Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
EFECT OF REACTOR FEEDING PATTERN ON
PERFORMANCE OF AN ACTIVATED SLUDGE SBR
By
Francisco José Cubas Suazo
Thesis submitted to the faculty of the Virginia Polytechnic Institute and State University in
partial fulfillment of the requirements for the degree of
Master of Science
In
Environmental Engineering
Dr. John T. Novak, Chair
Dr. Gregory D. Boardman
Dr. Matthew J. Higgins
September 6, 2006
Blacksburg, Virginia
Keywords: Activated sludge, cations, feeding pattern, sodium, bioflocculation, settling,
biopolymers, sequencing batch reactor, divalent cation bridging.
Copyright © 2006, Francisco José Cubas Suazo
EFECT OF REACTOR FEEDING PATTERN ON PERFORMANCE OF AN
ACTIVATED SLUDGE SBR
Francisco José Cubas Suazo
ABSTRACT
The possible effects of changes in the feeding pattern on activated sludge properties related to
bioflocculation have been analyzed in lab scale sequencing batch reactors (SBR) in order to
determine if these changes in effluent water quality and settling and dewatering properties are
significant, so they can be considered in future studies or if they can be recommended as
crucial when operating and designing wastewater treatment plants. The activated sludge
process is widely used to treat wastewater from both industrial and municipal sources.
Biomass from industrial facilities containing high monovalent to divalent ion content usually
settles poorly, which leads to low quality effluents that fail to meet environmental
requirements. Therefore, the combined effect of feeding pattern plus the addition of sodium
to activated sludge reactors was studied in this experiment.
A series of SBRs were operated at different sodium concentrations that ranged from 1.5 – 15
meq/L and different feeding times that ranged from 1 minute to 4 hours. Biomass samples
were taken from each reactor to study the settling and dewatering properties and effluent
samples were used to analyze the amount of organic matter and exocellular polymeric
substances present due to deflocculation. As expected, the changes in feeding strategies
affected all of the properties measured. When the feeding time was maintained low (pulse
feed) the effluent quality and settling properties were the best. As the feeding time was
increased the effluent quality, settling, and dewatering properties increased suggesting that the
way in which the reactors were fed affected the overall bioflocculation process. The causes of
the high deflocculation observed are not well understood, but data suggest that a microbial
community change could have affected exocellular biopolymers which are believed to play an
important role on bioflocculation.
iii
This research demonstrates the importance of the interaction between cation content and
feeding pattern when operating a wastewater treatment plants and when reporting lab-scaled
results related to settling and bioflocculation.
iv
ACKNOWLEDGEMENTS
The author would like to express his appreciation to Dr. John T. Novak, academic and
research advisor, for his support, guidance and constant collaboration to make possible the
completion of this work. Special recognition to both committee members, Dr. Matthew J.
Higgins for his valuable time, recommendations, and contribution on the microbiological
analysis related to this study and Dr. Gregory D. Boardman for his contribution and ideas
used on the methodologies and analysis performed on this study.
Thanks to the staff of the OAS-LASPAU program, especially to Mrs. Jennifer Havlicek,
Placement Advisor, and Ms. Juliana Vanegas, Program Advisor, for her continuous support.
Thanks to the staff of the National Autonomous Service of Aqueducts and Sewage (SANAA)
of Honduras, especially to the authorities for providing the support through out the entire
scholarship. Gratitude to Dr. Nancy Love for providing input on this research. Special thanks
to Julie Petruska and Jody Smiley for their continuous help and guidance through out the
entire experimental analysis phase, Jeff Parks for the analyses on the Inductively Coupled
Plasma instrumentation, Paolo Scardina for performing Zeta Potential analysis. Special
recognition to José Cerrato for his friendship, technical assistance, and exceptional
contribution on the statistical analysis performed in this study and Nestor Murray for his help
and support. Special recognition and gratitude to all of my lab mates and friends Sathya
Easwaran, Chris Wilson, Chris Muller, and Chul Parks for their technical support, help,
support, and contribution to all the lab skills learned through the entire research process.
The author wants to dedicate this work to the Cubas family for their perpetual support and for
being the source of inspiration always; Marcela Girón for providing the strength, knowledge,
and wisdom to his life; and the impoverished people of Honduras for providing the motivation
to pursue and contribute to a better global water quality and environment. And above
everything mentioned the author wants to thank and dedicate his entire work to God. Funding
for my scholarship was provided by the Organization of American States and the help of
Virginia Polytechnic Institute of Technology through its fellowship program OAS-LASPAU.
Any opinions, findings, conclusions or recommendations expressed in this material are those
v
of the authors and do not necessarily reflect the views of the Organization of American States
and LASPAU.
vi
Table of Contents
CHAPTER I................................................................................................................................1
Literature Review .......................................................................................................................1
1.1 Background information on wastewater treatment.....................................................1
1.2 Solid-Liquid separation ..............................................................................................3
1.3 Bioflocculation ...........................................................................................................4
1.4 Role of Filamentous Bacteria and SRT ......................................................................7
1.5 Cations and there effect in settling .............................................................................9
1.6 Sequencing Batch Reactors (SBR) and Feeding Pattern..........................................14
CHAPTER II ............................................................................................................................18
Combined Effect of Reactor Feeding Pattern and Cations on the Performance of an Activated
Sludge SBR ..............................................................................................................................18
Introduction ..........................................................................................................................19
Materials and Methods .........................................................................................................22
Results and Discussion.........................................................................................................26
Implications: .........................................................................................................................50
Engineering and Scientific Significance. .............................................................................52
Conclusions ..........................................................................................................................53
References ................................................................................................................................55
Appendix ..................................................................................................................................58
vii
List of Tables
Table 1-1 Minimum national standards for secondary treatment a............................................2
Table 2-1 SBR operating times ...............................................................................................22
Table 2-2 Cations used in the feed. .........................................................................................23
Table 2-3 Sodium concentrations for each of the three phases. ..............................................24
Table 2-4 Average pH for each reactor in all 3 phases ...........................................................27
Table 2-5 Average MLSS for each reactor with an SRT of 18 days.......................................28
Table 2-6. Summary of effluent quality and settling parameters for steady state conditions .30
Table A-1. Cations concentration in bactopeptone used on feed ............................................58
List of Figures
Figure 1-1. Capsules and slime layers produced by bacteria. ...................................................5
Figure 1-2. Potential role of cations in bioflocculation...........................................................13
Figure 1-3. Typical sequence of events in a SBR ...................................................................15
Figure 2-1. Activated sludge and settling properties as a function of time for phase II. Steady
state conditions between day 30 and day 55. ...........................................................................25
Figure 2-2. MLSS vs. time in the reactors with the highest sodium concentration ................29
Figure 2-3. Effluent TSS comparison......................................................................................32
Figure 2-4. Effluent COD (mg/L) for A-Phase I, B-Phase II, and C-Phase III.......................33
Figure 2-5. Total effluent COD for each phase. ......................................................................34
Figure 2-6. Effluent soluble COD for each phase ...................................................................35
viii
Figure 2-7. Effluent quality parameters comparison for a M:D ratio of 5.2 ...........................36
Figure 2-8. Settling properties as a function of feeding time and sodium concentration........37
Figure 2-9. CST comparison for three different feeding patterns ...........................................39
Figure 2-10. CST comparison for a M:D ratio of 5.2..............................................................40
Figure 2-11. Optimum dose for phase II and phase III ...........................................................40
Figure 2-12. Zeta-potential values for phase II and phase III .................................................41
Figure 2-13. Effluent polysaccharides (mg/L) for A-Phase I, B-Phase II, and C-Phase III....43
Figure 2-14. Effluent proteins (mg/L) for A-Phase I, B-Phase II, and C-Phase III ................44
Figure 2-15. Total effluent polysaccharides for each phase....................................................45
Figure 2-16. Soluble effluent polysaccharides for each phase ................................................45
Figure 2-17. Total effluent proteins for each phase ................................................................46
Figure 2-18. Soluble effluent proteins for each phase.............................................................47
Figure 2-19. Proteins and Polysaccharides comparison for a M:D ratio of 5.2 ......................48
Figure 2-20. Effect of feeding pattern on bioflocculation for 1.5 meq/L Na+.........................50
Figure 2-21. Effect of feeding pattern on bioflocculation for 6.0 meq/L Na+.........................51
Figure 2-22. Effect of feeding pattern on bioflocculation for 15 meq/L Na+..........................51
1
CHAPTER I
Literature Review
1.1 Background information on wastewater treatment
Wastewater treatment is a combination of unit operations and processes to treat polluted water
collected from municipalities and industries in order to be returned to receiving waters or to
the land for future reuse. Water treatment consists of the removal of pollutants that can harm
the aquatic environment. The ultimate goal of wastewater engineering is the protection of
public health while maintaining a balance with the environment considering social,
economical and political aspects.
Unit operations and processes, which are linked together in a process train, involve physical,
chemical, and biochemical reactions that achieve the destruction, stabilization, collection or
transformation of soluble or suspended pollutants some of which are classified as oxygen-
demanding or nutrients that contribute oxygen depletion or eutrophication of water bodies.
For this reason most of the unit operations used for the stabilization of organic matter are
biochemical which can be classified, within a wide range of options, as attached or suspended
growth, with activated sludge the most common process within the suspended growth
bioreactors.
Effluents discharged into a receiving body of water are regulated by the authorities whose
goals are to protect human health and the environment. As research has become more
extensive and as technology for analyzing specific constituents have become more
comprehensive, more stringent limits have been imposed, requiring higher water quality and
the use of the latest available technology in order to achieve the standards. Table 1-1 shows
an example of minimum national standards for secondary treatment. (Metcalf and Eddy,
2003)
2
Table 1-1 Minimum national standards for secondary treatment a
Characteristics of discharge
Unit of
measurements
Average 30-day
concentration
Average 7-day
concentration c
BOD5 mg/L 30 d 45
Total Suspended Solids mg/L 30 d 45
Hydrogen-ion concentration pH units 6.0 – 9.0 all times e 6.0 – 9.0 all times e
CBOD5 mg/L 25 40 a Federal Register (1988, 1989). c Not to be exceeded. d Average removal shall not be less than 85 percent. e Only enforced if caused by industrial wastewater or by in-plant inorganic chemical addition. f May be substituted for BOD5 at the option of the permitting authority.
Most municipal wastewater is generated from domestic sources but there has been increasing
amounts of industrial wastewater discharged to these municipal collection systems
introducing large amounts of heavy metals and synthetic organic compounds. Many of the
compounds generated from industrial processes are difficult and costly to treat; therefore,
effective pretreatment becomes an essential part of water quality management (Metcalf and
Eddy, 2003).
Differences exist between the characteristics of industrial and municipal wastewaters. Most
of these waters differ in the amount of cations (monovalent, divalent, trivalent), chemical
oxygen demand (COD) or biological oxygen demand (BOD) in the influent. It has also been
suggested that there is no substantial concentrations of proteins and polysaccharides in the
influent stream of many industrial plants while their presence has been observed in the
effluent (Murthy and Novak, 2001). Differences are also found in the effluent quality.
Researchers have found that the effluent quality at municipal plants is generally better than
most of industrial plants containing similar concentrations of monovalent and divalent
cations, and also it has been suggested that activated sludge from industrial facilities often
exhibit poor flocculation compared to domestic wastewater systems and this is attributed to
the absence of extra cellular polymeric substances (Park et al., 2006).
3
One characteristics of some industrial wastewaters is the high concentrations of sodium, with
concentrations up to 2,000 mg/L (Murthy et al., 1998). The high sodium concentration is the
result of adding many sodium based compounds such as sodium hydroxide (NaOH) or
sodium bicarbonate (NaHCO3) for manufacturing processes, prevention of volatilizations of
acetic acid or for pH control which requires some base addition. The effect of cations in
wastewater treatment will be discussed later.
1.2 Solid-Liquid separation
The solid-liquid separation is a physical process which involves the removal of solid particles
that can be in a suspended or colloidal state after every stage of treatment. The removal of
this suspended and colloidal material primarily composed of biological particles from the
biochemical operations is mostly done by gravity sedimentation. The efficiency of the
activated sludge treatment process is correlated to a good solid-liquid separation, which is
strongly determined by the settling properties (Govoreanu et al., 2003). For a successful
separation the microorganisms must clump together to form flocs of a defined size, porosity,
density and strength to allow them to settle and compact well without leaving a high
concentration of suspended solids in the effluent. Since single bacterial are relatively small (≈
0.5 – 1.0μ) (Madigan et al., 2003), it would be practically impossible to remove them in this
way if they grew individually. Fortunately, for practical and operational aspects, bacteria in
suspended cultures under the appropriate growth conditions grow and attach to each other
forming clumps or gelatinous particles called bioflocs. The bacteria responsible for this
phenomenon are called floc-forming bacteria and a variety of species fall into this category.
An ideal biofloc is one that is strong and compacts well so that it settles properly producing a
dense sludge for recycle to the bioreactor and a high quality effluent (Grady et al., 1999).
Unfortunately, not all bacteria present in suspended growth environments are beneficial when
related to floc formation. One type of organism is filamentous bacteria that grow in long
strands that become intermeshed with the floc particles and affect sedimentation, although
4
some number of filaments is necessary to give strength to the floc. These filaments can act as
a backbone that holds the floc together.
1.3 Bioflocculation
Bioflocculation is one of the most important steps in building a big, strong, dense, and
compact settleable floc in order for the wastewater treatment process to achieve a good solid-
liquid separation and therefore reach the desired effluent water quality. For a better
understanding on microbial flocculation, the structure and components of the floc must be
defined in order to propose means to change their properties and improve the settling and
dewatering properties of the activated sludge.
Activated sludge floc consist primarily of biopolymer produced by the cells, cations
(monovalent, divalent, and trivalent), microorganisms (floc forming bacteria and filamentous
bacteria), and other substances or particles such as polymeric substances that are part of the
influent wastewater or any suspended solid or debris trapped within the floc that would affect
its properties. Bioflocs are highly hydrated and very heterogeneous (Park et al., 2006) and
flocs with very different properties and morphologies may occur, depending on the conditions
in the activated sludge treatment plant and wastewater composition (Wilén et al., 2003).
Bioflocs are held together by means of exocellular polymers and divalent cations that interact
each other by electro potential forces (Bruus et al., 1992). Although most of the biopolymer
is incorporated within the activated sludge floc matrix a portion of the biopolymer remains
unattached in solution as biocolloids (Murthy and Novak, 1999).
Researchers have used the concept of microstructure and macrostructure to describe the
formation of bioflocs. The microstructure of the floc is considered to be the matrix composed
of microbes and exocellular polymers substances (EPS) bound together (bioflocculation
process), while the macrostructure consists of the microorganism such as filamentous bacteria
that provide the backbone for developing a larger and stronger floc. The literature suggests
that an improvement in the microstructure of the floc may overcome the negative impacts of
having too many filamentous organisms in the macrostructure that causes bulking. This area
5
needs further research in order to establish a well defined relationship among the EPS, cations
and filamentous microorganisms (Higgins et al., 2004b). Other authors suggest that two
different polymers exist; one that is very firmly bound to cells and within micro colonies of
cells and one that is more loosely bound in the floc matrix (Wilén et al., 2003).
Complex microbial aggregates, as activated sludge flocs, are poorly described and understood
in terms of the individual components and the flocculation mechanisms; however, there is a
point of agreement in that exocellular polymer substances (EPS) are the central to aggregation
of individual bacteria into floc particles (Grady et al., 1999). Researchers suggest that EPS,
both in terms of quantity and quality, are very important for the floc properties of the
activated sludge.
Figure 1-1. Capsules and slime layers produced by bacteria.
The EPS are mainly composed of carbohydrates, proteins, nucleic acids, lipids and humic
substances. The production of EPS is believed to be dependent on the growing phase of the
bacteria and the growth environment. The biopolymers produced by the cells form a matrix
which encapsulates the microbes and aids in the aggregation of the microorganisms or are
excreted into the surrounding medium as slime. Figure 1-1 illustrates these capsules formed
by bacteria. The floc formed is very heterogeneous and is made up of microbial colonies that
are embedded in cloud of EPS where a number of intermolecular interactions contribute to the
binding of the flocs components. Researchers agree that these interactions are hydrophobic,
steric and are a consequence of cations interactions (Wilén et al., 2003) and have been
6
explained by either the Double Layer Theory (DVLO), Alginate Theory, and Divalent Cation
Bridging Theory (Sobeck and Higgins, 2002).
The sum of the total proteins, humic compounds, carbohydrates, and DNA is considered to
represent the total mass of EPS in a floc. These materials have been found to represent up to
80% of the mass of activated sludge (Sobeck and Higgins, 2002). Some studies have tried to
define the chemical composition of sludge and EPS suggesting that protein was for most
sludges the major EPS component (19–45%) followed by humic compounds (19–45%) and
carbohydrates (7–32%). Uronic acids make up 1–3% of the EPS or 8–26% of the
carbohydrates; and, in some cases, high concentrations of DNA (0-32%) have been detected.
To obtain these results, 10-30% of the sludge’s organic fraction was extracted as EPS (Wilén
et al., 2003).
These results are in accordance with many studies that show that several types of EPS’s are
involved in bioflocculation. Polysaccharides are one of the most important structural
components of this polymer. Other recent studies suggest that proteins also play an important
role in bioflocculation (Grady et al., 1999). Some researchers have suggested polysaccharides
play a major role in flocculation; however, others suggest that polysaccharides forming only a
small part of the floc matrix are not as important as proteins for the aggregation on bacteria.
Since proteins are the most abundant macromolecule in EPS, several authors have tried to
identify the different classes that are present in the exocellular environment of bacteria. These
include extracellular enzymes, proteinaceous S-layers, lectins, or polypeptide capsular
material. The exocellular protein extracted from activated sludge samples could be from a
combination of these sources, and of these possible proteins, lectins are one of the most likely
types to be involved in bioflocculation (Higgins and Novak, 1997b). Lectins are
nonenzymatic proteins that bind sugar residues and play a role in attachment and colonization
of bacteria in both animals and plants, and other microorganisms. The lectins produced by
bacteria are typically located on appendages such as the pili and fimbriae of bacteria
(Madigan et al., 2003). The data suggest lectins or bacterial fimbriae play a very important
role in flocculation in activated sludge systems.
7
Researchers have proposed that biopolymers originate from release of bacterial growth, decay
and lysis, and from the influent wastewater, creating a matrix in which microorganisms can be
aggregated (Park et al., 2006). The most important sources of ECP are metabolism and cell
lysis products produced by protozoa and bacteria (Grady et al., 1999). Biopolymers have a
number of functional groups such as hydroxyl and negatively charge carboxyl groups that
contribute to the binding of floc constituents by means of cationic bridges in a way that
biopolymers can bind through specific protein and polysaccharide interaction, hydrophobic
interactions, hydrogen bonding, and ionic interactions (Park et al., 2006). EPS are very
hydrophilic due to their negative charge causing a negative effect on bioflocculation, but it
has been shown that the interactions between EPS and divalent cations can bind the
biopolymers to microbial cells and to other biopolymers. Therefore, because the majority of
exocelluar biopolymers are negatively charged, cations then become an important structural
component as a binding agent within the biopolymeric network (Novak et al., 1998). These
biopolymers are thought to be the glue that holds bioflocs together.
The sole presence of ESP’s is not the only mechanism to ensure a strong floc. Since bacteria
have a negative charge, the environment that surrounds the microorganisms, the bioreactor
configuration, cations, ionic strength, solids retention time, and suspended growth play an
important role.
1.4 Role of Filamentous Bacteria and SRT
As noted earlier in this document the relative proportion of floc-forming bacteria and
filamentous bacteria compose the macrostructure of flocs. Some of the filamentous bacteria
are enmeshed inside the floc providing a solid structure. If the filamentous bacteria grow and
extend beyond the biofloc, the particles compact poorly, increasing the sludge volume index
(SVI) and negatively impacting the solid-liquid separation process. In suspended growth
reactors the SRT affects the macrostructure of the floc. Studies show that low SRTs between
0.25 and 2 days produced large dispersion of suspended growth biomass producing
inadequate flocculation, and relative high SRTs between 9 and 12 produce irregularly shaped
pin point floc (Grady et al., 1999).
8
Different types of filamentous bacteria that are found in activated sludge have high affinities
for different limiting nutrients or dissolved oxygen which makes them out-compete the floc-
forming bacteria. Therefore, to better understand the communities of microorganisms present
in activated sludge, some knowledge of the growth kinetics of filamentous and floc-forming
bacteria is needed. For example when a particular substrate is present, floc-forming bacteria
have higher specific growth rate coefficient (μ) and half saturation coefficient for substrate
(Ks) values than filamentous bacteria which means that the floc-forming bacteria will grow
faster when the substrate concentration is high, but when filamentous bacteria have a higher
affinity for the substrate it can grow faster when the substrate concentration is low (Grady et
al., 1999).
When comparing growth kinetics of floc forming bacteria and filamentous bacteria based on
the Monod equation different tendencies are observed. If a substrate gradient is allowed in a
way that the influent concentration is high, floc-forming bacteria can grow at the expense of
the filamentous bacteria. A configuration of this type can be obtained in many different ways;
for example, a reactor can be set to work like an ideal plug-flow reactor, or by adjusting the
feeding pattern to obtain the desired influent substrate concentration.
Analysis suggests that filamentous bacteria have less capacity to accumulate and store carbon
reserves and have a lower substrate uptake rate than floc formers. Floc-forming bacteria have
the ability to store substrate when exposed to a high substrate concentration environment,
balancing the growth rate at a more constant rate, which gives them an advantage under
dynamic conditions (Martins et al., 2003). The latter should not be considered as an absolute
rule and also it remains uncertain whether the proliferation of filamentous bacteria is due to a
higher substrate affinity or just to its morphology; therefore, further analysis is required.
Under low substrate concentrations there is also a substrate diffusion limitation inside the
floc.
9
Under this condition filamentous bacteria can easily protrude into the bulk liquid and gain
access to the substrate outside the floc, and if the filamentous bacteria grow beyond the floc
they can out-compete the floc forming bacteria. It is believed that these substrate diffusion
gradients inside the flocs are an important factor for the growth of filamentous microorganism
within the floc, whereas in an environment where substrate diffusion is not an issue the
filamentous bacteria remain mostly inside the floc (Martins et al., 2003).
1.5 Cations and there effect in settling
Before examining in detail the effect of cations (monovalent, divalent, and trivalent) in the
performance of an activated sludge process three proposed theories should be evaluated.
a. Double layer theory (DLVO). Named after there developers (Derjaguin, Landau,
Verwey, and Overbeek) is a classical colloidal theory (Sobeck and Higgins, 2002).
In a colloidal suspension there cannot be a net imbalance in the overall electrical
charge and therefore the primary charge of the particle must be counterbalanced in
a system. The primary charges accumulate in an interfacial region together with
the opposite charged ions forming an electrical double layer (AWWA, 1999). The
first layer is comprised of a tightly associated layer of counterions, and the second
layer (diffuse layer) results from the electrostatic attraction of ions of opposite
charge to the particle (less tightly associated counterions). An electrostatic
potential exists between the surface of the particle, where the potential is
maximum, and the solution surrounding the particle. This potential decreases with
distance from the surface until the concentration of ions equals the bulk solution.
When colloidal particles approach each other, their diffuse layers begin to interact
and their charges create a repulsive potential energy that inhibits aggregation. As
the ionic strength increases, the double layer decreases and the repulsion potential
decreases allowing other forces, like the van der Waals forces, to interact and
promote aggregation.
b. Alginate Theory. This theory was first proposed by Bruus (Bruus et al., 1992).
Alginate is a polysaccharide generally made up of glucuronic acids and repeating
10
mannuronic produced by bacteria that result in a formation of alginate gel when
calcium is present. This model implies that when high concentrations of sodium
are present in activated sludge and calcium is replaced by sodium within the floc,
this result in deterioration of the biofloc. Alginate aggregation is exclusively by
calcium so researchers stated that calcium induced aggregation of alginate is
important (more than magnesium) for the bioflocculation process (Sobeck and
Higgins, 2002).
c. Divalent cation bridging theory (DCBT). The first researchers to propose this
theory were McKinney and Tezuka (Sobeck and Higgins, 2002). According to
this theory divalent cations, such as calcium and magnesium, work as a bridge and
link negatively functional groups with the EPS and by this means create a more
stable biopolymer matrix enhancing bioflocculation. This model proposes that
non-specific binding occurs rather than a specific interaction of gel formation
between calcium and alginate. Some researchers agree and have contributed with
some studies that support this theory (Higgins and Novak, 1997). They
demonstrated that sodium addition caused a deterioration of flocs because of the
displacement of divalent cations from binding sites within the floc.
Cations significantly affect bioflocculation and alter the settling and dewatering
characteristics of activated sludge systems. Cations imbalances are a common cause of
sludge settling problems especially in industrial activated sludge plants where these
unbalances are most likely to occur (Higgins and Novak, 1997a). It is believed that cations
interact with the negatively charged biopolymers in activated sludge and change the structure
of the floc matrix through an exchange of biopolymers between flocs and the environment.
Altering the cations in the wastewater has been to be an economical way to enhance settling
and dewatering properties (Higgins et al., 2004a) and a decrease in microbial soluble product
would result in a lower effluent COD. This was demonstrated in many studies where
monitoring the cations and the changes in biofloc characteristics affected the effluent water
quality through an exchange of biopolymers between flocs and the environment (Murthy et
al., 1998). Most of these studies refer to a surrounding environment where filamentous
11
bacteria are not the main cause of the deterioration in settling and all of the settling properties
area attributed to biopolymer and cations interaction.
Historically, most of the research done is related to monovalent and divalent cations in
activated sludge. The divalent cations that are mostly considered are calcium (Ca2+) and
Magnesium (Mg2+). The divalent cations bridge across negatively charged sites on the
biopolymers to form a dense, large, and compact floc structure more resistance to shear and
therefore promotes bioflocculation and enhance settling and dewatering properties (Higgins
and Novak, 1997c, Murthy and Novak, 2001). Both of these divalent cations are required for
the adhesion of certain bacterial monocultures depending mainly on the sensitivity of each
type of bacteria in the absence of calcium and magnesium (Higgins and Novak, 1997c).
Researchers suggest that cations help bioflocculation by bridging the negative sites on
exocellular biopolymers (DLVO) while others agree in that cations induced flocculation by
the double compression theory or by alginate theory, the latter when dealing with calcium
only (Park et al., 2006, Sobeck and Higgins, 2002). Most of the data that exists in the
literature report that when activated sludge is fed with either Ca2+ or Mg2+ the settling
properties improve in a similar manner, which lead to the conclusion that cation bridging best
explains the role of divalent cations in the floc structure (Higgins and Novak, 1997c) (Park et
al., 2006).
One of the questions that arose from these previous studies was the amount of divalent cations
that must be present in order to have good settling and good dewatering properties. To this
question Higgins and Novak suggest a value between 0.7 – 2.0 meq/L of calcium and
magnesium necessary for acceptable settling and dewatering, suggesting that increasing the
feed concentrations of these cations above these levels improved the floc strength, settling,
dewatering properties, and increased the bound protein concentration (Higgins and Novak,
1997c). Divalent cation addition has also been proved to decrease the polymer demand used
for conditioning by 30-75% (Higgins and Novak, 1997a). Other data also suggest that adding
calcium and magnesium to an activated sludge system can help reduce the settling properties
and effluent problems related to bulking and filamentous bacteria, while not controlling the
12
filamentous organism concentration (Higgins et al., 2004b). Calcium and magnesium ions are
also related to lectin interactions, enhancing the activity of these proteins within the
extracellular matrix.
One concern that resulted from previous studies was the point of application of the divalent
cations in order to produce a better effect in settling and dewatering. In many laboratory
studies improvements in settling properties were observed when calcium and magnesium
were added to the feed rather than directly to the reactor (Murthy and Novak, 2001). This can
be explained as follows: when cations are applied in the feed, they are able to become
enmeshed in the polymer network while the bacteria are producing the biopolymer and
therefore can be incorporated to the floc as it is formed (Novak et al., 1998). In contrast, when
cations are applied in the reactor as a slug dose, the incorporation of cations occurs only in the
outer portions of the floc. Therefore, the incorporation of cations during floc formation must
be considered important (Murthy and Novak, 2001).
It has been demonstrated that high concentrations of sodium present in activated sludge
results in a deterioration of properties such as floc density, capillary suction time (CST),
sludge volume index (SVI), effluent TSS and effluent COD. Sodium in the influent
wastewater also causes an increase in proteins and polysaccharides in the effluent and
therefore increases the total suspended solids (TSS) and effluent COD concentration (Murthy
and Novak, 2001). This increase in the TSS is considered to be a sign of weak floc and it is
attributed to a decrease in the bonding strength of exopolymer functional groups. Like the
case of divalent cations, a concentration value that represented the amount of sodium that will
create a considerable deterioration in the settling properties was needed. Research suggests
that at values lower than 10 meq/L sodium did not greatly impact the settling properties of
activated sludge, but at concentrations higher than 10 meq/L there may be problems related to
poor settling (Higgins and Novak, 1997c). Other authors suggest that problems related to
settling and dewatering will appear on activated sludge systems when the concentration of
monovalent to divalent cations exceeded 1:1 and will get worst when exceeding 2:1,
expressed on an equivalent basis (Higgins and Novak, 1997a). Therefore, the poor settling
and dewatering can be improved by raising the concentration of divalent cations (adding
13
calcium and magnesium). Some authors had also observed in some cases that addition of
sodium can improve settling. This might be due to the fact that flocs become more compact
and therefore denser, or there might be an increase in permeability that can reduce drag while
settling, by allowing water to flow better through the floc structure. This phenomenon can be
associated to the physiological reaction of bacteria to sodium (Novak et al., 1998). Therefore
a certain amount of sodium and other monovalent cations could be beneficial for
bioflocculation. Figure 1-2 depicts the relationship between EPS and cations. Other
monovalent ions considered to cause problems related to settling in activated sludge systems
are ammonium and potassium. It has been suggested that these cations, act like sodium,
replacing divalent cations and creating problems with settling (Novak, 2001).
Figure 1-2. Potential role of cations in bioflocculation
Recently, and since most of the studies has been related to monovalent and divalent cations,
specialist have studied the effect of trivalent cations in settling and dewatering properties.
Some of the trivalent cations that have been studied include aluminum (Al3+) and iron (Fe3+)
because they are abundant in activated sludge and because it is assumed that due to their
higher valence, their contribution to floc stability would be better (Park et al., 2006). Recent
studies have demonstrated that as the concentration of iron and aluminum increased, the
biopolymer found in solution decreased, suggesting that Al and Fe are good absorbers of
negatively charged organic particles. Related studies show that poor biopolymer binding was
observed when there was an absence of Al and Fe (Park et al., 2006). These studies suggest
Repulsive Force
14
that Al and Fe have considerable impacts on activated sludge characteristics regardless of the
relationship between monovalent and divalent cations.
As the M:D ratio increases, the effluent TSS and COD increases as a result of microorganism
and biopolymers that are released from the floc into solution. This COD concentration is
believed to be proportional to the biopolymer in solution (Murthy and Novak, 2001).
Therefore, the retention of biopolymers within the floc matrix will result in an improvement
in effluent quality measured as oxygen demand (Higgins et al., 2004a). The absence or
removal of beneficial cations from wastewater can result in a release of the primary
biopolymer constituents such as proteins, polysaccharides, DNA, RNA, and lipids into
solution (Murthy and Novak, 2001). Some authors observed that when divalent cations
increase, the bound protein concentration also increases and that there is little effect on the
polysaccharide bound concentration. These data suggest that divalent cations bind together
exocellular protein with flocs explaining the relationship biopolymers and cations and
highlighting that that proteins and not the polysaccharides are dominant in bioflocculation
(Higgins and Novak, 1997b). Following this statement it can be inferred that calcium and
magnesium bind lectins, and lectins bind polysaccharides within the floc matrix depending on
the type of microorganism present and the lectin produced. And with this a extension of the
previous model can be added, stating that lectinlike proteins would bind polysaccharides that
are cross-linked to adjacent proteins (Higgins and Novak, 1997b). Meanwhile other authors
believe that the main biopolymer constituent released into solution is polysaccharides.
1.6 Sequencing Batch Reactors (SBR) and Feeding Pattern
Sequencing batch reactors (SBR) are those reactors that are operated in a sequence of steps.
Figure 1-3 shows the typical sequencing of a SBR. One of the most important steps in a SBR
is the fill period. One of the aspects that are affected by the filling period, which is governed
by the length of time that is required to reach a specified volume, is the process loading factor
which is part of the hydraulic characteristic of the bioreactor. A high and instantaneous
process loading factor will occur when the filling period is short. In this short time the
biomass in the reactor will receive a high initial amount of organic matter and nutrients and
15
the concentration of these constituents will reduce over time after the filling period is over. In
this case the SBR will be analogous to a continuous feed system with a configuration of tanks
in series or a plug flow reactor. Conversely, if the filling period is long, then the process
loading factor will be small and the SBR will be analogous to a completely mixed reactor
(Grady et al., 1999). This analogy is valid, providing an equal SRT for both types of reactors.
An activated sludge system with a selector placed at the head of the system (for bulking
control) followed by the aeration basin can also be simulated by a SBR. It is important to take
in account that most of the reactions such as substrate utilization and biomass growth will take
place during the filling period.
Figure 1-3. Typical sequence of events in a SBR
According to Grady et al., (1999), the process loading factor is the mass of the substrate
applied per unit time divided by the mass of microorganisms in the bioreactor. As mentioned
earlier in this document, the process loading factor will influence the competition between the
microorganisms and will affect the settling properties.
SBR are chosen in many laboratory models because they can be operated in many different
ways that replicate many situations and processes if the SRT and HRT (hydraulic retention
time) are the same. Also the variation of the length of the filling period will allow the SBR to
operate in a range between a plug flow reactor and a completely mixed reactor. Historically it
has been observed that a SBR generally produced better settling biomass than other
completely mixed systems, especially when the filling ratio was minimal and this resulted to
be a turning point in the study of wastewater treatment using SBRs. Many studies have been
Influent
Fill React Settle Decant
Effluent
16
made in SBRs to find out how the communities of microorganism, that conform the activated
sludge, respond to the operation and conditions of the SBR. One of the main objectives of
these studies is to determine a link between changes in activated sludge settling properties,
floc structure, microbial community dynamics and the operation of a SBR. This combination
should lead to a more detailed understanding of the treatment process performance.
In a recent study (Govoreanu et al., 2003), a series of SBRs were operated for approximately
200 days under stable condition in which they were able to observe three different stages
characterized by changes in floc structure and microbial community. In the first stage, a
predominant presence of floc-forming bacteria was observed which controlled the floc
structure. The second stage was a short period where filamentous bacteria appeared in almost
the same amounts as floc-forming bacteria, creating a good floc that improved settling
properties. A highly dynamic microbial community started to emerge at this point. Finally,
during the last stage, a high amount of filamentous bacteria was observed and therefore
bulking was an issue (Govoreanu et al., 2003). This study suggested a dynamic microbial
environment in this type of reactors which had a great influence in settling properties and
therefore in effluent quality.
As mentioned earlier in this chapter, the competition and selections of microorganism inside a
reactor will be associated to properties such as growth rate, substrate intake rate, and substrate
affinity. Experiments suggest that the filling ratio in an SBR may have an impact on the
settling properties of the sludge. Increasing the length time of the feed which creates a low
substrate concentration can negatively affect the settling properties of the activated sludge.
However, when the fill time is short, a high substrate gradient is present which promotes the
substrate intake close to the maximum specific rate of bacteria, leading to good settleability.
Under these conditions, floc-forming bacteria prevail and filamentous organisms are less
abundant or are incorporated within the floc. This is consistent with the idea that the feeding
pattern had an influence on the microbial population dynamics and kinetics of activated
sludge (Martins et al., 2003).
17
When the feeding pattern was changed from a long fill time to a pulse feed (short fill time), a
decrease in the abundance of filamentous bacteria was observed. The few filaments present
were mostly around the floc and suddenly incorporated within the floc instead of growing far
beyond the floc structure (Martins et al., 2003). Also, a gradual change in the morphology of
the flocs was observed. Large flocs (with filaments incorporated in the floc) were observed
during the pulse feed; these flocs were strong, round and compact. When the feeding pattern
was changed to a large filling time, the flocs were more irregular and porous suggesting that
morphology changes in flocs also affected the settling properties of sludge.
It is obvious that small difference in the aerobic fill time will have a dramatic effect on the
sludge setting properties, considering that other factors such as aeration and nutrients
provided will not affect the community of microorganism present in the activated sludge
system.
18
CHAPTER II
Combined Effect of Reactor Feeding Pattern and Cations on the Performance of an Activated Sludge SBR
ABSTRACT: Laboratory scale sequencing batch reactors (SBR) were used to study the effect
of feeding pattern on activated sludge containing different concentrations of monovalent
cations. Effluent wastewater quality and settling and dewatering properties were analyzed to
determine if extending the feeding time to an SBR would have an impact on the effects of
adding high amounts of sodium (Na+) to an activated sludge system, simulating wastewater
treatment plants containing high quantities of monovalent cations in their influent. Data
suggest that excess amounts of sodium caused poor flocculation in general, but when the
feeding time was raised from 1 min (pulse feed) to 1hr and 4hr, an increase in deflocculation
was observed. Deflocculation caused by the combined effect of sodium and feeding pattern
was less when the sodium concentration was low. As sodium was increased to 6 meq/L, the
negative effects on bioflocculation caused by the feeding strategy were offset. As the amount
of sodium was further increased, a dramatic effect on settling and effluent water quality was
observed reflected by an increment in effluent total suspended solids (TSS) and effluent
chemical oxygen demand (COD) accompanied by an increase in sludge volume index (SVI)
and capillary suction time (CST). An increase in effluent biopolymers (proteins and
polysaccharides) was also observed. A better understanding of the effects produced by
changing the feeding strategy on bioflocculation will help wastewater plant operators meet
effluent requirements when struggling against high amounts of monovalent cations in their
influent. Also, when performing analysis related to settling and dewatering in lab scale
activated sludge reactors, the feeding pattern should always be reported in order to
standardize procedures and compare results where the effects of changing the feeding
strategies have been considered.
KEYWORDS: Activated sludge, cations, feeding pattern, sodium, bioflocculation, settling,
biopolymers, sequencing batch reactor.
19
Introduction The activated sludge process is widely used to treat both industrial and municipal wastewater.
While most municipal treatment plants perform well, industrial facilities often struggle to
meet effluent requirements. One reason for this is that the biomass from industrial facilities
settles poorly. One reason for this poor performance has been attributed to a high monovalent
to divalent ion ratio in many industrial wastewaters (Higgins and Novak, 1997).
The quality of the effluent from activated sludge treatment plants is highly dependent on the
efficiency of the solid-liquid separation process. If this process is poor, wastewater treatment
will be ineffective, resulting in failure to achieve regulatory effluent requirements. Solid-
liquid separation is a physical process which involves the removal of solids particles that can
be in a suspended, colloidal or soluble state. This is usually accomplished by gravity
sedimentation. For a successful separation the microorganisms must clump together to form
flocs of a defined size, porosity, density and strength to allow them to settle and compact well
without leaving a high concentration of suspended solids in the effluent. Bacteria in
suspended cultures, under the appropriate growth conditions, are able to grow and attach each
other to form flocs. Flocs with very different properties and morphologies may occur,
depending on the conditions in the activated sludge treatment plant and wastewater
composition (Wilén et al., 2003).
Bioflocculation is one of the most important steps in building large, strong, dense, and
compact settleable floc. For a better understanding on microbial flocculation, the structure
and components of the floc must be defined in order to propose means to change their
properties and improve the settling and dewatering properties of the activated sludge.
Activated sludge floc consists primarily of biopolymer, cations, microorganisms (floc forming
bacteria and filamentous bacteria), and debris trapped within the floc. These biopolymers
referred to as exocellular polymeric substances (EPS), are produced when the active biomass
converts complex organic matter into low molecular weight compounds (Sponza, 2004),
forming a matrix that encapsulates the microbes and aids in the aggregation of the
microorganisms.
20
The production of EPS, which is mainly composed of carbohydrates, proteins, nucleic acids,
lipids and humic substances, is believed to be dependent of the growing phase, decay and
lysis of bacteria. Researchers suggest that EPS, both in terms of quantity and quality, are very
important for the floc properties of the activated sludge (Liao et al., 2000). A combination of
floc-forming bacteria and filamentous bacteria compose the macrostructure of flocs, with
some of the filamentous bacteria enmeshed inside the floc to provide a solid structure. The
relative abundance of filamentous organisms will depend on several factors such as solids
retention time and substrate concentration. Several studies also suggest that feeding patterns
also have a strong influence on the microbial community of activated sludge (Martins et al.,
2003).
Cations significantly affect bioflocculation and alter the settling and dewatering
characteristics of activated sludge systems. Cations imbalances are a common cause of
sludge settling problems especially in activated sludge plants related to industrial activities
(Higgins and Novak, 1997a). High concentrations of monovalent cations such as sodium
(Na+) are detrimental to activated sludge properties (Park et al., 2006). It has been
demonstrated that high concentrations of sodium present in activated sludge result in a
deterioration of sludge properties such as, capillary suction time (CST), sludge volume index
(SVI), effluent TSS and effluent COD. Sodium in the influent wastewater also causes an
increase in proteins and polysaccharides in the effluent (Murthy and Novak, 2001).
Researchers suggested that at values lower than 10 meq/L sodium did not impact greatly the
settling properties of the activated sludge, but at concentrations higher than 10 meq/L, poor
settling will occur (Higgins and Novak, 1997c).
It has been proposed that the divalent cations act as a bridge between the negatively charged
particles within the biopolymer network but in the presence of monovalent cations they are
displaced by an ion exchange process that reduces the ability of biopolymer to bind and form
a good floc matrix (Higgins and Novak, 1997c). Higgins and Novak (1997b) suggested that a
relationship between the sum of monovalent and divalent cations (M:D) could be a good
indicator to evaluate the problems related to settling and dewatering; they suggested that with
21
a ratio greater than 2 there would be considerable problems in effluent quality associated to
settling (Higgins and Novak, 1997a).
The above-mentioned studies describe many of the mechanisms involved in bioflocculation,
which consider several aspects that will influence the floc formation process, but these are not
the only factors that will affect effluent water quality, settling and dewatering properties of
activated sludge. While performing lab-scale studies related to cations in sequencing batch
reactors (SBR), Higgins and Novak (personal communication) observed that the way in which
the SBR was fed influenced the results obtained for analysis related to effluent quality and
stated that differences in feeding pattern will probably have a dramatic effect on the sludge
setting properties in SBRs. They observed that by increasing the length of time of the feed,
which creates a low substrate concentration in the reactor, affected negatively the settling
properties of the activated sludge. However, when the fill time was short, a high substrate
gradient was present which results in the substrate intake to be close to the maximum specific
rate of bacteria and appears to lead to good sludge settleability. These studies supported the
idea that the feeding pattern had an influence on the microbial population dynamics and
kinetics of activated sludge (Martins et al., 2003). The competition and selection of
microorganism inside a reactor will be associated with properties such as growth rate,
substrate intake rate, and substrate affinity, which will then affect the quality and quantity of
EPS produced.
The objective of this study was to determine if the effluent quality, settling and dewatering
properties of lab-scale SBR reactors at different sodium concentrations were to experience
any change when subject to different feeding patterns, and to determine which feeding
configuration yields the best results regarding effluent quality. If the deterioration was
eminent, then considerations in controlling influent flows should be taken when operating
industrial or municipal plants where the M:D ratio might be high due to the usage of sodium
based chemicals for pH control, and to encourage other researchers to include the impact of
feeding pattern when performing studies related to solid liquid separation in activates sludge
systems.
22
Materials and Methods Experimental Approach. Experiments were conducted using four sequencing batch reactors
(SBR), each one containing a 1 L volume. Since the purpose of this study was to analyze the
effect of the feeding pattern in the effluent quality and sludge settling properties when
exposed to different sodium concentrations, different feed configurations were required. A
phase, as denoted, represents the period of time when four reactors containing different
amounts of sodium are operated under the same flow conditions. Each phase has a unique
feeding pattern which is maintained constant through the entire operational period. Each
phase is characterized by different filling time duration, as shown in table 2-1. A pulse feed
was used for phase one (feeding time: one minute), for phase two, one hour, and for phase
three, a four hour feed time was used.
The reactors were seeded with mixed liquor obtained from the Blacksburg, VA., municipal
wastewater treatment plant. The SBR were operated with two cycles per day, each cycle
lasting for 12 hours. The reactors were operated using a 1 day hydraulic retention time (HRT)
and an 18 day solids retention time (SRT). This SRT was chosen to maintain a mixed liquor
suspended solid concentration of around 2,000 mg/L to allow nitrification and therefore
reduce the effects of the ammonium (NH4+) ion, and to reduce the effect of SRT in floc
formation.
Table 2-1 SBR operating times
Fill time Reaction time Settling time Decant time
Phase I 1 min 11 hrs 45 min 15 min
Phase II 60 min 11 hrs* 45 min 15 min
Phase III 240 min 11 hrs* 45 min 15 min
* Aeration was applied for 40 min of the filling time for phase II.
** Aeration was applied for 220 min of the filling time for phase III.
Oxygen was provided using compressed air fed through diffuser stones to allow a dissolved
oxygen concentration greater than 2.0 mg/L and to provide enough mixing to keep the
biomass in suspension without disturbing the process of floc formation. The oxygen input
was regulated in a way that it did not favor the growth of filamentous organisms. The pH was
23
monitored but was not controlled (one of the added salts contained HCO3 which added some
buffer capacity). The temperature for the entire study was maintained at 20º C.
Feed and Cations. Bactopeptone, a microbiological enzymatic digest of protein for use in
culture media, was used as feed (electron donor, carbon, and nutrient source) for the reactors
at a concentration of 300 mg/l as COD. Several COD (chemical oxygen demand) analysis
were done to determine the required bactopeptone concentration that would give the desired
value. It was determined from this analysis that 273 mg/L of bactopeptone provided a COD
of 300 mg/L. The low concentration of cations in the bactopeptone allowed control of cations
to the feed as described in table 2-2.
Table 2-2 shows the cations that were added to the feed. The concentrations of all the cations
were maintained constant during the three phases of the experiment. The inorganic salt
concentrations in the feed (excluding sodium) were chosen in a way that they had a minimal
impact on the settling and effluent properties of activated sludge following the recommended
values found in the literature (Higgins and Novak, 1997c, Murthy and Novak, 1998, Park et
al., 2006) .
Table 2-2 Cations used in the feed.
Cation Concentration Compound used Concentration
(mg/L)
Calcium (Ca2+) 1.5 meq/L CaCl2 · 2H2O 110
Magnesium (Mg2+) 1.5 meq/L MgSO4 · 7H2O 185
Potassium (K+) 0.5 meq/L KHSO4 68
Iron (Fe3+) 6 mg/L FeCl3 17.5
Aluminum (AL3+) 3 mg/L Al2(SO4)3 · 18H2O 37
The sodium concentration for each reactor is shown in table 2-3. These concentrations were
maintained constant during each of the three phases. In the third phase of the study the
sodium concentration in reactor 2 was increased from 3 to 10 meq/L. This change was done
24
because little difference was observed between the results obtained from the 1.5 and 3.0
meq/L sodium reactors for the first two phases; so, in order to have a better understanding of
the impact of high sodium concentrations, a higher value was selected.
Table 2-3 Sodium concentrations for each of the three phases.
Reactor ID NaHCO3
Conc. (meq/L)
NaHCO3
Conc. (mg/L) M:D ratio
Reactor 1 1.5 126 0.7
Reactor 2* 3.0 252 1.2
Reactor 3 6.0 504 2.2
Reactor 4 15.0 1260 5.2
* The concentration of this reactor was increased to10 meq/L in the third phase of the experiment.
The monovalent to divalent cation ratio for each of the different sodium concentrations was
calculated in order to compare it with values selected in the literature that would provide a
range of settling and dewatering properties.
Steady state determination. The reactors in each phase were operated until it was
determined that a steady state has been achieved. The system was considered to be at steady
state when visual inspections of settling properties plotted as a function of time were stable
and variability no greater than ≈ 25 % occurred. An example of steady state determination for
phase II in two of the reactors is shown in figure 2-1. For this example, the steady state
period was considered to be between about 30 and 55 days. Steady state for the reactors was
usually achieved at 2 SRT’s. The samples for analysis were taken once the steady state was
reached. The samples used for effluent quality evaluation were taken at the end of each SBR
cycle just before the decanting process, whereas the samples for MLSS analysis were taken at
the end of a cycle before the settling period started.
25
0
500
1000
1500
2000
2500
3000
0 10 20 30 40 50 60 70 80 90
Time (d)
MLS
S (m
g/L)
-
50
100
150
200
250
300
350
400
Na 6 meq/L MLSS Na 15 meq/L MLSSNa 6 meq/L SVI Na 15 meq/L SVI
Steady State
Figure 2-1. Activated sludge and settling properties as a function of time for phase II. Steady state conditions between day 30 and day 55.
Settling and dewatering properties. Total suspended solids (TSS), mixed liquor suspended
solids (MLSS), and volatile suspended solids (VSS) were analyzed using Method 2540D and
2540E, respectively, from Standard Methods (1995). The settling properties of the activated
sludge were characterized by the sludge volume index (SVI) and the dewatering properties
were determined using the capillary suction time (CST) analysis based on method 2710D and
2710G, respectively according to Standard Methods (1995). Soluble chemical oxygen
demand (COD) was measured based on the method 5220C of Standard Methods (1995). For
conditioning and dewatering of the activated sludge, a dry cationic polymer (clarifloc 3275)
was used. The polymer solution was prepared at 0.01% by weight. The optimum dose of
polymer was determined by plotting the values of the CST for each corresponding dose and
obtaining the minimum CST reading. This analysis was done under low shear conditions.
Soluble protein was measured using the Hartree modification of the Lowry method (Hartree,
1972). Polysaccharides were measured using the Dubois method (Dubois et al., 1956).
Bovine serum albumin and glucose were used as protein and polysaccharide standards,
respectively. For soluble analysis of COD, proteins, and polysaccharides samples were
26
filtered through 1.5 μm, 0.45 μm, and 1,000 Dalton (1-k) filter membranes. The 1-k ultra
filtration was performed at 55 psi.
Filamentous organism observation and particle charges. Microscopic observations of the
flocs were performed periodically on a regular basis to quantify filamentous organism in the
activated sludge. The filamentous organism content was quantified using the method of
Jenkins (Jenkins et al., 1986) where the presence of filamentous is rated on a scale of 0-6; 0
corresponds to no filamentous organisms present and 6 to excessive presence of filamentous.
Denaturing gradient gel electrophoresis (DGGE) analysis was used to examine temporal
differences within the activated sludge bacterial community. This method generates a
community fingerprint pattern, in which a band generally corresponds to one bacterial
ribotype (population), which allows the detection of changes in the presence of the dominant
bacterial population in the community. This procedure is based on the electrophoresis of
polymerase chain reaction (PCR) amplified 16S rDNA fragments, obtained form total sludge
DNA in a polyacrylamide gel (Muyzer et al., 1993). The particles charges within the
biopolymer were measured by a zeta potential analysis.
Statistical analysis. A series of pooled t-tests and analysis of variance (AOV) for more than
two populations means were used for statistical comparison and to determine differences in
the results obtained for the settling and dewatering properties of the activated sludge for each
phase and sodium concentrations. A type I error value of 0.05 was used (α = 0.05).
Results and Discussion Laboratory reactors were operated for approximately 90 days for each of the three phases of
the experiment. Each phase, as denoted, consisted of four reactors that were fed with
bactopeptone and a series of cations where sodium was applied at different concentrations
raging from 1.5 meq/L to 15 meq/L. For each phase the feeding pattern was different. A
pulse feed was used for phase one (feeding time: one minute), for phase two a one hour feed
time was used and for phase three, a four hour feed time was used. The shortest feed period
resembles the substrate gradient similar to the one occurring in a plug flow reactor, while the
longer feeding system can be compared to a completely-mixed flow reactor with low soluble
27
substrate concentration. The effect of the sodium concentration and the impact it has when
applied at different flows is analyzed in this study. Settling properties and effluent quality
parameter were measured and analyzed.
Table 2-4 Average pH and Na+ concentrations for each reactor
Reactor #
Na Conc.
(meq/L) pH Feed pH
Reactor 1 1.5 7.70 7.70
Reactor 2a 3.0 7.94 7.94
Reactor 3 6.0 8.47 8.20
Reactor 4 15.0 9.00 8.40 a The concentration for reactor 2 was increased from 3.0 to 10.0 meq/L in the third phase of the
experiment. The average pH for the reactors is shown in Table 2-4. It can be seen that the pH was
generally between 7.7 and 9.0. The pH generally increased as the sodium bicarbonate
concentration increased. The solids retention time (SRT) was maintained constant at 18 days
although the amount of wastage in those reactors where the deflocculation was high made it
hard to maintain a constant SRT for the high sodium system. The amount of biomass
maintained in the reactors in each phase is shown in Table 2-5.
A comparison of the mixed liquor suspended solids (MLSS) concentration was made between
the reactors fed with the same amount of sodium for each phase to eliminate any impact on
the results introduced by a difference in the amount of biomass present. A statistical analysis
was used to determine if there was a difference between these values. The MLSS
concentrations for the sodium feeds of 1.5, 6.0, and 15 meq/L were compared. The
corresponding p-values obtained for the analysis of variance (ANOVA) were found to be
0.418, 0.124, and 0.086, respectively. This shows that there is no significant difference in the
MLSS concentrations for the reactors containing the above stated amounts of sodium.
28
Table 2-5 Average MLSS for each reactor with an SRT of 18 days
Reactor
Na Conc
(meq/L) MLSS (mg/L)
1 - Phase I 1.5 1621 ± 266
2 - Phase I 3.0 1493 ± 301
3 -Phase I 6.0 1839 ± 254
4 - Phase I 15.0 2126 ± 398
1 - Phase II 1.5 1959 ± 297
2 - Phase II 3.0 1991 ± 344
3 - Phase II 6.0 2022 ± 481
4 - Phase II 15.0 1917 ± 255
1 - Phase III 1.5 1715 ± 643
2 - Phase III 10.0 2197 ± 285
3 - Phase III 6.0 1809 ± 291
4 - Phase III 15.0 1705 ± 375
At the end of each phase, the amount of suspended solids removed from the reactors through
the decant process plus the wastage was so high that it lead to a decline in the MLSS. This
was common after the 90 days of operation for the reactors with a low sodium concentration;
but, in the reactors containing higher amounts of sodium this period was even shorter, about
60 days. At this point the reactor data were not considered useful for analysis. Figure 2-2
shows how the MLSS concentration in the reactors containing the highest sodium
concentration changed over time for each phase; it can be seen from the plot that at the end of
each phase the amount of biomass began to decline.
The influence of filamentous organisms on the floc structure and properties was minimized by
several strategies, but not eliminated. The seed wastewater contained few filaments and there
was no evidence from the Blacksburg wastewater treatment plant personnel that they had
problems with bulking. The dissolved oxygen was not limiting (greater than 2.0 mg/L O2),
but also it was not in excess to allow nuisance organism to proliferate.
29
0
500
1000
1500
2000
2500
3000
0 10 20 30 40 50 60 70 80 90
Time (d)
MLS
S (m
g/L)
-
50
100
150
200
250
300
350
400
SVI (
mL/
g M
LSS)
Na 6 meq/L MLSS Na 15 meq/L MLSSNa 6 meq/L SVI Na 15 meq/L SVI
Steady State
Figure 2-2. MLSS vs. time in the reactors with the highest sodium concentration
A change in microbial community during each phase was expected because of the differences
in the F/M ratio introduced by the change in feeding pattern. Therefore, actions to reduce the
presence of nuisance microorganisms were taken only when the presence of these organisms
was too high. The presence of red worm was observed, especially in the phase were the
feeding time was greatest, but it was controlled by adding a few drops of bleach when their
presence was excessive.
Effect on filamentous organisms. Quantification of filamentous organisms was done in the
second and third phase when some episodes of filaments and nuisance organisms occurred.
This quantification was done by direct observation through a microscope and based on a
subjective scoring of filaments abundance described by Jenkins, et al. (1986). During the
second phase of this study the presence of filaments was constant between 1 and 2 while in
the third phase the presence of filamentous organisms ranged from 1 to 3. In this last phase,
the amount of filaments was higher in the reactors where the sodium concentration was lower,
usually between 2 and 3, and lower where sodium was highest. This suggests that the
filaments did not impact the settleability and effluent quality parameters of the activated
sludge when the sodium concentration was higher, but might have a small impact when the
sodium concentration was lower.
30
A - Phase I
Reactor- Na Conc. (meq/L)
Effluent TSS
(mg/L)
SVI (mL/g MLSS)
CST (sec)
Soluble Effluent
COD (mg/L)
Effluent COD
(mg/L)
Effluent Soluble Poly-saccharides
(mg/L)
Effluent Poly-
saccharides (mg/L)
Soluble Effluent Proteins (mg/L)
Effluent Proteins (mg/L)
Zeta Potential
(mv) 1 – 1.5 23.8 ± 10.3 74 ± 14 17.5 ± 4.4 17.4 ± 5.4 36.4 ± 5.2 6.85 ± 2.4 11.3 ± 6.3 2.6 ± 1.0 11.8 ± 3.5 ND 2 – 3.0 37.7 ± 14.5 97 ± 14 23.7 ± 3.9 ND ND ND ND ND ND ND 3 – 6.0 29.0 ± 5.5 76 ± 20 18.0 ± 0.1 22.1 ± 3.8 46.0 ± 2.2 5.75 ± 4.0 12.4 ± 7.4 3.5 ± 1.8 16.8 ± 3.5 ND
4 – 15.0 49.0 ± 10.8 90 ± 24 49.3 ± 4.7 23.5 ± 2.0 69.3 ± 5.2 6.21 ± 2.7 13.6 ± 3.2 6.0 ± 1.2 15.1 ± 2.4 ND ND – not determined
B - Phase II
Reactor- Na Conc. (meq/L)
Effluent TSS
(mg/L)
SVI (mL/g MLSS)
CST (sec)
Soluble Effluent
COD (mg/L)
Effluent COD
(mg/L)
Effluent Soluble Poly-saccharides
(mg/L)
Effluent Poly-
saccharides (mg/L)
Soluble Effluent Proteins (mg/L)
Effluent Proteins (mg/L)
Zeta Potential
(mv) 1 – 1.5 31.7 ± 11.4 76 ± 36 10.7 ± 1.8 22.5 ± 4.5 33.9 ± 0.5 5.73 ± 3.0 10.8 ± 5.9 6.1 ± 1.0 11.4 ± 1.0 -9.3 ±1.8 2 – 3.0 24.2 ± 13.1 112 ± 20 11.3 ± 1.3 20.4 ± 5.5 33.9 ± 8.7 7.71 ± 2.1 11.9 ± 3.4 6.4 ± 0.4 16.3 ± 3.0 -10.0 ± 1.9 3 – 6.0 34.2 ± 12.4 124 ± 41 21.9 ± 7.6 24.2 ± 1.9 39.8 ± 10.9 6.94 ± 4.7 10.4 ± 9.5 6.7 ± 1.0 13.2 ± 1.7 -13.1 ± 2.1
4 – 15.0 78.8 ± 29.7 90 ± 19 30.0 ± 3.3 30.1 ± 14.2 80.9 ± 22.2 7.61 ± 6.7 17.5 ± 0.7 12.4 ± 1.5 46.8 ± 0.7 -12.1 ± 2.5 C - Phase III
Reactor- Na Conc. (meq/L)
Effluent TSS (mg/L)
SVI (mL/g MLSS)
CST (sec)
Soluble Effluent
COD (mg/L)
Effluent COD
(mg/L)
Effluent Soluble Poly-saccharides
(mg/L)
Effluent Poly-
saccharides (mg/L)
Soluble Effluent Proteins (mg/L)
Effluent Proteins (mg/L)
Zeta Potential
1 – 1.5 53.3 ± 22.9 64 ± 18 10.3 ± 0.4 25.1 ± 4.6 46.7 ± 13.1 5.9 ± 1.8 13.5 ± 3.0 5.6 ± 1.6 13.2 ± 1.3 -9.9 ± 1.7 2 – 10.0 60.4 ± 16.8 95 ± 9 34.4 ± 0.2 28.0 ± 1.3 54.8 ± 8.6 7.0 ± 0.4 12.1 ± 1.6 8.8 ± 1.2 21.2 ± 12.1 -13.7 ± 3.1 3 – 6.0 34.6 ± 11.1 113 ± 24 10.5 ± 2.2 27.2 ± 5.0 36.6 ± 2.2 8.0 ± 2.5 11.6 ± 2.4 6.5 ± 1.2 13.9 ± 1.3 -15.2 ± 1.4
4 – 15.0 120.0 ± 44.1 63 ± 13 96.9 ± 4.5 46.4 ± 3.9 105.2 ± 21.1 8.8 ± 1.2 19.9 ± 2.3 15.1 ± 2.7 45.7 ± 9.4 -20.9 ± 3.2
Table 2-6. Summary of effluent quality and settling parameters under steady state condition for each phase
31
Effect on water quality, settling, and dewatering properties. Laboratory reactors were
operated until a steady state condition was achieved. A summary of the results obtained from
the laboratory reactors is shown in table 2-6. Table 2-6 A shows the results obtained for the
first phase. In this phase, after 35 days of operation, the biomass of the second reactor was
suddenly washed out of the system, resulting in a deterioration of the reactors properties. The
cause for this deterioration was unknown; therefore, this reactor was shut down and no further
data were collected for this reactor. Measurements of particle charge were not made in the
first phase because it was after analyzing the data from this phase that it was realized that it
would be of importance to analyze the particles charge and determine if there was a difference
in the charge for the flocs formed. The effects of the feed patterns are discussed in the
following sections.
Effluent TSS. The effluent total suspended solid (TSS) was measured from the beginning
until the end of each phase and was used to determine steady state conditions. Table 2-6
shows the average TSS results; these values represent the average during the steady state
period. Figure 2-3 shows a comparison of the TSS for each phase at different monovalent to
divalent ratios. A statistical difference was found among the values for the reactors
containing the lowest sodium concentration (M:D ratio of 0.7) with a corresponding p-value
of 0.0078. These data show that the third phase reactors with the longest feed time contained
the highest TSS values. The high effluent TSS concentration in these reactors was caused
mainly by deflocculation, although some filaments were observed.
The effluent TSS for each of the three phases decreased as the sodium concentration was
increased from 1.5 meq/L to 6.0 meq/L. At this concentration there was no difference among
all of the TSS values (p-value 0.50) suggesting that the feeding pattern did not have an impact
on the effluent TSS at values over a range of sodium from 1.5 to 6 meq/L or a M/D from 0.7 -
2.2. An improvement in TSS was observed when the M:D ratio was increased from 0.7 – 2.2
for the third phase, suggesting that some sodium is needed to enhance flocculation. As the
sodium concentration increased to above a M:D of 2.2, deflocculation was observed.
32
-
20
40
60
80
100
120
140
- 1.0 2.0 3.0 4.0 5.0 6.0
Monovalent : Divalent ratio (eq/eq)
TSS
(mg/
L)
Phase I Phase II Phase III
Figure 2-3. Effluent TSS comparison High TSS values were recorded for the reactors containing the highest sodium concentration
(M:D ratio of 5.2) and these values were different for each phase (p-value 0.0015). The
lowest TSS values were obtained when the feeding time was the shortest (pulse feed) and
increased as the feeding time was raised, suggesting that the negative impact on
deflocculation caused by high sodium concentration is magnified by the way the reactors are
fed. The effluent TSS increased by 70% and by more than 100% when changing from a pulse
feed to a four hour feed when the monovalent to divalent ratio in the reactors was 0.7 and 5.2,
respectively.
High effluent TSS is a sign of weak flocs which results in the formation of less dense particles
that are harder to settle (Novak et al., 1998). These small particles usually stay in suspension
and are washed out of the reactors through the wastage process which leads to a reduction in
the biomass of the system.
Chemical oxygen demand. Effluent soluble and total chemical oxygen demand
measurements were also used to assess the effluent quality of the reactors. Effluent samples
from each reactor were filtered through a 1.5 μm, 0.45 μm, and through a 1k Dalton
membrane. The results for each phase are shown in Figure 2-4; the error bars represent one
standard deviation.
33
A
-
20
40
60
80
100
120
1.5 6.0 15.0
Sodium Concentration (meq/L)
CO
D (m
g/L)
1 K Filter 0.45 μ Filter No Filtration
B
-
20
40
60
80
100
120
1.5 3.0 6.0 15.0
Sodium Concentration (meq/L)
COD
(mg/
L)
1 K Filter 0.45 μ Filter No Filtration
C
-
20
40
60
80
100
120
1.5 6.0 10.0 15.0
Sodium Concentration (meq/L)
COD
(mg/
L)
1 K Filter 0.45 μ Filter No Filtration
Figure 2-4. Effluent COD (mg/L) for A-Phase I, B-Phase II, and C-Phase III
34
Data collected showed little difference between the samples that passed through 0.45 and 1.5
μm filters; therefore, the results corresponding to the 1.5 μm samples are not shown. Figure
2-4 shows, for each phase, the COD decreased as the effluent passed through a smaller pore
size filter, demonstrating that the COD is a result of a combination of suspended organic
particles and soluble microbial products. It can be seen from this plot that the soluble COD
(samples passed through a 0.45 μm filter) is usually half the value of the total COD (unfiltered
samples), indicating that at least 50 % of the total COD is introduced by larger suspended
particles that result from deflocculation and poor settling. The difference between the
samples passed through the 1k membrane and the 0.45 μm filter is approximately 30% for the
reactors containing a sodium concentration less than 10 meq/L, but increases as the sodium
concentration is increased to 15 meq/L. These results were constant for all three phases.
-
20
40
60
80
100
120
- 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
COD
(mg/
L)
Phase I Phase II Phase III
Figure 2-5. Total effluent COD for each phase.
The results for the soluble COD (SCOD) and total COD (TCOD) for each phase are
summarized in Table 2-6. Figure 2-5 shows the different TCOD concentrations for the three
phases. It can be seen from this plot that the TCOD corresponding to the third phase is
slightly greater than for phase I and II at the lowest M:D ratio, but a statistical analysis reveals
that there is no significant difference among the three phases (p-value = 0.08). The small
COD increment observed in the third phase is due mainly to the change in feeding pattern, but
is not as significant because of the low sodium concentration. When the sodium
concentration is increased to 6 meq/L (M:D ratio of 2.2), the COD for each phase is similar, at
around 22 to 27 mg/L (p-value = 0.65).
35
-
10
20
30
40
50
- 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
COD
(mg/
L)
Phase I Phase II Phase III
Figure 2-6. Effluent soluble COD for each phase
As the M:D ratio is increased to 5.2, the effluent TCOD increased and the difference among
the three feeding strategies can be seen. A statistical analysis to compare the three phases was
not done because there were not enough data for the first phase; only the last two phases were
compared. No significant difference was found between phase II and III (p-value = 0.16), but
the trends shown in Figure 2-5 suggests that a difference between the phases exists. For this
sodium concentration, the highest TCOD value was obtained for the third phase, which
corresponds to the longest feeding time. These results show that the feeding pattern has a
greater impact in the effluent COD when the sodium concentration increases.
The soluble chemical oxygen demand (SCOD) shows a similar trend as the TCOD. This can
be seen in Figure 2-6. At the lowest M:D ratio, there is a slight difference between the first
phase and both phases II and III. This difference is due to the effect of the feeding strategy
because the sodium concentration is low in these reactors. As the sodium concentration is
increased to 6 meq/L (M:D ratio of 2.2), the SCOD concentration for all three phase
approaches a value of approximately 25 mg/L as COD. There is no statistical difference
between these values (p-value = 0.39). Once again, as the sodium concentration is increased
to 15 meq/L (M:D ratio of 5.2), the SCOD starts to increase for all three phases, but this
increment is greater in the phase where the reactors are under a longer feeding time. At this
sodium concentration there is no significant difference between phase II and III (p-value =
0.10), but Figure 2-6 reveals that there is a considerable difference between phase I and phase
36
III. This difference might be greater than the difference from the TCOD results, which
suggests that soluble microbial products like exocellular polymers contribute to the SCOD
and TCOD. Figure 2-6 also shows that the SCOD for the third phase is greater than the other
two phases.
0
40
80
120
160
Phase I Phase II Phase III
CO
D (m
g/L)
SCOD TCOD TSS
Figure 2-7. Effluent quality parameters comparison for a M:D ratio of 5.2
The results obtained for the COD and TSS analysis are strongly related, and both show
similar trends. When the sodium concentration and the feeding time are increased, the
amount of effluent TSS and COD also increases. In the third phase at lower sodium
concentration, the effluent TSS and COD concentrations are not the lowest. Both decrease as
the M:D ratio is increased to 2.2, but increase again as the sodium increases. The similarity in
these trends suggest that effluent COD is mostly representative of organics products produced
by microorganisms and released from the activated sludge floc to solution, rather than
undegraded influent. The large fractions of the effluent COD (> 50 %) from the reactors may
be generated from the biomass washed-out from the system as effluent TSS. An increase in
COD is likely related to an increase in the sum of solution protein and polysaccharides
(Murthy and Novak, 2001). When the amount of sodium present in the reactors was the
highest (M:D ratio of 5.2), the feeding pattern had a greater impact in the effluent water
quality parameter. Figure 2-7 shows that the effluent quality parameters measured as effluent
TSS, SCOD, and TCOD increased as the feeding time also increased. It can be seen from this
37
figure that the worst effluent quality parameters, at a sodium concentration of 15 meq/L, were
obtained when the feeding time to the SBRs was the longest (phase III).
Sludge Volume Index (SVI). The sludge volume index is a common parameter used to
report results obtained when evaluating floc and settling properties. Many researchers have
tried to correlate the SVI to many other effluent quality parameters. The SVI was also used to
determine steady state conditions. Table 2-6 shows a summary of the SVI values obtained
throughout the steady state period. Figure 2-8 shows a comparison of the SVI for each phase.
-
20
40
60
80
100
120
140
0 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
SVI (
mL/
g M
LSS)
Phase 1 Phase 2 Phase 3
Figure 2-8. Settling properties as a function of feeding time and sodium concentration.
The results from the SVI analysis show that there is a difference among the reactors for each
phase. For the reactor containing a M:D ratio of 2.2, there is a statistical difference (p-value =
0.06) between phases I and II, phases I and III, but there is no difference between phases II
and III. For the reactor containing a M:D of 5.2, there is a statistical difference (p-value =
0.056) between phases I and III, phases II and III, but no difference between phases II and III.
There was no statistical difference for the reactors containing the lowest M:D ratio (p-value =
0.40). This last result suggests that when there is low sodium, there is no significant
difference in the SVI when the feeding time is changed. The data for phase I follow an
expected trend, as the sodium concentration increases, the SVI increases showing a negative
38
impact on the settling properties. For phases II and III the same trend is observed until
reaching a concentration of 6.0 meq/L of sodium (M:D ratio of 2.2).
During phases II and III the SVI started to decrease when the M:D ratio > 2.2. This was not
the expected trend. This decrease was due to a change in particle properties (e.a., size,
porosity, and density) which affected the settling properties of the activated sludge. Particles
experience an increase in their porosity reducing the effect of water on settling by reducing
settling drag (Higgins and Novak, 1997c). The change in feeding pattern and the large
amount of sodium stimulated deflocculation that resulted in small, denser flocs that settled
and compacted very easily, leaving high amount of suspended solids in suspension. This
gives a false sense of the sludge volume index because only a part of the particles settled
which represents only a fraction of the biomass. This is in accordance with authors who state
that the SVI is an incomplete measurement of the settling behavior of activated sludge
(Novak, 2001). No analysis related to floc size and densities were done, but changes in
particle properties were so evident that a visual inspection of the flocs and a microscopic
analysis was sufficient to substantiate these changes.
This situation makes it almost impossible to estimate if changes on the feeding strategies
under high concentration of sodium affected the SVI. From this analysis it can only be stated
that the change in feeding pattern affected the SVI at sodium concentrations less than 6
meq/L, though it cannot be stated which phase introduced a greater deterioration because
there is no significant difference between the values from phases II and III at this sodium
concentrations. The presence of some filamentous organisms during the third phase did not
contribute to changes in the SVI for each reactor.
Capillary Suction Time. The dewatering characteristics of the biological suspension were
determined by capillary suction time (CST). The results obtained for each phase are
summarized in Table 2-6. Figure 2-9 shows a comparison among the reactors containing the
same concentration of sodium for each phase. When the M:D ratio was less than 2.2, the CST
was not affected by the change in feeding pattern, although Figure 2-9 shows little
improvement in the CST as the feeding time was extended. A statistical analysis was done to
39
demonstrate that there is no difference between the CST for the 1.5 and 6.0 meq/L sodium
reactors (M:D ratio of 0.7 and 2.2) for phase II and phase III (p-values = 0.73 and 0.10
respectively). There were not enough data to include phase I in the statistical comparison.
0
20
40
60
80
100
0 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
CST
(sec
) Phase 1 Phase 2 Phase 3
Figure 2-9. CST comparison for three different feeding patterns
at different sodium concentrations.
When the sodium concentration was raised from 6 - 15 meq/L or a M/D from 0.7 - 2.2, the
CST continued to increase in all three phases. Figure 2-9 reveals that phases I and II shared a
similar trend; when the amount of sodium was increased, the CST also increased. Phase I has
a higher CST for every sodium concentration when compared to phase II, but this difference
is not considered to be significant.
Phase III experienced a higher value in CST as the amount of sodium increased. Figure 2-10
shows that when the M:D ratio was equal to 5.2, the CST for phase III was more than twice
the value of phase I and II. The CST for phase III was considerably higher than for phase II
(p-value = 0.0001). These results suggest that an increase in the feeding time will affect the
dewatering properties of activated sludge containing a M:D ratio greater than 2.2. The high
deflocculation might be the cause of the deterioration in dewatering properties since smaller
particles are associated with poor dewatering.
40
0
20
40
60
80
100
120
Phase I Phase II Phase III
CST
(sec
)
Capillary Suction Time
Figure 2-10. CST comparison for a M:D ratio of 5.2 Figure 2-10 reveals that phase III has the highest CST as expected. This figure also shows
that as the feeding time was increase from 1 – 60 minutes the CST decreased at this M:D
ratio, but from the results obtained it was not possible to demonstrate if there was a significant
difference between the values from phases I and II.
0
0.05
0.1
0.15
0.2
0.25
0 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
Pol
ymer
Opt
imum
Dos
e (m
g/gT
SS)
Phase 2 Phase 3
Figure 2-11. Optimum dose for phase II and phase III
A CST and optimum polymer dose analysis was done to measure the impact that might exist
in conditioning due to the difference found in CST between phase II and III. After adding a
cationic polymer and gently mixing for a short period of time, the CST was measured, the
polymer dose versus CST profile was developed, and the optimum polymer dose was
determined as the dose corresponding to the minimum CST value. The results are plotted in
41
Figure 2-11. According to Figure 2-11, as the feeding time is reduced and it approximates to
a pulse feed (phase II), the polymer required to improve the dewatering properties of the
activated sludge decreased. An increase in polymer conditioning demand implies a greater
number of negative charged sites available in the floc and an increase in negative biological
colloids in solution, confirming that the feeding strategy influences dewatering properties.
Particle charge. A zeta-potential analysis was done to characterize the biopolymer by
determining the particle surface charge. Surface charge is related to the production,
composition and physical characteristics of EPS, and it is related to the ionizable groups
present on sludge surfaces. Surface charges increase the polar interactions of EPS with water
molecules. Therefore, the more charged the sludge surface, the lower the hydrophobicity.
-25
-20
-15
-10
-5
00 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
Z-P
oten
tial (
mv)
Phase II Phase III
Figure 2-12. Zeta-potential values for phase II and phase III
Table 2-6 summarizes the results obtained for the zeta-potential analysis performed only in
the last two phases of the study. A comparison of the results from phases II and III is shown
in Figure 2-12. Like most of the previous results, there is no difference between the values
that correspond to a M:D ratio of < 2.2, (p-value = 0.38 and 0.52), but as the amount of
sodium is increased, a difference in zeta-potential is observed. When the Na+ concentration
reaches a value of 15 meq/L (M:D ratio of 5.2) there is a significant difference between these
values (p-value = < 0.0001). Results show that the feeding pattern affected also the particle’s
charge when the M:D ratio was higher than 2.2. This result is related to previous analyses
42
were there was a deterioration on the settling, dewatering and effluent quality of the activated
sludge, as the feeding pattern and sodium concentration changes.
Effluent biopolymer substances. Unbound proteins and polysaccharides were measured to
determine if the feed Na+ and feeding strategy were related to sludge and effluent properties.
Biopolymer in solution phase and effluent can be used to characterize the extent of
biopolymer binding in activated sludge floc because biopolymer will remain in solution if
flocculation is poor. High soluble and colloidal biopolymers suggest that flocculation is poor
and that a substantial fraction of organic matter is being washed out of the system. Effluent
samples from each reactor were filtered through a 1.5 μm, 0.45 μm, and through a 1k Dalton
membrane and as in the COD analysis, the data related to the effluent that passed through the
1.5 μm filter are not shown. The results for each phase are shown in Figures 2-13 and 2-14;
the error bars represent one standard deviation.
Figures 2-13 and 2-14 show that the polysaccharides and proteins are reduced as they are
filtered through different openings. Like the COD, the soluble polysaccharide concentration
is approximately 50 % of the total polysaccharides, suggesting that the effluent contains half
soluble and half colloidal polysaccharides. That is not the case for the proteins concentration
where the amount of soluble proteins was less than 50 % of the total proteins, especially in the
reactors where sodium was high. The amount of polysaccharides that passed the 1k
membrane remained almost constant for each phase, showing little change as the sodium
concentration was increased. The protein concentrations has a similar trend with the
exception that at higher sodium concentrations the amount of protein that passed through the
1k membrane increased, and were magnified as the feeding time also increased.
The results for soluble and total polysaccharide and for soluble and total proteins for each
phase are summarized in Table 2-6. Figure 2-15 shows the different polysaccharide
concentrations for the three phases. It can be seen from this figure that at the lowest M:D
ratio the amount of total effluent polysaccharides is higher for the third phase, but no
statistical difference was found among the three phases (p-value = 0.78).
43
A
-
5
10
15
20
25
1.5 6.0 15.0
Sodium Concentration (meq/L)
Poly
sacc
harid
es (m
g/L)
1 K Filter 0.45 Filter No Filter
B
-
5
10
15
20
25
1.5 3.0 6.0 15.0
Sodium Concentration (meq/L)
Pol
ysac
char
ides
(mg/
L)
1 K Filter 0.45 Filter No Filter
C
-
5
10
15
20
25
1.5 6.0 10.0 15.0
Sodium Concentration (meq/L)
Poly
sacc
hari
des
(mg/
L)
1 K Filter 0.45 Filter No Filter
Figure 2-13. Effluent polysaccharides (mg/L) for A-Phase I, B-Phase II, and C-Phase III
44
A
-
10
20
30
40
50
1.5 6.0 15.0Sodium Concentration (meq/L)
Prot
ein
Conc
(mg/
L)
1 K Filter 0.45 Filter No Filter
B
-
10
20
30
40
50
1.5 3.0 6.0 15.0
Sodium Concentration (meq/L)
Pro
tein
Con
c (m
g/L)
1 K Filter 0.45 Filter No Filter
C
-
10
20
30
40
50
1.5 6.0 10.0 15.0
Sodium Concentration (meq/L)
Prot
ein
Conc
(mg/
L)
1 K Filter 0.45 Filter No Filter
Figure 2-14. Effluent proteins (mg/L) for A-Phase I, B-Phase II, and C-Phase III
45
-
4
8
12
16
20
- 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
Poly
sacc
harid
es (m
g/L)
Phase I Phase II Phase III
Figure 2-15. Total effluent polysaccharides for each phase
As the M:D ratio increases to a value of 2.2, there is a slight decrease in total effluent
polysaccharides for phase II and III. At this sodium concentration there is no apparent
difference among the polysaccharides concentrations for each phase (p-value = 0.9199),
suggesting that the feed pattern does not affect the amount of polysaccharides released into
solution when the M:D ratio is less than 2.2. When the M:D is further increased to 5.2 by
increasing the Na+ concentration, the total effluent polysaccharides start to increase and a
difference among the three feeding strategies is clear.
4
5
6
7
8
9
10
- 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
Poly
sacc
harid
es (m
g/L)
Phase I Phase II Phase III
Figure 2-16. Soluble effluent polysaccharides for each phase
46
The effluent soluble polysaccharide shows a different trend than the total polysaccharides. It
can be seen from Figure 2-16 that for phase I there is no change in soluble polysaccharides as
the sodium concentration increases. For phases II and III there is a slight increase in soluble
polysaccharides when the M:D ratio is increased from 0.7 to 1.2, followed by a sudden
decrease as sodium increases the M:D ratio to 2.2. Finally, the polysaccharide in solution
increases as the sodium reaches its highest value. No significant difference was obtained for
any of the three M:D ratios being compared (p-value = 0.92, 0.73, 0.77).
Figure 2-17 shows the effluent proteins as the M:D ratio changes for each phase. For phase I
the amount of proteins released into solution slightly varies as the sodium concentration is
increased, similar to what happened with the effluent polysaccharides. When the feeding
strategy is changed from a pulse feed to a slower feeding pattern, the amount of effluent
proteins start to increase as the sodium concentration is raised. When the M:D ratio is
minimal, there is no difference in the amount of proteins in the effluent among the three
phases (p-value = 0.13). There is also no statistical difference when M:D ratio increases to
2.2 (p-value = 0.99).
-
5
10
15
20
25
30
35
40
45
50
0 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
Pro
tein
s C
onc
(mg/
L)
Phase I Phase II Phase III
Figure 2-17. Total effluent proteins for each phase
However, when the M:D ratio is increased to a value greater than 2.2, the total effluent
proteins start to increase and a difference among the three feeding strategies is clear. There is
a significant difference in the amount of effluent proteins between phase I and phase II and
47
between phases I and III, when sodium reaches a value of 15 meq/L (M:D ratio of 5.2) (p-
value = 0.0006); there is no different between phases II and III. This result suggests that
increasing the feeding time magnified the effect of sodium in bioflocculation, as reflected by
the amount of proteins released into the effluent.
The effluent soluble protein shows a similar trend to the total effluent proteins. Figure 2-18
shows that for the three phases, the amount of soluble proteins increases as the sodium
concentration also increase with a slight difference in the rate for each phase.
-
2
4
6
8
10
12
14
16
0 1 2 3 4 5 6
Monovalent : Divalent ratio (eq/eq)
Prot
eins
Con
c(m
g/L)
Phase I Phase II Phase III
Figure 2-18. Soluble effluent proteins for each phase
According to this plot the rate of increase of soluble proteins versus sodium concentration is
higher for the third phase, followed by the second phase. This plot also reveals that the
concentration of soluble proteins for phase I is lower than for phase II and III for all sodium
concentrations. At a M:D ratio of 5.2 there is a significant difference between phase I and II,
and between phase I and III (p-value = 0.0026). There was no difference found between
phase II and III at any M:D ratio (p-value = 0.66, and 0.79). This result suggests that
changing the feeding pattern has also an impact in effluent soluble proteins, but with a lower
rate when comparing it to the increase in total effluent proteins.
Proteins and polysaccharides were both affected by the feeding pattern at the highest M:D
ratio. Figure 2-19 shows a comparison of the total proteins and polysaccharides for each
48
phase at the highest M:D ratio where the feeding pattern has the highest impact. Figure 2-19
shows that the total polysaccharides increased as the feeding time was increased and as stated
earlier, there is a significant difference among these values. The release of proteins to the
effluent seems to be more sensitive than the polysaccharides, this can be seen from Figure 2-
19 that increasing the feeding time from 1 – 60 minutes resulted in a high amount of effluent
proteins, but increasing further the feeding time did not resulted in a further increase of
proteins in the effluent.
0
10
20
30
40
50
60
Phase I Phase II Phase III
Bio
poly
mer
Con
c. (m
g/L)
Total Polysaccharides Total Proteins
Figure 2-19. Proteins and Polysaccharides comparison for a M:D ratio of 5.2
The results obtained for the effluent biopolymer substances analysis are strongly related since
the results for both proteins and polysaccharides showed similar trends. The amount of
proteins and polysaccharides released into solution was very sensitive to the change in feeding
pattern, with protein being the one that experienced a higher change. From our study it can be
determined that the effluent proteins and polysaccharides increased dramatically when
changed from phase I to phase II. When the feeding time was further increased from one hour
to four hours, a slight change in the amount of soluble proteins and polysaccharides was
observed. Statistical analysis demonstrated that this latter change in proteins and
polysaccharides concentrations was not significant. Although these values remained very
close as the substrate gradient was decreased, the effluent quality, settling, and dewatering
properties continued to deteriorate, as reflected by increases in effluent TSS and COD and
CST.
49
The changes in these properties could be related to the differences in EPS composition of the
activated sludge which depends on the microbial community. Researchers agree that the
quantity of EPS plays a very important role in flocculation, but may not be the only thing to
consider when relating EPS to settling (Sponza, 2004). The presence of a similar amount of
soluble proteins and polysaccharides along with a continuous deterioration on effluent quality
as the feeding time is increased agree with studies that state that settling and effluent quality
parameters depend not only on the quantity of EPS, but also on the composition and physical
configuration of specific EPS molecules. In some cases the composition of EPS is more
important than the quantity of EPS in activated sludge (Sponza, 2004). A change in the
microbial community, which is expected as a result from changing the feeding pattern due to
a change in substrate intake, might be expected to contribute to some change in the EPS.
Some authors (Liao et al., 2000) suggest that lower microbial growth rates can produce large
amount of cell lysis which contribute to proteins accumulation due to secretion and cell lysis.
This accumulation of proteins can change the proportions of biopolymers forming the EPS
and different EPS components may have different roles (Morgan et al., 1990).
The results obtained for the effluent biopolymers analysis reveals that that the proteins and
polysaccharides in solution contribute to the TCOD and SCOD measured for each phase.
Also, the higher amount of proteins and polysaccharides in the effluent coincide with the high
amount of solids in the effluent measured as TSS which is in accordance with many studies
which suggest that biopolymers in solution is a sign of poor flocculation (Park et al., 2006).
The poor floc formation results in a release of organic matter to the effluent, which is a
fraction of the biomass composed mainly by particles rich in proteins and polysaccharides
which are reflected in the effluent COD, and effluent proteins and polysaccharides. The
increase in microbial soluble products (exocellular proteins and polysaccharides) is a result of
a decrease in retention of biopolymers in the floc which increase the effluent COD, suggesting
that this increase is not due to a change in kinetics of microbial degradation (Higgins et al.,
2004b). Figures 2-11 and 2-12 also show an expected trend that is related to effluent proteins
and polysaccharides where an increase in the amount of EPS released in the effluent could be
directly related to increased values of negative surface charge of the flocs. This indicates that
50
the EPS contain mainly negatively charged groups that contribute to the binding of the
different floc constituents.
Implications: The effect of various cations on activated sludge characteristics was not
expected to be independent of the feeding pattern. It has been shown that excess sodium in
activated sludge systems has a negative impact on the characteristics related to effluent
quality, settling, and dewatering properties, explained by deterioration in bioflocculation due
to high concentration of monovalent to divalent cations present in the feed. Also, from the
results obtained, it can be stated that small difference in the aerobic fill time had a dramatic
effect on the activated sludge properties. In general, it can be stated from the overall results
that effluent quality parameter such as effluent TSS, effluent COD, and effluent biopolymers
along with the dewatering properties measured by the CST, were affected by the negative
effect of excess sodium and this effects were magnified by the rate at which the reactors were
fed, especially at higher sodium concentrations, usually above 6 meq/L (M:D ratio 2.2).
Figures 2-20 – 2-22 helps confirm these negative effects.
Reactor 1, Phase II - 4x Reactor 1 Phase III – 4x
Reactor 1, Phase II - 10x Reactor 1, Phase III – 10x
Figure 2-20. Effect of feeding pattern on bioflocculation for 1.5 meq/L Na+.
51
Reactor 3, Phase II - 4x Reactor 3, Phase III – 10x
Reactor 3, Phase II - 10x Reactor 3, Phase III – 10x
Figure 2-21. Effect of feeding pattern on bioflocculation for 6.0 meq/L Na+.
Reactor 4, Phase II - 4x Reactor 4, Phase III – 4x
Reactor 4, Phase II - 10x Reactor 4, Phase III – 10x
Figure 2-22. Effect of feeding pattern on bioflocculation for 15 meq/L Na+.
52
From these microscopic pictures, it can be seen that for an equal amount of sodium there is
change on the floc morphology introduced by the different feeding strategies. A greater
deflocculation can be observed in the phase where the feeding time was greater confirming
the negative effect on bioflocculation due to a high monovalent to divalent ratio and a larger
feeding time.
Data obtained reflect that changing the feeding strategy when the M:D ratio was 2.2 produced
no effect in the activated sludge properties. When comparing the three phases, there is no
significant evidence to prove that there is a difference in the effluent quality and dewatering
properties among the three phases. Some authors suggest that some quantity of monovalent
cations appears to be beneficial for several floc properties (Murthy and Novak, 2001),
although excessive amounts of monovalent cations can cause deterioration in most floc
properties (Higgins et al., 2004b). The sodium aided flocculation in phase III until reaching a
M:D ratio of approximately 2.2. Above this, flocculation deteriorated.
Engineering and Scientific Significance. The implications of the previous study are
significant for wastewater treatment plants that have high concentrations of monovalent
cations. When the M:D ratio is greater than 2, besides adjusting the amount of divalent
cations, plant operators can adjust their influent flows to improve floc properties. This can be
done by building equalizations basin to store certain amounts of wastewater and released it a
high flows in short periods of time resembling a pulse feed. It has been demonstrated that
considerable changes in feeding pattern results in noticeable changes in flocculation, but
considering the restrictions and economical constraints, designers should look for the
adequate balance between influent flows, storage, polymers aids, and effluent quality.
Designers should consider up to what extent they can reduce their influent flow without
increasing the costs of operation. Industrial wastewater treatment plants operating continuous
mixed reactors with a low influent flow rate (similar to a SBR with a very long feeding
period), should consider changing their reactor configuration or simply reduce their feeding
time in order to approximate to a plug flow reactor, which has been demonstrated to have
better effluents quality and better settling properties.
53
The implications of this study are also significant when performing research studies on
activated sludge flocculation. It has been demonstrated that the feeding pattern has an effect
on settling and dewatering properties and also in effluent quality; therefore, when reporting
data collected in lab-scale reactors, the feeding time should be reported in order to compare
results that have the same initial considerations. Also, if researchers want to reduce any effect
when performing any study on bioflocculation using lab-scaled reactors, they should consider
eliminating any variable that will affect their study, including the effects introduced by
changes in the feeding pattern which can be minimized by standardizing the influent flow and
feeding time.
Conclusions The results obtained from lab-scale reactors for this study demonstrated that changing the
feeding pattern will have an impact on the settling and dewatering properties of activated
sludge. Increasing the substrate gradient in the system by increasing the feeding time or
reducing the influent flow has a strong negative impact on effluent water quality, and the
settling and dewatering properties of the activated sludge.
Sodium, at higher amounts, have a negative effect on activated sludge properties and effluent
quality, measured as effluent TSS, COD, SVI, CST and soluble effluent biopolymers.
Deflocculation resulting from the presence of a high M:D ratio in the influent is magnified as
the feeding time is increased. Poor flocculation is a result of changes in particle properties,
filamentous organisms, and changes in the amount and quality of EPS as a result of microbial
community changes due to the variation in substrate rate intake by bacteria. Therefore,
linking these processes with the microbial physiology and microbial population dynamics
might be the key to explain the different degrees of sludge settleability and dewatering
improving effluent quality. In order to prove this, more bacterial community analysis should
be done and a deeper understanding in microbial kinetics is needed. A value of 6.0 meq/L of
sodium is suggested as a point where the change in feeding pattern, as the sodium
concentration is increased, will start to affect bioflocculation at a higher rate resulting in poor
settling and bad effluent quality.
54
The knowledge of this effect is relevant to define strategies to prevent poor effluent quality
and improve settling qualities. From an engineering point of view a better understanding of
how the feeding pattern affects flocculation might be useful to operate wastewater treatment
plants. By increasing the flow and reducing the feeding time through an equalization basin or
a storage tank, operators in a wastewater treatment plant can improve water effluent quality
without the aid of any flocculants especially in industrial wastewater treatment plant where
sodium concentration are usually high due to chemicals used for pH control. From a
laboratory and research point of view, reporting the feeding flow and feeding time of lab-
scaled reactors might be useful to a better understanding of the results and can be useful to
standardize procedures to better compare data.
55
References
Association, American Public Health, Association, American Water Works & Federation, Water
Environment (1995) Standards Methods for the Examination of Water and Wastewater,
Washington, D.C.
Awwa (1999) Water Quality & Treatment A Handbook of Community Water Supplies, McGraw-Hill,
New York.
Bruus, J.H., Nielsen, P.H. & Keiding, K. (1992) On the stability of activated sludge flocs with
implications on dewatering. Water Research, 26.
Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A. & Smith, F (1956) Colorimetric Methods for
Determiantion of Sugars and Related Substances. Analytical Chemistry, 28.
Govoreanu, R., Seghers, D., Nopens, I., De Clercq, B., Saveyn, H., Capalozza, C., Van Der Meeren,
P., Verstraete, W., Top, E. & Vanrolleghem, P.A. (2003) Linking floc structure and settling
properties to activates sludge population dynamics in an SBR. Water Science and Technology,
47, 9-18.
Grady, C.P. Leslie, Daigger, Glen T. & Lim, Henry C. (1999) Biological Wastewater Treatment,
New York.
Hartree, E.F. (1972) Determination of Protein. A Modification of the Lowry Method That Gives a
Linear Photometric Response. Analytical Biochemistry, 48, 422-428.
Higgins, Mathew J. & Novak, J.T. (1997a) Dewatering and settling of activated sludges: The case
for using cation analysis. Water Environment Research, 69, 225-232.
Higgins, Mathew J. & Novak, J.T. (1997b) Characterization of Exocellular Proteins and Its Role in
Bioflocculation. Journal of Environmental Engineering, 123, 479-485.
Higgins, Mathew J. & Novak, J.T. (1997c) The effect of cations on the settling and dewatering of
activated sludges: Laboratory Results. Water Environment Research, 69, 215-225.
56
Higgins, Mathew J., Sobeck, David C., Owens, Steven J. & Szabo, Lynn M. (2004a) Case Study II:
Application of the Divalent Cation Bridging Theory to Improve Biofloc Properties and
Indistrial Avtivated Sludge System Performance-Using Alternatives to Sodium Based
Chemicals. Water Environment Research, 76, 353-359.
Higgins, Mathew J., Tom, Lou A. & Sobeck, David C. (2004b) Case Study I: Application of the
Divalent Cation Bridging Theory to Improve Biofloc Properties and Industrial Activated
Sludge System Performance-Direct Addition Divalent Cations. Water Environment Research,
76, 344-352.
Jenkins, D, Richard, M.G. & Daigger, Glen T. (1986) Manual on the Causes and Control of Activated
Sludge Bulking and Foaming. Ridgeline Press, Lafayette, California.
Liao, B.Q, Allen, D.G., Droppo, I.G., Leppard, G.G. & Liss, S.N. (2000) Surface Properties of
Sludge and Their Role In Bioflocculation and Settleability. Water Research, 35, 339-350.
Madigan, Michael T., Martinko, John M. & Parker, Jack (2003) Brock Biology of Microorganisms,
Prentice Hall, New Jersey.
Martins, Antonio M.P., Heijnen, Joseph J. & Van Loosdrecht, Mark C.M. (2003) Effect of feeding
pattern and storage on the sludge settleability under aerobic conditions. Water Research, 37,
2555-2570.
Metcalf & Eddy (2003) Wastewater Engineering Treatment and Reuse, McGraw Hill, New York.
Murthy, Sudhir N. & Novak, J.T. (1998) Effects of Potassium Ion on Sludge Settling, Dewatering
and Effluent Properties. Water Science and Technology, 37, 317-324.
Murthy, Sudhir N. & Novak, J.T. (1999) Factors Affecting Floc Properties During Aerobic
Digestion: Implications for Dewatering. Water Environment Research, 71, 197-202.
Murthy, Sudhir N. & Novak, J.T. (2001) Influence of Cations on Activated-Sludge Effluent Quality.
Water Environment Research, 73, 30-36.
57
Murthy, Sudhir N., Novak, J.T. & De Hass, Robert D. (1998) Monotoring cations to predict and
improve activated sludge settling and dewatering properties of industrial wastewaters. Water
Science and Technology, 38, 119-126.
Muyzer, G., De Waal, C.E. & Uitterlinden, G.A. (1993) Profiling of complex microbial populations
by denaturing gradient gel electrophoresis of polymerase chain reaction-amplified genes
coding for 16S rRNA. Applied Environmental Microbiology, 59, 695-700.
Novak, J.T. (2001) The Effect of the Ammonium Ion on Activated-sludge Settling Properties. Water
Environment Research, 73, 409-414.
Novak, J.T., Love, N.G., Smith, M.L. & Wheeler, E.R. (1998) The effect of Cationic Salt Addition
on the Settling and Dewatering Properties of an Industrial Activated Sludge. Water
Environment Research, 71, 251-254.
Park, Chul, Muller, Christopher D., Abu-Orf, Mohammad M. & Novak, J.T. (2006) The Effect of
Wastewater Cations on Activated Sludge Characteristics: Effects of Aluminum. Water
Environment Research, 78, 31-40.
Sobeck, David C. & Higgins, Mathew J. (2002) Examination of three theories for mechanisms of
cation-induced bioflocculation. Water Research, 36, 527-538.
Sponza, Delia T. (2004) Properties of Four Biological Flocs as Related to Settling. Journal of
Environmental Engineering, 130, 1289-1300.
Wilén, B.M., Jin, B. & Lant, P. (2003) Relationship between flocculation of activated sludge and
composition of extracellular polymeric substances. Water Science and Technology, 47, 95-
103.
58
Appendix
Table A-1. Cations concentration in bactopeptone used on feed,
measured with an Inductively Coupled Plasma instrumentation
Constituent mg/L meq/L
Na+ 4.80 0.21
K+ 0.57 0.02
Ca2+ 0.16 0.01
Mg2+ 0.02 0.002
Al3+ 0.0017 0.0002
COD 300