熊本大学学術リポジトリ

Kumamoto University Repository System

Title Studies on fixed bed anammox process treating low

strength ammonium containing wastewater with high

…

Author(s) 劉, 成良

Citation

Issue date 2009-03-25

Type Thesis or Dissertation

URL http://hdl.handle.net/2298/14296

Right

Studies on fixed bed anammox process treating low strength

ammonium containing wastewater with high salinity

(高塩分含有低濃度アンモニア含有排水の固定床アナッモックスプロセスに関する研究)

Doctoral Dissertation

March ，2009

By

CHENGLIANG LIU

SupervIsor

Pro f. KENJI FURUKAWA

Department ofNew Frontier Science

Graduatc School of Science and Tcchnology

KUMAMOTO UNIVERSITY ， JAPAN

Acknowledgement

I would like to acknowledge with deep gratitude to all those who helped me finish this

proJ ect.

I want to thank my advisor, Kenji Furukawa, professor of the Graduate School of

Science and Technology in Kumamoto University, for his assistance, guidance and support

during my four years of study (guest researcher, one year; Doctor Study, three years).

Without his constant consultant in my writing this thesis, it would not have been possible.

Prof. Furukawa is a very warmhearted man, his kindness and emotion investment to his

Chinese students is famous in Department of Science and Technology. I can never forget that

it was professor Furukawa who supports me with his own research fund. I will not forget this

matter during all my life.

I am very gratefu1 to Prof. Shinichi Abe, Prof. Susmu Takio, Prof. Yoshito Kifazono of

Kumamoto University and Prof. Haobo Hou of Wuhan University, China, deeply appreciate

for their encouragements and supports in my study and care on my life in Japan. Thanks ProÅí

Takao Fujii and Dr. Takashi Nishiyama for their kind helps to my experiment.

' I have a real debt of gratitude to Miss Seiko Seitoh and Miss Aya Fujimoto for their

enthusiastic assistance ofmy applying arbiter bill and experimental help, respectively.

Thanks for help of Mr. Xiaochen Xu, who always gave me encouragement. He is the

man who has the willing to reason and comprehend someone not by appearance phenomenon

but by the essence of spirit at early time he contact. I am particularly indebted to Chunfang

Zhang, Mr. Chunlai Li, Mrs. Jiali Yang, Mrs. Yingiun Cheng, Mr. Yuanhua Xie, Mrs. Yunling

Feng, Mr. Cheng Chen, Mr. Zhigang Li, Mr. Jiachun Yang, Mr. Xiangsheng Cao, Mr. Wenjie

Zhang, Mrs. Li Zhang, Mr. Yongguang Ma, Dr. Hu jing, Mrs. Yanning Gao, Mr. Chuangming

Zhou and Vice prof. Changcun Chao for their kindly supports and encouragements in usual

life and study in Japan.

My heartfelt thanks are due to Japanese students of Furukawa's laboratory for their

kindly help and warm friendship. Thanks M). Taichi Yamamoto for his kind advice in my

research work. I am particularly indebted to Mr. Lai Minh Quan, Mr. Hashimoto Yoshinori,

Mr. Do Phoung Khanh, Mr. shinohara Takehhiko, Mr. wakamatsu Shingo, Mr. Tran Thanh

Liem, Mr. Okuda Yutaro, Mr. kaneshiro Kosuke, Mr. Nakashima Ryosuke and Mrs.

matsumoto Noriko for their kindly supports and encouragements in usual life and study in

Japan.

Finally, I would like to give my special thanks to my wife and lovely son for supporting

my scientific career. Because of my busy study, all the housework was on my wife's

shoulders. But she had never complained. Her encourage in hard times and warmhearted

love enabled me to complete this work.

Abstract

The anammox process (!2Ltaerobic ammonium g2tgiidation), a new process for ammonium

removal from wastewater, can remove ammonium without oxygen and with nitrite as the

electron acceptor. The costs of conventional ammonium removal via

nitrification-denitrification from sludge reject waters are mainly associated with the

reduction in aeration-cost and supplementation cost for extemal carbon source such as

methanol in denitrification step. The anammox process provides an alternative for the

nitrification, with no requirement for an external carbon source. Combined with partial

nitritation (PN) process such as SHARON-process, the total aerati•on costs are reduced with

25 O/o. Compared to conventional nitrificationldenitrification, PN-anammox process can save

100 O/o of the required carbon source (i.e. methanol) and 50 O/o of the required oxygen. This

leads to a reduction of operational costs of 90 O/o, a decrease in C02 emissions of more than

100 O/o (the process actually consumes C02), and a decrease in energy demand. Wastewaters

that are suitable for the treatment with anammox is sludge reject waters ("sludge liquor") and

industrial wastewater (such as sour water) and brine wastewater from natural gas production.

The two processes can usually be engineered in two separate reactors, or in a single vessel

(the CANON-concept, "gompletely Autotrophic N-removal 9ver N"i'trite").

In this dissertation I have developed a novel anammox reactor using small carriers and

succeeded in its operation under high salts concentration.

The first research work was focused on "Anammox treatment of low-strength

ammonium-containing wastewater under room temperature". The anammox is commonly

used for the treatment of high-strength ammonia-containing wastewater at rather high

temperature of 30-40 Åé. Novel up-flow anammox fixed-bed reactor using a thin sheet of

polyester nonwoven biomass carrier could be established within 3 months of operation. This

fixed-bed anammox reactor was operated at room temperature for the treatment of

i

low-strength ammonium wastewater for 900 days. Influent containing ammonium and nitrite

in the forms of (NH4)2S04 and NaN02 with other nutrients and buffer was supplied to the

reactor in up-flow mode. The ratio ofNH4-N and N02-N in the influent was maintained at

'1.0 to ensure N02-N limiting conditions during treatment. At t-he first step, 8 strips of

-3original nonwoven were used and the maximum TN removal rate (TNRR) of 1.2 kg-TN m

d-,i was obtained. At the second step, 16 strips nonwoven was used as biomass carriers and

the maximum TN loading rate (TNLR) of2.4 kg-N m-3 d"i was obtained. At the third step,

small size biomass caniers composing nonwoven cutting average length of 1-2 mm were

used as biomass caniers and fi11ed in the gap of strips. High TNRR of 4.9 kg-TN m-3 d-i

was obtained finally. At the last step, eleven layers of small biomass beds were used, and the

maximum TNRR of 9.93 kg-TN m'3 d-i was obtained after 720 days of operation with TNLR

of 12.5 kg-TN m-3 d-i . When the system had been fu11y developed, a stable anammox

treatment lasted for three months. The anammox reactor was operated under N02-N limiting

condition for whole experimental period during which anammox treatment of a relatively

low-strength ammonium-containing wastewater at 20-25 eC was successfu11y demonstrated.

tt The second research was focused on "Effect of salt concentration in anammox treatment

' ' 'using non-woven biomass carrier". Effects of salt concentration on fixed-bed anammox

' 'reactor using non-woven biomass canier was investigated. A freshwater anammox sludge

'was used for the inoculum, and the salt concentration was gradually incrgased from 2.5 g 1-i

' tt tt 'to 33 g 1'i. The anammox treatment was stable at the salt concentration of 30 g 1-i and

'TNRR of 1.7 kg-N m-3 d-i was stably kept for 65 days. However, the NRR was sharply

/. t tt tt t. ' t .t.deteriorated above the salt concentration more than 30 g 1-i. The bacterial cgnsortia were also

examined by 16S rRNA gene analysis before and after acclimation of the anammox sludge to

ttt t t tthigh galt concentrations. Although all results of major clgne were rel.ateq, to .uncultured

'bacterium clone the two entries were affiliated with candidate division OPIO. And the

' t tt/ tfreshwater anammox bacteria KU2 and KSU-1 were detected at the salt concentration of30 pe tt tt . . .tt ' ttt t ttg 1- 1. Ip addition, Lysobacter, sp. belonging to 7 Proteobact,eria were revealed to coexist with

anarhmox bqcteria.

t/ .t The third research was focused on "Nitrogen removal capabilities of two kinds of

' tt t ttt t

ii

fixed-bed anammox reactors for treating partial nitrified brine wastewater". Fixed-bed

anammox reactors using non-woven biomass carrier and small size biomass caniers were

operated under salinity condition' comparable to sea water level. Fixed-bed anammox reactor 'using nonwoven biomass carrier could be successfu11y adapted to high salinity conditions by

elevating influent NaCl concentrations from 2.5 to 33 g 1-i stepwisely. This anammox

reactor could keep stable operation even under salt concentration comparable to sea water

( 30 g l-i ofNaCl), and the highest TNRR of 1.7 kg-N m-3 d-i was obtained After about one

year of continuous operation under salt concentration comparable to sea water, the synthetic

inorganic salt wastewater was replaced with the effluent from partial nitritation reactor

treating brine wastewater containing NH4-N from natural gas producing company. The

reactor was operated for more than 80 days by feeding the partial nitrified brine wastewater

and the maximum TNRR of 1.9 kg-N m-3 d]i was obtained. •It was difficult to increase TNRR

'more than this value in this fixed-bed anammox reactor. Then, the nonwoven biomass canier

was replaced with small size biomass carriers for getting higher TNRR. TNLRs of the

fixed-bed anammox reactor using small size biomass carrier could be successfu11y increased

stepwise from O.89 to 2.41 kg-N m-3 d-i within 40 days of continuous operation. TNRR of

'2.52 kg-N m-3 d-i was achieved under TNLR of 3.18 kg-N m-3 d-i without clogging and

' ' 'channeling phenomena, which were observed in the operation of fixed-bed reactor using

' tt tnonwoven biomass carrier. The color of anammox sludge in this fixed bed reactor using

small size biomass carriers was changed to weak red from black on day 85. TN removal

tt ' ' 'capability of fixed-bed anammox reactor using small size biomass carriers was much higher

'than that of fixed-bed anammox reactor using nonwoven biomass carrier, and the usefulness

of small size carriers as biomass attachment was verified.

' t/t There are two types of anammox sludge, freshwater type and salinity water type. The

tt t tt tt tabove mentioned experimental data indicated clearlY the way of acclimation for freshwater

t/ t ' t

anammox sludge to the ammonium•-containing wastewater with high salinity comparable to

.t/ /t 'sea water level. As to the freshwater anammox sludge, the difficult point in its establishment

'of stable anammox consortia is to increase TNRR under low ammonium concentrations

without temperature control. As to the anammox sludge operated uhder high salt

tt ttt t t tt tconcentrtations, the difficult point in its establishment of stable anammox consortia is to

iii

obtain high anammox reaction rates. These difficulties could be overcome in this present

studies. Our obtained results will give a new open and perspective road for the further

application of anammox reaction to different kinds of ammonium containing wastewaters.

iv

List of contents

Acknowledgements

Abstract

List of contents

List of abbreviations

List of Tables

List of Figures

Chapterl Introduction

1.1 Biologicalnitrogencycle

1.2 Anamniox

1.3 Anammoxcellstructureandbiochemistry

1.3.1 Planctomycetes

1.3.2 Anammoxcellstmcture

1.3.3 Anammoxlipids

1.3.4 Anammoxbiodiversity

1.3.5 Biochemicalmechanism

1.4 Traditionalbiologicalnitrificationldenirificationprocess

1.5 Anammoxprocess

1.6 ,Partialnitrifications/anammoxprocess

1.7 SHARONandanammoxprocess

1 .8 OLAND and anammox process

1.9 CANONprocess

1.10 SNAPprocess

1.11 Researchwork

1.11.1 Problemstatement

1.11.2 Obj ectives of this study

1.11.3 Researchplan

1.12 ReferencesChapter 2 Anammox Treatment of Low-strength Ammonium-containing Wastewater at Room Temperature

2.1 Introduction

v

!P:ag9

1

V

IX

x

xi

1

1

2

3

3

4

5

6

6

7

8

9

11

12

13

14

14

15

15

16

17

23

23

2.2 Materialsandmethods ./ 2.2.1 Experimentalsetup

2.2.2 Seedsludge

2.2.3 Biomasscanier

2.2.4 Attachimmobilizationofanammoxsludgeonbiomasscarrier

2.2.5 Syntheticinfiuent

2.2.6 Protocol for increasing in TN loading rate

2.2.7 Operationalconditioninwholeperiod

2.2.8 Analyticalitemsandmethods

2.3 Resultsanddiscussion

2.3.1 Observationoftheattachedbiomass

2.3.2 Changes in total nitrogen loading rates and removal rates

2.3.3 PerformancesofTotalnitrogenremovalefficiencies

2.3.4 NH4-Nremovalperfomiances

2.3.5 N02-Nremovalperformances

2.3 .6 Removal ratio of NH4-N, N02-N and N03-N

2.3.7 Comparisonofanammoxtreatmentcapabilities

2.3.8 The reasons forhigh removal rate

2.4 Conclusions

2.5 ReferepcesChapter 3 Effect of Salt Concentration in Anammox Treatment Non-Woven Biomass Carrier

3.1 Introduction

3.2 Materialsandmethods

3.2.1 Set up and operational conditions ofanammox reactor

3.2.2 Analyticalmethods

3.2.3 DNAextractionandPCRamplification

3.2.4 Cloning and sequencing of 16S rRNA

3.2.5 Denaturinggradientgel-electrophoresis(DGGE)

3.2.6 Nucleotidesequenceaccessionnumbers '

3.3 Resultsanddiscussion

3 .3 . 1 Reactor performances under high salt conditions

3.3.2 Bacterialcommunityundersaltcondition

Using

Vi

24

25

25

25

26

26

26

27

27

27

28

30

31

31

32

33

33

34

35

37

37

39

39

40

40

41

41

42

42

42

44

3.4 ReferencesChapter 4 Nitrogen Removal Capabilities of Two Kinds of Fixed-bed Anammox Reactors for Treating Partial Nitrified Brine Wastewater

4.1 Introduction

4.2 Materialsandmethods

4.2.1 Anammoxreactors

4.2.2 Biomasscaniers

4.2.3 Brine water used in nature gas company

4.2A Partialnitritationprocess

4.2.5 Influentofanammoxreactors

4.2.6 Start-up and operational strategy ofanammox reactors

4.2.7 Analyticalmethods

4.3 Results and discussion

4.3.1 Performanceofthefixed-bedreactors 4.3.2 Changes in biomass color in fixed-bed anarmnox reactors

4.3.3 Small size biomass caniers for enhancing anammox performance

4.3.4 Comparison of anammox treatment capabilities under high salinity

condition

4.4 Conclusions

4.5 ReferencesChapter 5 Study on Long-term Stable Operation of High-rate Fixed-bed Anammox Reactor Using Different Biomass Carriers at Moderately Low Temperature

5.1 Introduction

5.2 Materialsandmethods

5.2.1 Anammoxreactor

5.2.2 Nowoven biomass carriers and small size biomass carriers packed in small carrier bed

5.2.3 Inoculumsandsyntheticinfluent

5.2.4 Start-up and operational conditions offixed-bed anammox reactors

5.2.5 Analyticalmethods

5.3 Resultsanddiscussion

5.3.1. Perfotmanceofanammoxreactor

5.3.2 Observationoftheattachedbiomass

vii

47

51

51

53

53

54

55

55

56

56

58

59

59

63

64

66

67

68

71

71

73

73

73

75

75

77

78

78

83

5.3.3 Small size biomass caniers for enhancing anammox performance

5.3.4 Mathematic simulating model for increasing TNLR when using small size biomass canier

5.3.5 Comparisonofanammoxtreatmentcapabilities

5.4 Conclusions

5.5 References

Chapter6 ConclusionsandRecommendations

6.1 Conclusions

6.2 Recommendations

Appendix: Publications Related to This Dissertation

84

85

87

88

88

93

93

95

96

Viii

ANAMMOX

AOB

CANON

DGGE

DO

FBR

HRT

KSU-1

KU1, KU2

PN

NOB

OLAND

PCR

PE

TNLR

MRRSEM

SHARON

SNAP

TN

MLSS

MLVSS

List of abbreviations

' 'Anaerobic ammonium oxidation

' ' ' t tttAmmonium-oxidizing bacteria

Completely autotrophic nitrogen removal over nitrite

' 'Denaturing gradient gel electrophoresis

Dissolved oxygen

Fluidized'bed reactor

Hydraulic retention time

t.registered names of anammox strains in NCBI database

registered names of anammox strains in NCBI database

Partial nitritation

Nitrite-oxidizing bacteria

Oxygen-limited autotrophic nitrification-denitrification

Polymerase chain reaction

Polyethylene

Total nitrogen loading rate

Total nitrogen removal rate

Scanning electron micrograph

Single reactor system for high ammonium removal over nitrite

Single-stage nitrogen removal using anammox and partial nitritation

Total nitrogen

Mixed liquor suspended solids

Mixed liquor volatile suspended solids

ix

Table

Table 2-1

Table 2-2

Table 2-3

Table 3-1

Table 3-2

Table 4-1

Table 4-2

Table 4-3

Table 5-1

Table 5-2

Table 5-3

List of Tables

Title

Composition of synthetic wastewater

Minimum HRT for each Run

Comparison of anamniox removal capabilities

Composition of synthetic wastewater

Homology search results for 16S rRNA gene sequences ofbacterial

members in the community existing under salt condition (day 320).

Composition ofbrine water and sea water

Operational conditions during the whole experimental period

Comparison of different anammox reactors under high salinity

condtion

' 'Corpposition of synthetic wastewater

'Operational conditions during the whole experimental period

Comparison of different anammox reactors under different

operational conditiOns

:Pagg

26

30

33

39

46

55

58

67

75

77

87

x

Etigy!ere

Fig. 1-1

Fig. 1-2

Fig. 1-3

Fig.

Fig.

Fig.

1-4

1-5

1-6

Fig. 1-7

Fig. 1-8

Fig. 1-9

Fig. 1-10

Fig. 1-11

Fig. 1-12

Fig. 2-1

Fig. 2-2

Fig. 2-3

Fig. 2--4

Fig. 2-5

Fig. 2-6

Fig. 2-7

Fig. 2-8

List of Figures

Title

Biological nitrogen cycle

Anammox reaction pathway

The first full scale anammox reactor (2002) at the Dokhaven

wastewater treatment plant, Rotterdam, the Netherlands.( Photo:

Paques BV.)

Anammox granules from Dokhaven anammox reactor

Cell structure of anammox bacterium

TEM image of Candidatus "Brocadia anammoxidans" (photo John

FuerstlRick Webb)

Example of a chemical structure of a natural ladderane (ladderane Y)

from anammox cells. (above, left) , place of ladderanes X and Y in

anammox lipids electron microscopic image of two isolated(above,

middle) and anammoxosomes next to a panially lysed anammox cell

(above,right)

Biochemical mechanisms of the anammox process.

HH: hydrazine hydrolase, the hydrazine-forming enzyme;

HZO: hydrazine-oxidizing enzyme; NR: nitrite-reducing enzyme.

Nitrificationfdenitrification process

Partial nitrificationlAnammox

The SHARON process in a well mixed continuous flow reactor

Reactor configuration for PN (left) and anammox oxidatjon (rjght)

Schematic diagram ofexperimental system

Thin strips of nonwovens

Small-carrier of nonwoven

Color changes in anammox biofilm

Changes in TN loading/removal rate and N03-N production

Changes in influent and effluent TN concentrations and TN removal

efflciencies

NH4-N removal perfomances

N02-N removal performances

Xi

P;agg

2

3

3

3

4

4

5

7

8

9

11

12

23

25

25

27

28

30

31

32

Fig. 2-9

Fig

Fig

Fig

Fig

Fig

2-1O

2-11

2-12

3-1

3-2

Fig. 3-3

Fig. 3-4

Fig. 4-1

Fig. 4--2

Fig. 4-3A

Fig. 4-3B

Fig. 4-4A

Fig. 4-4B

Fig. 4-5

Fig. 4-6

Fig. 4-7

TN removal, N02•-N removal and N03-N production rates with

respect to NH4-N removal rate

Thin strips ofnonwoven carrier

Eight strips ofnonwoven carrier

Undoing nonwoven fiber mixed with anammox sludge

Schematic diagram of fixed bed anammox reactor

Time courses of nitrogen concentrations under various salt

concentratlons

Relationship between the nitrogen removal rate (NRR) and salt

concentration. Bars indicate standard deviation

Denaturing gradient gel electrophoresis (DGGE) profiles for 16S

rRNA gene sequences of bacterial members in the community

existing under high salt condition (day 320).

Fixed-bed anammox reactors

Original and split nonwoven biomass caniers (top view)

Nitrogen removal performance ofthe reactors. (Panel A)

Nitrogen removal performance of the reactors. (PanelB)

Ratios of TNRR, N02-N consumption and N03-N production rates to

NH4•-N consumption rate

Ratios of TNRR, N02-N consumption and N03-N production rates to

NH4-N consumption rate

Changes in color of anammox biomass during the whole experimental

period

Microscopic observations of anammox biomass attached on small size

carriers on day 148: (a) biomass attached on hydrated lime, (b)

biomass attached on nonwoven filaments, (c, d) biomass attached on

the mixture of small size caniers.

SEM images of small size biomass carriers and attached biomass. (A)

Hydrated small carriers without surface modification by hardcore-1,

without biomass attaclment; (B) hydrated small caniers after surface

modification by hardcore-1 without biomass attachment; (C, D)

biomass attaching on small size biomass carriers at different

magnifications on day 150

xii

32

34

34

34

39

43

44

45

53

54

60

60

62

62

63

65

66

Fig. 5-1

Fig. 5-2

Fig. 5-3

Fig. 5-4

Fig. 5-5

Fig. 5-6

Fig. 5-7

Fig. 5•-8

Fig. 5-9

Fig. 5-10

Fixed-bed'reactor

Original and split nonwoven biomass carriers (top view)

TN removal performance of the reactors

NH4-N and N02-N removal performance ofthe reactors

Ratios of TNRR, N02-N consumption and N03-N production rate to

NH4-N consumption rate. .External appearance of the anammox reactor during the whole

experlment

Appearance of the anammox biofilm token from the reactor

Size of small carrier bed

SEM images ofsmall size biomass caniers

Mathematic simulating model obtained in the end of the experiment.

X axis: time (days)

73

74

79

80

81

83

83

83

84

86

xiii

Chapter 1 Introduct .Ion

Concentrated wastewaters produced in many agriculture and food industries are now

commonly treated by anaerobic digestion process. High nitrogenous pollutants are also

produced from livestock farms, sites of aquaculture and landfi11 sites, etc. However, high

concentration of ammonium can not be removed in the anaerobic digestion process, thereby

yielding an effluent containing a high concentration of ammonia and low biodegradable

COD. The discharge of ammonium-rich wastewater caused nitrogen pollution for receiving

water bodies such as lake, inner sea and underground water. Nitrogen pollution is now the

critical environmental problem for the protection of water environment. These

ammonium-rich wastewaters are usually treated by biological nitrogen removal process, but

it is costly and troublesome. The original methods of nitrogen removal are physical and

ttchemical processes. Several phYsical and chemical processes, such as breakpoint

chlorination, air stripping and selective ion exchange have been used for nitrogen removal.

However, these processes are not applied widely due to their high treatment cost, operational

and maintenance problems and environmental concerns about the use of chemicals (Sedlak R.

et al., 1991). Therefore, traditional biological nitrificationldenitrification process is the next

option because of high potential removal efficiency, high process stability and reliability

(Sedlak R. et al., 1991, Watson S.W. et al., 1991). In the 1980s and 1990s, there were several

indications that nitrificationldenitrification are not the only processes that can remove

ammonium, but that there is an organism that can oxidize ammonium to nitrogen gas using

nitrite instead of oxygen as its terminal electron acceptor. This process is, therefore, called

"aLtaerobic ammonium gLtidation" (anammox). A lot of research works in this anammox

field are still going and its application to ammonium-rich wastewaters will contribute to the

prevention ofnitrogen pollution to receiving water bodies.

1.1 Biologicalnitrogencycle

The biological nitrogen cycle plays an

1

lmportant role in the mamtenanceof global

biosphere. It has been a focus of microbiological investigations since the late nineteenth

century. In the past 100 years, applied interests in the nitrogen cycle have shifted from

improving agricultural crop yields to concerning about surface water pollution, destruction of

the ozone layer and global wamiing. The biological nitrogen cycle, with the processes of

nitrification, denitrification, N-fixation and anammox, is shown in Fig. 1-1.

pt.ESX# Siee

xA '"

'

tisY"Xeeptfgee?dvSaj tpsIx"SS

Nesi

Fig. 1-1 Biological nitrogen cycle

1.2 Anammox

Anaerobic ammonlumOXIdation

es .R.

wt..

(anammox), i.e. the microbiological

"sSg,•pt&tt•ren

rdtwffTEImm. iEiam, iaj.

'

ee#si4

' t ttt ttt

Nge2ow

/

.t ttt

•• •• ••1•

•-

--•- -- eeltaj"•,

'

conversion of

uaftsr•

Fig. 1-2 Anammox reaction pathway

2

ammonium and nitrite to dinitrogen gas, is a very recent addition to our understanding of the

biological nitrogen cycle (Kuenen et al., 2001; Strous et al., 1999a). Discovered as late as

1986, it is so far the most unexplored part of the cycle. Given its basic features, the anammox

process is a viable option for biological wastewater treatment (Jetten et al., 1998;Jetten et al.,

2001;Strous et al., 1997b). Very recently, it was discovered that anammox made a significant

(up to 70 O/o) contribution to nitrogen cycling in the World's oceans (Thamdrup & Dalsgaard

2002). The anammox reaction pathway is shown in Fig. 1-2. The first fu11 scale anammox

plant has been installed in Dokhaven (Rotterdam, the Netherlands) for the ammonium

removal of digester liquor of domestic wastewater plant (Fig. 1-3). In this actual plant,

anammox granules were applied. Anammox granules from the Dokhaven anammox reactor

are shown in Fig. 1-4.

v

;.'"'•,E/.//./).,tt

""••'

ifl f•v"

Fig. 1-3 The first fu11 scale anammox Fig. 1-4 Anammox granules reactor (2002) at the Dokhaven Dokhavenanammoxreactor wastewater treatment plant, Rotterdam,

the Netherlands. ( Photo: Paques BV. )

1.3 Anammox cell structure and biochemistry

1.3.1 Planctomycetes

The planctomycetes are an interesting group of bacteria with many rare

3

from

or unlque

properties. The order Planctomycetes were described in detail (Schlesner H. et al., 1996) and

included only four genera (Planctomyces, Pirellula, Gemmata, and Isosphaera) with seven

validly described species (Fuerst J.A., 1995). They lack the otherwise universal bacterial cell

wall (Fuerst J.A. 1995, Konig, E. et al., 1984, Liesack W. et al., 1986) polymer

peptidoglycan, they divide by budding, and they have a differentiated cytoplasm, with

different membrane-bounded compartments apparently allocated to different cellular

functions. They are separated from other bacteria and amongst themselves by large

evolutionary distances. Recently, it appeared that planctomycetes might be the most ancient

group of bacteria, at the very root of the bacterial tree of life. Other species of

planctomycetes are aerobic chemoorganoheterotrophs, very different from the anammox

bacteria (which are anaerobic chemolithoautotrophs) (Fuerst J.A., 1995, Liesack W. et al.,

1986, Stackebrandt, E. et al., 1986)

1.3.2 Anammox cell structure

The bacterium has two main parts, inner and outer space separated by cell inner

membrane. There are nucleoid, cytoplasm, ribosome and anammoxosome enclosed by

bilayer in inner space, versus paryphoplasm, cell wall in outer space in anammox bacterium.

Anammox bacterium is a coccoid bacterium with diameter of less than 1 ptm. They are

physiologically different from the other known Planctomycetes and are anaerobic

chemolithoautotrophs (Strous M. et al., 1998). In anammox bacterium (Fig. 1-5), catabolism

),ljg'?. .

ssxf'l,k,.tX 'i' .

\$tKk•x-

...kl}.S..'gva'bet.v"rv"k•hrttl'llliibu••'es.".',•$ti•:.tfidi,,,

.•.x-'Svsff W•• ,'S;t'"ef ;' x ..I}}t? .x ...,,iiig,e•Z,,.,f.//,r,lllllli•ll/l//ii.,1///111ilii•lilillilili/ILI•,lilll!lillliilfil•illli.iiii.•i,l

',•,"','/.ut.tF-k".."svai,,'),;y.' ""'tt"':' in"i"• •f'g•n,E. {'+': ••.•.

.t tlt"trf

.tsts- '-'X'.

.I".;tl...:S'.ti.{.I

paryphoplasm z e=,pt,tt,fea,V-es

. .,pt-"•#s' cellnucleoid

{} hmaT,, ,' anammoxosome "k.."'4""//li.l,ii.l.ii/f/..-E....••••tse'e{$ts" intracytopiasm1

•N. lt, ',,, ',Sl. i .

J;'t•E-:,i'•,/z'ir'

i/i/I'•l',l,lll.l,ll:11,,/r:11'111i/11i'f////:///.Xikna•"iXeilii"i•

"•\i"'.:•Pil{•S..l':':,t.:.i,1/,,"..W•pt

"•{•-;t•rs•//1/,/1'k"'tww'g'.tr[/111SIii/..

t "t t.t "kufw w i."'/

ecr: yew.,.:twt•wapssc1,.:••;pt.A-.:f.r.rS.+

f.4I..-•-f-

ltt

lIx:

Ex'

' 't' ; tttk\t iv"n'w.z,vm / ''.s' k'v' u'•" ftsik•tii'w' '-" bilayer

ftf /k"sili'IL...g•.."l.k.si,rv. .gts"`S" rf

' ""+"t

.sS

t.

)e"tftt,tttstt.wnttxmss,mmtwe-

Fig. 1--5 Cell structure ofanammox bacterium

4

mner space

cell membrane

outer space

cell wall

•l

es

Fig. 1-6 TEM image of Candidatus

"Brocadia anammoxidans" (photo

John FuerstlRick Webb)

takes place in the compartments which are separated by a single bilayer membrane and is

called the anammoxosome. The cytoplasm in anammox bacteria is divided into three

compartments by single bilayer membranes: (1) the outer region is the paryphoplasm which

is bounded on the outer side by the cytoplasmic membrane and cell wall and on the inner

side by the intracytoplasmic membrane, (2) the riboplasm, containing the nucleoid and (3)

the inner ribosome-free compartment is the anammoxosome bounding by the

anammoxosome membrane (Van Niftrik L.A. et al., 2004). Scheme of anammox cell

structure is shown in Fig. 1-5. Fig. 1-6 indicates the TEM image of Candidatus "Brocadia

anammoxidans".

1.3.3 Anammox lipids

The anammoxosome is surrounded by a bilayer membrane that consists of unique and

bizarre lipids. The anammoxosome lipids contain 'ladderane' moieties, rigid and dense

ladders of concatenated cyclobutane rings. Both experimental evidence and molecular

modeling have shown that a membrane with ladderanes in its core is extremely impermeable

towards passive diffusion ofchemicals. From an evolutionary perspective, it is interesting to

note that part ofthe ladderane tails are ether-linked to the glycerol backbone; up to now, only

a minority of bacteria (either thermophiles or sulfate reducers) have ether linked to lipids.

Ladderane lipids will be used as biomarkers for past and present anammox activity in natural

ecosystems. Ladderane molecules with unique feature are now considered to be used as

opto-electronics. The anammox process is the only known natural source so far, and

,,i,..,,(i'

llg'

w

-]

t

t v y

1

k)

1

mo t,tS's..,,v,"tJVX,t,vvt--Lx-,,..y"za- .s.it.X C)r Y'

#gi"

"N.tf"OV-..""th'X..."S"-uwS-'X....t"""h.x

LL.o.f'N.......v"Kx".N'fNx."x"'x...x'X CY Y

!i

,'

" lli!!.-11N

Fig. 1-7 Example ofa chemical structure ofa natural ladderane (ladderane Y) from anammox cells. (above,

left) , place of ladderanes X and Y in anammox lipids electron microscopic image oftwo isolated(above,

middle) and anammoxosomes next to a partially lysed anammox cell (above,right)

5

ladderanes are very difficult to produce synthetically. Anammox bacteria contain a variety of

unconventional membrane lipids (Schmid M. et al., 2003, Kuypers M.M.M. et al., 2003,

Damste J.S.S. et al., 2003). Like all other Planctomycetes, anammox bacteria lack

peptidoglycan and have a proteinaceous cell wall instead ( Damste J.S.S. et al., 2002, Liesack

W. et al., 1986, Stackebrandt, E. et al., 1986). The lipids contain one, two or both of two

different ring-systems (X and Y) as shown in Fig. 1-7(Van Niftrik L.A. et al., 2004, Jetten

M.S.M. et al., 2003). Lipids from anammox bacteria are characterized by substantial}y lower

i3 C content than their carbon source(Schmid M. et al., 2003, Schouten S. et al., 2004). The

structure of the ladderane membrane lipids is unique in nature (Van Niftrik L.A. et al., 2004,

Damste J.S.S. et al., 2002) that can use as biomarkers for the presence of anammox cells in

environment (Schmid M. et al., 2003, Kuypers M.M.M. et al., 2003).

1.3.4 Anammoxbiodiversity

Currently, three genera of anammox bacteria have been discovered: Brocadia, Kuenenia

and Scalindua. The first two anammox bacteria ha' ve been found in wastewater treatment

systems. The latter, Scalindua, has also been detected in many marine ecosystems, such as

the Black Sea. The three genera share a common ancestor, but are evolutionally quite far

apart (less than 85 O/o sequence similarity on the 16S level). Still, all anammox bacteria seem

to be very similar phenotypically: they all grow at very slow rate (doubling time of about

1ldays), they all have an anammoxosome and ladderane lipids. The differences that must

exist among the three genera are the topic of ongoing research.

1.3.5 Biochemicalmechanism

The anammox pathway was determined by using i5N-labelling experiments, which

showed that hydrazine was an important intermediate (Jetten M.S.M. et al., 2003, Van de

GraafA.A. et al., 1997). The occurrence of free hydrazine in microbial nitrogen metabolism

was quite unique. The hydrazine oxidoreductase was purified from B. anammoxidans (Schalk

6

J., De Vries S. et al., 2000). Anammox catabolism takes place inside the anammoxosome.

The proposed biochemical mechanism of anammox reaction is shown in Fig. 1-8. (Schalk J.,

De Vries S. et al., 2000). In this mechanism, ammonium combined with hydroxylamine to

form hydrazine by hydrazine hydrolase (HH), a hydrazine-forming enzyme. Subsequently,

hydrazine is oxidized by a hydrazine--oxidizing enzyme (HZO), a key enzyme. HZO was

similar to hydroxylamine oxidoreductase (HAO) of Nitrosomonas europaea (Watson S.W. et

al., 1991). The oxidation of hydrazine produces dinitrogen gas, four protons and four

electrons. Consequently, these four electrons are combined with five protons from the

cytoplasm to reduce nitrite to hydroxylamine by a nitrite-reducing enzyme (NIR) (Kuenen J.

G. et al., 2001) as shown in Fig. 1-8.

cyteplasm - NH3

1N2}k"

NH20H

"Ng 4me

xxNe2

5H"

HN02 + H20 +N.K)" ptHNe3 +N.XDH,

Fig. 1-8 Biochemical mechanisms ofthe anammox process.

HH: hydrazine hydrolase, the hydrazine-forming enzyme;

HZO: hydrazine-oxidizing enzyme; NR: nitrite-reducing enzyme.

Anammox culture cell-free extracts showed a strong absorption at 468 nm in reduced

cytochrome spectra. The enzyme appeared to have some similarity to HAO ofNitrosomonas

europaea. HAO of Nitrosomonas europaea has a similar peak at 460 nm. Both enzymes

could oxidize both hydroxylamine and hydrazine (Schalk J. et al., 2000).

1.4 Traditional biological nitrification/denitrification process

Nitrogen elimination is commonly regarded as the conversion of biologically available

7

nitrogen compounds such as ammonium

(NH1), nitrite (NOi), or nitrate (NOi),

to dinitrogen gas (N2), which is released

into the atmosphere as a harmless

product (Fig.1-9). Today, wastewater

treatment plants commonly adopt a

biological nitrificationldenitrification

process for the purpose of nitrogen

removal from wastewater. Nitrification

means aerobic oxidation of ammonium

by nitrifying bacteria. Ammonium (NH;

nitrite (NO5) to nitrate (NOi).

. 20, (1 000/,) -

,,.,,,:••, , elgl{.paio

ti"i,in.x,..k,,:,.1///klj/i.i.l,,(e.gig•/

.{.l{-"f•*kg;•'

tt.:}}'.{tttti',t/•1' ..,+•,,,v,/•

ttt tt lttttuttttttt tt tt t

NMIicitiiim l,'

.•;-S.i •i;iti'i• NH;

2 e, (roo%}

k/,Eill/11Ii,lll,I/g•e•s,S••i••,i:•ii',r/'i•l,•,g•,l••g'scK•\••e•'T"r'5i"tw""•va•g•tt"`"ci;iii'ilva,i,,tll.i/s,,x•i.i'i:/l/11/i"i'I'iillllllii.,i}••,/it;.,,,.,.•.gN•,II,lll,i,,.,/I/,,.,,,,,•,,,,..•,,•t,eg.ili'1111•lg,;1/L/g.iigli:/:;

. . ."..... .ut +... t tt t ttt z';•••,': '•1/l•//,:?get•iv'//:ndgtiwh}tt.il"l'lil',i"//g'k•-"'i"':'''

getTe•i'X/',tE•(,:1'i'•"-•i)•s••"il•,".,/./)•••+•••-./l.,/3S'i/i.••. •it'sc-"

ill/lliPenitrifi,,,,,'I,,..l.,,,,g'I'g'S'iE•k"e,,y,.:"'/e'•,g,t//tts•,/k/,11/s,11ii•f•i,/,gi,/L/.ge..pt..i,?i'TS,g;i}l//;.,I-//,,,,./J//,r/:,/l•li•/E,//,,L,/,e.i•i-•/k,/•,//,••ee'•,,L:;•t,`iJif•y,Eii•i.

Fig. 1-9 Nitrificationldenitrification process

) is oxidized by oxygen (02) via the intermediate

Nitrification and denitrification reactions are as follows:

Organic - Carbon(e.g., methanol3.4kg l kgN)NH, ->NO, (1-1) -> O.5N2 Nitnfication Denitripcation

NH; +1.5 02 - NOi +H20+2H' (1-2)

N05 +02 - NOi (1-3)

Denitrification means nitrate respiration by denitrifying bacteria. Nitrate is reduced

under anaerobic conditions to dinitrogen gas (N2) with the addition of external organic

carbon source such as methanol.

NOi+OrganicCarbon . N2+2C02 (1-4)

1.5 Anammox process

Ammonium is oxidized to dinitrogen gas by nitrite under anaerobic condition through

anammox reactlon.

8

3•4kg.f}cag}.•il,,,.....,.,,{/f//E.i,,l,,//,g-/)ttew••xxii'III,,

. ... -t .. tt

s. '1ge.i//':

e.5 M2

+

s•f,di,.ee,i•"t2,•,gj,/L/1•i•l•l/

'"1'2P,i•,.{•.},••,.

,/t;•iil,S•kI.•••i

•..,tYl'l.Å}l':11•

NHI +NOi-N2+H20 (1-5) Anammox process (Fig.1-2) is a biological autotrophic process to removal ammonium

from wastewater. Ammonium is converted to dinitrogen gas with nitrite as electron accepter

under anaerobic conditions and without adding external organic carbon. The anammox

bacteria form a monophyletic cluster branching off deep in the order Planctomhycetales

(Fuerst J.A. 1995, Stackebrandt, E. et al., 1986), and were previously discovered in

wastewater treatment plants systems. Based on mass balances over anammox enrichnent

cultures, the anammox stoichiometry was estimated as following (Strous M., Heijnen J.J. et

al., 1998):

NH2 +1.32NOE +O.066 HCOi +O.13 H' .

1.02 N2+O.26 NOi +O.066 CH20o.sNo.s+2.03 H20 (1-6)

1.6 Partial nitrifications/anammox process

Two kinds of reactor are 'Pardal nibitadonrequired for ammoniumig?c?;oZipaXfa'7.g,,,,,.1",X.Rl,ll&i /NKZ x/?4g5%S

(about 50 O/o of influent

ammonium is converted to

nitrite) treatment is the

anammox process. For the

removal of nitrogen by nitrification-denitrification process, influent total organic carbon to

nitrogen ratio (CfN) is an important parameter to take into account for getting high nitrogen

removal efficiencies. When the available amount of carbon source is not enough to complete

the denitrification reaction, the addition of external organic carbon source is required. PN

process can reduce the oxygen requirements by 60 O/o compared with nitirification (Fig.

9

1-1O).

Nitrification is a two step biological processes mediated by two kinds of nitrifying

bacteria. First step is the conversion of ammonium to nitrite by ammonium oxidizing

bacteria (AOB) and the second step is the conversion of nitrite to nitrate by nitrite oxidizing

bacteria (NOB). Owing to the higher reaction rate of NOB compared with that of AOB,

accumulation of nitrite during nitrification reaction is quite rare. Nitrite is a substrate for

anammox reaction, so that inhibition ofNOB is essential for partial nitritation process. There

are different approaches for restricting NOB activity as follows.

1. Addition of specific inhibitors for NOB (Schmid M. et al., 2003). This option is not

easy to implement due to the difficulty to find a compound able to inhibit only the second

step of the nitrification. Besides the microorganisms can adapt to the inhibitor after long

periods of application, decrease its inhibitory efficiency must be considered. (Kuypers

M.MM, 2003).

2. Control of the dissolved oxygen concentration at low values to avoid the oxidation of

nitrite to nitrate due to the higher affinity for oxygen ofAOB than NOB ( Damste J.S.S. et al.,

2004).

3. The use of pure cultures of ammonia oxidizing bacteria is proposed. This method is

not usefu1 in practice, because the wastewater itself contains different kinds of

microorganisms, which quickly grow and contaminate the pure culture in the reactor.

4. To control the pH value and the ammonia concentration in the reactor for inhibition

of NOB according to the findings of Anthonisen et al., (Van Niftrik L.A., 2004). In this

approach, the adaptation of NOB to these controlled operational conditions will become a

problem. ' 5. Selection of nitrifying population based on the different growth rates of AOB and

NOB. This is the concept applied in the SHARON process. This process is carried out in a

chemostat operated ataHRT (equal to the SRT) of1day and 30 Åé which is favorable for

the growth of AOB. NOB are finally washed out from the reactor under these operational

condition. (Liesack W. et al., 1986).

PN lanammox process (Fig. 1-3) is shown as follows.

10

NH2 p..Oti'7.5192iit,OiOtX.')ti.. -'O'5NO5+O'5NH4' O'i5.0.in(ÅíO.'.YO) N'O'5N2 (1-6)

'

Compared to two step process ((1-1) & (1-6)), the PNIanammox process has several

'advantages over traditional nitrificationldenitrification process. Oxygen supply can be

reduced by 60 O/o. This means the reduction in the energy requirement for aeration. Also,

autotrophic anammox bacteria do not require the addition of external carbon source. The cell

yield of autotrophic anammox bacteria is quite low, so that the treatment cost for excess

sludge can be minimized. The new PN!anammox process uses not only less energy and fewer

resources, but it is also less expensive than conventional nitrificationldenitrification.

1.7 SHARON and anammox process

SHARON is an abbreviation of

Single reactor system for Uigh

Ammonium Removal Over Nitrite.

SHARONIanammox system iscomposed of two reactors. In the original

SHARON reactor, ammonium iscompletely transformed to nitrite (Fig.

1-11). Nitrite oxidation by NOB can be

prevented and heterotrophic denitrification

of nitrite using methanol as external

' 'carbon source occur under high 'temperature (35 O

of the carbon demand can be saved

.pt El 3

Hc•eS

S\es-k-asu'eZ'

c,e1,

ll

LXT•e2

S•L,WMGE

e es.

-p 't

.e2" g

-

"t

e

- tsns

" s- ev --

" 1te ,-

,-ms l

C) without sludge retention. Twenty five percent of the oxygen

compared

process. However, an external organic carbon for denitrificaton, i.e, methanol

'effective aeration system are still required (Stackeb

'from SHARON process and original ammonium containing wastewater mixed together by

ratio of 1 : 1 and is fed to the following anammox reactor.

11

Fig. 1-11 The SHARON process in a well

mixed continuous flow reactor

and 40 O/o

with complete nitrificationldenitrification

as well as an ' randt et al.,1986). Full nitrified effluent

In 2002, pilot scale experiment using PNIanammox process, which is shown in Fig.1-12,

for ammonium removal of sludge digester supernatant was carried out. Fifty eight percent of

the ammonium in the supernatant was converted to nitrite and nitrite production was O.35

kg-N m-3 d-i at temperature of 300C during nitritation process. The anammox process was

canied out in a SBR with a nitrogen removal rate of2.4 kg-N m'3 d'i. Over 90 O/o ofthe inlet

nitrogen load to the anammox reactor was removed and the sludge production was negligible

(Van Nifuik L.A. et al., 2004).

Elemonta}"itregen <Nal

dgedigestereMuentMixtureof({ammoniurn-rich)ammontumartdnitriteNH;

Hcoi+methanoiITIg'O.ooQ;oNeq'NH;••

e" fiN;su1{;egieNte`y"io

'1'/'t''/i"''''1'1/'''

tt tttt

Aeration {02>

"We<b.kgts}l.{igg6pte

, ' A

Treated vvater to the prs-settIement tank(low Åëoncant'ratiens of

NHS, N05, NOi}

d'i'),

ww"NC}i , eq ' t tttt tt blgeÅíxxypaggE ,,..,

Fig. 1-12 Reactor configuration for SHARON(left) and anammox oxidation (right).

1.8 OLAND and anammox process

OLAND is an abbreviation of 9xygen Limited Ammonium removal via

Nitrification-Denitrification.

OLAND and anammox reactions were canied out in one reactor. The present lab-scale

research revealed the potential of implementation of OLAND system with normal nitrifying

sludge as the biocatalyst for the nitrogen removal from nitrogen-rich wastewater in one step.

The current treatment capacity of the OLAND system is still low. However, the fact that the

inoculum can readily be grown in large quantities is an important factor which favors the

applicability of the OLAND system for practical purposes. Indeed, the seed nitrifying sludge

can easily be prepared from an activated sludge, and it can be applied directly to OLAND

system without adaptation. Moreover, operation of the OLAND system has no requirement

12

for an NO5 supply. An NH2 rich wastewater can be fed directly at a suitable Ioading rate.

Although the process requires limited dissolved oxygen concentrations, it does not require

strictly anaerobic conditions. Anammox reaction occurrs inside ofbiofilm, which ammonium

and nitrite exist under anaerobic condition. Therefore, inhibition by trace 02 exposure is not

a serious problem of concern in practice. The process operated by a pH controller is simple

and reliable for practical operation (Jetten M.S.M. et al., 2003).

tt '

1.9 CANON process

CANON is an abbreviation ofCompletely Autotrophic Nitrogen removal-Qver Nitrite.

In the CANON process, aerobic and anaefobic ammonia oxidizing bacteria cooperate to

remove ammonia in one oxygen-limited reactor ( Schmid M. et al., 2003). Recent studies

(Schouten S. et al., 2004) showed that the CANON biomass is very resilient against

disturbances in wastewater composition. CANON process becomes a good alternative to

existing nitrificationldenitrification systems for treatment of liquid waste rich in ammonia by

proper control of cultivation condition.

In CANON process, both partial nitrification and anammox reactions occur in one

reactor systems, where AOB and anammox bacteria coexist under oxygen-limited condition.

In this option, limited amounts of oxygen were introduced carefully into CANON reactor.

AOB consume all the dissolved oxygen, thus maintaining a very low dissolved oxygen

concentration in the aggregates of anammox sludge and produce nitrite as the electron

acceptor for anammox bacteria (Kuypers M.M.M. et al., 2003, Van Niftrik L.A. et al., 2004,

Damste J.S.S. et al., 2002). The CANON biomass was analyzed by FISH method and

showed that about 40 O/o ofbacterial community ofCANON process consisted ofAOB, and

about 40 O/e of bacterial community consisted of anammox bacteria. Nitrite oxidizing

bacteria (NOB) such as Nitrospira or Nitrobacter species were not detected in the bacterial

communlty.

Compared to SHARON-anammox reactor, CANON process uses one reactor, whereas

SHARON-anammox needs two reactors. In addition, the TN conversion of

13

SHARON-anammox is limited by the maximal strength of the wastewater being treated.

Compared to nitrificationldenitrification, the N-removal rate of CANON is lower (Damste

J.S.S. et aL, 2002).

1.10 SNAP process

SNAP (Single-stage pitrogen removal using anammox and partial nitritation) process is

a newly developed ammonium removal process using PN and anammox reactions. SNAP

process was successfuliy applied to the ammonium removal of synthetic landfi11 leachate.

During operation of partial nitritation, anammox bacteria grew inside of attached sludge on

the net-type acry1-fiber biomass carrier and enabled the short-cut conversion of ammonium

to dinitrogen gas in only one SNAP reactor. Under volumetric ammonium loading rate ofO.6

kg-N m-3 d"i, the SNAP process achieved nearly 90 O/o of ammonium conversion and about

80 O/o of TN nitrogen removal efficiency. Under higher volumetric loading rate of 1.0 kg-N

m-3 dLi,about 80 O/o of TN removal efficiency was obtained. The SNAP biomass was

demonstrated to be composed of AOB (close relatives of Nitrosomonas europaea), NOB

(close relatives ofNitrospira sp.) and anammox bacteria (close relatives of KU2 and KSU-1

'strains), in which AOB and anammox bacteria were dominant. These results showed the high

applicability of SNAP process to the ammonium removal oflandfi11 lechate.

1.11 Researeh work

A lot of research works were carried out for the development and applying for

anammox technology more than 15 years. Most of the researchers focused on basic studies

and got lots of valuable new information. I had started my research work on anammox about

three years ago after my enrolling Ph.D. study. My main work focusing on anammox

research is how to get a stable high-rate anammox treatment system under poor operational

conditions. With the help of Prof. Furukawa, I was able to obtain various excellent

achievements on anammox treatment process.

14

1.11.1 Problem statement

There are several problems on anammox application as follows:

1) In case of starting and operating an anammox system, the total nitrogen removal rate

goes through a course of increasing firstly and decreasing afterwards.

2) The anammox sludge did not fu1fi11 in reactor, but the sludge volume began to

reduce and anamniox activity was devitalization with sufficient nutrients influent.

3) The total nitrogen loading rate can not increase easily for low-strength

ammomum-contamlng at room temperature. . . 4) How to develop a high-rate anammox treatment system under high salinity

concentration comparable to sea water?

5) How to adapt the freshwater anammox sludge to salt concentration of 30 g 1-i for

short acclimation time?

6) How to solve the problem of system starting, defeat and restarting all the times.

All above mentioned problems puzzled me for about one year. I was able to overcome

those difficult problems with the help of Prof. Furukawa.

1.11.2 Objectives of this study

The objectives ofthis study are mentioned as follows:

1) To evaluate the nitrogen removal capabilities of anammox sludge in a fixed-bed

reactor using various biomass carriers, nonwoven, split nonwoven, splitted nonwoven and

nonwoven powders as caniers.

2) To develop an anammox process using hydrates material ofcement and harecore-1 as

biomass carrier and to compare the nitrogen removal capabilities with anammox process

usmg nonwoven. ' 3) To investigate the nitrogen removal capability of the anammox process using small

carriers (selfmade products) as biomass carrier.

4) To develop an anammox process under salinity concentration of30 g 1-i. First step is

15

to change freshwater type anammox sludge to salt resistance type anammox sludge.

4) To compare the nitrogen removal capabilities of nonwoven biomass caniers and self

made small carriers.

1.11.3 Research plan

This study focuses on the application of anammox process for low strength

ammonium-rich wastewater under low temperature and high salinity condition.

Part 1. Anammox treatment of low-strength ammonium-containing wastewater at room

temperature

In this study, fixed-bed anammox reactor was operated at room temperature for the

treatment of low-strength ammonium containing wastewater. The anammox reactor was

operated under N02-N limiting condition for the treatment of a relatively low-strength

ammonium-containing wastewater at 20-25 Åé.

Part 2. Effect of salt concentration in anammox treatment using non-woven biomass

carrler

This study focuses on the evaluation of anammox process using nonwoven biomass

carriers under high salt concentration. Effect of high salt concentration on the anammox

treatment was investigated for the establishent of acclimation strategy under high salt

concentration conditions. The bacterial community was examined by 16S rRNA gene

analysis and DGGE after the acclimation of the anammox sludge to high salt conditions.

Part 3. Nitrogen removal capabilities of two kinds of fixed-bed anammox reactors for

treating partial nitrified brine wastewater

This experiment was carried out to determine the nitrogen removal capacities of two

types ofbiomass caniers (nonwoven and selfmade small carriers). Two fixed-bed anammox

reactors using nonwoven and small size biomass carriers were operated respectively under

high salinity condition comparable to sea water level, and compared nitrogen removal

capacities ofthis two type biomass carriers.

Part 4. Study on long-term stable operation ofhigh-rate anammox biofilm reactor using

nonwoven carrier under moderately low temperature

' ' '

16

In this study, main task was to improve the activity of anammox sludge. Small biomass

technology for keeping plenty of anammox sludge in the reactor and chemical or physical

methods to enhance anammox activity by killing some part bacteria of anammox sludge was

applied. At last, mathematical methods to simulate and forecast the TNLR 5 or 7 days later

were developed.

1.12 References

Anthonisen AC, Loehr BC, Prakasam TBS, Srinath EG.: Inhibition of nitrification by

ammonia and nitrous acid, J. MPCF., 48(5), 835-852(1976). '

Broda E.: Two kinds of lithotrophs missing in nature, Z. Allg. Mikrobiol., 17 (6), 491493

(1977).

Damste J.S.S., Strous M., Rijpstra W,I.C., Hopmans E.C., Geenevasen JAJ., Van Duin

A.C.T., Van Niftrik L.A. and Jetten M.S.M.: Linearly concatenated cyclobutane lipids

form a dense bacterial membrane, Nature, 419, 708-7 12 (2002).

Fuerst J.A.: The planctomycetes: emerging models for microbial ecology, evolution and cell

' ' biology, Microbiology, 141, 1493-1506 (1995).

Fux C., Boehler M., Huber P,, Brutnner I., Siegrist H.R.: Biological treatment of

' ammonium-rich wastewater by partial nitritation and subsequent anaerobic ammonium

oxidation (anammox) in a pilot plant, 1. Biotechnol., 99, 295-306 (2002).

Ganido JM, Van Benthum WAJ, Van Loosdrecht MCM, Heijnen J. J.: Infiuence of dissolved

oxygen concentration on nitrite accumulation in a biofilm airlift suspension reactor,

Biotechnol Bioeng, 53,168-178 (1997)

Hellinga C., Schellen A.A.J.C., Mulder J.W., van Loosdrecht M.C.M., Heijnen J. J.: The

' Sharon process: an innovative method for nitrogen removal from ammonium rich

' wastewater, PVat. Sci. Tech., 37(9), 135--132 (1998).

Jetten M.S.M., Sliekers A.O., Kuypers M.M.M., Dalsgaard T., Van Niftrik L., Cirpus I.,

Van de Pas-Schoonen KT., Lavik G., Thamdrup B. et al.: Anaerobic ammonium

oxidation by marine and freshwater planctomycete-like bacteria, Appl. Microbiol.

17

Biotechnol., 63, 107-114 (2003).

Jetten M.S.M., Sliekers A.O., Kuypers M.M.M., Dalsgaard T., Van Niftrik L., Cirpus I.,

Van de Pas-Schoonen K.T., Lavik G., Thamdrup B. et al.: Anaerobic ammonium

oxidation by marine and freshwater planctomycete-like bacteria, Appl. Microbiol.

Biotechnol., 63, 107--114 (2003).

Jetten MSM et al.: The anaerobic oxidation of ammonium, FEMS Microhiol. Rev , 22,

421-437, (1999)

Jetten MSM et al., Microbiology and application of the anaerobic ammonium oxidation

('anammox') process. Curr. Opin.Biotechnol. 12, 283-288, (2001)

Konig, E., Schlesner, H. and Hirsch, P.: Cell wall studies on budding bacteria of the

Planctomyces/Pasteuria group and on a Prosthecomicrobium sp, Arch. Microbiol., 138,

200-205 (1984).

Kuenen J. G., Jetten M. S. M. : Extraordinary anaerobic ammonium-oxidizing bacteria, ASM

News, 67,456-463 (2001)

Kuypers M.M.M., Sliekers A.O., Lavik G., Schmid M., Jorgensen B.B., Kuenen J.G.,

Damster J.S.S., Strous M., Jetten M.S.M.: Anaerobic ammonium oxidation by

Anamniox bacteria in the Black Sea, Nature, 422, 608-61 1 (2003).

Liesack W., Konig H., Schlesner H. and Hirsch P.: Chemical composition of the

peptidoglycan-free cell envelopes of budding bacteria of the Pirella/Planctomyces

group, Arch Microbial., 145, 361-366 (1986).

Lieu P.K., Hatozaki R., Homan H. and Furukawa K.: Single-stage nitrogen removal using

anammox and partial nitritation (SNAP) for treatment of synthetic landfi11 leachate,

Japan. J. Mater Treat. Biol., 41 (2), 103-112 (2005).

Lieu P.K.: Nitrogen removal from landfi11 leachate using a single-stage process combining

anammox and partial nitritation, Ph. D. Dissertation, Kumamoto University, Japan

(2006).

Linping Kuai and Willy verstraete: Laboratory of Microbial Ecology, Department of

Biochemistry and Microbial Technology, State University of Ghent, 9000 Ghent,

Belgium, Ammonium Removal by the Oxygen-Limited Autotrophic

Nitrification-Denitrification System, Appl. andEnviron. Microbial., 4500-4506(1998)

18

Michael Nielsen , Annette Bollmarm , Olav Sliekers , Mike Jetten , Markus Schmid , Marc

Strous , Ingo Schmidt , Lars Hauer Larsen d, Lars Peter Nielsen , Niels Peter Revsbech:

Kinetics, diffusional limitation and microscale distribution of chemistry and organisms

in a CANON reactor, FEMS Microbiology Ecology, 51, 247-256 (2005)

Santos VA, Tramper J, Wijffels RH.: Integrated nitrification-denitrification with immobilized

microorganism. In: Biofilms-science and technology. Dordrecht, The Netherlands.:

Kluwer Academic Publishers, 449-54(1992).

Schalk J., De Vries S., Kuenen J.G. and Jetten M.S.M.: Involvement of a novel

hydroxylamine oxidoreductase in anaerobic ammonium oxidation, Biochemistty, 39,

5405-5412 (2000).

Schlesner H. and Stackebrandt E.: Assignment of the genera Planctomyces and Pirella to a

new family Planctomycetaceae fam. nov. and description of the order Planctomycetales

ord. nov, Syst. Appl. Microbiol., 8, 174-176 (1986).

Schmid M., Walsh K., Webb R.I., Rijpstra W.I.C., Van de Pas-Schoonen K.T., Verbruggen

M.J., Hill T., Moffert B., Fuerst J.A. et aL: Candidatus "Scalindua brodae", sp. Nov.,

Candidatus "Scalindua wagneri", sp. Nov., Two New Species ofAnaerobic Ammonium

Oxidizing Bacteria, Syst. Appl. Microbiol., 26, 529-538. (2003).

Schnidt I., Sliekers O., Schmid M., Cirpus I., Strous M., Bock E., Kuenen J.G., Jetten

M.S.M.: Aerobic and anaerobic ammonia oxidizing bacteria - competitors or natural

partners?, FEMS Microbiol. Ecol., 39, 175-180 (2002).

Schouten S., Strous M., Kuypers M.M.M., Rijpstra W.I.C., Baas M., Schubert C.J., Jetten

M.S.M. and Sinninghe Damste J.S.: Stable carbon isotopic fractionations associated

with inorganic carbon fixation by anaerobic ammonium-oxidizing bacteria, Appl.

Environ. Microbiol., 70, 3785-3788 (2004).

Sedlak R.: Phosphorous and nitrogen removal from municipal wastewater: principles and

practice, Metcalf& Eddy, Inc., McGraw-Hill, New York (1991)

Sliekers A.O., Third K., Abma W., Kuenen J.G. and Jetten M.S.M.: CANON and Anammox

in a gas-lift reactor, FEMS Microbiol. Lett., 218, 339-344 (2003).

Stackebrandt, E., Wehmeyer, U. and Liesack, W.: 16S ribosomal RNA- and cell wall

analysis of Gemmata obscuriglobus, a new member of the order Planctomycetales,

19

FEMS Microbiol. Lett., 37, 289-292 (1986).

Strous M, Van Gerven E, Kuenen J.G & Jetten M.: Effects of aerobic and microaerobic

conditions on anaerobic ammonium-oxidizing (Anammox) sludge, Appl. Environ.

Microbiol., 63, 2446-2448(1997).

Strous M, Van Gerven E, Ping Z, Kuenen J.G & Jetten MSM.: Ammonium removal from

concentrated waste streams with the Anaerobic Ammonium Oxidation (Anammox)

process in different reactor configurations, Mat. Res., 31,1955-1962(1997).

Strous M., Heijnen J.J., Kuenen J.G. and Jetten M.S.M.: The sequencing batch reactor as a

powerfu1 tool for the study of slowly growing anaerobic ammonium-oxidizing

microorganisms, Appl. Microbiol. Biotechnol., 50, 589-596 (1998).

Strous M., Kuenen J.G. and Jetten M.S.M.: Key physiology of anaerobic ammonium

oxidation, Appl. Enviion. Microbiol. , 65, 3248-3250(1999)

Thamdrup B & Dalsgaard T.: Production of N2 through Anaerobic Ammonium Oxidation

Coupled to Nitrate Reduction in Marine Sediments, AppL Environ. Microbiol.,

68,1312-1318(2002).

Thamdrup B. and Dalsgaard T.: The fate of ammonium in anoxic manganese oxide-rich

marine sediment, Geochim. Cosmochim. Acta, 64, 4157-4164 (2000)

Third K.A., Sliekers O., Kuenen J.G., Jetten M.S.M.: The CANON System (Completely

Autotrophic Nitrogen-removal Over Nitrite) under Ammonium Limitation: Interaction

and Competition between Three Groups ofBacteria, Syst. AppL Microbiol., 24, 588-596

(2001).

Third, K.A., Sliekers, A.O., Kuenen, J.G and Jetten, M.S.M. The CANON system

(competlely autotrophic nitrogenremoval over nitrite) under ammonium limitation:

interaction and competition between three groups of bacteria. System. Appl.

Microbiol, 24, 588-596(2001).

Turk O, Mavinic D.S.: Stability of nitrite build-up in an activated sludge system. J. WIPCF,

61(8), 1440-1448(1989).

Van de Graaf A.A., De Bruijn P., Robertson L.A., Jetten, M.S.M. and Kuenen, J.G.:

Metabolic pathway of anaerobic ammonium oxidation on the basis ofN-15 studies in a

fluidized bed reactor, Microbiology UK, 143, 2415-2421(1997).

20

van Dongen U, Jetten MSM, Van Loosdrecht M.C.M.: The SHARON ANAMMOX process

for treatment ofammonium rich wastewater, Mater Sci. Technol., 44(1):153-60(2001).

Van Niftrik LA., Fuerst J.A., Damst J.S.S., Kuenen J.G., Jetten M.S.M., Strous M.: The

anammoxosome: an intracytoplasmic compartment in anammox bacteria, FEMS

Microbiology Lett., 233, 7-13 (2004).

Van Niftrik L.A., Fuerst J.A., Damst J.S.S., Kuenen J.G., Jetten M.S.M., Strous M.: The

anammoxosome: an intracytoplasmic compartment in anammox bacteria, FEMS

Microbiology Lett., 233, 7-13 (2004).

Watson S.W.,Valos F.W. and Waterbury J.B.: The family nitrobacteraceae., In The

prokaryotes, Lewis Publishers, New York (1991).

21

22

Chapter 2

Anammox Treatment ofLow-strength Ammonium-Containing Wastewater at Room Temperature

2.1 Introduction

Anaerobic ammonium oxidation (Anammox) is an anoxic microbiological process in

which ammonia, together with nitrite, is oxidized to gaseous nitrogen anaerobically (A. Olav

Sliekers et al., 2002). Anammox is commonly used for the treatment of high-strength

ammonia-containing wastewater at rather high temperature of30-40 Åé. Anammox process

is a new and promising alternative to the conventional nitrogen removal processes. The

application of anammox to nitrogen removal process in wastewater treatment would lead to a

significant reduction of costs for aeration and addition of exogenous electron donor as

compared to the conventional nitrification-denitrification process(van Dongen et al., 2001).

'

NH4' + 1.32 N02-+ O.13 H'+O.066 HC03- -i>

1.e2 N2+ O.26 N03- + 2.03H2Q + Q.66 BiQmass (Eq. 1)

' effluent

outlet

However, anammox process

requires a long start-up time due to

extremely slow growth rate of anammox

bacteria (van Dongen et al., 2001) (the

doubling time was reported to be about

11 days) (Strous et al., 1998). For the

prompt establishrnent of anammox

reactor, the selection of proper seed

sludge and carefu1 increasing protocol in

nitrogen loading rates become important.

temperaturecontroller nonwovens

carrier

entwin6d

heater

S,l,

g/l/l

l-

tpt/ ..

gl {--

influent tank peristaltic pump

Fig.2-1 Schematic diagram of experimental system

ts!t/.111v/'[vvt"':/x/--y'-'T'e .

1::,l•::•i•i,,liliEllEi!'iiii{l

e

[

fixed- edreactor

23

Also, the complete retention of slowly growing anammox bacteria in the reactor is important.

It was reported that it could successfu11y develop high-rate anammox biofilm reactors using

the up-flow fixed-bed biofilm column reactor by inoculating seeding sludge with high

'abundance of anammox bacteria for short periods, operate steadily and achieve a high

nitrogen removal rate of26.0 kg-N m"3 d-i within 250 days(Tsushima et al., 2007). But it was

operated at high temperature and high concentration of ammonium, and the experiment was

not under the control of stability. Up to now, anammox reaction was applied to the high

strength ammonium containing wastewaters such as digestion liquor and landfi11 leachate.

In order to apply this anammox reaction to the conventional nitrogen removal process,

the applicability of anammox process to low strength ammonium containing wastewaters at

room temperature must be evaluated. The objectives of this study were to examine the

start-up anammox reactor using low strength ammonium containing wastewater under room

temperature and make clear the anammox treatment potential of our newly invented up-flow

column reactor.

For these purposes, up-fiow fixed-bed biofilm column reactor with nonwoven fabric

sheets as biomass carrier was used for the cultivation of anammox bacteria and the reactor 'was operated for one and a halfyear.

22 MaterialsandMethods

2.2.1 Experimental setup

The reactor used in this experiment was shown in Fig.2-1. This reactor was made of

glass and the size is Q 95mm x 423mm, with the available volume of 2.8L. Influent was

supplied to the reactor by up-flow mode. Reactor was heated to about 20 OC in winter day

and kept at room temperature in other seasons. The reactor set was shaded from a light with

black vinyl sheet enclosures.

24

2.2.2 Seed sludge

Seed anammox sludge of 1.4g-VSS was obtained from another anammox reactor

cultivated in our laboratory. In order to improve the attachnent of anammox sludge on the

nonwoven biomass canier, granular of anammox sludge was crushed into small particles

using O.5 mm sieve.

2.2.3 Biomass carrier

The 8 strips of polyester nonwovens were outspread evenly by 16 strips appearing a

column liking inside of reactor. The purpose of using thin-nonwovens is to get more

superficial surface area where more biofilm can be attached on it. The nonwoven shape was

shown in Fig. 2-2. After anammox sludge developed fu11y in the reactor, undoing nonwoven

?ww•i/ttt#,/g.,///.i•

•ge

'

t••;l isÅ}:

Fig. 2--2 Thin strips ofnonwovens Fig. 2-3. Carrier ofnonwoven

fibers shown in Fig. 2-3 were added for bridging the gap between thin strips of nonwoven.

The small caniers are like sticks with length no more than 1 mm. The undoing nonwoven

fibers were prepared by smashing thin strips ofnonwoven with cutting method.

2.2.4 Attach immobilization of anammox sludge on biomass carrier

Thin strip of nonwoven carrier was set in the reactor. Three Liter of pretreated crushed

25

anammox sludge suspension was added from the bottom of reactor with flow rate of

1,OOO-2,OOO lh-i. Then, the reactor was purged with nitrogen gas for 20 min. At last, internal

circulation 10 1 h-i was carried out for 12 h. Through this procedure, seed anammox sludge

was evenly attach-immobilized on thin strip polyester nonwoven biomass carrier.

2.2.5 Synthetic influent

Influent wastewater was prepared by adding ammonium and nitrite in the forms of

(NH4)2S04 and NaN02, with other nutrients and buffer according to the composition given in

Table 2-1.

Table2-1 Compositionofsyntheticwastewater

Composition Concentration

(NH4)2S04 70Å}5(mg-N1-')

NaN02 70Å}5(mg-N1-i)f

EDTA 5.0(mg1-i)

FeS04.7H20 9.0(mg1-i)

KHC03 125.1(mgri)

KH2P04 54.4(mg1-')

2.2.6 Protocol for increasing in TN loading rate

Total nitrogen loading rate (TNLR) was increased stepwise by O.05 kg-N m-3 d-i - O.1

'kg-N m"3 d-i every step from O.l kg-N m"3 d-i to 1.2 kg-N m-3 dffi . After loading rate

reached to 1.2 kg-N m-3 d-i , TN loading rate was increased by O.2Å}O.05 kg-N m-3 d-i for

each step. Effluent nitrite concentrations were always kept below 20 mg-N 1-i.

2.2.7 Operational condition in whole period

Temperature was kept at 20Å}2 OC in winter season and at room temperature in other

26

seasons. Concentration of (NH4)2S04 and NaN02 in synthetic wastewater was 75Å}5 mg-N 1-i

and the total nitrogen concentration was 150Å}10 mg-N l-i. TNLR was increased only by

increasing flow rate.

2.2.8 Analytical items and methods

The concentrations ofN02-N and N03-N were measured by the colorimetric method in

accordance with the Standard Methods (APHA, AWWA, WPCF 1995). NH4-N concentration

was measured by the modified phenate method using ortho-phenyl phenol (OPP) (Kanda, J.,

1995). Absorbance and pH values were measured using a spectrophotometer (U-2010;

Hitachi, Tokyo, Japan) and a pH meter (B-21 1; Horiba, Kyoto, Japan), respectively.

2.3 Results and Discussion

2.3.1 Observation of the attached biomass

The changes in anammox biofilm shape and color were shown in Fig. 2-4. The biofilm

developed in thicker biomass and the color of anammox biofilm changed to deep red color

gradually. But after 13 months of operation, the color was not as red as before because of

addition ofundoing nonwoven fibers for bridging the gap between thin strips ofnonwoven in

the anommox reactor.

•• l•Sl,ii1.,,LGg

"'''i'I'}'//'

•k

"txxl'

beeeww.,l

ge

-wwiigeigi

Fig.2-4 Colorchangesinanammoxbiofilm

27

2.3.2 Changes in total nitrogen loading rates and removal rates

Fig. 2-5 showed daily changes in total nitrogen loading rate (TNLR), total nitrogen

removal rate (TNRR) and N03-N production rate. From this graph, whole experiments can

be divided into 5 Runs. The term of Run 1 was days O to 205. The maximum TNLR was 1.63

kg-N m-3 dri and the maximum total nitrogen removal rate (TNRR) of 1.3 kg-N m'3 d"i was

obtained after 205 days of operation. The establishment of anammox reactor using low

strength of influent nitrogen (influent NH4-N N02-N was about 70Å}5 mg 1-i) was proved

possible without temperature control from these results.

Ig .6

,""KTv" ezdn

MK...i

eN

'i6

gE2

Å~tu

sRRz-

.::...3

.3.,i,

l.S

igl -e

I II III IV v

o-A

TN loading ratet( kg-N m-3 d-i)

TN removal rate( kg-N m'3 di) @No3-N production( kg-N m-3 d-i) )L

@- l i ;,,- ,' 'lg'.,./:, ..

x'-" ,,k •-,/ .,./ ':l-ll-ll-!- i" L• i

.i ,-x k 1 ,r, ., r,r,n ' ,',, i' I,l" -lats

s•s•/ 'g: if-}

Fig. 2-5

t, eV

-

@Smallga,rr,ie.r,

•

i-i

"F

Y

'

e,{Fii ,i,i,,,,ii"i'

f

""

6

,E"

L'

'

5

i'.S ,

e l{ie :•gbex :4•rs 2gf,• :,;,tk seci• ajee 4s(: wwe spt.y'

Time(days) -Changes in TN loadinglremoval rate and N03-N production rate

g.,4

'3 .fs•

O.4

.'-'N

Tv9gzda

Mvpoe:

,9689nz

:"oZ

The term of Run 3 was days 235 to 370. We were able to increase TNLR to 2.6 kg-N

m-3 d-i on day 320, but TNRR fluctuated and could not get the stable anammox activities.

After day 365, the reactor system deteriorated quickly due to clogging and channeling inside

of the anammox reactor. This phenomenon was supposed to be caused by no intentional

sludge withdrawal. Thus, 6.5g-VSS of anammox sludge was withdrawn from reactor on day

370.

28

The term of Run 3 was days 235 to 370. 0n day 400, reactor temperature was increased

accidentally to 50 OC owing to failure in temperature control at 20 Åé in winter season.

The color of anammox sludge turned to white color and lost most of its anammox activity.

This dead anammox sludge was taken out from the reactor and restarted experiment under

low loading rates. Just at this time, we added small-caniers into the gaps of thin strips of

nonwoven.

The term of Run 4 was days 41O to 520. TNLR and TNRR were able to increase quickly.

TNLR increased from O.78 kg-N m-3 d-i to 6.5 kg-N m-3 d-i in about 3 months of operation.

Then, the reactor was fu11y occupied with anammox sludge as shown in Fig. 2-4. Maximum

TNLR and TNRR were 6.5 kg-N m-3 d-i and TNRR is 4.68 kg-N mH3 d-i , respectively in this

Run. On day 530, the reactor was stopped for the measurement of anammox sludge grown in

the reactor. Nonwoven biomass carrier was taken out from the reactor and detached the

attached sludge on nonwoven and suspended in 31 ofefiluent, and measured its concentration.

The weight of anammox biomass ofnonwoven small-carrier in the reactor was 40.7 g-MLSS

and 3 1 .2 g-MLVSS. The weight of anammox biomass attached on thin strips of nonwoven in

the reactor was 31.0 g-MLSS and 21.7 g-MLVSS. So, the total weight of anammox

biomass in the reactor was 71.7 g-MLSS and 52.9 g-MLVSS. VSS content of our

cultivated anammox sludge was 73.8 O/o. This VSS value was unexpectedly high compared

with another autotrophic sludge like nitrifying sludge.

Finally, the nonwoven canier was installed into the reactor again. Then, 3 1 of anammox

sludge was fi11ed in the anammox reactor and culture liquid was recycled at 10 1 h'i for

attachment of anammox sludge on nonwoven biomass carrier. After 32 hours of recirculation,

the effluent was almost clear. Then, continuous operation was restarted. Initial anammox

sludge concentration was 19.8 g-MLSS 1-i. Run 5 was started from day 520. Stable and quick

anammox start-up was also realized in Run 5 like Run 4.

Table 2-2 shows the maximum influent flow rate and 'minimum HRT for each Run.

Owing to the low influent nitrogen concentration, high influent flow rate had to be applied

for getting high TNLR. In Run 4, extremely short HRT (O.56 h) was applied. This is the first

report to succeed the stable operation of anammox reactor under such an extremely low HRT

under room temperature.

29

Table 2-2 Minimum HRT for each Run

Run 1 2 3 4 5

Maximuminfluentflowrate(1d

MinimumHRT(h)

-i ) 29.0

2.32

66.2

1.01

14.4

4.67

120

O.56

98

O.69

2.3.3 Performances ofTotal nitrogen removal efficiencies

TN removal performances during whole experiments were shown in Fig. 2-6. At the

starting periods up to 50 days, influent TN concentration was kept at the low level of 50 to

70 mg-N 1-i. From day 51 to day 530, influent TN concentration was kept at 150 mg-N 1-i but

the effluent total nitrogen concentrations were changed depending on the operational

conditions. Before day 340, the effluent total nitrogen concentrations were stable and ranged

from 20 mg l-i to 30 mg-N 1-i But the effluent total nitrogen concentrations were ranged from

30 mg INi to 40 mg-N 1-i after day 340. The reason of these high effluent TN concentrations

was supposed to be the formation of water channel under short HRT. The total nitrogen

removal efficiencies mainly ranged from 70 O/o to 80 O/o before day 340, while ranged mainly

from 60 O/o to 70 O/o after day 340.

g?,$.

'Oi

-aneu.9retsg8gzeu8EEII

a8

ÅëE

a2, 6

g'19

se2.

g,i

•5S

.32

34

'g'7,

it

"-

,gkllk s-.

}il es

g

s

ge'w

Influent TN concentration(mg 1-i)

Effluent TN concentration(mg 1-i)

TN removal efficiency(O/o)

TN removal rate kg-N mt3 d'i

1-i

c'

iitseci

t'

1

:`'' :

I.

"tli s.. :

Å}

s

vatt.

i-

t

pm

g,g)

'7, a)

tw

.)-,a)

:";'t

LTg

gij

/gft

AeXO

Å~8

•g

ur..o

-,ge

oE9zE---,

vg-A

ivveegRm

•# 4iC

Fig. 2- 6

at$ giipt,1 :7L.l o- :.s-e•.g llltg,e 32,tr s6,gj ec.ee `#E#if 4gg) u•,s -sec,

Time (days)Changes in influent and effluent TN concentrations and TN removal efficiencies

30

2.3.4 NH 4-N removal performances

Fig. 2-7 showed daily changes in influent and effluent NH4-N concentrations, NH4-N

removal efficiencies and NH4-N removal rates during whole experiments. NH4-N removal

efficiencies ranged from 75 O/o to 85 O/o and effiuent NH4-N concentrations were ranged from

10 mg 1-i to 25 mg 1-i before day 340. While NH4-N removal efficiency ranged from 60 O/e to

55 O/o and effluent NH4-N concentrations ranged from 20 mg 1-i to 30 mg 1-i after day 340 as

shown in Fig.2-7.

j-.-S

ATv"ez

tSi)

MveE

re>ofi

9zk

=z

3.2

:.-s

2.,S

i.s

g.4

g,x

g.'7,

,g..4

gl .e

,iil)

vaiiiiiidi

tw

ADee

gl}S

Infiuent NH4-N(mg 1-i)

NH4-N removal efficiency(O/o)Effiuent NH4-N(mg 1-i)

NH4-N removal rate( kg-N m-3 d'i)

"

e 4g] gij lg2,ga sgi) :.tw

Fig. 2-7.

k"gck 2-ffo s2g) r,6cf #tt Time(days)NH4-N removal perfomances

4gct k..g soo

pa

gij -A,

,, g-so

ljgfi" fi' ag

wtr-". ev'i#i .St:r•# g•E"

.. gg US',,'i

1ge

2.3.5 NO2-N removal performances

Fig. 2-8 showed daily changes in influent and effluent N02-N concentrations, N02-N

removal efficiencies and N02-N removal rates. N02-N removal efficiencies ranged from 80

to 95 O/o and effluent N02-N concentration were kept below 20 mg-N 1-i before day 340.

After day 340, N02-N removal efficiencies were averaged at 77 O/o and effluent N02-N

concentration were kept below 25 mg-N IMi. Compared with NH4-N removal, N02-N removal

performances were better than NH4-N removal performances in this study. We have operated

our anammox reactor under N02-N limiting condition for feeding equal concentration of

31

NH4-N and N02-N.

T"Ab

v" Ez

e{!)

MUts:ge

oE•9

z &oz

3.j

3.2

2.S

L,.,S

e.g,

E.4

g.a

fi•ij -

pts,ft,,C ,e-

=,I

.e." •-- ,c{

,r

f r'

A Influent N02-N(mg IJi)

O N02-N removal efficiency(O/o)A Effluent N02-N(mg 1-i)

wa N02-N removal rate( kg-N m-3 d-i

'

lc

"

,ll , wt

e-

r

`

r"li f=-

y,

'

L

"

kl,t1

'

"

. F:L

-i

-

-

,g

-

eff.-

va

it s$

(liix

ge .g2.e a•$i) 2•.ee 4-sie )-lx, )-ffe

lle

•ge"

sg

S5

;,s'

3Ll

'l S

Fig. 2-8

24e 2sg ,3Lr,e.4 ,3iec, 4oo tssig{,

Time(days)N02-N removal performances

,$

A-v'

vegbE

Z",5"

gu•tds

ggo

Ae.

oh

g'6.a-Et

ge

o8e

zk

ozvg

2.3.6 Removal ratio of NH4-N, N02-N and N03-N

K2o,

kÅé,

s•E

g•.pa

fi •g,

482thO"n

z

rl.-.-•,

i'("

g.--',

-c,

'i

.cfi

.5

.s

.5

.I}

'j

e.o

Amp

D

TN removal rate( kg-N m'3 d-i)

N02-N removal rate( kg-Nm'3 d "i)

N03-N production( kg-N m-3 d-i )

v= l,.'ge9x,x

,TRf = iO.99i2 "- "

v •== l.;•t g x

R2 = g.gT• r}

-

'

eedwee

rf e

, ,IE' " 'e• ""' '-' il' lf'/ f'

tat. ..t

pt

g.t ="/.Lr23x

,R' = 'i-kt' m'9'$"2'

fi..{p

Fig. 2-9

e-s ge g..i ?...5

Ammonium (NH4-N) removal rate ( kg-N m'j d'`)TN removal, N02-N removal and N03-N production rates with ' respect to NH4-N removal rate

and

Fig. 2-9 showed the relationship between NH4-N removal rates, N02-N removal rates

N03-N production rates. The ratios of TN removal, N02-N removal and N03-N

32

production rates to NH4-N removal rates in this study were shown in Fig. 2-9. The ratio was

1.98 : 1.20 : O.22, which was slightly lower value compared to the theoretical anammox

reaction ratio (Eq. 1)

2.3.7 Comparison of anammox treatment capabilities

We have operated our anammox reactor under poor operational conditions (low nutrient

concentrations, low operational temperature and short HRT), but good treatment

performances were obtained even under poor operational conditions. Our obtained results

were comparable with another results for same type of anammox reactors, which were all

operated under ideal operational conditions with high nutrient level, high operational

temperature(35 OC) and long HRT. Table 2-3 showed that comparison of the total nitrogen

removal capabilities for various kinds of anammox reactors operated in our laboratory. Even

under poor operational conditions, high TN removal rate was successfu11y obtained in this

study.

Table 2-3 Comparison of anammox removal capabilitiesTypeofbiomass Inf.TN Operational TNremovalrate

.carrler cocentration(mg-Nl`i) Temperature( oC)( g-Nm-3d-i)

Nonwoven 250-300 35 3.1

Acrylicresinfiber 90-120 35 2.2(Fujiietal.,2002)

Nonwoven 150 20 4.68thisstud

2.3.8 The reasons for high removal rate

In this study, high nitrogen removal rate was obtained even under poor operational

conditions. There are several reasons responsible for this high total nitrogen removal rate

obtained in this study.

Firstly, 16 thin strips nonwoven biomass carrier provided a high surface area for the

attachment of anammox sludge. Second was the particle size of seed anammox sludge. Our

used small size of seed anammox sludge (<O.5 mm) enabled the efficient entrapment of

33

anammox sludge inside nonwoven biomass carrier. Anammox sludge grew inside of strips

firstly and then grew outside ofthe surface of strip as shown in Fig.2-1O. We have used the 8

strips nonwoven shown in Fig.2-1 1 for long time as the biomass canier for anammox sludge.

Big size of anammox granules were formed on the surface of eight strips of nonwoven and

could not grow inside of strips. After the fu11 development of anammox biomass on

nonwoven biomass carrier, about 10 g-MLSS anammox sludge was taken out from the

reactor and mixed with 30 g of undoing nonwoven fiber shown in Fig.2-12 and added to the

reactor for bridging the gap between thin strips of nonwoven. The gap was fi11ed with

anammox sludge after 145 days ofoperation and extremely high anammox activities of4.68

kg-N m-3 d'i were obtained under HRT ofO.56 hours.

#

Fig. 2-10. Thin strips ofnonwoven carrier

fit t yt

-'tig/ ''f

Fig. 2- 1 1 . Eight strips of nonwoven carrier

Fig. 2-12. Undoing nonwoven fiber in anammox sludge

2.4 Conclusions

We have succeeded in the establishment of anammox reactor using low strength influent

substrate of 150 mg-N 1-i under low temperature of20-250C. TNLR and TNRR reached to 6.5

kg-TN m-3 d-i and 4.68 kg-TN m-3 d-i, respectively. The TNRR achieved in this study was

34

about two times higher than that of others anammox reactor operated in our laboratory. The

results of this study on the establistment of anammox process using low strength TN

concentration under low temperature would be significant in practical industrial application.

2.5 References

A. Olav Sliekers, K.A. Third, Abma,J.G. Kuenen, M.S.M. Jetten.: CANON and Anammox in

a gas-lift reactor, FEMS Microbiol. Letter, 218 ,339-334(2003).

Ikuo Tsushima Yuji Ogasawara, Tomonori Kindaichi, Hisashi Satoh, Satoshi Okabe.:

' Development ofhigh-rate anaerobic ammonium-oxidizing(anammox) biofilm reactors,

Mat. Res., 41,1623 - 1634 (2007).

Strous, M., Heijnen, J.J., Kuenen, J.G,, Jetten, M.S.M.: The sequencing batch reactor as a

powerfu1 tool to study very slowly growing micro-organisms, Appl. Microbiol.

Biotechnol. 50, 589-596( 1998).

Takao Fujii, Kenji Furukawa.: Characterization ofthe microbial community in an anaerobic

ammonium-oxidizing biofilm cultured on a nonwoven biomass carrier, J. Biosci. and

Bioeng., 94(5), 412-418(2002). •

van Dongen, U., Jetten, M.S.M., van Loosdrecht, M.C.M.: The SHARON-anammox process

for treatment of ammonium rich wastewater, Mater Sci. Technol. 44 (1) , 153-160

( 2001).

35

36

Chapter 3

Effect of Salt Concentration in Anammox Treatment Using

Non-woven Biomass Carrier

3.1 Introduction

Nitrification-denitrification process has been widely used for nitrogen removal from

wastewater. However, not only a large amount of oxygen for nitrification and addition of

external carbon sources (i.e., methanol) for denitrification are required but considerable

amount of excess sludge is also produced. In order to overcome these factors, attempts have

been made to develop the alternate biological nitrogen processes. One of the processes

recently developing was the partial nitritation-anammox process, which has many potential

advantages over conventional nitrogen removal processes (van Dongen, U. et al., 2001). In

this nitrogen process, 60 e/o of influent ammonium is converted to nitrite in the aerobic panial

nitritation treatment and then anammox bacteria oxidize remaining ammonium to nitrogen

gas using nitrite as an electron accepter under anoxic conditions, with their growth occurring

by carbon dioxide fixation. For these reasons, the amounts of oxygen supply, external carbon

source, and excess sludge production can be reduced in this partial nitritation-anammox

process (van Dongen, U. et al., 2001).

Since partial nitritation-anammox process was successfu11y applied to the treatment of

'sewage sludge digester liquor in Netherlands (van der Star, W.R.L. et al., 2007), it opened

doors for application to many kinds of wastewater treatments such as industrial wastewater,

livestock wastewater, and landfi11 leachate. However, these wastewater contain high

concentrations of salts which have been considered as an inhibiting factor in biological

nitrogen removal process. Thus, effects of high salt concentration on nitrification and

denitrification have been previously investigated (GIass, C. et al., 1999; Campos, J. L. et al.,

2002; Moussa, M.S. et al., 2006). It was repored in these studies that nitrification and

'denitrification activities were sustained by gradual acclimation of freshwater sludge to high

37

salt conditions until certain salts levels. Halophilic denitrifying bacteria were isolated from

the long-term acclimated sludge, and higher denitrification performances were demonstrated

when the long-term acclimated sludge was used as inoculum (Yoshie, S. et al., 2006).

Furthermore, Furukawa et al., (Furukawa, K. et al., 1993) reported that nitrifying sludge

taken from night soil treatment plant employing a sea-water dilution in summer season could

adapt more smoothly to high salt condition than the case of freshwater sludge. In other

studies, marine anammox bacteria belonging to "Scalindua" genus have been detected in

natural surroundings (Thamdrup, B. et al., 2002; Kuypers, M.M.M. et al., 2003; Schmid,

M.C. et al.,2007) and very recently Nakajima et al., (2008) enriched them from an enclosed

coastal sea in Japan using a continuous culture system. These results suggested that the

anammox bacteria inherently prefening to the culture containing high concentration of salts

and living in the high salt habitats would be accumulated in the cultivations and available for

industrial application. In other hand, there is an inconsistent experiment. Kartal et al., (2006)

adapted the anammox sludge which consisted of 50 O/o of Candidatus "Kuenia

stuttgartiensis" and 50 O/o of Candidatus "Scalindua wagneri" to the salt concentration of 30

g 1-i. Although it would be the culture condition suitable to the growth of Candidatus

'"Scalindua, he reported that the major anammox bacteria after the acclimation were

Candidatus "Kuenenia stuttgartiensis" enriched from freshwater condition. Because Kartal et

'al., (2006) used the seed sludge containing marine anammox bacteria besides freshwater

anammox bacterium, the result that major anammox bacteria at high salt condition were

freshwater anammox specie is open to question. In addition, Kartal et al., (2006) focused on

only the population of anammox bacterium species without the evaluation of coexistent

bacteria community.

In this study, the effect ofhigh concentration of sodium chloride on anammox treatment

was investigated. We used the anammox fixed-bed reactor using non-woven biomass carrier

with a seed containing only freshwater anammox bacteria, strain KU2 and KSU-1, differing

from those used by Kartal et al (2006). The salt concentration was increased stepwise from

2.5 g 1-i to 33 g 1-i. In addition, the bacterial community was also examined by 16S rRNA

gene analysis after the acclimation ofthe anammox sludge to high salt condition.

38

3.2 Materials and methods

3.2.1 Set up and operational conditions of anammox reactor

An up--flow fixed-bed column reactor with an inner diameter of 95 mm, and a height of

400 mm (an effective voiume of2.8 1) was used in this study as shown in Fig. 3-1. Aporous

polyester non-woven fabric canier (1.0 L, Japan

Vilene, Tokyo, Japan) was used as support material

at a packing ratio of 35.7 O/o. The reactor was

Non-woveninoculated with 1.0 g-VSS of freshwater anammox B osludge, which was taken from another anammox : waterjacket

reactor treating synthetic wastewater without salt 'addition (Qiao, S., ph.D. thesis, Kumamoto Settiingarea

University, Kumamoto, 2007).

The reactor temperature was maintained at P lnfluent2soc during the entire operational period• The PH Of Fig. 3-1 schematic diagram of fixed

bed anammox reactor.the reactor was not adjusted, and the pH of influent

and effluent were around 7.6 and 7.7, respectively. The composition of the synthetic

wastewater used in this study is shown in Table 3-1.

Table 3-1. Composition of synthetic wastewater

Efflu

ee

sr.

Composition Concentration

(NH4)2S04 75mg-N1-i

NaN02 75mg-N.1-i

KHC03 125mgl'l

KH2P04 54mg1-1

T.element o.smlrl

NaCl O-33.og1-i

T. element: FeS04- 7H20, 18g 1-i

'

39

•EDTA•,

2Na, 10 mg 1-i

The synthetic wastewater was flushed with nitrogen gas to decrease the dissolved

oxygen (DO) concentration below 1.0 mg 1'i. The hydraulic retention time (HRT) was varied

from 36 h to 1.7 h depending on the nitrogen loading rate (NLR). The reactor was fed with

synthetic wastewater without salt addition until the nitrogen removal rate (NRR) increased to

1.6 kg-N m-3 dHi (day 102). After that, the NLR was maintained at 2.2 kg-N m-3 d-i and the

infiuent salt concentration was increased stepwise from 2.5 g 1-i to 33 g lmi.

3.2.2 Analytical methods

The concentrations ofN02-N and N03-N were measured by the colorimetric method in

accordance with the Standard Methods (APHA, AWWA, WPCF 1995). NH4-N concentration

was measured by the modified phenate method using ortho-phenyl phenol (OPP) (Kanda, J.

1995). Absorbance and pH values were measured using a spectrophotometer (U-2010;

Hitachi, Tokyo, Japan) and a pH meter (B-211; Horiba, Kyoto, Japan), respectively.

3.2.3 DNA extraction and PCR amplification

A sludge sample was taken from the reactor after its acclimation to high salt

concentration (day 320). The sample was suspended in 1 ml of TE buffer with 1 pt1 of

Ready-Lyse Lysozyme Solution (EPICENTRE, U.S.A.) and incubated at 37 OC for 1 h. The

solution was added with 1.5 mg of achromopeptidase (Wako, Osaka, Japan) and incubated at

37 OC for 30 min. Then, bacteriolysis was performed by addition of 100 pl of 10 O/o sodium

dodecyl sulfate solution. Proteins in the supernatant prepared by centrifugation were

decomposed with Proteinase K (Wako) treatment. The supernatant was prepared by

centrifugation and meta-genomic DNA in it was purified by phenoYchloroform extraction.

The DNA was recovered by ethanol. The meta-genomic DNA was dissolved in TE buffer,

treated with RNase A, precipitated by ethanol with 13 O/o PEG8000-1.6 M NaCl, and

dissolved again in TE buffer. The amplification of 16S rRNA gene was performed with KOD

-plus- DNA polymerase (TOYOBO, Osaka, Japan) using eubacterial primers 6F (forward

40

primer: 5'-GGAGAGTTAGATCTTGGCTCAG-3') and 1492R (reverse primer:

5'-GGTTACCTTGTTACGACT-3') (Lane, D.J. 1991; Tchelet, R. et al., 1999). PCR

(polymerase chain reaction) was carried out according to the following thermocycling

parameters: 2 min initial denaturation at 94 OC, 25 cycles of 15 sec at 94 OC, 30 sec at 60 OC,

and 30 sec at 68 OC. The amplified products were electrophoresed on a1 O/o agarose gel and

the DNAs in excided agarose gel were purified using Wizard SV Gel and PCR CIean-Up

System (Promega, U.S.A.).

3.2.4 Cloning and sequencing of 16S rRNA

The purified fragments were ligated into the Hincll site of pBluescript II KS+

(Stratagene, U.S.A.). E. coli DHIOB was transformed using the constructed plasmids. The

plasmids were extracted by the alkaline method from 33 clones carrying them. The

nucleotide sequences ofinserted DNA in them were determined with 3130xl genetic analyzer

and BigDye terminator v3.1 cycle sequencing kit (Applied Biosystems, U.S.A.). The

sequences determined in this study were compared with those in nr database by the basic

local alignment search tool (BLAST) program on the NCBI web site.

3.2.5 Denaturing gradient gel electrophoresis (DGGE)

Partial 16S rRNA gene was amplified by PCR using a eubacterial primer set, 357F

(forward primer: 5'-CCTACGGGAGGCAGCAG-3') and 534R (reverse primer:

5'-ATTACCGCGGCTGCTGG-3') (Muyzer, G et al.,1993), and the extracted meta-genomic

DNA as template. The amplified fragments were purified and added with the GC-clamp

(5'-CGCCCGCCGCGCCCCGCGCCCGTCCCGCCGCCCCCGCCCG-3') at the 5' termini

by second PCR using a primer set, 357F with GC-clamp and 534R. The products were

resolved by DGGE for 16 h at 100V at 60 OC using DCode system (Bio-Rad, U.S.A.). An 8

O/e acrylamide gel with a 30-65 O/o denaturing gradient was used, where 100 O/o denaturant was

defined as 7 M urea and 40 O/o form amide. The gel was stained with SYBR-Gold solution

(Invitrogen, U.S.A.), and visualized using FLA-2000 system (Fuji Photo Film, Tokyo, Japan).

41

Five bands were excised from the gel to determine the sequences. The gel piece was crushed

with disposable polypropylene pestles and soaked in the DNA extraction buffer (500 mM

sodium acetate (pH 5.2), 1 mM EDTA). After ovemight incubation at 4 OC, the mixture was

centrifuged and the supernatant containing the extracted DNA was transferred into a new

tube. The DNA was ethanol precipitated, dissolved in TE buffer and amplified by PCR with

a primer set, 357F and 534R. The amplified fragment was directly sequenced.

3.2.6 Nucleotide sequence accession numbers

The partial 16S rRNA gene sequences of Operational taxonomic unit (OTU) 1, OTU 2,

OTU 3, OTU4, OTU 5, OTU 6, and OTU 7 were submitted to the DDBJ database under

accession numbers AB434253-AB434256, AB434257, AB434258, AB434259, AB434260,

AB434261, and AB434262, respectively.

3.3 Results and discussion

3.3.1 Reactor performances under high salt conditions

After achieving a NRR of 1.6 kg-N m-3 d-i (day 102) without salt addition, the TNLR

was maintained at 2.2 kg-N m"3 d-i and salt addition was initiated. Fig.3-2 shows the time

courses of nitrogen concentrations under various salt concentrations. When the salt

concentration was set at 5 g 1'i, the effluent NH4-N and N02-N concentrations were increased.

However, the effluent NH4-N and N02-N concentrations decreased by decreasing the salt

concentration from 5 g 1-i to 2.5 g 1'i. Thereafter, although the salt concentration was

increased to 5 g 1-i, the effluent NH4-N and N02-N concentrations did not increase. The salt

concentration was then increased stepwise to 30 g 1-i. During this period, the effluent NH4-N

and N02-N concentrations remained constant demonstrating successfu1 performance of

anammox treatment under high salt conditions. The anammox treatment was successfu11y

maintained at the salt concentration of 30 g 1"i between day 195 to 203. However, when the

salt concentration was further increased above 30 g 1-i, the effluent NH4-N nd N02-N

42

concentrations increased. Consequently, the salt concentration was decreased to 28 g IJi until

the efficiency of anammox treatment was recovered. After recovery, the salt concentration

was again increased to 30 g 1-i. As shown in Fig. 3-2 the anammox treatment then became

stable at the salt concentration of 30 g 1-i and the T]NFRR of 1.7 kg-N m-3 d-i was achieved for

the ensuing 65 days (day 272-337).

-r--l

z/da

S.E'

gg:&.g

scz

foo

ee

ge

'7g

6e

•5if

4ff

3"

21g

Ie

g

Sak c"nc.: 3e g l-a

" gls i?ec sf Lag ,ti/ini• iE x x• -

tUbli va

st

t "tltt

ue':=ajtt,ts"t}tsl'lsl""',t.,"th,k=""ca'fi/fta'wtt .-'xx,ei-.,s"',t've

asx. iiliasrfinv"geaslj.,ge.va-,rm• ee nc ut ee N. S•, *. ili ee lfi

'

p ' "] ewi tsbptP, . S di0 6>

'pm'

iilll81II}}SS/ilii,li,l.le.,.l,CS'"ll.ll•ll,ik:2,inv•IIi•lii:•ts,tl'illtstkmu:,.••.t/,.L,at.ts-ti:h.f.t,:st,:,,tag,i•l;!,.c,kmamul••"kl'th"s•tg"s•"'••ft,'"C'O'•'b•

35

3•e

2-5

2ts

f•5

fie

5

-s

-rpmseti-su

•=ets

stusc

vaiiiEtlR,twngtw"'4-•L'th•,i.ep/ts.,i{[/t/s,tik,,t,t.

fee fsg 2eic 2-sg 3cg ' 3sc Tl lme {d/}

Fig. 3-2 Time courses of nitrogen concentrations under various salt concentrations. Symbols: D influent NH4'; Xinfluent N02-; Q effiuent

NH4'; O effluentNOi; A effluentN03';solidline,salt.

Dapena-Mora et al., (Dapena-Mora, A. et al., 2006) applied anammox treatment to the

digester liquor from fish canning industry using a SBR, and reported that the NRR was O.3

kg-N m-3 d-i at a salt concentration of 10 g 1-i. Kartal et al., (2006) operated a SBR for

anammox treatment with salt addition, and reported that the NRR was 1.0 kg-N m-3 d"i at a

salt concentration of 30 g 1-i. In contrast, the TNRR at the salt concentration of 30 g 1'i was

1.7 kg-N m-3 d-i in this study. Relatively lower NRRs reported by Dapena-Mora et al., (2006)

and Kartal et al., (2006) may be attributed to the fact that because water density was

increased by increasing salt concentration, the sludge with lesser density might have been

washed out of the SBR. On the other hand, high amount of anammox sludge was

consistently maintained in the fixed-bed reactor using non-woven biomass canier in this

study resulting in higher TNRR of 1.7 kg-N m-3 d-i. An Anammox reaction ratio (NH4-N

' 43

consumption, N02-N consumption, N03-N production) of 1 : 1.32 : O.26 is generally

accepted by researchers (Strous, M. et al., 1998; Dapena-Mora, A. et al., 2004). In this study,

the anammox reaction ratio at the salt concentration of 30 g 1-i (day 103-337) was 1 : 1.29 :

O.24, which was close to the accepted ratio (Strous, M. et al., 1998).

The relationship between the TINIRR and

the salt concentration is illustrated in Fig.

3-3. The TNRR was maintained atapproximately 1.65 kg-N m-3 d-i under salt

concentrations ranging between 10 g 1-i and

30 g 1-i. However, the TNRR sharply

declined at a salt concentration of more than

30 g 1-i. Kartal et al., (2006) also observed

that the anammox activity of the SBR was

completely lost when the salt concentration

was increased from 30 g 1-i to 45 g 1-i. Based

on these results, it is suggested that

a

that saltwater acclimated sludge might

concentrations in excess of 30 g 1-i. For

denitrifying sludge which was acclimated to

resulted in high denitrification

Therefore, it is lik

ofmore than 30 g 1-i

"T,

vny:

Ezile,

egif

-N'

>eEfl

EzaEntz

2.e

fe5

1.e

ij.s

e,.e

ll • t l -l ""ti k l

+

freshwater

pplied to anammox treatment at salt concentrations higher th

be

example,

performance

ely that the anammox sludge could p

by long-term acclimation at elevated salt concentrations.

3.3.2 Bacterial community under salt condition

Thirty three 16S rRINA gene sequences ofbacterial members in the community existing

under high salt condition (day 320) were determined. Table 3-2 shows homology search

results for 16S rRNA gene sequences. The major clone in OTU 1 had 100 O/o sequence

identity with uncultured bacterium clone AnDHS-2 (AB430333) which isolated from an

44

e 5 fg {5 2•e 25 •3ts 35 SaEt cenen. (9 }-")

Fig.3-3. Relationship between the nitrogen removal rate (TNRR) and salt concentration. Bars indicate standard deviation.

acclimated anammox sludge can not be

an 30 g 1-i. However, it appears

capable of treating wastewater with salt

Yoshie et al., (2006) reported that a

the salt concentration of 20 g 1-i for 3 years

even at a salt concentration of 100 g 1"i.

rovide treatment at salt concentration

anammox reactor, and this clone might be affiliated with candidate division OPIO (Crocetti,

GR. et al., 2002). Five clones in OTU 2 and 1 clone in OTU 3 had 100 O/o sequence identity

with KU2 and 99 O/o with KSU-1, respectively. These bacteria have been recognized as

anammox bacteria and enriched in the anammox reactor treating synthetic wastewater

without salt addition (Fujii, T. et al., 2002). Two clones in OTU 4 had 99 O/o sequence identity

with Lysobacter sp.

Fig. 3-4 shows DGGE profiles for 16S rRNA gene

sequences of bacterial members in the community

existing under high salt condition (day 320). Four

explicit bands and some minor bands were detected

after the resolution of amplified DNA fragments by

DGGE. The nucleotide sequences of the explicit bands

were directly sequenced. Except for the primer regions,

the determined sequences ofBands 1, 2, and3 were in

agreement with the 16S rRNA sequences of OTUs 4

(AB434259, nts 331-490), 1 (AB434253-AB434256,

nts 319-454), and 2 (AB434257, nts 365-525),

respectively. The nucleotide sequence of Band 4 was

not involved in any sequences of 16S rRNA genes

cloned in this study, and showed low identities (<90 O/o)

to the 16S rRNA gene sequences ofbacteria belonging

to phylum Chloroflexi (data not shown). In addition,

because the intensity of Band 2 was the highest, OTUI

might be dominant in the reactor.

From these results, it was confirmed that

the freshwater anammox bacteria (KU2) and

concentration of 30 g 1'i.

method,

3.

the primer sets in the cloning and DGGE.

45

i'iilii,li

.,I,!-/thi,illlllll

.-.

I' ix•.1

unidentified bacteria which perhaps belong to candidate division OPIO were dominant

Lysobacter

Although Band 4 was not detected and not identified by the cloning

the intensity of the band seemed to be almost the same as those of Band 1 and Band

This result was likely caused by the affmities differences for the 16S rRNA gene between

Fig. 3 - 4. Denaturing gradient gel

electrophoresis (DGGE) profiles for

16S rRNA gene sequences of bacterial

members in the community existing

under salt condition (day 320).

, and

sp. existed under a high salt

ChloroLflexi bacteria and unidentified bacteria K4-33 (AY793665) coexisted with

anammox bacteria (KU2) in the inoculums (Qiao, S., Ph.D. thesis, Kumamoto University,

Kumamoto, 2007). On the other hand, unidentified bacteria which perhaps belong to

candidate division OPIO and Lysobacter sp. coexisted with anammox bacteria under a salt

concentration of 30 g 1'i. Thus, it is assumed that the microbial community shift occurred by

increasing salt concentrations. Since the bacterial members in the community was examined

only at the salt concentration of 30 g 1"i in this study, fUrther research is required to know the

process of microbial community shift. In addition, although anammox bacteria were

normally dominant in the freshwater condition, unidentified bacteria which perhaps belong

to candidate division OP1O were dominant under salt concentration of30 g 1-i (Fujii, T. et al.,

2002; Kindaichi, T. et al., 2007). Thus, the unidentified bacteria seem to be important role in

the reactor. However, since the metabolism of the unidentified bacteria is unknown, further

investigation must be required.

Table 3-2. Homology search results for 16S rRNA gene sequences ofbacterial members in the

communi exjstin undersaltcondition(da 320).

OTU Taxon (Accession No.) Identity

( o/o)

Numberof clones

1

Uncultured bacterium clone AnDHS-2 (AB430333)

Uncultured bacterium clone YCB35 (EF205445)

Uncultured bacterium clone FCPT516 (EF515998)

Uncultured bacterium clone vf23 (DQ975217)

Uncultured candidate division OP1O bacterium clone

HAVOmat40 (EF032775)

99

91

90

90

90

21

2Anoxic sludge bacterium KU2 (AB054007)

Kuenenia stuttgartiensis (CT573071)99 5

3Planctomycete KSU-1 (AB057453)

KIST-JJYOOI (EF515083)99 1

4 Lysobacter sp. (AJ298291) 99 2

5 Mesorhizobium sp. Ala-3 (AM491621) 98 1

6

7

Unclassified

Unclassified

2

1

Total 33

OTU: Operational taxonomic unit. Unclassified: The sequence does not have homology more than85 O/o identity to any one in nr database.

46

There appear to be two plausible explanations of our obtained results that the anammox

treatment was maintained under high salt conditions. One is that anammox bacteria in the

freshwater sludge were replaced by other salt-tolerant anammox strains such as marine

species under high salt conditions. The other is that freshwater anammox bacteria existing in

the inoculums were originally salt-tolerant and could survive even under high salt

concentration of 30 g 1-i. Kartal et al., (2006) adapted the anammox sludge which consisted

of 50 O/o of Candidatus "Kuenia stuttgartiensis" and 50 O/o of Candidatus "Scalindua

wagneri" to the salt concentration of 30 g IHi, and reported that the major anammox bacteria

after the acclimation were Candidatus "Kuenenia stuttgartiensis" enriched from freshwater

condition. Although our cultivation method was different from that of Kartal et al., (2006),

we also confirmed that the major anammox bacteria after the acclimation to high salt

condition were KU2 enriched from freshwater condition. Therefore, it can be deduced that

there is a possibility that freshwater anammox bacteria had the salt-tolerant characteristics

and cloud survive even under high salt conditions.

3.4 References

APHA, AwwA, WPCF: Standard method for the examination ofwater and wastewater, 19th

ed. American Public Health Association, Washington, D.C. (1995).

Campos, J. L., Mosquera-Corral, A., Sanchez, M., Mendez, R., and Lema, J. M.:

Nitrification in saline wastewater with high ammonia concentration in an activated

sludge unit, UlaterRes., 36, 2555-2560 (2002).

Crocetti, G.R., Banfield, J.E, Keller, J., Bond, P.L., and Blackall, L.L.:

Glycogen-accumulating organisms in laboratory-scale and fu11-scale wastewater

treatment processes, Microbiology, 148, 3353-3364 (2002).

Dapena-Mora, A., Campos, J.L., Mosquera-Corral, A., and Mendez, R.: Anammox process

for nitrogen removal from anaerobically digested fish canning efftuents, Mater Sci.

Technol., 53, 265-274 (2006),

Dapena-Mora, A., Campos, J.L., Mosquera-Corral, A., Jetten, M.S.M., and Mendez, R.:

Stability ofthe ANAMMOX process in a gas-lift reactor and a SBR, J. Biotechnol., 110,

47

159-170 (2O04).

Fujii, T., Sugino, H., Rouse, J. D., and Furukawa, K.: Characterization of the microbial

community in an anaerobic ammonium-oxidizing biofilm cultured on a nonwoven

biomass canier, J. Biosci. Bioeng., 94, 412--418 (2002).

Furukawa, K., Ike, A., and Fujita, M.: Preparation of marine nitrifying sludge, J. Ferment.

Bioeng., 76, 134-139 (1993).

Glass, C. and Silverstein, J.: Denitrification ofhigh-nitrate, high-salinity wastewater, Mater

Res., 33, 223-229 (1999).

Kanda, J.: Determination of ammonium in seawater based on the indophenol reaction with

o-phenylphenol (OPP), varateTRes., 29, 2746-2750 (1995).

Kartal, B., Koleva, M., Arsov, R., van der Star, W., Jetten, M.S.M., and Strous, M.: Adaption

of a freshwater anammox population to high salinity wastewater, J. BiotechnoL, 126,

546-553 (2006).

Kindaichi, T., Tsushima, I., Ogasawara, Y., Shimokawa, M., Ozaki, N., Satoh, H., and Okabe,

S.: In situ activity and spatial organization of anaerobic ammonium-oxidizing

(Anammox) bacteria in biofilms, Appl. Environ. MicTobiol., 73, 4931-4939 (2007).

Kuypers, M.M.M., Sliekers, A.O., Lavik, G, Schmid, M., Jorgensen, B.B., Kuenen, J.G.,

Sinninghe Damste, J.S., Strous, M., and Jetten, M.S.M.: Anaerobic ammonium

oxidation by anammox bacteria in the Black Sea, Nature, 422, 608-6l1 (2003).

Lane, D.J.: 16S123S rRNA sequencing. In Nucleic Acid Techniques in Bacterial Systematics

ed, Goodfellow, M. Chichester, UK, Wiley, 115-148 (1991).

Moussa, M.S., Sumanasekera, D.U., Ibrahim, S.H., Lubberding, H.J., Hooijmans, C.M.,

Gijzen, H.J., and van Loosdrecht, M.C.M.: Long term effects of salt on activity,

population structure and floc characteristics in enriched bacterial cultures of nitrifiers,

MaterRes., 40, 1377-1388 (2006).

Muyzer, G., de Waal, E. C., and Uitterlinden, A. G: Profiling of complex microbial

populations by denaturing gradient gel electrophoresis analysis of polymerase chain

reaction-amplified genes coding for 16S rRNA, AppL Environ. Microbiol., 59, 695-700

(1993).

Nakajima, J., Sakka, M., Kimura, T., Furukawa, K., and Sakka, K.: Enrichment of anammox

bacteria from marine environment for the construction ofa bioremediation reactor, Appl.

48

Microbiol. Biotechnol., 77, 1159-1166 (2008).

Schmid, M.C., Risgaard-Petersen, N., van de Vossenberg, J., Kuypers, M.M.M., Lavik, G.,

Petersen, J., Hulth, S., Thamdrup, B., Canfield, D., Dalsgaard, T., Rysgaard, S., Sejr,

M.K., Strous, M., den Camp, H.J.M.O., and Jetten, M.S.M.: naerobic

ammonium-oxidizing bacteria in marine environments: widespread occurrence but low

diversity, Environ. Microbiol., 9, 1476-1484 (2007).

Strous, M., Heijnen, J. J., Kuenen, J. G, and Jetten, M. S. M.: The sequencing batch reactor

as a powerfu1 tool for the study of slowly growing anaerobic ammonium-oxidizing

microorganisms, Appl. Microbiol. Biotechnol., 50, 589-596 (1998).

Tchelet, R., Meckenstock, R., Steinle, P., and van der Meer, J.R.: Population dynamics of an

introduced bacterium degrading chlorinated benzenes in a soil column and in sewage

sludge, Biodegradation, 10, 113-125 (1999).

Thamdrup, B. and Dalsgaard, T.: Production of N2 through anaerobic ammonium oxidation

coupled to nitrate reduction in marine sediments, Appl. Environ. Microbiol., 68,

1312-1318 (2002).

van der Star, W.R.L., Abma, W.R., Blommers, D., Mulder, J.W., Tokutomi, T., Strous, M.,

Picioreanu, C., and van Loosdrecht, M.C.M.: Startup reactors for anoxic ammonium

oxidation: Experiences from the first fu11-scale anammox reactor in Rotterdam, Mater

Res., 41, 4149-4163 (2007).

van Dongen, U., Jetten, M. S. M., and van Loosdrecht, M. C. M.: The SHARON-Anammox

process for treatment of ammonium rich wastewater, Mater Sci. Technol., 44, 153-160

(2001).

Yoshie, S., Ogawa, T., Makino, H., Hirosawa, H., Tsuneda, S., and Hirata, A.: Characteristics

of bacteria showing high denitrification activity in saline wastewater, Lett. in Appl.

Microbiol. , 42, 277-283 (2006).

49

50

Chapter 4

Nitrogen Removal Capabilities of Two Kinds of Fixed-bed Anammox

Reactors for Treating Partial Nitrification Brine Wastewater

4.1. Introduction

Natural gas and iodine dissolved in brine water are now recovered commercially in

Chiba Prefecture, Japan. In the process of natural gas and iodine production, the brine

wastewater containing NH4-N is produced as by-product. The salinity of this brine

wastewater is almost the same level of sea water. For the removal ofNH4-N from this brine

wastewater, the development of an economical nitrogen removal processris urgently

required.

Compared with the traditional biological nitrogen removal process such as

nitrification-denitrification process, anammox process was now regarded to be a novel,

promising, and cost effective alternative. Anammox process has many advantages, e.g., no

requirement of external carbon sources, low oxygen demand, minimized excess sludge

production and reduction in C02 emissions (Op Den Camp et al., 2006; Liu et al., 2008).

Single-reactor high-activity ammonium removal over nitrite (Sharon) process was

successfu11y applied for partial nitritation process, which is the requisite pretreatment before

applying anammox process for NH4-N removal. The Sharon-Anammox system was proved

to be an economical nitrogen removal process for the treatment of wastewaters with low

carbon to nitrogen ratio (C/N), and the running costs and C02 emissions during this new

NH4-N removal process could be decreased up to 60 O/o and 90 O/o, respectively (van Dongen

et al., 2001). In the partial nitritation process, about 50 O/o of influent NH4-N must be nitrified

to nitrite nitrogen (N02-N), and this effluent is ideally suited as the influent for subsequent

anammox process. In the autotrophic anammox process, NH4-N is oxidized by N02-N under

anaerobic conditions and produce dinitrogen gas and a small amount of nitrate nitrogen

51

(N03-N). The generally accepted stoichiometry of anammox process was depicted as follows

(Strous et al, 1998).

NH4' + 1.32N02- + O.13H' + O.066HC03- -)

1.02N2 + O.26N03- + 2.03H20 + O.66Biomass (Eq.1)

Salinity is generally an important factor for wastewater treatment, because many

industrial wastewaters, such as landfi11 leachate, leather industry wastewater, fish-canning

wastewater, tannery wastewater and the above mentioned brine wastewater from natural gas

producing plant, contain both high concentrations of NH4-N and salt. The high salinity of

these kinds of wastewaters had been considered to exert negative effects on biological

nitrogen removal (Chen et al., 1971; Panswad and Anan, 1999; Dincer and Kargi, 1999). The

presence ofhigh salinity in wastewater induces salinity stress to the microbial species, results

in the inhibition of many enzymes, decreases cell activity, and eventually leads to

plasmolysis (Uygur, 2006). Especially for the treatment of brine wastewater from the natural

gas plant, the salinity level is almost the same level as of sea water. However, recent studies

on the effects of high salinity on partial nitritation and anammox processes have illustrated

that both processes could be operated stably after the suitable acclimation period of

freshwater bacterial community to high salt condition. For partial nitritation process, the

Sharon reactor could work stably under high salt concentrations up to 427 mM (about 25 g 1'i)

of NaCl after the adaptation of microorganisms to saline environment (Mosquera-Corral et

al., 2005). On the other hand, for anammox process, the mass cultivation of marine

anammox sludge was not established yet, though Nakajima et al.,(2008) succeeded in the

enrichment of marine anammox bacteria belonging to "Scalindud' genus from an enclosed

coastal sea in Japan (Nakajima et al., 2008). However, there were few reports on the

acclimation of freshwater anammox bacteria to high salt condition. Very recently, some

anammox processes using fixed-bed reactor, sequencing batch reactor (SBR), and rotating

biological contactor (RBC) in an oxygen-limited sutotrophic nitrification-denitrification

(OLAND) system had been successfu11y operated under high salinity condition (Liu et al.,

2009; Kartal et al., 2006; Windey et al., 2005). In these reports, all anammox reactors were

operated with synthetic inorganic saline wastewater and adaptation experiments of

freshwater anammox sludge to cultural condition with high salts concentration were carried

52

out by stepwise increase in influent salt concentrations. Only Dapena-Mora et al.,(2006)

reported a SBR anammox reactor treating real partial nitrification fish-canning wastewater

under relatively low salinity of 10 g NaCl 1-i (Dapena-Mora et al., 2006). Among all these

reactors, the highest TNNR of 1.7 kg-N m-3 d-i was reported under salinity level of 30 g

NaCl 1-i (Liu et al., 2009).

The objectives of this research are to make clear the applicability of two kinds of

fixed-bed anammox reactors to the treatment of partial nitrification brine wastewater and

compare the treatment capabilities of anammox reactors using different kinds of biomass

canier for this wastewater.

4.2. Materials and Methods

4.2.1. Anammox reactors

The reactors used in the r-----;E:;ffiuent rtEffiuent

fixed-bed anammox processes

are shown in Fig. 4-1. The

reactor was made of Perspex

with available volume of 2.81 (rp

95 mm Å~ 423 mm). The influent

was supplied to the reactor by

up-flow mode. The reactor

temperature was maintained at

about25Å}2 OC

and kept at room temperature in

vinyl sheet enclosure.

anammox reactor (a)) for 3 months

biomass carriers

l2u

'x

x, tt•I

i.,{•ecst•i

.ttlg.

•Å}?tt•

}ll•ili,x

"x

{a}

Nen-woven Small carriers

ti

lnfluent lnfluent {b)

Fig. 4-I Fixed-bed anammox reactors ,

oC

sheltered from light by black

by heating through water jacket when the temperature was lower than 25

other days. The reactor was

After a continuous operation of the reactor using nonwoven biomass canier (fixed-bed

, this reactor was changed to a reactor using small size

(fixed-bed anammox reactor (b)). A point of difference for two reactors is

only the difference in applied biomass canier.

53

4.2.2. Biomass carriers

Nonwoven porous polyester coated with a pyridinium-type polymer (Japan Vilene, US

patent 5,185,415; 1993) was used as biomass canier for fixed-bed anammox reactor (a). The

original nonwoven has 8 strips. For the purpose of increasing specific surface area and

nutrients diffusion rate of nonwoven

biomass carrier, 8 strips of nonwoven

was split into 16 strips using razor.

The split column-like nonwoven

canier (100 mm diameter, 2.5 mm

thickness for one trip, 395 mm length,

shown in Fig. 4-2) was mounted into

fixed-bed anammox reactor (a).

In fixed-bed anammox reactor (b),

biomass caniers. The small size

cement (Portland blast-furnace slag),

hardcore-1

,.ffss

Fig. 4-2 Original and split nonwoven biomass carriers (top view)

small size caniers were used instead of nonwoven

biomass caniers consisted of nonwoven small filaments,

calcined lime, steel slag (smelt steel slag) and

(War industry of military department product without opened). Nonwoven small

filaments were prepared with a mean size of lmm length and 50 pm width by untying and

cutting nonwoven strips. The cement was neutralized by soaking it in running tap water for 2

weeks. The steel slag, whose original diameters were from 1 to 5mm, was sieved to obtain a

mean diameter of1 mm. The hardcore-1 was a special high active agent, which was a

powerfu1 nanophase material.

360g of these upper mentioned small materials were mixed together (the dry weight

ratio ofcement:calcined lime & slag:nonwoven filaments:hardcore-1 was 200:200:50:

5) and put into 1.0 1 tap water. This prepared mixed slurry was named as small size biomass

caniers, which was put into fixed-bed anammox reactor (b) as biomass canier after mixing

well with anammox inoculums.

54

42.3. Brine water from Nature Gas Company

Table 4-1 shows the compositon ofwater qualities of the brine water and sea water. The

concentrations, especially of I- and Br- , are obvious different from that of sea water. After

the removal ofusefu1 materials (natural gas and iodine) from this brine water, the remained

brine wastewater contains high concentration of NH4-N. The mean salinity, NH4-N

concentration and pH value of this brine wastewater are 30 g 1-i, 200 mg-N 1-i and 6.9,

respectively. The salinity of the brine wastewater used in this experiemnt is almost the same

as that of sea water.

Table 4-1 Composition ofbrine water and sea water

Component Brine Water(mg 1-i) Sea Water(mg ri)

rCl-

Br-

Na'

K'Ca2'

Mg2'

So42-

HC03- 2-C03

C02HBo32-

Fe (Total)

11O-130

18,OOO-19,500

120

1O,OOO

3OO

190

500

1,OOO

10-30

10

2-5

O.1

18,OOO

60

9,OOO

350

370

1,200

2,500

1OO

5.9

22

02

4.2.4. Partial nitritation process

Nitritation treatment of the brine wastewater was carried out for providing suitable

influent to sebsequent anammox treatment. The partial nitritation process was applied by a

'swim bed reactor using thread type acrylic fiber biomass carrier.

55

4.2.5. Influent of anammox reactors

The partially nitrification brine wasterwater was supplied to the fixed-bed anammox

reactors as influent. The water qualities of the partial nitritation effluent were not stable in

some cases. Therefore, a 300 1 storage tank for collecing the effluent from the partial

nitritation reactor was utilized and the storage nitirified brine wastewater was used as the

influent of the fixed-bed anammox reactors. The storage partial nitrified brine wastewater

was supplied to the anammox reactors directly after measuring its water quality during stable

panial nitritation treatments. In case of unstable partial nitritation periods, the N02-N and

NH4-N concentrations of the partial nitirified brine wastewater was adjusted to a proper

concentration ratio for the subsequent anammox reactor by adding sodium nitrite and

ammonium sulfate. The composition ofboth unadjusted and adjusted influents was 75Å}15

mg-NH4-N l-i, 90Å}20 mg-N02-N 1"i, 160Å}25 mg-TN 1"i (as shown in Table 4-2). The average

ratio of NH4-N to N02-N of the influent fed to anammox reactor during the whole

experiment period was about 1:1.2.

The influent N03-N concentration of the fixed-bed anammox reactors was also fluctuant

sometimes. The influent N03-N is not engaged in the anammox reaction, so that this N03-N

concentration was ignored and excluded in this study. Consequently, the TN concentration

was calculated by the sum ofNH4-N and N02-N concentrations for the influent and plus the

producing N03-N concentration for the effluent. The TNLRs applied to the anammox

reactors were calculated based on the influent TN (NH4-N + N02-N) concentration. The

TNRRs of the anammox reactors were determined by taking into account the producting

N03-N concentration from the anammox reaction.

4.2.6. Start-up and operational strategy of anammox reactors

The fixed-bed anammox reactor using nonwoven biomass carrier was operated

sequentially after the previous study of this reactor treating synthetic high saline wastewater.

This reactor was started-up with the initial TNLR of O.8 kg-N m-3 d-i and a hydraulic

56

retention time (HRT) of 5.6 hr. After the operation of fixed-bed anammox reactor (a) was

stopped, both the anammox biofilm detached from the nonwoven carrier and small amount

of settled sludge in the bottom part of the up-flow reactor were collected. The detached and

collected anammox sludge was mixed completely with the prepared small size biomass

carriers manually, and then the carriers and sludge were put into the reactor. The initial

TNLR of the fixed-bed anammox reactor (b) was O.89 kg-N m-3 d-i with a HRT of 5.17 hr.

The initial TNLRs for both fixed-bed anammox reactors were set at relatively high values

owing to the ' high initial acclimated anammox sludge concentrations. The TNLR was

increased only by increasing flow rate (or reducing HRT). The protocol for increasing in

TNLR was increasing TNLR stepwise by O.18--O.2 kg-N m-3 d-i each time in case of the

effluent N02-N concentration below 20 mg 1-i most oftime.

The experiments were divided into four Runs, the former two Runs were the experiments

for fixed-bed anammox reactor (a) and the later two Runs were the experiments for fixed-bed

anammox reactor (b). In Run I (from day O to 58), reactor (a) was operated with a quick

increase in TNLR. In Run M (from day 85 to 133), fixed-bed anammox reactor (b) was

operated with a quick increase in TNLRs. In Run ]V (from day 134 to 174), the increasing

rate of TNLR was lower than that of Run M due to the shortage of the partial nitrified brine

wastewater and low substrate concentrations. The operational conditions during the whole

experiment were shown in Table 4-2.

The influent dissolved oxygen (DO) concentrations during the whole experiment were

ranged from 5 to 7 mg 1"i (Tabel 2). DO is toxic to anammox consortium, because anammox

activity can be expected only under strict anoxic conditions (Jetten, 2001). However, some

coexisted oxygen consuming microbial consortia, such as aerobic nitrifiers and floc forming

heterotrophs, could protect the anammox bacteria from the effect of oxygen (Fujii et. al.,

2002; Qiao et al., 2008; Liu et al., 2008 b). In order to save the high cost required for keeping

absolutely anaerobic condition, the influent with DO concentration of 5-7 mg 1-i was

supplied directly to the fixed-bed anammox reactors without nitrogen gas flushing.

57

Table 4-2 Operational conditions during the whole experimental period

Parameters

Time range (day)Fixed-bed anammox reactor

Biomass carrier

HRT (hr)InfluentNH4-N (mg-N1-i)

InfluentN02-N (mg-N1-i)

Influent TN (mg-N1-i)

TNLR (kg-N m-3d-i)

Temperature

( .C)

Salinity (g r')DO influent (mg l")

Run I Run ll Run M Run IV

O-58

(a)

Nonwoven5.6-2.17

73'v84

94-103

167-186

O.8-2,12

25Å}2

30

5r"7

59-84

(a)

Nonwoven 2.12-1.64

72-84 97rvl07

184-187 2.12-2.78

27Å}2

30

5-7

Small carriers

85-133

(b)

5.17-1.76

72"v89

88--103

168--188

O.69-2.57

27Å}2

30

5-7

134-174

(b)

Small carriers

1.73-1.12

64ew78

70'v82

137-169 2.51-3.07

25Å}2

30

5-7

There is no pH adjustment to the influent though the pH ofpartially nitrified wastewater

was fluctuant. The temperature was heated to about 25Å}2 @ in cool days and kept at room

temperature in other days. The influent salinity was unchanged at 30 g 1-i in the entire

experimental period.

4.2.7. Analytical methods

The concentrations of N02-N and N03-N were measured by the colorimetric method in

accordance with the Standard Method (APHA, 1995). NH4-N concentration was measured

colorimetrically by the phonate method using ortho-phenylphenol as a substitute for liquid

phenol (Kanda, 1995). DO concentration of the influent was measured by using a digital,

portable DO meter (D-55, Horiba). The concentration of mixed liquor volatile suspended

solids (MLVSS) of sludge samples was deterrnined in accordance with Standard Methods

(APHA, 1995). pH value was determined by using a pH meter (IM-22P, TOA EIectronics).

58

Biomass attached on small biomass caniers was observed by an electron microscope (Nikon

Eclipse E600, Japan) and a digital camera (Nikon 4500, Japan).

Scanning electron microscopy (SEM) was used for the observation of small carriers

with attached biomass, the procedure is as follows. Samples were first washed in a O.1 M

phosphate buffer solution (pH 7.4) for 5 min, and then hardened for 90 min in a 2.5 O/o

glutaraldehyde solution prepared with the buffer solution prepared with the buffer solution.

Next, samples were washed in the buffer solution three times for 1O min once and then fixed

for 90 min in a 1.0 O/o Os04 solution prepared with the buffer solution. After washing

samples three times for 10 min each in the buffer solution, they were dewatered for 10 min

each in serially graded solutions of ethanol at concentrations of 10, 30, 50, 70, 90, and 95 O/o.

SEM observations were conducted using a scanning electron microscope (JEOL, JSM-5310

LV, Japan).

4.3. Results and Discussion

4.3.1. Performances of the fixed-bed anammox reactors

A continuous operation of the anammox reactor applying nonwoven biomass carrier

and small size biomass carriers was carried out for a period of 174 days. During the entire

experimental period, effluent pH values were always O.5--O.7 higher than influent values.

Increase in pH indicated the occurrence of anammox reaction (Yamamoto et al., 2008). The

N03-N production rates were from O.1 to O.3 kg-N m-3 d-i in most ofperiods (Fig. 4-3, Panel

A) which expressed the growth of anammox bacteria (van de Graaf et al., 1996).

The reactor performances in terms ofnitrogen removal during the whole experimental

period is depicted in Fig. 4-3 (Panels A and B). The data in Runs I and ll show the results

for fixed-bed anammox reactor (a), and the data in Runs M and TV present the results for

fixed-bed anammox reactor (b).

In Run I , fixed-bed anammox reactor (a) was operated with a quick increase in TNLR,

from O.8 to 2.12 kg-N m-3 d-i, only in 58 days. The TNRR increased synchronously from O.7

to 1.81 kg-N m-3 d'i due to the high TN removal efficiency ranged from 83 O/o to 88 O/o.

59

tte?

fi

4.tsua

e

S6i

Eg

.so

89za

4oth

z.

Ezi.

MIii

F. .fi

4.5

4,e

:'i

].e

:..s

[•.g

l.S'

'1.C

o.s•

t}.g

1 f)o

l8e

l {i{)/

Eilf)

E:,g

1•ge

ge

tw

40

2, iO

o

.siko-pt.iI ll m N

fi

A

'

evuavaeqwhI op. I l

i ] ?• "f ;t q tizXiipE' -- i,

' ' '

l

es

Ir'

l ' rf il l i F, ,J , t, .Yv'' 'ss# l,

iMdeqrvta.

I

C''"''

"''

i,tL,,,P,,,,l'rred$'bi,,:,,tisc,[SSi"

} r 1 IV ff,' 'Y ••'!N,, ,.N •t'' . v,

11e 1[o lso 1-to !so loo 1]o leo

lcn

zz

s"

70

6"

)--ij

4e

s"

:•e

le

o

!"-se.

b8

'6ur.Et

9oee

z-

-Ai-Ztsi)

s

vg'ktsgg8

z

o lo :c 3a 4f) i.g sf" ?o s.o gf) 1•oo Time (days)

I I ll I. M ,,.,,,.,"-'lit"- - ':i Cb•"V l" -"r "-" ' l. tlS{61S EiS},liSSi{(iii,t,i.?

B t l., l'i,, tJ?/J,.illl.iliiltwIiiilli.liriiilll.lg.•illtliiiillliliiii-••.,th,.ieei,,.L71i.iiiiiliiiillliii..wt,,NvlliiltlSi,lilililiiil,...,.ili,//L.,.l"l,li

: lst itl

' ''' i. T?., ,nt tt't 's. '/' 'lt. i'"'/ {1'' 'i'S pt ,. r StS 'k : '11'- ''t' '.,' /'./l./,. '''/l'Å}'i" '}" lt {'i ""X 'X","'? ag

I " .,• •' "' w, ', ' .. ' .t

l N

100

sc

d,if

4a

2.g

o

(o)

( t••fi• •• )

Ae.

bg

'6ur

.o..ge

oEopt

O ID l.O 3't) 40 50 6" 7e ge 9e 10e 110 120 1.aa- l40 15e L60 17e 1.Re Time (days) Fig.4-3. Nitrogen removal performance of the reactors. Panel A: (tw ) TN removal efficiency;

TNLR; Cll'liiil"'•) TNRR; (i'',) N03-N production rate. Panel B: (O) infiuent TN; (-) N02-N removal

efficiency; (*) NH4-N removal efficiency; (de,) influent N02-N; (O) infiuent NH4-N; effluent TN; (ig•) effluent NH4-N; (A) effluent N02-N.

The effluent TN concentrations were stable and ranged from 20 to 30 mg-N 1'i. Both

effluents ofNH4-N and N02-N concentration were lower than 10 mg-N 1-i with the average

of 5 mg-N 1-i. This perfect TNRR yielded in only two month confirmed that the

saline-resistant anammox culture, which was acclimated to synthetic influent supplemented

with only NaCl (30 g 1-i) (Liu et al., 2008), could adapted quickly to the practical brine

wastewater with same level of salinity, though which contained many other ions (Br-, K',

Mg2' and Ca2', as shown in Table 4-1).

In Run ll , the TNLR was increased to 2.78 kg-N m-3 d-i gradually but the TNRR did

not change much and maintained at a nearly definite value. The effiuent TN concentrations

60

increased from 30 to 60 mg-N 1-i with the TN removal efficiency decreasing to 63 O/o at the

end ofthis Run, when the effluent NH4-N and N02-N concentrations increased to 15 and 40

mg-N 1-i, respectively. The reason ofthis decrease in TN removal efficiency was supposed to

be the limitation of anammox biomass and the occurrence of clogging inside ofthe attached

biomass on nonwoven biomass carrier and some bigger granules, which was indicated by the

color change of anammox biomass from brownish red to black (Fig. 4-5, day 84) in

fixed-bed anammox reactor (a). In Runs I and ll , the highest TNRR reached 1.9 kg-N m-3

d"i was obtained.

Afterwards, a fixed-bed anammox reactor (b) using small size biomass canier was

applied for obtaining higher TNRRs from day 85. The whole anammox sludge in fixed-bed

anammox reactor (a) was transferred to fixed-bed anammox reactor (b), so that the quantity

of anammox sludge was not changed too much between two anammox reactors.

In Run M, the TNLR was exponentially increased from O.89 to 2.41 kg-N m'3 d-i within

40 days, and the TNRR successfu11y increased from O.69 to 1.96 kg-N m'3 d-i during this

period. The TN removal efficiency recovered quickly to 75e-85 O/o with both effluent

NH4-N and N02-N concentrations keeping around 10 mg 1-i. Comparing to the experimental

results for fixed-bed anammox reactor (a), the TNRR for fixed-bed anammox reactor (b) was

increased more quickly. This result indicates that the anammox treatment performance was

improved by using fixed-bed anammox reactor (b). The color of anammox biomass changed

from black to brown gradually, as shown in Fig. 4-5 (dayl33).

Increasing trend in the TNRRs of Run IV was not constant owing to the following two

reasons. One is the improper balance ofpartial nitrified effluent ofthe brine wastewater, and

the other is the insufficient amount of the partial nitrified effluent for the operation of the

anammox reactor under high TNLRs. In Run IV, the maximum TNLR was increased to 3.17

kg-N m'3 d-i and the maximum TNRR reached to 2.52 kg-N m-3 d-i which was higher than

that in Run M. The TN removal efficiency was similar to that obtained in Run M. The

average N02-N removal efficiency was about 88 O/o and the effluent N02-N concentrations

were kept below 1 1 mg 1-i. The color of anammox biomass changed from brown to weak red

finally (Fig. 4-5, day 170). Owing to the lower influent TN concentration around 140 mg-N

61

1-i (Fig. 4-3, Panel B), the maximum influent fiow rate was 60 1 d-i in Run N with the

'shortest HRT of 1.12 hour.

The reaction ratio of TNRR, N02-N consumption and N03-N production rates to

NH4-N consumption rate was 1.81:1.01:O.20 for fixed-bed anammox reactor (a) (Fig. 4-4,

Panel A), and 1.95:1.15:O.20 for fixed-bed anammox reactor (b) (Fig. 4-4, Panel B). The

reaction ratios in two reactors were both slightly lower than the stoichiometric ratios of

anammox reaction as shown in Eq. 1, but the closer reaction ratio with more reliable

correlation were obtained for fixed-bed anarrmiox reactor (b). This indicates that the better

cultural conditions for anammox bacteria were provided and the more stable anammox

treatment performance was obtained in fixed-bed anammox reactor (b) by using small size

biomass caniers.

-A"'E 2' -4 ik.2 ., o.g3 ")E :-4 R"=oys.ua .?-• .{S Mua 2.fig',.i, ge Iiklueagin g',., ;i=-iS,5S,111.Z•li• geS""`5eS"'"/i,,llli,,11,i,,1-f 11i.iklil'• ,..<cc"co" "K,{-:,k,go',,;

g ,,, .. , i' Il'I - 8. i.i. ii,... ., ii, ,.i,.--•ii• iii. ••• •• •'i-i-• g.{I o.e O.4 g.f {k.6 e.7, O.8 O.9 1.0 1.1 1.2 O.r, if.-st C}.5 •O.6 O.? {S.S B.9 1.{e 1.I 1.2 l.3 1.4

NH4-N consumption rate ( kg-N mj day-i) NH4-N consumption rate ( kg-N mj day-i)

Fig.4-4. Ratios of TNRR, N02-N consumption and N03-N production rates to NH4-N consumption rate. Panel A: fixed-bed anammox reactor (a). Panel B: fixed-bed anammox reactor

(b). (op) TNRR; (O) N02-N consumption rate; (l,l,i) N03-N production rate.

Dapena-Mora et al., reported the value for N02-NfNH4-N consumption ratio and

production ratio of N03-N production and NH4-N consumption were 1.67 and O.28,

respectively, in an anammox SBR treating the partial nitrified fish canning effluents under

the salinity of 8•-JIO g NaCl 1-i at 35 OC (Dapena-Mora, 2006). These different experimental

data for anammox reaction ratios indicates that anammox reaction ratios will change

depending on the cultural conditions (substrate composition, salt concentration and

temperature etc.) and the responsible anammox bacteria.

62

4.3.2. Changes in biomass color for fixed-bed anammox reactors

Anammox biomass changed its colour depending on cultural conditions because of the

unique cytochrome content in anammox bacteria (van De Graaf et al., 1996; Ahn et al.,

2004).

day O day 84 day 95 day l03 day l15 day 124 day 133 day 148 day 170 Fig.4-5. Changes in color of anammox biomass during the whole experimental period

The initial inoculums derived from this fixed-bed anammox reactor using nonwoven

carrier by feeding synthetic wastewater with high salinity as influent presented brownish red

color (Fig. 4-5, day O). The color of anammox biomass in fixed-bed anammox reactor (a)

changed to black after using the partial nitrified brine wastewater as influent (Fig. 4-5, day

84). Yamamoto et al., also reported the color of anammox biomass changed from red to

grayish black after switching the influent from synthetic wastewater to the partial nitritation

piggery wastewater (Yamamoto et al., 2008). There are two possible reasons for the color

changes in anammox biomass in this reactor. One reason is the color component such as

fumid acid and many ions contained in the partial nitrified the brine wastewater, and the

color change accounted for the adaptation of anammox bacteria to the brine wastewater. The

other possible reason is the clogging and channeling of biofilm originated from the deeper

part of nonwoven carrier and the inner of some bigger granules, which caused the poor

diffusion of substrate to these part of biomass and consequently decreased the anammox

activity. Ifthese phenomena were kept for long time, the color ofbiofilm tumed to black.

By changing biomass carrier from nonwoven to small size biomass carriers, the color of

anammox biomass changed from black to brown in Run III (Fig. 4-5, day 133), then to weak

red finally (Fig. 4-5, day 170). One reason for this color change is the dissolving the

limitation factor in increasing anammox activity for fixed-bed anammox reactor (a), by the

better living microenvironment of anammox bacteria provided by small size biomass carriers.

63

Another possible reason is the special function of the small size biomass carriers containing

some nanophase materials, which surrounding and enwinding the anammox bacteria could

prevent or relieve the permeation of the color components and ions in the brine infiuent to

anarnmox cells. This assumption should be verified by further study.

In this study, the color of anammox biomass changed from brownish red to black in

fixed-bed anammox reactor (a), and successfu11y rebounded to weak red with high anammox

activity in fixed-bed anammox reactor (b) using small size biomass carriers. But the detailed

mechanism ofcolor change in anammox biomass could not be verified.

4.3.3. Small size biomass carriers for enhancing anammox performance

For the fixed-bed anammox reactor (a), the TNRRs more than 1.9 kg-N m-3 d-i could

not be achieved owing to the limitation of elevating anammox biomass and the decrease in

anammox activity caused by clogging and channeling phenomena. Even though the biomass

carrying capacity of nonwoven carrier was very high (Furukawa, 2003), it was difficult to

increase the amount of anammox biomass in the reactor after reaching maximum value of

sludge retaining capacity of nonwoven carrier. In order to increase the total amount of

anammox biomass in the reactor and reactivate anammox activity, nonwoven biomass canier

was replaced with small size biomass carriers.

The different spatial distributions ofbiomass between in fixed-bed anammox reactor (a)

and in reactor (b) can be distinguished from Fig. 4-6. Anammox biomass mainly attached on

the chrysanthemum shape nonwoven canier in fixed-bed anammox reactor (a), which only

occupied small proportion of the available volume of the reactor. On the contrary, anammox

biomass distributed evenly over the whole space ofthe fixed-bed anammox reactor (b).

Fig. 4-6 shows the microscopic observation ofbiomass attached on small size carriers

on day 148. The brown, brownish red, red anammox biomass attached on hydrated white

lime (Fig. 4-6, a), nonwoven filaments (Fig. 4-6, b) and mixture of small size carriers (Fig.

4-6, c and d). These attachments confirmed the even distribution of anammox biomass in the

whole reactor, deprived from the initial well mixture ofthe inoculums and small size carriers

64

and the continuously homogeneous disturbance of the whole caniers imposed by the

produced micro bubbles from anammox reaction, which were observed at the top ofreactor

all the time.

Fig.4-6. Microscopic observations of anammox biomass attached on small size caniers on day 148: (a)

biomass attached on hydrated lime, (b) biomass attached on nonwoven filaments, (c, d) biomass

attached on the mixture of small size caniers such as cement, lime and nonwoven powders.

The hardcore-1 was a significant component in the small size caniers. After the surface

modification of the small size crystal mixture (hydrated cement, slag and lime) by the

enwrapping ofthis nanophase material, the smooth surfaces and edge angles ofit changed to

coarse surface with passivity angles (Fig. 4-7, a and b). The spatially wrinkled structure with

many needle fibers and micro channels produced after surface modification was in favorable

of the bacterial immobilization and substrate transfers, which consequently resulted in the

high anammox activity. By the surface modification, the whole of small size biomass caniers

presented a very evenly dispersed status rather than irregular aggregations. No clogging and

channeling phenomena were observed during the whole period of fixed-bed anammox

reactor (b). Moreover, the surrounding of nanophase material on the small mixture could

prevent the exudation ofsome heavy metal ions which was supposed to be toxic to biomass.

SEM images of C and D (Fig.4-7) visualize the biomass adhered the modified small

size carriers. Comparing with the microscopic photos in Fig. 4--6 (b), anammox biomass

attached to inorganic small carriers prior to nonwoven filaments. The nonwoven filaments

performed the main function of connecting these inorganic carriers and transfening the

substrates due to its hydrophilicity. Because of the high mean density of the small size

biomass caniers

Attached by anammox biomass tightly, there was almost no any washout of anammox

sludge in Runs III and IV. The MLVSS concentration of anammox biomass (excluding

nonwoven filaments) in fixed-bed anammox (b) was 6,265 mg 1-i on day 148. Dapena-Mora

et al., reported a biomass concentration between 1,OOO and 2,700 mg-VSS 1'i in a similar

65

study (Dapena-Mora et al., 2006).

concentration was much higher. This

small size biomass caniers.

In contrast to this result, our obtained biomass

result verified the excellent sludge retaining ability of

•ts

gtw"

pu

Fig.4-7. SEM images of small size biomass carriers and attached biomass. (A) Hydrated

small carriers without surface modification by hardcore-1, without biomass attachnent; (B)

hydrated small carriers after surface modification by hardcore-1 without biomass

attachment; (C, D) biomass attachment on small size biomass carriers at different

magnifications on day 150.

4.3.4. Comparison of anammox treatment capabilities under high salinity condition

Table 4-3 shows the comparison of our obtained results with other results obtained from

anammox treatment under high salinity concentration.

Our anammox reactors were operated under unfavorable operational conditions, i.e.,

salinity concentration of 30 g NaCl 1-i, practical industrial influent, moderately low nutrient

concentrations, room temperature, no removal of influent DO and short HRT. Two

fixed-bed anammox reactors in this study yielded higher TNRRs than that in others reports.

66

Table 4-3 Comparison of TNRR for different condition anammox reactors under high salini condtion

Process ReactorSalinity(g 1-i)

TNRR( kg-Nm-3d-i)

Substrate Reference

Anammox

OLAND

Anammox

Anammox

Anammox

Anammox

SBR

RBC

SBR

Fixed-bed

Fixed-bed (a)b

Fixed-bed (b)C

10

30

30

30

30

30

O.34

O.61

1.oa

1.7

1.9

2.52

Fish-canning

wastewaterSynthetic

Synthetic

Synthetic

Brine

wastewaterBrine

wastewater

Dapena Moraet al.,(2006)

Windey et al.,(2005)Kartal et al.,(2006)

Liu et al.,(2009)

This study

This study

aThis is the TNLR. In all other cases, the TNRR is presented.

bFixed-bed anammox reactor using nonwoven biomass canier.

CFixed-bed anammox reactor using small size biomass carriers.

Especially, the highest TNRR of 2.52 kg-N m-3 d'i obtained in fixed-bed anammox

reactor (b) using small size biomass carriers was several times higher than the reported

TNRRs by other researchers. Moreover, the limitation of the TNRR in fixed-bed anammox

reactor (b) was caused by the shortage of the influent. Without this limitation, much high

TNRR might be obtained. The biomass concentration in fixed-bed anammox (b) was reached

at 6,265 mg 1-i MLVSS on day 148, and the specific anammox activity was calculated to be

320 mg-N g-MLVSS'i d-i on this day. This high biomass concentrations and high anammox

activity contributed to the excellent treatment result and indicated the advantages of the

application of small size biomass carriers to fixed-bed anammox reactors. By using this

small size of biomass carriers, we were able to obtain stable anammox treatment

performances without clogging and channeling phenomena which were the disadvantages for

fixed-bed anammox reactor using nonwoven biomass carriers.

4.4. Conclusions

During a period of 174 days, we have succeeded in the establishment of anammox

67

reactors treating the partial nitrification brine wastewater, whose salinity was 30 g 1-i, at

about 25 OC. The maximum of TNRR for the fixed-bed reactor using nonwoven biomass

'carrier reached to 1.9 kg-N m-3 d]i, but could not increase more than this value owing to the

clogging, channeling and limitation of retaining biomass in the reactor. The TNRR for the

fixed-bed anammox reactor using small size biomass carrier reached to 2.52 kg-N m-3 d-i and

stable anammox operation was achieved in this reactor. The color ofbiomass in the fixed-bed

anammox reactor using small size caniers turned to weak red from black color finally. For

completely understanding the excellent performance in this fixed-bed anammox reactor

using small size biomass caniers, microbiological community analyses and the batch activity

evaluations of anammox bacteria should be perfomied in future study.

4.5 References

Ahn, Y.H., Hwang, I.S., Min, K.S.: ANAMMOX and panial denitritation in anaerobic

nitrogen removal from piggery waste, Mater Sci. Technol. 49, 145-153(2004).

APHA Standard Methods for the Examination of Water and Wastewater.: American Public ' Health Association, Baltimore, MD. (1995).

Chen, M., Canelli, E., Fush, G.W.: Effect of salinity on nitrification in East River, J. Mater

Pollut. ControlFed. 42, 2474-2481(1971).

Dapena-Mora, A., Campos, J.L., Mosquera-Corral, A. and Mendez, R.: Anammox process

for nitrogen removal from anaerobically digested fish canning effluents, nlater Sci.

Technol. 53 (12), 265-274(2006).

Dincer, A.R., Kargi, F.: Salt inhibition of nitrification and denitrification in saline wastewater,

Environ. Technol., 20 (11), 1147-1153(1999).

Fujii, T., Sugino, H., Rouse, J.D., Furukawa, K.: Characterization of the Microbial

Community in an Anaerobic Ammonium-Oxidizing Biofilm Cultured on a Nonwoven

Biomass Carrier, 1. Biosci. Bioeng., 94 (5), 412-418(2002).

Furukawa, K., Rouse, J.D., Yoshida, N., Hatanaka, H.: Mass cultivation of anaerobic

ammonium-oxidizing sludge using a novel nonwoven biomass canier, J. Chem. Eng.

Jpn., 36, 1163-1169(2003).

68

Jetten, M.S.M., New pathways for ammonia conversion in soil and aquatic systems. Plant

Soil, 230, 9-l9(2001).

Kanda, J.: Determination of ammonium in seawater based on the indophenol reaction with

o-phenylphenol (OPP), MaterRes., 29, 2746-2750(1995).

Kartal, B., Koleva, M., Arsov, R., van der Star, W., Jetten, M.S.M., and Strous, M.: Adaption

of a freshwater anammox population to high salinity wastewater, J. Biotechnol., 126,

546-553(2006).

Liu, C., Yamamoto, T., Nishiyama, T., Fujii, T., and Furukawa, K.: Effect of Salt

Concentration in Anammox Treatment Using Non-woven Biomass Carrier. J. Biosci.

Bioeng., 107 (5) (2009) (in press)

Liu, S., Yang, F., Xue,Y., Gong Z., Chen H., Wang, T., Su, Z.: Evaluation of oxygen

adaptation and identification of functional bacteria composition for anammox

consonium in non-woven biological rotating contactor, Bioresoun Technol. 99,

8273-8279(2008).

Mosquera-Corral, A., Gonzalez, F., Campos, J.L., M6ndez, R.: Partial nitrification in a

SHARON reactor in the presence of salts and organic carbon compounds, Process

Biochem., 40, 3109-3 118(2005).

Nakajima, J., Sakl<a, M., Kimura, T., Furukawa, K., and Sakka, K.: Enrichment ofanammox

bacteria from marine enviroument for the construction of a bioremediation reactor,

Appl. Microbiol. Biotechnol., 77, 1159-1166(2008).

Op Den Camp, H.J.M., Kartal, B., Guven, D., van Niftrik, L.A.M.P., Haaijer, S.C.M., Van

Der Star, W.R.L., Van De PasSchoonen, K.T., Cabezas, A., Ying, Z., Schmid, M.C.,

Kuypers, M.M.M., Van De Vossenberg, J., Harhangi, H.R., Picioreanu, C., Van

Loosdrecht, M.C.M., Kuenen, J.G., Strous, M., Jetten, M.S.M.: Global impact and

application of the anaerobic ammonium-oxidizing (anammox) bacteria, Biochem. Soc.

Trans., 34 (1), l74-178(2006).

Panswad, T., Anan, C.: Specific oxygen, ammonia, and nitrate uptake rates of a biological

nutrient removal process treating elevated salinity wastewater, Bioresoun Technol., 70

(3), 237-243(1999).

Qiao, S., Kawakubo, Y., Cheng, Y., Njshiyama, T., Fujii, T., Furukawa, K.: Identification of

69

bacteria coexisting with anammox bacteria in an upflow column type reactor,

Biodegradation (in press) (2008).

Strous, M., Heijnen, J.J., Kuenen, J.G., Jetten, M.S.M.: The sequencing batch reactor as a

powerfu1 tool for the study of slowly growing anaerobic ammonium-oxidizing

microorganisms, AppL Microbiol. Biotechnol., 50, 589-596(1998).

Uygur A.: Specific nutrient removal rates in saline wastewater treatment using sequencing

batch reactor, Process Biochem., 41 (1), 61-66(2006).

van de Graaf, A.A., de Bruijn, P., Robertson, L.A., Jetten, M.S.M., Kuenen, J.G.: Autotrophic

growth of anaerobic ammonium-oxidizing microorganisms in a fluidized bed reactor,

Microbiology, 142 (8), 2i87-2196(1996).

van Dongen, U., Jetten, M.S.M. and Loosdrecht, M.C.M.: The SHARON@-Anammox@

process for treatment of ammonium rich wastewater, Mater Sci. Technol., 44(1),

153-160(2001).

Windey, K., De Bo, I., Verstraete, W.: Oxygen•-limited autotrophic

nitrification-denitrification (OLAND) in a rotating biological contactor treating

high-salinitywastewater, MaterRes., 39, 4512-4520(2005).

Yamamoto, T., Takaki, K., Koyama, T., Furukawa, K.: Long-term stability of partial

nitritation of swine wastewater digester liquor and its subsequent treatment by

Anammox, Bioresouz Technol., 99, 6419-6425(2008).

70

Chapter 5

Study on Long-term Stable Operation of High-rate Fixed-bed

Anammox Reactor Using Different Biomass Carriers at Moderately Low Temperature

5.1 Intreduction

Anaerobic ammonium oxidation (anammox) process was first experimentally

demonstrated and documented, in a denitrifying pilot plant at Gist-Brocades, Delft, the

Netherlands in 1995 (Mulder et al., 1995). The anammox reaction converts ammonium to

dinitrogen gas with nitrite as electron acceptor, catalyzed by the planctomycete-like bacteria

involving hydrazine as an intermediate. These bacteria have a unique prokaryotic organelle

(anammoxosome) surrounded by ladderane lipids, which exclusively contains the hydrazine

oxidoreductase as the major protein to combine nitrite and ammonia in a one-to-one fashion

(Jetten et al., 2005). Compared with the traditional biological nitrogen removal process such

as nitrification-denitrification process, anammox process is now regarded to be a novel,

promising, and cost effective alternative, which has many advantages, e.g., no requirement of

external carbon sources, low oxygen demand, minimized excess sludge production and

reduction in C02 emissions (Op Den Camp et al., 2006; Liu et al., 2008). In recent years,

there has been an enormous increase in research efforts dedicated to the anammox process,

and a few fu11-scale treatment plants using anammox process have been established (Abma et

al., 2006; van der Star et al., 2007). However, the stringent operation conditions and

extremely slow growth rate of the anammox bacteria has restricted the application of

anammox process (Strous et al., 1999).

One of the stringent conditions is the high optimal temperature for the maximum

anammox activity, which was between 30-35 OC, 37 OC and 35-40 OC in different reports

(Yang et al., 2006; Isaka et al., 2008; Dosta et al., 2008). So most of the anammox studies

were normally carried out at temperature high than 30 OC (Strous et al, 1997; Furukawa et al.,

71

2003; Tsushima et al., 2007). However, very recently, several researches focused on the

effects of moderately low temperatures on the stability of this process (Dosta et al., 2008;

Isaka et al., 2008). Their results indicated that the application of anammox process could not

be restricted to effluents with temperatures around 30 OC. Nitrogen removal activity was

observed even at 6 OC, but gradually decreased with decrease in temperature. The apparent

activation energy was calculated as 93 kJ mol-i and 33 kJ mol"i at temperature range from 22

to 28 OC, and from 28 to 37 OC respectively (Isaka et al., 2008). Moreover, at 20-22 OC, Isaka

et al., achieved stable nitrogen conversion rate of 2.3 kg-N m-3 d-i in 320 days, and high

nitrogen conversion rate of 8.1 kg-N m-3 d-i was reported but this activity was maintained on

this value only for 15 days.

Another drawback of anammox process is the requirement of a long start-up period due

to mainly extremely slow growth rates ofanammox bacteria. The doubling time of anammox

bacteria was reported to be approximately 11 days (Strous et al., 1998), and 9 days under

optimal conditions (Strous et al., 1999). Therefore, the immobilization of anammox bacteria

would be attractive for faster start-up and obtaining high nitrogen removal performances

(Isaka et al., 2007). Up to date, the highest nitrogen removal rate reported in the literature

was 26.0 kg-N m-3 d-i, obtained in a high-rate anammox biofilm reactors using the up-flow

fixed-bed biofilm column reactor within 250 days (Tsushima et al., 2007). But this reactor

was operated at high temperature (37 OC), and such so high removal rate was only challenged

once and decreased quickly. Recently, fixed-bed reactors using nonwoven fabric carriers

became attractive for stably attached immobilization of anammox sludge (Funkawa et al.,

2003; Isaka et al., 2007; Tsushima et al., 2007).

There are many industrial wastewaters containing high concentration of NH4-N with

less organic rnaterial, such as wastewaters coming from fertilizer industry, explosive industry

or some pharmaceutical processes (Wiesmann, 1994) and developers used in

photolithography contain tetramethyl ammonium hydroxide (TMAH). Anammox process

has the great potential for the treatment of these industrial wastewaters. In case of TMAH

containing wastewater, TMAH could be completely degraded to NH4-N through

ammonification (Chang et al., 2008). After proper partial nitritation treatment, such as

Single-reactor high-activity ammonium removal over nitrite (Sharon) process, of these

72

wastewaters, partially nitrified effluent is fed to anammox process. This combined PN and

anammox process has an advantage of cost-saving and environnient friendliness over

physico-chemical method, such as catalytic oxidation process for treating this wastewater

containing TMAH (Hirana et al., 2001). The total nitrogen (TN) concentration of wastewater

containing TNAH from some semi-conductor plants is about 150 mg-N 1-i. Therefore, in the

present experiment, influent TN concentration was set to this value, which is lower than that

used in most studies, especially in high rate anammox reactors, such as approximately 330

and 900 mg-N 1-i in the literatures (Isaka et al., 2007; Tsushima et al., 2007).

In this study, a fixed-bed anammox reactor using different configurations of nonwoven

biomass canier and small size biomass carriers was developed and evaluated for pursuiting

the long-term stability of high nitrogen removal performance using low strength wastewater

under moderately low temperature (19-25 OC).

5.2 Materials and Methods

52.1. Anammoxreactorf

The schematic diagram of fixed-bed anammox reactor used

in this experiment was shown in Fig. 5-1. This reactor was made

up of glass and the size is rp 95 mm x 423 mm, with the available

volume of 2.8 1. Influent was supplied to the reactor by up-flow

mode. The reactor temperature was maintained to about 20 OC by

heating through heater coil outside when the temperature was

lower than 20 OC and kept at room temperature in other days. The

reactor was sheltered from light by black vinyl sheet enclosure.

5.2.2. Nowoven biomass carriers and small size biomass carriers

Nvwy

ar4

'

Fig. 5-1. Fixed-bed reactor

Nonwoven polyester coated with a pyridinium-type polymer (Japan Vilene, US patent

5,185,415; 1993) was used as biomass carrier for fixed-bed anammox reactor. The original

73

strips usmg originalsplitted chrysanthemum shape nonwoven

caniers (100 mm diameter, 5 and 2.5 mm

thickness for one trip, 395 mm length, as

shown in Fig. 5-2) were mounted into

fixed-bed anammox reactor sequentially.

Afterwards, in order to further

immobilize more biomass in the gap

between thin strips or nonwoven, 30

size of 1mm length and 50 pm

In the last period of this experiment,

nonwoven biomass caniers and packed in

nonwoven (shown in Fig.5-7). There are

The small size biomass caniers consisted

blast-furnace slag), and hardcore-1 (War

opened). Nonwoven small filaments were

was neutralized by soaking it in running

were from 1

self-made special high active agent,

cement : nonwoven small filaments :

water. This prepared mixed slurry was

chrysanthemum shape nonwoven carrier has 8 strips, which was used in the first period of

experiment. Then, for the purpose of increasing specific surface area for biomass attachment

and nutrients diffusion rate ofnonwoven biomass canier, 8 strips ofnonwoven was split into

16 ' ' razor. The '' and

Fig. 5-•2 Original and split nonwoven biomass

carriers (top view)

g of undoing nonwoven small filaments with a mean

diameter were added in to the reactor.

small size biomass carriers were used instead of

a kind of small carrier bed made of plastic and

11 layers of beds with 3.6cm height in reactor.

Every bed size was 7.0cmÅ~6.3cm and net made of nonwoven srips was 2cmÅ~2cmÅ~3cm.

of nonwoven small filaments, cement (Portland

industry of military department product without

the same as that above mentioned. The cement

tap water for 2 weeks, whose original diameters

to 5mm, was sieved to obtain a mean diameter of 1 mm. The hardcore-1 was a

which was a powerfu1 nanophase material. These above

mentioned small materials (270 g dry weight) were mixed together (the dry weight ratio of

hardcore-1 was 200:60:10) and put into 1.0 L of tap

named as small size biomass carriers, which was

packed homogeneously into a self-made small canier bed made ofplastic after mixing well

with anammox inoculums collected from the total biomass in the previous period. The small

carrier bed consisted of 1 1 units with the same size (tp 94.5 mm x 36 mm). The cylindrical

unit was hollow and of no cover, with a square open (30 mm x 5 mm) near the circle and

74

many small open holes (tp 2 mm) on the floor. These units packed with mixed small size

biomass carriers and seed sludge were mounted into the reactor sequentially with a stagger

arrangement of the square opens on adjacent ones. The unique structure of the small carrier

bed could produce a main horizontal baffling and subsidiary vertical steam of the current in

the reactor, consequentially provide a perfect niche for anammox consortium with sufficient

contact with subtract.

5.2.3. Inoculumandsyntheticinfluent

The seed anammox sludge was taken from a laboratory scale anammox up-flow column

reactor (Funkawa et al., 2003) and seeded with of 500 mg 1-i. In order to improve the

attachment of anammox sludge on the nonwoven biomass carrier, granular anammox sludge

was crushed into small particles using O.5 mm sieve.

After the pretreated crushed anammox sludge WaS Table s-1

filmed evenly onto nonwoven strips, the reactor was Compositionofsyntheticwastewater

purged with nitrogen gas for 20 min. Then, internal

circulation was carried out for 12 h. Through this

procedure, seed anammox sludge was evenly

attached-immobilized on the nonwoven biomass

carrler.

Synthetic influent was prepared by adding NH4-N and N02-N in forms of (NH4)2S04

and NaN02, with other nutrients and buffer according to the composition given in Table 1.

The pH of the synthetic influent was 7Å}O.2 all the time. The reactor was operated under

N02-N limiting condition by feeding equal concentration ofNH4-N and N02-N, both about

75 mg-N 1-i during the whole experimental period, except the lower concentrations with

same ratio in the first 50 days.

5.2.4. Start-up and operational conditions of fixed-bed anammox reactor

This reactor was started-up with the initial TNLR of O.05 kg-N m-3 d-i with a hydraulic

75

Components Concentration

(NH4)2S04 75Å}5(mg-N1-i)

NaN02 75Å}5(mg-N1-i)

EDTA 5.0(mg1"')

FeS04.7H20 9.0(mg1-')

KHC03 125.1(mg1-')

KH2P04 54.4(mg1-')

retention time (HRT) of 26.7 hr. The experiment was divided into five Runs according to the

usage of different biomass carriers.

In Run I (from day O to 206), the original column-like nonwoven biomass carrier was

used. From day 207 to 234, efforts for recovering the activity of anammox bacteria were

devoted with fluctuant results (data not shown). Then in Run ll (day 235 to 365), the reactor

was operated using splitted chrysanthemum shape biomass carrier. Subsequently, some black

color anammox sludge was removed from the reactor and the TNLR was decreased initially

in Run M (day 366 to 404). Afterwards, nonwoven small filaments were added in to the

reactor in Run N (day 405-518). At last, in Run V (day 519-898), small size biomass

carriers packed in the small carrier bed were used instead of nonwoven biomass caniers.

Except Run M, the anammox biomass in the previous Run was all collected and packed into

the reactor in the next Run. The operational conditions during the whole experiment were

shown in Table 5-2.

In the first 50 days of operation, the TNLR was increased stepwise by O.05 to O.1 kg-N

m-3 d-i every step. After day 50, the TNLR was increased only by increasing flow rate (or

reducing HRT), and the protocol for increasing in TNLR was increasing TNLR stepwise by

O.18-O.2 kg-N m-3 d-i each time in case ofthe effluent NOi -N concentration below 20 mg-N

ri most oftime.

Dissolved oxygen (DO) is toxic to anammox consortium, because anammox activity

can exert only under strict anoxic conditions (Jetten, 2001). In Run I, influent DO

concentrations were reduced to O.5-1 mg IMi by flushing nitrogen gas prior to supplying

influent to the reactor. However, some coexisted oxygen consuming microbial community,

such as aerobic nitrifiers and floc forming heterotrophs, could protect the anammox bacteria

from the effect of oxygen (Fujii et. al. 2002; Qiao et al., 2008; Liu et al., 2008 b). Therefore,

after Run I , influent wastewaters with DO concentration of 5-7 mg 1-i were supplied

directly to the anammox reactor for evaluation of effect of DO on anammox treatment

performances.

76

5.2.5 . Analyticalmethods

Table 5-2 Operational conditions during th ewhole experimental period

Parameters

Time range (day)

Biomass carriers in

reactor

HRT (hr)

Inf. NH4-N( mg-N 1"')

Inf. N02-N( mg-N 1-i)

Influent TN( mg-N

1-i)

TNLR ( kg-N m-3 d"i)

Temperature (OC)

infiuent DO (mg IHi)

Run I Run ll Run M Run IV Run V

O-206

CNBca

26.67-2.32

27.7-79.2

26.8-79.0

54.5-158.9

O.O5--1.63

19-25

O.5-1

235-365

Split CNBC

2.22-1.01

71.0-79.1

71.0-81.5

i412-160.6

158-3.72

19-25

5-7

366-404

Split CNBC

1.77-1A2

702-75.0

71.0-76.4

141.6-151.4

1.9S-2.56

19-25

5-7

405-518

Split CNBC& NSFb

4.67-O.56

72.5-76.5

76.3-81.0

150.1-153.1

O.78-6.56

19-25

5-7

519-898

SSBcc din SCB

1.22-O.32

74.9--83.5

73.3-84.2

149.9-162.1

2.98-12.48

19-25

5-7

CNBC: column-like nonwoven biomass carrier b NSF: nonwoven small filaments

C SSBC: small size biomass canier

dscB: small canier bed

The concentrations ofN02-N and N03-N were measured by the colorimetric method in

accordance with the Standard Method (APHA, l995). NH4-N concentration was measured

colorimetrically by the phonate method using ortho-phenylphenol as a substitute for liquid

phenol (Kanda, 1995). DO concentration of the influent was measured by using a digital,

portable DO meter (D-55, Horiba). The concentration of mixed liquor volatile suspended

solids (MLVSS) of sludge samples was determined in accordance with Standard Methods

(APHA, 1995). PH value was determined by using a pH meter (IM-22P, TOA EIectronics).

Biomass attached on small biomass carriers was observed by an electron microscope (Nikon

Eclipse E600, Japan) and a digital camera (Nikon 4500, Japan).

Scanning electron microscopy (SEM) was used for the observation of small carriers

with attached biomass. The procedure was as follows. Samples were first washed in a O.1 M

phosphate buffer solution (pH 7.4) for 5 min, and then hardened for 90 min in a 2.5 O/o

77

glutaraldehyde solution prepared with the buffer solution prepared with the buffer solution.

Next, samples were washed in the buffer solution three times for 1O min once and then fixed

for 90 min in a 1.0 O/o Os04 solution prepared with the buffer solution. After washing

samples three times for 10 min each in the buffer solution, they were dewatered for 10 min

each in serially graded solutions of ethanol at concentrations of 10, 30, 50, 70, 90, and 95 O/o.

SEM observations were conducted using a scanning electron microscope (JEOL, JSM-5310

LV, Japan).

5.3. ResultsandDiscussion

5.3.1 Anammox reactor performances

An operation of the fixed-bed anammox reactor using different configurations of

nonwoven biomass carrier and small size biomass carriers was sequentially canied out for

900 days. During the entire experiment period, the N03-N production rate was from O.1 to

1.2 kg-N m"3 d-i in most of periods (Fig. 5-3, Panel A) which expressed the growth of

anammox bacteria (van de Graaf et al., 1996).

The TN removal performances in terms of nitrogen removal during the whole

experimental period is depicted in Fig. 5-3 (Panels A and B). Fig. 5-4 shows the NH4-N and

N02-N removal performances in Panel A and Panel B, respectively.

In Run I , the TNLR of the fixed-bed anammox reactor increased from O.05 to 1.63

kg-N m-3 d'i gradually. The TNRR increased synchronously from O.04 to 1.3 kg-N m-3 d-i

with the TN removal efficiency ranged from 63 O/o to 85 O/o. The effluent TN concentrations

ranged from 20 to 50 mg-N 1-i with fluctuations resulting from the intermittent increase of

TNLR. Effluent NH4-N and N02-N concentrations were below 25 and 19 mg-N 1"i

respectively, also with fluctuations. The establishment of fixed-bed anammox reactor using

low strength influent nitrogen was proved possible under moderately low temperature (19-25

Oc) from this result.

However, clogging and channeling phenomena of the anammox biofilm occurred after

200 days of long term operation.The black color of anammox biofilm was observed in deep

78

part of nonwoven biomass canier, but the outside anammox biofilm was still in red color, as

shown in Fig. 5-6 (day 145) and Fig.5-7 (day 200). The black anammox biofilm was

supposed to be cased by the insufficient substrate supply under high sludge concentration.

Therefore, from day 207 to 234, some efforts were devoted to recover the activity of

anammox sludge, such as removing the black sludge from the reactor, increasing to 350C

IS

"Ai

cV.ie

zrtsua

d

vee

G•e-

8ga4o-z.st

&Mfi

&

l2

11

IC)

9

s

7,

6

;.

4

3

L7

1

o

)• Ai i i i I l I I i l i' ` k--

I #lii liij lli jli ilj iil iil iii ili iii iil lill hrlll, ,im " -'-'e'

}

R v i '=:-;---=ft-"t" l .--L l .z--V"'"-" l a- :t l r- l -:- l --s 1- m.4 : ,. ----i .s ',''

:- lf L-, . I•.,----

.ii

..i'i F--

l' ,

:-.-::---

ro

2.7t

-":4

'2l

lg

15

1:,.

9

e

]

o

{e #{.o 1ee lse 2•ts] 2•-{o ,3{wa lso 4•oo 4-{.e )".{m sse pt{y) eso ,so ?tro sac} s.s',e g[>m

Time (days)

g-pt

=

I,'-sy g•"'''i"SIBIIkl,l 11

l

","l R v i

:, za i j i i i i i

\ aj

t 'atl

iilii

llii}ii

i

::;l

;I

l}

•v: s r'st

giiiii

i

,,,

,g i

l i i fi siggil /

' i

}ll}

s

i

eti

I

{

lsg

lao

Eee

120

l{X]

3ij

6t"j

4Cl

2• .O

o

F'o

g8

•g

urrSt

gz&>

-;

zer

u'-o.

e-

88g

iii

1. s•o lo•as !iio :•.e{} 2fi,so 3ivg 3io gen 4ie iee ssas 6ijo 6sij 7,pt] 7,-sio ggo gsff g{)o

Time (days)

Fig.5-3. TN removal performance of the reactors. Panel A: (-) TNLR; (ua) TNRR; (A) N03-N production rate. Panel B: (Å~) infiuent TN; ( .th ) TN removal efficiency; (D) effluent TN; (-) HRT.

suddenly and returning to 20 OC after 12 hr, inducing internal recycle for one day, changing

influent concentrations, etc. The combination of these methods retumed the black part of the

biofilm to red color for certain period.

i9

In Run ll , by switching the original chrysanthemum shape nonwoven biomass to split

one, the TNRR reached 2.51 kg-N m'3 d-i with a short HRT of 1.01 hr. The effluent TN

concentrations ranged from 22 to 40 mg 1-i before day 340, but even increased to 80 mg-N 1-i

on day 350 with a low TN removal efficiency of only 50 O/o. A few fioating anammox

granules, as shown in Fig. 5-7 (day 350), were washed out from the reactor and the biofilm

inside the nonwoven biomass carrier turned to black as same as in Run 1 (Fig. 5-7, day 365).

Ieo 1ias

p gn,

vrp ?,8zela 'pt

MUsEg aj'E

a3:g L,,o 1

o

I

s / g if

A

l -.Aili,, .•. il.f' Etl{;$illljet '• l,.'l•

l!ll •iiii [i /II llIl/ .,II[ Iii

m••/ e/ .• t.I tt z' ... t ilIl F.

rv i

ig

v

'

,

rf,juigig"o

ti

l•

ifa

twasgtwaj

ltl

t

'

{

iggab

buGnfaj

1{

9

ttgCle 74,.tsua

d

UfN

g4'g

81:g :,O 1 o

e ie 1fio sg :,ttse :•so 3'tza sso 4/uo 4se s.go ss" "off 6su ?tva 7,s.o gew gss gf?oTime (days)

9{}

g#

?c

6e

50

`if)

30

2,g

IO

fi

bg

'6utgij

rSt

ge

g&>4ge

u•e-

g.

g88

Z-=Z

I

1

ii

i

'

i

!

ii!',l}iii

! ll-ll ill!,." rv{

vkof,ec&twee

?e"i

..

B

ii

lafi)pa-8ij

i

!

l

Iiiiiiiiii

•4sl

4tiIZt

wt

Akst.i?,llillik,,gZftk,l$.fS',,

Al l,

-

-/ f's ...tf'i i.elj1.' ,,g••,.. 'iza s

g'"x,

s••

's

u.pmt.)

ll

j'

' ' j

!t

iilll

l

l

ii

iIlill

1",k'"SA

s-

lil}li

'' i

l

k

J/,rt•s,, .,.

tt1/

i"•v. .> .

EI

".

.<.. )

..

Vtk .

1,

Etsg)

ttst- )/

l

g•E'

•Rgajnjut}iin.,

•-

E,eatt?i..

•i" /t/'ttl•i?g-. 'ts',sSts•I-.te. ..."-.

1tw

sc

go

7Cj

tw

S•iL)

40

rso

2•iLl

IO

e

bg

'6.ac

rSl

gg&>

-:

4gu•e-

-di-•

Eu8

4OCN

z

" S.O likO lie 2.(su ].SO Z{"S 35ft- 4eif) 45.0 iffO 5.i,e fitt4 SEig Tee 7JhPij Sts] g5.0 •eet}

Time (days)Fig.5-4. NH4-N and N02-N removal performance of the reactors. Panel A: (A) NH4-N removalefficiency; (me) influent NH4-N; (D) effluent NH4-N; ( >< ) NH4-N consumption rate. Panel B: ( de..pt"• ) N02-N

removal efficiency; (A) influent N02-N; (O) effluent N02-N; (me/) N02-N consumption rate.

The reason for this deteriorative reactor performances were supposed to be the clogging

and channeling of the biofilm under short HRT. From day 250, the influent was supplied

directly to the reactor without nitrogen gas flushing. However, the TN removal performance

80

was stable before day 340, which elucidated that the biomass in this reactor could adapt the

abrupt increase in infiuent DO in a short time. Compared with the report by Liu et al., (Liu et

al., 2008) with requirement of step-wise increase in influent DO concentration the quick

adaptation of anammox biomass to DO observed in this experiment was possibly attributed

to the biodiversity of anammox bacteria in our system.

For recovery ofthe anammox activity, in Run M, 6.5 g (dry weight) ofblack anammox

biomass was removed from the reactor on day 370. The TNLR was reduced first then

increased gradually to 2.56 kg-N m-3 d-i. The TN removal efficiency increased to 80 O/o with

the NH4-N and N02-N concentrations in the effluent lower than 15 and 10 mg-N 1-i

respectively on day 404. Just on this day, reactor temperature was increased accidentally to

500C owing to failure in temperature control at 200C in winter season. The color of outside

anammox sludge turned to white color and lost most of its anammox activity at one night.

This died anammox sludge was removed from the reactor and restarted experiment under

low TNLR in Run IV.

In Run IV, the nonwoven small filaments (30 g) were added in to the reactor for

increasing attached-immobilized anammox sludge. The quick recovery of anammox activity

was indicated by the quick growth of anammox biofilm on nonwoven small filaments which

was almost fu11 of the reactor and the change in color from white to distinct red color (Fig.

5-6, day 430). The TNLR was

increased from O.78 to 6.56 kg-N

m-3 dmi during 113 days of

operation with a synchronously

increase of TNRR from O.46 to

4.48 kg-N m-3 d-i. The shortest

HRT in this period was O.56 hr.

No clogging and channeling of

biofilm occurred in this Run.

Compared with the previous

Runs, the high TNLR (or the low

rA-V'E

4yx5'

Ng

-o

g9-g

.tt-

E:go

21

IO

9

.g

7

6

5.

4

3

2

1

g

i{lx'

}; =2.05 •x

tsR-= g.99g

k g ..- --{' xz cb

.' ` . ..'

t

A geEi

v= 1.2 9x '1 RL = O.9S •4 b. ..Sts .+, .f:itt-g•'.•"•pt.

t' re "' '

y = oi .L) rt sx- .:

RJ- = O.963

a

"

mv

g.ij O.5 t.ij l.5 2,g L7.5. Ll.fi 3.5 4.0 4.5 tr,.e 5..i NH4-N consumption rate ( kg-N m3 d-i)

Fig. 5-5. Ratios of TNRR, N02-N consumption and N03-N

production rate to NH4-N consumption rate. (A) TNRR; (])

N02-N consumption rate; (-) N03-N production rate.

81

HRT) prevented substrate transport limitation in the biofilms throughout the reactor

(Nicolella et al., 2005). However, all effluent nitrogen concentrations were fluctuanting (Fig

5-3 and Fig. 5-4), which suggested a better operation strategy should be considered in the

next period. The measured MLSS concentration was reached to 18,890 mg 1"i on day 518. '

' '

In Run V, small size biomass carriers packed in the small carrier bed were applied in

the reactor. All ofbiomass at the end of Run IV was used as seed of this Run. The reactor

was operated by a unique operational strategy of increasing the TNLR according to a

mathematic simulating model. The TNLR was exponentially increased from 2.98 to 11.5

kg-N m-3 d'i within 180 days ofoperation, and the TNRR successfu11y increased from O.46 to

8.3 kg-N m-3 d-i during this period. The HRT was shortened to O.32 hr on day 700 and

maintained this short HRT in the following operational period. TN removal efficiencies were

kept about 75 O/o with effluent TN concentrations ranging from 36 to 50 mg-N 1"i. Effluent

NH4-N and N02-N concentrations were below 25 and 10 mg 1"i, respectively. Compared

with former experimental results, effluent nitrogen concentrations and TN removal

efficiencies were surprisingly stable, which demonstrated the favorable application of

mathematic simulating model for increasing strategy in TNLR and the well anammox

performance deriving from the application of small size biomass carriers. From day 700 to

898, the operational condition was not changed, and stable TN removal performances were

achieved for 198 days. In this stable operational period, the TNRRs were kept stable at 8.2 to

9.93 kg-N m"3 d-i under the TNLR from 11.3 to 12.48 kg-N m-3 dHi. Effluent TN

concentrations were around 36 mg 1"i and effluent N02-N concentrations were always

below 1O mg 1-i with an average of3 mg 1-i.

The reaction ratio of TNRR, N02-N consumption and N03-N production rates to

NH4-N consumption rate during the whole experiment was 2.05:1.29:O.23 (Fig. 5-5), which

was slightly lower value compared to the well accepted anammox reaction ratio of

2.06:1.32:O.26 (Strous et al., 1998). However, the ratio of N02-N to NH4-N consumption

rates was 1.29 at TN concentration of 150 mg 1-i in this study, which was very similar to 1.3

at the TN concentration of 140 mg 1-i (Strous et al., 1999). The close reaction ratio with

82

reliable correlation elucidated the good anammox activity under moderately low

temperature.

5.3.2. 0bservationoftheattachedbiomass

Fig. 5-6. External appearance ofthe anammox reactor during the whole experiment

Generally, anammox biomass is apt to change its color (brownish red, red, or dark red

colors) depending on the cultural conditions because of the unique cytochrome content in

Ifi ,.ge.x . t. .. +.ft't}.etlv " .+ [ .. .-I"ts ;-- "/' :"' th ' ' ge "•k•:ilt- "T/

day 350 day 365

;

and 5-7 showed the changes in

color and shape during the whole

reactor from day 404, the

that for chrysanthemum shape

nonwoven filaments in increasing

of anammox biomass using small

83

day 200 day 498 Fig.5-7. Appearance ofthe anammox biofilm token from the reactor

anammox bacteria (van De Graafet al., 1996 Ahn et al., 2004).

Figures 5-6

anammox biomass

experiment. After adding small nonwoven filaments to

the fixed bed anammox samount of attached immobilized anammox biomass was :' emuch larger than

nonwoven strips. This verified the high sludge retammg

ability of small

biomass. The color

-

x

,A

ge:-l-T

B

day 898

7.ecm "K/ sc

s- 6.3cni

2cmx2cmx3cm

c'

Fig. 5-8. Size of small canier bed

size biomass carriers was changed fromgrey black (cement color) to red (Fig. 5-6, day 523,

575, 898), indicating the novel high sludge retaining characteristics of our invented small

size biomass carriers. Comparing the colors and surface structure of anammox biomass using

small size nonwoven filaments and small size caniers (Fig. 5-7, day 498 and 898), the

former was brownish red and loose structure, and the latter was much redder, denser and

homogeneous.

ewtm..

s

,ec

Fig.5-9. SEM images of small size biomass caniers. (A) Hydrated small cement carriers without

surface modification by hardcore-1; (B) hydrated small cement caniers after surface modification

by hardcore-1.

The clogging and channeling of biofilm occurred in the deeper part of nonwoven

biomass carrier and the inner part of big size anammox granules after fu11y developed

anammox reactor. These phenomena caused the poor substrate diffusion and finally resulted

in the decrease of anammox activities. If these poor cultivating conditions were kept for long

time, the color of anammox biomass changed to black (Fig. 5-7, day 200 and 350). It was

confirmed that black color of anammox biomass could return to red color by applying proper

cure for recovering deteriorated anammox activities.

5.3.3. Small size biomass carriers for enhancing anammox performances

The homogeneously spatial distributions of biomass in the anammox reactor using

small size biomass caniers was regarded to originate from the initial well mixing of seed

sludge with small size caniers and the continuous disturbance of reactor content by micro

84

bubbles of nitrogen gas from anammox reaction.

The hardcore-1 was a significant component in the small size carriers. After the surface

modification of the small size hydrated cement by the enwrapping of this nanophase material,

the smooth surfaces and passivity edge angles. (Fig. 5-9, A and B). The spatially wrinkled

structure with many needle fibers and micro channels produced after surface modification by

hardcore-1 was beneficial to the bacterial attached immobilization and substrate diffusion,

which consequently resulted in the high anammox activity. By this surface modification, the

whole of small size biomass caniers presented a homogenous dispersed status rather than

irregular aggregations without surface modification. Moreover, the surrounding ofnanophase

material on the small mixture may prevent the exudation of some heavy metal ions which

were supposed to be toxic to anammox bacteria. No clogging and channeling phenomena

were observed during the whole operational period for fixed-bed anammox reactor using

small size carriers. The nonwoven filaments performed the main function of connecting

these inorganic carriers and transfening the substrates due to its hydrophilicity. Because of

the high density of small size biomass carriers, anammox biomass was attached tightly to

small size biomass caniers, washout of anammox sludge from the reactor could reduce to a

great extent.

5.3.4. Simulation model for increasing TNLR using small size biomass carrier

In Run V, simulation model was used for increasing the TNLR. This model was

derived from two-dimensional nonlinear regression analysis using initial experimental results.

In this simulation model, time and TNLR are independent variables and TNRR was

independent variable. The predicted TNLR could be obtained by this simulation model, and

then this model was corrected by the actual experimental results. The corrected model was

used for further prediction. Based on the actual results at the end of this Run, the final

regression equation was defined as shown in Fig. 5-10. Fig. 5-10 demonstrated that the

TNLR to TNIUI presents a linear relationship, while TNLR and TNRR present exponential

relationship with time. By increasing TNLR based on this simulation model, the reactor

could reach a stable operational condition under which high anammox activities could be

85

realized. The further evaluation of this simulation model and the elucidation

relationship between this model and the growth rate of anammox bacteria are still

mvestlgatlon.

of the

under

Fig.5-ny10. Mathematic simulating model obtained in the end of the experiment. X axis: time

(days); Y axis: TNLR ( kg-N m'3 d-]); Z axis: TNRR ( kg-N m-3 d-').

The formula ofTNRR is Z==Z(X, Y) produced by software of 1stOpt 1.0. Whole formula

is shown as following.

z=exp((1+pl"log(x+3)"O.5+p2"log(y+3)"O.5+p3*log(x+3)+p4"log(y+3)+p5"log(x+3)"1.5+

p6*log(y+3)"1.5+p7*log(x+3)A2+p8'log(y+3)"2+p9"log(x+3)"2.5+plO*log(y+3)"2.5+pll

"log(x+3)"3+p12*log(y+3)"3+p13*log(x+3)AO.5"log(y+3)"O.5+p14"log(x+3)"O.5"log(y+3)

+p15"log(x+3)"O.5"log(y+3)"1.5+p16"log(x+3)*log(y+3)"O.5+p17*log(x+3)"log(y+3)+pl

8"log(x+3)"log(y+3)Al.5+p19"log(x+3)"1.5"log(y+3)"O.5+p20*log(x+3)"1.5"log(y+3)+p2

1"log(x+3)"1.5"log(y+3)"1.5)!(p22"log(x+3)AO.5+p23"log(y+3)"O.5+p24"log(x+3)+p25"lo

g(y+3)+p26"log(x+3)Al.5+p27"log(y+3)"1.5+p28"log(x+3)"2+p29"log(y+3)"2+p30"log(x

+3)"2.5+p31"log(y+3)"2.5+p32"log(x+3)"3+p33*log(y+3)A3+p34"log(x+3)"O.5"log(y+3)"

O.5+p35*log(x+3)"O.5*log(y+3)+p36"log(x+3)AO.5*log(y+3)"1.5+p37"log(x+3)"log(y+3)"

86

O.5+p38*log(x+3)"log(y+3)+p39"log(x+3)"log(y+3)"1.5+p40"log(x+3)"1.5*log(y+3)"O.5+

p41*log(x+3)"1.5"log(y+3)+p42"log(x+3)Al.5*log(y+3)"1.5+p43))-3;

There are 43 parameters. p(43,1)= [-541.17 -308.15 -12.85 639.3 44.00 546.23 22.79

141.88 4.88 -100.10 -1.1615.07 -16.74 147.5281.30-6.90-5.01--3123-20.35

-37.07 -41.06 35.95 86.26 -18.17 12.68 -13.84 44.22 -5.11 73.55 -O.93 51.17 O.42 -30.64

7.59 -3.30 -O.91 -O.50 -3.30 -9.40 1.04 -1.23 -7.91 390.19].

5.3.5. Comparisonofanammoxtreatmentcapabilities

In this study, four kinds of biomass carrier were applied sequentially, i.e., original

chrysanthemum shape with aid of simulating model for increasing TNLR. This result

demonstrate that the small size biomass caniers highest TNRR and the most stable

performance was obtained in Run V applying small size biomass carriers nonwoven

biomass carrier, split chrysanthemum shape nonwoven biomass carrier, splitted

chrysanthemum shape nonwoven biomass canier with undoing nonwoven small filaments ,

' Table 5-3 Comparison of different Anammox reactors under different operational conditions

Process Reactor Temperature TNRR.,, Stabletime (TNRRa) HRT.i. InfluentDO Reference(Oc) (kg-N m"3 d'i) (days) (kg-N m'3 d") (hr) (mg l-')

Anammox Fixed-bedreactor 37(2007)

Anammox ABF 20-22

AnammoxC RBCd 17Anammox Fixed-bedreactor" 18-26

26.0

10.1

O.5

9.93

1

103

15

320

200

(26)

(5•8m14)

(8.1)

(2.3)

(8.2-9.93)

O.24

O.32 hr

<O.5 Tsushima et al.

Se.5 Isaka et al. (2007)b

Cema et aL, (2006)

5-7(most) Thisstudy

aThis value referred to the average TNRR or TNRR scale at the stable operational time. bThe results showed in this study was nitrogen conversion rate calculated from the sum of

the NH1 -N and NOI -N removal rates, which should be lower than the TNRR due to discarding of the NOi -N production rate. CThis anammox process was accompanied with autotrophic nitrification and heterotrophic denitrification. d Rotating biological contactor. eFixed-bed anammox reactor using small size biomass carriers packed in small carrier bed.

and small size biomass carriers. The low-strength wastewater under moderately low

temperature have the advantage over nonwoven biomass carrier in the fixed-bed anammox

reactor treating the relatively Table 3 shows the comparison of our obtained experimental

87

results with other results obtained from various kinds of anammox reactors operated under

different conditions. In spite of our reactor's operation under unfavorable operational

conditions, i.e., moderately low substrate concentrations and operational temperature; report

on successfu1 operation of fixed bed anammox reactor with high TNRR under moderately

low temperature. carriers showes higher TNRRs as shown in Table 3. Especially, after

increasing TNLR based on simulation model, high-rate anammox treatment of TNRRs

ranged from 8.2 to 9.93 kg-N m'3 d'i was lasted stably for 200 days of operation. This result

will provide basic understanding for development of stable high-rate anammox treatment

using fixed-bed anammox reactor under moderately low temperature. To our knowledge, this

paper is the first high influent DO concentration and short HRT, novel fixed-bed anammox

reactor using small size biomass

5.4. Conclusions

During an experimental period for 900 days, we have succeeded in the establishment of

fixed-bed anammox reactor treating low strength influent wastewater of 150 mg-N 1-i under

moderately Iow temperature of 19-25 OC. Using nonwoven biomass canier, the TNRR

reached to 4.48 kg-N mr3 dHi with fluctuant TN removal efficienceies due to intermittent

increase of TNLR. By the application of small size biomass carriers, the TNRR could

increase to 8.3 kg-N m-3 d-i in 180 days of operation with a HRT of O.32 hr, and maintained

from 8.2 to 9.93 kg-N m-3 d-i with stable TN removal efficiency in the following 198 days of

operation. The results of this study on the stable high-rate anammox process treating

relatively low strength wastewater under moderately low temperature would be significant in

practical industrial application of anammox process.

5.5 References

Abma, W.R., Schultz, C.E., Mulder, J.W., Loosdrecht, M.C.M., Star, W.R.L., Strous, M.,

Tokutomi, T.: Full-scale granular sludge anammox process, In: Proceedings of the

International Conference on Biofilm Systems VI, Amsterdam, The Netherlands,

88

171-178(2006).

Ahn, Y.H., Hwang, I.S., Min, KS.: Anammox and partial denitritation in anaerobic nitrogen

removal from piggery waste, ifater Sci. Technol., 49, 145-153(2004)

APHA, Standard Methods for the Examination of Water and Wastewater, American Public

Health Association, Baltimore, MD. (1995).

Cema, G, Wiszniowski, J., Zabczyfiski, S., Zablocka-Godlewska, E., Raszka, A., and

Surmacz-G6rska, J.: Biological nitrogen removal from landfi11 leachate by

deammonification assisted heterotrophic denitrification in a rotating biological

contactor, Mater Sci. Technol. , 55, 35-42(2007).

Chang, K.-F., Yang, S.-Y., You, H.-S., Pan, J.R.: Anaerobic Treatment of Tetra-Methyl

Ammonium Hydroxide (TMAH) Containing Wastewater. Semiconductor

Manufacturing, IEEE Trans, Semicond. Manuf, 21 (3), 486-491(2008).

Dosta, J., Fernandez, I., Vazquez-Padin, J.R., Mosquera-Corral, A., Campos, J.L.,

Mata-Alvarez, J., Mendez, R.: Short- and long-term effects of temperature on the

Anammox process. , J. Hazard. Mater., 154 (1-3), 688-693(2008).

Fujii, T., Sugino, H., Rouse, J.D., Furukawa, K.: Characterization of the Microbial

Community in an Anaerobic Ammonium-Oxidizing Biofilm Cultured on a Nonwoven

Biomass Carrier, J. Biosci. Bioeng., 94 (5), 412-418(2002).

Furukawa, K., Rouse, J.D., Yoshida, N., Hatanaka, H.: Mass cultivation of anaerobic

ammonium-oxidizing sludge using a novel nonwoven biomass canier, J. Chem. Eng.

Jpn., 36, 1163-1169(2003).

Hirano, K., Okamura, J., Taira, T., Sano, K., Toyoda, A., Ikeda, M.: An efficient treatment

technique for TMAH wastewater by catalytic oxidation. Semiconductor

Manufacturing, IEEE Trans, Semicond. Manuf, 14 (3), 202-206(2001).

Isaka, K., Date, Y., Kimura, Y., Sumino, T. & Tsuneda, S.: Nitrogen removal performance

using anaerobic ammonium oxidation at low temperatures, FEMS Microbiol. Lett., 282

(1), 32-38(2008).

Isaka, K., Y., Sumino, and Tsuneda, S.: High nitrogen removal performance at moderately

low temperature utilizing anaerobic ammonium oxidation reactions, J. Biosci.

Bioeng. ,103 (5), 486-490(2007).

89

Jetten, M.S.M.: New pathways for ammonia conversion in soil and aquatic systems, Plant

Soil, 230, 9-19(2001).

Jetten, M.S.M., Cirpus, I, Kartal, B., van Niftrik, L., van de Pas-Schoonen, KT., Sliekers, O.,

Haaijer, S., van der Star, W., Schmid, M., van de Vossenberg, J,, Schmidt, I., Harhangi,

H., van Loosdrecht, M., Kuenen, J.G., Op den,Camp, H., Strous, M.: 1994-2004: 10

years ofresearch on the anaerobic oxidation ofammonium, Biochem.. Soc. Trans., 33,

119-123(2005).

Kanda, J.: Determination of ammonium in seawater based on the indophenol reaction with

o-phenylphenol (OPP), MaterRes. , 29, 2746-2750(1995).

Liu, S., Yang, F., Xue,Y., Gong Z., Chen H., Wang, T., Su, Z.: Evaluation of oxygen

adaptation and identification of functional bacteria composition for anammox

consortium in non-woven biological rotating contactor, Bioresour. Technol., 99,

8273-8279(2008).

Mulder, A., van de Graaf, A.A., Robertson, L.A., and Kuenen, J.G.: Anaerobic ammonium

oxidation discovered in a denitrifying fluidized bed reactor, FEMS Microbiol. Ecol.,

16, 177-183(1995).

Nicolella, C., van Loosdrecht, M.C.M., Heijnen, S.J.: Paricle-based biofilm reactor

technology, Trends BiotechnoL, 18, 3 12-320(2000).

Op Den Camp, H.J.M., Kartal, B., Guven, D., van Niftrik, L.A.M.P., Haaijer, S.C.M., Van

Der Star, W.R.L., Van De PasSchoonen, K.T., Cabezas, A., Ying, Z., Schmid, M.C.,

Kuypers, M.M.M., Van De Vossenberg, J., Harhangi, H.R., Picioreanu, C., Van

Loosdrecht, M.C.M., Kuenen, J.G., Strous, M., Jetten, M.S.M.: Global impact and

application of the anaerobic ammonium-oxidizing (anammox) bacteria, Biochem. Soc.

Trans., 34 (1), 174-178(2006).

Qiao, S., Kawakubo, Y., Cheng, Y., Nishiyama, T., Fujii, T., Furukawa, K.: Identification of

bacteria coexisting with anammox bacteria in an upflow column type reactor,

Biodegradation (in press) (2008).

Strous, M., Heijnen, J.J., Kuenen, J.G., Jetten, M.S.M.: The sequencing batch reactor as a

powerfu1 tool for the study of slowly growing anaerobic ammonium-oxidizing

microorganisms, Appl. Microbiol. Biotechnol.. 50 (5), 589-596 (1998).

90

Strous, M., Kuenen, J.G., Jetten, M.S.M.: Key physiology of anaerobic ammonium oxidation,

Appl. Environ. Microbiol., 65 (7), 3248-3250(1999).

Strous, M., van Gerven, E., Zheng, P,, Kuenen, J.G., Jetten, M.S.M.: Ammonium removal

from concentrated waste streams with the anaerobic ammonium oxidation (Anammox)

process in different reactor configurations, MaterRes., 31 (8), 1955-1962(1997).

Tsushima, I., Ogasawara, Y., Kindaichi, T., Satoh, H., Okabe, S.: Development of high-rate

anaerobic ammonium-oxidizing (anammox) biofilm reactors, Mater Res., 41,

1623-1634(2007).

van de Graaf, A.A., de Bruijn, P., Robertson, L.A., Jetten, M.S.M., Kuenen, J.G.:

Autotrophic growth of anaerobic amnionium-oxidizing microorganisms in a fluidized

bed reactor, Microbiology, 142 (8), 2187-2196(1996).

van der Star, W.R.L., Abma, W.R., Blommers, D., Mulder, J.-W., Tokutomi, T., Strous, M.,

Picioreanu, C., van Loosdrecht, M.C.M.: Startup of reactors for anoxic ammonium

oxidation: Experiences from the first fu11-scale anammox reactor in Rotterdam, Mater

res., 41 (18), 4149-4163(2007).

Wiesmann, U.: Biological nitrogen removal from wastewater, Adv. Biochem. Eng. Biotech.,

51, 113-154(1994).

Yang, Y., Zuo, J.-E, Shen, P., Gu, X.-S.: Influence of temperature, pH value and organic

substance on activity of ANAMMOX sludge, Huan Jing Ke Xue/Environmental Science , 27 (4), 691-695 (in Chinese). (2006).

91

92

Chapter 6

Conclusions and recommendations

6.1 Conclusions

There are several key points for favorable cultural conditions of anammox process.

They are the characteristics of infiuent wastewater (organic and inorganic carbon, salinity,

ammonium, DO concentrations), operational conditions (temperature, pH), suitable biomass

caniers and reactor types. The researchers who are engaged in anammox process have

already obtained extremely high nitrogen removal rates, but the long term and stable

maintenance of these high nitrogen removal rates were not reported. On the other hand, few

research works are canied out for anammox treatment under unfavorable operational

conditions.

Our experimental studies focused on the nitrogen removal by anammox process under

unfavorable operational conditions and evaluated their anammox treatment capabilities

experimentally.

1) Through the long-term stable operation of high-rate anammox biofilm reactor using

nonwoven biomass canier under moderately low temperatures and moderately low strength

ammonium-containing wastewater, we obtained the following results.

(a). The reactor system could operate for about three years without much disturbing the

treatment system.

(b) We have developed the novel small size biomass carriers composed of nonwoven fiber,

cement and hardcore-1. By using this small size biomass carrier technology, high stable

anammox activities were kept in the fixed bed anammox reactor.

(c) The maximum total nitrogen loading rate (TNLR) of 12.5 kg-N m-3 d'i and the maximum

total nitrogen removal rate (TNRR) of9.93 kg-N m-3 d-i with TNR efficiency of 80 O/o were

obtained for anammox reactor using small size biomass caniers. This high anammox activity

was kept for long time showing the stable sludge retaining capability of small size biomass

'carrlers.

93

(d) The hardcore-1 was a very important material in this small size biomass technology.

When the hydrates particles of cement were remodeled by hardcore-1, anammox sludge

could attach tightly on the layers of anaphases surface wrinkled coarse structure with many

needle fibers and micro channels. The high-rate nitrogen removal rates were realized using

application of hardcore- 1 .

(e) The TNRR could be controlled by the empirical mathematic model for increasing TNLR

according to partial derivative.

2) The effect of salt concentration on anammox treatment was experimentally evaluated

using fixed-bed anammox reactor with non-woven biomass canier. Anammox sludge

dominating freshwater anammox bacteria, strain KU2 and KSU-1, was used as seed sludge

of anammox reactor. Influent salt concentrations were increased stepwise from 2.5 g 1-i to

33 g IHi. Bacterial community was also examined by 16S rRNA gene analysis after the

acclimation of the anammox sludge to high salt condition. Based on these results, it was

revealed that freshwater anammox sludge could not be applied to the anammox treatment at

salt concentrations higher than 30 g 1-i. However, it was demonstrated that freshwater

anammox sludge could acclimate to salts concentration comparable to sea water level of30 g

1-1.

3) Nitrogen removal capabilities of two kinds of fixed-bed anammox reactors (one was using

nonwoven biomass carrier and another was small size biomass carriers) for treating partially

nitrified brine wastewater were compared.

(a) At the first stage of experiment, the fixed-bed reactor using nonwoven was fed with

partially nitrified brine wastewater and operated with a quick increase in TNLR, from O.8 to

2.12 kg-N m-3 d-i, only in 58 days. The TNRR increased synchronously from O.7 to 1.81

kg-N m-3 d-i due to the high TN removal efficiencies ranged from 83 O/o to 88 O/o. These

'results showed that fresh anammox sludge acclimated to NaCl concentration comparable to

'sea water level (30 g 1-i) could adapt quickly to the panial nitrified brine wastewater whose

salts concentration was comparable to sea water level. The color of freshwater anammox

sludge in fixed-bed anammox reactor using nonwoven changed to black color after anammox

treatment ofpanially nitrified brine wastewater.

(b) At the second stage experiment, fixed-bed anammox reactor using nonwoven biomass

94

carrier was changed to the fixed-bed anammox reactor using small size biomass carriers,

Maximum TNLR was increased to 3.17 kg-N m-3 d-i and the maximum TNRR reached to

'2.52 kg-N m-3 d-i, which was higher than that obtained in first stage experiments. The

average N02-N removal efficiency was about 88 O/o and the effluent N02-N concentrations

were kept below 11 mg 1ffi. The color of anammox biomass changed from black to weak red

color.

6.2 Recommendations

1) Anammox reactor using small size biomass canier technology has been proved to be an

efficient anammox reactor without clogging and channeling phenomena, which are the

disadvantages of fixed bed anammox reactor using nonwoven biomass carrier. Development

of suitable operational procedure for this novel anammox reactor is required for proper

application of this process to fu11 scale plant.

2) The preparation of nanophase materials of hardcore-1 must be improved in the future

study. It is very important to make hardcore-1 reacting with cement and others minerals

effectively. The favorable chemical bonds and physical connection between hardcore-1 and

cement must be investigated.

4) The development of novel economical nitrogen removal process such as our invented

fixed bed anammox reactor using small size biomass carriers is becoming increasingly

important as more stringent effluent discharge requirements are imposed. In order to answer

these requests, application test of newly developed nitrogen removal process must be canied

out in the fu11 scale plant.

95

Appendix: Publications related to this dissertation

t"inS(

kt Z Dissertation name: rStudies on fixedbed anammox process treating low strength ammonium

containing wastewater with high salinityj

(iE'fiifi'LpNE JSl[k7eeE.,\. 7' )/ ie= 7kJfltik7tc (1) pm J'k-Erttk Ji' }' )>t gi ))2 t7 - 70u -l >< ecFee7v3-6iilFts'Å}i]U)

1 mu

7AYiikl O JijS21 • H{}Jce

ag 2k

1 . Anammox treatment of low-strength ammonium-containing wastewater at room temperature

(5YM v(fk Anammox Er 7iE fB L vc 7' >i e =7JtkgeiftaJtses7j< rb) 6 a)gltS SA3< i21 k PeeH -9' 6 itst)

INTERNATIONAL SYMPOSIUM ON ENVIRONMENTAL SCIENCE AND TECHNOLOGY.

SECTION SIX: Water Pollution and Water Quality Control (653). Beijing, China, November

13-16, 2007. (International conference proceedings). pp.653-659 lff.

SIZ JSZ 19 qi 1 1 fi 1 3 -- 1 6 H rp MdkSl, M

;Ji"-l lg2i: gfJJi"t Elt. thJi(,ta-coVE?

2 . Anammox application to low strength ammonium containing wastewater (Anammox ErVaM Lv(

7 )/ =e J : 7- Jbtilg$kJfi JAt 7S< rb > C> Og$ ve ik eC Fes -g- 6 tfi Etf)

SIZJ5L 1s cEliRÅ}7k"tri L"kilijizzisZgKfiJ}5'I" IEEIIk/Akfi,ii;E,t S:;R CO-ROM. pp.875-876

iPJN 19 4G}li (:Ndi

ZJ2iZ: giJJISZR. diJliXd.,VA

eg 3 if

Effect of salt cncentration in aammox treatment using non-woven bomass crrier (tfi7ee Jxil? rb }" Z< $&

Jfii igreTlsc 2 Ltc Anammox aJLLmpCIJftX"g4Sl""-.,ts'xOimEt)

Journal ofBioscience and Bioengineering, Vol. 107, No. 5 (2009)

(Printed)

Zg21: vas,idJillZR. illlsJJA<-. iiiEiillwa. ecX:Z;Eft. 'Ei-JllR..tep

96

eg 4e

1 . Nitrogen removal capabilities of two kinds of fixed-bed anammox reactors for treating

partial nitrified brine wastewater. ( 2 Tptlfi. (D Anammox maErttK V 7 P l9 Er 7SM Lv( . gK'/z) t-lffk'

mauTL 8 nkza rb>ink vxog$ ua i21E2Obt ptk, )

Japanese journal of water treatment biology( H JzN 7LI< iSQ{lt ]Egl l lttig ':iil` t21 E','us.)

(Submitted)

gg Z : pa,, ti JilSt Ell . Jli] fi ij HJ] . tk HftR. -iliJIl pt.. , i?fi

eg s il3lr

1. Study on long-term stable operation of high-rate anammox biofilm reactor using nonwoven

carrier at moderately low temperature. (i[k?MExl\v(fs7i<filtli Erti;Eo fe Anammox gltbuEes V 7P 6i

Er l]2 .lgl e9 fo >- ) i-(t re'-t=' L k 6([Lt mp e c Pce7 -94 6 JIfi st)

Japanese journal ofwater treatment biology( EI JzN7Ll<pmÅ}llEEIt ltkif,ii!:`kE',tulrN.)

(Submitted)

llli :Z;L Z , M,, il JSL I:l . mu fi il HJ] . tk HjiiER. 'iSJII fiu VA

97

98