrerad process med biofilm på fast bärarmaterial och aktivslam) som matades med kommunalt avloppsvatten som förbehandlats i en UASB-reaktor.

De olika driftstrategierna omfattade olika temperaturer, luftningsstrategier, syrekoncentrationer,

slamåldrar och hydrauliska uppehållstider. Syftet med driftstrategierna var att främja AOB- och

anammoxbakteriers tillväxt för att i framtida studier kunna erhålla en förbättrad

deammonifikationsprocess. Syftet i denna studie var dock i första hand att förbättra den bakteriella

konkurrensen och göra den lättare att mäta, inte att uppnå bästa möjliga kväverening. Den

driftstrategi som gav bäst resultat i denna studie innebar att hålla en syrehalt på 1,5 mg/l med 10

minuter luftning följt av 20 minuter utan luftning vilket säkerställde en normal kväveavskiljning och

samtidigt möjliggjorde övervakning av samtliga fyra bakteriegrupper. Totalkväveavskiljningen var

över 50 % och ammoniumavskiljningen över 95 % medan kvävereningsaktiviteten ökade till 30 g

N/m ^3- d och systemet hade en övergripande effektivitet på 75 %. Studien visade att under rätt

förutsättningar kan de nödvändiga bakteriegrupperna selekteras fram och deammonifikation av

kommunalt avloppsvatten kan utföras på ett framgångsrikt sätt.

Abstract

Deammonification is a widely used technology for side stream treatment with rich ammonium streams at relatively high temperatures, such as, the reject water coming from dewatering units in treatment of digested sludge and industrial wastewaters. The deammonification process has lower operational costs than conventional systems, consumes less energy, enables the increase of biogas production and it is easy to implement. However, this technology has not yet been applied in full- scale mainstream treatment due to its restrictions in coping with high C/N ratios, low temperatures, and the need for post-treatment processes. These conditions are allegedly negative to the growth and performance of anammox bacteria affecting the bacterial groups’ behavior in the process.

This master thesis project aimed to evaluate the feasibility of using deammonification to remove nitrogen from mainstream wastewater, which was studied by monitoring the bacterial activity in a pilot scale reactor. The different bacterial groups involved (AOB, NOB, heterotrophs, and denitrifiers) were monitored by weekly measuring their activity in batch activity tests. The results allowed the evaluation of different operational scenarios and their impact by following up on the changes in the bacterial competition. The study was conducted for six months in a single-stage IFAS (integrated fixed-film activated sludge) pilot-scale reactor located in Stockholm and fed with pretreated (with a UASB) municipal wastewater.

The different operational scenarios involved changes in temperature, aeration patterns, DO concentration, SRT, and HRT. The adjustment of these features was done in the interest of promoting AOB and anammox bacterial growth, leading to an improvement of the deammonification efficiency in future studies. However, the chosen operational conditions were to enhance bacterial competition and facilitate its visualization, not to maximize nitrogen removal. Thus, the most suitable scenario found during this study included DO concentration of 1.5 mg/L with 10 aeration-20 non-aeration pattern and ensured nitrogen removal rates within normal values while allowing the monitoring of all the bacterial groups. TN removal reached a value above 50% and NH4-N above 95%, whereas nitrogen Removal Rate (NRR) increased to 30g/N/m ³ -d and the system had an overall nitrogen removal efficiency of 75%. Nevertheless, it was proven that in the right environment, the necessary bacterial groups can be selectively accumulated and successfully perform deammonification and reduce nitrogen levels in mainstream wastewater.

Keywords

Ammonia Oxidizing Bacteria (AOB), Anammox, Bacterial activity, Deammonification, Heterotrophic

bacteria, IFAS, Mainstream wastewater, Nitrite Oxidizing Bacteria (NOB), Nitrogen Removal.

Acknowledgements

I would first like to thank my thesis supervisor Professor Elzbieta Plaza and extend my deepest gratitude for the patience she still has with me and for all her help during this process. Letting me be part of this project helped me reassure my passion for water and inspired a new interest in me for research. It was an honor to pursue this project at Sjöstadsverket research facility and I enjoyed tremendously the work I did. I would also want to thank dr. Karol Trojanowicz who was involved during most part of the process and suffered with me the hottest summer in Stockholm while doing experiments in the lab. Thank you for helping me along this process and teaching me that research can be terrific and exciting. Thanks to dr. Jozef Trela for being part of the project.

I want to also express my gratitude to all members and now colleagues at Hammarby Sjöstadsverk for all your support and for teaching me so much, especially to Gabriel for everything. Mayumi, thank you for all your help and patience. Mila, Jesper, Niclas, I keep receiving all your help and am still learning so much from you, thank you.

To my parents, I do not have enough words to thank you for everything you have done for me and all the love we share, thank you for helping me chasing my passions. V thank you for always pushing me to be a better person and for always being there for me, even while living in different continents.

Thanks to all my family for your support.

Finally, I thank everyone I have met during these years in Stockholm, you have made this city a home to me, this accomplishment would not have been possible without you. Thank you.

Andrea Carranza Muñoz

Table of Contents Table of Contents

Sammanfattning ... i

Abstract ... iii

Acknowledgements ... v

Table of Contents ... vii

List of figures ... ix

List of tables ... ix

Abbreviations ... xi

1 Introduction ... 1

2 Background ... 2

2.1 Nitrification / Denitrification ... 2

2.2 Anammox ... 3

2.3 PN/A – single stage deammonification ... 4

3 Research Question and Scope ... 6

4 Thesis Outline ... 6

5 One stage - Mainstream Deammonification ... 8

5.1 Bacterial groups involved in the process ... 8

5.2 Process Configurations ... 9

5.3 Mainstream conditions ...10

5.4 Factors influencing the process ...10

6 Materials and Methods ... 13

6.1 Pilot scale Single-Stage Mainstream Deammonification ... 13

6.1.1 System Characteristics ... 14

6.1.2 Online Monitoring ... 15

6.1.3 Operation and Control Strategies ... 16

6.1.3.1 Temperature ... 16

6.1.3.2 Aeration patterns and DO concentration ... 17

6.1.3.3 Flowrate ... 17

6.1.3.4 Suspended Solids Concentration MLVSS ... 18

6.2 Chemical Laboratory Analyses ... 18

6.3 Bacterial Activity Tests ... 19

6.3.1 Experimental set up and devices ... 19

6.3.2 Schedule and Sampling ... 20

6.3.3 Oxygen Uptake Rate (OUR) ... 22

6.3.4 Specific Anammox Activity (SAA) and Nitrate Uptake Rate (NUR) ... 24

7 Results and Discussion ... 26

7.1 Influent Characteristics ... 26

7.2 Nitrogen Removal Performance and Nitrogen Concentrations ... 29

7.2.1 Influence of Temperature ... 33

7.2.2 Aeration Patterns ... 33

7.2.3 Flowrate ... 34

7.3 Microbial Activity tests - OUR and SAA ... 34

7.3.1 Biofilm-Carriers ... 35

7.3.2 Activated Sludge ... 38

7.4 Microbial Activity tests – NUR ... 40

8 Conclusions ... 42

References... 44

 Annexes ... 48

o Annex A – Online Data ... 48

o Annex B.1 – Example Calculations for OUR ... 49

o Annex B.2 – Results OUR ... 50

o Annex C.1 – Example Calculations for SAA ... 56

o Annex C.2 – Results SAA ... 57

List of figures

Figure 1 Nitrogen removal, complete cycle (Daims et al., 2016) ... 3

Figure 2 Autotrophic processes for N removal in Biofilms (Modified Henze et al., 2017) ... 5

Figure 3 Bacterial groups involved in the process ... 8

Figure 4 Configuration IFAS Pilot (Modified. Robiglio, 2018) ... 14

Figure 5 Oxygen relation with bacterial behavior (Modified Henze et al., 2017) ... 17

Figure 6 COD/NH4-N ratio (Inflow) ... 27

Figure 7 Alkalinity and pH values (Inflow) ... 28

Figure 8 NH4-N comparison online-Lab values (in) ... 29

Figure 9 Performance of the IFAS system with temperature variation a) TN and NH4-N removal. b) Changes in NH4-N, NO3-N and NO2-N concentrations ... 31

Figure 10 Nitrogen Removal Rate and Nitrogen Loading Rate (gN/m3-d) showing the performance of the system ... 32

Figure 11 NO3-N produced / NH4-N removed ... 32

Figure 12 OUR and SAA activity tests carried out with Biofilm-carriers ... 35

Figure 13 OUR and SAA activity tests carried out with Activated Sludge ... 39

Figure 14 NUR and SAA activity tests carried out with Biofilm-carriers ... 40

Figure 15 NUR and SAA activity tests carried out with Activated Sludge ... 41

List of tables Table 1 Anammox technologies ...4

Table 2 Schedule... 18

Abbreviations

AOB Ammonia Oxidizing Bacteria

Anammox ANaerobic AMMonia OXidation

C/N Carbon / Nitrogen ratio

COD Chemical Oxygen Demand

DO Dissolved Oxygen concentration

HRT Hydraulic Retention Time

IFAS Integrated Fixed-Film Activated Sludge

mg/L milligram per liter

mg/m ² -d milligram per square meter per day mg/gVSSd milligram per gram of VSS per day

MBBR Moving Bed Biofilm Reactor

MLSS Mixed Liquor Suspended Solids

N2 Nitrogen Gas

N/DN Nitrification / Denitrification

NH4-N Ammonium nitrogen

NLR Nitrogen Loading Rate

NO2-N Nitrite nitrogen

NO3-N Nitrate nitrogen

NOB Nitrite Oxidizing Bacteria

NRR Nitrogen Removal Rate

NUR Nitrate Uptake Rate

OUR Oxygen Uptake Rate

PN/A Partial Nitritation - Anammox

RAS Returned Activated Sludge

SAA Specific Anammox Activity

SRT Solid Retention Time

SSV Sjöstadsverket

TSS Total Suspended Solids

TN Total Nitrogen

UASB Upflow Anaerobic Sludge Blanket

VSS Volatile Suspended Solids

WAS Waste Activated Sludge

WWTP Wastewater Treatment Plant

1 Introduction

Increasing population in urban areas around the world have led to higher nitrogen loads in WWTPs’

effluents. Their contribution to water bodies’ pollution and eutrophication concerns governments worldwide, who have tried to address this problematic through the introduction of increasingly strict nutrient loading limits. These higher standards have driven the WWTP industry to seek more economically feasible technologies to remove nutrients. Costs related to energy consumption, large area requirement and carbon sources addition are some of the obstacles that new technologies are trying to cope with.

Some of the new technologies that are currently under investigation take advantage of the shortcuts on the nitrogen cycle to effectively remove nitrogen through a biological pathway and at the same time, reduce some of the disadvantages of traditional processes. One good example is deammonification, which has been successfully implemented in side stream treatment processes and has shown high efficiencies at low operational costs. The advantages of Partial Nitritation/Anammox or deammonification process include: low energy consumption, maximization of energy recovery in terms of biogas, no need of external carbon source and its compatibility with existing infrastructure from other processes (Malovanyy et al., 2015).

Deammonification uses the shortcut of partial nitritation process where Ammonium Oxidizing Bacteria oxidize the ammonium into nitrite, and it is followed by Anammox bacteria transforming ammonium to Nitrogen gas without an external carbon source (O’Shaughnessy, 2015). Although the implementation of deammonification in mainstream treatment has been recently researched, it hasn’t been yet established as a main Nitrogen removal technology due to some uncertainties on the bacterial performance under mainstream conditions.

This thesis performs and discusses an evaluation of the deammonification process applied to

mainstream treatment. The evaluation and monitoring of the process and the bacterial activity

behavior under different conditions were done by Bacterial Activity Tests. Furthermore, different

strategies that were applied for the out selection of some bacterial types and the achievement of high-

efficiency performance are discussed and analyzed.

2 Background

Nitrogen levels present in water have been exponentially increasing in the past decades due to increase in population, use of nitrogen-based fertilizers and industrial activities. It is known that high concentrations of Nitrogen lead to environmental problems related to eutrophication and oxygen depletion, affecting the animal and plant lifecycle in surface water bodies.

Since the second half of the twentieth century, several methods and water treatment technologies for Nitrogen removal have been researched, being Nitrification /Definitrification (N/DN) the most worldwide used up until now.

2.1 Nitrification / Denitrification

This process is characterized by having two steps in which ammonium is oxidized into nitrites under aerobic conditions (nitrification) and then converted to nitrogen gas under anoxic conditions in the presence of organic carbon (denitrification). The denitrification phase can be located before the aerobic zone (pre-denitrification) or after (post-denitrification) (Ali, 2015).

The Nitrification is carried out in two parts Nitritation and Nitratation. Nitritation refers to the oxidation of Ammonia into Nitrite (1) and it is performed by Nitrosomonas (Ammonia Oxidizing Bacteria - AOB). Nitratation is the oxidation of nitrite into nitrate (2) by Nitrobacter and Nitrospira (Nitrite Oxidizing Bacteria - NOB). The total nitrification step (3) needs 4.57 g O2 per gram of NH4-N to complete the oxidation process. The other stage of this process is the oxidation of Nitrate (NO3) into Nitrogen gas (N2) by using a carbon source as a substrate (4). In this reaction 1 g NO3-N is equivalent with 2.86 g O2, hence 2.86 g COD is needed to oxidize 1 g of NO3-N (Randall et al., 1992, Metcalf & Eddy (2014))

𝑁𝐻 + 1.5 𝑂 → 𝑁𝑂 + 𝐻 𝑂 + 2𝐻 (1)

𝑁𝑂 + 0.5 𝑂 → 𝑁𝑂 (2)

𝑁𝐻 + 2 𝑂 → 𝑁𝑂 + 𝐻 𝑂 + 2𝐻 (3)

0.2 𝑁𝑂 + 1.2 𝐻 + 𝑒 → 0.1 𝑁 + 0.6 𝐻 𝑂 (4a)

𝑁𝑂 + 𝐻 + 4𝑔𝐶𝑂𝐷 → 0.5 𝑁 + 1.5 𝑔 𝑆𝑙𝑢𝑑𝑔𝑒 (4b)

However, the conventional nitrogen removal (N/DN) has some disadvantages including high energy

consumption in the aerobic phase, the need of an external carbon source to complete the

denitrification process and the requirement of vast areas for construction and operation (Kartal,

2010).

The study of the different bacteria responsible for the transformations that occur in the nitrogen cycle (Figure 1) disclosure several technologies that have taken advantage of the short cuts of the nitrogen cycle in order to perform a more efficient nitrogen removal. One of the most relevant recent discoveries is the Anaerobic Ammonium Oxidation (Anammox), bacteria that in combination with other processes can perform an efficient Nitrogen Removal.

2.2 Anammox

Anammox is an autotrophic process discovered on the 80’s but just until the 90’s was started to be used for water treatment purposes (Henze et al., 2017).

Anammox is one of the multiple processes that uses a shortcut on the nitrogen cycle (Figure 1) to effectively remove N from water. In this process, Anammox uses CO2 as a carbon source to produce biomass (CH 2 .O 0.5 N 0.15 ) and nitrites as an electron acceptor for ammonium conversion to N2 (5). And at the same time nitrites are used as an electron donor for CO 2 reduction (6) (Kuenen, 2008) 𝑁𝐻 + 𝑁𝑂 → 𝑁 + 2 𝐻 𝑂 (5)

𝑁𝐻 + 1.32 𝑁𝑂 + 0.066 𝐻𝐶𝑂 + 0.13 𝐻 → 1.02 𝑁 + 0.26 𝑁𝑂 + 2.03 𝐻 𝑂 + 0.066 𝐶𝐻 𝑂 _. 𝑁 _. (6)

Since the first studies in Anammmox, it was clear that Anammox was strictly anaerobic. One of the first to focus his studies in this process was Strous et al. (1997), who proved that bacterial growth was inhibited even by low oxygen concentrations (<0.5% air saturation), no matter how many different bacteria groups responsible for the process. Despite its extremely low growth rates (Approx. 2 weeks), Anammox advantages are clear, the decreasing in energy consumption, no external carbon source requirement and the decreasing in space are just the most important ones.

The use of Anammox after a nitrification process leaves a Nitrate residual concentration that is costly to remove. Thus, several combinations were researched trying to reach the highest anammox performance, but it was just after the development of the SHARON technology that it could be achieved. SHARON ideally produces a 50:50 nitrite ammonium rate, which decreases the NO3 presence on the effluent and increases the Anammox activity.

Sharon technology converts 50 % of the ammonium to nitrite through a nitritation process (7) performed by AOA and AOB. Under specific conditions, Sharon successfully completes the NOB Nitrobacter wash out and the exclusive NH4 oxidation to NO2 (Biological WWT, 2017, Kuenen, 2008, Gonzales et al, 2018).

𝑁𝐻 + 1.5 𝑂 → 𝑁𝑂 + 𝐻 𝑂 + 2𝐻 (7)

Figure 1 Nitrogen removal, complete cycle (Daims

et al., 2016)

The combination of SHARON with the autotrophic denitrification based on Anammox is known to work efficient and it is done in two different reactors. SHARON/Anammox has been proven to be efficient and cost/ effective with N removal ratios from 26 up to 76 kg N/m ³ /d, operating at high temperatures and without sludge retention (Hellinga et al., 1998, Gonzales et al., 2018).

After Sharon, different technologies that could be combined with the anammox faculties were patented. They can be divided in three main groups by process type as follows.

Table 1 Anammox technologies

2.3 PN/A – single stage deammonification

This process is called “aerobic de-ammonification” and converts NH4 into N2 gas without requiring an electron donor (Verstraete and Philips, 1998). It is a two -step biological process where first AOB oxidize the ammonia in aerobic conditions and the anammox oxidize the ammonia using nitrite (previous section).

Though Anammox bacteria were thought to be completely sensitive to aeration, and that its growth inhibition was imminent when exposed to even low DO concentrations, it was later proven by Strous et al. (1997) that performing both nitrtitation and anammox processes on the same reactor can be done by setting intermittent aeration patterns (Lackner, 2010). This opened a door for single reactors processes (Table 1).

Deammonification is a more economically feasible option in terms of energy consumption and reactor volume, but the immobilization of the biomass has to be considered due to the known low growth rate of anammox bacteria. This is a determinant parameter that can be accomplished by using biofilm, granulated sludge or gel beads ((Ali et al., 2014; Ge et al., 2009; Isaka et al., 2008a; Magrí et al., 2012, Metcalf & Eddy, 2014).

Figure 2 is a representation of the processes of the autotrophic bacteria in biofilm. It shows the

anammox development on the deepest part of the biofilm due its anaerobic nature (Henze et al, 2017).

This thesis will focus however, in the single stage deammonification in IFAS reactors.

Currently this technology is used for side stream treatment at WWTP for supernatant However, some publications and concerning investigations including the present thesis have been aiming to use this process to treat mainstream wastewater at lower costs and smaller areas.

Figure 2 Autotrophic processes for N removal in

Biofilms (Modified Henze et al., 2017)

3 Research Question and Scope

This thesis project aimed to evaluate and monitor the feasibility of mainstream deammonification process tested in an IFAS pilot reactor and describe how bacterial groups behave under different operational conditions. Monitoring of the process was performed by bacterial activity tests (OUR, SAA, NUR) carried out in the biofilm carriers and in the sludge. The pilot scale IFAS reactor was located in Hammarby Sjöstadsverket research facility in Stockholm and it was fed with mainstream inflow pretreated with a UASB reactor.

The evaluation of the process was carried out during a period of six months. Several operational conditions were tested in order to find the most suitable combination to achieve the highest N removal. The strategies implemented were: enhancing the Anammox growth and promoting the AOB over the NOB by the adjustment of the operational parameters based on the close follow up of the bacterial competition.

A favorable performance of the deammonification process can open an entirely new chapter of more cost/effective technologies for nitrogen removal in urban wastewater treatment.

4 Thesis Outline

The 5 ^th chapter of this thesis contains a description of the mainstream deammonification process, its biochemical characteristics and main constrains. A recap on the key operational conditions and factors affecting the involved processes and their limitations are also presented. Along this literature review chapter, a special emphasis was given to lessons learnt from previous pilot systems and other studies, as well as to processes performances and comparable technologies. Theses parameters were used as initial parameters for this study.

Further in this thesis, on chapter 6 ^th a bacterial characterization of all the groups involved on the process is presented with their expected behavior under mainstream conditions; followed by a background on the methods used for the activity tests. This helps to put this study into context and to allow a correct interpretation and analysis of the results.

The next chapter describes in detail all the methods used during this project and the calculations performed to obtain the results. A description of the pilot reactor, laboratory procedures, schedule of the tests and sampling, a broad summary of the activity tests and the online monitoring program are mentioned in this section.

The results and discussion chapter outline step by step the most relevant data along with an

interpretation of the results. Graphic representation of the system performance trough time was

meant to be easily compared with the bacterial activity, giving a complete overview of the

deammonification process. During the results chapter, all parameters and operational conditions are discussed based on bacterial behavior and the performance of the IFAS reactor.

The limitations that were found during this thesis project, which involved methodological procedures

regarding sampling, instruments performance, external factors and process uncertainties are

included in the results chapter.

5 One stage - Mainstream Deammonification

One stage deammonification system Configuration is briefly described in chapter 2. This chapter explains the process in more detail and focuses on mainstream conditions.

5.1 Bacterial groups involved in the process

Deammonification is also referred to as PN/A which states for Partial Nitritation / Anammox, the two biological processes that are present in the reactor. Although the successful performance of these two processes in a single reactor is necessary to reduce the NH4 concentrations, the involvement of several bacteria group makes it difficult to achieve this goal. The principal bacteria groups involved are AOB, NOB, Anammox and Denitrifiers and perform as follows:

Figure 3 Bacterial groups involved in the process

As it is shown in Figure 3, the first key parameter for PN/A process to be successful is to stop the nitrification process halfway, letting AOB oxidize around 50% of NH4 ⁺ into NO2 (Nitritation) and at the same time wash out or inhibit NOB activity to avoid oxidation of NO2 ^- into NO3 ^- (Niratation).

Nitritation as a part of the nitrification process, needs oxygen for the oxidation (equation 1) and it is

mainly performed by the bacteria Nitrosomonas. On the other side, Anammox are completely

anaerobic bacteria and the creation of the two environments on the same reactor is the second

important parameter in this process. The last key parameter involves the different growth rates

between the AOB and anammox bacteria. Anammox is known to have a considerably low growth rate

and thus, it needs to be immobilized in the reactor to ensure its high performance.

5.2 Process Configurations

There are two principal process types for deammonification; Suspended growth systems and Biofilm systems. The first one includes suspended growth biomass in which the flocculent or granular sludge keep the microorganisms in the system in form of flocs (Cao et al., 2017, Malovanyy, 2014). The bacteria groups AOB and NOB have higher growth rates than anammox in these kinds of systems due to their facility to grow in suspended biomass, enhancing the nitritation performance (Lemaire et al., 2013). On the other side in biofilm systems, biomass is retained and grows on the surface of different materials called carriers. This layer formed on the carriers is known as biofilm and it is explained in Figure 2.

MBBR

Moving Bed Biofilm Reactor - MBBR is a biofilm system in which carriers remain in suspension in the reactor while biomass grows in their surface. These carriers are characterized for having a small size but a high specific surface per unit of volume, allowing more biomass to be present in the reactor.

MBBR has the capacity to retain more anammox biomass, overcoming the low growth rate of these bacteria at low temperatures (Gilbert et al., 2015).

Malovanyy et al. (2015) shows results from an MBBR PN/A reactor treating mainstream at 25°C. This pilot had a high average efficiency, but it was producing significant amount of NO3 ^- even with high temperatures, which are not a main characteristic of mainstream water.

One stage MBBR PN/A for mainstream application has not been fully proven yet. There are several control strategies proposed by different authors based on: COD/N ratio, DO/TAN ratio, FA inhibition, DO control, among others. But none of these have entirely ensured partial nitritation in mainstream conditions (Kowalski et al., 2019).

IFAS

Integrated Fixed film Activated Sludge – IFAS is a hybrid reactor. It combines suspended growth systems and biofilm systems in the same reactor, this bring advantages for different bacteria groups optimizing the process. The suspended solid concentration is then higher than in other systems, which enhances the AOB and NOB growth and at the same time the system retains anammox in the carriers.

Other advantages of combine this two processes described by Veuillet et al. (2015) and Kowalski et

al. (2019), is the improvement of biomass retention, the distribution of the bacteria groups between

the sludge phase and the carriers, and the thinning of the biofilm layer due to its interaction with the

biomass, which allows a better DO control strategy for biomass wash-out.

In summary, IFAS has been proven to increase the efficiency of the PN/A process (Malovanyy, 2015, Veuillet et al., 2015, Gilbert et al., 2015, Kowalski et al., 2019 and Lemaire et al., 2013). However, the main constrain in these kinds of processes are the mainstream conditions.

5.3 Mainstream conditions

Three key parameters were mentioned to be considered in order to achieve high process performance:

Intermittent aeration patterns, immobilization of biomass and NOB inhibition (Cao et al., 2017).

These parameters however, are harder to control in a mainstream system due to the water limiting initial conditions.

Mainstream water contains relatively low NH4-N concentrations, lower temperatures and higher COD concentration than side stream. These conditions determine the bacterial behavior in the process, mainly contributing to the NOB growth. NOB bacteria are characterized by their higher affinity with higher oxygen concentrations, their higher growth in low temperatures and their strong competition with Anammox for the NO2-N available in the system. These characteristics along with the low performance of AOB under mainstream conditions make the NOB inhibition harder than in side stream treatment or in systems with two stage configurations.

Intermittent aeration plays a more important role in mainstream system due to the bacterial sensitivity to changes at low activity ratios, including the anaerobic performance of Anammox bacteria (Vilpanen, 2017). Furthermore, immobilization of bacteria should be performed by the utilization of biofilm or granules ((Ali et al., 2014; Ge et al., 2009; Isaka et al., 2008a, 2007; Magrí et al., 2012; Quan et al., 2011, Metcalf & Eddy, 2014). Usually the NOB wash out in these kinds of systems is achieved by high temperatures, low SRT and/or low DO, but that cannot be applied in mainstream treatment.

Another important issue for mainstream application of deammonification is the post-treatment that could be necessary to polish effluent water quality in order to meet the effluent discharge standards.

High NO3-N and COD concentrations can be present in the process effluent.

Despite the challenges of a one stage mainstream deammonification system, the process has lower energy consumption, maximize energy recovery from biogas production and it is relatively easy to implement in already existing processes (O’Shaughnessy, 2015). Furthermore, the reduction of an external carbon source dose has been proven.

5.4 Factors influencing the process

Several factors directly affect the performance of the mainstream deammonification process and

represent limitations for bacterial activity growth. pH, temperature and C/N ratio significantly vary

between reject water and mainstream conditions and these parameters affect the process the most.

Furthermore, operational conditions including DO concentration, HRT and SRT also define bacterial growth and the process success (Robiglio, 2018).

Temperature

Mainstream influents are affected by the temperature of the environment. In countries with seasons, the flow temperature can decrease to 15°C during wintertime, and in Nordic countries like Sweden down to 10 - 9°C in the coldest weeks of the year. This means that the treatment process is directly affected by the changes in the temperature along the year and with it, the bacterial activity. Bacterial groups behave different under temperatures changes as it was shown in the previous sections.

Temperature is one of the first conditions used for bacterial selection. Ammonium oxidation, for example, is more affected by temperature changes than Nitrite oxidation, and it is also proven that at 20+ºC AOB have a higher k (bacterial growth rate) than NOB (Metcalf & Eddie, 2014); these growth rates make possible a selective wash out of NOB in suspended biomass systems for partial nitration by adjusting the SRT (Hellinga et al., 1998). Furthermore, temperature directly affects Anammox activity and its growth rates (He et al., 2018) making the PN/A process difficult at low temperatures (mainstream).

DO

Perform deammonification process in a single reactor is a challenge due to the variety of bacterial groups involved. Nitrosomonas, Nitrobacter and Nitrospira are the bacterial groups that perform the Nitrification and all of them need oxygen to carry out the different processes; on the other side, Anammox performs strictly under anaerobic conditions. Furthermore, AOB overcome over NOB should be ensured for a successful Nitrogen removal.

Aeration patterns of on and off intervals are proven to create an anoxic and aerobic environment. This is one of the central strategies when creating an IFAS PN/A single reactor, the changes between aerated and non-aerated periods ensure the overcome of AOB over NOB, at least in side stream treatments. After several studies were conducted, different strategies have reached different ideal aeration patterns that are summarized by Robiglio, 2018:

 According to several authors, low DO concentrations (0.15-1 mg/L) are favorable for the NOB suppression and AOB growth.

 Anammox bacteria are anaerobic organisms, thus low DO concentrations do not contribute to its growth. Even if it positively contributes to the suppression of NOB, it is maybe not the most suitable strategy.

 Previous investigations conducted on an IFAS pilot by Malovanyy (2015) and on a

comparable one by Trojanowicz (2016) suggested that one suitable DO set point is 1.3mg/L

with 20/40 in aerated and non-aerated pattern; results showed that this pattern usually reaches an acceptable nitrogen removal.

 Set the control value in the ammonium concentration on the reactor instead of on the aeration pattern. When this value was reached, the aeration was interrupted and so on.

Aeration patterns not just depend on the overall aeration period per day (i.e. 8h/day) but the intervals defined (i.e. 20 aeration /40 non-aeration or 10/20 which is the same as 8h/day). These intervals ensure that some bacteria, in this case AOB, can grow more comfortable than others (NOB).

Furthermore, the bacterial growth kinetics have a unique trend for each group (k: growth rate), this k can be related with the DO concentration and the time the reactor takes to get to this DO set point.

Though in a complex process as the PN/A, other parameters have to be considered because the DO

set point cannot affect other processes’ performances.

6 Materials and Methods

6.1 Pilot scale Single-Stage Mainstream Deammonification

The pilot was located in Hammarby Sjostadsverket (SSV), a research facility focused on the development of new and improved water treatment methods and technologies. SSV facilities are located above one of the two wastewater treatment plants in Stockholm, Henriksdal WWTP. Some inflow from the inlet channel to Henriksdal is pumped to Sjöstadsverket to be used in the research projects.

Before this thesis, several investigations and different set up configurations were carried out by different students and researchers in SSV. The pilot was first set up in 2008 with two moving bed film reactors (MBBR) of about 200L each, with K1 Kaldness biofilm carriers, a mixing stirrer and controlled DO concentration and temperature. These reactors were meant to treat continuously fed supernatant. After its set up, a number of research activities took place in this pilot including, side- stream treatment by deammonification process, N2O emissions and mainstream treatment by deammonification, being the last one tested with a high ammonium concentrated inflow after an ion exchange process, also evaluation of temperature and low concentrations effects and combination of anaerobic treatment with UASB and mainstream deammonification process were tested (Deammonification report 2014, IVL).

The reactor was then switched to an IFAS configuration based on different researchers results including Malovanyy et al. (2015), that proved a hybrid system (MBBR – Activated Sludge) could improve the performance of a deammonification process. The IFAS set-up improved results in terms of nitrogen removal efficiency compared to MBBR systems.

Since 2015 the pilot has run as IFAS and some master thesis focused on different parameters have

been developed. Along this document, some of these results are mentioned and discussed.

6.1.1 System Characteristics

Figure 4 Configuration IFAS Pilot (Modified. Robiglio, 2018)

As it is mentioned in the previous section, the deammonification process was preceded by a UASB reactor used for organic matter removal. This reactor had a working volume of 2.5 m ³ (Total Volume:

3.2 m ³ ) and it was fed with primary settled municipal wastewater from the Henriksdal. The average flowrate was ≈0.88 m ³ /h and the reactor was operated at temperatures between 20 and 28° C.

The flow was then pumped through a system of four filters that removed suspended solids and some of the non-soluble organic matter. After this step, the water was stored in an equalization tank with variable retention time before being pumped into the IFAS reactor. The flowrate pumped from the equalization tank will be referred as Inflow (IN) in this document and it was variable during the experimental period, the parameters considered to define the inflow rate are mentioned further in the document.

The IFAS reactor had a volume of 0.2 m ³ and was running in continuous mode for most of the experimental time, for some short periods it was set to run in batch mode, especially when fast recovery was needed (after operation problems). The reactor was filled with Kaldnes K1 biofilm carriers (500 m ² / m ³ ) up to 40% of its volume and because it was in complete mix, the biofilm carriers were in constant suspension as well as the activated sludge (Robiglio, 2018). Two sedimentation tanks in series were located after the IFAS (Figure 4).

The pilot included a control system with online signals from sensors and pumps, facilitating the control and monitoring of the process. The complexity on the microbial processes of the PN/A in compare to other systems, required the use of several sensors. The control system (PID) included the following signals:

2,5

 Inflow:

o Ammonium (NH4) sensor o Conductivity sensor

 Reactor

o DO sensor

o Nitrate (NO3-N) sensor o NO3-N and NH4-N sensor o pH meter

o Conductivity sensor

o Suspended solids (MLSS) sensor o Rod-Cooler to control the temperature

The Inflow had a flowrate varying from 120 mL/min to 200 mL/min, low concentration of Ammonia (30 – 50 mg/L), COD/NH4-N ratio between 1.4 and 3.2 and low temperature during winter season (around 15°C) (typical mainstream values). Specific water characteristics are widely explained in the results section further in the document.

The key parameters to control the process included aeration patterns and DO concentration, Temperature, inflow rate and RAS rate in the reactor.

All these characteristics were changed during the experimental time to find the most suitable combination to contribute towards the suppression of NOB’s and further optimization of the PN/A process.

6.1.2 Online Monitoring

Throughout the months of this thesis project, the control system played an important role on the monitoring and control of parameters on the PN/A system. The importance of the online monitoring relied on the control that the system provided, making the defined strategies stable, reliable and robust.

The software used at SSV was Uni-view 9.01 - open

source. Uni-view stores data every minute for every

impute signal, it allows to create any type of graph

and/or combine different signals with comparison

purposes, it calculates average, minimum and

Control System

maximum values in any length of periods and it allows an effective visualization of the process, among other characteristics.

Figure 2 shows the software visualization of the PN/A process. The input signals mentioned in earlier chapters are located just in the reactor and first sedimentation tank and their specific values are shown in the screen as it can be seen.

Online monitoring contributed to have a proper overview of the process and apply the control strategies defined along the study.

Data collection was done at least once a week and saved in an excel file where then was analyzed.

6.1.3 Operation and Control Strategies

PN/A is a well stablished technology and it is widely applied for reject water worldwide. Lackner (2014) summarized the main problems in full-scale PN/A processes, giving greater relevance to mixing problems followed by influent pump troubles, blower failure and other oxygen related problems. In the third group of problems are pH shock, temperature variation and influent solids concentration. Even though this thesis is investigating mainstream PN/A, the complexity of the microbial community kinetics and behaviour can be treated as the same, since the critical part of the process is to successfully achieve the short-cut in the nitrification process. Therefore, the difficulties reported by several authors including the one mentioned before were addressed and the control strategies were design to overcome these problems.

6.1.3.1 Temperature

The main goal of this thesis was to monitor and show the variation between different bacterial groups’

interaction in Mainstream PN/A, thus due the importance of the temperature in the bacterial behavior, it was chosen as a key parameter for the evaluation of the mainstream deammonification process.

In order to perform the evaluation, especially in cold weather where most groups have less activity

and growth, it was necessary to simulate the temperatures in the pilot reactor. The reactor started at

20°C for the bacterial activity tests and then it was decreased gradually during the experimental

period until 15°C were reached.

The temperature was changed according to the following schedule:

 21°C -> 11 ^th June – 1 ^st July

 21-24 °C -> 1 ^st July – 15 ^th August*

 20°C ->15 ^th August – 24 ^th September

 18°C ->24 ^th September – 5 ^th November

 15°C ->5 ^th November – 12 ^th December

(*) Due to unusual high temperatures during the summer of 2018, there was a period with an un- planed temperature increase in the reactor (23-24°C). Unfortunately, because it was during the stabilization phase no bacterial activity tests were carried out (June and July).

6.1.3.2 Aeration patterns and DO concentration

Previous projects developed in the same pilot tested different aeration patterns and DO concentration set points on the PN/A process. Projects developed by Robiglio (2018), Los (2018) and Malovanyy (2015) were considered when deciding the best DO starting point.

However, after starting the bacterial activity tests and comparing the growth kinetic rates between NOB and AOB, other patterns were tested. DO concentration and Aeration pattern varied as follows:

 20 min aeration – 40 min non aeration (recommended by previous studies)

 10 min aeration – 20 min non aeration

 5 min aeration – 10 min non aeration

 10 min aeration – 20 min non aeration (most suitable pattern according to bacterial behavior)

Figure 5 Oxygen relation with bacterial behavior (Modified Henze et al., 2017)

6.1.3.3 Flowrate

The flowrate that is entering the reactor is part of the Hydraulic Retention Time (HRT)’s calculation.

The HRT define the time the influent stays in the system and therefore, the Nitrogen loading rate and

the behavior of the AOB and Anammox bacteria according to (Feng, et al., 2017).

6.1.3.4 Suspended Solids Concentration MLVSS

MLVSS is the sludge suspension in the reactor on a biological treatment process. There are two other key concepts related to this parameter that define the biomass solids in the sludge (TSS: as Total Suspended Solids) and the Volatile Suspended Solids (VSS).

During previous studies developed in the same IFAS reactor, it was defined that the concentration of solids is one of the most important parameters when achieved high nitrogen removal efficiencies. The high TSS content is well known and studied, but the process and operational parameters in order to maintain a high concentration of solids in the reactor was the challenge during this project.

Suspended solids as a control strategy was defined as: Suspended solids should be higher than 1000 mg/L at all times. However, loss of sludge due to sludge flotation in the sedimentation tank, clogging of the pipes, inadequate retention time and stops in the recirculation pumps were some of the problems faced by the reactor during the project development and thus, this strategy was not entirely achieved as it will be further explained in results.

6.2 Chemical Laboratory Analyses

Despite all the online measurements available to monitor the process, it was necessary to schedule sample analyses in order to obtain nitrogen concentration profiles on the system and mass balance, furthermore to check, calibrate and follow the online signals’ performance. During the 6 months duration of the project, it was scheduled to take samples and make the chemical laboratory analyses twice per week for the inflow and outflow. As it is mentioned before inflow is defined as the water that comes into the reactor from the equalization tank and outflow as the flow leaving the reactor.

However, due to functionalities offered by the system’s design, the samples marked as outflow were taken inside the reactor, assuming that the values of the measured parameters did no change during the sedimentation process.

Furthermore, the chemical analyses were carried out using cuvettes. The results were used to graph the process performance and calibrate the instruments. These results are shown and evaluated in the results section 8.1 and 8.2 of this report.

Table 2 Operational schedule

The schedule including chemical analysis is shown in Table 2. Tuesdays and Fridays were strategically

chosen for analysis and subsequent calibration of the instruments, this following other studies

experience where Mondays is a recovery day after weekend, thus Tuesday, and that the process should be left as precise as possible before the weekend, thus Friday. The chemical analysis schedule was followed almost during the whole experimental period except two weeks during July 2018 where the reactor was not working properly.

The parameters measured were for IN –> COD Total and Dissolved, Total Nitrogen (TN), NH4-N, Alkalinity, Potassium (K) and pH; and for REACTOR/OUT –> COD D, TN, NH4-N, NO3-N, NO2- N, Alkalinity, K, Chloride (Cl) and pH. Furthermore, the DO in the reactor was measured with a hand DO-meter twice per week and compared to the online value.

The procedures followed to perform the analyses were standard for the used cuvettes. Previous filtration through a 45µm filter was necessary in order to remove suspended solids in every sample (except COD T), before carrying out the actual procedure described. Lastly the samples were measured in the spectrophotometer and the results were collected.

The inflow sample was taken from the influent and filtered directly. On the other hand, the Outflow sample was taken inside the reactor assuming no reactions were happening in the sedimentation tank;

this sample was then passed through a coffee filter in order to remove a percentage of the solids and facilitate the required filtration (45µm) before chemical analysis.

Total Suspended Solids (TSS) and Volatile Suspended Solids (VSS) were included in the laboratory analyses schedule. They were carried out following the Standard Methods twice per week for the reactor and once per week on the washed sludge used for activity tests.

6.3 Bacterial Activity Tests

6.3.1 Experimental set up and devices

During the experimental period, two different set-up Configurations were needed. The first one for the Oxygen Uptake Rate (OUR) test and the second one for the Specific Anammox Activity (SAA) and Nitrate Uptake Rate (NUR) tests.

OUR tests

The laboratory equipment included one water bath kept at 25° C, one submerged magnetic stirrer, one three neck glass bottle batch reactor (Woulff type) with a volume of 1 dm ³ , one Dissolved Oxygen meter (DO meter) with data recorder (HQ40D / brand: Hach instruments), two syringes 5 c m ³ with needles and an air compressor.

The following chemical solutions were used in the OUR tests:

Chemical Concentration

 Sodium Chlorate (NaClO3) 580 g/L

 N – Allylthiourea (ATU) 1.25 g/L

 Ammonium Chloride (NH4Cl) 13.36 g/L

 Sodium Acetate (CH3COONa) 14.60 g/L

 Potassium Dihydrogen Phosphate (KH2PO4) Used for phosphate buffer

 Dipotassium hydrogen phosphate (K2HPO4) Used for phosphate buffer

All the solutions were mixed in Sjostadsverket’s laboratory under safety work related regulations and following the standard methods.

SAA and NUR tests

The laboratory equipment included one water bath kept at 25 ° C during all the experiments, one submerged magnetic stirrer, four gas proof glass batch reactors with a volume of 300 ml, Nitrogen gas, silicone and a pressure meter (Handheld Pressure-Meter water-proof, with data logger - GMH 5150 / brand: Greisinger).

The following chemical solutions were used in the SAA and NUR tests:

Chemical Concentration

 Ammonium Chloride (NH4Cl) 13.36 g/L

 Sodium Acetate (CH3COONa) 14.60 g/L

 Sodium Nitrite (NaNO2) 17.24 g/L

 Sodium Nitrate (NaNO3) 9.24 g/L

 Potassium Dihydrogen Phosphate (KH2PO4) Used for phosphate buffer

 Dipotassium hydrogen phosphate (K2HPO4) Used for phosphate buffer

6.3.2 Schedule and Sampling

The experimental period performing activity tests started on august 2018 when the reactor was stable

and were terminated on December 12 of the same year. At the beginning of the thesis project, just

OUR and SAA were being performed because it was believed that the data collected from those results

was enough to describe the bacterial behavior. With just two activity tests to perform per week and

based on the reactor’s behavior, it was decided that the most suitable days to carry out the

experiments were Wednesdays and Fridays; however, some SAA experiments were carried out on

Thursdays.

Further on time NUR tests were decided to be necessary and when they started to be performed, it was necessary to moved SAA to Thursdays, leaving Fridays to perform NUR during the month of October.

The IFAS reactor performs the PN/A process through biofilm or carriers and sludge, thus the activity tests had to be carried out in both. Sludge and carriers sampling was done at the same time and with the same procedure. Furthermore, the experiments were conducted in duplicates for SAA and NUR and triplicates for OUR, in order to have better accuracy in results and data collection.

The tests were scheduled to have duration of between 5 and 8 hours for the total of 4 batch reactors (sludge and carries in duplicate) for SAA / NUR and 6 reactors for OUR. Though this time depended mainly on the bacterial activity, temperature and experimental set up, which varied during the project.

During the first phase of the project that consisted on the two first weeks of activity tests (OUR and SAA), the sludge was washed the same morning of the experiment. However, the amount of time necessary to wash the sludge ended up being higher than predicted (around 3-4 hours depending on the SS concentration) and the time was not enough. Thus, the procedure had to be modified and sampling of sludge and carriers was then done one day before the OUR test was carried out (Tuesday’s).

The sample was taken, washed (*) and kept covered and aerated for the rest of the experiments. This procedure varied due to problems with the bacterial activity, especially on the sludge due to changes in the environment Temperature and DO concentration. These changes occurred mainly during the first 5 weeks of the project, however they will not be mentioned in this thesis since the data shown further in the document corresponds to the week 7 and beyond.

It is important to mention that during November, on the last stage of this thesis project the experiments were suspended for two weeks and a half.

The detailed schedule is shown in Table 2.

(*) Procedure for washing the sludge: The activated sludge (AS) sample is left for sedimentation and further decantation. The extra water is purred out through a siphon and the remaining decanted AS is then transferred into a 1 dm ³ beaker. Phosphate buffer solution is added and mixed, after decantation the extra water is purred out. The procedure should be done at least 3 times.

Procedure for washing the carriers: Carriers are taken from the reactor with a strainer where they

are rinsed with water. They are then put in a container filled with phosphate buffer, then it is

shacked and emptied. The procedure should be done at least 3 times.

6.3.3 Oxygen Uptake Rate (OUR)

The method used for the determination of AOB and NOB in the biofilm and activated sludge was OUR determination by respirometric method taken from KTH documents. The test consists on oxidation rate measurements of ammonium-nitrogen (N-NH4), Nitrite-Nitrogen (N-NO2) and COD in a batch reactor. Evaluation of AOB, NOB and heterotrophs is the main goal.

The inhibition of the NOB and AOB metabolism is made through the addition of chemicals. Sodium chlorate is a selective inhibitor of NOB and ATU (Allylthiourea) inhibits both AOB and NOB, which allow the determination of bacterial presence in either carriers or AS by the measurement of the Oxygen Uptake Rate. Oxygen is needed by bacteria as a substrate for growth, thus the DO concentration rate decreases proportionally to the activity and the type of bacteria growing.

In order to ensure that other factors were not affecting the OUR, the temperature was kept constant (at 25°C) and other substrates were limited and controlled. As a result, the DO concentration decreases lineally, and the calculations were carried out.

Taking into account the graphs derived from the DO values against time, the slope of each linear regression applied to the three phases of the experiment was calculated:

1. Initial (all bacteria)

2. After NAClO3 addition (NOB inhibition) 3. After ATU addition (AOB and NOB inhibition)

The OUR value is then calculated from ammonium ions oxidation to nitrites (AOB activity) and nitrites to nitrates (NOB activity).

1. OUR NOB = OUR AOB+NOB+H -OUR AOB+H =OUR T -OUR NaClO3

2. OUR AOB = OUR AOB+H -OUR H =OUR NaClO3 -OUR NaClO3+ATU

Calculation of oxygen uptake rate:

𝑂𝑈𝑅 = 𝑎 ∙ (60 ∙ 60 ∙ 24) ∙ 𝑉 𝑚𝑔𝑂 𝑑 Where:

OUR – oxygen uptake rate [mgO2/*d],

a – slope value of the curve (calculated with linear regression method) [mgO2/dm ³ *s], VR – volume of the batch reactor [dm ³ ].

Laboratory Procedure:

Around 5 litters of influent water were sampled directly from the equalization tank at the beginning

of the experiment. After preparing all the set-up regarding bath tub, temperature of water (25°C),

mixer and computer connected to the DO meter, the preparation of the reactors were done as follow:

 107 carriers or 100ml of activated sludge were added along with a magnetic mixing stirrer.

 Inflow wastewater was added until up to 1L.

 Chemical addition:

o 25mL of NH4Cl o 15mL of CH3COONa

The reactor was then aerated with an air compressor and the DO probe was put on top of the reactor.

Then the measurements were started and after 6 mins 5 ml of NaCLO3 were added. After 6 other minutes 5 ml of ATU were added and the DO was being measured for 6 minutes more.

The data was then saved in the computer and each experiment was done in triplicate.

a. Biofilm – Carriers

The previous section showed the general method for calculating and determining the OUR. However, when developing the method for sludge and carriers there were things to take into account.

The formula contains a specific area, which in the carriers is determined by the area inside each carrier where biofilm is attached. For matters of this calculation, it was assumed (found in literature and measured by the supplier) that each carrier had an effective area of 0.046729 m ² and it was defined an experimental area of 0.05 m ² of biofilm. Thus, for each experiment 107 carriers were counted and placed in the reactors.

Then the formula was modified and is as shown below:

𝑆𝑂𝑈𝑅 = 𝑂𝑈𝑅 1000 ∙ 𝐴

𝑔𝑂 𝑚 ∙ 𝑑 Where:

OUR S – specific oxygen uptake rate [gO 2 /g s.m. *d], A . – area of biofilm in the batch reactor [ m ² ].

b. Activated Sludge

The activity was calculated based on the amount of biomass on the reactor (X ’vss ); which was determined carrying out a regular VSS test on the sludge.

The amount of Sludge poured on the reactor was defined as 100 ml of washed sluge; and the formula was used as it is shown below:

𝑆𝑂𝑈𝑅 = 𝑂𝑈𝑅

1000 ∙ 𝑋

𝑔𝑂

𝑔𝑉𝑆𝑆 ∙ 𝑑

Where:

SOUR – specific oxygen uptake rate [gO2/gVSS*d], X VSS – amount of biomass in the batch reactor [gVSS].

6.3.4 Specific Anammox Activity (SAA) and Nitrate Uptake Rate (NUR)

The method for SAA and NUR was calculated in a similar way and its differences relied on the chemicals used when developing the experiment and will be further explained. This method was extracted from KTH documents.

The manometric method was chosen for evaluating the Specific Anammox Activity and Nitrate Uptake Rate through the measure of the nitrogen removal capacity of Anammox, denitrifying and heterotrophic bacteria. As it is manometric, the method bases its calculations on the overpressure produced in the batch reactor. The results are then expressed in the amount of nitrogen that can be removed by 1 g of activated sludge (VSS) or one square meter of biofilm during 24 h.

In order to express the pressure values in nitrogen removal rate it is necessary to use the following formula where the thermodynamics universal gas constant is included as well as the slope of the graph obtained from the overpressure changes on the reactor and the bacterial presence in a specific area or gr of VSS in the sludge.

Laboratory Procedure:

This experiment did not require as much attention during the actual development of the experiment but the preparations before the recording took around 1.5 - 2 hours. Thus, in this section the procedure will be summarized and in order to avoid over explaining some steps will not be mentioned.

Four batch reactors were prepared in duplicate as follows:

The reactors were completely sealed in order to ensure the anaerobic conditions through the entire experiment. Each batch reactor was measured during 2 hours keeping a constant temperature of 25°C.

The procedure for NUR was the same though 5 mL of CH3COONa was added replacing NH4 ⁺ and 5mL of Nitrates (NO3) instead of Nitrites (NO2). The rest of the procedure was carried out as explained previously.

As mentioned in the OUR procedure, there were differences in calculations for carriers and sludge, they are mentioned in the next section.

a. Biofilm – Carriers

The formula used for calculation of SAA and NUR is the same and thus, the same assumptions were made. 107 carriers representing 0.05 m ² of biofilm were used and the calculations done using the following formula.

𝑆𝐴𝐴/𝑁𝑈𝑅 = 𝑎

1000 ∙ 𝑀 ∙ 𝑉

𝑅 ∙ 𝑇 ∙ (60 ∙ 24) ∙ 1 𝐴

𝑔𝑁 𝑚 ∙ 𝑑 [gN/m ² -d],

a – slope value of the curve (calculated with linear regression method) describing changes of pressure in the batch reactor in the function of time [mbar/min],

M N2 – molecular weight of dinitrogen [g/mole] (28 g/mole), Vg – head-space volume in the batch reactor [cm ³ ],

R – universal gas constant [bar*cm ³ /mol*K] (82.077 [bar*cm ³ /mole*K]), A- Total biofilm area in the batch reactor(m2)

b. Activated Sludge

200 mL of sludge were used for the calculation of SAA and NUR.

𝑆𝐴𝐴/𝑁𝑈𝑅 = 𝑎

1000 ∙ 𝑀 ∙ 𝑉

𝑅 ∙ 𝑇 ∙ (60 ∙ 24) ∙ 1 𝑋

𝑔𝑁 𝑔 . ∙ 𝑑 [gN/gVSS*d],

a – slope value of the curve (calculated with linear regression method) describing changes of pressure in the batch reactor in the function of time [mbar/min],

MN2 – molecular weight of dinitrogen [g/mole] (28 g/mole), Vg – head-space volume in the batch reactor [cm ³ ],

R – universal gas constant [bar*cm ³ /mol*K] (82.077 [bar*cm ³ /mole*K]),

X VSS – amount of biomass in the batch reactor [gVSS].