The Application of PH and Orp Process Control AEROBIC DENITRIFICATION

Download as pdf or txt
Download as pdf or txt
You are on page 1of 192

THE APPLICATION OF pH and ORP PROCESS CONTROL

PARAMETERS WITHIN THE AEROBIC DENITRIFICATION


PROCESS

A THESIS SUBMITTED IN FULFILMENT

OFTHE

DEGREE OF DOCTOR OF PHILOSOPHY IN CIVIL ENGINEERING

BY JOSEPH BERNARD HOLMAN

DEPARTMENT OF CIVIL ENGINEERING


THE UNIVERSITY OF CANTERBURY
NEW ZEALAND

AUGUST 2004
ACKNOWLEDGEMENTS

I would like to give sincere thanks to my Supervisor Dr David Wareham for his
guidance and encouragement throughout my study. The advice and knowledge
imparted by David is much appreciated.

Special thanks also to Dave MacPherson and Richard Newton (technical staff at
the University of Canterbury) who provided so much input to the project from
2000-2004.

The financial support offered by the University of Canterbury through a research


grant is gratefully acknowledged.

Thanks is given to the Foundation for Research Science and Technology whose
support through a Bright Future Scholarship made this research possible.

Finally, thanks to my family who have made significant personal sacrifices in


order for this research to be undertaken.

2 MAY 2005
TABLE OF CONTENTS
ACKNOWLEDGEMENTS
TABLE OF CONTENTS
LIST OF FIGURES AND TABLES
ABSTRACT

Chapter 'I LITERATURE REVIEW


1.1 NITROGEN IN WASTEWATER
1.2 NITRIFICATION 2
1.3 DENITRIFICATION 4
1.4 AEROBIC DENITRIFICATION 6
1.4.1 BIOLOGICAL EXPLANATIONS FOR AEROBIC DENITRIFICATION 8
1.4.2 PHYSICAL EXPLANATION FOR AEROBIC DENITRIFICATION 14
1.5 OPERATION OF PROBES 18
1.5.1 OXIDATION REDUCTION POTENTIAL (ORP) 19
1.5.2 HYDROGEN ION CONCENTRATION (pH) 20
1.5.3 DISSOLVED OXYGEN CONCENTRATION [DO] 22
1.6 ONLINE REAL TIME PROFILES (ORT PROFILES) 23
1.7 APPLICATION OF ONLINE PARAMETERS 28
1.8 PROJECT NEED, OBJECTIVES, AND METHODOLOGY 32
1.8.1 PROJECT NEED 32
1.8.2 PROJECT OBJECTIVES 33
1.8.3 PROJECT METHODOLOGY 34

Chapter 2 EXPERIMENTAL HARDWARE AND OPERATION


2.1 REACTORS 36
2.2 AIR SUPPLY SYSTEM 38
2.3 PROCESS CONTROL SYSTEM 41
2.4 FEED SUPPLY SYSTEM 44
2.5 EXPERIMENTAL OPERATION 46
2.6 REACTOR SEED 47
2.7 REACTOR FEED 47
2.8 REACTOR OPERATION AND CONTROL PARAMETERS 50
2.9 REACTOR COMISSIONING 52

Chapter 3 EXPERIMENTAL TESTING


3.1 TESTING UNDERTAKEN 53
3.2 SAMPLE PRESERVATION AND STORAGE 56
3.2.1 COD 56
3.2.2 TPN 56
3.2.3 NH 3 -N 57
3.2.4 N0 3-N 57
3.2.5 N0 2-N, ORTHO P, AND TP 57
3.3 ON-LINE MONITORING 57
3.3.1 ORP 58
3.3.2 pH 58
3.3.3 DO 58

Chapter 4 EXPERIMENTAL RESULTS SUMMARY


4.1 INTRODUCTION TO EXPERIMENTAL RESULTS 59
4.2 SUMMARY AND EXAMPLE OF OPERATIONAL DISSOLVED OXYGEN CONCENTRATION 4.0 mg/L 64
4.3 SUMMARY AND EXAMPLE OF OPERATIONAL DISSOLVED OXYGEN CONCENTRATION 2.5 mg/L 70
4.4 SUMMARY AND EXAMPLE OF OPERATIONAL DISSOLVED OXYGEN CONCENTRATION 1.0 mg/L 77
4.5 SUMMARY AND EXAMPLE OF OPERATIONAL DISSOLVED OXYGEN CONCENTRATION 0.5 mg/L 83
TABLE CONTENTS
Chapter 5.0 DISCUSSION

5,1 (Objective 1) Confirm the presence of aerobic denitrification activity 89


(Objective 2) Elucidate some operational aspects of aerobic denitrification, in particular comment on the
5,2 91
nitrification, denitrification, and sludge production rates,

5,2,1 Nitrification rates 91

5,2,2 Denitrification rates 93


5,2,3 Sludge production rates 97

(Objective 3) Comment on the requirements for air relative to conventional separate stage nitrification
5,3 98
denitrification processes,

(Objective 4) Identify the dissolved oxygen conditions necessary for aerobic denitrification and for its
5.4 101
optimisation,

(Objective 5) Comment on the need for soluble organic carbon for the removal of nitrogen in the aerobic
5,5 107
denitrification process,

(Objective 6) Confirm if the online profiles have unique features with respect to the aerobic denitrification
process, in particular the ammonia elbow on the ORP prOfile and the ammonia valley on the pH profile,
5,6 113
Correlate these online features with measured biochemical events such as the depletion of organic carbon
or ammonia nitrogen,

5,6,1 ORP 113

5,6.2 pH 118
5,6,3 Dissolved oxygen [DO] 119

5,6.4 Air flow rate 120

(Objective 7) Develop and demonstrate control algorithms that use online features to control the aerobic
5,7 denitrification process, (Le, indirectly detect the biochemical events), In doing so demonstrate the reliability 121
of pH and ORP control algorithms based upon relative rather than absolute values,

5,7,1 ORP 121


5,7,2 pH 122
5,7,3 Disolved oxygen [DO] 122
5,7.4 Air flow rate 123

Chapter 6.0 CONCLUSIONS AND RECOMMENDATIONS


6,1 CONCLUSIONS 124
6,2 RECOMMENDATIONS 128

APPENDIXA1
A1 TRACK STUDY ONE TO TRACK STUDY SEVEN DOSP 4,0 mg/L 130-136
A2 TRACK STUDY ONE TO TRACK STUDY SEVEN DOSP 2,5 mg/L 137-143
A3 TRACK STUDY ONE TO TRACK STUDY SEVEN DOSP 1,0 mg/L 144-150
A4 TRACK STUDY ONE TO TRACK STUDY SEVEN DOSP 0,5 mg/L 151-157

APPENDIX 81
B1 PHOTOS OF EXPERIMENTAL HARDWARE 158-161
B2 SEQUENCING BATCH REACTOR OPERATIONAL PARAMETERS 162
B3 DETERMINATION OF AERATION SYSTEM EFFICIENCY 163
B4 PROCESS CONTROL SOFTWARE AND CONTROL ALGORITHMS 164-168

REFERENCES
REFERENCES 169-181
LIST OF FIGU

Figure Description Page number

Schematic of possible dissolved oxygen profile within microbial floc (Adapted from Pochana
Figure 1.4-1 15
and Keller 1999).
Representative view of sludge floc under a microscope with suggested zones for nitrification
Figure 1.4-2 16
and denitrification. (Adapted from Demoulin et al (2001)).
Figure 1.5-1 Diagram of an ORP electrode and operation 19

Figure 1.5-2 Diagram of a pH electrode 22

Figure 1.6-1 Generalized schematic of ORP, pH, and DO profiles, main bending points illustrated. 23

Figure 1.6-2 Typical ORP vs time profile showing nitrate breakpoint (adapted from Wareham et al (1993)) 26

Figure 1.7-1 Typical ORP profile for AIO BNR plant (adapted from NZWWA, 1997) 31

Figure 2.1-1 Diagram of one sequencing batch reactor 36

Figure 2.2-1 Aeration system for set pOints down to 1.0 mg/L (Duplicated for second reactor). 39

Figure 2.2-2 Aeration system for set points below 1.0 mg/L (Duplicated for second reactor) 40

Figure 2.3-1 Process control system (with attached hardware)* 43

Figure 2.4-1 Heating of wastewater 46

Figure 2.5-1 Representative configuration of the SBR process. 47

Schematic of Bromley wastewater treatment plant with the location of experimental feed
Figure 2.7-1 48
collection indicated, (Adapted from Christchurch City Council public information brochure).

Figure 2.7-2 Organic carbon degradation during wastewater storage 50


Effect of DO concentration on SRT required to achieve effluent ammonia concentrations less
Figure 4.1-1 59
than 1.0 mg/L in a completely mixed activated sludge system (adapted from Metcalf and Eddy
Figure 4.2-1 Mixed liquor suspended solids concentration DOSP 4.0 mg/L 64

Figure 4.2-2 Mixed liquor wastage DOSP 4.0 mg/L 65

Figure 4.2-3 Solids residence time DOSP 4.0 mg/L 65

Figure 4.2-4 pH and [soluble nitrogen] 68

Figure 4.2-5 Air flow rate [NH3-N] and [COD] 68

Figure 4.2-6 ORP and [~O] 69

Figure 4.3-1 Mixed liquor suspended solids concentration OOSP 2.5 mg/L 70

Figure 4.3-2 Mixed liquor wastage DOSP 2.5 mg/L 71

Figure 4.3-3 Solids residence time DOSP 2.5 mg/L 71

Figure 4.3-4 pH and [soluble nitrogen] 74

Figure 4.3-5 Air flow rate [NH3-N] and [COD] 74

Figure 4.3-6 ORP and [~O] 75

Figure 4.4-1 Mixed liquor suspended solids concentration DOSP 1.0 mg/L 77
LIST OF FI URES

Figure Description Page number

Figure 4.4-2 Mixed liquor wastage OOSP 1.0 mg/L 78

Figure 4.4-3 Solids residence time OOSP 1.0 mg/L 78

Figure 4.4-4 pH and [soluble nitrogen] 81

Figure 4.4-5 Air flow rate [NH3-N] and [COD] 81

Figure 4.4-6 ORP and [~O] 82

Figure 4.5-1 Mixed liquor suspended solids concentration OOSP 0.5 mg/L 83

Figure 4.5-2 Mixed liquor wastage OOSP 0.5 mg/L 84

Figure 4.5-3 Solids residence time OOSP 0.5 mg/L 84

Figure 4.5-4 pH and [soluble nitrogen] 86

Figure 4.5-5 Air flow rate, [1\lH3-N] and [COD] 87

Figure 4.5-6 ORP and [~O] 88

Figure 5.2-1 Nitrification rate versus dissolved oxygen concentration 91

Figure 5.2-2 Extension in aeration time required to oxidise ammonia nitrogen (relative to OOSP 4.0 mg/L) 92

Figure 5.5-1 Cycle illustrating nitrogen removal rate versus soluble organic carbon availability 107

Figure 5.5-2 PHS (poly-B-hydroxybutyrate) 109

Figure 5.6.1-1 Illustration of ORP elbow in an AlO process. (Figure Adapted from Yu et al (1998) 114

Figure 5.6.1-2 Illustration of ORP elbow in an AlO process. (Figure Adapted from Ra et al (2000) 115

Illustration of ORP ammonia elbow in an aerobic denitrification process, Holman and


Figure 5.6.1-3 Wareham (2003), (note anoxic phase for first 10 minutes). *First ORP elbow resulted from 115
organic carbon depletion, not shown.
Illustration of ORP elbow in aerobic sludge digestion process. (Figure adapted from Wareham
Figure 5.6.1-4 116
et al 1993).
Figure 5.6.1-5 Illustration of absence of ORP ammonia elbow when dissolved oxygen concentration
116
maintained at a consistent level (taken from TS1 OOSP 4.0 mg/L).

Figure 5.6.2-1 Typical pH profile, (aeration starts at time = 0) 119

Figure 5.6.4-1 Typical air flow profile 120


LIST OF TABLES
Table Description Page number

Table 2.7-1 Influent wastewater characteristics 49

Table 2.7-2 Wastewater degradation while in cold storage 49

Table 3.1-1 Chemical parameters monitored 53

Table 3.1-2 Chemical parameters analyzed using colorimetric methods 54

Table 3.1-3 HACH nitrogen tests used in this research 55

Table 4.1-1 Summary of results DOSP 4.0 mg/L - 0.5 mg/L 63

Table 4.2-1 Summary of track study data [operational dissolved oxygen] 4.0 mg/L 66

Table 4.2-2 Summary of HRT values DOSP 4.0 mg/L 66

Table 4.3-1 Summary of track study data [operational dissolved oxygen] 2.5 mg/L 72

Table 4.3-2 Summary of HRT values DOSP 2.5 mg/L 73

Table 4.4-1 Summary of track study data [operational dissolved oxygen] 1.0 mg/L 79

Table 4.4-2 Summary of HRT values DOSP 1.0 mg/L 79

Table 4.5-1 Summary of track study data [operational dissolved oxygen] 0.5 mg/L 85

Table 4.5-2 Summary of HRT values DOSP 0.5 mg/L 85

Table 5.2-1 Aeration time relative to dissolved oxygen concentration 92


Comparison of biochemical transformation rates between conventional and
Table 5.2-2 95
aerobic denitrification processes.
Table 5.2-3 Normalized sludge production 97

Table 5.3-1 Total air required for ammonia oxidation 98

Table 5.4-1 Effect of dissolved oxygen concentration on aerobic nitrogen losses 101
ABSTRACT
Aerobic denitrification is a biological nitrogen removal process in which
nitrification and denitrification occur simultaneously (in the same reactor) under
identical environmental conditions. This contrasts to traditional separate stage
nitrification denitrification in which the nitrification and denitrification processes
occur sequentially in different reactors under opposing environmental conditions.
While aerobic denitrification has long been identified in other ecosystems (such
as the nitrogen cycle within soil) it was not thought possible within wastewater
treatment processes. The existence of aerobic denitrification within wastewater
treatment systems was first identified in the early 1980's following nitrogen mass
balances that suggested unexplained nitrogen losses were occurring in the
aeration tanks of many full-scale biological nutrient removal facilities (total
nitrogen losses of up to 30% were frequently occurring U.S.EPA (1987)). Since
then researchers and engineers have attempted to elucidate the mechanics
behind the aerobic denitrification phenomenon and the conditions required for its
optimization. It is thought that aerobic denitrification may offer advantages and
possible savings when compared to alternative traditional separate stage
nitrification denitrification processes.

The use of real time parameters such as ORP, pH, DO and airflow rate (oxygen
demand) can provide immediate insight into a biological treatment process. This
knowledge can be used to ensure optimum performance in terms of real time
pollutant concentrations and hydraulic loads. This research aimed to elucidate
some operational aspects of the aerobic denitrification phenomenon, to
investigate opportunities for several types of real time control (ORP, pH, DO, and
airflow), and to develop a process control system using the online parameters.

An activated sludge process was established within two lab-scale sequencing


batch reactors. The reactors were operated under a range of conditions using
raw domestic wastewater as the feed. ORP, pH, DO, and airflow were measured
online in real time while other biochemical parameters (such as the various forms
of nitrogen) were measured periodically using HACH photometric procedures.
Dissolved oxygen concentration was the operational variable (dissolved oxygen
set points (DOSP) 4.0-0.5 mg/L), other parameters such as MLSS concentration
and feed strength were maintained (where possible) at a consistent value (-3000
mg/L and - 600 mg/L COD respectively). The system was operated under both
nitrification and aerobic denitrification conditions with the dissolved oxygen
concentration determining the degree to which aerobic denitrification existed
(-40% TN removal at DOSP 0.5 mg/L).

The biochemical event of interest was the depletion of ammonia nitrogen. The
key online profiles of interest were the ORP-time profile and the pH-time profile.
The research sought to demonstrate the credibility of ORP and pH as real time
control parameters for the depletion of ammonia nitrogen in the aerobic
denitrification process. To achieve this a microprocessor-software based process
control system was developed by using the relationship between online
measurements and biochemical events.

The results indicated the ORP-time profile does not provide any feature for the
depletion of ammonia nitrogen when the dissolved oxygen is maintained at a
fixed concentration. That is the previously identified "ammonia elbow" is probably
the result of dissolved oxygen concentration breakthrough rather than nutrient
depletion. The lack of an ammonia depletion elbow meant that ORP could not be
used for process control within the aerobic denitrification process. The pH-time
profile showed an "ammonia valley" feature at the point of ammonia depletion.
This feature was consistently present in both the nitrification and aerobic
denitrification processes. The research incorporated the feature into the process
control system and successfully used it to control the length of the aerobic
denitrification treatment sequence. With respect to elucidating some operational
aspects of the aerobic denitrification phenomenon the main variable of interest
was the dissolved oxygen concentration. The results indicated the aerobic
denitrification process has an optimum dissolved oxygen concentration around
but probably below 0.5 mg/L. The process probably does not have an optimum
concentration but an optimum range. It is likely this range is influenced by
variables such as the biomass concentration and the release of reducing power
in terms of the ability to hydrolyze stored carbon polymers.

A secondary objective of this research was to elucidate advantages of aerobic


denitrification relative to alternative traditional separate stage nitrification
denitrification processes. For example it has been proposed that aerobic
denitrification may require smaller treatment reactors, require less air for
nitrification, produce less sludge per unit of wastewater treated (relative to a
traditional nitrification-denitrification process), and have less dependence on
organic carbon for denitrification.

The results suggested the low dissolved oxygen concentrations required for the
aerobic denitrification process significantly inhibit the nitrification process. This
causes a considerable extension in the required aeration times for the oxidation
of ammonia nitrogen (-300% increase in aeration time relative to a traditional
nitrification process). The longer aeration times suggest the process may not
offer savings in terms of aeration requirements (aerobic denitrification required
-200% more air per unit of wastewater treated relative to a traditional nitrification
process) or treatment tank sizes (relative to traditional separate stage
processes).

A reduction in the quantity of sludge produced (per unit of wastewater treated) of


over 30% was demonstrated for the aerobic denitrification process. While aerobic
nitrogen removal has been achieved under certain conditions autotrophically by
other researchers this work found the process is probably undertaken
predominantly by heterotrophic micro-organisms. The low dissolved oxygen
concentrations required for the process also appear to favor heterotrophic
denitrification using stored intracellular carbon (biosorption). This research
demonstrated the aerobic denitrification process was able to remove nitrogen
with less dependence on organic carbon (the organic carbon requirements for
aerobic denitrification were not quantified but experimental data suggests a
possible 40% saving) either by the use of a shortened nitrification-denitrification
pathway and/or the ability to use stored carbon.
Chapter 1 LITERATURE REVIEW

1.1 NITROGEN 11\1 WASTEWATER


Untreated wastewater typically contains nitrogen in one of five forms. These are
organic nitrogen, ammonia nitrogen (I\IH3-N) ammonium nitrogen (NH/-N), nitrite
nitrogen (N0 2-N), and nitrate nitrogen (N0 3-N). The organic and ammonia
nitrogen forms predominate. Organic nitrogen is converted to ammonia nitrogen
as cells die and bacterial decomposition of matter occurs. Ammonia nitrogen
exists either as the ammonium ion or as ammonia depending upon the pH, at pH
levels below 9 the ammonium ion predominates (Metcalf and Eddy (2001 )).

Primary and secondary treatment facilities do not remove nitrogen from


wastewater. Typically nitrogen remains in the organic, ammonia nitrogen, or
nitrate nitrogen forms. In some instances the discharge of nitrogen in treated
effluents may pose an environmental or public health risk. Nitrogen in the form of
organic nitrogen, ammonia, or the ammonium ion can exert an oxygen demand
on the receiving water body as it is transformed to an oxidized form. Nitrogen in
the ammonia and nitrite forms is known to be toxic to fish. Nitrate is a
biostimulant which may cause eutrophication of water bodies and has also been
linked to methemoglobinemia or blue baby syndrome in infants when present in
drinking water supplies (particularly contaminated ground water sources)
(Spalding and Exner (1993), Azov ef al (1995), Meinardi et al (1995), Agrawal ef
al (1999), Addiscott (2000)).

Due to the potential environmental and public health impacts the removal of
nitrogen from wastewater is becoming a requirement in many developed
countries. For example, countries within the European Union (EU) are
increasingly specifying nitrogen removal as a treatment objective (Fuerhacker et
a/ (2000)). Typical nitrogen removal rates (such as those required in Austria)
include 60-70% nitrogen removal (BGBI 210/1996) with a decrease in NH4-N and

1
N03-N to below 10 mg/L. The removal of nitrogen is also an economic necessity
for some European wastewater plants. New environmental protection legislation
has seen penalty taxes applied to nitrogen in wastewater effluents. For example
in Denmark the discharge of nitrogen from treatment plants is now taxed at a rate
of $US2.40/kg (Cecil (2003». Cecil (2003) investigated the Ejby Molle plant in
Denmark and reported possible penalty tax savings of $US 48,000 per year for
every 1mg/L of N that could be removed. The removal of nitrogen is typically
achieved in two separate stages, these are called nitrification and denitrification.

1 NITRIFICATION
Nitrification involves two steps the first being the biological oxidation of the
ammonium ion, (NH/) to nitrite nitrogen (N02 -) as shown in equation (1.2-1).

Step One - (Nitrification)


Nltrosomonas
5NH/ +7602 +109HC03

Oxidation state of nitrogen Oxidation of nitrogen :::


-3 + +4 ::;; +1 + -4::;; -1

Equation 1 First stage of nitrification.

Within this simplified step the oxidation state of the nitrogen goes from in the
ammonium to +3 in the nitrite. This has involved the donation of electrons from
the nitrogen atom (or oxidation) where oxygen has acted as the electron acceptor
(has been reduced). ((Metcalf and Eddy (1991».

This reaction proceeds through the action of a group of approximately six


bacteria commonly referred to as Nitrosomonas. These bacteria are
chemoautotrophic meaning they obtain their energy and generate their cell tissue
from chemical reactions involving inorganic material (Snoeyink and David

2
(1980». The bacteria act as catalysts in the reaction by reducing the activation
energy required for the ammonium ion to be oxidized (Thain and Hickman
(1996».

The second nitrification step involves the oxidation of nitrite (N02-) to nitrate
(N0 3-) as shown in equation (1.2-2).

Step Two - (Nitrification)


Nilrobacter

Oxidation state of nitrogen:: +3 Oxidation state of nitrogen:: +5


+3 + -4 ='1 +5 +"6::'1

Equation 1 Second stage of nitrification.

Within this simplified step the oxidation state of the nitrogen goes from +3 in the
nitrite to +5 in the nitrate. This has involved the donation of electrons from the
nitrogen atom (or oxidation). Again oxygen acts as the electron acceptor.

This reaction proceeds through the action of a group of bacteria, commonly


referred to as Nitrobacter. These bacteria are also chemoautotrophic. As with the
previous stage these bacteria act as catalysts reducing the activation energy
required for the reaction to occur.

Note that the term nitritation is often used to refer to the nitrification step from
!\IH3-N to N02-N while the term nitratation refers to the nitrification step from N02-
N to N03-N. It should also be noted that the nitrification process consumes
approximately 7.14 grams of alkalinity per 1 gram of ammonia oxidized, thus a
pH drop can occur if the wastewater lacks sufficient alkalinity for the nitrification
of ammonia to nitrate. From equations 1.2-1 and 1.2-2 it can be seen that
3
approximately 4.3 mg of O2 is required for the oxidation of each mg of ammonia
nitrogen to nitrate nitrogen. In a biological treatment process additional oxygen
would be supplied (by aeration) to the wastewater for these reactions. The
research of Painter (1977) found the DO concentration for nitrification should be
higher than 2.0 mg/L for nitrification to occur without the availability of oxygen
being rate limiting to the process. Bliss and Barnes (1986) found that nitrification
ceased to occur at residual DO concentrations below 0.2 mg/L..

It should be noted that the nitrification steps described in equations 1.2-1 and
1.2-2 do not facilitate nitrogen removal from the wastewater but they do eliminate
its oxygen demand. However the nitrogen is now in the form of nitrate nitrogen
(N03-N). To remove nitrogen another step called denitrification is required.

1.3 DENITRIFICATION
Denitrification involves the conversion of nitrate nitrogen (N03-N) to nitrogen
gases (which are released to the atmosphere). This is accomplished biologically
under anoxic conditions by a group of chemoheterotrophic bacteria.
Chemoheterotrophic means the bacteria obtain their energy and/or generate their
cell tissue from chemical reactions involving organic carbon material (Thain and
Hickman (1996)). The denitrifying microorganisms are part of the heterotrophic
group of microorganisms that may use oxygen as the electron acceptor in the
(biologically catalysed) oxidation of organic substrates. These organisms are
facultative in that they can live under altered conditions using either oxygen or
nitrate nitrogen as an electron acceptor (Henze (1991 )). In an aerobic
environment they tend to use oxygen as the preferred acceptor, however when
DO concentrations are very low or when only chemically bound oxygen is
available (Le. an anoxic environment) they have the ability to use nitrate nitrogen
as an electron acceptor. When anoxic conditions are provided along with an
organic substrate (requiring oxidation) this results in the reduction of the N0 3-N
to nitrogen gas as electrons are transferred from the substrate to the N03-N. It is

4
thought that the electron transfer pathway for the transfer of electrons from
organic material to the electron acceptor is similar regardless of whether oxygen
or nitrate acts as the final electron acceptor (Christensen and Harremoes (1977)
and Haandel et al (1981 )).

A simplified illustration of the denitrification reaction is shown in equation 1.3-1.

Equation 1.3-1 Simplified illustration of denitrification reaction.

The last three compounds are gaseous products (nitrogen monoxide or nitric
oxide, di-nitrogen oxide or nitrous oxide, and nitrogen gas) that are released to
the atmosphere.

The simplified denitrification reaction shown in equation 1.3-1 can be illustrated in


an example-using methanol as the carbon source as shown in equations 1.3-2,
1.3-3, and 1.3-4. Equations 1.3-2 and 1.3-3 show the energy creation and growth
components of the denitrification reactions with methanol while equation 1.3-4
shows the overall combination of the two reactions (Metcalf and Eddy, 1991).
Example, Denitrification (using methanol as the carbon source)

6NOJ - +2CHPH -+6N0 2 - +2C0 1 +4H 2 0


6N02 - +3CH 30H -+ 3N2 +3CO a +3H 20+60H-

~
/
Oxidation stale of nitrogen = +5
'5 +"61:'1
1 \ ""dot'" ..... of Mm,... 0

Oxidation stale of carbon:: " 2 Oxidation state of carbon = +4


'2 + '3 + '2 + +1 1: 0 +4 +"4:: 0

Equation 1.3-2 Energy reaction

5
Equation 1.3-3 Biosynthesis reaction

N03- +l.08CHPH +H+ -+O.065CsH 7 N02 +0.47N2 +O.76C02 +2.44H20

Equation 1.3·4 Overall reaction

Note in equation 1.3-2 that the oxidation state of nitrogen goes from +5 to O.
Therefore the nitrogen atom has accepted electrons or has been reduced. The
electrons have come from the organic carbon source. Note also the oxidation
state of the carbon has gone from -2 to +4 i.e. it has donated electrons or has
been oxidized.

Until recently denitrification within the wastewater treatment environment was


considered as being exclusively anoxic. However it has been found that (under
certain conditions) denitrification can occur in aerobic environments. This
phenomenon has been called aerobic denitrification.

1.4 AEROBIC DENITRIFICATION


Aerobic denitrification other wise known as simultaneous or co-current
nitrification and denitrification (SND) (co-N/DN) implies that nitrification and
denitrification occur simultaneously in the same treatment reactor under identical
environmental conditions.

Aerobic denitrification has attracted the interest of wastewater treatment plant


designers and operators because it offers
1. Potential to save on a second anoxic tank either by a reduction or by the
elimination of the need for anoxic facilities
2. May result in a reduction in the treatment time required for a nitrification
and denitrification process

6
3. Potential advantages in terms of energy savings by reducing the aeration
requirements required for the oxidation of ammonia nitrogen
4. Potential advantages in terms of reducing the need for variable
anoxic/aerobic zones and the associated energy requirements
5. Potential "biological" as well as conventional energy savings. For example,
research by Helmer and Kunst (1998) and Stross (2000) suggests that
under certain conditions the micro-organisms responsible for aerobic
denitrification may be autotrophic. Other research Third (2004) suggests
that under certain conditions aerobic denitrification may be undertaken
using stored intracellular carbon. If either of these claims are true the need
to supply a readily degradable organic carbon source (as a source of
electrons/energy) for denitrification may be reduced or overcome. This
may allow a reduction in the quantity of flow recycled or the need for an
additional external carbon source.

According to the traditional concept of nitrification and denitrification, aerobic


denitrification is not possible as nitrification relies on the presence of oxygen
where as denitrification requires the absence of oxygen. Initially the denitrification
process was thought of as truly anoxic, however from the mid 1980's it was
noticed that some full-scale facilities were experiencing nitrogen losses that could
not be accounted for by the traditional nitrogen and denitrification concepts. For
example a 1985 investigation of a full-scale SBR plant at the Grundy Center
VVVVTP in Iowa revealed the SBR operating without an anoxic or anaerobic phase
unintentionally removed approximately 80% of the inorganic nitrogen and over
50% of the phosphorus from the system (Irvine (1987». The U.S.EPA (1987)
reported that total nitrogen losses of up to 30% were frequently occurring in the
aeration tanks of many full-scale BI\lR processes. The documented losses in full-
scale facilities were complemented by researchers such as Kugleman and
Spector (1988) who identified nitrogen losses from an Ala treatment plant at

7
Pontiac Michigan in which full-scale aeration tanks reported total nitrogen
reductions of 30%.

Other early reports of the existence of aerobic denitrification activity include


Moriyama et 81 (1990) who demonstrated aerobic denitrification in biological
contactor units and Masuda et 81 (1991) and Gupta et 81 (1994) who
demonstrated aerobic denitrification in fixed film RBC units. Halling et 81 (1992)
demonstrated aerobic denitrification within up flow fixed bed reactors and
Watanabe et 81 (1992) and Munch et 8/ (1996) who demonstrated the
phenomenon within sequencing batch reactors.

There have been various explanations for the mechanisms by which nitrogen is
reduced within aerobic treatment systems to nitrogen gas. These explanations
sL1ggest aerobic denitrification is the result of either a biological or a physical
phenomenon. It is possible that a number of different processes all contribute in
some way to the reduction of oxidized ammonia.

1.4,1 BIOLOGICAL EXPLANATIONS AEROBIC


Biological explanations include microorganisms continuing to use nitrogen after
an anoxic phase, microorganisms simultaneously using oxygen and nitrate as
receptors, inhibition of oxygen respiration by nitrite resulting in the use of NOx-I\l,
that traditional BNR theory is too simplistic, and the physiological variety of
microorganisms is greater then previously thought.

An early theory proposed by KL1gleman and Spector (1988) was that some of the
microorganisms responsible for denitrification during an anoxic phase might be
able to continue reducing nitrogen for an undefined period once oxygen
concentrations had increased Le. in the aerobic zone/tank following an anoxic
period. They found the aerobic denitrification phenomenon seemed to be
restricted to the biomass which developed in Ala or anoxic/aerobic style
systems. The theory was also indirectly supported by Payne (1981) who found
8
the denitrifying enzymes of some bacteria were inactivated by oxygen, while in
others, synthesis was suppressed, but the existing enzymes disappeared only
gradually meaning it might be possible for some denitrifiers to continue
denitrifying after an increase in DO concentration. Munch et al (1996) suggested
that heterotrophic bacteria coming from the anaerobic/anoxic phase might for a
limited period still produce enzymes that use NOx-N as a final electron acceptor
even though readily available dissolved oxygen is present. Patureau et al (2000)
also identified improved aerobic denitrification activity just after an anoxic period
and suggested it was due to the "expression of" enzymes synthesized during the
anoxic period.

However there have been many instances of the existence of aerobic


denitrification in systems that do not contain anoxic or anaerobic environments.
Thus while these explanations cannot be ruled out it is now clear that there are
also other mechanisms involved. Another explanation for the aerobic nitrogen
reduction is the suggestion that it might be possible for some microorganisms to
simultaneously use oxygen and nitrate/nitrite as electron acceptors in the
oxidation process (Zhao et al (1999».

In the denitrification process nitrate and nitrite generally serve as electron


acceptors in the respiratory transport chain in the same way oxygen does in the
nitrification process. This is achieved with only a small modification to the
metabolic system of the micro-organisms. One explanation for the nitrogen
losses is that the acceptance of electrons by the nitrate/nitrite in solution during
an aerobic process could result in the transformation of these elements to a
gaseous nitrogen form. Some researchers have questioned this explanation on
the basis that nitrifiers have been shown to prefer using dissolved oxygen as an
electron acceptor even when the dissolved oxygen concentration is as low as 0.1
mg/L (Knowles (1982». However other researchers such as Kugleman et al

9
(1991) have suggested a possible inhibition of oxygen respiration by nitrite,
resulting in the use of NOx-N as an electron acceptor.

When aerobic denitrification is active the [NOTN] is often elevated (Hayward


(1998), Yoo et al (1999)). This is unusual as the traditional two stage nitrification
process is usually rate limited by the oxidation of NH3-N to N02-N resulting in the
rapid conversion of N02-1\I to N03-I\I, (and a distinct absence of
detectable/measurable N02-N). The higher levels of N02-N suggest the second
step of nitrification (nitration undertaken by Nitrobacter) may be inhibited.

Free ammonia or elevated pH levels have been known to inhibit Nitrobacter,


other inhibitors include reduced temperatures, operating solids content, and
acute process loadings (Anthonisen et al (1976), Alleman (1984), Abeling and
Seyfried (1992), Ballmelle et al (1992), Rhee et al (1997), Yu et al (1998)).

A number of researchers have proposed that nitrogen removal may take place
via a shortened nitrite pathway. In some instances this is accompanied by
elevated nitrite concentrations. For example Pochana and Keller (1999)
conjectured that elevated N02-N levels experienced in the aerobic denitrification
process may be linked to reduction of nitrogen directly from the N0 2-N form.
They found that when nitrification activity in the aerobic denitrification process
was inhibited nitrite was not further oxidized to nitrate even after ammonia
oxidation was complete. They proposed this situation accompanied a shortened
nitrogen removal pathway with the reduction of N02-N directly to nitrogen gas.
Kim et al (2003) demonstrated aerobic denitrification within a biofilm airlift reactor
operated specifically under conditions in which low dissolved oxygen created
Nitrobacter inhibition. Biological nitrogen removal via the nitrite pathway was
also demonstrated by Strous et al (1997). He identified a bacterium belonging to
the Planctomycetes group called ANAMMOX (acronym for Anaerobic Ammonium
Oxidation) that could oxidise ammonium anaerobically using nitrite as its electron

10
acceptor (converting it to dinitrogen gas and trace amounts of nitrate). The
ammonium was converted to dinitrogen gas and small amounts of nitrate. Stross
(2000) found that ANAMMOX bacteria can be active in the aerobic denitrification
process and that under conditions of low dissolved oxygen can reduce nitrite
directly to dinitrogen gas. He found that under conditions of low dissolved oxygen
concentration traditional nitrifiers such as Nitrosomonas and Nitrosospira could
work cooperatively with ANAMMOX bacteria. That is nitrification was limited to
the production of nitrite by Nitrosomonas and Nitrosospira which was then
simultaneously reduced. This is a completely autotrophic process which Dijkman
and Strous (1999) named CANON (an acronym for Completely Autotrophic
Nitrogen-removal Over Nitrite). The CANON process can be considered in two
stages which occur simultaneously, first the oxidation of ammonium to nitrite
(equation 1.4-1) and the reduction of nitrite to nitrogen gas (equation 1.4-2). The
combined reaction is illustrated in equation 1.4-3.

+ +H 2 0

Equation 1 Half step nitrification reaction under low dissolved oxygen


conditions

NH/ + 1. 3N02 ~ N2 +O.13N03- +2H2 0

Equation 1.4-2 Anammox mediated reaction

Equation 1.4-3 CANON reaction - Autotrophic removal of nitrogen under aerobic


denitrification conditions

Others to demonstrate denitrification from systems via the nitrite pathway include
Yu (1998) who showed denitrification from a continuous flow SBR system in
which full pathway nitrification had become fully inhibited, (minimal nitrate was

11
formed) and Abeling and Seyfried (1992), Yang and Alleman (1992), Akunna et
al (1993), Ho (1994), Helmer and Kunst (1998), Gejlsbjerg et al (1998), and Yoo
et al (1999) who identified coupled NH/-N oxidation and N02-N reduction as an
important source of N20 production under nitrifying conditions in activated
sludge.

From the mid 1980's it was also found that nitrate may be reduced under certain
aerobic conditions in place of oxygen. Poth (1986), Abeliovich and Vonshak
(1992), and Bock et al (1995) found that under oxygen limiting conditions some
autotrophic nitrifiers were able to reduce nitrate or nitrite into nitric and nitrous
oxide and nitrogen gas. More recently Patureau et al (1996a) demonstrated co-
respiration of oxygen and nitrates through kinetic experiments.

The use of nitrate nitrogen as an electron acceptor (in place of oxygen) may also
be considered from a thermodynamic perspective. The potential energy available
to microorganisms from the transfer of electrons depends upon several variables

including the final electron acceptor. The free energy change (pe o) between the
-2
use of 02' N03-N, and S04 shows the free energy change available through the
use of oxygen is only slightly higher than when N03 -1\J is used. This might
suggest the 02 electron acceptance pathway is only slightly preferred over the
N0 3-N pathway (i.e. the additional energy available to microorganisms through

the use of 02 as opposed to N03 -N may not be that significant). The pe o values
-2
for 02' N0 3 -N, and S04 respectively are +21.5 , +21, +5.75 (Snoeyink and
David (1980)).

Other likely explanations for the occurrence of aerobic denitrification are that the
range of microorganisms involved in biological nutrient removal may be of
greater physiological variety than previously thought, traditional nitrification

12
denitrification theory is too simplistic, and some previously identified organisms
have a more complex respiratory system than was initially recognized.

In recent years some researchers have suggested the physiological variety of


microorganisms involved in biological wastewater nutrient transformation is
greater than previously thought. For example Drysdale et al (1999) undertook an
investigation into the microbial species responsible for denitrification in activated
sludge processes and found the microbial consortium was so diverse that he
recommended further research to identify constituent bacteria. Delgenes and
Patureau (2004a) while investigating the ability of M. aerodenitrificans to nitrify
and aerobically denitrify attempted to isolate other aerobic denitrifiers, they found
many aerobic denitrifying strains. Frette et al (1997) isolated one hundred and
sixty-nine bacterial strains from an alternating aerobic/anaerobic activated sludge
wastewater treatment basin. They found sixteen strains from a sub sample of 23
nitrogen oxide reducers were true respiratory denitrifiers, and all denitrified under
both anaerobic and aerobic conditions.

Traditional theory also believed that (for the purposes of wastewater treatment
ecosystems) nitrification was undertaken by autotrophic microorganisms under
aerobic conditions, and denitrification was undertaken by heterotrophic
microorganisms under anoxic conditions. For practical purposes it was believed
that dissolved oxygen was the only electron acceptor in aerobic processes, that
is other elements such as nitrates could not be reduced in aerobic systems as
oxygen was thought to be preferentially reduced. However recent research has
shown the traditional theories to be too simplistic, for instance the presence of
heterotrophic nitrifiers in wastewater treatment processes has been
demonstrated (Verstraete (1975), Castlgnetti and Hollocher (1984), Kuenen and
Robertson (1987), Robertson (1988), Robertson et al (1988), Robertson et al
(1989), Van Niel (1991), Robertson and Kuenen (1992), Van Niel et al (1992),
and Zhao et al (1999». For example Robertson (1988) and Van Niel (1991)

13
identified heterotrophic microorganisms able to both nitrify in aerobic
environments and denitrify in both aerobic and anoxic environments and Zhao et
al (1999) found that heterotrophic nitrification could contribute a significant
fraction of ammonia oxidation under favorable conditions, such as low dissolved
oxygen concentration and relatively high organic loading. It was suggested that
autotrophic nitrifiers became inhibited under these conditions. The presence of
autotrophic denitrifiers has also been demonstrated Poth (1986), Abeliovich and
Vonshak (1992), and Bock et al (1995). They identified autotrophic denitrifiers
able to reduce nitrate or nitrite into nitric and nitrous oxide and nitrogen gas
under conditions of low dissolved oxygen concentration.

Since the mid 1980's it has also become evident that some of the previously
identified BNR species have a more complex respiratory system than was initially
recog nized. For example Robertson and Kuenen (1984) and Lloyd et al (1987)
demonstrated how traditional nitrifier species such as Nitrosomonas eutropha
and Nitrosomonas europea were able to denitrify in the presence of small
amounts of oxygen.

1 PHYSICAL EXPLANATION DENITRIFICATION


Another explanation is that aerobic denitrification results from anoxic micro zones
inside activated sludge flocs (Rittmann and Langeland (1985), Metcalf and Eddy
(2001 ». This theory suggests that aerobic denitrification is a physical
phenomenon that can be explained using the traditional concepts of autotrophic
nitrification on the exterior of sludge flocs and heterotrophic denitrification using
adsorbed organic carbon within an anoxic micro zone at the centre of the sludge
floc. The corner stone to this theory is the belief that denitrification could occur
within flocs due to a dissolved oxygen concentration gradient within the floc as
shown in Figure 1.4-1.

14
DISSOLVED OXYGEN
PROFILE INSIDE FLOC

Y-:;:,
--~- \ANOXIC

/K-- OXYGEN DIFFUSION

Figure 1.4-1 Schematic of possible dissolved oxygen profile within rnicrobial floc
(Adapted from Pochana and Keller 1999),

Typical floc sizes within the activated sludge process were found to range from 5-
200 IJm in diameter by Andreadakis (1993), 10-70 IJm by Pochana and Keller
(1999), and 50-110 IJm Third (2004). The floc diffusion theory was tested by
Pochana and Keller (1999) who illustrated that a reduction in median floc size
from 80 to 40 IJm by high speed blending of the biomass before a cycle resulted
in a reduction in aerobic denitrification nitrogen losses from 52 to 21 %. This
occurred while nitrification rates remained constant suggesting that the drop in
aerobic denitrification was not due to blending operations affecting the viability of
the biomass. The studies suggested that floc size may play an important part in
aerobic denitrification activity. Pochana et al (1999) also developed a
mathematical model to simUlate internal floc diffusion mechanisms and found
that while the outer part of flocs had a DO level above 0.2 mg/L (thereby
reducing denitrification activity) the central sections of flocs had considerable
time with virtually no DO providing an environment suitable for anoxic reduction
to occur. This finding supported the theory that aerobic denitrification may be
partly attributed to traditional denitrification within sludge flocs.

From a full-scale perspective it has been reported that the Potsdam wastewater
treatment plant in Germany (90,000 p.e,) achieves aerobic denitrification via

15
internal floc denitrification. Demoulin et al (2001) illustrated how the plant used
online measuring of the specific oxygen uptake rate to control the aerobic
denitrification process. Typical floc profiles were as illustrated in Figure 1.4-2.
They found the rate of nitrate diffusion into the floc was on the order of ten times
that of oxygen, thus under aerated conditions there was no nitrate limitation
within the floc. They also reported that sufficient carbon provision for
denitrification was achieved through carbon storage (biosorption) and the control
of dissolved oxygen which minimized the use of substrate carbon by oxic
metabolism.

NITRIFICATION

DENITiRIIFICATION

Figure 1.4-2 Representative view of sludge floc under a microscope with


suggested zones for nitrification and denitrification. (Adapted from Demoulin et al
(2001)).

Heinen and Norgaard (1998) reported on the application of floc diffusion theory to
the full-scale "Symbio process". The Symbio process is a nitrogen removal
technology which has been developed on the basis of using dissolved oxygen
concentration gradients within microbial flocs, The process attempts to maintain
aerobic and anoxic zones within flocs through the control of a number of
biological parameters (mainly low dissolved oxygen concentration).

16
The floc oxygen diffusion theory may be indirectly supported by a number of
researchers (refer discussion 5.4) who have suggested the optimum dissolved
oxygen concentration for aerobic denitrification would be around 0.5 mg/L. The
general consensus is that the process has an upper limit of around 1.0 mg/L and
an optimum just under 0.5 mg/L. It should be noted these are theoretical values,
a review of the literature found no research that has experimentally determined
the optimum dissolved oxygen concentration range for the aerobic denitri'fication
process.

However the floc diffusion explanation has also been questioned by a number of
researchers including Helmer and Kunst (1998) who achieved significant aerobic
denitrification nitrogen losses in a reactor (with dilute mechanically homogenized
biomass) in which large cell clusters were eliminated ensuring all cells were
supplied equally with DO during batch tests. The earlier research of Kugleman et
al (1991) also creates some doubt as to the degree of nitrogen loss that can be
attributed to anoxic micro zones within flocs. In a lab scale (AlO) type process
with significant aerobic nitrogen losses they increased the DO concentration to a
constant 10 mg/L in order to reduce the volume of any possible anoxic/anaerobic
micro zones within the interior of floc particles. After 14 days of operation the
aerobic nitrogen losses were identical to the losses they experienced while
operating at 5 mg/L (around 30%). Kugleman et al (1991) supported their
findings with research undertaken by Robertson and Kuenen (1984) and
Kawakami et al (1985) who both established aerobic denitrification within pure
cultures maintained under highly aerobic conditions. Patureau et al (1996b) also
demonstrated simultaneous nitrate and oxygen reduction within a chemostat
culture maintained under conditions of high dissolved oxygen. The floc diffusion
theory may also be questioned by the work of Bang et al (1995) who studied
aerobic denitrification activity in the biofilms attached to a partially submerged
RBC, they found the numbers of denitrifiers in the surface layer of the biofilm

17
were of 1 to 2 orders of magnitude higher than those in the middle and bottom
layers. From this they suggested the surface layers had a higher denitrifying
activity. They also found aerobic denitrification was active at dissolved oxygen
concentrations up to 6 mg/L.

It is clear that to date, no definitive explanation for the loss of nitrogen under
aerobic conditions has been agreed upon. It is likely that aerobic nitrogen losses
may be the result of a number of mechanisms such as biological conversion by a
broad range of microorganisms under a range of conditions, as well as traditional
biological conversion resulting from a physical phenomenon such as anoxic
micro zones contained within flocs. As to the debate over the conditions required
for aerobic denitrification the generally accepted consensus is that aerobic
denitrification is principally a low dissolved oxygen concentration process (:::; 1.0
mg/L), the optimum dissolved oxygen concentration range for aerobic
denitrification has not been determined experimentally. The requirements for
organic carbon remain unclear.

The measurement of biochemical parameters such as dissolved oxygen and pH


is important for control and monitoring of the nitrification, denitrification, and
aerobic denitrification processes, this is typically achieved using insitu probes.

1 OPERATION OF PROBES
The commonly used insitu control parameters for activated sludge are ORP
(Oxidation Reduction Potential), DO (Dissolved Oxygen) and pH as the probes
are inexpensive and easy to use (Peng et al (2002)). In this research all
measurements for ORP, pH, and DO have been undertaken using probes as
follows:

18
1.5.1 OXIDATION REDUCTION POTENTIAL (ORP)
ORP cannot be determined experimentally, rather it is always measured using an
ORP probe. The measurement of ORP allows the ratio of oxidants to reductants
prevailing within a solution to be established. The actual redox potential of a
solution may be considered as the tendency to give up or take up electrons (Le.
to be oxidized or reduced).

The ability of a solution to give up or take up electrons can be measured by


immersion of a platinum electrode along with a reference electrode (such as
Ag/AgCI). Figure 1.5-1 provides an illustration of an ORP probe and its
components. When the probe is immersed into a solution the flow of electrons (in
an attempt to create an equilibrium) results in a measurable millivolt difference
between the two electrodes.

COAXIAL CABLE TO METER


OR DATA LOGGER - CONTROL
SYSTEM

AglI'(JCI
REFERENCE
ELECTRODE-
CATHODE

INTERIOR Of THE CELL IS


EQUATION 1-3 REDUCTION OF Agel FILLED WITH ELECTROLYTIC
AgCI!,) + e- _ ~+ a- GEL SATURATED WITH
4.2 NKCI

CERAMIC PORUS GLASS INTERIOR


JUNCTION ALLOWS
PASSAGEOfK+
AND Cl- IONS T O " ..
MAINTAIN '"
ELECTRONUTRALITY 0" :::...------- PlATINUM BAND (NOBLE METAL)
ANODE

Reducing environment
/\ Highly oxidising environment (Flow
(illustrated) Le. reduction of of electrons not illustrated) i.e.
hydroxylamine to nitrite ... during the aeration phase ...
NH,OH + H,O .... NO,- +5H+ +4.- 0, ~ 4,- +4H' -> 2H,0

Figure 1.5-1 Diagram of an ORP electrode and operation

19
To illustrate, within a denitrification process, there is a lack of freely available
dissolved oxygen. Organic material within this environment is continuously
subjected to degradation by redox reactions catalyzed by enzymes. The organic
material is oxidized in that it donates electrons and these electrons are accepted
by nitrate nitrogen which is itself reduced to nitrogen gas. When an ORP probe
is in this solution some of the electrons from the oxidized organics will gravitate
along the platinum wire to the electrode (cathode) since the Ag/AgCI reference
electrode has a large positive electrode potential. The flow of electrons will
cause the Ag/AgCI to be reduced forming solid silver and free chloride irons. In
this environment when the flow of electrons is from the solution to the reference
electrode the recorded ORP value will be negative. Thus a reducing environment
(I.e. during an anoxic phase) will cause the ORP mV reading to fall or have a
negative slope over time (BIOLAB Scientific (2003)).
When oxygen is present in solution the oxygen has a larger positive potential
than the reference Ag/AgCI electrode. That is the electrons from the oxidation of
materials such as organic carbon and ammonia which would have traveled to the
reference electrode wi" instead travel to the oxygen (and reduce the oxygen to
water). Thus material such as organic carbon and ammonia nitrogen are oxidized
and donate electrons while oxygen is reduced or accepts electrons. In addition
the oxygen has a greater attraction for electrons than the probe's Ag/AgCI
electrode and this results in a flow of electrons from the electrode to the oxygen
in solution. When the flow of electrons is from the reference electrode to the
solution the recorded ORP value will be positive. Thus an aerobic environment
(I.e. during an aeration phase) wi" cause the ORP mV reading to increase or
have a positive slope over time (BIOlAB Scientific (2003)).

1.5.2 HYDROGEN ION CONCENTRATION (pH)


The pH can be determined experimentally but is usually measured with a pH
probe. The inside of a pH probe contains a HCI solution of fixed (H+] which acts
as one half of a cell, the other half of the cell being a reference electrode

20
composed of material such as Ag/AgCI. The probes have a thin glass (bulb
shaped) membrane located at the tip. The glass membrane separates the
internal HCI solution from the material being measured but allows an H+ ion
exchange to take place. The electrode works by allowing hydrogen ions to
exchange at the glass surface. The flow of H+ between the internal solution and
the solution being measured changes the cr concentration in the HCI solution.
This is in turn balanced by the Ag/AgCI electrode. Equation 1.5-1 illustrates the
action of the Ag/AgCI reference electrode depending upon the cr concentration
in the probe HCI solution.

AgCI(s) +e ¢:> Ag(s) + cr


Equation 1 Action of reference electrode

The Ag+ ion either forms an ionic bond with the cr ion or alternatively can accept
one electron to balance its valence charge. Thus there is a flow of electrons
similar to the situation described for the ORP probe (Snoeyink and David (1980),
Tchobanoglous and Schroder (1987), Oxford dictionary of chemistry (1996)).
Readings of the electron flow can be made on a high resistance voltmeter or pH
meter (which is essentially a high resistance voltmeter with a scale to read in
pH). This reading can be used to determine the difference between the constant
internal [H+] and the external variable [H+]. An illustration of a pH probe is
provided in Figure 1.5-2.

21
COAXIAL CABLE TO METER

AII/AgCI
COATONA
PLATINUM WIRE
REFERENCE
ELECTRODE

INTERNAL
SOLUTION
0.1 mol HCI ~ H+ SENSITIVE GLASS

Figure 1.5-2 Diagram of a pH electrode

1 DISSOLVED OXYGEN CONCENTRATION [DO]


The dissolved oxygen concentration within a solution can be measured
experimentally via chemical titration or alternatively using a DO probe and meter.
A DO probe contains a sensor which is surrounded by KCI solution. A membrane
separates the sensor from the solution being measured. Dissolved oxygen is free
to pass from the solution being measured into the probes KCI solution. A DO
meter is able to take a reading by passing a polarizing voltage across the sensor,
this causes oxygen which has passed throUgh the membrane to react at a
cathode causing a current to flow. The membranes and KCI solution are
disposable and usually require replacement every week if the probe is used
continuously. For low dissolved oxygen measurements such as in the aerobic
denitrification process thin membranes can be used which allow increased
sensitivity and greater accuracy of measurement (BIOlAS Scientific (2003)).

The output from probes such as dissolved oxygen, oxidation reduction potential,
and pH may be logged. When plotted over time the data produces curves known
as profiles. If plotted in real time the data produces online real time profiles.

22
1.6 ONLINE REAL TIME PROFILES (ORT PROFILES)
The present research used four online parameters, these were oxidation
reduction potential (ORP), pH, dissolved oxygen concentration [DO], and air
demand. Figure 1.6-1 illustrates a generalized schematic of the online profiles
that could be expected from an AIO process treating domestic wastewater (note
the air demand profile is not illustrated).

pH
1

ORP

Time

Figure 1.6 1 Generalized schematic of ORP, pH, and DO profiles, main bending
m

points illustrated.

The profiles contain features that correspond to changes in biochemical


conditions. These features are of interest as they indicate when certain events
have occurred and can provide insight into the treatment process. To understand
the use of real time parameters it is necessary to have knowledge of the various
features, thus the following section will provide some background information to
this end.

23
With respect to the ORP profile in Figure 1.6-1, the ORP profile has four main
bending points; these are labeled A, B, C and 0 respectively, (note that bending
points Band C also appear on the DO profile).

At point A the air is switched on which causes the ORP to change from a
negative (downward facing) slope to a positive slope, (Le. as discussed in section
1.5.1 electrons start flowing from the ORP electrode to the oxygen in solution).
The strong attraction of oxygen for electrons results in an interrelationship
between the ORP and DO profiles. ORP is linked to dissolved oxygen
concentration (or any electron acceptor/donor for that matter) by a logarithmic
relationship (Paul et al (1998». Thus the change in ORP is not proportional to
electron acceptor concentration but the log of its concentration. This means that
large changes in concentration result in small changes in redox potential (Mosey
(1985».

For a nitrification type process the ammonia nitrogen concentration will be


depleted after a certain aeration period. Point B is thought to occur when the
ammonia nitrogen is exhausted resulting in the elbow feature on both the
dissolved oxygen and the ORP profile, (often referred to as the "ammonia elbow"
Yu et al (1998), Ra et al (2000), Holman and Wareham (2003». Dissolved
oxygen breakthrough typically occurs at point B as oxygen is no longer required
to act as an electron acceptor for ammonia. This means that additional oxygen is
available to accumulate and this increases the residual dissolved oxygen
concentration. Note that the logarithmic relationship between DO and ORP
means that under conditions of high dissolved oxygen concentration a small
variation in the dissolved oxygen concentration may not be reflected on the ORP
profile, thus point B will only be clearly shown on the ORP profile if the oxygen
concentration is relatively low (- < 2 mg/L).

24
The logarithmic relationship between the dissolved oxygen concentration and
ORP allows the bending point B on the ORP profile to be used to indicate
whether the aeration rate is appropriate for the oxygen demand. That is the
appearance of the ORP bending point typically indicates the oxygen supply rate
is suited to the oxygen requirements at that time. Under aeration will result in the
dissolved oxygen not breaking through in the aeration time of the process (which
means point B does not appear) while over aeration will also result in Point B not
appearing on the ORP profile. The absence of point B can also indicate problems
with the nitrification process, such as nitrification inhibition (Paul et 81 (1998)).

Once aeration is terminated the dissolved oxygen concentration starts to fall


rapidly as biological activity consumes the remaining residual dissolved oxygen.
Once the freely available dissolved oxygen is fully depleted the system makes
the transition from using dissolved oxygen to using chemically bound oxygen,
that is the system goes from aerobic to anoxic activity. At the point where the
system becomes anoxic the ORP changes from a relatively level plateau to a
steep negative slope (illustrated in Figure 1.6-1 as point C). The flow of electrons
changes due to the shift to a reducing environment, that is electrons start flowing
to the ORP cathode. It should be noted that this part of the ORP profile
effectively represents the actual respirometric activity of the sludge and ORP will
decrease strongly only when dissolved oxygen has been consumed completely
(Zipper et 81 (1998)).

In the anoxic environment, nitrate nitrogen is reduced to nitrogen gas. The nitrate
acts as the electron acceptor (is reduced) while organic carbon is oxidized
(donates electrons). Once the NOx-N is depleted there is a transition from anoxic
to anaerobic conditions. At this point the bacteria pass from highly efficient
respiration (Le. using nitrate as a terminal electron acceptor in the electron
transport chain) to less efficient (e.g. anaerobic fermentative) processes
(Wareham et 81 (1993), Plisson-Saune et 81 (1996)). This transition from anoxic to

25
anaerobic conditions results in a nitrate breakpoint feature (Peng et al (2002),
Cecil (2003)). This is Illustrated as point 0 on Figure 1.6-1 and also in more detail
in Figure 1.6-2. The nitrate breakpoint feature is often called the "nitrate knee"
and it signals the point where the denitrification reaction is complete (i.e. all
nitrates have been eliminated).

5
:::::J' 4
c::n :3 Point of Zero
~ 2 Nitrate Concentration
<:
~
h 1
/
.-
<3"
0
-1
("'j
"S;:- -2
~ -3
--ORP
A NOs-N (mg/L)
&:
0
-4
-5
-6
Reducing Conditions
-7
0 2 4 6
(hours)

Figure 1.6-2 Typical ORP vs time profile showing nitrate breakpoint (adapted
from Wareham et al (1993))

Points Band D (Figure 1.6-1) are of particular interest to those working with
online real time systems as they represent the points at which nitrification and
denitrification (respectively) are complete. Identification of these points can
potentially allow treatment sequences to be terminated resulting in benefits such
as savings in aeration, greater treatment capacity, and improved process
stability.

With respect to the pH profile on Figure 1.6-1, the profile has 6 main bending
points or features that are correlated to changing biochemical conditions. At the
start of aeration there is typically a sudden sharp rise in the pH (points 1-2). This

26
is thought to be due to CO2 stripping from the system caused by the initiation of
aeration (Hao and Huang (1996)). After the initial increase the pH decreases
again (points 2-3), as nitrification consumes alkalinity as part of the nitrification
process (7.14 mg HC03 consumed per mg of ammonia). At point 3, the pH curve
shows the feature termed "ammonia valley" (Andreottola et al (2001)) Point 3
occurs when ammonia is essentially depleted (and nitrification is complete), note
that point 3 corresponds to point B on the ORP profile.

Following the ammonia valley at point 3 continued aeration will result in an


increase in the pH (points 3-4). Hao and Huang (1996) found that the mechanism
behind this increase in pH was complex and not fully understood. They
suggested it may be caused by the hydrolysis (or cleavage of organic nitrogen
molecules with water) of organic nitrogen after the exhaustion of the ammonia.

For a short period following the termination of aeration the pH can fall. This
occurs as the stripping of CO 2 from the system (caused by aeration) ends and
there is an increase in the CO 2 concentration (points 4-5) which may occur until
anoxic conditions develop. Once the conditions are anoxic the pH increases
again as alkalinity (consumed during nitrification) is partially released back into
the environment (points 5-6). The increase in pH occurs until denitrification is
complete, i.e. all nitrates have been reduced to nitrogen gas. At this point there is
a pH breakpoint commonly referred to as the "nitrate apex", (point 6) (Peng et al
(2002)). The nitrate apex is related to the ORP "nitrate knee" in that it signifies
the pOint at which all nitrates have been depleted and the system makes the
transition from anoxic to anaerobic conditions.

Following the nitrate apex the pH profile tends to plateau and then decrease
again as anaerobic conditions develop (Points 6-1). Anaerobic conditions occur
in three main stages, hydrolysis, acidogenesis, and methanogenesis.
Acidogenesis results in the breakdown of products into simple organic acids, the

27
most common of which is acetic acid. The breakdown of products into organic
acids results in the fall in pH observed between points 6-1.

Points 3 and 6 are of particular interest to those working with online real time
systems as they represent the points at which nitrification and denitrification
(respectively) are complete. As with the ORP features the identification of these
points can potentially allow treatment sequences to be terminated resulting in
benefits such as savings in aeration, greater treatment capacity, and improved
process stability.

1.7 APPLICATION OF ONLINE PARAMTERS


Stricter wastewater effluent requirements and issues such as conservation of
energy and a limitation on the available area for treatment facilities has
emphasized the need for incorporating new technology into wastewater
treatment processes in order to increase their performance (Peng et al (2002)).
The move towards greater efficiency may be considered both in terms of greater
pollutant removal but also in terms of regulating plants to ensure optimum
performance in terms of real time pollutant concentrations and hydraulic loads,
that is to minimize inefficiencies caused by plants designed for operation at a
static design load (Risholt et al (2002), Baeza et al (2002)). In the last few
decades the establishment of intemational commissions has resulted in the
implementation of new wastewater treatment regulations within the European
Union (EU). These regulations have drastically increased the need for real time
monitoring systems within European wastewater treatment plants (Gunatilaka
and Dreher (2003)). Regulatory authorities are now frequently requiring biological
nutrient removal (BNR) strategies to be implemented at wastewater treatment
facilities, for example new regulations within parts of the EU require nitrogen
removal as a treatment objective. Common nitrogen removal requirements (such
as those required in Austria) include 60-70% nitrogen removal (BGBI 210/1996)
with a decrease in NH4-N and N03-N to below 10 mg/L (Fuerhacker et al (2000)).

28
There is also a growing need within the EU for improved online real time control
systems to monitor and control pollution in rivers and water bodies into which
treated wastewater is discharged, that is to monitor and take into account
pollutant emissions when determining control strategies for minimizing the impact
of compounds such as ammonia nitrogen on urban water bodies (Meirlaen ef al
(2002), Frehmann ef al (2002». Online real time control strategies are a way of
achieving the new treatment objectives and facilities such as new BNR plants
offer the opportunity to incorporate real time control as part of the plants ICA
system (Cecil (2003».

pH is an established form of online measurement often used in the wastewater


treatment industry to pinpoint the end of nitrification by identifying the ammonia
depletion breakpoint (AI-Ghusain ef al (1994), AI-Ghusain and Hao (1995), Ra ef
al (2000), Andreottola ef al (2001 ». ORP is also an established form of online
measurement used to pinpoint the end of denitrification «Koch ef al (1985),
Wareham ef al (1993), Yu ef al (1997a), Charpentier ef al (1998), Yu ef al (1998),
Peng ef al (2002), Cecil (2003».

pH and ORP have shown themselves to be low maintenance, reliable, and


relatively inexpensive when incorporated as process control tools in wastewater
treatment processes. The probes are not particularly sophisticated and not
usually associated with the number of practical problems that can be
encountered with the use of sophisticated sensors in the harsh and dirty
conditions of a wastewater treatment process (Vanrolleghem and Lee (2003».
Alternatives such as the latest biosensors can have life times as short as 6
weeks and are highly dependent upon variables such as effective sterilization
during construction and can be susceptible to mechanical damage (Nielsen ef al
2002».

29
ORP measurement was first used in wastewater treatment from the early 1940's
following the development of practical ORP electrodes for this application
(Rohlich (1948)). However with the subsequent development of the dissolved
oxygen electrode engineers and plant operators lost interest in redox (Cecil
(2003)). Interest in ORP was revived when nutrient removal became an objective
in the 1980's (Koch (1985)). Nutrient removal required the inclusion of treatment
phases in which no dissolved oxygen was present. The profiles from pH, and
ORP can operate in anoxic and anaerobic conditions and have the capacity to
provide useful information under these conditions with respect to nitrogen
transformation and removal. This is primarily because in low dissolved oxygen
and/or anoxic-anaerobic environments significant changes in pH and redox
potential can occur while conventional control parameters such as the dissolved
oxygen probe may show little or no change.

The use of the ORP profile in aerobic-anoxic-anaerobic conditions is illustrated in


Figure 1.7-1. The figure shows aeration/non-aeration phases of an AlO type BNR
plant. From the start of aeration the ORP profile increases steadily until aeration
is terminated after which the ORP decreases. While the operating range of the
dissolved oxygen probe stops when the dissolved oxygen is depleted the ORP
profile continues providing information during the anoxic-anaerobic reducing
conditions. pH and ORP have now become promising contenders for the
automated control of process trains that incorporate a micro/non-aerated zone.

Aeration On
Off
r::==::::J.__...r:==L.._J:====L___.L
1'""

Relative
Oxygen
Concentration ¥-----I-_......L.~_.L-_.........~_..L_ _ _~

ORP

Time (hours)

Figure 1.7-1 Typical ORP profile for A/O BNR plant (adapted from NZJJVWA,
1997)
30
The use of pH and or ORP profiles for control purposes requires the online data
to be analyzed so that process control decisions can be made. Specific profile
features are correlated to known biochemical events, when these profile features
appear their detection by a control system provides an indirect way of identifying
when the biochemical event has occurred. Control decisions can be based upon
either the use of set point (absolute) values (that is using set points at which the
event is thought to occur) or relative values. Relative values involve
measurement of the profile slope so that changes in slope can be used to identify
the feature.

Profile features may be considered relative in the sense that different probes may
show the breakpoints (features) occurring at different absolute values due to
irregularities in the mixing regime, individual probe characteristics, and localized
oxidation-reduction reactions occurring in the vicinity of the tip of the probe
(ORP). However most probes will tend to show the breakpoints occurring at the
same moment in time.

Initial concerns regarding the usefulness of ORP measurements in biological


systems Stumm (1966), Morris and Strumm (1967) and Harrison (1972) have
been alleviated in light of the knowledge that the emphasis in some situations
can be transferred from the absolute ORP value to the relative ORP value (Le.
the ORP variation with time or dORP/dt).

The use of relative ORP variations to monitor nutrient depletion has been
demonstrated by numerous researchers including Charpentier et al (1989),
Peddie et al (1990), Wareham et aJ (1994). Paul et aJ (1998), and Ra et al
(2000). For example Ra et al (2000) operated a two stage sequencing batch
reactor treating piggery wastewater and used relative ORP to detect the point of
nitrate depletion via identification of the nitrate knee feature. The change in slope

31
or derivative of the ORP profile has also been employed in some modern
process control systems and aspects such as the nitrate knee are now being
regularly detected by on-line systems which use the information in real time to
make changes to the process control (Le. initiation or termination of aeration).
For example, the recent adoption of the nitrate knee in the Zullig ORP control
system is reported to provide savings in energy of up to 30% over a strictly timer-
based system (Water and wastes in NZ (1997)).

The use of relative pH or pH and ORP variations to monitor nutrient depletion


has also been demonstrated by numerous researchers including AI-Ghusain and
Hao (1995), Yu et al (1997), Kim and Hao (2001), and Peng et al (2003). For
example Peng et al (2003) investigated the variation of pH with time during a
nitrification process treating brewery wastewater. They attempted to use the pH
ammonia valley to determine the point of ammonia depletion. Kim and Hao
(2001) operated an alternating aerobic-anoxic continuous flow activated sludge
process and used relative pH to detect the point of ammonia depletion via the
ammonia valley and relative ORP to detect the point of nitrate depletion via the
nitrate knee.

Aspects such as the application of online parameters for monitoring and possibly
control of the aerobic denitrification process, the optimum dissolved oxygen
concentration, and the requirements for organic carbon are some areas that
require further research and clarification.

1.8 PROJECT NEED, OBJECTIVES, AND METHODOLOGY


1.8.1 PROJECT NEED
To date, little research has been undertaken in the area of leA (Instrumentation
control and automation) or real time control of the aerobic denitrification process.
Exceptions include Zhao et al (1999) who used absolute ORP as a real-time
control parameter for aerobic denitrification, Demoulin et al (1997) who illustrated

32
how ORP measurement and its rate of change throughout a cycle could
successfully control dissolved oxygen set points as well as optimize nitrogen and
phosphorus removal in a full-scale aerobic denitrification sequencing batch
reactor facility, Holman (2000) who investigated online ORP and DO profile
features within the aerobic denitrification process and highlighted possible
features for process control, and Third (2004) who used online OUR
measurements as a process control parameter for the aerobic denitrification
process.

A review of the literature suggests there is still a need to experimentally


determine the optimum dissolved oxygen concentration for the aerobic
denitrification process. It is likely the optimum dissolved oxygen concentration for
aerobic denitrification will not be a fixed value but an optimum range dependent
upon variables such as the wastewater composition and the biomass
concentration.

The requirements for organic carbon in the aerobic denitrification process needs
to be clarified. For example the removal of nitrogen via autotrophic denitrification
(including aerobic denitrification) was identified in 2002 as an area that requires
further research by WG 4 biological processes (working group 4) a division of
COST (European cooperation in the field of scientific and technical research)
(Cost (2003)).

1.8.2 PROJECT OBJECTIVES


The objectives of this work were to elucidate some operational aspects of the
aerobic denitrification phenomenon (relative to alternative traditional separate
stage processes). That is to provide some comparisons on aspects such as
nitrification, denitrification, and sludge production rates. To comment on the
effect of dissolved oxygen concentration and the need for organic carbon, and to

33
investigate opportunities for several types of real time control (ORP, pH, DO, and
airflow).

To achieve this the experimental work had the following specific objectives:

1. Confirm the presence of aerobic denitrification activity


2. Elucidate some operational aspects of aerobic denitrification, in particular
comment on the nitrification, denitrification, and sludge production rates
3. Comment on the requirements for air relative to conventional separate
stage nitrification denitrification processes
4. Identify the dissolved oxygen conditions necessary for aerobic
denitrification and for its optimisation
5. Comment on the need for soluble organic carbon for the removal of
nitrogen in the aerobic denitrification process
6. Confirm if the online profiles have unique features with respect to the
aerobic denitrification process, in particular the ammonia elbow on the
ORP profile and the ammonia valley on the pH profile. Correlate these
online features with measured biochemical events such as the depletion of
organic carbon or ammonia nitrogen
7. Develop and demonstrate control algorithms that use online features to
control the aerobic denitrification process, (Le. indirectly detect the
biochemical events). In doing so demonstrate the reliability of pH and
ORP control algorithms based upon relative rather than absolute values

1.8.3 PROJECT METHODOLOGY


Lab scale sequencing batch reactors were operated treating domestic
wastewater under conditions likely to produce nitrification and aerobic
denitrification.

34
An intensive chemical parameter testing program (series of track studies) was
undertaken along with on-line monitoring of ORP, pH, DO, and air demand to
provide real-time analysis of process conditions. Specific cycles were singled out
for intensive testing of parameters such as COD, TN, NH3-N, N0 2-N and N03-N.
Samples for each of the parameters were taken periodically throughout a cycle.
The track studies provided an indication of the presence of aerobic denitrification
activity (in the event that a loss of NH3-N occurred without a corresponding
increase in N0 2-N or N03-N, once nitrogen assimilation had been considered).

Following the confirmation of the presence of aerobic denitrification activity, its


performance was monitored as the operational dissolved oxygen concentration
was decreased.

The automatic collection of probe readings at intervals of up to sixty seconds


provided real-time profiles rich in information. The outputs from the probes were
used for analysis of both trends and specific biological events.

nally the research attempted to correlate online features with biochemical


events and to develop control routines which used these features.

35
Chapter 2 EXPERIMENTAL HARDWARE AND OPERATION

2.1 REACTORS
Biological treatment of the wastewater was undertaken in two sequencing batch
reactors (SBRs). Figure 2.1-1 illustrates a schematic of a reactor.
TORQUE INDEPENDENT
CONSTANT SPEED MIXER _ _ _ (FILLING) EMERGENCY
, CUTOFF FLOAT swrrCH

o
o

o
o

SAMPLING PORTS - - -
HEIGHT ADJUSTABLE
DECANT PIPE MOUNTED
IN BASE OF REACTOR

25mm DIA SPHERE

SHAPEDAQUARIUM
STONE POROUS ----1=~~ii:
AIR DIFFUSER

REACTOR FILLING PORT


LOCATED IN BASE

Figure 2.1-1 Diagram of one sequencing batch reactor

The reactors were built with the intention of using one reactor as a control,
however the consistency of the treatment process (from the perspective of online
readings and biochemical parameter testing) led to the decision to run the
reactors independently.

After reviewing various reactor materials (e.g. glass) the decision was made to
construct the reactors from 6mm thick perspex. This material was chosen for its
high impact resistance and because it was transparent. Transparency was
important to allow visual monitoring of the biological process and the assessment
of aspects such as flocculation and sedimentation. Other components within the
reactors were made from stainless steel, brass, or polyethylene. This eliminated

36
the use of materials which may have degraded in the potentially corrosive reactor
environment.

The decision was made to construct the reactors following similar designs used
by other researchers (Munch et al (1996), Yu et al (1998), Andreottola et al
(2001), Peng et al (2002)). The size of the reactors was based upon the likely
cycle lengths and the quantity of feed that could be stored in a modified freezer.
As the Christchurch City wastewater treatment plant was 20 km away it was
decided the design should not require the collection of feed more than once per
week. The usable freezer storage volume was about 450 L, with two reactors this
limited the design to a maximum through flow of about 225 L per week per
reactor (refer Appendix B2 for design details).

Mixing was undertaken by mixing blades. The motors that turned the mixing
blades were required to operate up to 20 hours per day, 7 days per week.
Stepper motors were used as they did not contain armature brushes which could
have worn out with continuous use. In order to overcome any overheating
problems the stepper motors were mounted in aluminum cases with thermally
conductive paste, (allowing the motor case to act as a heat sink). In addition a
secondary cooling fan was also incorporated within the motor case to provide
forced cooling. The use of stepper motors also provided the added benefit of
being torque independent, i.e. the RPM was constant regardless of the torque
experienced allowing for constant mixing.

Simple, open-shut, solenoid valves (230 volt) were used for control of clarified
effluent decanting. The solenoids were normally shut but would open upon
supply of 230 volts from the process control system.

37
Aeration of the activated sludge within the reactors was provided by compressed
air sourced from the laboratory reticulated air supply. A single air diffuser was
mounted in the base of each reactor.

2.2 AIR SUPPLY SYSTEM


Supply of air to the reactors was necessary for the aerobic period in which
carbon, ammonia, and nitrite nitrogen were all oxidized. The reactors were
operated at various dissolved oxygen set points (DOSP) from 4.0 mg/L to 0.5
mg/L. For the dissolved oxygen set points from 4.0 mg/L to 1.0 mg/L an aeration
system was developed as illustrated in Figure 2.2-1.

The upstream side of the aeration system started with a regulator to reduce the
mains air pressure from 80 PSI to 40 PSI allowing for pressures within the range
specified by manufactures of the downstream equipment (such as variable rate
solenoids and air flow meters. The regulator was followed by a Burket 0-10 Bar
230 Volt on/off solenoid and then a Burket variable rate solenoid, (Burket 0-9
L/min, 0-20mA, 0-2.8 Bar Model 6021). The purpose of the first on/off solenoid
was to remove air pressure from the variable rate solenoid when the air supply
was off.

38
PROCESS CONTROL
COMPUTER
MANUAL AIR FLOW
_ INTERNET AND RATE METER
INTRANET LINK

. ..
SEQUENCING
BATCH REACTOR
.. . ELECTRONIC AIR
FLOW RATE METER
WITH AERATION PROPORTIONAL
AIR VALVE
CONTROLLER
VARIABLE AIR FLOW
RATE SOLENOID
(ORT) [DO]

ON/OFF AIR SOLENOID

MAIN (MICROPROCESSOR BASED)


PROCESS CONTROL SYSTEM

REGULATOR

RETICULATED
AIR SUPPLY
(80 PSI)

Figure 2.2~1 Aeration system for set points down to 1.0 mg/L (Duplicated for
second reactor).

The variable rate solenoid was opened and closed by the proportional air valve
controller which determined the difference between the actual dissolved oxygen
concentration within the reactor and the desired concentration (specified in the
process control software). This difference corresponded to the degree to which
the air valve would open. As the difference became smaller the valve would
slowly close. Readings were updated every six seconds thereby allowing
responsive accurate control of the reactors dissolved oxygen concentration.

Downstream from the variable air flow rate solenoid there was an electronic
Honeywell mass air flow sensor. The flow rate sensor was a Model AWM5000
micro switch series/micro bridge able to detect flow accurately between 0-20
Llmin. The electronic flow rate sensor was linked to the process control system
and the computer to provide online real-time readings of air flow rate.

39
The variable rate solenoids used for the dissolved oxygen set points 4.0 mg/L to
1.0 mg/L were unable to accurately control the low air flow rates required for the
set points below 1.0 mg/L. To obtain accurate control of the low air flow rates it
was necessary to use a micrometer based needle valve.

A needle valve was selected that provided a linear response (that is a linear
increase in flow rate as the needle valve was opened). As the system had to be
automated the needle valve was opened and closed by a stepper motor
connected to the process control system. The 8 Bit control system allowed the
motor to control the valve by 256 graduations from the closed to the fully open
position.

The flow rate from the valve was dependent upon the input pressure but an input
pressure of 25 PSI provided for a flow rate from 0 Llmin to 9 Llmin. The 256
graduations allowed this flow rate to be controlled theoretically in steps of 36
mLimin. The aeration system for the set points below 1.0 mg/L is illustrated in
Figure 2.2-2.

PROCESS CONTROL MANUAL AIR FLOW


COMPUTER RATE METER
INTERNET AND
INTRANET LINK
ELECTRONIC AIR
FLOW RATE METER
SEQUENCING
SATCH REACTOR • 0
(ORT) MICROMETER
WITH AERATION BASED FLOW RESTRICTOR wmr
LINEAR RESPONSE NEEDLE
VALVE • ADJUSTED BY S VOLT
STEPPER MOTOR

ON/OFF AIR SOLENOItl

MAIN (MICROPROCESSOR BASED)


PROCESS CONTROL SYSTEM
REGULATOR

RETICULATED
AIR SUPPLY
(SO PSI)

Figure 2.2~2 Aeration system for set points below 1.0 mg/L (Duplicated for
second reactor)

40
Diffusion of oxygen into the reactor was achieved with a single 25 mm diameter
sphere-shaped aquarium diffuser. Experimental work (Appendix 83) determined
there was little increase in oxygen transfer efficiency when multiple stones or
other diffuser types (such as sock diffusers or porous aeration discs) were used.
The oxygen transfer efficiency was found to be 3 % (in fresh water at 20°C). This
compares to values ranging from 27-39 % commonly experienced in full-scale
facilities (Metcalf and Eddy (1991)). The main impediment to increasing the
transfer efficiency was the depth of the reactor relative to full-scale facilities. Full-
scale aeration facilities frequently have depths in excess of 4.5 m while the depth
of the lab scale reactor was only 0.38 m. The depth of the aeration tank
translates directly into contact time between the diffused air and mixed liquor. As
the contact time could not be modified it was necessary to have high air flow
rates in order to compensate for the low dissolved oxygen transfer efficiency.

The duration of the aeration phase was determined either by online real time
readings or alternatively upon default timer values loaded into the process control
system.

Note that low dissolved oxygen concentrations can be maintained by either direct
control (as used in this research) or indirect control of the air flow rate. Direct
control relies upon readings from online real time dissolved oxygen sensors while
indirect control relies upon other process parameters such as ORP or NADH
fluorescence. Researchers to use direct control include Osada et al (1991) and
Suwa et al (1992); and use of ORP has been undertaken by Moriyama et al
(1993) and Collivignarelli and 8ertanza (1999); while NADH fluorescence has
been used by Helmo (1993).

2.3 PROCESS CONTROL SYSTEM


Automated control of the reactors was necessary to allow the experimental setup
to run continuously unassisted when no operator was present. A microprocessor

41
based control system as illustrated in Figure 2.3-1 was developed in house at the
University of Canterbury. The process control system was composed of two key
elements, a central control box and a personal computer.

The central control box was microprocessor based and contained 16 analogue
inputs, 16 digital inputs, and 16 digital outputs. This allowed complete and
independent control of all the required reactor operations. Included within the
system was an independent "watch dog" microprocessor whose purpose was to
maintain control of the reactor on a limited basis in the event of a computer
malfunction. The central control box checked the status of the personal computer
every six seconds. In the event of identifying a computer problem the watchdog
microprocessor was programmed to act as a stand-alone unit operating on a
predetermined timer basis until being manually overridden by an operator. This
provided additional operational security by reducing the possibility of a
"biological" upset, if the computer had problems during an unattended period.

The personal computer ran a software program developed with the Lab View
package from National Instruments. The software allowed the computer to
control and monitor the process using either an onboard NO card or via one of
the computer communication ports. Other researchers to have used Lab View for
this type of application include Yu et al (1998) and Andreottola et al (2001).

Appendix B1 contains photographs of the experimental hardware including the


process control system while Appendix B4 provides a description of the process
control software and the algorithms used.

42
EMERGENCY
OVERFLOW
_ _ _ SWITCHES

AIR

(4)
TREATED
EFFLUENT

WASTE
SLUDGE

MUNICIPAL
SEWAGE

KEY
(1 ) Process controller, microprocessor control, controlled and monitored on AMD 1 GHz Athlon PC running
Windows XP. Box incorporates 16 analogue inputs, 16 digital inputs, and 16 digital outputs.
(2) Online real-time data reports for the analysis of treatment process
(3) Network link allowing for remote access to process control system, for example via Intranet and intemet.
(4) Regulated compressed air supply
(5) 500Lcapacity cold wastewater storage with submersible feed pumps
(6) Immersion heated water bath
(7) (PAC) Proportional air valve controller
(8) Waste solenoids
(9) Air solenoid
(10) Variable rate air solenoids
(11 ) (FRS) Electronic flow rate sensors
(12) 8L capacity sequencing batch reactors
(13) Mixers
(14) DO meters
(15) pH meters
(16) ORP cables
(17) Fill control pressure transducers

Figure 2.3-1 Process control system (with attached hardware)*


43
The process control software displayed and recorded data for ORP, pH, [~O]

and airflow (oxygen demand). It was possible to input algorithms to allow the
software to interpret the data for process control decisions. For example if the
profiles displayed certain features that were correlated to known biochemical
events the algorithms could identify these features and then initiate some form of
action.

The use of a personal computer allowed some forms of modern communication


technology to be incorporated. For example a permanent internet connection
permitted remote control and monitoring of the process and provided access to
WAP, (Wireless Application Protocol) technology. Access to WAP technology
meant that emails or SMS (Short Message Service or txt-text messages) could
be sent directly to any (WAP) cell phone allowing the operator to be alerted to
conditions while being away from the process. A live Internet based camera was
also used to provide remote visual checking of the process performance when
the system was unattended. This visual data could also be accessed with any
intemet connected PC, WAP phone, or pocket PC (with an appropriate display).

As the process control system was developed on the fly, it was periodically
necessary to put the reactors under the control of a PLC (programmable logic
controller) while maintenance or changes were made to the main system.

2.4 FEED SUPPLY SYSTEM


To store the wastewater and limit degradation, a 500 L chest freezer was
purchased and fitted with an external temperature controller built in-house at the
University of Canterbury. The temperature controller required the input of a
desired temperature set point, it then simply turned the freezer on or off
depending upon the actual temperature as measured by a submersed
thermostat.

44
Tile freezer contained two submersible feed pumps and one submersible mixer
pump. The wastewater mixer was connected to a timer which switched the mixer
on for 3 minutes of each 15 minute period. This ensured the wastewater
composition was as consistent as possible. The mixer pump was located so that
mixing could be undertaken without the pump acting as an aerator (thus reducing
biological degradation). To prevent the growth and acclimatization of bacteria in
the freezer it was emptied and washed thoroughly before each fresh batch of
wastewater was added.

By trial and error, it was found that the lowest temperature the wastewater could
be safely stored at without risk of forming ice on the sides of the freezer (and
possibly damaging the pumps) was 3.8 DC. Since the wastewater temperature at
Bromley was typically between 15-20 DC, the large thermal mass of the
wastewater (-450 L at 15 DC) meant the freezer typically took 15-20 hours to
lower the temperature to the desired 3.8 DC.

The operating temperature of the reactors was approximately 20 DC and since


the wastewater was stored at 3.8 DC it was necessary to heat the influent before
input into the reactors as low temperatures could slow biological activity. To
achieve the required temperature increase, the feed lines from the freezer to the
reactors were passed through a water bath (2350 Watt unit produced by Grant
Instruments). The polyethylene feed lines were connected to glass heat transfer
coils submerged in (potable) water that was heated to approximately 50 DC. The
glass coil contact time was a matter of seconds due to the feed line flow velocity.
The water bath contained an in-built stirring mechanism which aided the heat
transfer. Figure 2.4-1 depicts the water bath used.

45
EfFLUENT
WASTEWATER
fEEDTD
REACTORS
-21 DEGREES
CENTIGRADE

INFLUENT
WASTEWATER
REMOVABLE LID FEED LINES TO
TO REDUCE EVAPORATION GLASS HEAT TRANSFER
COILS, If/PUT TEMP - 3,8

\ DEGREE CENTIGRADE

WATER BATH
FmEO WITH BULL CocK
DICITALOlsPLAY MAINS PRESSURE WATER
THERMOSTAT AND
SYSTEM To REpLACE WATER
TEMPERATURE CONTROL
LOST TO EVAPORATION
LINKED TO I1EA'TING COIL
ANOMIXER

Figure 2.4-1 Heating of wastewater

It is also important to note that the laboratory was air-conditioned to provide a


constant air temperature of 20°C. This resulted in stable/consistent reactor
temperatu res.

EXPERIME NTAL
The SBR operation had four sequential steps to each cycle namely (1) fill & mix,
(2) react (aeration), (3) settle (sedimentation/clarification) and (4) draw (decant).
Figure 2.5-1 illustrates the four main steps to each cycle.

46
-- -
FILL MIX SElTLE DECANT
AEROBIC
INFLUENT EFFLUENT

MIXED LIQUOR RECYCLE


SLUDGE WASTING

Figure 2.5-1 Representative configuration of the SBR process.

2.6 REACTOR SeED


Activated sludge from the Lyttleton sewage treatment plant, (near Christchurch,
New Zealand), was obtained by collecting a 20 L bucket of mixed liquor from the
oxidation ditch. The bucket contents were allowed to settle for 30 minutes before
the effluent was decanted to provide the seed material (activated sludge). A
sample of approximately 200 mL of concentrated sludge (~5000 mg/L) was then
used to seed the reactor. Once seeded the reactor was filled to the top with fresh
wastewater. To acclimatize and grow the mixed liquor the reactor was initially
operated with a default time value of 150 minutes aeration at a DOSP of 4.0
mg/L (instead of online readings).

2.7 REACTOR FEED


The reactor was fed with fresh municipal wastewater obtained from the inflow
region of the primary sedimentation tanks at the main Christchurch wastewater
treatment plant located at Bromley. Figure 2.7-1 shows a schematic of Bromley
wastewater treatment plant with the location of feed collection point.

47
Point of wastewater collection at the
Christchurch city wastewater treatment
plant (Bromley WWTP).

Wastewater collected
from the h,)ad of the primary CLARIFIER
sedimentation tanks, (just following
the aerated grit chambers),

Figure 2.7-1 Schematic of Bromley wastewater treatment plant with the location
of experimental feed collection indicated, (Adapted from Christchurch City
Council public information brochure).

The wastewater at this point had been through coarse bar screens to remove
large objects such as rags and sticks and had been through an aerated grit
chamber sequence to remove most of the inert coarse grit. The composition of
the wastewater is illustrated in Table 2.7-1 which shows the average values from
tests taken over the duration of the work.

48
Table 2.7-1 Influent wastewater characteristics

TCOD seOD NH4 - N N02-N N03-N TP TPN


TCODfTPN
(mg/L) (mg/L) (mg/L)

Average 700 350 55 0 0 10 65 11


total soluble total total total total total
±er 91 71 9 0 0 2 9

The wastewater was obtained once or twice a week depending on requirements.


Checks to monitor the wastewater composition while in cold storage were carried
out on a regular basis. An example set of tests are presented in Table 2.7-2,
these results show the wastewater composition over a seven day period while
Figure 2.7-2 illustrates the decline in [COD].

Table 2.7-2 Wastewater degradation while in cold storage

Component mg/L Mon Tue Wed Thur Fri Sat Sun

COD Filtered 380 330 326 280 242 225 191


NH3-N Filtered 26 33 33 33 33 32 34
NOTN Filtered 0 0 0 0 0 0 0
N03 -N Filtered 0 0 0 0 0 0 0
SS 186 154 157 156 151 145 148

Table 2.7-2 shows the COD declining at an almost linear rate and the ammonia
nitrogen increasing slightly as organic nitrogen is converted to the ammonium
cation NH/ (wastewater pH was typically between 6.8-7.4). Note the decline in
unfiltered COD appears to be largely caused by the fall in soluble COD. The
BODslTN ratio for the wastewater was 11 based upon a BOD 5 concentration of
domestic wastewater generally around 50% of the unfiltered COD concentration
(0.4-0.8 BODs/COD Metcalf and Eddy (1991), -0.5 BODs/COD WRC (1984».

49
800

700

600
'§, 626
593
TOTAL COD
E
z 500 544
0 525 524
;:: 496
IiI- 400
Z
W
u
z 300 330 326 SOLUBLE COD
0
u 280
0
0 200 242
u 225
191
100

0
1 2 3 4 5 7
DAYS IN STORAGE

Figure 2.7-2 Organic carbon degradation during wastewater storage

2.8 REACTOR OPERATION AND CONTROL PARAMETERS


The first reaction step involved the introduction of oxygen that was maintained at
a specific set point value for the entire aeration phase. This set point was
achieved by the variable input of air dependent upon a reading from a dissolved
oxygen probe. The actual dissolved oxygen concentration in the reactor relative
to the desired concentration determined how much air the control system would
supply to the system. The actual air flow rate was measured continuously with an
inline electronic air flow meter.

A specific dissolved oxygen concentration was required because it was


necessary to investigate biochemical conditions (such as ammonia oxidation
rates, nitrite and nitrate reduction rates, development or inhibition of aerobic
denitrification) that resulted from operating at specific dissolved oxygen
concentrations. With respect to ORP it was necessary to differentiate between
ORP changes that resulted from nutrient depletion and those that resulted from
changes in the dissolved oxygen concentration. The maintenance of a fixed
dissolved oxygen concentration eliminated the possibility of ORP changes
resulting from a dissolved oxygen breakpoint.

50
The duration of the aeration phase was determined by the process control
system detecting an online profile feature which had been correlated to the
depletion of ammonia nitrogen. There was also a default aeration time which
terminated the aeration phase in the event that the online real time control
system failed. The default aeration time depended upon the dissolved oxygen set
point and the likely time required to oxidize the ammonia nitrogen.

The settle and decant steps were 15 and 5 minutes respectively and were
controlled by a timer within the process control system. The fill step was
dependent upon a fill level sensor; however this phase typically lasted 2-3
minutes.

The quantity of the clarified effluent that was decanted at the end of each cycle
was set at 50 % of the reactor volume or approximately 4.0 L (Total used reactor
volume 8.0 L). Sludge wasting was undertaken manually during the fully mixed
aeration cycle. It was not included as one of the main process steps as there was
no set time period within each cycle dedicated to wasting. No wasting was
undertaken during a track study.

The amount and frequency of sludge wasting was determined by taking regular
[MLSS] readings with the objective being to maintain the reactor within +/- 10% of
the desired operational [MLSS] of 3000 mg/L, (i.e. 2700-3300 mg/L). The desired
operational [MLSS] of 3000 mg/L was chosen following classification of the
wastewater as medium-high strength. With reference to Table 2.7-2 the
wastewater total COD while in cold storage was typically between 500-650 mg/L,
this corresponds to a medium-high strength wastewater. Typical design
parameters for sequencing batch reactors detailed in MetCalf and Eddy (2001)
give an operational MLSS range of 2000-5000 rng/L dependent upon factors
such as the COD concentration. The SRT was an independent parameter that
changed depending upon the relative growth rates of the bacteria.

51
2.9 REACTOR COIVIISSIONING
The reactors were run for nearly 18 months. The first 7 months were used for
development of the hardware. Following the developmental phase the reactors
were seeded and operated for six months before the official runs were
undertaken, during this time the reactors were operated at a range of dissolved
oxygen concentrations between 0.5 - 5.0 mg/L. The first run was initiated
subsequent to this six-month period using the acclimatized reactor biomass. The
reactors were operated continuously from the point of initial seeding to the point
of decommissioning.

There were no replicate runs undertaken in an attempt to duplicate the results,


this was largely due to time constraints. The experimental work experienced
considerable delays due to difficulties experienced during the research.
Difficulties included software bugs, hardware problems, and solving electrical
problems (particularly ground loops).

52
Chapter 3 EXPERIIVIENTAL TESTING

3.1 TESTING UNDERTAKEN


Table 3.1-1 illustrates the parameters that were monitored during the
experimental work.

Table 3.1 ~1 Chemical parameters monitored

Chemical parameters monitored

Ammonia nitrogen NH3-N


Nitrite nitrogen N02 -N
Nitrate nitrogen NOa-N
Total phosphorus TP •
Ortho phosphorus Ortho-P •
Chemical Oxygen Demand COD
Suspended Solids SS
(On-line) Oxi-Red potential ORP
(On-line) Hydrogen ion concentration pH
(On-line) Dissolved DO

* Measured in the influent wastewater only

All samples were taken while the reactor contents were fully mixed and were
vacuum filtered through Whatman glass fiber filters (GF/C) (grade C particle
retention 1.2 um) followed by Millipore griddled membrane filters made from
mixed esters of cellulose optimized for microbiological analysis of water (type HA
0.45 um). Retained activated sludge from the filtering process was returned to
the reactor. Certificates of quality for the filters used confirm Standard Methods
compliance. With the exception of MLSS and the on-line parameters, samples
were analyzed using HACH Test 'N Tube products. Readings were taken on a
HACH DR/2000 spectrophotometer as it provided consistently accurate results.
Table 3.1-2 provides a description of the analysis methods used.

53
Table 3.1 Chemical parameters analyzed using colorimetric methods

: Parameters tested with HACH


Test 'N Tube products Method Number Method Description Range Designed for testing
I PhosVer3 with acid persulfate Water - wastewater -
Total phosphorus 8190 digestion 0-3.5 mg/L sea water

Water - wastewater -
i
Orthophosphate (reactive) 8048 Ascorbic acid method 0-2.5 mg/L sea water

Water - wastewater -
Oxygen demand COD 8000 Colormetric measurement 0-1500 mg/L COD sea water

Total nitrogen TPN 10071 Persulfate digestion method 0-25 mg/L N Water· wastewater
Water - wastewater-
Ammonia 10031 HR Salicylate method 0-50 mg/L NH,-N sea water
0-150 mg/L NO,
Nitrite 8153 HR Ferrous sulfate method (0-45 mg/L N0 2-N l Water wastewater

HR Cadmium reduction Water wastewater -


Nitrate 8039 method 0-30 mg/L NOs-N sea water
'HR - High Range version of the tesl
• When sample concenlrations exceeded the test roMe they were diluted With dam/neret/sed water

The filtered sample sizes required for analysis on the DR/2000 varied between
0.2 25 mL. For a complete analysis of COD, TN, NH3-N, N0 2 -N, and N0 3-N an
initial sample size of approximately 70 mL would be withdrawn from the reactor
which would then be filtered before sub-samples for the various chemical
parameters were taken. Experimental work was usually limited to specific testing
runs where a single cycle would be intensively analyzed during a track study, i.e.
samples would usually be taken every 10 15 minutes with the main emphasis
upon COD and the various forms of nitrogen.

The test for TN involved the use of a HACH alkaline persulfate digestion method
known as TPN. The Total Persulfate Nitrogen (TPN) measurement is a test
offered by HACH that provides a measure of nitrogen similar to the traditional
Total Kjeldahl Nitrogen (TKN) test except the procedure is faster and eliminates
hazardous Kjeldahl-method materials. The HACH method is based upon the
procedure in Standard Methods for the Examination of Water and Wastewater
20 th Edition Method 4500-P ,p. 4-146. An evaluation of the TPN method was
undertaken by Langner and Hendrix (1982),

54
The TPN test is finding application in the monitoring of total nitrogen within
wastewater treatment facilities. The TPN test uses an alkaline persulfate
digestion to convert all nitrogen forms to nitrate nitrogen. The nitrate reacts with
chromotropic acid under strongly acidic conditions to form a yellow complex with
an absorbance maximum at 410 nm, measurement of the yellow complex is
undertaken on a spectrophotometer. The test measures all forms of nitrogen
such as soluble forms of ammonia, nitrate and nitrite as well as soluble and
insoluble forms of organic nitrogen (BIOlAB Scientific (2003) Product technical
information). Table 3.1-3 provides an illustration of the types of nitrogen
measured in both the TPN and other nitrogen tests used in this research.

Table 3.1-3 HACH nitrogen tests used in this research

TPN Test 'N Tube

Ammonia Test 'N

Nitrite Test 'N Tube

Nitrate Test 'N Tube

While the TPN test measured insoluble forms of nitrogen all samples in this
research were filtered to the 0.45 um level making it unlikely that insoluble
particulate matter was present in significant quantities.

Sampling for [MlSS] was undertaken on a daily basis. Sample sizes were
between 90 -100 ml; these were vacuum filtered through a Whatman glass fiber
filter (GF/C) (grade C particle retention 1.2 um) that had been oven dried at 110
°c for 24 hours prior to the test. The filtered samples were then dried at 110°C
for 24 hours and the MlSS was determined by calculating the retained/filtered
solids following standard methods 208 D.

55
3.2 SAMPLE PRESERVATION AND STORAGE
Samples were collected in clean plastic or glass bottles. The bottles for
phosphorus samples were acid cleaned with 1: 1 hydrochloric acid solution and
rinsed with demineralised water. Detergents containing phosphate were avoided
when cleaning sample bottles used in the tests.

All samples taken from the reactor were tested immediately. Some samples of
wastewater from Bromley wastewater treatment plant and from the feed cooler
were unable to be tested shortly after sampling and were preserved prior to
storage. The preservation was necessary to prevent degradation of the sampled
material and loss of test accuracy. Samples preserved by adding liquids had to
have adjustments made to their final test results to allow for the dilution
effects/volume changes of the original sample. It should be noted that
preservation was rarely undertaken due to the possible detrimental effects on
result accuracy and complications involved in preserving small sample sizes.

Duplicate analyses were undertaken on a "casual basis" to ensure confidence in


the results. A statistical analysis of the duplicate tests was not undertaken. The
following storage and preservation methods were used.

3.2.1 COD
For sample storage periods up to 28 days, sulfuric acid was added to the filtered
sample until the pH was less than 2.0. The sample was then placed in cold
storage at 4 °C or below. Immediately prior to analysis the preserved sample was
warmed to room temperature.

3.2.2 TPN
Preservation involved reducing the pH to 2 or less with concentrated sulfuric
acid. The samples were then stored at 4 °C or less for up to 28 days.
Immediately prior to analysis the samples were warmed to room temperature and
the pH was neutralized with 5 N sodium hydroxide.
56
3.2.3 NH3-N
The sample was preserved for up to 28 days by reducing the pH to 2 or less and
refrigerating at 4 DC or less. Immediately prior to analysis the sample was
warmed to room temperature and the pH was neutralized with 5 N sodium
hydroxide.

3.2.4 N03 -N
Samples could be preserved at 4 DC or lower, providing they were analyzed
within 48 hours. Prior to testing, the samples were warmed to room temperature.
For longer storage periods, the pH was adjusted to 2 or less with sulfuric acid
and the sample was refrigerated. Before testing the preserved sample was
warmed to room temperature and the pH was neutralized with 5.0 N sodium
hydroxide standard solution.

N02 -N,ORTHO AND


If samples were to be analyzed within 48 hours (24 hours for TP) then cold
storage at 4 DC or lower was satisfactory. For longer storage periods mercuric
chloride solution was added in the ratio of 4.0 mLl1000 mL of sample however
sample refrigeration was still required. Both methods required the samples to be
warmed to room temperature before any test work was undertaken. The samples
preserved with mercuric chloride had to have a sodium chloride level of 50 mg/L
or higher to prevent mercury interference in the test; samples low in chloride
could be spiked with approximately 1 mg sodium chloride.

3.3 ON-LINE MONITORING


On-line monitoring of ORP, pH, and [~O] was undertaken continuously with each
reactor having its own associated electrodes/probes "permanently" fitted. Results
were logged automatically on a dedicated data-logging PC. A real time plot was
also automatically produced on the logging PC allowing monitoring of biological
activity to be undertaken while a cycle was in progress. Readings for calculation
57
purposes were taken every six seconds while for plotting purposes every sixty
seconds.

3.3.10RP
Platinum band electrodes were used as described in the literature review. The
ORP electrodes were connected directly to the process control system. The ORP
probes were cleaned frequently (which involved wiping with a damp sponge),
Failure to regularly clean ORP probes can result in the variation rate and the
variation range of the ORP reading being affected (Paul et al (1998),
Collivignarelli and Bertanza (1999». The probes were also periodically checked
in quinhydrone buffer solution following the method specified by BIOlAB
Scientific (2003) Product technical information.

3.3.2 pH
The pH was measured on EDT (RH357Tx) microprocessor controlled meters
fitted with "field grade" glass body electrodes; both meters provided a mV output
which was fed to the data logging PC.

Both the pH and the dissolved oxygen meters had the ability to measure the
temperature within the reactors. This was checked periodically to ensure the
water bath was working correctly.

3.3.3 DO
The presence and concentration of dissolved oxygen within the reactors was
measured using YSI Model 57 dissolved oxygen meters (one for each reactor).
The probes used membrane-covered clark-type polaragraphic sensors which
were covered with a thin permeab!e membrane. High sensitivity membranes (half
the thickness) were used to allow accurate readings at low DO concentration
leve!s. The meters provided a mV output which was fed to the process contro!
system.

58
Chapter 4 EXPERIMENTAL RESULTS SUMMARY

4.1 INTRODUCTION TO EXPERIMENTAL RESULTS


The following section provides a summary of the results. A full set of results
including plots for operational variables such as [MLSS] and SRT are presented
in Appendices A 1 to A4.

The investigation of aerobic denitrification was undertaken in a series of periods


called runs, each run represented the operation of the reactors at a different
dissolved oxygen concentration. The following aspects should be noted before
review of the runs.

1) One operational objective was to maintain the mixed liquor suspended


solids concentration at 3000 mg/L +/- 10 % (i.e. 2700-3300 mg/L). It was
not possible to maintain a target SRT as it was dependant upon the
growth rate of the biomass. The slower biomass growth rates at lower
dissolved oxygen concentrations resulted in a considerable increase in
SRT. The effect of DO concentration on SRT can be illustrated in Figure
4.1-1 which shows the SRT required to achieve an effluent ammonia
concentration less than 1.0 mg/L (at 20°C) in a completely mixed activated
sludge system (adapted from Metcalf and Eddy (2001)).
35

30

25

'" 20
f-""
Iii 15

10

0
0 0.5 1.5 2.5 3
00 Concentration, mgll

Figure 4.1-1 Effect of DO concentration on SRT required to achieve effluent


ammonia concentrations less than 1.0 mg/L in a completely mixed activated
sludge system (adapted from Metcalf and Eddy (2001)).
59
2) The SRT profiles for the set points 1.0 mg/L and 0.5 mg/L (Figures 4.4-3
and 4.5-3) displayed almost a cyclic pattern. This pattern was the result of
the low MLSS growth rates and the effects of the wasting regime. The
amount of biomass wasted each event was kept similar to the previous
runs however the low dissolved oxygen concentration meant it took the
system allot longer to recover. The greater period between wasting events
resulted in a significant impact on the calculated SRT value and the
production of a cyclic (saw toothed) pattern on the SRT profile.

3) The wastewater temperature at Bromley was typically between 15-20 DC.


When a new load of feed was obtained the large thermal mass of the
wastewater (~450 L at 15°C) meant the storage freezer typically took 15-
20 hours to lower the temperature to the desired 3.8 DC. The implication of
a higher wastewater temperature during this period was that the influent
temperature to the reactor was also slightly higher (~24 °c as a result of
the water bath). It was noted that slightly higher MLSS growth rates
following the delivery of fresh wastewater were sometimes experienced. It
was thought that these higher growth rates may have been linked to the
warmer reactor temperatures during this period (a 10°C increase in
temperature can result in doubling of metabolic and growth rates (Metcalf
and Eddy (1991».

4) The settleability of the sludge was excellent throughout the entire period of
operation with the sludge volume index between 50 -100 mUg resulting in
a low effluent suspended solids content, (typically less than 5 mg/L
following a 15 minute settle period).

5) During each run an intensive chemical parameter testing program (a


series of track studies) was undertaken along with on-line monitoring of
ORP, pH, DO and air demand to provide real-time analysis of conditions.

60
The track studies included samples for TN, I\IH3-N, N0 2-N, N03-N, and
COD, which were taken periodically through out a cycle. The chemical
results were then plotted against the online profiles for each run in which
the operational dissolved oxygen concentration was the variable. All the
track study results are primarily limited to the aeration phase of a
treatment cycle.

6) The nitrogen loss calculations in this study were based upon the change in
[NH3-N], the possible nitrogen assimilated, the change in [N0 2-N] and the
change in [N03-N]. Organic nitrogen was not included. Measurements for
TPN (which includes organic nitrogen) were undertaken but only to
provide a check/confirmation of the nitrogen mass balance calculations.
An example calculation for the estimated assimilation is provided in
appendix A 1.1.

7) Nitrogen mass balance calculations required an estimate of assimilated


nitrogen to be undertaken. A literature review indicated that others
operating similar lab-scale batch reactors incorporating aerobic
denitrification have not incorporated nitrogen assimilation within their
nitrogen loss calculations (Munch et al (1996), Helmer and Kunst (1998),
Pochana and Keller (1999)). This is most probably because the nitrogen
loss to assimilation is considered small compared to that lost to biological
oxidation-reduction reactions. Research by Patureau et al (1998) found
that the amount of nitrogen lost to assimilation in a nitrifying sequencing
batch reactor was around 1%. Assimilation calculations in this research
were based upon the change in the volatile mixed liquor suspended solids
concentration and estimates of the likely nitrogen component of the new
cell material produced.

61
8) Inaccuracies in measurement of the MLSS concentration have contributed
to inaccuracies in the assimilation estimates. A series of MLSS
concentration measurements revealed the test was only accurate to -
200mg/L.

9) This research did not take into account the ammonium nitrogen produced
from organically bound nitrogen.

10) Depletion of the COD refers to the depletion of readily biodegradable


COD. While the COD results sometimes suggest the presence of
oxidisable products beyond the point of "COD depletion" these products
can most probably not be readily oxidised by the biomass.
Thayalakumaran et al (2003) also experienced a similar residual effluent
soluble COD. They operated a lab scale SBR in a nitrogen removal mode
treating meat processing effluent. They defined the residual COD as
"soluble non-biodegradable" as reflected by no further soluble COD
reduction following prolonged aeration in their system.

11 )With reference to the analysis of the pH profiles the pH scale had a 0.8
unit difference for the 4.0 mg/L and 2.5 mg/L DO set points. However the
1.0 mg/L and 0.5 mg/L set pOints have a 0.4 unit difference. This should
be noted before comparing plots from different runs. Prior to the
discussion of results examples of one track study from each operational
dissolved oxygen set point follow. For a full set of results refer to the
appendices.

12)While the ORP, pH, and airflow profiles were all evaluated for their
potential use as process control tools only the pH profile was eventually
used, that is an algorithm was developed and demonstrated to detect the
point of ammonia depletion on the pH profile (ammonia valley).

62
Chapters 4.2 - 4.5 provide a summary of the specific dissolved oxygen set
points. The "finer nuances" of the results are presented in Table 4.1-1.

Table 4.1-1 Summary of results DOSP 4.0 mg/L - 0.5 mg/L

2.5 12 12 2985 490 175 26 8 1 NO


1.0 11 28 3165 322 283 16 15' 2 5 YES

* The SRT for the aerobic denitrification process in this research ranged
between 15-29 days, this corresponds to values of 15-25 days reported by
Munch et al (1996) .

63
4.2 SUMMARY AND EXAMPLE OF OPERATIONAL DISSOLVED OXYGEN
CONCENTRATION 4.0 mg/L

The reactor was operated at a dissolved oxygen set point of 4.0 rng/L and
allowed to run for an acclimatization period of 14 days. Since the SRT for this run
was approximately 9 days this period was slightly less than two sludge ages.
Seven track studies were undertaken at dissolved oxygen set point 4.0 mg/L, the
track studies were undertaken over a one-month period following the 14-day
acclimatization period. The online profiles were consistent within 2-3 days (their
stability suggesting acclimatization of the system). This section provides a
summary and example of the results while Appendix A 1 contains a full set of
results. Figure 4.2-1 shows the [MLSS] over the duration of the run.

[MLSS]

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Figure 4.2~1 Mixed liquor suspended solids concentration DOSP 4.0 mg/L

64
The wasting rate required to keep the [MLSS] close to 3000 mg/L is illustrated in
Figure 4.2-2, while Figure 4.2-3 shows the SRT of the system.

M M ..,
Wastag e (L)
.., ..,
N N N N N • N NN N
t
N N
t
........... .... ...

o 5 10 15 20 25 30 35 40 45

OPERATIONAL DAYS FOR THE RUN

Figure Mixed liquor wastage DOSP 4.0 mg/L

SRT (days)

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Figure 4.2-3 Solids residence time DOSP 4.0 mg/L

65
Table 4.2-1 provides a summary of the seven track studies undertaken at DOSP
4.0 mg/L while Table 4.2-2 shows the average values corresponding to the
hydraulic residence time.
Table 4.2-1 Summary of track study data [operational dissolved oxygen] 4.0
mg/L
Parameter Units TS1 TS2 TS3 TS4 TS5 TSa 157 7Wii
[MLSS} mg/L 3300 3586 3115 3074 I 2972 3639 3179 3266
Initial NH,·N mg/L 20 22 20 19 21 20 19 20
Final NH,.N mglL 0 0 0 0 0 0 0 0
Del NH,·N (A) mglL 20 22 20 19 21 20 19 20

Inilial NO,.N mglL NO,·N 0 2 1 0 1 1 0 1


Final NO,.N mglL NO,·N 1 1 0 1 0 1 1 1
Del NO,·N (B) mglL NO,-N 1 -1 -1 1 -1 0 1 0

Initial NO,·N mgILNO,·N 4 5 5 6 6 5 8 6


Final NO,-N mglL NO,-N 16 23 20 20 24 18 21 20
Del NO,-N (C) mglL NO,.N 12 18 15 13 17 13 13 15

(D) (A-
Nitrogen loss (B+C» mg/LN 7 5 7 5 4 7 5 6

Estimated assimilated nitrogen (E) mg/L N 5 5 4 2 1 6 4 4

Unaccounted for nitrogen loss (F) (D-E) mglLN 2 0 3 3 3 1 1 2

Nitrification rate mglL.min N 0.17 0.21 0.14 0.18 0.16 0.14 0.19 0.17

mg/L.min.mg MLSS 5,E-{lS 6,1:...Q5 4,E",S 6.E-05 6.E..{)5 -'i.E-OS fLE-05 S.E"'S

Rate unaccounted N loss mg/L.mln N 2£..02 O.E+OO 2.E"'2 :.lE-02 2.E·02 6.E---03 O.E-03 1.E.oZ

% Initial TN removal % 29 17 25 19 15 27 17 21
% TN assimilated % 21 17 15 7 4 24 14 14
% Inilial TN loss unaccounted % 9 0 10 12 11 3 3 7

TPN Initial (G) mg/L N 27 31 29 29 30 29 30 29


TPN Finat (H) mg/LN 19 25 22 24 26 22 25 23
TPN loss (I) = (G-H) mglLN 8 5 7 5 4 7 5 6

COD Initial (J) mglL N 125 175 161 145 159 152 161 154
COD Final (K) mglL N 19 24 23 22 21 22 20 22
COD loss (L)=(J-K) mg/LN 106 151 138 123 138 130 141 132

Table 4.2-2 Summary of HRT values DOSP 4.0 mg/L

HRT days 0.2 0.2 0.3 0.2 0.2 0.2 0.2 0.2

66
With reference to Table 4.2-1 the average influent soluble nitrogen (TN) was 27
mg/L (not including soluble organic N, this figure composed of 20 mg/L NH3-N, 1
mg/L N0 2-N, and 6 mg/L N03-N, refer * Table 4.2-1). Fourteen percent of the
influent soluble nitrogen was assimilated into new cell tissue; the remainder was
mostly transformed to NOx-N (refer ** Table 4.2-1).

Seven percent (refer *** Table 4.2-1) of the initial total nitrogen could not be
accounted for in the nitrogen mass balance procedure.

The main cause of error came from the calculations regarding assimilation; these
required the measurement of [MLSS] at the start and end of the cycle. The
difference in the two readings was then taken to calculate the nitrogen
assimilation. A series of tests on mixed liquor (MLSS concentration 3000 mg/L)
showed that two tests on the same material had a difference of up to 200 mg/L.
The actual change in [MLSS] over the length of a cycle was usually 20-40 mg/L,
thus an accurate determination of assimilation was not possible. See discussion
5.1, it is possible the assimilation calculations also over estimated the amount of
nitrogen assimilated into new cell material.

It is unlikely that the unaccounted for nitrogen can be attributed to aerobic


denitrification due to the high dissolved oxygen concentration. The pH of the
reactor was also too low (-6.8-7.8) for air stripping to be a pathway for ammonia
loss.

The online real time plots for pH and air flow provided some useful features that
could be correlated to biochemical events. Figure 4.2-4 shows a typical pH
profile with the ammonia valley (point 1) indicating the point of ammonia
depletion, (all profiles taken from TS1 DOSP 4.0 mg/L).

67
7.7

1.S

7.5
TPN
7.4
:z:
7.3 '"
7.2

o 10 20 30 40 50 GO 70 80 90 100 110 120 130


SBR eyelelime (minute.)

Figure 4.2-4 pH and [soluble nitrogen]

The pH curves provided profiles with consistently clear features. The ammonia
valley, (as described in section 1.6), was so distinguishable and consistent that
an algorithm was developed so that the process control system could use the
feature to determine the point of ammonia depletion. The process control system
detected the point of ammonia depletion using the pH algorithm in 419/420
cycles giving a successful detection rate equal to 99.8% reliability. Figure 4.2-5
shows the air flow profile with ammonia and COD data.

T9
-'-B

:r ·7
C,
£ 20 6 '2
'j§
!if
a 5 ~
0 15 !:
~
Z
4 c
'"
:£ 10 3 ~
e. COD/5
2
5

o 0
o 10 20 30 40 50 60 70 80 90 100 110 120 130
SBR cycle time (minutes)

Figure 4.2-5 Air flow rate [NH3-N] and [COD]

The depletion of COD was indicated by a transition to a somewhat level gradient


on the air flow profile (point 1 Figure 42-5) while the depletion of ammonia

68
appeared as a drop in the air demand at the end of the cycle (point 2 Figure 4.2-
5). A visual inspection of the 420 profiles resulted in an estimate that the air flow
profile could have been used for determination of the point of COD and ammonia
depletion in 80% and 40% of the cycles respectively. The ammonia depletion
feature was inconsistent in that the feature was often not distinct and the shape
of the feature did not lend itself to be reliably detected by an algorithm. The 40%
detection rate would suggest the ammonia depletion feature is unlikely to provide
a reliable process control tool however it could be used for manual interpretation
by an experienced operator. No algorithms were developed in this research to
provide automated process control from the air flow profile. Figure 4.2-6 shows
the ORP and dissolved oxygen profiles.
9 200
8
ORP

7 -------------- 100
6
...J
a, ;;-
E
6'
5

4
~~~
DO
0 -
E
ll.
e. oc
0
3

2 -100

o 400
o 10 20 30 40 50 60 70 80 90 100 110 120 130
SBR cycle time (minutes)

Figure ORP and [DO]

The ORP profile did not show any distinguishing features that would enable it to
be used as a process control tool. The operation at a fixed dissolved oxygen
concentration meant that changes to ORP could only result from changes in
oxidation reduction potential resulting from nutrient depletion rather than by
changes in dissolved oxygen concentration (i.e. the ORP profile did not change
due to factors such as DO breakthrough).

69
4.3 SUMMARY AND EXAMPLE OF OPERATIONAL DISSOLVED OXYGEN
CONCENTRATION 2.5 mg/L

The reactor was set to a dissolved oxygen set point of 2.5 mg/L and allowed to
run for an acclimatization period of 19 days. Since the SRT for this run was
approximately 8 days this period was slightly greater than two sludge ages. As
with the previous DOSP consistent behavior (in terms of online profiles) was
established well before the end of the acclimatization period. Seven track studies
were undertaken at dissolved oxygen set pOint 2.5 mg/L. The track studies were
obtained over a one-month period following the 19-day acclimatization period.
This section provides a summary and example of the results while Appendix A2
contains a full set of results. Figure 4.3-1 shows the [MLSS] over the duration of
the run.

[MLSS] mg/L
~~
'IDID
, MM

;~
i<:t'<t
m~
C')i

,'<I'
, It)
'10
N

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Figure 4.3-1 Mixed liquor suspended solids concentration DOSP 2.5 mg/L

70
The wasting rate required to keep the [MLSS] close to 3000 mg/L is illustrated in
Figure 4.3-2 while Figure 4.3-3 shows the SRT of the system.

Wastage (L)
0 0
...; ...;

It) It) It)

NN N
""
N
0 0
N N
IX!
..-
It!
..... It!
.....
...."! "!
.... '<I;
....
<"!
.,...
q q q q
.... .,... .... ....

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Mixed liquor wastage DOSP 2.5 mg/L

SRT (days)

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Figure 4.3-3 Solids residence time DOSP 2.5 mg/L


71
Table 4.3-1 provides a summary of the seven track studies undertaken at DOSP
2.5 mg/L while Table 4.3-2 shows the average values corresponding to the
hydraulic residence time.

Table 4.3-1 Summary of track study data [operational dissolved oxygen] 2.5
mg/L
Units T51 T52 T53 -T54 T55 T56 T57 Ave
[MLSS] mglL 2504 2901 3620 3944 2834 3194 3594 3227

Initial NH,-N mglL 19 21 20 18 19 16 20 19


Final NH,·N mg/L 0 1 2 0 0 1 1 1
Del NH,·N (A) mglL 18 20 17 18 18 15 19 18

Initial NO,-N mg/L NO,-N 2 1 0 0 0 1 0 1


Final NO,.N mglL NO,-N 1 1 1 0 1 1 0 1
DeINO,·N (B) mg/LNO,.N -1 0 1 0 1 0 0 0

Initial NO,·N mglL NO,·N 5 6 5 2 6 4 4 4


Final NO,·N mg/L NOa·N 20 18 18 16 17 12 15 17
Del NO,·N (C) mg/L NOJ-N 15 13 13 14 12 8 11 12

(D)=(A.
Nitrogen loss (B+C» mglL N 5 8 3 4 6 7 8 6

Estimated assmilated nitrogen (E) mglLN 1 3 2 3 3 3 5 3

Unaccounted for nitrogen loss (F) = (D·E) mg/LN 3 5 1 1 3 4 3 3

Nitrification rate mglLmin N 0.12 0.12 0.13 0,16 0.14 0,14 0,11 0.13
mg/L,min,mg
MLSS 5.f.,OS 4.E-05 4.E-05 4.E-OS 5,E-05 4,E-05 3.E,05 4,E,05

Rate unaccounted N loss mglL.min N 2.E-02 3,E..[)2 1,E-02 3.E-03 2.E-D2 3£"[)2 2.E-02 2,E-02

% Initial TN removal % 18 27 13 22 24 34 34 25
% TN assimilated % 5 10 8 15 11 16 20 12
% Initial TN loss unaccounted % 13 17 5 7 12 18 14 12

TPN Initial (G) mglL N 29 31 28 23 27 23 27 27


TPN Final (H) mglL N 24 23 25 19 21 16 19 21
TPN loss (I) = (G·H) mglL N 5 8 3 4 6 7 8 6

COD Initial (J) mg/LN 149 180 147 121 128 115 170 144
COD Final (K) mg/LN 19 19 28 12 19 22 19 20
COD loss (L)=(J·K) mglLN 130 161 119 109 109 93 151 125

72
Table 4.3-2 Summary of HRT values DOSP 2.5 mg/L

HRT days 0.2 0.3 0.3 0.2 0.2 0.2 0.3 0.2
8.4

With reference to Table 4.3-1 the average influent soluble nitrogen (TN) was 24
mg/L (not including soluble organic N, this figure composed of 19 mg/L NH 3-N, 1
mg/L N0 2 -N, and 4 mg/L N0 3 -N, refer * Table 4.3-1). Twelve percent of the
influent soluble nitrogen was assimilated into new cell tissue; the remainder was
mostly transformed to NOx-N (refer ** Table 4.3-1).

Twelve percent (refer *** Table 4.3-1) of the initial total nitrogen could not be
accounted for in the nitrogen mass balance procedure, the unaccounted for
nitrogen was within the accuracy limits of the mass balance calculations. It is
unlikely the unaccounted for nitrogen can be attributed to aerobic denitrification
as the dissolved oxygen concentration was maintained at 2.5 mg/L at all times
and the pH was between 6.8-7.8 which was too low for air stripping of the
ammonia to have occurred. In comparison to the previous run the average time
taken for the complete oxidation of the ammonia increased from 153 minutes to
175 minutes. Thus a 1.5 mg/L drop in the operational dissolved oxygen
concentration increased the aeration time by nearly 15% or 22 minutes.

The online real time plots for pH and air flow continued to display some useful
features that were correlated to biochemical events. Figure 4.3-4 shows a typical
pH profile with the ammonia depletion feature (point 1). All the profiles were
taken from track study six (TS6) DOSP 2.5 mg/L.

73
30 7.7
pH

7.6
::J 25
l
Z
7.5
~ 20
7.4
~ TPN
~ 15 7.3 X
Q.

~
o 7.2
~ 10

7.1

7.0

6.9
20 40 60 80 100 120 140
SBR cycle time (minutes)

Figure 4.3-4 pH and [soluble nitrogen]

The pH profiles were consistent and the ammonia valley feature (point 1 Figure
4.3-4) was always clear and well defined. During the run the SBR completed
approximately 370 cycles. The process control system using the ammonia valley
detection algorithm correctly identified the valley (and terminated the aeration) in
360/370 cycles giving a successful detection rate equal to 97% reliability. The 10
cycles where the algorithm failed to detect correctly the point of ammonia
depletion were due to either a flat slope following the valley or an earlier pH spike
(with false trigger) caused by noise in the signal. Figure 4.3-5 shows the air flow
profile with ammonia and COD data.

30 -

~ 20
.§. c:
!i?
a
0 15
5!
;:
!d.
4,ll
~
I10 3

coo

~--~--~--~----~--~--~-L~o
20 40 60 80 100 120 140
SBR eyet.lime (minutes)

Figure 4.3-5 Air flow rate [NH3-N] and [COD]


74
The depletion of COD was indicated by the transition to a level gradient (point 1
Figure 4.3-5) and ammonia depletion was indicated by a sudden fall in the air
flow at the end of the cycle (point 2 Figure 4.3-5). A visual inspection of the 370
cycles resulted in an estimate that the air flow profile could have been used for
determination of the point of COD and ammonia depletion in 80% and 40% of the
cycles respectively.

The ammonia depletion feature remained relatively inconsistent in that the


feature was not of a repetitive shape and was often not that distinct. For this
reason an algorithm is unlikely to be able to reliably use the profile to detect the
point of ammonia depletion as a consistent predictable shape is a prerequisite of
successful automated process control. Figure 4.3-6 shows the ORP and
dissolved oxygen profiles.
9 200
8

7
ORP 100
6
J
.s""5 ~
o
5'4
e.
3 DO
2

0 +----,-----r--.----.---,---,--~_+ -200
0 20 40 60 80 100 120 140
SBR cycle time (minutes)

Figure 4.3-6 ORP and [DO]

The ORP profile did not show any distinguishing features that could be correlated
to the depletion of nutrients. The response of the ORP does not appear to
correspond directly with features on the dissolved oxygen profile (as one would
expect). For example the change of the ORP from a negative to a positive slope
(point 1 Figure 4.3-6) occurred 18 minutes after the initiation of aeration, also the

75
ORP apex (point 2 Figure 4.3-6) appeared to occur slightly before the dissolved
oxygen concentration actually reached zero or before anoxic conditions had
developed. Reasons for this may include the dissolved oxygen probe not
accurately reflecting the true dissolved oxygen concentration (out of calibration),
and/or the ORP probe itself may have been fouled slightly leading to a slow
response.

76
4.4 SUMMARY AND EXAMPLE OF OPERATIONAL DISSOLVED OXYGEN
CONCENTRATION 1.0 mg/L

The reactor was operated at a dissolved oxygen set point of 1.0 mg/L and
allowed to run for an acclimatisation period of 15 days. Since the SRT for this run
was approximately 15 days this period was equivalent to one sludge age. Seven
track studies were undertaken at dissolved oxygen set point 1.0 mg/L, the track
studies were undertaken over a one-month period following the 15-day
acclimatisation period. Following the switch to DOSP 1.0 mg/L the reactor
displayed consistent conditions in terms of the online parameters within 3 days.
The acclimatization period of 15 days was considered adequate before sampling
analysis was undertaken. This section provides a summary and example of the
results while Appendix A3 contains a full set of results. Figure 4.4-1 shows the
[MLSS] over the duration of the run.

;
,':
[MLSS] mg/L " ;

, - _... ,-, --·T·-~~···

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Figure 4.4-1 Mixed liquor suspended solids concentration DOSP 1.0 mg/L

77
The wasting rate required to keep the [MLSS] close to 3000 mg/L is illustrated in
Figure 4.4-2, while Figure 4.4-3 shows the SRT of the system.

o
,..;

1
wastage (l) 0 0 <.:>
N N N
'Xl
If!
..... If!
....
--
q q
..... ..-

I I

o 5 10 15 20 25 30 35 40 45

OPERATIONAL DAYS FOR THE RUN

Figure 4.4-2 Mixed liquor wastage DOSP 1.0 mg/L

SRT (days)

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Figure 4.4-3 Solids residence time DOSP 1.0 mg/L

78
Table 4.4-1 provides a summary of the seven track studies undertaken at DOSP
1.0 mg/L while Table 4.4-2 shows the average values corresponding to the
hydraulic residence time.

Table 4.4-1 Summary of track study data [operational dissolved oxygen] 1.0
mg/L
~q,q",o.o,
Units TS1 TS2 TS3 TS4 TS5 TS6 TS7 Ave
[MLSS] mg/L ""~ i 3008 3105 3469 3012 3507 3793 3258

Initial NH3-N mglL 20 21 18 20 21 19 18 20


Final NH,-N mg/L 0 1 0 0 1 1 1 1
Del NH3-N (A) mg/L 19 20 18 20 20 18 17 19

Initial NO,-N mg/LNO,-N 3 2 2 3 2 2 2 2


Final NO,-N mg/L NO,-N 6 5 5 6 4 5 6 5
Del NO,-N (B) mg/L NO,-N 3 2 3 4 3 3 4 3

Initial NO,-N mg/L NO,-N 3 5 3 5 4 2 4 4


Final NO,.N mg/L NO,-N 10 11 7 11 10 10 9 10
Del NO,-N (C) mg/L NO,-N 6 6 4 6 6 8 5 6

(D) = (A-
Nitrogen loss (B+C)) mg/L N 10 11 11 11 11 8 B 10

Estimated assmilaled nitrogen (E) mg/LN 2 2 2 2 5 4 2 3

Unaccounted for nitrogen loss (F) = (O-E) mg/LN B 10 9 9 6 3 6 7

NitrifIcation rate mg/LmlnN 0.08 0.09 0.07 0.08 0.08 0.08 0.07 0.08

mg/Lmln.mg MLSS 3.E-05 3£-05 2,E·05 H·05 3.E..oS Z.E-OS 2.E-05 2.E-05

Rate unaccounted N loss mg/L.min N 3,E'()2 5,E-02 4 E-02 3£·02 3,E-02 1.E-02 2.E-02 3.E·02

% Initial TN removal % 38 41 47 38 44 33 32 39
% TN assimilated % 8 6 8 8 19 18 7 11
% Initial TN loss unaccounted % 31 34 39 31 24 15 25 28

TPN Initial (G) mg/L N 28 31 26 31 29 26 27 28


TPN Final (H) mg/LN 18 19 15 20 17 18 19 18
TPN loss (I)" (G-H) mg/LN 10 12 12 11 12 8 8 10

COD Initial (J) mg/L N 130 182 144 169 180 161 155 160
COD Final (I<) mg/LN 18 26 24 26 23 23 20 23
COD loss (L)=(J-K) mg/L N 112 156 120 143 157 138 135 137

Table 4.4-2 Summary of HRT values DOSP 1.0 mg/L

HRT days 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4

79
With reference to Table 4.4-1 the average influent soluble nitrogen was
approximately 26 mg/L (not including soluble organic N, this figure composed of
20 mg/L NH 3-N, 2 mg/L N0 2-N, and 4 mg/L N0 3-N, refer * Table 4.4-1). There
was significant nitrogen lost from the system which could not be accounted for
via the nitrogen mass balance and there was a build up of N0 2-N suggesting a
possible inhibition of the nitrification process (elevated nitrite levels often
accompany aerobic denitrification activity (Ho (1994), Yoo et al (1999)).

Eleven percent of the influent soluble nitrogen was assimilated into new cell
tissue; the remainder was mostly transformed to NOx-N. Twenty eight percent
(refer *** Table 4.4-1) of the initial total nitrogen could not be accounted for in the
nitrogen mass balance procedure. The build up of N0 2-N, and the unaccounted
for nitrogen loss suggests that aerobic denitrification activity was present in the
system.

The time taken to complete the oxidation of the ammonia increased from 175 to
283 minutes. Thus a further 1.5 mg/L drop in the operational dissolved oxygen
concentration increased the aeration time by nearly 62% or 108 minutes.

The online profiles for pH and air flow continued to provide some features that
were correlated to the depletion of ammonia and COD. Figure 4.4-4 shows a
typical pH profile with the ammonia valley (point 1) indicating the point of
ammonia depletion, (all profiles taken from TS6 DOSP 1.0 mg/L).

80
30 7.60

::r 7.55
'i3> 25
.s
za. 7.50
t:. 20
7.45
5?
0
~ 15
NH,-N pH
7.40 t
0
Z
. 7.35

--
10
~
7.30
5? NO,·N
-£ 5
~ 7.25

a 20 40 60 60 100 120 140 160 160 200 220 240 260 260 300
SBR cycle time (minutes)

Figure 4.4-4 pH and [soluble nitrogen]

The online pH profiles consistently illustrated the ammonia valley feature, (point 1
Figure 4.4-4). The process control system correctly identified the ammonia valley
in 214/228 cycles giving a successful detection rate equal to 94% reliability.
Figure 4.4-5 shows a typical air flow profile with ammonia and COD data.

30

25

~
OJ
.s 20 6

~
0
5
"
'E
=l
g 15

'"~"
0
4
5?
-£ 10
t:.
2
5

+~~~~~,---~~~~~~J····T~-+O
o 20 40 60 80 100 120 140 160 160 200 220 240 260 260 300
SBR cycle time (minutes)

Figure 4.4-5 Air flow rate [I'JH3-N] and [COD]


The depletion of COD (point 1 Figure 4.4-5) and ammonia nitrogen (point 2
Figure 4.4-5) were indicated by a transition to a level gradient and a drop in the
air demand respectively. A visual inspection of the 228 cycles resulted in an
estimate that the profile could have been used to identify the points of COD and
ammonia depletion in 80% and 30% of the cycles respectively. The lower

81
operational dissolved oxygen set point resulted in the ammonia depletion feature
being less distinct. Figure 4.4-6 shows the ORP and dissolved oxygen profiles.
9 200

7 OR? 100
6
2'OJ >'
g 5

6' 4
0 -a.E
e. 0::
0
3

2 ·100

o 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
SBR cycle time (minutes)

Figure 4.4-6 ORP and [DO]

As with the previous operational dissolved oxygen set points the ORP profile was
devoid of features (correlated to either the transformation or the depletion of
nitrogen or organic carbon) that would enable it to be used as a process control
tool.

82
4.5 SUMMARY AND EXAMPLE OF OPERATIONAL DISSOLVED OXYGEN
CONCENTRATION 0.5 mg/L

The reactor was set to a dissolved oxygen set point of 0.5 mg/L and allowed to
run for an acclimatization period of 15 days. The SRT for this run was
approximately 29 days thus the acclimatization period was equivalent to half a
sludge age (the SRT was double the previous SRT value experienced at DOSP
1.0 mg/L). Consistent behavior (in terms of online profiles) was established within
3 days following the decrease in the DOSP from 1.0 mg/L to 0.5 mg/L. Seven
track studies were undertaken at dissolved oxygen set point 0.5 mg/L. The track
studies were obtained over a one-month period following the 29-day
acclimatization period. This section provides a summary and example of the
results while Appendix A4 contains a full set of results. Figure 4.5-1 illustrates the
[MLSS] over the duration of the run.

[MLSS] mg/L """


(D

~~
""" .\
!;i!<'),/'
~i
en
....
I()

M ,
,
,.;.
~
~/

~
en
~m
il()
."""
\C'I/
,/
"

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Figure 4.5-1 Mixed liquor suspended solids concentration DOSP 0.5 mg/L

83
The wasting rate required to keep the [MLSS] close to 3000 mg/L is illustrated in
Figure 4.5-2, while Figure 4.5-3 illustrates the SRT of the system.

Wastage (L)

o 5 10 15 20 25 30 35 40 45

OPERATIONAL DAYS FOR THE RUN

Figure 4.5-2 Mixed liquor wastage DOSP 0.5 mg/L

SRT (days)

o 5 10 15 20 25 30 35 40 45
OPERATIONAL DAYS FOR THE RUN

Figure 4.5-3 Solids residence time DOSP 0.5 mg/L

84
Table 4.5-1 provides a summary of the seven track studies undertaken at DOSP
0.5 mg/L while Table 4.5-2 illustrates the average values corresponding to the
hydraulic residence time.

Table 4.5-1 Summary of track study data [operational dissolved oxygen] 0.5
mg/L
Q.a",a,a. Units TS1 TS2 TS3 TS4 TS5 TS6 TS7 Ave
[MLSS] mg/L 3160 80 2953 3343 3344 3256 3195

Initial NH,-N mg/L 21 22 21 20 20 18 23 21


Final NH,-N mg/L a 0 1 0 0 1 0 0
Del NH,-N (A) mg/L 21 22 20 20 20 17 23 20

Initial NO,-N mglL N02-N 5 4 4 6 5 6 5 5


Final NO,.N mg/L N02-N 11 10 8 11 11 10 10 10
Del NO,-N (S) mgILN02-N 6 6 4 5 6 5 5 5

Initial NO,-N mg/L NO,-N 3 1 2 1 2 2 2 2


Final NO,-N mg/L NO,-N 5 3 3 2 3 2 2 3
Del NO,-N (C) mglL NO,-N 2 2 1 1 1 0 1 1

(D)=(A.
Nitrogen loss (B+C» mg/LN 12 14 15 14 13 12 17 14

Estimated assmilated nitrogen (El mglLN 2 3 3 4 2 5 3 3

Unaccounted for nitrogen loss (F) = (D.E) mg/LN 11 11 12 10 11 7 14 11

Nitrification rate mg/L.min N 0.05 0.06 0.05 0.05 0.05 0.04 0.06 0.05
mg/L.min.mg
MLSS 2.E·Q5 2.E·05 2.E·05 2.E·05 2.E·05 1.E·OS 2.E·05 2.E·05
i
Rate unaccounted N loss mg/L.min N 3.E·02 3.E·02 3.E'{)2 2,E·02 3.E·Q2 1.E.02 i 3.E·02 3.E-02

% Initial TN removal % 43 52 54 52 49 46 57 50
% TN assimilated % 6 11 11 15 9 19 12 11
% Initial TN loss unaccounted % 38 41 43 37 41 28 46 40

TPN Inilial (G) mglLN 31 30 30 29 29 29 32 30


TPN Final (H) mglLN 19 16 15 15 16 16 15 16
TPN loss (I) (G·H) mg/LN 13 14 15 14 13 13 17 14

COD Inilial (J) mg/LN 159 171 136 152 161 106 170 151
COD Final (K) mg/LN 25 23 26 28 23 20 23 24
COD loss (L)=(J·K) mg/LN 134 148 110 124 138 86 147 127

Table 4.5-2 Summary of HRT values DOSP 0.5 mg/L

Units TS1 T52 1S3 154 TSS TS6 rs7 Ave


Cydeilme minutes 434 441 421 470 424 452 471 445
HRT days 0.6 0.6 0.6 0.7 0.6 0.6 0.7 0.6

85
With reference to table 4.5-1 the average influent soluble nitrogen was
approximately 28 mg/L (not including soluble organic N, this figure composed of
21 mg/L NH3-N, 5 mg/L N0 2-N, and 2 mg/L N03-N, refer * Table 4.5-1). Eleven
percent (refer ** Table 4.5-1) of the influent soluble nitrogen was assimilated into
new cell tissue; the remainder was mostly transformed to NOx-N in particular
there was a significant increase in the N0 2-N concentration. Forty percent (refer
*** Table 4.5-1) of the initial total nitrogen could not be accounted for in the
nitrogen mass balance procedure.

The significant increase in the amount of unaccounted for nitrogen and the
increase in nitrite nitrogen would suggest an increase (relative to DOSP 1.0
mg/L) in aerobic denitrification activity at the 0.5 mg/L dissolved oxygen level.
There was a significant increase in the time required for the oxidation of the
ammonia. In comparison to the previous run the time taken for the complete
oxidation of the ammonia increased from 283 minutes to 445 minutes. Thus a 0.5
mg/L drop in the operational dissolved oxygen concentration increased the
aeration time by nearly 57% or 162 minutes.

The online real time plots for pH and air flow displayed some useful features that
were correlated to biochemical events. Figure 4.5-4 shows a typical pH profile, all
the profiles were taken from TS1 DOSP 0.5 mg/L.

7.70

7.65

7.S0

7.55

7.45

7.40
------"
7.35

7.30
o 100 200 300 400
SBR cycle time (minutes)

Figure 4.5-4 pH anq [soluble nitrogen]

86
The pH profiles were consistent and the ammonia valley feature, (Point 1 Figure
4.5-4) was well defined. During the run the SBR completed 145 cycles. The
process control system using the ammonia valley detection algorithm correctly
identified the valley in 133/145 cycles giving a successful detection rate equal to
92% reliability. Figure 4.5-5 shows the air j:low profile along with the profiles for
ammonia and COD.
30 9

8
25
7

:J' 20 6
OJ .::
.§.
5 ~
5' 15 ;J:
o 4,g·
2. NH,·N ...
~ 10 3 :.{
£
~
2
5
, 1

o 50 100 150 200 250 300 350 400 450


SBR cycle time (minutes)

Figure Air flow rate, [NH3-N] and [COD]

The depletion of COD (point 1 Figure 4.4-5) was indicated by a transition to a


somewhat level gradient (There was no feature identifying the point of ammonia
depletion in this figure). A visual inspection of the 145 cycles resulted in an
estimate that the profile could have been used to identify the point of COD and
ammonia depletion in 80% and 15% of the cycles respectively. The results
suggest that the ammonia depletion feature becomes less apparent as the
operational dissolved oxygen set point is lowered. Figure 4.5-6 illustrates the
ORP and dissolved oxygen profiles.

87
9 200

8
ORP
7 100
6
:;;:-
~ E
§. 5 0
Q.
o 4 0::
0
8.
3
·100
2

DO

0 I ·200
0 50 100 150 200 250 300 350 400 450
SBR cycle time (minutes)

Figure 4.5-6 ORP and [~O]

ORP failed to show any distinguishing features that would enable it to be used as
a process control tool.

88
Chapter 5.0 DISCUSSION

5.1 (Objective 1) Confirm the presence of aerobic denitrification activity.

Operation of the reactors at the 4.0 mg/L and 2.5 mg/L dissolved oxygen level
resulted in a 95-100% transformation of ammonia nitrogen to nitrate nitrogen. At
these dissolved oxygen concentrations the presence of nitrite nitrogen was < 1
mg/L and the nitrogen mass balance procedure was able to account for
approximately 90% of the influent nitrogen. That is 90% of the influent nitrogen
went to known forms and could be accounted for with the mass balance
procedure. However once the dissolved oxygen concentration was lowered to 1.0
mg/L and 0.5 mg/L the nitrogen mass balance identified significant losses from
the system. At these dissolved oxygen levels 30% - 40% of the influent nitrogen
could not be accounted for with the nitrogen mass balance procedure. That is the
loss of ammonia nitrogen could not be balanced by an increase in nitrite, nitrate,
or assimilated nitrogen.

The nitrogen mass balance typically assumed ~10% of the influent nitrogen was
assimilated during each cycle. This estimate is higher than that reported by
Patureau et al (1998) who found the amount of nitrogen lost to assimilation in a
nitrifying sequencing batch reactor was around 1%. The removal of nitrogen from
the system under aerobic conditions suggests aerobic denitrification activity was
present, (volatilization of ammonia was not possible as the pH was too low at
around 7.0).

While there was minimal nitrite nitrogen present when the dissolved oxygen
concentration was maintained at 4.0 mg/L and 2.5 mg/L the nitrite nitrogen levels
were as high as 6 mg/L and 11 mg/L when the dissolved oxygen concentration
was lowered to 1.0 mg/L and 0.5 mg/L respectively. The presence of elevated
nitrite levels is another indicator that aerobic denitrification was occurring. Ho

89
(1994) suggested that aerobic denitrification may be accompanied by the
inhibition of the second step of nitrification involving the oxidation of nitrite to
nitrate. Elevated levels of nitrite nitrogen are common in aerobic denitrification
processes and have been reported by other researchers such as Randal and
Buth (1984), Hanaki and Wantawin (1990), Ho (1994), Munch et at (1996), Yoo
et al (1999), and Delgenes and Patureau (2004b)).

A likely explanation for the nitrite build up is that Nitrobacter can be inhibited by
low dissolved oxygen concentrations. With reference to double Monod kinetics
the half saturation coefficient for dissolved oxygen is 0.25-0.5 mg/L for
Nitrosomonas, this is lower than that of 0.72-2.84 for Nitrob a cter. «Randal et at
(1992), Munch et al (1996), Helmer and Kunst (1998)).

This means that Nitrobacter is more susceptible to low dissolved oxygen


concentrations than Nitrosomonas, and operation at a dissolved oxygen
concentration of 1.0 mg/L and 0.5 mg/L could have inhibited the Nitrobacter
activity leading to a build up of nitrite. Munch et al (1996) operated a lab scale
aerobic denitrification SBR and found that nitrite accumulated within the system
when the oxygen concentration was maintained lower than the half saturation
coefficient of nitrobacter.

In summary the presence of aerobic denitrification activity at dissolved oxygen


concentrations of 1.0 mg/L and below was supported by the development of
elevated nitrite levels and from mass balance calculations.

90
5.2 (Objective 2) Elucidate some operational aspects of aerobic denitrification, in
particular comment on the nitrification, denitrification, and sludge production
rates.

5.2.1 Nitrification rates


With reference to Figure 5.2-1 the biokinetic nitrification rates were considerably
lower when the operational dissolved oxygen was :s; 1.0 mg/L (the range required
for aerobic denitrification). A traditional nitrification process is likely to be
operated at> 2.0 mg/L and is likely to have a nitrification rate three times that of
a system operated around the 0.5 mg/L aerobic denitrification range.

5.0 0.18
0.16
:i:f 4.0 0.14 2'2
.sI!II
"'!II ~ .-e;
~:J 0.12 c:
,g::s 3.0
0.10
0...1
+:i_
::lz
" ""
'u_
~:z
::: til 2.0
0.08 E~
z~
z§. 0.06

1.0 0.04
0.02
0.0 +----,.-----,---.----.----1- 0.00
o 2 3 4 5
DOSP(mgIL)

Figure Nitrification rate versus dissolved oxygen concentration

The decrease in the nitrification rate translates directly into increased aeration
time for aerobic denitrification processes. Table 5.2-1 and Figure 5.2-2 illustrate
the increase in aeration time required as the operational dissolved oxygen
concentration was lowered.

91
Table 5.2-1 Aeration time relative to dissolved oxygen concentration

oxygen set Average dissolved oxygen -Relative increase in


point aeration time concentration treatment time

4.0 153
2.5 175 1.5 14
1.0 283 1.5 60
0.5 445 0.5 60

The rate of nitrification experienced at the 0.5 mg/L aerobic denitrification level in
this experimental work corresponds well with that reported by Demoulin et al
(2001) who investigated the 90,000 p.e. aerobic denitrification plant in Potsdam
Germany. They reported a nitrification rate of 1.1 mg N/g MLSS.h. This
experimental work had a slightly higher value at 1.2 mg N/g MLSS.h. (1.4 mg N/g
MLVSS.h, tests showed the biomass was approximately 84% volatile).

3.5

~----

0.0 0.5 1.0 1.5 2.0 2.5 3,0 3.5 4,0 4.5
DOSPmg/L

Figure 5.2-2 Extension in aeration time required to oxidise ammonia nitrogen


(relative to DOSP 4.0 mg/L)

Munch et al (1996) reported a nitrification rate of 1.47 mg N/g MLVSS.h for an


aerobic denitrification system in which the average dissolved oxygen
concentration during the aeration phase was 1.2 mg/L.

92
Metcalf and Eddy (2001) reported the rates of both nitrification and denitrification
in aerobic denitrification processes are likely to be a function of the reaction
kinetics, floc size, floc density, floc structure, rbCOD loading and bulk DO
concentration. Because of the complex physical factors the process is still to be
modeled accurately. However Metcalf and Eddy (2001) reported nitrification rates
for the aerobic denitrification process to be lower than conventional processes.
This is thought to be due to the lower dissolved oxygen concentrations and
because only a portion of the nitrifying bacteria contained within the flocs are
active.

Metcalf and Eddy (2001) reported typical nitrification rates for conventional plants
being in the order of 0.78 - 21 mg N/g MLVSS.h. It would appear that the
nitrification rates for aerobic denitrification facilities are at the extreme lower end
of the values reported for conventional processes.

5.2.2 Denitrification
Full denitrification was not achieved in this work. The aerobic denitrification
process only managed -50% nitrogen removal (including any assimilation). If the
50% nitrogen removal value is converted into a rate it would equate to
approximately 0.71 mg N/g MLVSS.h

It is possible the incomplete nitrogen removal was a result of a low COD/N ratio
and/or the complexity of the wastewater organic carbon compounds. Research
by Third (2004) investigated aerobic denitrification within sequencing batch
reactors and suggested it was a heterotrophic process which relied on the
storage of carbon polymers (for use as intra-cellular carbon) for the reduction of
oxidized nitrogen. Third (2004) found that the quantity and types of storage
polymers depended upon the quantity and type of organic carbon present in the
feed. That is simple organic carbon compounds such as glucose or acetate (not
typically present in wastewater but present in synthetic experimental feed) are

93
more easily stored as polymers and produce simple polymers such as PHB.
Barker and Dold (1996), Wentzel and Ekama (1997) suggested a COD/N ratio
>12 is required to adequately support the denitrification process while Third
(2004) suggested a COD/N ratio (in influent feed) of >10 to achieve complete
aerobic denitrification. This research used a domestic wastewater that typically
had a COD/N ratio of around 5. It is possible the low COD/N ratio combined with
aspects such as raw wastewater producing storage polymers that did not easily
hydrolyze and release "reducing power" may be contributing factors to the
incomplete nitrogen removal achieved in this research.

With respect to the values reported by Demoulin et al (2001) the German


Potsdam facility is required to comply with E.C. directive 91/271 which requires
strict effluent standards in particular for nutrients. Since the nitrification rates for
the Potsdam aerobic denitrification facility were essentially the same as those
experienced in this work, assuming a 90% nitrogen removal objective for the
Potsdam plant the rate of denitrification can be estimated as a value of 0.9 x 1.1
mg Nlg MLSS.h =1.0 mg N/g MLSS.h (or approximately 1.2 mg N/g MLVSS.h).

lVIetcalf and Eddy (2001) reported values for conventional denitrification range
from 2.5 3.8 mg Nlg MLVSS.h for a single sludge anoxic reactor under non
limiting conditions (at 20°C) and 0.56-1.6 mg N/g MLSS.h for systems using
endogenous carbon sources. Values for aerobic denitrification processes have
yet to be established however Metcalf and Eddy (2001) reported aerobic
denitrification process would have lower denitrification rates than conventional
processes because of substrate consumption in the aerobic portion of flocs.

It would appear that the rate of denitrification for the aerobic denitrification
process is less, probably around half of that experienced in traditional separate
stage anoxic denitrification facilities. From an experimental perspective there are
conflicting reports with respect to aerobic denitrification rates. Patureau et a/

94
(1998) suggested the rate of aerobic denitrification will largely be limited by the
rate of nitrification at the low dissolved oxygen concentration. Neef et al (1996)
indicated that micro-organisms capable of aerobic denitrification may have
slower denitrification rates than anoxic denitrifying bacteria. However Delgenes
and Patureau (2004a) identified numerous strains of microorganisms able to
aerobically denitrify. Upon further examination they found the aerobic denitrifying
performances of the consortium were comparable to traditional anoxic
denitrifying performances.

A summary of the nitrification and denitrification rates for both traditional and
aerobic denitrification facilities is shown in Table 5.2-2. These values are
approximates sourced from this experimental work and calculated from Munch et
al (1996), Metcalf and Eddy (1991), and Demoulin et al (2001).

Table 5.2-2 Comparison of biochemical transformation rates between


conventional and aerobic denitrification processes.
-Nitrification rate -Denitrification rate

mg N/g MLSS.h mg N/g MLSS.h

Aerobic denitrification process 1.0 1.0


Conventional N/DN process <21 <4

With reference to Table 5.2-2 the lower rates of biochemical transformation for
the aerobic denitrification process suggest the overall treatment reactor volume
will need to be larger. This finding is supported by Metcalf and Eddy (2001) who
reported aerobic denitrification was only suitable for processes with sufficient
volume to accommodate the lower nitrification and denitrification rates. That is
aerobic denitrification will require a longer detention time and larger treatment
reactors than a conventional process.

95
This conflicts with some reports from full scale aerobic denitrification facilities
which suggest an overall saving in treatment capacity. For example Collivignarelli
and Bertanza (1999) partitioned off the equivalent of 2,500 p.e. of a full-scale
440,000 p.e. facility and converted it into an aerobic denitrification process. A two
year trial showed the aerobic denitrification process offered a possible 20%
saving in treatment tank volume compared to the conventional process (when
results are translated on a flow weighted basis). This conclusion is also
supported by the research of Demoulin et al (1997) who compared the efficiency
of the aerobic denitri'fication process to a conventional nitrification denitrification
process. Parallel operations lasting one year were undertaken at the Austrian
GroBarl WWTP in 1995. It was found that the aerobic denitrification process
demonstrated superior treatment efficiency when operated at the same loading
conditions as the conventional process. Both nitrogen and phosphorus removals
were consistently around 30% higher in the aerobic denitrification batch process.
It was estimated that the aerobic denitrification process offered a volumetric
saving of about 30% when compared to the conventional treatment technology.

However it is difficult to compare the results of Collivignarelli and Bertanza (1999)


and Demoulin et al (1997) with the experimental data from this research as their
papers appear to lack firm design data (such as comparative biochemical
transformation rates). Another important variable that has been overlooked by
Collivignarelli and Bertanza (1999) and Demoulin et al (1997) is the need for
additional post aeration facilities to oxidise any remaining nitrite nitrogen from the
main treatment process. This may be necessary in situations were elevated
effluent nitrite levels could be problematic. For example elevated nitrite nitrogen
levels can be particularly troublesome for plants that use chlorination for
disinfection as nitrite is readily oxidized by chlorine requiring 4g chlorine/g N02-N
(Metcalf and Eddy 2001). Additional aeration facilities to oxidise (periodic)
elevated nitrite levels may also need to be considered when sizing I designing an
aerobic denitrification facility.

96
In summary this experimental work suggests the nitrification and denitrification
rates for aerobic denitrification are significantly lower than those of conventional
separate stage processes. It appears that savings in terms of achieving opposing
reactions in the same time/space are likely to be offset somewhat by the reduced
rate of biochemical transformation. While it is possible aerobic denitrification may
offer opportunities for simplification of the treatment process (I.e. less recycled
flow) the experimental results obtained in this research suggest it is unlikely
aerobic denitrification will offer a reduction in the size or number of reactors
required. These results appear to contradict reports from some full scale plants
which suggest aerobic denitrification may offer the potential for treatment tank
volumetric savings.

5.2.3 Sludge production rates


Table 5.2-3 illustrates the normalized sludge production rates. It shows the mass
of sludge created per volume of wastewater treated. The aerobic denitrification
process produced significantly less waste sludge than the traditional nitrification
process.
Normalized sludge production

Wastage *Wasted Treated Waste (mg/L)


3
DOSP (L) Cycles Solids (mg) (L) (kg/m )
4.0 57 420 171000 1680 102
2.5 42 370 124200 1480 84
1.0 25 228 74400 912 82
0.5 12 145 34800 60

*Assuming average wastage [] = 3000mg/L ** 4 L wasted I cycle x number of cycles

It should be noted that the higher dissolved oxygen set points only nitrified the
wastewater while the aerobic denitrification process also denitrified. Therefore
the true savings in sludge production arising from the aerobic denitrification
process will be greater than suggested by Table 5.2-3. That is the nitrification

97
only systems would also have a sludge contribution from subsequent anoxic
facilities (required for denitrification). Further research is required to determine
the sludge production rates in an optimized aerobic denitrification process (one
that achieves full denitrification).

There is little reference in the literature as to the sludge production rates for the
aerobic denitrification process. However assuming the aerobic denitrification
process uses a shortened denitrification (nitritation type) pathway Turk and
Mavinic (1986) and Turk and Mavinic (1987) found this nitrogen removal pathway
can result in lower biomass yields. Oelgenes and Patureau (2004b) also
suggested a shortened denitrification pathway achieved through the control of
dissolved oxygen should result in less biomass production. Stouthamer et al
(1997) investigated some of the microorganisms responsible for aerobic
denitrification and found that some like Thiosphaera pantotropha produced a
lower growth yield.

(Objective 3) Comment on the requirements for air relative to conventional


separate stage nitrification denitrification processes.

Table 1 shows the total amount of air required to oxidise the influent
ammonia nitrogen.

Table 5.3-1 Total air required for ammonia oxidation

Dissolved oxygen set Average air


point required Average air flow rates
l air I mg NH,·N limin

2.5 2.0 2.0


1.0 2.9 1.8
0.5 3.5 1.3

98
With reference to Table 5.3-1 the 4.0 mg/L DOSP required more air per unit of
ammonia oxidized than the 2.5 mg/L DOSP because of excessive aeration and
inefficiencies at the 4.0 mg/L level. The increase in the amount of air required at
DOSP 1.0 mg/L and 0.5 mg/L is due to inhibition of the nitrification process
caused by the low dissolved oxygen concentrations.

The data suggests a traditional nitrification process run at - 2.5 mg/L would
require - half as much air per unit of ammonia oxidized compared to an aerobic
denitrification process run at 0.5 mg/L.

With respect to the average flow rates required to reach the dissolved oxygen set
points it can be seen the 0.5 mg/L aerobic denitrification process required
approximately half the air flow rate of the traditional 2.5 mg/L nitrification process.

Thus it appears as if the aerobic denitrification process may require twice as


much air supplied at half the air flow rate, (relative to a traditional nitrification
process run at a dissolved oxygen concentration of 2.5 mg/L). This might
translate into higher overall aeration costs for the aerobic denitrification process
as efficiencies in terms of smaller blowers or the number of units required may
not offset the cost in terms of $/unit of air required.

Experiences from full scale facilities suggest a reduction in aeration


requirements. For example Hayward (1998) investigated an aerobic
denitrification and phosphorus removal plant at Cabooltura (Australia). He said
the aerobic denitrification process had the ability to minimize the operational
costs associated with supplementary chemical treatment and aeration (reduced
aeration times were reported). However it should be noted that the SBR type
plant also incorporated an online real time control system at the time of
transferring to an aerobic denitrification process. It is possible that the reported

99

THE LIBRARY
UNIVERSITY OF CANTERBURY
CHRISTCHURCH, N.Z.
aeration savings were a result of the new control system also providing increased
efficiency.

Heinen and Norgaard (1998) reported results from full-scale facilities modified to
incorporate nitrogen removal through aerobic denitrification. Specifically they
noted the low dissolved oxygen concentration required for the process ensured
that existing aeration facilities were usually sufficient in upgraded plants, while
they expected proposed new facilities would require less oxygenation equipment.
Pedersen et a/ (2003) reported on the conversion of the industrial wastewater
treatment plant at the CPKeico ApS pectin plant in Denmark. The industrial
WWTP is one of the largest in Denmark and was converted from a conventional
nitrification/denitrification process to a nitritation-denitritation process. That is the
process was converted to one of ammonium oxidation to nitrite followed by
controlled reduction of nitrite to N2 . They reported the shortened pathway
process decreased the oxygen required for oxidation by 25%.

The development of a 2,500 p.e. aerobic denitrification trial facility within a larger
440,000 p.e. plant by Collivignarelli and Bertanza (1999) allowed operating data
to be obtained over a two year period. They undertook a cost evaluation of both
conventional and aerobic denitrification facilities. Based on costs calculated in
Italian lire it was found the construction of conventional facilities required an
additional 10% capital investment along with an ongoing 30% higher operational
cost per year. Specifically it was found that a 50% saving in the electrical energy
expenditure could be attained with an aerobic denitrification process.

From an experimental perspective Yu et a/ (1998) reported that appropriate


online control combined with the savings attainable from an aerobic denitrification
shortened nitritation pathway could result in aeration energy savings of up to
45%. Abeling and Seyfried (1992) reported on aeration savings resulting from
the use of a shortened nitritation pathway. They found nitritation required only

100
75% of the oxygen resources compared to full pathway nitrification. This was
supported by Oelgenes and Patureau (2004b) who reported a possible 25%
reduction in theoretical air requirements for nitritation followed by nitrite type
denitrification. Stross (2000) reported on the CANON process (refer literature
review) in which nitrification is limited to nitrite type nitrification. He reported a
63% decrease in the total air requirements.

In summary, the experimental data indicated the aerobic denitrification process


required a lower air flow rate and this may provide opportunities in terms of
smaller blowers (or less units) and corresponding lower capital costs. However
the process also required twice as much air in total (per unit of ammonia
oxidised). This may indicate the aeration running costs for an aerobic
denitrification system could be higher.

5.4 (Objective 4) Identify the dissolved oxygen conditions necessary for aerobic
denitrification and for its optimisation.

A major variable influencing the rate of aerobic denitrification appears to be the


dissolved oxygen concentration as shown in Table 5.4-1.

Table 5.4~1 Effect of dissolved oxygen concentration on aerobic nitrogen losses

Dissolved oxygen set Average % of influent


point nitrogen lost aerobically

2.5 10% •
1.0 26%
0,5 40%

* The nitrogen mass balance procedure was an approximation. Limitations of the


[MLSS] testing procedure meant that a precise determination of assimilation was
not possible, losses here may be within accuracy limits of the mass balance.

101
There was a drop in aerobic denitrification from 40% at a dissolved oxygen
concentration of 0.5 mg/L to 0-7% at a dissolved oxygen concentration of 4.0
mg/L. The nitrogen losses at the 1.0 and 0.5 mg/L dissolved oxygen
concentrations are classified as "aerobic denitrification losses".

It is important to note that the losses can probably be ascribed to both traditional
anoxic denitrification and aerobic denitrification. That is a combined effect of
aerobic denitrification and classic anoxic denitrification, refer literature review 1.2-
1.4).

The aerobic denitrification activity appeared to become suppressed in


environments where the dissolved oxygen concentration was above 1.0 mg/L.
Nitrogen losses increased as the dissolved oxygen concentration was lowered to
0.5 mg/L suggesting 0.5 mg/L was closer to an optimum than 1.0 mg/L.

The research attempted to operate at a set point of 0.3 mg/L however the
hardware was unable to reliably measure and control the dissolved oxygen
concentration at this level. The experimental hardware was limited to the use of
traditional dissolved oxygen probes which were not suited to this application. As
the aeration system relied upon a reading from the probe to adjust the amount of
air supplied a small drift in the probe reading at the low set point could result in a
(proportionately) significant reduction in aeration. That is the probe could read
-0.3 mg/L when the actual conditions were anoxic, in this situation the aeration
system would assume the set point of 0.3 mg/L had been reached and the air
flow rate was reduced to -zero. There was also a degree of instability as a result
of operating a nitrification system at such a low dissolved oxygen concentration.
That is the nitrification process became significantly inhibited most probably by
the lack of dissolved oxygen. Dissolved oxygen concentrations below 0.5 mg/L
are reported to result in considerable inhibition of the nitrification process (Metcalf
and Eddy (2001)).

102
While this research suggested the optimum dissolved oxygen concentration
appeared to be around 0.5 mg/L research by Third (2004) suggests the optimum
dissolved oxygen concentration for aerobic denitrification is unlikely to be a fixed
value, rather it will depend upon variables such as the biomass concentration
and the release of reducing power in terms of the ability to hydrolyze stored
carbon polymers.

The elevated nitrite and nitrate levels and absence of effluent ammonia nitrogen
at the 0.5 mg/L dissolved oxygen set point in this work may suggest the optimum
(in this instance) was lower than 0.5 mg/L. Researchers such as Munch et al
(1996), Demoulin et al (1997), Yoo et al (1999), and Collivignarelli and Bertanza
(1999) found that dissolved oxygen concentrations above the aerobic
denitrification optimum resulted in the effluent containing higher concentrations of
N02-N and N03-N while dissolved oxygen levels too low resulted in higher NH3-N
concentrations in the effluent.

Munch et al (1996) suggested a dissolved oxygen value of 0.5 mg/L was suitable
to achieve a nitrification rate equal to the denitrification rate ensuring complete
aerobic denitrification. However they only reported the 0.5 mg/L dissolved
oxygen level as a theoretical value and did not operate at that level. Instead they
established nitrification and denitrification rates at higher dissolved oxygen
concentrations and then determined a theoretical dissolved oxygen concentration
that could produce the same rate of nitrification and denitrification. This was done
by combining a Monod type kinetic expression for the rate of nitrification and an
IAWQ model for the rate of denitrification.

Pochana et al (1997) developed a mathematical model to determine the optimum


dissolved oxygen concentration for aerobic denitrification. The model was based
on the hypothesis of diffusion limitations in activated sludge flocs. They found

103
that a dissolved oxygen concentration of 0.2 mg/L may result in incomplete
nitrification with elevated effluent ammonia concentrations along with complete
denitrification of any oxidized ammonia. At a dissolved oxygen concentration of
0.5 mg/L the model found that nitrification would improve but a build up of nitrite
and nitrate could be expected. Thus the model essentially predicted the optimum
dissolved oxygen concentration would be somewhere between 0.2-0.5 mg/L. The
model also predicted that aerobic denitrification could achieve over 90% nitrogen
removal at very low dissolved oxygen concentrations (0.2 mg/L) where as at
higher dissolved oxygen concentrations (> 1mg/L) the aerobic denitrification
reaction would be strongly inhibited.

In general the experimental findings obtained in this research complement those


of other researchers including Pochana et al (1997). That is this research found
the 0.5 mg/L set point had elevated effluent nitrite and nitrate concentrations
while the 0.3 mg/L set point had residual ammonia nitrogen, (no data presented
as system unstable). The data also illustrated an inhibition of aerobic
denitrification at dissolved oxygen concentrations above 1.0 mg/L.

A review of the literature would suggest the optimum dissolved oxygen


concentration (range) for the activated sludge aerobic denitri'fication process has
not been determined experimentally. Reported optimum values are either
theoretical or from studies involving cultures (with plans to extend this to more
complex systems). Attempts to experimentally isolate the optimum dissolved
oxygen range have been imprecise. In the late 1990's a central figure in the
attempts to determine the optimum aerobic denitrification dissolved oxygen
concentration was Jurg Keller of the Advanced Wastewater Management Centre
Queensland Australia. Keller was involved with the research of Pochana et al
(1997), Pochana and Keller (1999) and with the earlier work by Munch et al
(1996). In experimental terms Pochana and Keller (1999) attempted to validate
earlier theoretical work. This was partially successful however as with other

104
studies the optimum dissolved oxygen concentration (range) was not determined.
They worked with dissolved oxygen concentrations down to O.S mg/L and found
that aerobic denitrification rates continued to increase down to and possibly
beyond those concentrations (lab-scale SBR activated sludge process).

With respect to studies involving cultures attempts have been made to identify
the specific optimum conditions for aerobic denitrifiers within culture studies with
the aim of transferring this knowledge to experiments with activated sludge
processes. For example Heijnen and Van Loosdrecht (2004) are currently
undertaking research at the Delft university clean technology institute to
determine the process conditions which lead to optimal aerobic nitrogen
elimination rates in wastewater treatment processes. This research is based
upon earlier experiments with the model organism Thiosphaera pantotropha
which have indicated the specific optimum conditions for which aerobic
denitrification is likely to occur.

Other culture studies include those of Chen et al (2003) who undertook


experiments with cultures of Pseudomonas aeruginosa (ATCC 9027) and
demonstrated an approximate linear increase in aerobic denitrification activity as
the dissolved oxygen concentration was lowered from 1.3 mg/L to 0.1 mg/L. The
rate at 1.3 mg/L was only 1/S of the maximum rate (1 mmol/g of cells.h). In
order to measure the dissolved oxygen concentration at such low levels their
work required the use of optical sensors that could accurately determine very low
dissolved oxygen concentrations based upon oxygen-quenched luminescence.

Patureau et al (2000) extended earlier work with culture studies and trialed an
aerobic denitrifier organism mixture within an alternating aerobic/anoxic process
(in the presence of nitrate). They evaluated the aerobic denitrifying performance
at various dissolved oxygen concentrations and found that aerobic denitrification
activity was present up to oxygen saturation conditions (7 mg/L dissolved

105
oxygen), however the denitrifying enzymes became increasingly active below a
0.35 mg/L threshold.

From a full scale perspective Goronzy et al (1997) examined several processes


incorporating aerobic denitrification. The important aspect was ensuring that the
process loading was sufficient so that for at least half of the aeration phase the
oxygen demand exceeded the oxygen supply capability of the aeration system,
such that the bulk dissolved oxygen concentration was <1.0 mg/L. This provided
the opportunity for substantial nitrogen removal via aerobic denitrification.

Hayward (1998) investigated an aerobic denitrification SBR wastewater


treatment facility in Caboolture South Australia. To achieve full ammonia
transformation and effluent total nitrogen concentrations under 5 mg/L the
aeration system was operated so that the first half of the aeration phase had a
dissolved oxygen concentration below 0.5 mg/L. Following that the dissolved
oxygen concentration was allowed to slowly climb to an ultimate concentration of
2 mg/L at the end of the aeration cycle (DO breakthrough). Hayward (1998)
found that maintenance of the dissolved oxygen below 0.5 mg/L was key to
ensuring low effluent nitrogen levels via aerobic denitrification (NOx-N < 2.5
mg/L). Short periods of dissolved oxygen concentrations as high as 2 mg/L at the
Caboolture plant did not appear to inhibit or affect the aerobic denitrifiers.

Thus while the aerobic denitrification process requires a low dissolved oxygen
concentration it would appear that "short" periods of high DO to oxidise remaining
ammonia may be acceptable. Demoulin et al (1997) demonstrated that once
established an aerobic denitrification process could continue for a limited period
in the presence of high dissolved oxygen concentrations however they also found
that continued operation at high dissolved oxygen concentrations required a
significant re-acclimatization period (~10 days) before removal efficiencies were
"fully restored". Munch et al (1996) found that "extended" operation at high

106
dissolved oxygen concentrations resulted in long term damage to the aerobic
denitrifying microorganisms, that is the removal efficiencies were not fully
restored following a period of extended operation at high dissolved oxygen
concentrations.

In summary the general consensus is that the activated sludge aerobic


denitrification process has a dissolved oxygen concentration optimum
somewhere close to but below 0.5 mg/L. While stable operation at 0.5 mg/L was
achieved in this research the elevated nitrite and nitrate levels at 0.5 mg/L
suggested the optimum was probably below this. Reliable experimental data for
the 0.3 mg/L dissolved oxygen set point was not obtained.

5.5 (Objective 5) Comment on the need for soluble organic carbon for the
removal of nitrogen in the aerobic denitrification process.

This research demonstrated the removal of nitrogen under aerobic conditions


continued at approximately the same rate even when soluble organic carbon
sources had been depleted. This is depicted in Figure 5.5-1 which shows the rate
of nitrogen removal (in terms of TPN) as well as the soluble organic carbon
concentration measured as COD.

TPN

~ 15
o
!:!.
Z' 10
~

50 100 150 200 250 300 350 400 450


SBR cycle tlmolmlnutes)

Figure 5.5-1 Cycle illustrating nitrogen removal rate versus soluble organic
carbon availability
107
Two explanations for the nitrogen loss include autotrophic denitrification, and
heterotrophic denitrification possibly using stored intracellular carbon.

Researchers to suggest that wastewater nitrogen reduction may be achieved


autotrophically include Strous et al (1997), Helmer and Kunst (1998), Dijkman
and Strous (1999), Stross (2000), and Holman (2000). For example Helmer and
Kunst (1998) proposed that newly formed N0 2-N may act as the electron
acceptor in the autotrophic conversion of ammonia to N2 . They demonstrated
90% aerobic nitrogen removal from a largely autotrophic micro-organism
population degrading landfill leachate without addition of any organic substrate.
Holman (2000) also demonstrated aerobic nitrogen losses in the absence of
soluble organic carbon in a lab experiment using synthetic wastewater. In this
case the rate of nitrogen removal did not diminish once the carbon source was
depleted. In some instances near total nitrogen removal was achieved in an
aerobic process without (measurable) soluble organic carbon. (Refer to literature
review for the work of Dijkman and Strous).

Researchers to suggest that aerobic nitrogen reduction may be due to


heterotrophic activity include Patureau et al (1998), Zhao et al (1999), and
Pochana and Keller (1999). Patureau et al (1998) identified a particular strain of
M aerodenitrificans capable of aerobic denitrification, they reported that M
aerodenitrificans has a high affinity for acetate (making it heterotrophic). Zhao et
al (1999) reported that the addition of acetate significantly improved their aerobic
denitrification rates and speculated that this was due to improved heterotrophic
denitrification. Pochana and Keller (1999) ran a sequencing batch reactor for 45
days and showed aerobic denitrification activity was significantly improved during
periods when the feed was spiked with readily degradable acetate. From this
they concluded that aerobic nitrogen reduction was undertaken by heterotrophic
microorganisms. It should be noted however that increased anoxic denitrification
or traditional heterotrophic activity has also been shown to occur following

108
acetate addition. For example Bilanovic et al (1999) found the addition of acetate
as a carbon source helped eliminate accumulated nitrite in an anoxic/aerobic
reactor. Gerber et al (1986) and Tam et al (1992) also found that acetate gave
higher anoxic denitrification rates than either methanol or glucose. Henze (1989)
found that the denitrification rate using wastewater as the carbon source in a
traditional separate stage denitrification process was only one third of the value
obtained when acetate or methanol was available.

Another explanation for nitrogen loss in the apparent absence of soluble carbon
may be heterotrophic denitrification using stored intracellular carbon. The
sequencing batch reactor is characterized by fluctuations in substrate
concentrations over time (feast famine conditions). Research by Third (2004) at
the University of Murdoch Western Australia has found that some bacteria can
adapt strategies to cope with the large fluctuations in electron donor/acceptor
concentrations. For example some microorganisms have been found to be able
to store soluble organic substrate as polymers. Aerobic denitrification research
by Third (2004) suggests that the SBR can select microorganisms which have
metabolic strategies such as storage capabilities. When these microorganisms
are cyclically exposed to high concentrations of substrate but limited electron
acceptor they are able to take in substrate far in excess of the amount required
for growth and respiration. This excess uptake of substrate can be converted into
a storage polymer such as PHB (poly-B-hydroxybutyrate) as illustrated in Figure
5.5-2. The PHB is essentially an energy reserve that is subsequently used when
an electron acceptor such as oxygen or nitrate nitrogen becomes available.

-O-CH-CH -C -
I I 2
CH 3 0
n

Figure 5.5-2 PHB (poly-B-hydroxybutyrate)

109
Third (2004) correlated PHS concentrations and aerobic denitrification rates and
found denitrification ceased when stored carbon in terms of PHS was diminislled
suggesting aerobic denitrification may be a heterotrophic process using stored
intracellular carbon. Data presented at the SSR3 conference Noosa Queensland
Australia March 2004 suggests that under suitable conditions up to 70% of the
soluble acetate can be converted to PHS. At this stage the trials have been
limited to a lab scale SSR run on synthetic wastewater, (with acetate as the
carbon source). The type of storage polymer formed depends upon the carbon
source. For example acetate is known to form PHS, the use of propionate and
acetate will yield PHV, glucose tends to yield glycogen. There is little data on
polymer storage with respect to aerobic denitrification systems operated with raw
domestic wastewater. To complement the lab scale experiments the University
of Murdoch is currently developing a 26 cubic meter pilot plant as well as a full-
scale experiment.

If municipal wastewaters are shown to inhibit the aerobic denitrification process


(due to a lack of simple organic carbon compounds) it is possible that the
inclusion of prefermenters to transform complex organics to simple organic
compounds may alleviate the problem.

At this stage it is thol1ght polymer formation is "favoured" by the use of simple


organic compounds such as acetate as compared to the more complex organic
compounds found in raw wastewater. If acetate is favoured this may explain the
reported increase in aerobic denitrification rates following the addition of acetate
by researchers such as Pochana and Keller (1999). It would also explain the
results obtained by Holman (2000) in which an aerobic denitrification process
treating synthetic wastewater (including acetate) within a lab scale SSR achieved
-98% nitrogen removal while a similar process run with this research but with
raw domestic wastewater has achieved much lower nitrogen removal rates
-50%.

110
The carbon storage-heterotrophic explanation is also supported by some full-
scale aerobic denitrification facilities. For example Demoulin et al (2001)
reported on the Potsdam wastewater treatment plant in Germany (90,000 p.e.) in
which aerobic denitrification is achieved via internal floc denitrification using
carbon stored through biosorption. They reported the low dissolved oxygen
concentration of the process minimizes the use of substrate carbon by oxic
metabolism.

With respect to the data obtained with this experimental work Figure 5.5-1 shows
a constant rate of nitrogen removal (~ 0.95 mg N/g MLVSS.h). It is possible the
nitrogen reduction rate would decrease during a cycle as the stored organic
carbon was depleted or the rate of nitrogen removal would have been higher at
the start of a treatment cycle when soluble organic carbon was present. Third
(2004) found that one variable governing the rate of aerobic denitrification was
the release of reducing power from storage polymers. It is possible the use of
(complex relative to synthetic feed) domestic wastewater results in complex
storage polymers that provide a slow sustained release via polymer hydrolysis.
That is it is possible the polymers were not depleted within the cycle lengths of
this research. Third (2004) also found that the aerobic denitrification process had
a higher nitrogen removal rate after the carbon had been taken up as a storage
polymer. This would explain why nitrogen removal rates were not greater at the
start of the cycles in this work when soluble organic carbon was present.

While the mechanism behind the nitrogen removal remains somewhat unclear,
the ability to remove nitrogen in the apparent absence of soluble organic carbon
may offer economic advantages in terms of organic and operational savings.
Semi-treated wastewater may not need to be exposed to raw incoming
wastewater, (carbon source), or supplied with an expensive external carbon
source such as methanol. Traditional ("relatively" high dissolved oxygen
concentration) processes can require the addition of an external carbon source
when the insitu organic carbon supply is insufficient. Bilanovic et al (1999)
111
undertook a comparison of different external carbon sources for denitrification
and found the cheapest source by far was methanol. However some US$0.011
of methanol was still required for the denitrification of each m3 of wastewater
containing 30 mg/L of N03-N. This means that the use of an external carbon
source for denitrification is still an expensive operation in the wastewater
industry.

A review of the literature indicates that quantification of the carbon requirements


for the aerobic denitrification process have not been conclusively determined. Yu
et al (1998) obtained full denitrification with a COD/NOx-N ratio of -3 when their
aeration phase was restricted to nitrite type nitrification. This compares to tile
research of Tam et al (1992) who proposed a COD/NOd\! ratio of -6 for
traditional denitrification. Delgenes and Patureau (2004b) proposed a possible
40% reduction in the organic carbon requirement for denitrification following
nitrite type nitrification. Patureau is currently undertaking research (2004) that will
quantify the organic carbon savings from a (low dissolved oxygen controlled)
nitrite type nitrification aerobic denitrification process. Pedersen et al (2003)
reported on the conversion of the industrial wastewater treatment plant at the
CPKeico ApS pectin plant in Denmark to a nitritation/denitritation process in
which nitrite type nitrification-denitrification decreased the carbon required for
reduction by 40% (when compared to the previous conventional
nitrification/denitrification process).

In summary while some researchers have demonstrated that aerobic


denitrification may be achieved autotrophically under certain circumstances. The
general consensus is that the process is heterotrophic requiring the presence of
organic carbon. It is also generally agreed that the process has a lower
stoichiometric requirement for organiC carbon due to the low dissolved oxygen
concentration resulting in nitrification inhibition and shortened pathway
nitrification-denitrification. It is also felt that intra-cellular carbon storage can play

112
a role due to the low dissolved oxygen concentrations minimizing the use of
substrate carbon by oxic metabolism and/or microorganism metabolic storage
strategies resulting from batch type processes. The experimental work
demonstrated the ability of aerobic denitrification to remove nitrogen from
wastewater without the need for supplementary carbon, thus there exists a
possibility for savings in operational costs in terms of the opportunity to remove
nitrogen with less dependence on organic carbon. The mechanism behind the
nitrogen removal (in the absence of soluble carbon) and "quantification" of the
requirements for organic carbon remain to be "conclusively" determined.

5.6 (Objective 6) Confirm if the online profiles have unique features with respect
to the aerobic denitrification process, in particular the ammonia elbow on the
ORP profile and the ammonia valley on the pH profile. Correlate these online
features with measured biochemical events such as the depletion of organic
carbon or ammonia nitrogen.

5.6.10RP
In an aerobic nitrifying environment electrons are transferred from the ammonia
nitrogen to oxygen which acts as the electron acceptor. At the same time the
ORP profile shows a positive slope as electrons are also transferred from the
ORP reference cathode to the oxygen (which has a greater positive potential). As
the ammonia is depleted and the flow of electrons from the ammonia nitrogen to
the oxygen diminishes this would also be expected to affect the rate of electron
flow from the ORP probe. Effectively more oxygen should be available to act as
electron acceptors for the ORP reference cathode and this could theoretically be
expected to cause an elbow or increase in the positive ORP gradient.

Some researchers have in fact made a correlation between the ORP ammonia
elbow and the point of ammonia depletion with the implication being the increase
in ORP and the depletion of the ammonia are directly related (i.e. the change in

113
ORP is due to changes in oxidative reductive potential resulting from the
depletion of ammonia nitrogen). For example Yu et al (1998) suggested the
elbow resulted from ammonia depletion in an AlO process as shown in (Figure
5.6.1-1).

ORP
100 ORPE.JbllNal
pelnt of 21M'Cn1•
depl.~nn 7.0
100
\
30 0."
'§,
E INH:J-NI
z 0
10
:r: :r:
6.6 a.
Z

~
!l. 0 6.4
'"0
-100

-200
Asrohk:· ;1iron "

Time (hours)

Figure 5.6.1 ~1 illustration of ORP elbow in an AlO process. (Figure Adapted from
Yu et al (1998)

Another example of a correlation between the ORP elbow and the point of
ammonia depletion was evident in the work of Ra et al (2000). They operated an
aerobic-anoxic SBR process treating piggery wastewater and identified the ORP
elbow at the point of ammonia depletion. They then used ORP to terminate
aeration. In this instance they detected the ORP plateau that follows the elbow
feature (Figure 5.6.1-2).

114
Ra CIt <fI2000 deluded the ORP
plateau and loon ferminated i3e:aUon 10.0
{ drop In ORP her~ .caused by Mf~iOn
;rminallonl
50
9,5

·50
0.0

.'00
1
"- ·150 8.5 :a
a:
a
·200
80
.2£0

·300

.350

·400
0 12 15 21

Time (hours)

Figure 5.6.1-2 illustration of ORP elbow in an AlO process. (Figure Adapted from
Ra et al (2000)

Research by Holman and Wareham (2003) indicated the existence of an ORP


elbow at the point of ammonia depletion within an aerobic denitrification process
treating readily biodegradable synthetic wastewater Figure 5.6.1-3.

ORP elbow at 400


6
point of ammonia
ORP
depletion \
300
[NH,·N)

\ 200

100 ~
:i!o
[DOJ
o

·100

+-~i---.----r-----'---"--::::::;===----+ ·200
20 40 60 60 100 120
Time (minutes)

Figure 5.6.1-3 illustration of ORP ammonia elbow in an aerobic denitrification


process, Holman and Wareham (2003), (note anoxic phase for first 10 minutes).
*First ORP elbow resulted from organic carbon depletion, not sllown.

The existence of an ORP ammonia elbow during the aerobic sludge digestion
process was also proposed by Wareham et al (1993) and Wareham et al (1994)
as illustrated in Figure 5.6.1-4.
115
NI-j"N

f--.l,l~r~-<>-----<>--~-~----.;~= [DO[

OR?
Elbuw at JWint of DO
break through fallowing

II
ammonIa depletioll .

I N{!rlilu, knee /
3 ho<Jts air on atpo!ntofnlttat~
depfallon

Tin'll! (HCUfS)

Figure 5.6.1-4 Illustration of ORP elbow in aerobic sludge digestion process.


(Figure adapted from Wareham et a/1993).

However in all the cases cited the dissolved oxygen concentration was allowed to
vary and find its own level. Figure 5.6.1-5 however shows a typical ORP profile
within an aerobic denitrification environment held at a fixed dissolved oxygen
concentration (0.5 mg/L). As can be seen dissolved oxygen breakthrough was
suppressed (as a result of the dissolved oxygen set point control system) and the
ORP showed no elbow at the point of ammonia depletion.
30 200

25 ORP
ioO

>E
o ;:
IS
NH,-N
·100

00
-I-d::::==;===;:=::;=:::::;:==~:::;===;:~~=i- ·200
50 100 150 200 250 3no 350 400 450
SBR cycle tIme (minutes)

Figure 5.6.1-5 illustration of absence of ORP ammonia elbow when dissolved


oxygen concentration maintained at a consistent level (taken from TS7 DOSP 0.5
mg/L).
As the ORP ammonia elbow does not appear when the dissolved oxygen is held
at a fixed concentration it appears that the ORP ammonia elbow results
exclusively from the dissolved oxygen breakpoint. That is the ORP elbow is a
116
response to dissolved oxygen concentration changes rather than a direct
response to the depletion of ammonia itself. For example if the aeration system
allowed the dissolved oxygen concentration to breakthrough following the
depletion of ammonia then the ORP would show a corresponding elbow and
increase at the point of dissolved oxygen increase.

In this work the ammonia elbow failed to appear for both the traditional
nitrification process when the dissolved oxygen concentration was maintained at
4.0 mg/L and 2.5 mg/L and the aerobic denitrification process when the dissolved
oxygen concentration was maintained at 1.0 mg/L and 0.5 mg/L.

The ORP dependence on the dissolved oxygen concentration may be due to the
logarithmic interrelationship between DO and ORP (section 1.6) and the strong
electron acceptance potential of the dissolved oxygen. Collivignarelli and
Bertanza (1999) reported the ORP value in the aerobic denitrification process
was determined by the 02/0H- equilibrium, as well as N03-INH/ N0 2-/NH/ etc.
However these experimental results suggest the 02/0H- equilibrium is probably
the dominant parameter such that in the presence of dissolved oxygen changes
in the other parameters do not result in detectable changes in the value of ORP.
In practical terms this means the ORP ammonia elbow probably results
exclusively from the DO breakpoint and it is misleading to imply the elbow is a
direct response to ORP changes resulting from ammonia depletion.

The dominance of dissolved oxygen and the effective suppression of an ORP


feature resulting directly from the depletion of ammonia probably raises doubts
as to the usefulness of the ORP profile in aerobic-aerated processes. Skepticism
of ORP measurements in aerobic processes has been expressed by other
researchers including AI-Ghusain et al (1994) and Csikor et al (1996) who
suggested the need for better online real time parameters such as pH. Cecil
(2003) reported on the development of a redox process control system at the

117
Odense City wastewater treatment plant in Denmark. Redox was clearly shown
to indicate the point of nitrate depletion during an anoxic phase. However redox
provided no useful information during the aerobic phases and an ammonium
sensor was necessary to determine the point of ammonia depletion during
aeration.

Thus while it would be possible to use the ORP elbow feature to identify the point
of ammonia depletion in a variable dissolved oxygen concentration system it
appears that a more direct pathway would be to use the dissolved oxygen
breakpoint feature itself.

5.6.2 pH
The nitrification process was known to produce a pH profile with an ammonia
depletion feature (ammonia valley Hao and Huang (1996». One objective of the
experimental work was to determine if the pH profile was also applicable to the
aerobic denitrification process; that is, did the ammonia valley identify the point of
ammonia depletion in a low dissolved oxygen aerobic denitri'fication process.

The results of this research found the pH profile provided a consistent and
reliable indication of the point of ammonia depletion for both the traditional
nitrification and the aerobic denitrification processes. The main pH feature
termed the ammonia valley was present at all dissolved oxygen levels. With
reference to Figure 5.6.2-1 the beginning of aeration caused a rise in pH
(between points 1 and 2); probably due to CO 2 stripping from the system caused
by the initiation of aeration. After the initial increase the pH decreased again
between points 2 and 3, caused by nitrification and the consumption of alkalinity
as part of the nitrification process (7.14 mg HC0 3 consumed per mg of
ammonia). At pOint 3, the pH curve shows the feature termed "ammonia valley"
signifying the point at which ammonia is essentially depleted and nitrification has

118
finished. An illustration of a pH profile from TS7 DOSP 0.5 mg/L is provided in
Figure 5.6.2-1
30

~ I''
7.65

r:
25

20
::J

'"
E
~ 15 NH 3·N pH 7.50 :a
J:
2!:. 7.45
10

5
~ 7.40

7.35

0 7.30
0 50 100 150 200 250 300 350 400 450
SBR cycle time (minutes)

Figure 5.6.2-1 typical pH profile, (aeration starts at time =0)

The essential shape and features of the pH profile were the same for both the
traditional nitrification and aerobic denitrification processes; however aerobic
denitrification produced a pH profile of half the magnitude, that is the pH scale
had a maximum 0.8 unit and maximum 0.4 unit difference for the nitrification and
aerobic denitrification processes respectively. This is quite possibly linked to a
shortened pathway nitrification-denitrification process consuming less alkalinity in
the aerobic denitrification process (section 1.2). Full pathway denitrification also
returns 3.5 mg/L of alkalinity per mg of nitrogen reduced. It is possible some
alkalinity was returned to the process via the aerobic denitrification process and
this may have acted as a buffer.

5.6.3 Dissolved oxygen [DO]


As the control system maintained a fixed dissolved oxygen concentration the
profile provided no information apart from the concentration itself. There were no
features that correlated to the depletion of biochemical parameters.

119
5.6.4 Air flow rate
As the control system adjusted the air flow to maintain a fixed dissolved oxygen
concentration the air flow rate profile essentially equated to the air demand
profile. An illustration of an air flow profile taken from TS7 DOSP 0.5 mg/L is
provided in Figure 5.6.4-1.

NH,·N

Airflow

o 50 100 150 200 250 300 350 400 450


SBR cycle time (minutes)

Figure 5.6.4-1 typical air flow profile

The air demand profile was expected to indicate events such as the depletion of
organic carbon or ammonia nitrogen as theoretically these events would result in
a fall in the oxygen demand. The research found the air demand profile provided
an indication of the point of (readily available) organic carbon depletion (point 1)
in 80% of the cycles. The pOint of ammonia .exhaustion was also evident in 15-
40% of the cycles (not evident in figure 5.6.4-1). The lower detection rates
occurred in the runs in which the DOSP was decreased (the aerobic
denitrification range). The profile was applicable to both the traditional nitrification
process and the aerobic denitrification process, that is the essential elements of
the profile were there in both cases. The features appeared as sudden drops in
the air flow.

120
In summary the research findings suggest ORP does not provide any features
that can be correlated to biochemical events within either the nitrification or the
aerobic denitrification processes. The previously identified ammonia depletion
elbow is probably a response to dissolved oxygen concentration changes (and is
reliant upon a breakpoint in the dissolved oxygen concentration). There is
therefore no ORP ammonia elbow when a system is operated at a fixed
dissolved oxygen concentration. The implication made by some researchers that
the change in ORP is related to changes in oxidative reductive potential resulting
from the depletion of ammonia nitrogen is probably misleading-incorrect.

The pH profile provided a useful ammonia depletion feature that was consistently
present over a range of operating conditions including those required for aerobic
denitrification, (The magnitude of the feature was approximately half for the
aerobic denitrification process).

The air demand profile did provide an indication of the point of organic carbon
depletion; there was also periodically a feature indicating the point of ammonia
depletion however this feature did not appear reliably under all operational
conditions.

5.7 (Objective 7) Develop and demonstrate control algorithms that use online
features to control the aerobic denitrification process, (I.e. indirectly detect the
biochemical events). In doing so demonstrate the reliability of pH and ORP
control algorithms based upon relative rather than absolute values.

5.7.10RP
ORP failed to provide any features that could be correlated to biochemical events
under fixed dissolved oxygen conditions. As such the ORP profiles were of no

121
further use for control of either the nitrification or aerobic denitrification
processes.

5.7.2 pH
The pH profile provided a feature at the depletion of the ammonia nitrogen
(ammonia valley) as illustrated in Figure 5.6.2-1. In order to detect the ammonia
valley an algorithm was developed in which the second derivative of pH was
calculated (algorithm based upon relative rather than absolute values). This
essentially equated to calculating the slope of the pH curve (dpH/dt) and then
calculating the change in slope over defined time periods (d 2 pH/de). For
example, the average slope of the pH profile over a 3D-minute period would be
determined (every six seconds) and when the second derivative changed beyond
some specific set point (typically set at 0.001) the control system identified the
event as the occurrence of the feature termed the ammonia valley and
terminated the aeration cycle, that is the air was switched off and the reactor put
into a settle phase. It was demonstrated that 92-99% of all cycles were able to be
controlled by the algorithm. This detection rate corresponds well with a similar
software system known as the Evolving Fuzzy Neural Network (EfuNN) system
developed by the University of Otago and Waste Solutions Ltd (both located in
Dunedin New Zealand). The EfuNN system obtained a 96% successful detection
rate for DO breakpoint determination with a laboratory SBR treating synthetic
wastewater (Cohen et al (2003». In this research the lower detection rates
occurred when the DOSP was decreased. Thus pH was demonstrated as a
monitoring and control tool for both the traditional nitrification and aerobic
denitrification processes. A description of the algorithms is given in Appendix B4.

5.7.3 Dissolved oxygen [DO]


As the control system maintained a fixed dissolved oxygen concentration the
profile provided no information apart from the concentration itself. There were no

122
features that correlated to the depletion of biochemical parameters and the
profile was of no use for process control purposes.

5.7.4 Air flow rate (effectively air demand)


The air demand profile contained up to two features that were correlated to the
depletion of (readily biodegradable soluble) COD and the depletion of ammonia
nitrogen (Figure 5.6.4-1). The experimental work showed 80% of cycles
provided profiles that would have enabled an algorithm to detect the point of
COD depletion while as little as 15 % of cycles showed a clear ammonia
depletion feature making the point of ammonia depletion unsuited for detection
by an algorithm. No air flow algorithms were developed in this experimental work.

123
Chapter 6.0 CONCLUSIONS AND RECOMMENDATIONS

6.1 CONCLUSIONS
The conclusions have been incorporated into the summary sections associated
with each fulfilled objective; however to reiterate, the following conclusions can
be drawn from this research .

., The presence of aerobic denitrification activity at dissolved oxygen


concentrations of 1.0 mg/L and below was supported by the development
of elevated nitrite levels and from nitrogen mass balance calculations.

(I; This experimental work suggests the nitrification and denitrification rates
for aerobic denitrification are significantly lower than those of conventional
separate stage processes. Nitrifying bacteria are known to be inhibited by
the low dissolved oxygen concentrations. It is also thought that only a
portion of the nitrifying bacteria contained within flocs are active in low DO
systems. The rate of denitrification is lower as denitrification is limited to
the anoxic zones of floc particles. Factors such as reduced anoxic floc
area and substrate consumption in the aerobic portion of the floc act to
red uce the denitrification rate (Met Calf and Eddy (2001)).

(I; It appears that savings in terms of achieving opposing reactions in the


same time/space are likely to be offset somewhat by the reduced rate of
biochemical transformation. While it is possible aerobic denitrification may
offer opportunities for simplification of the treatment process (i.e. less
recycled flow) the experimental results obtained in this research suggest it
is unlikely aerobic denitrification will offer a reduction in the size or number
of reactors required. These results appear to contradict reports from some
full-scale plants which suggest aerobic denitrification may offer the
potential for treatment tank volumetric savings.

124
• Calculated on the basis of a unit of sludge created per unit of wastewater
treated the results suggest the aerobic denitrification process offers a
lower sludge production rate. Further research is required to determine the
sludge production rates for an "optimized" aerobic denitrification
processes (that is an aerobic denitrification process that is achieving full
denitrification). There is little reference in the literature as to the sludge
production rates for the aerobic denitrification process.

• The experimental data indicated the aerobic denitrification process


required a lower air flow rate and this may provide opportunities in terms
of smaller blowers (or less units) and corresponding lower capital costs.
However the process also required twice as much air in total (per unit of
ammonia oxidized). This may indicate the aeration running costs for an
aerobic denitrification system could be higher.

• The general consensus is that the aerobic denitrification process has a


dissolved oxygen concentration optimum somewhere close to but below
0.5 mg/L. While stable operation at 0.5 mg/L was achieved in this
research the elevated nitrite and nitrate levels at 0.5 mg/L suggested the
optimum was probably below this. Reliable experimental data for the 0.3
mg/L dissolved oxygen set point was not obtained. Metcalf and Eddy
(2001) state that present models for the aerobic denitrification process are
inaccurate (further research is required), however DO concentrations for
the aerobic denitrification process are reported as being between 0.1 mg/L
and 0.4 mg/L.

.. While some researchers have demonstrated that aerobic denitrification


may be achieved autotrophically under certain circumstances. The general
consensus is that the process is heterotrophic requiring the presence of

125
organic carbon. It is also generally agreed that the process has a lower
stoichiometric requirement for organic carbon due to the low dissolved
oxygen concentration resulting in nitrification inhibition and shortened
pathway nitrification-denitrification. It is also felt that intra-cellular carbon
storage can play a role due to the low dissolved oxygen concentrations
minimizing the use of substrate carbon by oxic metabolism and/or
microorganism metabolic storage strategies resulting from batch type
processes. The experimental work demonstrated the ability of aerobic
denitrification to remove nitrogen from wastewater without the need for
supplementary carbon, thus there exists a possibility for savings in
operational costs in terms of the opportunity to remove nitrogen with less
dependence on organic carbon. The mechanism behind the nitrogen
removal (in the absence of soluble carbon) and "quantification" of the
requirements for organic carbon remain to be "conclusively" determined.

@II The research findings suggest ORP does not provide any features that
can be correlated to biochemical events within either the nitrification or the
aerobic denitrification processes. The previously identified ammonia
depletion elbow is probably a response to dissolved oxygen concentration
changes (and is reliant upon a breakpoint in the dissolved oxygen
concentration). There is therefore no ORP ammonia elbow when a system
is operated at a fixed dissolved oxygen concentration. The implication
made by some researchers that the change in ORP is related to changes
in oxidative reductive potential resulting from the depletion of ammonia
nitrogen is probably misleading-incorrect. As ORP failed to provide any
features that could be correlated to biochemical events under fixed
dissolved oxygen conditions the profiles were of no further use for control
of either the nitrification or aerobic denitrification processes.

126
• The shape and features of the pH profile were the same for both the
traditional nitrification and aerobic denitrification processes. The profile
provided a useful ammonia depletion feature (ammonia valley) that was
consistently present over a range of operating conditions including those
required for aerobic denitrification.

41) Aerobic denitrification produced a pH profile of half the magnitude, that is


the pH scale had a maximum O.S unit and maximum 0.4 unit difference for
the nitrification and aerobic denitrification processes respectively. This is
quite possibly linked to a shortened pathway nitrification-denitrification
process consuming less alkalinity in the aerobic denitrification process.

III In order to detect the ammonia valley an algorithm was developed in


which the second derivative of pH was calculated. The control system
identified the feature termed the ammonia valley and terminated the
aeration cycle, that is the air was switched off and the reactor put into a
settle phase. It was demonstrated that 92-99% of all cycles were able to
be controlled by the algorithm. The lower detection rates occurred when
the DOSP was decreased. Thus pH was demonstrated as a monitoring
and control tool for both the traditional nitrification and aerobic
denitrification processes.

• The air demand profile contained up to two features that were correlated
to the depletion of (readily biodegradable soluble) COD and the depletion
of ammonia nitrogen. The experimental work showed SO% of cycles
provided profiles that would have enabled an algorithm to detect the point
of COD depletion while as little as 15 % of cycles showed a clear
ammonia depletion feature making the point of ammonia depletion
unsuited for detection by an algorithm. No air flow algorithms were
developed in this experimental work.

127
• As the control system maIntained a fixed dissolved oxygen concentration
the profile provided no information apart from the concentration itself.
There were no features that correlated to the depletion of biochemical
parameters and the profile was of no use for process control purposes.

6.2 RECOMMENDATIONS
The following recommendations can be drawn from this research.

III Further research needs to be undertaken to clarify the reactor volumes


required for aerobic denitrification. The nitrification and denitrification rates
in aerobic denitrification processes are a function of the reaction kinetics,
floc size, floc density, floc structure, rbCOD loading and bulk DO
concentration. Because of the complex physical factors the process is still
to be modeled accurately (Metcalf and Eddy (2001 ».

III The sludge production rates for the aerobic denitrification process need to
be clarified. The results suggest the aerobic denitrification process may
offer a lower sludge production rate (relative to a separate stage
nitrification-denitrification process). There is little reference in the literature
as to the sludge production rates for the aerobic denitrification process.

• Further research is required to determine aeration requirements for the


aerobic denitrification process (the lack of accurate models mean the
aeration requirements are unknown). It is unclear if the aerobic
denitrification process offers the potential for savings in aeration costs.
• The optimum dissolved oxygen concentration range for the aerobic
denitrification process and the variables involved need to be determined.
For example research by Third (2004) suggests the optimum could be
influenced by variables such as the biomass concentration and the

128
release of "reducing power" in terms of the ability to hydrolyze stored
carbon polymers.

• The mechanism behind aerobic nitrogen removal (in the absence of


soluble carbon) and quantification of the requirements for carbon in the
aerobic denitrification process needs to be determined.

@ Further research is required to confirm the status of the ORP ammonia


elbow. (The implication made by some researchers that the change in
ORP is related to changes in oxidative reductive potential resulting from
the depletion of ammonia nitrogen is probably misleading-incorrect).

• Further work could be undertaken to determine the alkalinity requirements


for the aerobic denitrification process. In this research the aerobic
denitrification produced a pH profile of half the magnitude, that is the pH
scale had a maximum 0.8 unit and maximum 0.4 unit difference for the
nitrification and aerobic denitrification processes respectively. Reduced
alkalinity consumption could be used to support the shortened pathway
nitrification-denitrification explanation.

129
APPENDIX A1.1 - TRACK STUDY ONE (TS1) DOSP 4.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 1 follow

Table A1.1-1 Track study data and nitrogen mass balance calculations The pH profile had an ammonia valley feature at the point of ammonia

TS1 depletion, point 1. The detection algorithm successfully identified the


valley and terminated the aeration phase.
TRACl\STUDY

30

25
:::J
III
g 20 6 "2
In' '6
t)liROGEN MASS /iA'.A\lCE Ci 5 2
0 15
0.17 m!VL,fl\i" ~ ~
Sf.05 mgfl3;-,ln mg MLSB
~ "'~
£ 10
1~·jLSSJ ]J:){j ~I\.

439 mg't day /;.tlSS


~ CODI5
148 fll"l';uttffl (;'~'I';t,(;. BeWt" lQi'hnuh:~ ru"ndd(!t~I'lt)

(.,d,,-sld'dY 10
0"" MlSS I ;;)'c!~ IiO mwLcyt~

VS3lTSS
Dol Ml5S VC/d!i" o+.--~~--'--.--~--r--r--.-~--~~~~~L+
RIII:tlNir.~H,t.O, OJ2
Df!NH"N ~) m!}<t..t'.j'tleN l (~l "'k!SsCfUN" GalnofN:)
o 10 20 30 40 50 60 70 60 90 100 110 120
1'II(jiL.r.,'d1'! N {Iii" (J.a!nofU (.) =11)4$ c,j»; SBR cycle time (minutes)
tJ ..:t>;O),N \2 ~tl t:}'Ch~N !{+jvGahOr4(-}"IGMofNJ
Nloss' A'iS ..q rntI,L;;:ydeN
E,l'rn"lwanim'Ia!.ed!J mall N lct C,H;N01
UnJqQlm!td H j:x." mqfLN !(tj"!o360IN(")"GirnofN)
Figure A1.1-2 Air flow rate [ammonia nitrogen], and [COD]

The air flow rate profile provided an indication of the points of COD and
ammonia depletion, points 1 and 2.
The transformation and removal of ammonia appears to be balanced
by the production of nitrate and the likely assimilation. The nitrogen
mass balance procedure accounted for all but 9% of the systems
nitrogen.

*The assimilated nitrogen was calculated as follows, Change in volatile


[MLSS] per cycle = 40 mg/L.cycie 0.12 (ratio of N in biomass) =: 4.8
mg/L.cycie (rounded to 5 mg/L.cycie in Table A1.1-1).

200
30
ORP

7.6
100
7.5
TPN ;;-
7.4
.g, 5 DO E
E 0
'"
Q. 5' 4 tl.
7.3 e. II:
0
7.2
·100

o +-_,____;c--__.__~-.,..__,____,,____._____r-.,..__,____;r_.J ·200
o 10 20 30 40 50 SO 70 80 90 100 110 120 130
10 20 30 40 50 GO 70 80 90 100 110 120 130 SBR cycle time (minutes)
SBR cycle time (minutes)

Figure A1.1-1 pH profile and [soluble nitrogen] Figure A1.1-3 [DO] and ORP

The ORP profile failed to show any features that could be correlated to
the biochemical events.

130
APPENDIX A1.2 - TRACK STUDY TWO (TS2) DOSP 4.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 2 follow

Table A1.2-1 Track study data and nitrogen mass balance calculations 30

TS2

TRACKSTUPY

IfPNI ' •. !Slil TN1 Raleolo5$ RllbHO$:!I • Rille-j""s

0.21 pjIJ'Lmin
fE·OS ~''Lm!rLmoMlSS

3586 moll
ML!>!> r,hange t 24 hours 526 mgll,.tj,jyMLSS
20 30 40 50 60 70 80 90 100 110
10
C,e!~\ime 134 m~'~ (Aeroble·se!lJ&+lDfIll1utUr.~and<ltC3fll)
SBR eyel" time (minutes)
CFI&siday 11
DEl M;"SS I l;¥tIe 4!l 8gjlsycle
VSSfTSS
Dei MLSS Vc!3Ma 39 mQlLc)"de
012
!)'" 'JHrN ZZ my/LeydeN (("l "'Iou (lIN (.)<cGa!71 of N)
Figure A1.2-2 Air flow rate [ammonia nitrogen], and [COO]
Dei No;·~J mg,1..cycIaN \\+),"Gair,dNH"losSQrl~)
p.;< ~jO,.~ IB m9IL'}~N I (.. )= Gain 01 N H ",j05!< 01 tl)
The air flow rate profile provided an indication of the points of COD and
Eo;llfI\!I£das!'frila{ed t; 109/lHln c,H 1r,o)
Jr91;'Cfhl,"!!itd N~" rr,glLN «+1 "lussoftlH=Ga!nciN}
ammonia depletion, points 1 and 2.

The nitrogen mass balance accounted for all the systems nitrogen.

TPN

r 7,6

7,5

7.4
9· 200

100

J:
7.3
'"
0.
0:
7.2 o

7.1

NH,.N 7.0

:=::::;::===;:::==;~::::;==:;:=~=::;=~==::;:::=~w 6.9
10 20 30 40 50 GO 70 80 90 100 110 +---r--~-~-~-~---'~-_-~-~-~--+ ·200
SBR cycJe- time (minutes)
10 20 30 40 50 60 70 80 90 100 110
SBR cycle Ume (mlnut••)

Figure A1.2-1 pH profile and [soluble nitrogen]


Figure A1.2-3 [~O) and ORP
The pH profile had an ammonia valley feature at the point of ammonia
The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the
the biochemical events.
valley and terminated the aeration phase.

131
APPENDIX A 1.3 - TRACK STUDY THREE (TS3) DOSP 4.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 3 follow

Table A1.3-1 Track study data and nitrogen mass balance calculations
TS3 30

\
TfI!\CXSTUDY
7

NH,-N

COOlS

+--r-r-~-'-,-~-~-r·-r--r-'-.-~~~-'~~'+O
mg{~Jnln

mg/L'flln.mgMlSS
o 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160
3115 mc/l SBR cycle time (minutes)
305 mgfLda'l MLSS

'"
C"lMLSS.' cycle mgIU:ytl<1
VS5,TSS
Figure A1.3-2 Air flow rate [ammonia nitrogen), and [COD)
Del MLSS Vol~li1e
R;liol~jnC~HINo. 0,12
OeltJH,-N A mgl\.-oycle 11 ((+) "" "'$.1 of ~j H'" Gam of N ) The air flow rate profile provided an indication of the points of COD and
Ddt/Orr! mwLcyt:!eN ti+)"'Ga;n4ftJH"'\os50fN)
1(+)""Ga:~¢IN(-i"'lossofN)
15 mgfL,itJL1eN
mp/LqclcN
ammonia depletion, points 1 and 2. Point 2 was present but somewhat
Nb"" A-(8+C}
B(imilled assl1fi1!!1ki N mgillHo c!H,NO t
Uracc/;;!.IderlNIru$ mgllN j(t) .. !fJn(l/N(-)"" Gl.!incfN}
unclear.

The nitrogen mass balance accounted for all but 10% of the systems
nitrogen.

30 200
ORP

·100

;g; 5
;;-
§, E
00
5' 4
0-
e. '"0

·100
7.1

7.0

o -200
+--+~~~-~~~--~-~~-~~-,...::::::o,....--.--+6.9
W W • • • _ 00 _ n ill _ _ m m ~ ~
o 10 20 30 40 60 60 70 80 90 100 110 120 130 140 150 160

SSRcycle lime (m1nutes) SBR cycle time (minutes)

Figure A1.3-1 pH profile and [soluble nitrogen) Figure A1.3-3 [DO) and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase.

132
APPENDIX A1.4 - TRACK STUDY FOUR (TS4) DOSP 4.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 4 follow

Table A1.4-1 Track study data and nitrogen mass balance calculations
TS4

1R:ACKSTUDY

6 C

J
m
NH,-N

Airflow
:j !
tn~IU!11n
+---r--.---r--'---.---r--.---r--~--~~,L __--. 1
£E"()S tng:lLmln.rr'J MLSS 10 20 30 40 50 60 70 60 00 100 110 120 130
['~lSSl JO/4 I'TI\lIL SBR cycle time (minutes)
115 tn;ILday I,.'l5S
Cyd"limJ
Ci'd,,~jd!¥ 10
~'I,\lSSIC;y<;le

vssrrss 0,80
Dfl "1LSSV<lI~~i'- Figure A1.4-2 Air flow rate [ammonia nitrogen], and [COD]
Rd"l> N 'r, C~H>N{)) 0,1'2.
Dt.NH,..il m~jLt"rt:a N { (+) '" lau or tl H '" Gain mIi}
DdfiOrN M;)'Ltyc.Je-ti «(+)",GahclN(·l"!wcQO'.q The aeration system struggled in the initial stages of the cycle to reach
~INOyN 13 M~ILcycla 11 «t) '" Gill/l (If N (o} '" t~G6 or N )
11'I\l!Lt'f"laN the target DOSP and it is thought the DO transfer efficiency may have
E'lirraie31lt;l;jrri31tdfl tn9-ILNI¢ C,H,No"
[;~accv.,1dlla N bn fallen due to problems with the aeration stones. Following this track
study the aeration stones were replaced and this appeared to rectify
the problem. The air flow rate profile provided an indication of the point
With reference to table A 1.4-1 the nitrogen mass balance accounted
of COD depletion point 1. The point of ammonia depletion was not
for all but 12% of the systems nitrogen.
clearly visible.

30 7.7

r
7.6
25 o
Z 7.5
"-
t. ORP
20

~
z-
7.4
100
- -'
Z "'15 pH 7.3 i
a.§.
t:. :l
7.2
'"
.§. 5 ~
~
10
5' a.
e NH~-N 7.1 e. "'0
NO,·N 7.0
·100
<--~~~-"---··T--,--'--~~-..,..,.,=r=::::;:::'::5;""--.--+· 6.9
10 23 30 40 50 60 70 80 90 100 110 120 130
SBR cycte time (minutes)
o +---,--~.-~-,---,~~--~--~ --.----,---,----.---+ .~oo
10 20 30 40 50 60 70 80 90 100 110 120 130
Figure A1.4-1 pH profile and [soluble nitrogen]
SBR cycle tIme (minutes]

The pH profile had an ammonia valley feature at the point of ammonia


Figure A1.4-3 [DO] and ORP
depletion, point 1. The detection algorithm successfully identified the
valley and terminated the aeration phase. The ORP profile failed to show any features that could be correlated to
the biochemical events.

133
APPENDIX A1.S - TRACK STUDY FIVE (TSS) DOSP 4.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 5 follow

Table A1.5-1 Track study data and nitrogen mass balance calculations
TS5

TRAG'{STUbY

tllTROGEN MASS BAlANcE

AnJil!Wfilten'tr:(!Gat:on iliS mgllomln


5E-05 mg1Lmln.fT19 t,1lSS o 10 20 30 40 50 60 70 80 90 100 110 120 130
iMLSS) 2972 n;gll
115 mgit.dayMLSS SBR cycle time {minutes)
147 mirnd1)$ (Awcb~o t Satli/! .. 10 minutes f1Il!l!d :W:anlj
Cycl~i>/day

eel MLSS ! e)va


'0
" Figure A1.5-2 Air flow rate [ammonia nitrogenJ, and [CODJ
Del MtSS Volil.~'ec
"$Xl N In C~H!Nbl Q.12
~lH)"N A '21 mglL.cJ~leN ((+)''''It:lSsofNH=GainofNl
0;;: N01"N mgIL,c;yck>-N i l+j t Gl't-nol N (0)::: 10u::If N)
The air flow rate profile provided an indication of the points of COD and
D~I NOJ"N 11 mgtL.cy,:;lilN {{+," G&ino;tl (·)"jo$$QrN)
N ;oo;s A.(8'C) mgfLqt!e-N ammonia depletion, points 1 and 2.
Er.Uma\M e$slT0ia~d N ITl\llUHQ C!H,NQ,
JDzcc<ltiI'1!e:lNIcs$ mgllN ((+j"'loswotN(-)"'GI1'noIN)

The nitrogen mass balance accounted for all but 11 % of the systems
nitrogen.

30 7.7
200
TPN
;g: 25
7.6
g
7.5
Z
!l.
pH 100
t:, 20
7.4
Z :::r ;;;;
2; :r: OJ e
2::. 15 7.3 c. !. 00 0

~
!l.
Z 5' i>::
B- 0
2; 10
7.2
2::.
Z 7.1 ·100
:£ 5
2::. 7.0

0 6.9
o .200
10 20 30 40 50 60 70 80 90 100 110 120 130
0" W • 40 W • ro • • ~lH lWl.
SBR cycle time (minutes) SBR cycle lime (minutes)

Figure A1.5-1 pH profile and [soluble nitrogen]


Figure A1.5-3 [DOJ and ORP
The pH profile had an ammonia valley feature at the point of ammonia
The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the
the biochemical events.
valley and terminated the aeration phase.

134
APPENDIX A1.S - TRACK STUDY SIX (TSS) DOSP 4.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 6 follow

Table A1.6·1 Track study data and nitrogen mass balance calculations
TS6

TRACK STUDY 25
7

~
20 6 ::r
.§. en
5 .§.
!l!
Cl
15
~
lrn:$ Galn Gain
0
!2. '"
«
~

10
~
20 13
mglt ""It mj)il. m,,,- moJl mg/Lmln mg/l..mn mQlLmln
:r
~

t;YTROO£N MASS BALANCE

Il14 rug/Lmln
4,£·05 mglLmln.lfl9 MlSS
!MlSSl 3639 mg.'\. o 10 20 30 40 50 60 70 BO 90 100110120130140150160
MLSS ehiU1gn 124 hOUfl> 54<1 m!)lLday MLSS
SBR cycle lime (minutes)
Cyclelirnr: (A~.rllbl¢ t SeRki. 10 mktUtec fill and decant)
Cydeslday
nt: MlSSfc.ydt:
VSsrtss 0,80
Di'J MLS3 Volali!J! 50 tngtLll'{cle Figure A1.6-2 Air flow rate [ammonia nitrogen], and [COD]
Ra!;;) N In C1,H1NO. 0.12
fMNrh·N A 20 mgfLeyeleN «().) "'loliiottlH"'GII:inofN}
fMNOyN trIg/L.eycle N ((+) :eGal'lofN(.)""lonofn) The air flow rate profile provided an indication of the points of COD and
omNorN h'1p/l.cydc N ({tl '" Gam of N (-):::: lott of N)
Ii ~OS& A-(B,.q mlYLq'CieN ammonia depletion, points 1 and 2.
Eslim>lled ll.$$lfI"ohrlcd t~ mUll N to C~H1N~
UM.cc:OU!1\&d N btl< mglLN ((+)"!oMllfN(·)~G,.btcrNl

The nitrogen mass balance accounted for all but 3% of the systems
nrtrogen.

30 7.7

7.6 200
::J 25
0. TPN ORP
g 7.5
Z
20
b 7.4 100

~
Z
J; j5 7.3 ~
t'.
~,~-- ~
~ 5 DO
7.2
10 0 4 c.
t'.
7.1
Cl
'o"
'? NH,-N

t'. 7.0 ·100

t:::::=;==-.--~--.~~=;:::=;:::=:=:;::=;:==;==;;=:::;:S~~" 6.9
o 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160
o ".. -~-~-----,----~~----- __ ._ ~ _____ ____
~ ~ ___ ~_._

SBR cycle time (minutes) o 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
SBR cycle lime (minutes)

Figure A1.6·1 pH profile and [soluble nitrogen]


Figure A1.6-3 [DO] and ORP
The pH profile had an ammonia valley feature at the point of ammonia
The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the
the biochemical events.
valley and terminated the aeration phase.

135
APPENDIX A1.7 - TRACK STUDY SEVEN (TS7) DOSP 4.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 7 follow

Table A1.7-1 Track study data and nitrogen mass balance calculations 9

TS7

TH/lCKSniOY

,hera!!OIl ICOOI rrPNJ • '" {SclTHl R/I!e-!.'>'U RattHtlu • R3hHOSt;

,
"
mgll Il'gIL f!'i01l,m/n m(JILlT'ln m9L.mn

UITROGf:J ~ASS aAlA/-ICE

0,19 nlqlLml~

6E.05 rngilm'n.mg MlSs 50 60 70 ao 90 100 110 120


[ML5Sj SBR cycle time (minutes)
MLSS croln'll! f 7.~ h,n.1rI
Cy:leUrt0 146 (lleroble i SelLe + 10mlnutet fillllllddeG>Hl\)
Cyd~fd,fj'
r;.,lMlSSfeycle "
Figure A1.7-2 Air flow rate [ammonia nitrogen], and [COD]
fk,. MlSS Vo;ali111 31 FT".JiILcyde
R~:'c N I~ C~H,NDl 0,12
0';; UHl,N A rt".QIl,e,-eleN {!+) "'-Iou ()fN {-) "'G11n ofN ) The air flow rate profile provided an indication of the point of COD
O~I riOl'N ff'glLe)'tleN {(+)"'Ga:llofNt-l Jt lau[lrN)
Del NOyli ~J trglLcycleN {(+)"'Ga~ofN(-)"'!OuI)IN) depletion, point 1. There was no clear indication of the point of
hlatC)
E~l:n';4Irul ?ulm12ted N ammonia depletion.
'J.,:u:ctl!l(l!eqNkou ((t)., leu ~I N (_)::l Ollilnef N)

The nitrogen mass balance accounted for all but 3% of the systems
nitrogen,

200
30 7.7
::i'
0, 7.6
.s <5 TPN 100
Z 7.5
0..
t::. <0
Z 7.4 :l 5 :>
'" §.
d; :x:c. §. DO "-
~ 15 6' 4 '0"

r
e.
~~ 10
7.2
·100
~ 7.1
2
:E
~
7.0

6.9 10 20 30 40 50 60 70 ao 90 100 110 120 130


0 10 20 30 40 50 60 70 ao 90 100 110 120 130 Time (minutes)
SBR cycle tlma (minutes)
Figure M.7-3 [DO] and ORP
Figure A1.7-1 pH profile and [soluble nitrogen]
The ORP profile failed to show any features that could be correlated to
The pH profile had an ammonia valley feature at the point of ammonia the biochemical events.
depletion, point 1. The detection algorithm successfully identified the
valley and terminated the aeration phase.

136
APPENDIX A2.1 - TRACK STUDY ONE (TS1) DOSP 2.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 1 follow

Table A2.1-1 Track study data and nitrogen mass balance calculations 30

TS1
25

T~ACK6TUOY

Aerai:th ieOD} (TPNJ • -ISol TNI Rale4vss

130
"
,.

A~lIf",ge rale nitrifoea!hm 0,12 mglL.m!r.


5E·OS mgIL'Nn.mgMlSS
20 40 60 80 100 120 140
[MlSSj :.:504 mgIL SBR cycle tIme (minutes)
104 mg/Ldl.,' tAl5S
170 (Aeroblt; + Sell:.. -+- lOl1llMtEsf;!lar.dd'eeant)
Cyd»::I/d;r,'
De! MLSS f C)'Cte
vssrrss 0.110
Figure A2.1-2 Air flow rate [ammonia nitrogen], and [COD]
Jfl MLSS V~latil~
Ratio N if' C~h~Ol 0.12
OeltiH,N Ie mglLcyclell «(+)" iossorN (.)=GlI!s)afti)
The air flow rate profile provided an indication of the points of COD and
OeINO,-N 1 mglLcycle t-i ((+) .. Ga.I" elN H'=folisafN)
DMoo,-1i 15 mg!L:)'C~eN (+) "'Oalnof N{_)"'wssa(N}
ammonia depletion, points 1 and 2.
Nln:is A.(I.'HC) m!l!L~)'l::I!!N

E5Ulli"\.e;:I asslmi!ated N

The nitrogen mass balance procedure accounted for all but 11 % of the
systems nitrogen.

30 1.7 200

r2 25
7.6

Z 7.5 100
n.
t. 20 6
Z 7.4
:J :>
~ 15 7.3 a ~5
0
E

~ 10
04 Q.
rr:
7.2 B 0
EO.
Z 7.1
·100

b 7.0
NO,-N
6.9
0 20 40 60 ao 100 120 140
SBR cycle time (minutes)
o 20 40 60 ao 100 120 140
SBR cycle time (minutes)

Figure A2.1-1 pH profile and [soluble nitrogen]


Figure A2.1-3 [DO] and ORP
The pH profile had an ammonia valley feature at the point of ammonia
The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the
the biochemical events.
valley and terminated the aeration phase.

137
APPENDIX A2.2 - TRACK STUDY TWO (TS2) DOSP

The biochemical data, nitrogen mass balance calculations, and online data for track study 2 follow

Table A2.2-1 Track study data and nitrogen mass balance calculations
TS2 30

r
25

::l
r!li' 20 :5
£
0
0
l:1.
1. ~
0

'"
Z'
4 :.'1
Loss G~ln
i: 10
,m 13
2S
mg/l <'nOll mgIL mgil
Airflow

r,:rROGH~ MASS BAlMJCE

0.12 mgk.!'nin
4E~ m;jilmn.lTyMlSS 20 40 60 80 100 120 140 160 180
!MlS51 z001 fY>OJL SBR cycle time (minutes)
;;:«): tnglL.6ayMlSS
'207 minutes {A'$ol;!lc .. SeWe: .. 10 mlnutli!'s flU anti d~anI)
Crcl~s Ida)' 7
OliIMLSS!cyck! 29 ll'g:'U:;yc!e
vssrrss Figure A2.2-2 Air flow rate [ammonia nitrogen], and [COD]
OM MlSS Vol3~lv
Rat't N In c!H 1rl0 1
Jet NHr~~ A 20 mgll.cycll'!N i(+l"'!csscrt.l(.)::GalnoIN) The air flow rate profile provided an indication of the point of COD
;)elIJOI"'1 mllfLc,'ele f,I I {"Ill< Gain orN H '" loss 01 N)
DdNOr·'i 13 mglLCiCleN (\~J"'Gi!lJnol}l{-I"~noIU) depletion, point 1.
lib.;,. A·(B---C) (t.gllcyd"'N
Eo.tim3teJ .,uml~ted I"l mgil N to C~Hlt--I01
Uf"I<:e:.'Jtl\ed N:lo~s m\l'tl~ {{-+J"'l.:Y..-acfNH'" GainQIN)

The nitrogen mass balance procedure accounted for all but 14% of the
systems nitrogen.

30
200

100

:; ;;-
0,5 E
S
D..
5'4 0:
B- 0

·100
NH,·N

20 40 60 so 100 120 140 160 180


o 20 40 60 80 100 120 140 160 160
SBR cyc'c time (minutes) SBR cycle time (minutes)

Figure A2.2-1 pH profile and [soluble nitrogen] Figure A2.2-3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to

depletion, point 1. The detection algorithm successfully identified the the biochemical events.

valley and terminated the aeration phase.

138
APPENDIX A2.3 - TRACK STUDY THREE (TS3) DOSP 2.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 3 follow

Table A2.3-1 Track study data and nitrogen mass balance calculations
9
T83

25
TRACK' STUDY

fTPNJ ' "ISo! ml RahHclS R<rle-ms 'Ri'lte..Joss.

In
o
o
15
2.
~ 10
."
ffijj/l
17
IDglL mglLm!n t=.

tJlTRCGEN t.l'ASS BALANCE

0.132 mglt.rr!n
4E~ mglLmlnJng MLSS o 20 40 60 80 100 120 140 160
iMlSSj 3fl20 mg/l Time (minutes)
157 mg/Llla,\, MLSS
C.,.'C)etime 191 rrlnures (A~jblc.Se(l\4j+1llm!/'IlIt~!!lin&nd$!CMt)

C)'d~slday 6
;to mgll.,,~!e

VSS/TSS
Figure A2.3-2 Air flow rate [ammonia nitrogen], and [COD]
Dd MLSS Vcl~.i~e

Aal1) N In C!H,NO l 0.12


Did IfHrN A 11 mglLq-.::iIlN il+)"klllllofN(-}=GalnofNJ The air flow rate profile provided an indication of the points of COD and
D~I NO,.N mgll...-ey;:hlN «(+}'"GalnotNH,"jcssotNj
fl'l\PLeyd~N i(il"'G~lnol,.jH"lcllsefh)
13 ammonia depletion, points 1 and 2.
N ~:lfi:i A-(B-tC)
BtimlIenl!SlirrJhriedti mglL N 10 c!H,No,
Uraccn.mted tllo~t mylLN \(+j"!OliliCIN{-;",Qai1C1fN)

The nitrogen mass balance procedure accounted for all but 4% of the
systems nitrogen.

30 7.7 9, 200

:J 7.6
t;, 25 ~ TPN
..sz ' 7.5 ORP
100
t 20 NO,·N
7.4
!?[
0 :x:
15
~ 7.3
'"
Z !l2
a
~
10
7.2 o

~
7.1
5
t=. 7.0
NO,·N
0 6.9
20 40 60 80 100 120 140 160 20 40 60 80 100 120 140 160
SBR cycle time (minutes) SBR cycle time (minutes)

Figure A2.3-1 pH profile and [soluble nitrogen] Figure A2.3-3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion. point 1. The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase. Note the test samples taken
at 160 minutes were lost due to an experimental error however points
indicate the ammonia valley occurred at the point of ammonia
depletion.

139
APPENDIX A2.4 ~ TRACK STUDY FOUR (TS4) DOSP 2.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 4 follow

Table A2.4-1 Track study data and nitrogen mass balance calculations
9 200
T84

TRACK ST\)DY
100
rrPN) • - {So! TN] R,IIH1)lll RallHa~,

~
.§.
5

5" 4 8:
o
B

·100

I] 156 mgIl.JTtn +-----,------r-----r----~------r_----._~--+.200

:rAts!>j
AE.oo
3944
mlllLrrlnSl'lG I.1ls$
l1:>gtl
o 20 40 so 80 100 120 140
MlSS ..iullg!!! 24hDl1f~ 3m ITgfLdaF' MlSS SBR cycle lime (minutes)

10
DJt !.iLS$" f cyc!e
Figure A2.4-2 [DO] and ORP
V$.S{fSS
~IMLSSYofah!1I

R~tt" I~ In C,HrNO. 0,12


The ORP profile failed to show any features that could be correlated to
OfltJH,·n A i8 lfiy,L:;yt:1. N \ (-+)" lou or N (-)" Gain ct t.!)
Del No,·tj mg'~Ij,>N i(")"GainofNH"'icSGcIN)
the biochemical events.
O"I'~~-Il 14 mg/kcydeN «(f)"G .. rn..,rN(-j"I"uoINl
fi I~~s A-HHCj :3 mqIl.c}'dl! N
Er1,w"ld auimFated N

The nitrogen mass balance procedure accounted for all but 1% of the
systems nitrogen.

30 1.1

~ 2'
1.S
E
ie
"!!.0 20 TPN
Z 7.4
t 'i
Z 15 7.3
d"'
i!i.

~
7.2
10
i!S.
7.1
!f

i!'.. 7.0

6.9
20 40 so ao 100 120 140
SBR cycle tlmo (minutes)

Figure A2.4-1 pH profile and [soluble nitrogen]

The pH profile had an ammonia valley feature at the point of ammonia


depletion, point 1. The detection algorithm successfully identified the
valley and terminated the aeration phase. There was a problem with
the logging of the air flow rate and data was not obtained for this track
study. As the air flow rate profile was not available the [COD] data was
included on the pH profile.

140
APPENDIX A2.S - TRACK STUDY FIVE (TSS) DOSP

The biochemical data, nitrogen mass balance calculations, and online data for track study 5 follow

Table A2.5-1 Track study data and nitrogen mass balance calculations
9 200
T84

TRAC,(STllOY 7
100
[TPNj • ~ [Sol TN! Rate-lasli Rate.j(l$" • Rd!~

<:! ;;:-
§' 5 S
Q.
6' 4 0::
e. DO
0

!V -100

tHTROGEI,j MASS aALANCE

+-----~----r_----.-----r_----.-----.---__+.200
0,14 rtll1lLmln
5£4)5 mgIL'llm.iTlIIMLS$ o 20 40 60 80 100 120 140
!MlSS! l'lS"ll
SBR cycle time (minutes)
MLSScltl."gelZ4hpur$ 2M: mgtt.d~y MlSS
1&2 mlnLrtes (Anoble ~ Sellla .. 10mt'lules fI!I anljdecanlj
G)'clesl d~~
D~r MLSS t cycle
Figure A2.5-2 [DO] and ORP
VaS/TSS
Z2
Ra\JoUb,C!H,NO., 0.12
The ORP profile failed to show any features that could be correlated to
OtINK.\-14 18 rnglL.c)'C!IjN {(+)'" losJlo[N H"'GainorN)
De: ~JOrN molLcyete N «(+).,. GlIln of tl H "'!oJ, of N) the biochemical events.
DeIN0l't~ 12 mglLcycleN «(t}"'GiI;)!}oINH "'!oilnl N)
IHJS3 "dOtel m9lLcrclnf/
ESLmEledas"f"'i:al\1dfl
Un~cco,n;etlNj*,f>

The nitrogen mass balance procedure accounted for all but 10% of the
systems nitrogen.

r
30

- 7.6

1.5

1.4

l:
1.3 c.

7.2

1.1

7,0

20 40 60 60 100 120
N0 2-N
...
140
SBR cycle time {minutes}

Figure A2.5-1 pH profile and [soluble nitrogen]

The pH profile had an ammonia valley feature at the point of ammonia


depletion, point 1. The detection algorithm successfully identified the
valley and terminated the aeration phase. There was a problem with
the logging of the air flow rate and data was not obtained for this track
study, As the air flow rate profile was not available the [COD] data was
included on the pH profile,

141
APPENDIX A2.6 - TRACK STUDY SIX (TS6) DOSP 2.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 6 follow

Table A2.6-1 Track study data and nitrogen mass balance calculations
30
TS6

rQAC';;STUDY

COD
f;;TR'OGEN MASS BALANCE

0.12 mglLrrlr;
4E,OS rnj}lLmlruflg MLSS
It!~sSj "04 o+-----,-----~----~- ___~-----.------~-L--+
mgil day o.,1LSS o 20 40 so 80 100 120 140
Cyc;",jr-.e
SBR cycle time (minutes)
CyclES I d1.y
Del MlSSlc;'t'e
vssrrss
ntl 1,'lSS Vol:W!"
Ra!ln NIn C~H1NOl 0.1;( Figure A2,6-2 Air flow rate [ammonia nitrogen], and [COD]
OelNHl,N MgILcyd1!N (('t)"'loslIvN{-j"-Ga!nofN)
O"IHOr-/l1 f't\ljILeydetl {it)"'OalnIlfHH:.luuotN)
OalNOrtl l11g/LcycieN {(+l"'Ga:'netNH=lo$SorN) The air flow rate profile provided an indication of the points of COD and
A-jEte)
("lm'raled 3u!m'laled N Mgt\.. N!o ClHrN~ ammonia depletion, points 1 and 2.
Una""':&I1!ed U los,. mpil N «+) =kin 01 tH')" Gain orN l:

rhe nitrogen mass balance procedure accounted for all but 16% of the
systems nitrogen.

30 7.7
pH 200

7.6
:J' 25
en
~ 7.5 100
Z ORP
~ 20
7.4 ;;-
Sf E
0 15
TPN
r 0
<S 7.3 0.
Sf NH,·N
" 0::
0
0 7,2 DO
,::;. 10
·100
~
<S
.. 7.1
5
7.0
o +----~... --_,_----_,_-....-_,_----_,_----_,_---"-+ .200
0 6.9 o 20 40 60 80 100 120 140
0 20 40 SO 80 100 120 140 SBR cycle time (minutes)
SBR cycle Urns (minutes)

Figure A2.6-3 [DO] and ORP


Figure A2.6-1 pH profile and [soluble nitrogen]

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to

depletion, point 1. The detection algorithm successfully identified the the biochemical events.

valley and terminated the aeration phase.

142
APPENDIX A2.7 - TRACK STUDY SEVEN (TS7) DOSP 2.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 7 follow

Table A2.7-1 Track study data and nitrogen mass balance calculations
30
TS6

TRACI(STUOY

jCODJ /NH:HIJ [N02.Nj {TPNI • - (Sol TN) Ra!e4aEIl Rale-It)S$ • Rille-lou

151 11
mgll moll.

NITROGEN MASS eALANCE

0.12 ~.roin O+---~--~----~--~--~----r---~--~----~~


o 20 40 60 00 100 120 140 160 180
!MlSSJ !"',)ll SBR cycle time (minutes;
r:'liJ/L.rbyMLSS
C}cleNN~ '"
200 mlrMlIl'i

Cjl;lt~lo")' 7
De! MLSSJcycr,;
vss;-rss o.ro Figure A2.7-2 Air flow rate [ammonia nitrogen], and [COD)
,::llel Mi.SS VQ~tile
Kalin N In C~HlriO~
OflNH1"N HI f!1SlLJ;~I>lN «+)"'ka;JI;<J(N[-i=cG!iflQINi The air flow rate profile provided an indication of the points of COD and
(j mgil~yd!)N {(+) "'/33In c.i N (-j"'Io",QfN)
D~I tlC,.N C I'IIllILq:cle N {f+)=-Galn al N (-) ><lonolN) ammonia depletion, points 1 and 2.
h'GH A-(a+q
f$llr..netfanlmil:r,!).tftl mg/l Ntl! C~H1f'tOt
Ura:roun\ed N 1011$ '«\lit N {(-+-) '" l.:Iu cf II {-J .. Gt'n oj N)

The nitrogen mass balance procedure accounted for all but 11 % of the
systems nitrogen.

30

7.6
25 200
::1
I 7.5 8
Z
0.
20
to 7.4
7
100

~~ 15
Z ~5 :;:-
E
0• 10
§ 0
~ 0 4 0..
0::
Z
b
e. 0
:l: 3 00
~
-100

40 60 80 100 110 140 160 180


SBR cycle time (minutes)
o +---~--~""-T---~----r----r---,---,----,---L -200
20 40 60 80 100 120 140 160 180
SBR cycle lime (minutes,

Figure A2.7-3 [DO] and ORP


Figure A2.7-1 pH profile and [soluble nitrogen]
The ORP profile failed to show any features that could be correlated to
The pH profile had an ammonia valley feature at the point of ammonia
the biochemical events.
depletion, point 1. The detection algorithm successfully identified the
valley and terminated the aeration phase.

143
APPENDIX A3.i - TRACK STUDY ONE (TSi) DOSP 1.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 1 follow

Table A3.1·1 Track study data and nitrogen mass balance calculations
30
TS1

25
1
;J
.,
g
10'
20
.~
C 15
0 NH,·N ~
~
;'i'"
,
Gam GJlin
\0
'0 ~i!'. 10
mgil mil'L mg;' m>' "'<Pl fllglLm,n mgfLrnIn mg/L!rin

tH'''ROGENMASS BALANCE

rng/lm'll
31£-0:5 m;vL min IY\: MlSS
l Ml5SJ 2912
20 40 60 80 100 120 140 160 180 200 220 240 260 230 300
!1'\ff'Ld~y MlSS
SBR cycle tlma (minutes)
rn:~IJI~" (A!?fO!:!:; • Sel1'e .. 10/T<'llule$ ffiI a~d decant)
Cp::~s I q~y

Db MLSSle'tcl"
VSSI;SS '"
0,60
mwLt$le

,jelMlSS W!31~~ 15 mglLtytle


Figure A3.1·2 Air flow rate [ammonia nitrogen], and [COD]
,,;rtia N In Cj HIND, <til
D~' tlHrN A ({+)"lc$$Q(N!-)=GeinofN r
((.''''Gain!lf'\l(-)cloS50W}
rIQlLcydtli
The air flow rate profile provided an Indication of the points of COD and
D~:!,~N c mgfLc],!;lolN (l+j'" Gain or N ("j = IO$.$oPl )

£5kMtel1
A-(fl-.<::)
~s~lm~llled N
'v fT'9'LcydeN
mglL N to C!H1NO,
ammonia depletion, points 1 and 2.
lIn;;;:col,mltd'~I""l;; mgiLN ((+)"'ltSilcti/(.) "GiJlno(t~)

The nitrogen mass balance procedure accounted for all but 27% of the
systems nitrogen.

7,60 . 200

:J
7,55
:1
1 TPN 71 ORP 100
Z
E
Sf
a '!i
~5 >"E
;;. E- o
Sf: o 4 0.
0::
0
~ 10 0

Z 31

e 5 ·100

O+~-~~r--··'-~~···-c-~~~~~-~~--r~~_~r~,~+
o 20 40 60 lIO 100 120 140 160 180 200 220 240 260 280 300
SBR cyGln tIme (mlnules)
o 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
SBR cycle time !minutes)
Figure A3.1·1 pH profile anQ [soluble nitrogen]
Figure A3.1·3 [DO] and ORP
The pH profile had an ammonia valley feature at the point of ammonia
depletion. point 1. The detection algorithm successfully identified the The ORP profile failed to show any features that could be correlated to

valley and terminated the aeration phase. the biochemical events.

144
APPENDIX A3.2 - TRACK STUDY TWO (TS2) DOSP 1.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 2 follow

Table A3.2-1 Track study data and nitrogen mass balance calculations
30
TS2

25
T=<O\CKSTUDY

~ 20
C
g'"
15
~
0 NHJ~N
~
*"
!,!.
? '"
~
-£ 10
i!:.

0)18 'TIgllJnin
'JE.{J5 mg/L 1l'n,mgf,llS$
1~'\lSSI ~912 rrg!1.
MLSS dn'">jJ"j14 h01,;(5 111 lTIgILdlJ,MLSS
20 40 60 ao 100 120 140 160 180 200 220 240 260 280 300
SBR cycle Urne (mInutes)

ZO mglLtycje
VSSIlSS 0.60
DellhSS Vclane If mgJLcYC!1!
Figure A3.2-2 Air flow rate [ammonia nitrogenl. and [COO]
mglLq'CleN (('t)""!ass<:ffIH"GlI'incfH)
mglLC'ttleU \(");:G.,!nofNH"lossc/~J)
The air flow rate profile provided an indication of the points of COD and
OeltmrN rr,gILcyr.:ieN ((+)""GaJnoiNH"'IDSl!O(~J)

tjlOS$ A-{B+C)
2s!imalwa'f1'i,mhledN m!yt.. N 10 c,!i.NO:
ammonia depletion, points 1 and 2. The point of ammonia depletion
UI'I3cco<l!'lI;eu N 10m> mgflfl i(~)"'lcs~QrNH"'G;Jrj'1cfN)
was slight.

The nitrogen mass balance procedure accounted for all but 35% of the
systems nitrogen.

ORP 100

~5
g
04
o

-100

20 40 60 80 100 120 140 160 180 200 220 240 260 2aO 300
SBR cycle lime (mInutes] o 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
SSR cycle time (minutes)

Figure A3.2-1 pH profile and [soluble nitrogen]


Figure A3.2-3 [001 and ORP
The pH profile had an ammonia valley feature at the point of ammonia
The ORP profile failed to show any features that could be correlated to
depletion. pOint 1. The detection algorithm successfully identified the
the biochemical events.
valley and terminated the aeration phase.

145
APPENDIX A3.3 - TRACK STUDY (TS3) DOSP 1.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 3 follow

Table A3.3·1 Track study data and nitrogen mass balance calculations
TS3 30

TRACK6TUOY
25
7
~
g 20
@'
0
0 15
~
!Z
ltl .. \.,
;I:
11
" 2:.

0.07 m;YLttin

Il.1lSSJ 3105 frgil o 20 40 60 BO 100 120 140 160 180 200 220 240 260 280 300
MlSS dUl(j\lil! /24 t.~1i mg/LdiIY !,lUIS
SBR cycle time {minutes)
Cyd"N/dllY
Dtl r.\LSSf~f';e
VSSISS
DciMlSS V~~e
1",\", N In C\'M~O. OX, Figure A3.3-2 Air flow rate [ammonia nitrogen], and [COD]
0", !<Hrtl 1& mgll..~tlN t(+)"'iOSfQfNO'"GamoIN)
O~l HO,.N mWLcye:eN (lfj"Gatnaffli-l"'louoIN)
Odil3j ,f( l'tIQfLcye!etl j(+l "'G:I'nofN H""buoIN) The air flow rate profile provided an indication of the points of COD and
,'.Ikm A·(IHC) rngiLeytieN
f~limale4 ",.,ltnilat!l6 N mglL U !¢ CtH,t.'O; ammonia depletion, points 1 and 2.
tt",,'lll {(.)=.jes$<llNH"'GalnorJ~i

The nitrogen mass balance procedure accounted for all but 37% of the
systems nitrogen.

200

ORP
100

:::
o

DO -100

~---.-~-~- T-r'-~~-~--'---'-'::::~-,--~~~ 7.20


f-r-'-'--'-r--r·-'--'-~--·.--r~,-.-~--+-200
o 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
o 20 40 60 80 100 120 140 160 180 200 220 240 260 2BO 300
S8R cycle time (minutes)
SBR cycle time (minutes)

Figure A3.3·1 pH profile and [soluble nitrogen] Figure A3.3·3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase.

146
APPENDIX A3.4 - TRACK STUDY FOUR (TS4) DOSP 1.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 4 follow

Table A3.4-1 Track study data and nitrogen mass balance calculations
TS4

TRACK STUDY
:~
20
F
T7
~ \ '2
'5
§. 5 ;;;t
i5
l1f
0
~
.,
0

!1.
~
~
Lo!>s
143 11 "'"
11
10 NH"N
M;l'lVltS mg/L fT'gfL rnglLmln mglLmlh mgfUrnn ~

t/tlRcGEN ,,\ASS BALANCE

Q,Qa ml}'Lrn!n
A,f'ugemleniirilicllLcn
o 20 40 60 ao 100 120 140 160 180 200 220 240 260 280 300
2E·05 mg/l mln.rt'19 MlSS
SBR cycle time (minutes)
II..-'lSS!
Ml.ss.ch3nQ!l/24 flOUr..; i06 mgfLday ~),lSS
299 mlnlJlali {Attlll>in + Settle 1" 16 tr\!nlItes mand decanl)
Cy.:lulday
D":~LSS/C'fd1!, 22 mglLtyde
vSS/TSS a.an
Figure A3.4·2 Air flow rate [ammonia nitrogen], and [COD]
mg/Leyde
Del 'ilSS V\l!itUIe
Ratio Ii 10 C~HINOl "
(U2
The air flow rate profile provided an indication of the points of COD and
QeI~lH,"N A 20.4 mglLcycl!!N ((+):: loss of N (-) "" Ga~r; of N)
!lei ~JOI"N ,.S mglU:11'IeN ((+)-= Ga:nOrNi-l '" losso(N j
ammonia depletion, points 1 and 2,
Del tiOrN B, IfIgll.q'Glef.l {(+) '" Galn offj (-) "'-loss of N}
t1'oss A·(A+C) 1.;17 trg/d:yc;leN
£s!imlled as:;imi!aled N 22 mpll N 10 CsthHC J
Uh\ln:oiJn{ed 11 loss a.5 mglLN ((+) '" loss ()f N (-) "" Gain of N)

The nitrogen mass balance procedure accounted for all but 28% of the
systems nitrogen,

- 200
30 7.60

7.55
:j
Z 25 1 ' ORP
a.. 100
t:.
'f
7.50
6 J
20
<5
z
pH 7.45 ~ 5 ~
~-
.E.
z~
0..
15 =s. 0" cr:
o~
~ E
7.40
e. 4
0

f:. 7,35
10
'f ,100
:£ 7.30
f:.
7.25
~···,,-~·~-r--'--.-~-'--r--'--.- _ _'---r--+,200
+-~~'--~-r--'--.-~--,---r-,---,--."'::::'>r--"'r'---J. 7.20
o 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
o 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
SBR cycle time (minutes) SBR cycle time (minules)

Figure A3.4-1 pH profile and [soluble nitrogen) Figure A3.4·3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase.

147
APPENDIX A3.5 - TRACK STUDY FIVE (TS5) DOSP 1.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 5 follow

Table A3.5-1 Track study data and nitrogen mass balance calculations
T85

TRA.CKSTUD¥
~
g 20 '2'
'E
~
0
~
15 !<
!:l. 0
0::
~
:J: 10 ~
Ga'r, Loss
e.
" 5

3£·05
mgll-.m'n
rrglL.mln.moMLSS
o 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
3012 rr'lVl SBR cycle time (minut•• )
258 mqlLday MlSS

Gicffis!:ia),
Ott MlSS J c~cle 52 mGILcydc
V!.isrrss OBi) Figure A3.5-2 Air flow rate [ammonia nitrogen], and [COD)
De' "l~SS VV~\Le< 41 '1'I{I'1.,c\,cle
Rill!i,rHn CsHINo, 0.)2
Dtlt,'1,--N A 10 rrg;Lcycill N ((tj=I():$$Cffl (-) '" Gain 01 N) The air flow rate profile provided an indication of the points of COD and
0eliJD 1-N mglLC)'tltt~ ({+)"'Gai!loINi-j"IOUofUj
DilNOrN mJlL.C',t:leN (l+j"'GainoINH"'louoIN) ammonia depletion, points 1 and 2.
(~M'q!\!d llu1mH:Ilet ~ mail f~ 1::1 c.,H,I,o,
Ul'a~f.()unled N lass ((+) "1I;;.,.,nfUH"'Oain o[N)

The nitrogen mass balance procedure accounted for all but 22% of the
systems nitrogen.

30 7.60
200

7.55 8·
25
~ ORP
.§. 7.50 7
100
Z 20
"- 6
t:. 7.45
Z ::J >"
c\' 0,5 E
t:..
15 7.40 i g a
Z 5'4 !l.
~
c\' 10
7.35 e. 0
t:..
~ NO,·N
7.30
-100
:£ 2
t:.. 7.25

20 40 60 80 100 120 140 160 180 200 220 240 260 260 300
SBR cycle time (mInutes)
o 20 40 60 80 100 120 140 160 180 200 220 240 260 260 300
SBR cycle time (minutes)

Figure A3.5-1 pH profile and [soluble nitrogen) Figure A3.5-3 [DO) and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase.

148
APPENDIX A3.6 - TRACK STUDY SIX (TS6) DOSP 1.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 6 follow

Table A3.6-1 Track study data and nitrogen mass balance calculations 30
T36
25
7
TRACK STuDY
~
AefafM ICOD) g 20 6 ?
~
!;j'
a0
15 ~
!:!. 4 0

5[ ""
L
;;:
G;tln
:x: 10
lo"" ~
Mgll rrglL 2

"'TROGEN fYASS uAtAllCE


+-~--~~--~-,--'--'--'--.--~·-.--r--.--r--+O
13.0& Mgil.nin
2E--05 mg!Lmln.rngMlSS 20 40 60 80 100 120 140 160 180 200 220 240 260 2BO 300
)51)7 moll
SBR cycle time (minutes)
MLSS chanoe i 24 hntJIS 212 mgiL.nyMLSS
CI'del;m~

C,,:les/day ,
2ra mir~es (Ae,"oolc .. Seltle ~ 10 rrJrlJ!1m fi!l a'!d declmlj

DelMLS:Slc~cJe 41 mgILq'<:lr-
vssrrss Figure A3.6-2 Air flow rate [ammonia nitrogen], and [COD]
Dei MlSS V'o'aIW
Rtl.lkl N If! c,H1't0 1 Q,12
Cd NHy~. 15 mg!l.c~'JN !(+}"'!\tnI)(NH:Ga!nor~)
The air flow rate profile provided an indication of the point of COD
ON l~bt"N mll,Lc)'Glet1 i (+) = Gain of N (.)", kll.s uf N)
Defl10r N nglLq'l.:I!!:/i ((.)"'G/lmofNN"'losscfN)
depletion, point 1.
Nlo!!. A·(B+C) mglLC)"Cleti
Fsliroul«isssjrrJ,atedI. rrQ!l N to C!li;tiOl
UnaccO'"Tte<f N m'S mg/tll (t~),.ro$'(;'NO"'GlIh"rt~)

The nitrogen mass balance procedure accounted for all but 12% of the
systems nitrogen.

30 7.60 200

7.55
25
J
rZ 7.50
7 ORP 100
Q. 20
t. 7.45
~ ;;-
~ TPN 5
g E
15 7.40 i
~ 5' 4 0-
r.:
Z
6
B 0
0
7.35
10 3

r"
"'-
~
£
"'- DO

o 7.20 - -200
o 20 40 60 80 100 120 1411 160 180 200 220 240 260 280 300 o 20 40 60 80 100 120 140 160 180 200 220 240 260 2BO 300
SBR cycJe time (minutes)
SBR cycle time (minutes)

Figure A3.6-1 pH profile and [soluble nitrogen]


Figure A3.6-3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia


The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the
the biochemical events.
valley and terminated the aeration phase.

149
APPENDIX A3.7 - TRACK STUDY SEVEN (TS7) DOSP 1.0

The biochemical data, nitrogen mass balance calculations, and online data for track study 7 follow

Table A3.7-1 Track study data and nitrogen mass balance calculations
30
TS7
25
TRACK STUDY

lCO:l! ITPNI • - [Sol TN! Rat;l,·lou; Rale-IGt.S • Rale-'cs$: 2'


en 20
S '2
lO'
0
0 15
! ~
S!- o
'"~
~
2:.
10

II:TROOEN MASS BALANCE

A~ElilgB fIlle nr{,:;caFen O.oJ rr9iLmln


2E~ mgil mln.rng Ml!>!> 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
'I19/L
14 mglt<JayMlSS SBR cycle lima (minutes)
306 'T]inulES (Aerobic. 5e\1:", .. ;0 mlnllilfl fill and dIiC3i'\1)
CF~"5! ;Jay
D!:4MtSS I cycle
VSSfTSS 0,&)
Oe! MlSS Vm~ga 13 -"gtLcyde Figure A3.7-2 Air flow rate [ammonia nitrogen], and [COD)
Ralml4!r,C f H,ND} 0.12
Del Nh··-/ 17 mglL.eydeN {(+)"':(\$sQIN{-)=G/!:norN)
De! NCT~t mg!l..cyc!eli i(1)=G:ll'iofNl-)'"k1sSQ(N) The air flow rate profile provided an indication of the points of COD and
;)e,lNO)TN mglLqcb'N ((.) .. Gai;tc(f~H"'los~ofNj

Ij 11l~~ A·(B.C) m~Lcycl~tl


ammonia depletion. points 1 and 2.
E:;!amlt'd assimilated N rrmlL N 10 C,M,NOr
mg'LN (('-l"'ln$$ctN{")=Gmr1orN)

The nitrogen mass balance procedure accounted for all but 24% of the
systems nitrogen,

91 200
30

25 ~
roo
7,55
ORP
100
~
7.50
g'" TPN
20
Z
~ :;-

I'"
jl,
t:. E
~
S
a 15 7.40 :a. 5' 4 1::
~ e. 0
zh 10
+7.35
0
~ -100
7,30 2
~

~ 7,25

+-'~~'~'~-'~-~-'~~-"'--'-~~~-~-"''''''';F-+ 7.20
20 40 60 80 100 120 1.40 160 180 200 22Q 240 260 280 300 o 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
SBR cycle time (minutes) SBR cycle lime (minutes)

Figure A3.7-1 pH profile and [soluble nitrogen]


Figure A3.7-3 [DO) and ORP
The pH profile had an ammonia valley feature at the point of ammonia
The ORP profile failed to show any features that could be correlated to
depletion, point 1, The detection algorithm successfully identified the
the biochemical events,
valley and terminated the aeration phase,

150
APPENDIX A4.1 - TRACK STUDY ONE (TS1) DOSP 0.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 1 follow

r:
Table A4.2-1 Track study data and nitrogen mass balance calculations ~o

T82

~ L
25

rR,\CK sruOY

~ 20
g
-'-5;5
C' 15
0
!:!.
i 4 ~
t;::
NH,·N
3~
"4
L,,~~

21
Gain O~'11 LDn
13
~2:. 10

mg'L mgll mgll "'i", "


moJl rngtL.mIn mgll..mln Iri;'lmln
5 -"----x

Nl1fmG:EN MASS SA.ANCE.


O*---~--,-I--~--~--~----r--~~~L--+
C.05 m:vLrrln
:lE·05 IngiLmfnJ'i9 ML$$
o 50 100 150 200 250 300 350 400 450
3160 SBR cycle lime (minutes)
IMLBS(
1.',5$ ;:hange 124 h~':I" 53
"<>'
I1'l{1ILd~y r..iLSS

Del MLSS j t)';;I"


'",., mlnlks (AI!roblc .. Sertle t \0 minutes fin MId~;
Figure A4.2-2 Air flow rate [ammonia nitrogen]. and [COD]
fIlg,'Leyde
VosITsS OBO
D,,;IJ.LSS Vo'ali'a 13 'TIgfLc.ycle The air flow rate profile provided an indication of the point of COD
Ral'" N in c,;N~O, 0,12
21 ! (.. ) =IonolNH=-GA'i'I'ofNJ
Df'ijjH),N mglL.cydaU
depletion, point 1.
Dell~OrN mgILI'¥"Ie:~ (1+)" GahofNH '" toS!iol tt)
Delf~O,·N mglU:ye!el1 ((t)""GaJ\ofN'(.)" IoUofN)
12 m:)'Lcycwt!
F,limat:das.simUalt:dN MaJl N !o C~H!No,
\)rlatTr~nted t~ ""'5 11 miYU~ ((+j ,,10$$ vI N (-) <> Gairl 01 N)

The nitrogen mass balance procedure accounted for all but 36% of the
systems nitrogen.

30 : 7.70
TPN 6
7.65
~
0)
25 ORP
g 7.60
Z
a. 20 S-
t:. 7.55 f:
0
!?f
0 15 &:0
~

~2:. 10
7.45
.100
7.40
'f
-£ DO
~ , 7.35
·200
0 +-----..,.-----,---_-.__-=:::-,-_...L 7.30 50 100 150 200 250 300 350 400 450
0 100 200 300 400 SBR cvcle time (minutes)
SBR cycle tim. (minutes)

Figure A4.2-1 pH profile and [soluble nitrogen] Figure A4.2-3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to

depletion, point 1. Tile detection algorithm successfully identified the the biochemical events.

valley and terminated the aeration phase.

151
APPENDIX A4.2 - TRACK STUDY TWO (TS2) DOSP 0.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 2 follow

Table A4.2-1 Track study data and nitrogen mass balance calculations
1
TS2

TRACK STUDY

~
§.
30

25

20}
i r, 7

NITROGEN MASS BALANCE

0Jj, mglLmln
-loss

" \4
mg;L "'91Ll'l'In InglLmln n".g.'LlT'in
5
-- 2

2£·0'5 mglLmln_tll9MlSS 50 100 150 200 250 300 350 400 450
(MlSSj "29 m,~
S13R cycle time (minutes)
MLSS tJnnge/24Imi.!f5 100 mgILdayMlSS
441 mirrut:&1l (Aerobic: + SmJ,;. t 10 minutes fill ~d de<::MI}
crdl;!> I day
Cd MlSS I CyClE 31 m!JIL~1a
VSSrTSS 0.80
Figure A4.2-2 Air flow rate [ammonia nitrogen], and [COD]
DeIMLS$VDI5:;lil 25 mgtLc~a
Rs,I 0 Ii ,0 C;H1NOl 0.i2
Illgf!"'cyc l eN
Dk~ NHrU A ((+)=IQ$SI)IN(-)-CalncfN)
The air flow rate profile provided an indication of the point of COD
mglLcydl'!N {(tj'=Gaifl!>fN{·)=klssnfN)
OF.INCA>j
ICi~S
mglL.eydeN {(t):;;Ga1nofNH=lostflfN)
depletion, point 1.
!J A,(B1Cl mgl~C;'!:leN
r,,\lfll'01J!o:{ assWaled N " mgll N 10 ~HltJOl
LJna<';"Q<H!!e;jtl!a~ ((+) "IoMtfti{.)"'GJll'jtIN)
" mg'lN

The nitrogen mass balance procedure accounted for all but 37% of the
systems nitrogen.

30 7.70

7.65
;g: 25
S 7.60
Z 100
~ 20
7.55
~
<5 TPN
i
tE, 15 7.50
Z
7.45
&
o
6' 10
tE,
z 7.40 ·100

tE, 5
2
7.35 DO
+--~ .......~--~~-'--~--..---r----'-~"'....--+ 7.30 ¥'<--,----,-"'--~ --r--~--r--~-~--+i ·200
50 100 150 200 250 300 350 400 450 50 100 150 200 250 300 350 400 450
S13R cycle time (minutes) S13R cycl. time (minutes)

Figure A4.2-1 pH profile and [soluble nitrogen] Figure A4.2-3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase.

152
APPENDIX A4.3 - TRACK STUDY THREE (TS3) DOSP 0.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 3 follow

Table A4.3-1 Track study data and nitrogen mass balance calculations
T83
30
,RACK STUDY

f,llrilLcn (COD] rrPNl • -[Sol TNI RalH'!lilO Rate-Iou • Ra[.:Hoss 25


7

i 20 6_
<::

!?
8 15
5£ ~
15
" ~ 4 0
mg.'L mall mlJlL mgfl m.~ mg/L.mln mwt.·mJn mgfLmh
!f ""~

3 <;;
:E 10
i!:.
NITROGEN t,!ASS l1ALANCE

0.05 mg!L.rr.'n
ZE·OS rr.gtLmlnmg MLss
rMLSS) 3:'(10 m<ll
~,IU.:S "hilng,..124 hour~ 10. mgll.dily MlSS
421 mlrtul(lS (Aerobic ~ Sellle + lOm!rrJles fin and da:anI;
50 100 150 200 250 300 350 400 450
Op)MLSSi{;';'I':14 32 my1ley"'!'! SBR cycle time (minutes)
VSB/TS:; a.aD
Del MlSS Vru:!;,.. 25 mg.'Lt~c~
FilliQ N in Cl~h~lOl o.lZ
D?II'.1I1l,N 2D mg!U::ycIeN (\t)'" lot$ effi (-):G'-'n olNj Figure A4.3-2 Air flow rate [ammonia nitrogen], and [COD]
DelN01AJ msJ'L.q'ClttN ({+}=Ga{)ofNH'''105!1otNj
Del NO,N mglLpYde N ({1) "'-Gain tit N H=los.s.oHi)
rn!JrLcycl~N The air flow rate profile provided an indication of the point of COD
f!i1'na!ed a$sJm!lated ~l
" mg/L Nlo C~H1NOI
Ufl3ttO-lnted N1IJss 12 rn<J1LN i (t/" l(jss~f ~l (-j'" Gab 01 N) depletion, point 1. There was also a slight indication of the point of
ammonia depletion, point 2.

The nitrogen mass balance procedure accounted for all but 46% of the
systems nitrogen.

30 T7.70 200
7.65
~ 25
.§.
7.60 ORP
Z 100
b 20
7.55
;>
TPN
e
7.50 ~
!l1
0
7.45

7.40 ·100

NO,·N 7.35

o +--.. - - , - - - , - - - , - - - r - - - - , - - , - - - - , - - : - , . - . t 7.30
~~~--·--~--~----r---_r----.----:=--- --4·200 __
50 100 150 200 250 300 350 400 450
50 100 150 200 250 300 350 400 450
SBR cycle time (mlnutesj
SBR cycle time (minute.)

Figure A4.3-1 pH profile and [soluble nitrogen] Figure A4.3-3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to

depletion, point 1. The detection algorithm successfully identified the the biochemical events.

valley and terminated the aeration phase.

153
APPENDIX A4.4 - TRACK STUDY FOUR (TS4) DOSP 0.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 4 follow

Table A4.4-1 Track study data and nitrogen mass balance calculations
30 \
TS4
25 ~\
TRAGKSTUOY

ICOO)

'"
mgil
l~'
2l)
mglL
0310

mg/l
,.
rnglL
,. MglLml,
wc¥l J:l;Yl.m!n fflUlLwJIl
"" 2
5
fJITROGEH MASS BALANCE

0.05 rrg!Ln'~fi
O+----r----r---.----.----~--~----~----,c~__+O
2E.05 fI".;tLmlrH"'!1ML.<;S
:'JLS!ij 295) 50 100 150 200 250 300 350 400 450
MlSS t'lanqf/2<:. hpl.Il~ 122 Jll9ILdayMLSS
SBR cycle time (minutes)
Cyck"!~:ly
,
470 mlnuiulO (/mable .. Se\t1,,-I-lllmiflVU;!l.(.;] !l;nd d~lIo'>tj

Ott"MlSS/cyCh'<
VSS!TSS "" mg/Lcycle

O~II,RSS \lcla/i,c Figura A4.4-2 Air flow rate [ammonia nitrogen), and [COD]
~1!Lo N In C~HINOl "
0.12
I"flfl't.cyde

O<l'lNHl'N 20 m!llLcyc~N i (+}'" IllsscfN (-) .. GalnofN)


OellJOrN
OalNorN
mWL_"
mjjlLqcleN
(+):GU!ofN(-) .. tonofN)
{(+}"'Ga!ncl NH~lonofN)
The air flow rate profile provided an indication of the point of COD
Niess, A.·le+CJ 14 rrgfl.c;)'C'eN depletion, point 1,
~,.\'n:rioo lI:falm,lat.eQ N mglL ~~ 10 ~HINOI
UnillCOl.lflle4 N jos,.
'0 ",,,,IN ((i)'" IoS1RfN {-)'" Gamol N)

The nitrogen mass balance procedure accounted for all but 39% of the
systems nitrogen,

30

r"
200
:::J . 7,65
'in 25
£ 7,SO ORP
!iF 100
a. 20
t::.
-c 7.55
;if ';;!
0
t:. 15 7.50 ?i r Ii
:>
E
0 a.
;if a 4
0 7.45 '"
0
10
t:.
;if 7.40
i!
i!E. 7.35
N0 2,N
-----~-~ --~.~.~~ ..
o +----;---~----,----,..._--,.---_,_---.----~-+ 7.30
o 50 100 150 200 250 300 350 400 450 50 100 150 200 150 300 350 400 450

SBR cycle tim. (minutes) SBR cycle time (minutes)

Figure A4.4-1 pH profile and [soluble nitrogen) Figura A4.4-3 [DO] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion, point 1, The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase.

154
APPENDIX A4.S - TRACK STUDY FIVE (TSS) DOSP O.S

The biochemical data, nitrogen mass balance calculations, and online data for track study 5 follow

Table A4.S-1 Track study data and nitrogen mass balance calculations
30 9
T85

25
TRACK STUDY 7
[COOl fTPtfl • -[BotlN] Rllht-lalls Rale-IMs • RallHOGS
~ 20 6_
.s ~

5~
II0 15 4
;:
0
!:!.
""
~
L,,, Ga'l' Galn
3 <C
". '"
rrj)'L mg!!. mY'l
13
rnJ!L "
moll rrglLlrln 1T1\jIL.mlfl rrgtLmtn
J:
~
10


I.JITROOEN MASS B4lANCE

O,(l$ mglLw ln
~g{~,mln.m9 MLSS 0
100 150 250 300 350 400 450
""
83
mglL
mll1L31ayMlSS
50 200
SBR cycle time (minutes)
Cyc'Blirrle rrinU!es (Atroblc. 'I- Sf'ffil'! + 10 rrlnWHI fril and decl'll)
Cydu I u"r
OEI',\LSst')'(l~ 24 Ma'l,c~eli!'
vssrrss 060
rnglLcyele
Del MLS5 Vola'.'i"
Figure A4.S-2 Air flow rate [ammonia nitrogen), and [COD]
[(alioilin C;I-<,I.;O) 0.12-
DeIUH.-N ISUI mgll.cyde N (!.}"'!i;)'5l:lfN(.J=Oa.... (jfN)
Ol!-jNOyN 56 IT>lf1tcychlN 1 (+)= Ga'(\o!fl (-i'" km: of Nl
p, The air flow rate profile provided an indication of the point of COD
Oe:NorN mglLcyd\tN ((+} "'OjllnalH H '" Ios.S(.n~)
r. ,c,,~ A·(B~C-I 1J.l mglLcyc'.. N
EslillUl&d 1'I~$bnlalfd N
depletion, point 1.
Un3cc(>!m\c4 N h,lst; "
1Q})
Il".g!LNlc C$H,tlO1
""LN ((+)"'lvss\JfN(·): GMofNj

The nitrogen mass balance procedure accounted for all but 41 % of the
systems nitrogen,

30 7.70

200

i 25
7.65

7.60
[ 20 ORP
100
t::. TPN 7.55
Z
Ch5 7.50 =a ~5 :>
2;. .§. E
Z NO,-N 0.
7.45 04
J' 10 '"
~ --~p B. 0

z 7.40
:£ 5 ·100
~ ~NO,.N 7.35

o +--~--r---;---'--r--~'---r':::"'-r---l- 7.30
50 100 150 200 250 300 350 400 450 ... ~.....~~-'..... ...~~..~ ..~....~-.-.~ ..~_~ ....~--+ -200
SBR cycle time (minutes) 50 100 150 200 250 300 350 400 450
SBR cycle Urne (minutes)

Figure A4.S-1 pH profile and [soluble nitrogen)


Figure A4.5-3 [DO] and ORP
The pH profile had an ammonia valley feature at the point of ammonia
The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the
the biochemical events.
valley and terminated the aeration phase.

155
APPENDIX A4.S - TRACK STUDY SIX (TSS) DOSP 0.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 6 follow

Table A4.6-1 Track study data and nitrogen mass balance calculations
30 9
T86
25
TRAcKsruOY 7
::J
0, 20 6_
S
!1r 5~
"
c 15
0 II:
B 0

- ~ ""
11 :£ 10 ~
rr91L !'!Jllil
""'''
1J
m!>', " n'lgfl...mfl m~.mI1l m?'Lmh ~

MTROOE.N MASS SALANCE:

(MlI'iSj
MLSS ll'lanQ"o 1241\OIJt&
Cyd~tlIr'It

cyclwtday
'elMLSSfcyel_
VSS/TSS
0,04

,,..
"'"
-
.. "'-
''''
""
oro
...."un'"
mgiLmhm\:l M~SS

f11QILdll}'MLS5

mgtl.t:lycla
(Aw.>bI",~St«.Io-+1{\mlr.J..tesMand~~
o 50 100 150 200 250 300
SBR cycle time (minutes)
350 400 450

CdIJLSSV<t,zt;. Figure A4.6-2 Air flow rate [ammonia nitrogen], and [COO]
fl-atio NIII C,H1NCtt "
0.12
Ill}'tcyeVi

'TlC,L,C)'CIt!N ((i;"'ID~'i"(lIN{·)"'G'f\llm}
OI:lWlrN A
" tnQiLqt!e N I (+j-" Ge1i'\t>fN(-j,.louMN) The air flow rate profile provided an indication of the point of COD
D.!NDj',N rrGil~.,'i ((t)"'GmmN{.)'" k1nofN 1
U~t A.:ihC) 12 mgA.A::ye'e N depletion, point 1.
E&!lmtltdaMImiJatadtJ 5 mgI\. N to C,lI,Ho,
Une.::t:<'lut1lOON :0)$1 InllilN ((+1'" IonoINH"Gll.lflofN)

The nitrogen mass balance procedure accounted for all but 32% of the
systems nitrogen.

30 .
f70
T200
7.65

7.60 ORP 100


6
7.55
:J
0,5 >E
7,50 i. E.
04
e. &
0
7.45
3
7.40 ·100
2

7.35

7.30
o 60 100 150 200 250 300 350 400 450 o 50 100 150 200 250 300 350 400 450
SBR cycle timo {minutes)
SBR cycle time (minutes)

Figure A4.6·1 pH profile and [soluble nitrogen] Figure A4.6-3 [~O] and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase.

156
APPENDIX A4.7 - TRACK STUDY SEVEN (TS7) DOSP 0.5

The biochemical data, nitrogen mass balance calculations, and online data for track study 7 follow

Table A4.7-1 Track study data and nitrogen mass balance calculations
30
TS7
8
25
TfiACKSTlJDY

::i'
0, 20 6_
g c
!!!
g 15 5~
~
S<. 4 0
'? '"
.!:
147 2'
Gain Lou
11
"".
17
:£ 10 3«
mgll mgll m,1I. ,,"S/l mil;!. mg/Lmln m!O"l.tT;!n m(VL!T~n
~

COD

006 mQILmln
~lSS

lMlSSj ""'"
J25tl
rrglLmlruTlII
m,1e 50 100 150 200 250 300 350 400 450
1m; m\l'l..day MLES
471 mmtJie$ (ACi .. bie: + S.Et',ia + ,:.1 min..1ln foil and dl!eanl) SBR cycle time (minutes)
Crd"l<laa~

Pa: MLSs I cyc:a 35 ffiGlLtyde

~: MLSS Valatie
O,H,tlO,
Rehc ,.., ,n
2'
R12
Ing!t.,eye!e
Figure A4.7-2 Airflow rate [ammonia nitrogen). and ICOD)
CdrJHj"N mg.'Ley::!eN i{+)::::tH&cfN(-)"'GainofN}
Oclflo,·ll mglLtyeklN ({+}"'G.tJn 01 N (.)"'JouofN)
DaiNo,.N mglLcycleN i (tl "" Ga!not h H "fMC. of N) The air flow rate profile provided an indication of the point of COD
Nlef.t A-{tHC) 17 mg1leydeN
e<lunated a.lJ!'milat..u N mllil N Ie CtH,N01 depletion, point 1.
UnilL1:oJrliedN"",s \3 n\{/'t..ti 1(+),. loGs;;!! N (.J'" Gain crN)

The nitrogen mass balance procedure accounted for all but 46% of the
systems nitrogen.

30 7.70
200
7.65
~ 25 8
g 7,60 ORP
~
b
~
~ 15
20
t
7.55

'5.
::i'
0,5
6
100

;;;-
E
7.50 g
~ 0'4 0-

a
~
10
7.45 B
3
Ct:
0

~
:c
NO,-N
JD 7.40 -100
5
~ 7.35 DO

0 +--~-~--"'---"---r--~--'~-'-~~-!-' 7.30 0 -200


0 50 100 150 200 250 300 350 400 450 0 50 100 150 200 250 300 350 400 450
SBR cycle time (minutes) SBR cycle time (minutes)

Figure A4.7-1 pH profile and [soluble nitrogen) Figure A4.7-3 [DO) and ORP

The pH profile had an ammonia valley feature at the point of ammonia The ORP profile failed to show any features that could be correlated to
depletion, point 1. The detection algorithm successfully identified the the biochemical events.
valley and terminated the aeration phase.

157
APPENDIX 81 PHOTOS OF EXPERIMENTAL HARDWARE

To help identify items some photos also have numbers beside specific pieces of hardware, these numbers
can be cross referenced to the system diagram below to obtain a description,

OATA
REPORTS
(2)

(13)

PROCESS
CONTROL
SYSTEM
(1)

WASTE
SLUDGE

,~{~ J~~ ~
TREATED
EFFLUENT

AIR
SUPPLY
(4)
--~--~----~)-------~ (9)
i:i
(liD)
0..
~
(5)

i
BROMLEY WASTEWATER TREATMENT PLANT - CHRISTCHURCH
MUNICIPAL SEWAGE

KEY
(1) Process controlier, microprocessor control, controlled and monitored on Athlon PC. Box incorporates 16 analogue inputs, 16 digital inpuls, and 16 digital outputs,
(2) Online rea Hime data reports for Ihe analysis of lreatment process
(3) Network link allowIng for remote access to process control system, tor example via intranet and interneL
(4) Regulated compressed air supply
(5) 500Lcapacily cold waslewaler storage with submersible feed pumps
(6) Irnmersion Ilea ted water bath
(7) (PAC) Proportional air valve conlroller
(8) Waste solenoids
(9) Air solenoid
(10) Variable rate air solenoids
(11) (FRS) Electronic flow rale sensors
(12) 10L capacity sequencing batch reaclors
(13) Mixers
(14) DO meters
(15) pH meters
(16) ORP cables
(17) Fill control pressure transducers

Figure 81-1 System overview

158
APPENDIX 81 PHOTOS OF EXPERIMENTAL HARDWARE

Figure 81-2 Main control board with reactors

Figure 81-3 Close up of both reactors in operation

159
APPENDIX 81 PHOTOS OF EXPERIMENTAL HARDWARE

Figure 81-7 Electronic air flow rate sensor


(Box duplicated for second sensor)
Figu re 81-4 Microprocessor based process
control system

Figure 81-5 Switched 230 volt outputs (Also Figure 81-8 On/Off upstream air solenoids
part of process control system)

Figure 81-6 Proportional air valve controller Figure 81-9 Variable rate air solenoid (1 of
2)

160
APPENDIX 81 PHOTOS OF EXPERIMENTAL HARDWARE

Figure 81-10 Computer linked to process Figure 81-13 Mixing and wasting hardware
control system

Figure 81-11 Feed storage and supply


equipment

nu l .. EVE i
CONTFO!.
PRESSURE

Figure 81-12 Fill control equipment

161
APPENDIX 82 SEQUENCING BATCH REACTOR OPERATIONAL PARAMETERS

Two reactors were built, each had a usable volume of OIl Hydraulic residence time (HRT)
8L. This allowed for a wastewater consumption of day 24hours
=6hours 1
{ 8Lvo/llme - rct.-FloW'--f
32L day
around 32 L/day or 224 Llweek. With the operation of
Recommended HRT for full-scale (nitrifying) sequencing batch reactor
two reactors in parallel this amounted to the
Range 12-50 hours
consumption of approx 450 Llweek of wastewater.
Volumetric loading
The largest wastewater cooler that could be purchased 280m! BODs. 32L Flow.J.... Volume- rct.-g _.~. 1000L' = 1.12 kg BOD, - applied
{ L day EL JOOOmg lOOOg ml J mJ,day hi-volume
was 500 L, thus the decision to use an 8 L reactor Recommended volumetric loading for full-scale (nitrifying) sequencing batch reactor

limited the need for wastewater collection from the main Range 0.08 - 0.240 .::.:=-=:.:.
--?-....

wastewater plant to frequencies of about once per week.

" Food to microorganism ratio (FfM)


Operational parameters for the sequencing batch
reactors used in this research follow,
Recommended FIM range for full-scale (nitrifying) sequencing batch reaclor

II Percentage of volume decanted each cycle = Range 0.05 - 0.30 ~ ... '::::-CL=r:.'=

50%.
.. Estimated length of Cycle =3 hours, thus cycles
It was not physically possible to operate a lab scale reactor
_ {24hours CYCle} = Scycles within the ranges suggested for full-scale facilities. The
per day - day' 3hours day
ranges for full-scale facilities are provided as a comparison
" Estimated flow per day =4L per cycle, thus flow only and were obtained from Metcalf and Eddy (2001)

per day 8CYcles 4L} 32L


{ ~'cycle day

to The measured influent total [COD] was 650


mg/L. Therefore the initial [COD] = 650 mg/L x
50% reactor retained 325 mg/L. The initial
[BOD 5 ] -325 mg/L x 75% -280 mg/L.

.. Operational [MLSS] was around 3000 mg/L.


Recommended [MLSS] range for full-scale
nitrifying sequencing batch reactors is 1500-
5000 mgfL

162
APPENDIX 83 DETERMINATION OF AERATION SYSTEM EFFICIENCY

The transfer of air from the gas to the liquid phase was necessary
to supply oxygen to the processes. Air was pumped into the
reactor through an aquarium type porous stone diffuser.

The gas transfer properties of the system were analyzed to assess


aspects such as the rate and efficiency of oxygen transfer. The
transfer efficiency of the system was determined using the
procedure outlined in Metcalf and Eddy (2001). As the aeration
rate was < 3 Llmin a series of aeration rates from 0-3 Llmin were
selected. To enable accurate readings to be taken a Gas
Chromatograph (Ge) lab grade micro air flow controller was fitted.
The actual flow rates were determined manually by the water
Figure 63·2 Dissolved oxygen concentration during the re-
displacement method as illustrated in Figure 83-1.
oxygenation process.
'lWA.V
flOW
VALVE

Equation 63-1 Oxygen mass transfer coefficient in clean


water.

Table 63-1 Experimental data for 857 mUmin flow rate.

Time [DO] Cs-C t

2 1.7 7.6
Figure 63·1 Experimental hardware used to determine oxygen 5 3.7 5.6
transfer parameters.
12 5.6 3.7
16 6.7 2.6

The procedure involved filling the reactor with clean water. The
dissolved oxygen was removed from the water by the addition of Cs == 9.3 mg/L @ 20 °c
sodium sulphite. The water was then re-oxygenated to near
Example calculation Determination of Kla for 857 mUmin
saturation level. This procedure was repeated for a series of DO
flow rate.
flow rates while readings were manually taken for [DO] versus
time. Figure 83-2 illustrates the oxygen transfer rates. The Using equation 83 -1 Kla:: 2.303(IOg7.6-logI.8!(60)
23-2 )
dissolved oxygen transfer rates were determined from the slope of
the profiles; the slope from 0-5 minutes was taken to calculate a
transfer rate in terms of mg/L.min. 8y difference of air supplied to
Table 63-2 Summary of the main dissolved oxygen transfer
air transferred the efficiency was able to be obtained directly. The
parameters.
oxygen mass transfer coefficient was determined using equation
83-1 from Metcalf and Eddy (1991). Table 83-1 shows the
experimental data for the 857 mLlmin flow rate while Table 63-2
contains a summary of the oxygen transfer parameters for the
different flow rates.

163
APPENDIX 84 PROCESS CONTROL SOFTWARE AND CONTROL ALGORITHMS

A software package was developed using the Lab The second derivative of pH has also been used by
View software package from National Instruments. Peng et al (2003) who detected the depletion of nitrate
nitrogen by using the second derivative of pH to identify
The software package allowed for ORT monitoring and the point at which the nitrate apex occurred. The pH
control via pH, air flow, [DO], and ORP. Control was control algorithm in this research allowed for the input of
achieved by the input of appropriate algorithms which three variables. The first variable was the line length;
provided for the detection of certain profile features this specified the size of the window the second
that had been correlated to biological events. derivative would cover, (usually set to around 30
minutes). The next variable was the second derivative
pH CONTROL ALGORITHM trip level or set point. When the calculated value
ORT control was only implemented into the mix- exceeded the set point it indicated the pH profile had
aeration phase and this was achieved by the use probably reached the ammonia valley. The trip level was
2
d pH usually set to about 0.001. The size of the trip level was
of 2
important as a low value could result in false triggers as
dt
a result of spikes or noise on the pH profile while a high
The sequence behind the algorithm is value could result in failure to detect the ammonia valley.

illustrated in Figure 84-1.


The final variable was the hold off time. This allowed for
a redundant period at the start of the cycle in which the
algorithm would not be applied. This was important
because the pH initially increased due to CO 2 stripping
from the reactor and then decreased again due to
nitrification activity; thus, the application of the algorithm
from time zero could result in a false trigger.

Included within Figure 85-1 is the aerobic max time, a


variable set as a default aeration time for use in case the
YES YES

algorithm failed to detect the ammonia valley.


NO
TIme> hold off time ?

DO CONTROL ALGORITHM
• YES
Another sequence within the software allowed control of
the dissolved oxygen concentration, illustrated in Figure
Terminate aeration
84-2. Note that this sequence also included the output
from the pH control algorithm.

Figure 84-1 Sequence behind the pH control


algorithm

164
APPENDIX 84 PROCESS CONTROL SOFTWARE AND CONTROL ALGORITHMS

The numbers for the dump were generated from the


difference between the actual [DO] and the DOSP. This
generated a number called the DO proportional gain.

For example if the DOSP was 3.0 mg/L and the actual
[DO] was 0.6 mg/L the (difference) DO proportional gain
would be calculated as 3.0 - 0.6 :::: 2.4. Providing a
larger number than say DOSP 3.0 mg/L [DO] 2.5 mg/L
= 0.5. A larger DO proportional gain resulted in a larger
valve opening-air flow rate.

A proportional gain adjustment factor allowed the


proportional gain to be adjusted I.e. the DO proportional
Figure 84-2 Sequence used to control dissolved gain would be calculated (say 3.0 mg/L - 0 mg/L :::: 3.0)
oxygen concentration and then multiplied by the factor i.e. 50% " 3.0 1.5.

The control of the dissolved oxygen concentration The integral gain factor was used to maintain the DO
required the input of a number of variables; these have flow rate when the proportional gain reached zero, (i.e.
been presented in Table 84-1. when the DOSP was reached).

Table 84-1 Variables used to control the dissolved When the [DO] was higher than the DOSP the negative
oxygen concentration scale factor was implemented. (it activated when the
proportional gain was negative). The negative scale
MaxImum aeration time Terminate aeration If ORT control fails
factor was multiplied by the proportional gain before the
Proportional gain adjustment factor Used to adjust air flow When [DOJ < OOSP
Integral gain factor Used to adjust air flow When [DO] = DOSP
number was added to the dump, in effect adjusting the
Negallve scale factor Used to adjust air flow When [~O] > DOSP
rate at which the air flow rate was reduced and the rate
at which the [DO] was reduced.

To control the dissolved oxygen concentration the In some situations the proportional gain method could
system had to manage the air flow rate to the reactor not ramp up the air flow fast enough. For example in
by adjusting a variable rate air solenoid. This was situations were a low DOSP was specified the resulting
achieved by the use of a "number dump"; the degree proportional gain was also low. To ensure the DO flow
to which the valve would open was dependent upon a rate increased fast enough in these conditions a variable
value generated from this dump. Using this value the called the integral offset factor was included providing
control software could adjust the air flow rate from 0-9 an initial boost to the dump to jump start the process.
Llmin over 256 graduations (8 Bit system).

165
APPENDIX 84 PROCESS CONTROL SOFTWARE AND CONTROL ALGORITHMS

SELF ANALYTICAL ROUTINES


The control software also allowed for the development of self analytical routines (i.e. routines that could allow
the computer to grade its own performance and make adjustments to the process control parameters) that is,
it could detect if a cycle did not reach a set point within a reasonable time and set a new more appropriate set
point for future cycles.

GRAPHICAL USER INTERFACE (GUI) TO CONTROL SYSTEM


Control of the system was achieved via a graphical user interface (GUI). The main window displaying the
ORT profiles and immediate values is shown in Figure B4-3. The ORT parameters were plotted in real time
on screen; the software included various tools for viewing online parameters more closely, (zoom windows
and scrolling tools) . The immediate values for the ORT parameters were also shown in the column on the
right hand side. The ORT pH profile has been highlighted for Figure B4-3, note the operational [DO) at this
time was 7.0 mg/L as this photo was taken during the developmental stages of the research.

Figure 84-3 Main GUI display with ORT pH profile made bold

From the main control window there were icons to access other parts of the software; for example, the
ammonia valley detection algorithm variables are shown in Figure B4-4. This window could be opened from
the main screen via the "Air stop vals" icon, (bottom right on main GUI).

166
APPENDIX 84 PROCESS CONTROL SOFTWARE AND CONTROL ALGORITHMS

Figure 84-4 Ma in GUI display with ammonia valley detection window open

The primary control variables and calibration factors were accessed through the "Change values" icon, (lower
right main GUI) ; this opened another window as shown in Figure 84-5. The current sequence along with the
control variables could be changed from this window, (located at the bottom are the calibration factors for the
ORT parameters).

The force state icon (as shown at the bottom left of Figure 85-4) could be used to access a system over ride
window. This is illustrated in Figure 85-6. The over ride window enabled manual control of the main reactor
hardware, the manual control decisions superseded those sent by the main control software. This control
window was included for maintenance and calibration purposes. With reference to Figure 84-6 the variable
rate air valve control goes from 0-255 , there were 256 graduations (8 bit) but the scale went to 255 as the
software had to use the number 0 as the first graduation .

167
APPENDIX 84 PROCESS CONTROL SOFTWARE AND CONTROL ALGORITHMS

Figure B4-5 Main GUI display with primary control window open

Figure B4-6 Main GUI display with force state control window open

168
REFERENCES USED IN THIS RESEARCH
Alleman J.E. (1984) Elevated nitrite occurrence in biological
Abeling U., Seyfried C.F. (1992) Anaerobic-aerobic treatment wastewater treatment systems. Waf. Sci. Tech. 17,409-419.
of high-strength ammonia wastewater, nitrogen removal via
nitrite. Wat. Sci. Tech. 26(5-6), 1007-1015. Alleman J.E., Sweeny M.W., Kamber D.M. (1989) Automation
of batch waste water treatment systems using programmable
Abeling U., Seyfried C.F. (1993) Anaerobic-aerobic treatment logic controllers. Waf. Sci. Tech. 10(11),1271-1282.
of potato-starch wastewater. Wat. Sci. Tech. 28,165-176.
Anderson I.C., Levine J.S. (1980) Relative rates of nitric oxide
Abeliovich A, Vonshak A (1992) Anaerobic metabolism of and nitrous oxide production by nitrifiers, denitrifiers, and
Nitrosomonas europea. Arch. Microbiology. 158,267-270. nitrate respirers. Applied Environmental Microbiology. 51, 938-
945.
Addiscott T. (5 February 2000) Making a meal of it. New
Scientist- Opinion point-www.newscientist.com 48-49. Anderson I.C., Homstead J., Burdige D. (1993) A comparison
of NO and N2 0 production by the autotrophic nitrifier
Agrawal G.D., Lunkad S.K., Malkhed T. (1999) Diffuse Nitrosomonas europaea and the heterotrophic nitrifier
agricultural nitrate pollution of groundwater's in India Wat. ScI. Alcaligenes faecalis. Applied Environmental Microbiology. 59,
Tech. 39(3), 67-75. 3525-3533.

Akunna J.C., Bizeau C., Moletta R. (1993) Nitrate and nitrite Andreottola G., Foladori P., Ragazzi M. (2001) On-line control
reductions with anaerobic sludge using various carbon of a SBR system for nitrogen removal from industrial
sources: glucose, glycerol, acetic acid, lactic acid and wastewater Waf. Sci. Tech. 43(3), 93-100.
methanol. Waf. Res. 1303-1312.
Anthonisen AC., Loehr R.C., Prakasam T.B.S., Srinath E.C.
Alex J., Jumar U., Schutze M. (2003) On-line measurement (1976) Inhibition of nitrification by ammonia and nitric acid. J.
data of wastewater systems via WAP mobile phones. Wat. Sci. Waf. Poll. Con. Fed. 48, 835-853.
Tech. 47(2), 205-211.
Amant P.P., McCarty P.L (1969) Treatment of high nitrate
AI-Ghusain lA, Huang J., Hao O.J., Lim B.S. (24-30 July waste. Journal of the American Wastewater Association. 61,
1994) Using pH as a real-time control parameter for 659-663.
wastewater treatment and sludge digestion processes. In 1ih
Biennial Conference of the International Association on Water Andreadakis AD. (1993) Physical and chemical properties of
Quality, Budapest (Hungary). Water Quality International. pp. activated sludge flocs. Wat. Res. 27, 1797-1714.
159-168.

AP.HA (1998), Standard Methods for the Examination of


AI-Ghusain lA, Hao O.J. (1995) The use of pH as control
Water and Wastewater. 1ih ed. American Public Health
parameter for aerobic/anaerobic sludge digestion.
Association, Washington, D.C.
Environmental Engineering. 121 (3),22-235.

Alleman J.E., Irvine RL. (1980) Storage-induced denitrification


using sequencing batch reactor operation. Waf. Res. 14, 1483-
1490.

169
Asano and levine. (1996) Wastewater reclamation, recycling Bliss P., Barnes D. (1986) Modeling nitrification in plant scale
and reuse: past, present, and future. Wat. Sci. Tech. 33(10- activated sludge. Wat. Sci. Tech. 18(6),139-148.
11),1-15.
Blume D.J.w., Speece RE. (1991) A database of chemical
Asher B., Chozick R, Irvine RL. (1992) Treatment of a high- toxicity to environmental bacteria and its use in interspecies
strength, mixed phenolic waste in an SBR Wat. Env. Res. 64, comparisons and correlations. Research Journal WPCF. 63(3),
128-133. 198-207.

Azov Y., Tregubova T. (1995) Nitrification processes in


Bock E., Schmidt I., Stuven R, Zart D. (1995) Nitrogen loss
stabilisation reservoirs Wat. Sci. Tech. 31(12),313-319.
caused by denitfying Nitrosomonas cells using ammonium or
hydrogen as electron donors and nitrite as electron acceptor.
Baeza J.A., Gabriel D., Lafuente J. (2002) In-line fast OUR
Arch. Microbiology. 163(1), 16-20.
(oxygen uptake rate) measurements for monitoring and control
ofWWTP. Wat. Sci. Tech. 45(4-5),19·28.
Bonin P., Gilewicz M. (1991) A direct demonstration of co-
respiration of oxygen and nitrogen oxides by pseudomonas
Baker S.C., Ferguson S.J., lundwig B., Page D.M., Richter
nautical: some spectral and kinetic properties of the respiratory
O.H., Van RJ. (1998) Molecular Genetics of the Genus
components. FEBS Microbiology. Lett. 80, 183-188.
Paracoccus: Metabolically Versatile Bacteria with Bioenergetic
Flexibility. Microbiology and Molecular Biology Reviews. 62(4)
Brockett aD. (1972) Further aspects of the microbiological
1046-1078.
activity of the Mangere oxidation ponds. Report to the
Auckland regional authority research advisory SUb-committee.
Bang D.Y., Watanabe Y., Noike T. (1995) An experimental
Post Doctoral Studies University of Auckland Sep 1971-Aug
study on aerobic denitrification with polyvinyl alcohol as a
1973.
carbon source in biofilms. Wat. Sci. Tech. 32(8), 235-242.

Carter J.P., Hsiao YH, Spiro S., Richardson D.J. (1995) Soil
Ballmelle B., Nguyen KM., Capdeville B., Cornier J.C., Deguin
and sediment bacteria capable of aerobic nitrate respiration.
A (1992) Study of factors controlling nitrite build-up in
Applied Environmental Microbiology. 61 (8),2852-2858.
biological processes for water nitrification. Wat. Sci. Tech.
26(5-6),1017-1025.
Castignetti D., Hollocher T.C. (1984) Heterotrophic nitrification
among denitrifiers. Applied Environmental Microbiology 47,
Barker PS., Dold P.l. (1996) Denitrification behaviour in
620-623.
biological excess phosphorus removal activated sludge
systems. Wat. Res. 27, 617-624.
Cecil D. (2003) Controlling nitrogen removal using redox and
ammonium sensors. Wat. Sci. Tech. 47(11),109-114.
Bilanovic D., Battistoni F., Cecchi F., Pavan P., Mata-alvarez J.
(1999) Denitrification under high nitrate concentration and
Chambers B. (1993) Batch operated activated sludge plant for
alternating anoxic conditions, Wat. Res. 33(15), 3311-3320.
production of high quality effluent at small works. Wat. Sci.
Tech. 28(10), 251-258.
BIOlAB Scientific (2003) Product technical information.

Charpentier J., Florentz M., David G. (1987) Oxidation


Reduction Potential (ORP) Regulation: A way to optimize

170
pollution removal and energy savings in the low load activated Clark H.M" Cooper D., Cutler R. (1996) Dictionary of
sludge process, Wat, Sci, Tech. 19, 645-655, Chemistry (Text) Third Edition. Oxford University Press. ISBN
0-19-280031-0.
Charpentier J" Godart H" Martin G" Mongo Y. (1989)
Oxidation Reduction Potential (ORP) Regulation as a way to Collivignarelli C., Bertanza G, (1999) Simultaneous nitrification-
optimize aeration and C,N and P removal: Experimental basis denitrification process in activated sludge plants: Performance
and various full scale examples, Wat. Sci. Tech, 21, 1209- and applicability, Wat. Sci. Tech. 40(4-5),187-194.
1223,
Cost (2003) European cooperation in the field of scientific and
Charpentier J" Menardiere M" Vachon A" Martin G, (1991) tech nical research. http://www. ensic.inpl-nancy Jr/COSTl!VVVTP I
ORP as a control parameter in a single sludge biological
nitrogen and phosphorus removal activated sludge system, Cohen A, Hegg D., Michele M.de" Song Q" Kasabov N,
Water SA 17(2), 123-132, (2003) An intelligent controller for automated operation of
sequencing batch reactors, Wat. Sci. Tech. 47(12), 57-63.
Charpentier J" Martin G" Wacheux H" Gilles p, (1998) ORP
Regulation of Activated Sludge 15 years of experience, IAWQ Davis L.M., Cornwell AD. (1998) Introduction to environmental
19th Biennial Conference of Vancouver. June 1998, engineering, Third Edition. McGraw Hill. ISBN 0-07-015918-1.

Chen F, Xia Q" Ju Lu-Kwang, (2003) Aerobic denitrification of Delgenes J.P., Patureau D, (2004a) Laboratory of
Pseudomonas aeruginosa Monitored by Online NAD(P) H environmental biotechnology. The laboratory specializes in the
Fluorescence, Applied Environmental Microbiology, 69(11), research and development of biological processes for treating
6715-6722, effluent and solid waste coming from agricultural, industrial or
urban activities. Director: J.P, Delegenes Assistant-Director:
Chen, S, K., Juaw, C,K" Cheng, S,S, (1991) Nitrification and D, Patureau LBE-INRA Avenue des Etangs 11100
denitrification of high strength ammonium and nitrite NARBONNE Phone: +33 (0) 468.425.151 Fax: +33 (0)
wastewater with biofilm reactor. Wat, Sci. Tech. 23,1417-1425, 468.425.160 http://www.montpellier.inraJr/narbonne/GMGB03.htmi

Christensen I\I1,H" Harremoes p, (1977) Biological Delgenes J.P., Patureau D. (2004b) Laboratory of
denitrification of sewage: A literature review, Prog, Wat, Tech, environmental biotechnology. LBE-INRA Avenue des Etangs
8,509-555, 11100 NARBONNE Phone: +33 (0) 468.425.151 Fax: +33 (0)
468.425.160 http://www,montpeilier.inra.fr/narbonne/GM_GB02,html
Chudoba J, (1989) Activated sludge bulking control.
Wastewater treatment technology Encyclopedia of Demoulin G., Mervyn Goronszy K.W" Forsthuber E,
environmental centrol technology, Chereminsoff, Gulf (1997) Co-current nitrification I denitrification and biological P-
Publishing, Vol 3, Chapter 6. Removal in cyclic activated sludge plants by redox controlled
cycle operation, Wat. Sci. Tech. 35(1),215-224.
Csikor Z" Czako L" Mihaltz P., Holo J, (1996) Complete
nitrogen removal from waste and drinking water in a fluidized- Demoulin G., Rudiger A, Goronszy M.C., (2001) Cyclic
bed bioreactor. Food. Sci. Tech. Int, 2(3), 165-171. activated sludge technology - recent operating experience with
a 90,000 p.e. plant in Germany. Wat. Sci. Tech. 43(3). 331-
337.

171
Dijkman H., Strous M. (1999) Process for ammonium removal nitrification, and denitrification by Thiosphaera pantofropha.
from wastewater. htlp:llwww.anammox.com/application.html. Patent Biotech. Bioeng. 36, 921-930.
PCT/NL99/00446. Paper covering this process also available in

Environmental Science and Biotechnology. (2002). 1, 51-63. Gerber A, Mosert ES., Winter CT, de Villiers R.H. (1986)
The effect of acetate and other short-chain carbon compounds
Downes MT (1988) Aquatic nitrogen transformation at low on the kinetics of biological nutrient removal. Waf. SA. 12, 7-
oxygen concentrations. Applied Environmental Microbiology. 12.
54,172-175.
Glass C., Silverstein J. (1998) Denitrification, Kinetics of high
Drysdale G.D., Kasan H.C., Bux F. (1999) Denitrification by
heterotrophiC bacteria during activated sludge treatment. Wat. nitrite concentration water: pH effect on inhibition and nitrite

SA 25(3), 357-362. accumulation. Waf. Res. 32(3), 831-839.

Focht D.O., Verstraete W. (1977) Biological ecology of Goronzy M.C., Demoulin G., Newland M. (1997) Aerated

nitrification and denitrification, Adv. Microb. Ecol. 1,135-214. denitrification in full-scale activated sludge facilities. Wat. Sci.
Tech. 35(10), 103-110.

Francis Garrido (2000) Mechanisms of nitrogen oxide


production (N 2 0 and NO) by soils in aerobic conditions. Thesis Goronzy M. (1989) Selector design for high COD wastewaters,
ADEME/Conseil Regional de Bourgogne December 08, 2000 a case history. Waf. Pol/uf. Control. Fed. Proceedings of the
at Amphi Orbigny, Faculty of Sciences Gabriel, University of industrial waste symposium, San Francisco.
Bourgogne, 21000 Dijon.
Groffman P.M., Verchot LV (2004) Denitrification:
Frehmann T., Niemann A, Ustohal P., Geiger w'F. (2002) Underestimation in temperate forest ecosystems. Research
Effects of real time control of sewer systems on treatment plant report by an institute of ecosystem stUdies scientist report
performance and receiving water quality. Waf. Sci. Tech. 45(3), available at vV\Nw.ecostudies.org/research/reports/grofrep3.html
229-237.

Frette L., Gejlsbjerg B., Westermann P. (1997) Aerobic Gunatilaka A, Dreher J. (2003) Use of real-time data in

denitrifiers isolated from an alternating activated sludge environmental monitoring: current practices. Waf. Sci. Tech.

system. FEMS Microb. Ecol. 24, 363-370. 47(2),53-61.

Fuerhacker M., Bauer H., Ellinger R, Sree U., Schmid H., Gupta S.K., Raja S.M., Gupta AB. (1994) Simultaneous

Zibuschka F., Puxbaum H. (2000) Approach for a novel control nitrification and denitrification in a rotating biological contactor.

strategy for simUltaneous nitrification/denitrification in activated Environmental Technology. 15, 145-153.

sludge reactors. Wat. Res. 34(9), 2499-2506.


Haandel A C., Ekama GA, Marais G.R. (1981) The activated

Gejlsbjerg B., Frette L., Westermann P. (1998) Dynamics of sludge process-3 single sludge denitrification. Waf. Res. 15,

N2 0 Production from activated sludge. Wat. Res. 32(7), 2113- 1135-1152.

2121.
Halling S.B., Hjuler H. (1992) Simultaneous nitrification and

Geraats S.G.M., Hooijmans C.M., Van Neil EW,J., Robertson denitrification with an upflow fixed bed reactor applying

L.A, Heijnen J.J., Luyben K.CHAM., Kuenen J.G. (1990) The clinoptilolite as media. Water Treatment. 7, 77-88.

use of a metabolically structured model in the study of growth,

172
Hanaki K., Wantawin C. (1990) Nitrification at low levels of Henze M. (1989) Carbon sources for denitrification of
dissolved oxygen with and without organic loading in a wastewater. Seminar on nutrient removal from municipal
suspended growth reactor, Waf. Res, 24(3), 297-302, wastewater. 4-6 September 1989.Tampere University. Finland.

Hanaki K., Wantawin C., Ohagaki S. (1990) Waf. Res. 24, 289- Henze M. (1991) Capabilities of biological nitrogen removal
296, processes from wastewater. Waf. Sci. Tech. 23, 669-679.

Hao O.J., Huang J, (1996) Alternating aerobic-anoxic process Ho K.M. (1994-5) Biological nutrient removal in activated
for nitrogen removal: Process evaluation. Wat, Env. Res. sludge processes with low F/M sludge bulking control. Ph.D.
68(1), 83-88. Thesis. The University of Queensland Australia.

Harrison D.E.F. (1972) Physiological effects of dissolved


oxygen tension and redox potential on growing Holman J.B., (2000) Investigation of simUltaneous nitrification

microorganisms. Journal of Applied Chemistry and and denitrification within sequencing batch reactors and the

Biotechnology. 22, 417-440. use of ORP as a process control and monitoring parameter.
ME Thesis. The University of Canterbury New Zealand.

Heduit A., Thevenot D.R (1989) Relation Between Redox


Potential and Oxygen Levels in Activated-Sludge Reactors. Waf.
Holman J.B., Wareham D.G. (2003) Oxidation-reduction
Sci. Tech. 21,947-956.
potential as a monitoring tool in a low dissolved oxygen
wastewater treatment process. Environmental Engineering.
Heijnen J.J., Van Loosdrecht M.C.M. (2004) Delft University
129(1), 52-58.
clean technology institute, Research into aerobic denitrification
and heterotrophic nitrification. Study to identify the optimum
Huang J.C. (1995) Prevention and control of filamentous
process conditions required for aerobic denitrification. The
growth in activated sludge systems. Water SA. 21 (3), 229-236.
2004 research is based upon previous experiments with the
model organism Thiosphaera pantotropha which indicated the
Inamori U, Takahashi 1., Sudo R (1986) Effects of anaerobic
specific process conditions in which aerobic denitrification is
conditions on activated sludge processes. J.Jap. Sew. Wks.
likely to occur. (Experimental work is based on lab scale SBR
Ass. 23, 61-69.
activated sludge process). http://www.interducUudelft.nl

Irvine RL., Ketchum (Jr) L.H., Arora M.L., Barth E.F. (1985)
Heinen N., Norgaard P. (1998) Sequence Sequel. WQI May-
An organic loading study of full-scale sequencing batch
June.
reactors. Wat. Poll. Con. Fed. 57, 847-853.

Helmer C., Kunst S. (1998) Simultaneous nitrification I


Irvine RL., Murthy D.V.S., Arora M.L., Copeman J.L., Heidman
denitrification in an aerobic biofilm system. Wat. Sci. Tech.
JA (1987) Analysis of full-scale SBR operation at Grundy
30(4-5),183-187.
Centre, Iowa. Waf. Poll. Con. Fed. 2(59), 132-138.

Helmo K. (1993) Increasing wastewater plant capacity by


Irvine R.L., Wilderer PA, Flemming H.C. (1997) Controlled
online measurement of the microorganism activity and oxygen
unsteady state processes and technologies-an overview. Wat.
demand. In: Preprints of Upgrading of Wastewater Treatment
Sci. Tech. 35(1), 1-10.
Plants. Berlin. 21-24 Sept. 317-326.

173
Jenkins D., Sezgin M., Parker D.S. (1978) A Unified theory of
filamentous activated sludge bulking. Waf. Poll. Con. Fed. 50, Ketchum L.H. (1997) Design and physical features of
362-381. sequencing batch reactors. Wat. Sci. Tech. 35(1), 11-18.

Jenkins D, Palm J.C., Parker D.S. (1980). Relationship Khalil M.A.K., Rasmussen R.A. (1988) Nitrous oxide:trends
between organic loading and dissolved oxygen concentration and global mass balance over 300 years. Annuals of
and sludge settleability in the completely mixed activated glaciology. 10,73-79.
sludge process. Wat. Poll. Can. Fed.. 52, 2482-2506.
Jenkins C.J., Mavinic D.S. (1989) Anoxic-aerobic digestion of Killham K. (1986) Heterotrophic nitrification, Nitrification (lRL
waste activated sludge: Part II-Supernatant characteristics, Press Oxford). 117-126.
ORP monitoring results and overall rating system.
Environmental. Tech. Lett. 10,371-384. Koch FA, Oldham WK. (1985) Oxidation reduction potential a
tool for monitoring control and optimization of biological nutrient
Kawakami Y., * (1985) * Ferment. Tech. 63, 437-442. *other removal systems. Wat. Sci. Tech. 17, 259-279.
information unavailable.
Koch FA, Oldham WK., Wang HZ, (1988) ORP as a tool for
Kim H., Hao O.J. (2001) pH and Oxidation-reduction potential monitoring and control in bio-nutrient removal systems.
control strategy for optimization of nitrogen removal in an ASCEICSCE Environmental Engineering. Conference. July 12-
alternating aerobic-anoxic system. Wat. Env. Res. 73 (1), 95- 19, Vancouver BC. 162-170.
102.
Knowles R (1982) Denitrification. MicrobRev. 46, 43-70.
Kim D.J., Chang J.S., Lee 0.1., Han DW., Yoo I.K., Cha G.C.
(2003) Nitrification of high strength ammonia wastewater and Krishnamachari S., Clarkson WW. (1992) Effluent nitrite
nitrite accumulation characteristics. Wat. Sci. Tech. 47(11), 45- accumulation in the heterotrophic denitrification of high
51, strength industrial wastewaters. Environmental Engineering.
370-375.
Hayward K. (1998) BNR's Control Key. WQI Sep-Oct 1998
issue. Pages 14-15. Krul J,M. (1976) The relationship between dissimilalatory
nitrate reduction and oxygen uptake by cells of an Alcaligenes
Ketchum (Jr) L.H., Irvine R.L., Breyfogle RE., Manning (Jr) strain in flocs and in suspension and by activated sludge flocs.
J.F. (1987) A comparison of biological and chemical Wat. Res. 10(4), 337-341.
phosphorus removals in continuous and sequencing batch
reactors. Wat. Poll. Can. Fed. 59, 13-18. Kshirsagar M., Gupta A.B., Gupta S.K. (1995) Aerobic
denitrification studies on activated sludge mixed with
Ketchum (Jr) L.H" Irvine RL., Dennis RW (1979) Sequencing Thiosphaera pantotropha, Environmental Technology. 16, 35-
batch reactors to meet compliance. Water-1978, Amer. Inst,. of 43.
Chem Engr. Symposium Series. 75(190), 186-191.
Kugleman I.J., Spector M. (1988) Nutrient removal in staged
rd
biological reactors. Proceedings 3 WPCF joint technical
Ketchum (Jr) L. H., Liao P.C., Irvine R.L. (1978) Economics
seminar on sewage treatment technology May 12-14 1988
evaluation of sequencing batch reactors. Proc 33rd Industrial
Tokyo Japan. 4-27.
Waste Conf Purdue University, West Lafayette, Ind, 357-376.

174
Kugleman I.J., Spector M., Harvilla A, Parees D. (1991) aerobic rotating biological contactors. Waf. Sci. Tech. 23,
Aerobic denitrification in activated sludge. Environmental 1355-1363.
engineering special conference (Edited by Krenkel PAl. 312-
319. Am. Soc. Civil. Engrs. Matsuo Y., Hosobora K (1988) Proc 3rd WPCFIJSWA Joint
tech sem on sewage treat tech. Tokyo. 84-121.
Kuenen J.G., Robertson LA (1987) Ecology of nitrification and
denitrification. Cambridge University Press. pp 161-218. Meinardi C.R., Beusen AH.W., Bollen M.J.S., Klepper 0.,
Willems w.J. (1995) Vulnerability to diffuse pollution and
Langner C. L .. , and Hendrix P. F. (1982). Evaluation of a average nitrate contamination of European soils and
persulfate digestion method for particulate nitrogen and groundwater Waf. Sci. Tech. 31(8), 159-165.
phosphorus. Waf. Res. 16, 1451-1454.
Meirlaen J., Assel V.J., Vanrolleghem PA (2002) Real time
Lo K.V., Liao P.H., Van Kleeck R.J. (1990) A full scale control of the integrated urban wastewater system using
sequencing batch reactor treatment of diluted swine simultaneously simulating surrogate models Waf. Sci. Tech.
wastewater. Can. Agric. Eng. 33, 193. 45(3), 109-116.

Lo C.K., Yu C.w., Tam N.F.Y., Traynor S. (1994) Enhanced Metcalf and Eddy (1991) Wastewater engineering treatment
nutrient removal by oxidation reduction potential (ORP) disposal and reuse. Third Edition (Text). McGraw Hill
controlled aeration in a laboratory scale extended aeration Publishers. ISBN 0-07-041690-7.
system. Wat. Res. 28(10), 2087.
Metcalf and Eddy (2001) Wastewater engineering treatment
Lloyd D, Boddy L., Davies K.J.P. (1987) Persistence of and reuse. McGraw Hill Publishers. ISBN 0-07-112250-8.
bacterial denitrification capacity under aerobic conditions: The
rule rather than the exception. FEMS Microb. Ecol. 45, 185- Moriyama K., Sa to K, Harada Y, Washiyama K., Okamoto K.
190. (1990) Renovation of an extended aeration plant for
simultaneous biological removal of nitrogen and phosphorus
using oxic-anaerobic-oxic process. Wat. Sci. Tech. 22(7-8), 61-
Lukow T., Diekmann H. (1997) Aerobic denitrification by a 66.
newly isolated heterotrophic bacterium strain TL 1. Bio. Tech.
Lett. 19( 11), 1157-1159.
Moriyama K, Takahashi M., Harada Y. (1993) Retrofitting and
operation of small extended aeration plants for advanced
treatment some experiences in Japan. In: Preprints of Small
Massone AG., Gernaey K., Bogaert H., Vanderhasselt A,
Wastewater Treatment Plants. Trondheim (Norway). Tapir
Rozzi A., Verstraete W. (1996) Biosensors for nitrogen control
Publisher.
in wastewaters. Water Quality International 1 at h Biennial
Confenerce and Exhibition 23-28 June 1996 Singapore. 1,
Morris J.C., Stumm W. (1967) Redox equilibria and
183-190.
measurements of potentials in the aquatic environment. Adv.
Chem. Series. No. 67, Article 13, 270-285.

Masuda S., Watanabe Y., Ishiguro M. (1991) Biofilm


properties and simultaneous nitrification and denitrification in Mosey F. (1985) Redox potential in wastewater treatment. The
Chemical Engineer. May 1985. 21-24.

175
Munch EV., Lant p" Keller J, (1996) Simultaneous nitrifiation units with intermittent aeration process. Wat. Res. 25(11),
and denitrification in bench scale sequencing batch reactors, 1377-1388.
Wat. Res, 30(2), 227-284,
otlanabo N.L., (1993) Denitrification of ground water for
Nakanishi H" Fukagawa M., Takeuchi M" Ishikawa M., Ukita potable purposes. WRC report No.40311193.
M" Murakami S. (1990) Biological nitrogen removal in a
complete mixing type aeration with ORP control. Wat. Sci. Oxford dictionary of chemistry (1996) Third edition. Oxford
Tech. 22(7-8), 269. University Press ISBN 0-19-280031-0,

Neef et al (1996). This reference used in discussion on page Painter HA, (1977) Microbial transformations of inorganic
97" Please contact the author for further details with respect to nitrogen, Waf. Sci. Tech. 8(4-5), 3-29.
this reference, [email protected] (0064-27-241-
7043). Paul E., Plisson-Saune S., Mauret M" Cantet J. (1998)
Process state evaluation of alternating oxic-anoxic activated
Ng w'J., Sim T.S., Ong S.L., Ng KY., Ramasamy M., Tam sludge using ORP, pH and DO, Wat. Sci. Tech. 38(3), 299-
K.N. (1993) Efficiency of sequencing batch reactor (SBR) in 306.
the removal of selected microorganisms from domestic
sewage. 27(10),1591-1600. Payne w,J. (1981) Denitrification. Wiley. New York,

Nielsen w'M., Jansen C,J. (1996) The role of the sludge in Patureau D., Davison J., Bernet N., Moletta R. (1994)
nitrification inhibition tests, Wat. Sci. Tech. 33(6),93-100. Denitrification under various aeration conditions in
Comamonas sp., strain Sgly2. FEMS. Microb, Ecol. 14(1), 71-
Nielsen M" Revsbech N.P., Larsen L.H., Jensen LA (2002)
78.
Online determination of nitrite in wastewater treatment by use
of a biosensor. Waf. Sci. Tech. 45(4-5), 69-76,
Patureau D" Bernet N., Moletta R. (1996a) Study of the
denitrifying enzymatic system of Comamonas SP" strain
Norcross K,L" (1992) Sequencing Batch Reactors - An
SGL Y2, under various aeration conditions with a particular
overview, Waf, Sci. Tech. 26(9-11),2523-2526.
view on nitrate and nitrite reductases, Curr, Microb. 32(1), 25-
32,
Novak L., Larrea L, Wanner J" Garcia Heras ,I.L. (1993) Non-
filamentous activated sludge bulking in a laboratory scale
Patureau D., Bernet N., Moletta R. (1996b) Effect of oxygen on
system. Wat. Res. 27(8), 1339-1347.
denitrification in continuous chemostat culture with
Comamonas sp., strain SGL Y2. Journal of Industrial
NZWWA (1997) Nitrogen elimination in sewage treatment
Microbiology. 16, 124-128,
plants, Water and Wastes in New Zealand. No 97, September.

Patureau D" Godon J,J., Dabert P., Bouchez T., Bernet N.,
O'Neil M., Horan N. (1995) Achieving simultaneous nitrification Delgenes J.P., Moletta R. (1998) Microvirgula aerodenitrificans
and denitrification of wastewater at reduced cost. Waf. Sci. gen. nov., sp. nov., a new gram-negative bacterium exhibiting
Tech. 32(9-10), 303-312.
co-respiration of oxygen and nitrogen oxides up to oxygen-
saturated conditions. International Journal of Systematic
Osada 1., Haga K, Harada Y., (1991) Removal of nitrogen and Bacteriology. 48, 775-782.
phosphorus from swine wastewater by the activated sludge

176
Patureau D., Bernet N., Delgenes cI.P., Moletta R. (2000)
Effect of dissolved oxygen and carbon-nitrogen loads on Pochana K., Keller J., Lant P. (1997) Effect of level and range
denitrification by an aerobic consortium. Applied Microbiology of dissolved oxygen (DO) on simultaneous nitrification I
and Biotecl1nology. 54(4), 534-542. denitrification (SND). Proceeding of BNR3 conference.
Brisbane. Australia.
Patureau D., Helloin E., Rustrian E, Bouchez T., Delegenes
J.P., Moletta R. (2001) Combined phosphate and nitrogen Pochana K., Keller J. (1999) Study of factors affecting
removal in a sequencing batch reactor using the aerobic simultaneous nitrification and denitrification (SND). Wat. Sci.
denitrifier Microvirgula aerodenitrificans. Wat. Res. 35(1), 189- Tech 39(6), 61-68.
197.
Pochana K., Keller J., Lant P. (1999) Model development for
Peddie CC., Koch FA, Jenkins C.J., Mavinic D.S. (1988) simultaneous nitrification and denitrification. Wat. Sci. Tech
ORP as a tool for monitoring and control of SBR systems for 39(1),235-243.
aerobic sludge digestion. In proceedings of tl1e joint
ASCEICSCE Environmental Engineering Conference, July 12- Poth M. (1986) Dinitrogen production from nitrite by a
19 Vancouver Canada. p.171. Nitrosomonas isolate. Applied Environmental Microbiology.
52(4),957-959.
Peddie C.C., IVlavinic D.S., Jenkins C.J. (1990) Use of ORP for
monitoring and control of aerobic sludge digestion. Ra C.S., Lo K.V., Shin J.S., OH J.S., Hong B.J. (2000)
Environmental Engineering. 116(3), 461-471. Biological nutrient removal with an internal organic carbon
source in piggery wastewater treatment. Wat. Sci. Tech 34(3),
965-973.
Pedersen D.P., Jensen K., Lyngsie P., Johansen H. (2003)
Nitrogen removal in industrial wastewater by nitration and
Randal CW., Buth D. (1984) Nitrite build-up in activated
denitration 3 years of experience. Wat. Sci. Tech 47(11),181-
sludge resulting from temperature effects. Wat. Poll. Con. Fed.
188.
56, 1039-1044.

Peng Y.Z., Gao J.F., Wang S.Y., Sui M.H. (2002) Use of pH
Randal CW., Barnard J.L., Stensel H.D. (1992) Design and
and ORP as fuzzy control parameters of denitrification in SBR
retrofit of wastewater treatment plants for biological nutrient
process Wat. Sci. Tech 46(4-5),131-137.
removal. Water quality management library. Volume 5.
Technomic Pub. Co. Lancaster PA USA 25-56.
Peng Y.Z., Gao J. Wang S.Y., Sui M.H. (2003) Use of pH as
fuzzy control parameter for nitrification under different alkalinity
Rhee K.S., Lee J.J., Lee ST (1997) Nitrite accumUlation in a
in SBR process. Wat. Sci. Tech 47(11), 77-84.
sequencing batch reactor during the aerobic phase of
biological nitrogen removal. Biotechnology letters. 19(2), 195-
Pipes W.O. (1967) Bulking of activated sludge. Advanced
198.
Applied Microbiology. 9, 185-234.

Risholt L.O., Schilling W., Erbe V., Alex J. (2002) Pollution


Plisson-Saune S., Capdeville 8., Mauret M., Daguin A.,
based real time control of wastewater systems. Wat. Sci. Tech
Baptise P. (1996) Real-time control of nitrogen removal using
45(3),219-228.
three ORP bending-points signification control strategy and
results. Wat. Sci. Tech 33(1), 275-280.

177
Rittmann BE, Langeland w.E. (1985) Simultaneous
denitrification with nitrification in single-channel oxidation Sekine T., Iwahorik K, FUjimoto E., Inamori Y. (1985)
ditches. Wat. Poll. Can. Fed. 57(4), 300-308. Advanced control strategies for the activated sludge process:
instrumentation and control of water and wastewater treatment
Robertson LA, Kuenen J.G. (1983) Thiosphaera pantotropha and transport systems. In Proceedings of the 4th IA WPRC
gen. nov., a facultatively anaerobic, faculatatively autotrophic Workshop, Houston and Denver. USA. p.269.
sulphur bacterium. Gen. Microb. 129,2847-2855.
Stensel D.H., Coleman Denham B.W., Fleischman D.
Robertson LA, Kuenen J.G. (1984) Aerobic denitrification a (1995) Innovative process used to upgrade oxidation ditch for
controversy. Arch Microb. 139,351-354. nitrogen removal and SVI control. Presented at the Water
Environment Federation 68th annual conference and exposition
Robertson LA (1988) Aerobic denitrification and heterotrophic October 241995 Miami Beach Florida.
nitrification in Thiosphaera pantotropha and other bacteria.
Ph.D. thesis. Delft University of Technology. The Netherlands. Stouthamer AH., de Boer AP.N., Van der Oost J. (1997)
Emerging principles of inorganic nitrogen metabolism in
Robertson LA, Van Neil E.W.J., Torremans RAM., Kuenen paracoccus denitrificans and related bacteria. National institute
J.G. (1988) Simultaneous nitrification and denitrification in of informacia. Kluwer Academic Publishers 71 (1/2), p.33-41.
aerobic chemostat cultures of Thiosphaera pantoropha.
Applied Environmental Microbiology. 54(11), 2812-2818. Strous M., Van Gerben E., Kuenen J.G., Jettern M.S.M. (1997)
Effects of aerobic and micro-aerobic conditions on anaerobic
Robertson LA, Cornelisse R.DeVos P., Hadioetomo R., ammonium-oxidizing (Anammox). Applied Environmental
Kuenen J.G., (1989) Aerobic denitrification in various Microbiology. 63, 2446-2448.
heterotrophic nitrifiers. Antionie van Leeuwenhoek 56, 289-
299. Stross M., Van Gerven E., Zheng P., Kuenen J.G. Jetten
M.S.M. (1997) Ammonium removal from concentrated waste
Robertson LA, Kuenen J.G. (1990). Combined heterotrophic streams with the anaerobic ammonium oxidation (ANAMMOX)
nitrification and aerobic denitrification in Thiosphaera process in different reactor combinations. Waf. Res. 31, 1955-
pantotropha and other bacteria. Antonie van Leeuwenhoek. 57, 1962.
139-152.
Strous M., Fuerst J. A, Kramer EH.M., Logemann S., Muyzer
Robertson LA, Kuenen J.G. (1992) Nitrogen removal from G., van de Pas-Schoonen K, Webb R., Kuenen G.J., JeUen M.
water and waste microbial control of pollUtion. Cambridge S. (1999) Missing lithotroph identified as new planctomycete,
University Press. Nature. 29 July, 400, 446-449.

Rohlich GA (1948) Measurement and control of sewage Strous M. (2000) Microbiology of anaerobic ammonium
treatment process efficiency by oxidation reduction potential. oxidation. PhD Thesis. Delft University. The Netherlands.
Sewage Works Journal. 20(4), 650-653.
Stumm W. (1966) Redox potential as an environmental
Sasaki K, Yamamoto Y, Tsumuara K, Ouchi S., Mori Y parameter: Conceptual significance and operational limitation.
(1996) Development of 2-reactor intermittent-aeration activated Adv. in Wat. Poll. Res. Proc. of 3rd Int. Cont. of IAWPRC. Munich
sludge process for simUltaneous removal of nitrogen and Germany. 1, 283-308.
phosphorus. Waf. Sci. Tech. 34(1-2), 111.

178
Snoeyink V.L., David J. (1980) Water Chemistry (Text). John using a sequencing batch reactor. Wat. Sci. Tech. 47(10), 101-
Wiley and Son Publishers. ISBN 0-471-05196-9. 108.

Third K (2004) PhD research. University of Murdoch Australia.


Spalding R.F., Exner ME (1993) Occurrence of nitrate in
Supervised by Dr Ralf Cord-Ruwisch senior lecturer in
groundwater - A review. Environmental Quality 22, 392-402.
biotechnology, division of science and engineering. south
Street Murdoch Western Australia 6150. Email address of Dr
Subramaniam K, Keller J., Ho K.M., Johns M.R, Greenfield
Ralf Cord-Ruswisch [email protected] Tel: (618) 9360
PF. (1994) Efficient biological nutrient removal in high strength
2403.
wastewater using combined anaerobic-sequencing batch
reactor treatment. Wat. Sci. Tech. 30(6), 319-321.
Tiedje J.M. (1988) Ecology of denitrification and dissimilatory
nitrate reduction of ammonium. In biology of anaerobic
Subramaniam K, Keller J., Ho KM., Johns M.R, Greenfield
microorganisms ed A.J.B. Zehnder, pp 179-244. John Wiley
PF. (1994) Effect of pretreatment on the nutrient removal
and Sons. New York.
efficiency in high strength wastewater using SBR technology.
Proceedings of the second Australian conference on biological
Tonkovic Short G., Hansen P. (1997) Nutrient removal
nutrient removal from wastewater. Albury NSW Australia. 217-
experiences at Brushy Creek treatment plant -bulking control
226.
and enhanced denitrification. BNR3 conference proceedings.
pp297.
Sutherson S., Ganczarczyk J.J. (1986) Inhibition of nitrite
oxidation during nitrification some observations. Wat. Res. 21, Tomlins Z., Thomas M., Keller J., Audic J.M., Urbain V. (2002)
257-266. Nitrogen removal in a SBR using the OGAR process control
system. Wat. Sci. Tech. 46(4-5),125-130.
Suwa Y., Suzuki 1., Toyohara H., Yamagishi 1., Urushigawa Y.
(1992) Single-stage single-sludge nitrogen removal by an Turk 0., Mavinic D.S. (1986) Preliminary assessment of a
activated sludge process with cross-flow filtration. Wat. Res. shortcut in nitrogen removal from wastewater. Can. J. Civ. Eng
26(9),1149-1157. 13,600-605.

Tam N.F.Y., Wong Y.S., Leung G. (1992) Significance of Turk 0., Mavinic D.S. (1987) Selective inhibition: a novel
external carbon sources on simUltaneous removal of nutrients concept for removing nitrogen from highly nitrogenous wastes.
from wastewater. Wat. Sci. Tech. 26(5-6),1047-1055. Environmental Technology Letters. 8, 419-426.

Tchobanoglous G., Schroder ED. (1987) Water Quality (Text)


U.S. EPA (1987) Technology transfer design manual
Second Edition. Addison-Wesley Publishing. ISBN 0-201-
phosphorus removal. USEPAJ625/1-87/001 CERI Cinn., OH
05433-7.
45268.

Thain M., Hickman M. (1996) Dictionary of Biology (Text).


Van Niel EW.J. (1991) Nitrification by heterotrophic denitrifiers
Ninth Edition. Penguin Books NZ Ltd 182-190 Wairau Road
and its relationship to autrotrophic nitrification. Ph.D. thesis.
Auckland 10 New Zealand. ISBN 014-051·288-8.
Deft University of Technology.

Thayalakumaran N., Bhamidimarri R, Bickers P.O. (2003)


Biological nutrient removal from meat processing wastewater

179
Van Niel E.W.J., Sraber K.J., Robertson L.A, Kuenen J.G. Watanabe Y., Masuda S., Ishiguro M. (1992) Simultaneous
(1992) Heterotrophic nitrification and aerobic denitrification in nitrification and denitrification in micro-aerobic biofilms. Wat.
Alcaligenes faecalis strain TUD Antonie van Leeuwenhoek Sci. Tech. 26(3-4), 511-522.
62(3),231-237.
Water and Wastes in NZ (September 1997) page 39. Figure
Vanrolleghem PA, Lee D.S. (2003) On-line monitoring illustrating typical ORP profile, (Original figure slightly
equipment for wastewater treatment processes: state of the modified).
art. Wat. Sci. Tech. 47(2),1-34.
Wentzel M.C" Ekama GA (1997) Principle in the design of
Verstraete W (1975) Heterotrophic nitrification in soil and single sludge systems for biological removal of carbon,
aqueous media- A Review. Bull. Acad. Sci, U.S.SR. Bioi. Ser. nitrogen and phosphorus. Waf. Res, 69, 1222-1231,
4.515-530.
Westgarth WC., Sulzer DA, Okun DA (1964) Anaerobiosis
nd
Wanner J., Gran P. (1989) Identification of filamentous in the activated sludge process, Paper presented at the 2
microorganisms from activated sludge: A compromise between IAWPRC Conference, Tokyo.
wishes, needs and possibilities. Wat Res. 23(7), 883-891.
WRC (1984) Theory design and operation of nutrient removal
Wanner J, Kucman K., Ottoua U., Gran P. (1987) Effect of activated sludge processes. Water Research Commission of
anaerobic conditions on activated sludge in filamentous bulking South Africa.
in laboratory systems. Wat. Res. 21(12),1541-1546.
Yang L., Alleman E. (1992) Investigation of a batch wise nitrite
Wareham D.G., Hall K.J., Mavinic D.S. (1993) Real-Time control build-up by an enriched nitrification culture. Wat. Sci. Tech.
of aerobic-anoxic sludge digestion using ORP Environmental 26(5-6),997-1005,
Engineering. 119( 1),120-136.
Yoo H" Ahn KH., Lee H,J., Lee KH., Kwak YJ., Song K,G,
Wareham D.G., Mavinic D.S., Hall KJ. (1994) Sludge digestion (1999) Nitrogen removal from synthetic wastewater by
using ORP regulated aerobic anoxic cycles. Wat. Res. 28, 373- simUltaneous nitrification and denitrification (SND) via nitrite in
384. an intermittently-aerated reactor. Wat, Res, 33(1), 145-154,

Watanabe A, Miya A, Matsuo Y. (1984) Laboratory scale Yu R F., Liaw S.L., Chang C,N., Lu H.J., Chang WY (1997a)
study on biological phosphate removal using synthetic The monitoring and control using on-line ORP in the
wastewater. Newsletter of the IAWPRC study group on continuous flow activated sludge batch reactor system, Wat,
phosphate removal. 2, 40-43. Sci, Tech. 35(1),57-66.

Watanabe S., Saba K, Nogita S. (1985) Basic studies on an Yu RF., Liaw S.L., Chang C,N., Cheng W.Y. (1998) Applying
ORP/external carbon source control system for the biological real-time control to enhance the performance of nitrogen
denitrification process. Instrumentation and control of water removal in the continuous-flow SBR system, Wat. Sci, Tech,
and wastewater treatment and transport systems. Proceedings 38(3),271-280,
4th IAWPRC Workshop Houston and Denver, 641-644.
Zhao W,H" Mavinic D.S., William K.O., Frederic AK, (1999)
Controlling factors for simUltaneous nitrification and

180
denitrification in a two stage intermittent aeration process
treating domestic sewage. Waf. Res. 33(4), 961-970.

Zheng H., Hanaki K, Matsuo T (1995) Production of nitrous


oxide gas during nitrification of wastewater. Waf. Sci. Tech. *,
(*),103-111. "Information unavailable.

Zipper T., Fleischmann N., Hanerl R. (1998) Development of a


system for control and optimization of small wastewater
treatment plants using oxidation-reduction potential (ORP).
Waf. Sci. Tech. 38(3), 307-314.

181

You might also like