Aquacultural Engineering 34 (2006) 179–197

www.elsevier.com/locate/aqua-online

Nitrification kinetics of biofilm as affected by

water quality factors
Shulin Chen a,*, Jian Ling a, Jean-Paul Blancheton b
a
Department of Biological Systems Engineering, Washington State University,
Pullman, WA 99164-6120, USA
b
Station Ifremer de Palavas, 34250 Palavas les flots, France

Abstract

Various types of fixed film biofilters have been used in recirculating aquaculture systems under different water quality and
operating conditions. The effectiveness of the nitrification process can be evaluated by nitrification kinetics. Nitrification in the
bacterial film of the biofilter involves physical, chemical and biological processes that are governed by a variety of parameters
such as substrate and dissolved oxygen concentrations, organic matters, temperature, pH, alkalinity, salinity and turbulence
level. The impacts of these parameters upon nitrification kinetics make predicting the performance of a biofilter for a given
application an engineering challenge. Knowing the performance of a biofilter is critical for both designers and managers. This
paper summarizes the current knowledge on nitrification kinetics as affected by the aforementioned factors based on literature
and the results from the authors’ laboratories. These factors were ranked according to their significance of impact on biofilter
nitrification performance. The information presented can be used as a reference for the design and operation of biofilters in
recirculating aquaculture systems.
# 2005 Elsevier B.V. All rights reserved.

Keywords: Nitrification kinetics; Biofilters; Biofilm; Water quality parameters; Recirculating aquaculture systems; Review

1. Introduction nia to nitrate. Ammonia exists in aqueous solution in

two forms: NH3 and NH4+. Although both forms may
1.1. Process description be toxic to fish, unionized ammonia (NH3) is the more
toxic form at low concentrations (Meade, 1985). In
Biological filters are commonly used for ammonia general, the acceptable level of unionized ammonia
removal in recirculating aquaculture systems (RAS) nitrogen in fish culture waters is only 0.0125 mg L
1
through biological nitrification which converts ammo- (Colt and Armstrong, 1981; Meade, 1985). In a
different report, the design criterion for unionized
ammonia nitrogen concentration in RAS is determined
* Corresponding author. Tel.: +1 509 335 3743 (O);
fax: +1 509 335 3743. as 0.025 mg L
1 (Losordo and Westers, 1994). The
E-mail address: [email protected] (S. Chen). proportion of unionized ammonia relative to ionized

0144-8609/$ – see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.aquaeng.2005.09.004
180 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

ammonia (NH4+), for a given total ammonia (NH3 and the bulk solution. In a pure culture under a single
NH4+) concentration in an aqueous solution, depends limiting-substrate condition, the steady-state kinetics
on temperature, pH and salinity (Trussell, 1972). of substrate removal is usually described by the
Lower pH and temperatures increase the percentage of Monod-type expression (Srna and Baggaley, 1975;
ionized ammonia at the expense of unionized Rittmann and McCarty, 1980; Drtil et al., 1993):
ammonia (Huguenin and Colt, 1989). Thus, the total
X S
ammonia nitrogen (TAN) concentration instead of R ¼ mmax (4)
YS KS þ S
ammonia nitrogen is often used as a key limiting water
quality parameter in intensive aquaculture systems where R: substrate removal rate (g m
3 day
1); mmax:
design and operation (Losordo and Westers, 1994). maximum specific growth rate (day
1); X: bacterial
In a nitrification process, ammonia is first oxidized mass concentration (g cell m
3); YS: yield of bacterial
into nitrite (NO2
) by several genera of autotrophic mass per unit of substrate used (g cell g
1 sub-
bacteria, the most important being Nitrosomonas. strate
1); S: limiting substrate concentration
Nitrite is then oxidized to the much less toxic nitrate (g m
3); KS: half saturation constant (g m
3).
(NO3
) by several other genera of bacteria, the most This equation can also be used to describe
important of which is Nitrobacter. Eqs. (1) and (2) nitrification kinetics when ammonia is assumed as
show the basic chemical conversions occurring in a the growth-limiting substrate for Nitrosomonas while
nitrification process (USEPA, 1984; WPCF, 1983). nitrite as the limiting substrate for Nitrobacter. It was
reported that the growth rate of Nitrobacter is greater
NH4 þ þ 1:5O2 ! 2Hþ þ H2 O þ NO2
(1) than that of Nitrosomonas (WPCF, 1983), and
oxidation of ammonia is usually the rate-limiting
NO2
þ 1:5O2 ! NO3
(2)
step in the conversion of ammonia to nitrate. Thus, in
Energy released from the above conversions is used Eq. (4), values for ammonia oxidation are the rate-
by Nitrosomonas and Nitrobacter to drive their life limiting parameters in describing nitrification (Whea-
processes. In addition, these reactions require oxygen, ton et al., 1994). Both mathematical analysis and
produce hydrogen ions (lowering pH) and produce experimental data revealed two major characteristics
nitrite as an intermediate product. According to of Eq. (4). First, at a sufficiently high substrate
USEPA (1984), the complete nitrification process concentration, Eq. (4) becomes a zero-order expres-
can be expressed as: sion, meaning the reaction rate does no longer increase
with respect to substrate concentration. In this case,
NH4 þ þ 1:83O2 þ 1:98HCO3
factors other than substrate concentration become the
! 0:021C5 H7 O2 N þ 0:98NO3
þ 1:041H2 O limiting factor. Second, at a sufficiently low substrate
þ 1:88H 2 CO3
(3) concentration, the relationship becomes linear, fol-
lowing a first-order reaction.
where C5H7O2N presents the chemical expression of
the bacterial cell of Nitrosomonas and Nitrobacter. 1.3. Nitrification kinetics of biofilm
Eq. (3) can be used to estimate oxygen and alkalinity
requirements and the cell biomass produced. For every Biological nitrification can be accomplished in two
gram of TAN oxidized to nitrate nitrogen, approxi- types of systems: suspended and attached growth.
mately 4.18 g of oxygen (or 4.57 g according to Under a suspended growth environment, the micro-
Losordo and Westers, 1994) and 7.07 g of alkalinity organism is freely mobile in the liquid providing direct
(as CaCO3) are consumed and 0.17 g of bacteria contact between the bacteria cells and the bulk water.
biomass is produced. In an attached growth system, on the other hand,
microorganisms are grown in a viscoelastic layer of
1.2. Nitrification kinetics biofilm that are attached on the surface of a solid
support medium. Thus, this process is also called a
The rate of ammonia or nitrite oxidation depends fixed film process in which the individual bacteria are
strongly on the concentrations of these substrates in immobilized. Attached growth on a fixed biofilm
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 181

system offers several advantages when compared to nitrification reactions occur in the biofilm instead of in
suspended growth processes, such as handling the bulk liquid (Moreau et al., 1994), the substrate
convenience, increasing process stability in terms of utilization rate depends on local substrate concentra-
withstanding shock loading and preventing the tions within the biofilm. Nitrifying populations deep
bacteria population from being washed off (Fitch within the biofilm are maintained by endogenous
et al., 1998; Nogueira et al., 1998). Because most fish respiration under limited oxygen conditions, and
species tolerate only low level ammonia concentra- nitrifying populations on the surface are the only
tions, which is most suitable for the advantages of survivors under limited ammonia conditions (Horn,
attached growth processes, biofilters that operate on 1994). At local reaction sites, reactant concentrations
the principle of attached biofilm have been extensively are depressed, and product concentrations are elevated
applied in RAS. (Boller et al., 1994). Therefore, substrate concentra-
The kinetics of biofilm is more complex, as the tion is different at the different depths of the biofilm.
substrate supply into the layer-like aggregation of At a given point in the biofilm, Eq. (4) applies. A
bacteria film is a diffusion-controlled process driven precise analysis of nitrification kinetics as a function
by the concentration gradient across the biofilm of substrate concentration has to consider the variation
(Fig. 1). Next to the biofilm there is a water film that in substrate concentrations within the biofilm. How-
serves as the interface between the biofilm and the ever, for simplicity, attempts have been made to relate
bulk water. Therefore, the kinetics of biofilm reactions the nitrification rate directly to total ammonia
is influenced by mass transport (Rasmussen and concentration in the bulk liquid (Zhu and Chen,
Lewandowski, 1998). Factors that determine diffusion 1999). According to this approach, the ammonia
rate, such as the local chemical environment and flow concentration in the bulk liquid is treated as an
conditions, influence the rate of substrate supply and apparent concentration that is implicitly related to the
subsequently, the extent of biofilm growth. This may local concentration at a given point in the biofilm. The
have repercussions for biofilter operation in many nitrification rate can be expressed by the maximum
applications. The diffusion and transport process reaction rate with a correction term that reflects the
should be considered in order to better understand the effect of limiting substrate (Eq. (5)). Such treatment
nitrification kinetics of fixed film biofilters. For was demonstrated effective in evaluating biofilm
example, it was reported that liquid film diffusion nitrification rate (Zhu and Chen, 1999, 2002), thus will
has a considerable influence on the performance of a be used in this paper.
fixed film reactor (Tanaka and Dunn, 1982). Since
S
R ¼ Rmax (5)
KS þ S
where R: substrate oxidation rate (g m
2 day
1); Rmax:
maximum substrate oxidation rate (g m
2 day
1); S:
limiting substrate concentration (mg L
1 or g m
3); KS:
half saturation constant (mg L
1 or g m
3).

1.4. Nitrification rate in biofilm as affected by

water quality parameters

The nitrification rate in a biofilm is in equilibrium

between substrate demand created by the growth of
bacteria biomass and the rate of substrate supply
determined by diffusion transport limitation (Ras-
mussen and Lewandowski, 1998). The substrate
demand is determined by the factors that are related
to the characteristics of nitrifiers such as the amount of
Fig. 1. Conceptual profile of a fixed biofilm. nitrifier biomass, the specific growth rate and yield
182 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

coefficient. The substrate supply is determined by the biofilters used in a RAS should be designed based on
diffusion rate of essential nutrients. There are more than ammonia concentration (Wheaton et al., 1994). There
20 physical, chemical and biological factors that can are two fundamental questions related to nitrification
affect the growth or the substrate supply, thus ultimately processes: the minimum concentration that a biofilter
influencing the performance of nitrification biofilters can maintain and the relationship between nitrification
(Wheaton et al., 1994). The most significant factors can rate and TAN concentration. First, ammonia becomes
be classified into three major categories. The first toxic for many fish species at a very low concentration
category includes those that affect the biochemical level. The biofilters must be able to maintain a high
process of the microbes such as pH, temperature and quality of water with sufficiently low ammonia
salinity. The second category includes those that affect concentration. Second, these biofilters must process
the supply of nutrients to the biofilm such as substrate TAN at an adequate rate, at a given level, to prevent
concentration, dissolved oxygen (DO) and mixing TAN accumulation.
regime. The third category includes those that have This highest water quality that can be maintained in
impact upon both growth and nutrient supplies, such as RAS, in terms of ammonia, is defined by a minimum
the competition for either essential nutrients or space, substrate concentration that a biofilter can operate on a
represented by the level of organics. sustainable basis, Smin. The concept of a minimum
Nitrification has been a subject of numerous studies substrate concentration required to support a steady-
for various applications. The results of these research state biofilm was proposed and proven by Rittmann
efforts have advanced the understanding of the and McCarty (1980) and Rittmann and Manem
characteristics of the nitrification process in a biofilm. (1992). Rittmann and McCarty (1980) also mathe-
Techniques used for investigating biofilm reactor matically defined the Smin for a biofilm as:
kinetics have included mainly two types of analyses:
(1) local chemical and physical measurements inside b
Smin ¼ KS (6)
the biofilm using microsensors and (2) chemical mmax
b
analysis of bulk solutions. Microsensors have been
popular for biofilm investigations in recent years where Smin is the minimum substrate concentration
(Lewandowski et al., 1991; de Beer et al., 1996; (mg L
1) and b is the specific bacterial decay rate
Stoodley et al., 1997; Rasmussen and Lewandowski, (day
1).
1998). The microelectrode technique has been applied Zhu and Chen (1999) evaluated the minimum TAN
to DO concentration and its diffusivity measurements concentration for submerged nitrification biofilters in
near and within the biofilm (Lewandowski et al., 1991; a reactor series system and the mean value of the
Rasmussen and Lewandowski, 1998). In aquacultural minimum TAN concentration was found to be
engineering research, most studies were based on 0.07  0.05 mg L
1 at 27.2 8C. Fortunately, the
factors that can be measured in bulk water. minimum TAN concentration was much lower than
The purpose of this paper is to summarize the the toxic level for almost all the aquacultural species.
current knowledge on water quality related design and This result provided the theoretical support for safely
operational factors that affect the nitrification rate of a applying the fixed film nitrification process in RAS.
biofilter used for RAS. The information should The low concentration requirement in aquaculture
provide guidance for selecting and managing these systems determines that in most cases, TAN concentra-
factors to optimize the performance of the biofilter in tion is the rate-limiting factor of biological nitrification.
terms of both nitrification rate and reliability. The relationship between nitrification rate and TAN
concentration can be defined by Eq. (5). However, due to
the fact that aquaculture biofilters operate at lower TAN
2. Substrate concentrations concentrations compared with those used in industry
wastewater treatment, these equations can be simplified
The concentration of TAN as the substrate of for aquaculture applications.
nitrification is the most important factor to consider in At low ammonia substrate concentrations
biofilter design and operation. In general, nitrification (S  KS), nitrification kinetics can be simplified into
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 183

Fig. 2. Performance data for RBCs applied to recirculating aquaculture systems. Small: laboratory scale RBC (0.9 m diameter); pilot: large
laboratory scale RBC (1.8 m diameter); commercial: industrial application (3.7 m diameter) (Ester et al., 1994).

a first-order reaction model: tions and low organic loading rates. Liu and
Capdeville (1994) also detected a linear relationship
Rmax
R¼ ðS
Smin Þ (7) between influent ammonia concentration and ammo-
KS nia removal rate in RBC.
Eq. (7) shows that nitrification rates increase In the study of the nitrification performance of three
linearly with the increase of TAN substrate concen- types of biofilters (Ling and Chen, 2005), the authors
tration. This relationship has been confirmed with investigated substrate impact on nitrification rate at
experimental aquaculture systems. Ester et al. (1994) laboratory scale with chemical feed. The results
studied the performance of three rotating biological indicated that first-order reaction kinetics can also be
contactor (RBC) systems used for RAS where water established at low TAN concentrations (such as
temperature ranged from 24 to 30 8C and observed <1 mg L
1) according to Eq. (7).
first-order nitrification kinetics at low ammonia Table 1 shows the parameters of Monod kinetics
concentrations (Fig. 2). Other researchers (Watanabe and the influence of substrate concentration on
et al., 1980; Surampalli and Baumann, 1989) have also nitrification rates of three different types of biofilters
found that a first-order reaction can be developed for (floating bead, fluidized sand and submerged biocube
RBC reactors at extremely low ammonia concentra- filters). A Smin of 0.07 mg TAN L
1 was used in the

Table 1
Biofilters nitrification kinetic constants and first-order reaction rates at low ammonia concentration (TAN < 1 mg L
1) (Ling and Chen, 2005)
Temperature (8C) Biofilter type Rmax (mg m
2 day
1) KS (mg L
1) R (mg m
2 day
1) Reference
20 Floating bead 5000 8.5 588  S
41
Fluidized sand 3330 5.3 625  S
44
Submerged biocube 1670 5.5 345  S
24
15 Floating bead 5000 9.5 526  S
37
Fluidized sand 3330 6.0 556  S
39
Submerged biocube 1670 6.0 278  S
19
8 Submerged 1550 5.5 Zhu and Chen (2002)
14 Submerged 1690 2 Zhu and Chen (2002)
20 Submerged 1720 2 Zhu and Chen (2002)
27 Submerged 1860 2 Zhu and Chen (2002)
184 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

expression of ammonia removal rate (Eq. (7)). The reported (Wezernak and Gannon, 1967; Sharma and
linear expressions of reaction rate as a function of Ahlert, 1977).
substrate concentration confirm that the bead, the sand The effects of the DO concentration on nitrification
and the biocube filters all have first-order substrate rates had been reviewed by several authors (Sharma
removal kinetics similar to the RBC under low and Ahlert, 1977; Stenstrom and Poduska, 1980;
ammonia concentrations. At 15 and 20 8C, the Painter, 1986; Beccari et al., 1992) in both attached
nitrification rate of biocube filter was much lower and suspended growth systems. Table 2 lists the
(about one-half) than the sand filter and the bead filter. effects of oxygen on nitrification rates in pure culture
It can be speculated that this result is due to the and mixed culture fixed film nitrification processes.
difference in hydraulic conditions and operation With one exception, optimum DO of 4 mg L
1 for
among the different biofilters. The results also suggest maximum nitrification rate in activated sludge
the apparent half saturation constant, KS, varies with (Wuhrman, 1963), most of the investigations con-
respect to operation conditions. sidered the limitation of low DO on nitrification.
It needs to be pointed out that the biofilter reaction Schoberl and Engel (1964) studied DO uptake rates as
rates expressed in Table 1 were obtained with a function of DO concentration in pure culture
synthetic substrate solutions containing inorganic conditions and found that the growth rate of
ammonia and other necessary nutrients other than Nitrosomonas was independent of the DO concentra-
organic matter. However, aquaculture wastewater tion above 1.0 mg L
1, while the growth rate of
usually contains organic material that supports the Nitrobacter was independent above 2.0 mg L
1 of
growth of heterotrophic bacteria which compete with DO (Stenstrom and Poduska, 1980). This result is
nitrifiers for space and oxygen. It will be demonstrated much lower than the observations from mixed culture
later in this paper that the addition of organic matter systems. Knowles et al. (1965) studied nitrification
could decrease the nitrification rate compared to a pure kinetics with mixed culture reactors and reported that
culture condition. DO affected the growth rate of Nitrosomonas very
Substrate inhibitions on nitrification biofilter little at the level above 2 mg L
1, but Nitrobactor was
efficiency were reported with excessively high more sensitive to DO and showed a reduced growth
ammonia or nitrite concentrations (Sharma and rate with DO less than 4 mg L
1. Moreover, Downing
Ahlert, 1977; Carrera et al., 2004). However, it is et al. (1964) reported that nitrification ceases entirely
clear from the above discussion that substrate below a DO concentration of 0.2 mg L
1. Based on
limitation rather than substrate inhibition is often the available information from Table 2, the minimum
the major concern for biofilter designs in RAS due to range of DO concentration required to reliably
the low ammonia concentration in these systems achieve nitrification is between 0.6 and 3.4 mg L
1.
(Wheaton et al., 1994). Within the TAN concentration DO of 2 mg L
1 was suggested as the minimum
range that is common to RAS, the nitrification rate is oxygen level in aquaculture nitrification biofilters
proportional to the substrate concentration. (Wheaton et al., 1994).
However, most of the results in Table 2 are based on
research in pure culture conditions or activated sludge
3. Dissolved oxygen systems. Limited oxygen conditions in a fixed film
biofilter can be significantly different, as oxygen
For RAS design, knowledge of the relationship availability to nitrifying biofilm is subjected to
between nitrification rate and DO concentration is a diffusion limits. Zhu and Chen (2002) reported that
major interest. As demonstrated in Eq. (3), oxygen is a it was more important to maintain sufficient DO in the
requirement in ammonia oxidation. The theoretical fixed film process than in the suspended growth
oxygen requirements according to the nitrification processes due to the nature of diffusion transport with
stoichiometric equations are: 3.43 mg for oxidation of fixed film. DO concentration profiles within biofilms
1 mg NH3–N and 1.14 mg for oxidation of 1 mg NO2– were studied using a microelectrode technique and a
N, although a slightly lower ratio of oxygen consumed microslicing technique with heterotrophic biofilms,
to nitrogen oxidized in an experimental study was also heterotrophic–autotrophic biofilms and nitrifying
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 185

Table 2
Effect of DO on nitrification activity
Organism DO (mg L
1) Observation Conditions Reference
Nitrosomonas 1 Limiting Pure culture Schoberl and Engel (1964) a
2 Limiting Batches Knowles et al. (1965) a,b
Nitrobactor 2 Limiting Pure culture Schoberl and Engel (1964) a
4 Limiting Batches Knowles et al. (1965) a,b
Nitrifiers, 1 Limiting Pilot plant; Metcalf and Eddy (1973) b
general activated sludge
0.6–0.7 Limiting Percolating filter; Forster (1974)b
receiving seawater;
marine nitrifiers
4 Maximum Activated sludge Wuhrman (1963) a
rate of nitrification
1–1.5 Limiting for growth Activated sludge Wuhrman (1968) b
3.4 Limiting (without Circulation bed Nogueira et al. (1998)
pH controlled) reactor (CBR)
5.2 Limiting (with Circulation bed Nogueira et al. (1998)
pH controlled) reactor (CBR)
Up to 60 No inhibition; Submerged filter; Haug and McCarty (1972) b
no increase rate of receiving pre-
ammonia oxidation oxygenated feed
Limiting: rate of nitrification is concentration dependent below this value.
a
Cited in Stenstrom and Poduska (1980).
b
Cited in Sharma and Ahlert (1977).

biofilms (Zhang et al., 1995). An apparent decrease of layer (Fig. 3). According to Fig. 3, DO dropped
DO in the boundary layer as well as the base biofilm rapidly within the biofilm, and was in fact the limiting
layer was detected with all three types of biofilms factor as it was below 2 mg L
1 within the majority of
indicating the existence of diffusion resistance in both the biofilm despite the fact that the bulk solution had a
of the external boundary layer and internal diffusion sufficiently high DO concentration.

Fig. 3. Concentration profiles in the heterotrophic–autotrophic biofilms (Zhang et al., 1995).

186 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

Fig. 3 clearly demonstrates that DO was also a pH and temperature. The impact of these factors
limiting substance in nitrification for the substrate needs to be considered when determining the
level tested. Thus, the double Monod model should be oxygen requirement for the optimum nitrification
used to incorporate the effect of DO into nitrification rate.
kinetics to account for simultaneous ammonia and
oxygen limits (Stenstrom and Poduska, 1980; Beccari
4. Turbulence
et al., 1992; Picioreanu et al., 1997; Hagopian and
Riley, 1998), as shown by: It is critical to recognize that the mass transfer of
S C substrate and nutrients into the biofilm is a limiting
R ¼ Rmax (8) factor to the nitrification rate. Thus, the nutrient mass
KS þ S KC þ C
flux determines the efficiency of a fixed film biofilter.
where C is the concentration of DO in the reactor and Turbulence affects the thickness of the water film and
KC is the half saturation constant in terms of oxygen.
subsequently the transfer resistance of substrate from
The transition from ammonia or nitrite limitation to
bulk liquid into the biofilm. Therefore, turbulence has
DO limitation has also been studied (Gonenc and great impact on the mass transfer flux into biofilm as
Harremoes, 1985; Zhang et al., 1995; Nogueira et al., well as the nitrification rate. Stoodley et al. (1997)
1998). In an investigation on nitrification in rotating investigated the relationship between local mass
disc systems, the transition from ammonia rate transfer coefficients and fluid velocity in heteroge-
limitation to oxygen limitation was found to occur neous biofilms using microelectrodes and confocal
at a bulk DO ratio to bulk ammonia–N of approxi- scanning laser microscopy. They found that the effects
mately 2.5 (Gonenc and Harremoes, 1985). Nogueira
of biofilm heterogeneity on mass transport were
et al. (1998) reported that oxygen limitation occurred
strongly dependant upon the average flow velocity and
when the oxygen to ammonia concentration ratio was that the Sherwood numbers, Reynolds numbers and
below 1.5–2 g O2/(g NH4+) in a circulating bed Schmidt numbers were similar to values reported in
biofilm reactor for both laboratory and pilot scale the mass transfer literature on porous media. de Beer
studies. et al. (1996) measured DO concentration profiles on
Low DO concentrations can also cause nitrite heterogeneous biofilm and found that the thickness of
accumulation in nitrification biofilters. In a study of the mass transfer boundary layer on the film decreased
oxygen concentration effect on nitrite accumulation
exponentially with increasing flow velocity. Zhu and
in an airlift biofilter by mathematical modeling, it
Chen (2001) investigated the relationship between
was reported that nitrite accumulation occurred at a total ammonia nitrogen removal rate and the Reynolds
DO concentrations less than 2 mg L
1 (Picioreanu number (Re) in a steady-state nitrification fixed
et al., 1997). Zhang et al. (1995) detected an biofilm. In their study, Zhu and Chen (2001) followed
increased ammonia removal rate corresponding to the procedure that treated TAN removal rate and DO
the increase of DO of 0.5, 1.0 and 2.0 mg L
1 while consumption of a nitrification fixed biofilm in steady
the amount of nitrate production remained state (Fig. 1) as a two-step process, external mass
unchanged, implying that nitrite oxidizers were
transfer and internal mass diffusion.
seriously inhibited in the low DO environment.
Although there is no significant evidence to (1) External mass transfer: The transport of substrate
determine the optimum concentrations of bulk DO in moving liquid is governed by molecular
for nitrification reactors used in aquaculture sys- diffusion and advection (Lewandowski et al.,
tems, a DO greater than 2.3 mg L
1 within the 1992). The substrate transfer rate to the biofilm
biofilm is necessary for the purpose of preventing interface, due to the combination of these
nitrite accumulation which is a significantly negative processes, can be expressed as (Hamdi, 1995;
factor for the growth of aquaculture species. It also
Chen and Huang, 1996):
needs to be pointed out that oxygen available to
nitrifiers in the biofilter is affected by turbulence at DWS ShDWS
JS ¼ ðSW
SS Þ ¼ ðSW
SS Þ (9)
the surface of the biofilm as well as organic loading, L d
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 187

where JS: TAN transport flux rate or removal rate Reynolds number of the flow had a significant impact
(g m
2 day
1); DWS: TAN diffusion coefficient in on the ammonia removal rate (Fig. 4). Fig. 4 indicates
water (m2 day
1); L: thickness of water film layer that there was no significant difference in the
(m); SW: TAN concentration in bulk water maximum TAN removal rates between Re = 1668
(g m
3); SS: TAN concentration at biofilm surface and 4003. TAN removal rates corresponding to these
(g m
3); Sh: Sherwood number (dimensionless), two Reynolds numbers were much lower than those at
which is defined as the ratio of actual mass flux to higher Reynolds numbers. For a fixed TAN concen-
the rate of mass transfer that would occur if the tration, the TAN removal rate at Re = 66,710 was
same concentration difference were established about five times that at Re = 1668. This demonstrates
across a still water layer with the thickness of that the hydraulic condition of the biofilm surface was
characteristic length (day). a major factor affecting TAN removal rate.
(2) Internal mass diffusion: The mass balance The significant impact of Reynolds number on
equations within a biofilm are expressed below nitrification biofilm kinetics was further demonstrated
(Chen and Huang, 1996). It was assumed that for by Zhu and Chen (2003) in a separate study in which
the experimental system, the main limiting factors comparisons of nitrification performance were made
were TAN and DO concentration rates. The other between the system with the diffusers placed inside the
nutrients were supplied at adequate levels for reactors (D-in) and that with the diffusers placed outside
biofilm growth. the reactors (D-out). Experimental results indicated that
the turbulence caused by air mixing had a significant
d2 S S C impact on nitrification rate in the fixed film biofilters.
DS 2
¼ Vms (10)
dZ KS þ S KC þ C When TAN was not limited, the TAN removal rate in D-
where DS: TAN diffusion coefficient within a out reactors was only about 15% of that in D-in reactors.
biofilm (m2 day
1); S: TAN concentration within The results suggested that increasing turbulent air flow
a biofilm (g m
3); Z: biofilm depth perpendicular could be an effective method to improve the nitrification
to the support medium (m); Vms: maximum spe- efficiency of fixed film biofilters.
cific TAN removal rate (g m
3 day
1); KS: half These results were consistent with other reports.
saturation constant of TAN concentration Hsu et al. (2001) examined the kinetic behaviors of
(g m
3); C: oxygen concentration within a biofilm nitrogen compounds in biofilm channeling under
(g m
3); KC: half saturation constant of oxygen laminar and turbulent flow conditions. They found that
concentration (g m
3). the flow velocity significantly influenced the nitrifica-
tion and denitrification conversion rates. The effect of
Both theoretical analysis and experimental results turbulent flow on biofilm formation in potable water
in Zhu and Chen’s investigation (2001) show that the was studied by Percival et al. (1999). Their experi-

Fig. 4. Nitrification rate obtained with different Reynolds numbers (Re) (Zhu and Chen, 2001).
188 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

ments showed that the viable and total cell counts of with nitrifying bacteria for oxygen and growing space
the biofilm on stainless steel slides were higher at (Ohashi et al., 1995; van Benthum et al., 1997). With
higher flow rates (0.96 and 1.75 m s
1) than at the the addition of organic matter, fast-growing hetero-
lower flow rate (0.32 m s
1). Vieira and Melo (1999) trophic bacteria which use organic carbon as their
reported that biofilm formed in turbulent flow allowed energy source will out compete slow-growing nitrify-
a more stable reactor operation. This is particularly ing bacteria, resulting in a decrease in the nitrification
important for the biofilters applied in the aquaculture rate. It was reported that heterotrophic bacteria have a
industry because TAN concentration in aquaculture maximum growth rate of five times and yields of two
systems is usually very low (Wheaton et al., 1994). to three times that of autotrophic nitrifying bacteria
Therefore, there may be a way to improve the (Grady and Lim, 1980). Zhang et al. (1994, 1995) used
nitrification efficiency of a fixed film biofilter through a microelectrode technique and a microslicing
turbulent flow. technique to study the competition between hetero-
The significance of the effect of turbulence on trophs and autotrophs for substrate and space. It was
nitrification rate has also been demonstrated by other found that an increase of the organic loading rate
researchers. Kugaprasatham et al. (1991) studied the would result in a decrease of DO in the biofilm,
effect of hydraulic conditions on nitrifying biofilm thereafter inhibiting the nitrification rate due to the
grown under a low ammonia nitrogen concentration shortage of oxygen. The competition for substrate in
(about 1 g m
3) in a cylindrical reactor. When biofilm also results in a stratified biofilm structure as
turbulence intensity was changed and kept at the well as non-uniform spatial distributions of bacterial
new value for several days, filamentous-type biofilm populations and metabolically active bacteria. As a
with higher substrate flux was observed at high result, the ammonia removal rate will decrease as
turbulent intensities, but colony-type biofilm under organic loading increases (Pano and Middlebrooks,
low turbulent intensities showed reduced mass 1983). Grady and Lim (1980) observed that nitrifica-
transfer (Kugaprasatham et al., 1991). In addition, tion rates were maximized with a five-day biochem-
Chen and Huang (1996) found that chemical oxygen ical oxygen demand/total Kjeldahl nitrogen (BOD5/
demand (COD) removal flux rose with the increase of TKN) ratio of 0.25 and that nitrification rates were
rotation speed of a submerged RBC. Ling and Chen reduced at higher BOD5/TKN ratios. In studies of
(2005) also reported higher nitrification rates in staged RBC treating carbonaceous BOD loads,
biofilters with high turbulence levels. These results Antonie (1976) found that in the later stages, nitrifying
are important for the design and optimal operation of bacteria began developing once BOD fell below
biofilters, as they suggest that the nitrification rate 30 mg L
1. He suggested that at higher BOD5
may be significantly improved through increasing the concentration, the nitrifying organisms cannot com-
turbulence. pete with the faster growing carbon-oxidizing organ-
isms.
The presence of organics also affects the composi-
5. Organics tion of the microbial population. Ohashi et al. (1995)
found that the proportion of nitrifiers decreased with
The fecal material excreted by fish and uneaten an increasing C/N ratio. An exponential decrease of
feed is organic in nature; thus, the water in a RAS may the nitrification rate with an increased influent COD/N
have a significant amount of organics, both in ratio was also reported in a study on nitrogen removal
dissolved and particulate forms. Considerable from high-strength ammonia industrial wastewater
research related to the organic effect on the nitrifica- (Carrera et al., 2004). Using microelectrodes with high
tion process has been conducted, especially in spatial resolution and fluorescents in situ hybridization
domestic and industrial wastewater treatment. The (FISH), Satoh et al. (2000) evaluated ammonia-
most important impact of organics upon nitrification is oxidizing and heterotrophic bacteria competition for
due to the contribution of additional oxygen demand. oxygen in a biofilm and found that the competition for
Particulate and dissolved organics provide substrates oxygen increased with the increase of C/N ratio
for heterotrophic bacteria whose growth will compete resulting in a decline of ammonia-oxidizing bacteria
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 189

Fig. 5. Relationship between the nitrification rate and the influent COD/N ratio (Ling and Chen, 2005).

in the outer portion of the biofilm. Okabe et al. (1996) N  0.5), the nitrification rate decreased greatly with
discovered that a higher influent C/N ratio retarded the increase of COD/N ratio, but the organic inhibition
accumulation of nitrifying bacteria and resulted in a became less pronounced with an increase in the COD/
considerably longer start-up period for nitrification. It N ratio. As the COD/N ratio was higher than 3 (C/
was also reported that the regulation of nitrification by N  1), the nitrification rates of both systems tend to
organic carbon was not only dependent on the quantity remain unchanged at a minimum value. This indicates
but also the quality of organic carbon, where higher that the inhibition of organic matter on nitrification
quality organic carbon had a stronger negative impact could be maximized and maintained constantly as the
on nitrification (Strauss and Lamberti, 2000). growth of heterotrophic bacteria reached a saturation
In a laboratory study using a chemically fed reactor level with a certain COD/N ratio (COD/N = 3 in this
series system, the authors observed an exponential study). The relationship between a bead filter
decrease of nitrification rate with an increase of COD/ nitrification rate and influent COD/N ratio at 20 8C
N ratio in a study of nitrification rates of a floating can be defined by an exponential function shown as
bead filter, a fluidized sand filter and a submerged Eq. (11), which is similar to Eq. (12) that was
biocube filter. The biofilters were run at 15 and 20 8C developed based on a suspended growth system
with synthetic chemical solutions containing (Carrera et al., 2004).
10 mg L
1 TAN and various COD loadings of C/
N = 0, 0.5 and 2, which is equal to COD/N = 0, 1.4 and R ¼ 0:67 þ 2:27 e
1:38ðCOD=NÞ (11)
5.4 according to a COD/C = 2.68 with sucrose as the
Rnitrification ¼ 0:0323 þ 0:334 e
1:660ðCOD=NÞ (12)
organic carbon source (Zhu and Chen, 2001). The
relationship between COD/N and nitrification rate was Similarly, the relationship between nitrification
very similar to the results obtained from an activated rates and COD/N ratios can be developed for the bead
sludge system (Carrera et al., 2004). Fig. 5 shows the filter at 15 8C and the other two types of filters at both
effect of COD/N ratios on nitrification rates compar- 15 and 20 8C (Table 3).
ing a bead filter at (20 8C) and an activated sludge The results in Table 3 can be used as references in
system (25 8C). As COD/N increased from 0 to 1.5 (C/ the design of nitrification biofilters which are operated

Table 3
Expression of COD/N effect on nitrification rate (g m
2 day
1) (Ling and Chen, 2005)
Biofilter type Temperature (T, 8C)
20 15
Floating bead filter R = 0.67 + 2.27 e
1.38(COD/N) R = 0.85 + 1.88 e
1.36(COD/N)
Fluidized sand filter R = 1.0 + 1.16 e
0.92(COD/N) R = 0.90 + 0.93 e
1.21(COD/N)
Submerged biocube filter R = 0.33 + 0.65 e
1.38(COD/N) R = 0.51 + 0.62 e
2.22(COD/N)
190 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

Fig. 6. Effect of particulate organic matter concentration on the nitrification efficiency in a pilot scale biofilter.

in mixed culture conditions like aquaculture systems. had a volume of 8 L and the water flow was fixed at
The reduction of nitrification rates of the three types of 2 L min
1.
biofilters was about 60–70% under the tested This experiment clearly showed that the nitrifica-
conditions when the COD/N ratio increased from 0 tion efficiency decreased by 40% when the concen-
to 3. In a previous study, Zhu and Chen (2001) also tration of POM increased from 0.2 to 10 mg L
1,
reported a 70% reduction of nitrification rate when the while an additional drop of about 10% was observed at
C/N ratio increased to 1 (COD/N = 2.7) compared to higher concentrations of POM (Fig. 6).
C/N = 0 while the nitrification rate difference between The above summary clearly demonstrates the
C/N = 1 and 2 was not significant. The consistency of impact of organics on the nitrification rate. Therefore,
the two studies again confirmed the inhibition of as the general recommendation, organics should be
organic matter on biofilter nitrification performance. removed from RAS as soon as possible.
Consequently, the effects of COD/N ratio on biofilter
nitrification capacity need to be taken into account in
design and operations of RAS. It also needs to be 6. Temperature
noted that the COD/N versus nitrification rate
relationship reported by Ling and Chen (2005) was It has been well accepted that a higher temperature
obtained for an influent TAN = 10 mg L
1, which is enhances nitrification rate as the biochemical driven
much higher than the TAN concentration in a RAS. bacterial processes accelerate as temperature
Therefore, correction factors should be considered to increases. This is true in a suspended growth system.
compensate for substrate effects on nitrification rates For fixed film filters, however, the effects of
when the equations in Table 3 are applied for lower temperature on nitrification kinetics are also influ-
TAN concentrations. Based on this point, more work enced by other phenomena and parameters (Fdz-
needs to be done relative to the organic effect on Polanco et al., 1994), especially substrate diffusion
nitrification rates at a lower TAN concentration which and transport. A general conclusion on the relationship
is more applicable to aquaculture operations. between nitrification rate and temperature must also
The impact of particulate organic matter (POM) on include the effect of mass transfer and bacteria
nitrification efficiency was also evaluated by the third competition. However, the impacts of temperature
author using a series of two, pilot scale, submerged change on nitrification rate in fixed film biofilters were
filters, connected in a loop in a fresh water poorly understood (Okey and Albertson, 1989). Little
recirculation aquaculture system. These filters, filled information is available to quantify the effects of
with a pre-colonised mineral packing media (Bio- temperature on fixed film nitrification rate (Wheaton
grog), were progressively enriched with POM et al., 1994). Zhu and Chen (2002) studied the impact
(collected in the particle separator of the rearing of temperature on nitrification rate through laboratory
system and processed with in-house methods). Filters experiments, mathematical modeling and sensitivity
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 191

Fig. 7. Relationship between total ammonia nitrogen (TAN) concentration and removal rate at the temperature (T) of 8 and 14 8C (Zhu and Chen,
2002).

analysis. The results demonstrated that: (1) diffusion a more significant effect on nitrification rate in the case
mass transport played an important role in fixed film of TAN limitation than in the case of DO limitation.
nitrification processes and (2) the impact of tempera- The main reason for the reduction of temperature
ture on fixed film nitrification rate was less significant impacts on fixed film nitrification rate was due to the
than predicted by the van’t Hoff–Arrhenius equation. domination of the mass diffusion transport process in a
In other words, the impact of temperature on the fixed film biofilter. The DO saturation concentration in
nitrification rate in a fixed film processes was greatly the bulk water decreases as temperature increases.
reduced due to influence on bacterial growth rate Although the diffusion coefficient itself increases as
compared with its impact in a suspended culture temperature increases, the diffusion process limited
process. When oxygen was limited, the decrease in the mass flux into the fixed biofilm and thus the DO
saturation DO concentration, as temperature can become a limiting factor within the biofilm. The
increased, resulted in a negative temperature impact high sensitivity values for the diffusion coefficient
upon the nitrification rate. imply the importance of diffusion transport in a fixed
More specifically, Zhu and Chen (2002) showed film, which is much greater than the importance of the
that in the case of oxygen limitation, temperatures growth rate coefficient in mixed culture processes
from 14 to 27 8C had no significant impact on (Zhu and Chen, 2002). These conclusions apply to
nitrification rate in fixed biofilm filters (Figs. 7 and 8). systems with natural and mechanical aerations with air
A lower nitrification rate was observed only at the under atmospheric pressure. Use of pure oxygen or
lowest temperature they tested, 8 8C. Temperature had super saturated oxygen will help to overcome oxygen

Fig. 8. Relationship between total ammonia nitrogen (TAN) concentration and removal rate at the temperature (T) of 20 and 27 8C (Zhu and
Chen, 2002).
192 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

transfer limitation and thus enhance the impact of and 7.2 to 9.0 for Nitrobacter. According to Villaverde
temperature on nitrification rate. It is also well accepted et al. (1997), the optimal range of pH for nitrification
that a biofilm establishes much faster in a higher can be determined by the three different effects that the
temperature application than a lower temperature. pH can exert on nitrifying bacteria: (1) activation–
deactivation of nitrifying bacteria; (2) nutritional effect,
connected with alkalinity; (3) inhibition through free
7. pH ammonia and free nitrous acid.
Siegrist and Gujer (1987) studied the pH dependence
A great deal of investigation conducted has of the ammonia volumetric uptake rate (rvNS) with a
demonstrated the pH effects on nitrification. However, trickling filter and found a short term dependence of
poor agreement existed on how much, and at what point, ammonia oxidation can be modeled with Eq. (10),
pH begins to affect nitrification rates (Biesterfeld et al., _
rvNS ¼ rvNs ½1 þ 106:5
pH 
1 ðML
3 T
1 Þ (10)
2001). Based on the review provided by Sharma and
Ahlert (1977) and other researchers’ studies, the where rvNS is the volumetric ammonia uptake rate
_
optimal pH for the growth of nitrifying bacteria varies (ML
3T
1) and rvNs is the maximum volumetric
widely. It can be seen from Table 4 that the optimum pH ammonia uptake rate (ML
3T
1).
for nitrification can range from 7.0 to 9.0 with the The mechanisms by which pH affects nitrification
optimum pH range from 7.2 to 8.8 for Nitrosomonas are debatable among different researchers. Based on

Table 4
pH effect on nitrification activity
Organism pH Degree of inhibition Circumstance Reference
Nitrosomonas 8.5–8.8 Optimum Pure culture Metcalf and Eddy (1973)a
7.9 or 9.3 50%
8.0–8.5 Optimum Pure culture Buswell et al. (1954)a
8.3–8.6 Optimum Pure culture Lees (1954)a
7.2–7.9 Optimum Pure culture Engel and Alexander (1958)a
7.5–8.0 Optimum Pure culture, Loveless and Painter (1968)a
Nitrosomonas
isolated from
activated sludge
Nitrobactor 7.2–8.2 Optimum Pure culture Lees (1954)a
7.8 Optimum Boon and Laudelout (1962)b
9 Optimum Kawai et al. (1965)b
Nitrifiers, 7.0–8.0 Optimum Mixed culture in Haug and McCarty (1972) a
general submerged filter
8.2 Optimum Mixed culture in Villaverde et al. (1997)
submerged filter
8.4–8.5 Optimum Two-stage, activated- Rimer and Woodward (1972) a
sludge pilot plant
8.0–8.8 Optimum Batch activated sludge; Metcalf and Eddy (1973)a
laboratory scale
7.1 or 9.8 50%
7.8 Optimum Activated sludge Antoniou et al. (1990)
8 Optimum Percolating filter; laboratory scale; Forster (1974)a
mixed population; marine nitrifiers
5.9 50%
7.45 Optimum Mixed culture; marine nitrifying filter Srna and Baggaley (1975) a
a
Cited in Sharma and Ahlert (1977).
b
Cited in Wheaton et al. (1994).
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 193

the ability of free ammonia (NH3) and free nitrous acid

(HNO2) to penetrate the nitrifying organism, Antho-
nisen (1974) concluded that both NH3 and HNO2 are
more inhibitory to nitrifying bacteria than ammonium
and nitrite ions. However, Suzuki et al. (1974) and
Painter (1986) suggested that free ammonia instead of
ammonium is the substrate for ammonia-oxidizing
bacteria (Nitrosomonas) based on the observation of a
consistent Monod saturation constant under varying
free ammonia concentration. Therefore, reduced
nitrification activity at lower pH levels may result
indirectly from substrate limitation since the fraction
of NH3–N in total ammonia nitrogen decreases with
the decrease of pH (Allison and Prosser, 1993). The
variation in reported optimum pH ranges for
nitrification may therefore be due in part to variations Fig. 9. pH profiles in a nitrifying biofilm of thickness of 320 mm
in the TAN used. This is because when a higher TAN is (Zhang and Bishop, 1996).
used, a higher non-limiting concentration of NH3 may
be maintained at lower pH values (Biesterfeld et al.,
2001). Interestingly, to evaluate the pH effect on ratio of HCO3
/O2 (Fig. 9). With a mole ratio >5, the
ammonia oxidization activity, Groeneweg et al. (1994) maximum pH decrease within the nitrifying biofilm
measured ammonia oxidization rates at a constant was <0.4–0.6; when this ratio <3, a pH decrease of
NH3–N of 0.37 mg L
1 (varying TAN in accordance 1.4–1.6 could occurr within the biofilm. Substantial
with pH) and a constant TAN of 5 mg L
1 (NH3–N changes in the pH across the boundary layer and the
varies with pH) over a wide pH range (5–11). They biofilm were also predicted by using a steady-state
found that the maximum ammonia oxidization rate complexity nitrifying biofilm model and considering
was obtained between pH 6.7 and 7.0 (0.37 mg L
1 both pH effects and ion interactions (Flora et al.,
NH3–N) and pH 7.5 and 8.0 (5 mg L
1 TAN), while 1999). Consequently, Flora et al. (1999) mentioned
the ammonia oxidization rate decreased sharply that nitrification rates may be improved by increasing
outside the optimum pH ranges. The variation of bulk pH to high (alkaline) values, which results in
ammonia oxidization rate at a constant NH3–N optimum (neutral) pH within the biofilm. This is very
concentration and various pHs highlights the close significant for the operation of nitrification biofilters in
dependence of nitrification process on pH. aquaculture systems. Although a low pH is optimal in
Hydrogen ion toxicity at pH values of 5.7 and the culture tank to minimize the toxicity of the portion
below has also been proposed as a possible mechanism of unionized ammonia regardless of the optimal pH
for nitrification inhibition (Szwerinski et al., 1986). for the growth of aquaculture species, a pH higher than
This can also help to explain the observed reduction in the optimum in the biofilter is desirable for the
the nitrification rate when pH suddenly changed from improvement of the biofilters nitrification efficiency.
7.2 to 5.8, but not when it changed from 7.2 to 6.4 in an
activated sludge system (Shammas, 1986).
An inherent assumption in most of biofilm models 8. Alkalinity
was that the pH remains constant within the biofilm.
However, Flora et al. (1999) reported a difference The conversion of ammonia to nitrate consumes
between the pH in the bulk solution and in the biofilm alkalinity as shown in Eq. (3). In fact, alkalinity in the
by an experimental study. Zhang and Bishop (1996) form of carbonate and bicarbonate is a nutrient
observed that the pH decreased within a nitrifying element for nitrifying bacteria. In addition, alkalinity
biofilm of thickness 278 mm. They also found that the provides the buffering capacity that is necessary to
pH decrease within the biofilm depended on the mole prevent pH changes due to acid production in the
194 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

nitrification process. Therefore, the impact of alkali- (16 L/h) at a constant temperature (18 8C). The packing
nity on the nitrification rate is also related to that of the was conditioned and nitrification reached a steady state
pH. In a study on the pH effect upon the nitrification prior to experimentation. The average removal rate
efficiency in an up flow biofilter, it was reported that calculated on the values obtained when the TAN
the nitrification efficiency showed a linear increase of concentrations were between 0.4 and 6 mg L
1 was
13% per unit pH increase from pH 5.0 to 8.5 20 mg L
1 day
1 in salt water against 31.7 mg L
1
(Villaverde et al., 1997). The same authors also day
1 in fresh water, or 37% lower in the salt water. This
investigated the relationship between pH and alkali- ratio between salt and fresh water nitrification rates is
nity. They observed a linear correlation between the consistent with those commonly used by engineering
alkalinity (as mg CaCO3 L
1) and pH, with a companies and with values obtained from pilot scale
stoichiometry coefficient of 7.1 mg CaCO3 con- long term experiments with fresh and marine water
sumed/mg NH4+–N oxidized. Chen et al. (1989) recirculation fish production systems.
showed that the rate of nitrification would be reduced There is an agreement among researchers and
when alkalinity was below 40 g m
3. Gujer and Boller between laboratory research and commercial applica-
(1986) reported that in nitrifying biofilters used in tions in that salt water systems need a much longer
municipal wastewater treatment, an alkalinity level of start-up period in which nitrite accumulates. Adapta-
at least 75 mg L
1 (g m
3, or 1.5 meq/L) was needed tion of a fresh water biofilter to higher salinities was
to maintain maximum nitrification rate. Considering found to be possible, thus providing an effective tool
possible stratification of alkalinity and pH in a biofilm, for a shorter start-up period of a salt water system.
a higher alkalinity concentration of 200 mg L
1 is Nitrification rates were obtained which allow a
recommended here, especially for the applications rational design of intensive recirculating salt water
where the water exchange rate is minimal. fish culture systems (Nijhof and Bovendeur, 1990).

9. Salinity 10. Summary

Relatively limited information is available con- The most important feature of nitrification kinetics
cerning the effect of salinity on nitrification kinetics. of the biofilm in a biofilter is the fact that it is a diffusion-
There are discrepancies in the reports, probably due to controlled process. Many water quality factors affect
different experimental conditions. Nijhof and Boven- the nitrification rate of a biofilm, suggesting that there
deur (1990) compared the nitrification characteristics are various options for improving commercial biofilter
of salt water RAS with that of fresh water systems. The performance to approach the removal rate obtained
results indicated that the maximum nitrification under laboratory conditions. Some of these factors are
capacity in the salt water systems was considerably less flexible as they are related to the requirements of the
lower than in fresh water systems. At 24 8C, a species cultured. Other parameters can be varied during
maximum ammonia removal rate of 0.28 g m
2 day
1
NH4+–N was observed versus 0.69 g m
2 day
1 Table 5
NH4+–N in comparable fresh water systems. In a Ranking and recommendation of water quality parameters for
maximizing nitrification rate of a biofilter
separate laboratory study, however, Saucier (1999)
Parameters Ranking Recommendation
was able to obtain a sufficient nitrification rate that is
comparable with that reported in fresh water systems Turbulence/mixing 1 Adequate
Alkalinity 2 >200 mg L
1
under similar conditions (Zhu and Chen, 2002). In a
as CaCO3
recent study by the third author of this paper, a very pH 3 7.5–9
simple experiment was carried out with two similar Oxygen concentration 4 DO/TAN > 2.5
submerged biofilters (5 cm diameter and 1.5 m high) Organic matters 5 Minimize
filled with microporous mineral packing (Biogrog). Substrate concentration 6 Safe range
Temperature 7 Optimal
Each of the biofilters was connected to a peristaltic
Salinity 8 Optimal
pump, which circulated either fresh water or salt water
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 195

design and operation of RAS. Table 5 serves as a Downing, A.L., Painter, H.A., Knowles, G., 1964. Nitrification in
the activated sludge process. J. Inst. Sew. Purif. 63, 130–153.
summary that ranks the design and operation para-
Drtil, M., Németh, P., Bodı́k, I., 1993. Kinetic constants of nitrifica-
meters based on their importance on nitrification tion. Water Res. 27, 35–39.
kinetics and flexibility. The recommendations were Ester, C.C., Novak, J.T., Libey, G.S., Boardman, G.R., 1994.
made in a general sense and they should be re-evaluated Rotating Biological Contactor Performance in Recirculating
for a specific application. Aquaculture Systems. (cited in Wheaton et al., 1994).
Fdz-Polanco, F., Villaverde, S., Garcı́a, P.A., 1994. Temperature
effect on nitrifying bacteria activity in biofilters: activation
Acknowledgments and free ammonia inhibition. Water Sci. Technol. 30, 121–
130.
Fitch, M.W., Pearson, N., Richards, G., Burken, J.G., 1998. Biolo-
This paper includes results obtained from research
gical fixed-film systems. Water Environ. Res. 70, 495–518.
projects funded by the Western Regional Aquaculture Flora, E.M.C.V., Suidan, M.T., Flora, J.R.V., Kim, B.J., 1999.
Center, US Department of Agriculture (USDA) and Speciation and chemical interactions in nitrifying biofilms. I:
the USDA Special aquaculture grant program. model development. J. Environ. Eng. 9, 871–877.
Gonenc, I.E., Harremoes, P., 1985. Nitrification in rotating disc
systems. I: criteria for transition from oxygen to ammonia rate
References limitation. Water Res. 19, 1119–1127.
Grady, C.P.L., Lim, H.C., 1980. Biological Wastewater Treatment:
Allison, S.M., Prosser, J.I., 1993. Ammonia oxidation at low pH by Theory and Applications. Marcel Dekker, New York.
attached populations of nitrifying bacteria. Soil Biol. Biochem. Groeneweg, J., Sellner, B., Tappe, W., 1994. Ammonia oxidation in
25, 935–941. Nitrosomonas at NH3 concentrations near Km: effects of pH and
Anthonisen, A.C., 1974. The effects of free ammonia and free temperature. Water Res. 28, 2561–2566.
nitrous acid on the nitrification process. Ph.D. Thesis. Cornell Gujer, W., Boller, M., 1986. Design of a nitrifying tertiary trickling
University, Ithaca, NY. filter based on theoretical concepts. Water Res. 20, 1353–1362.
Antonie, R.L., 1976. Fixed Biological Surface-Wastewater Treat- Hamdi, M., 1995. Biofilm thickness effect on the diffusion limitation
ment: The Rotating Biological Contactor. CRC Press Inc., in the bioprocess reaction: biofloc critical diameter significance.
Cleveland, OH. Bioprocess Eng. 12, 193–197.
Antoniou, P., Hamilton, J., Koopman, B., Jain, R., Holloway, B., Hagopian, D.S., Riley, J.G., 1998. A closer look at the bacteriology
Lyberatos, G., Svoronos, S.A., 1990. Effect of temperature and of nitrification. Aquacult. Eng. 18, 223–244.
pH on the effective maximum specific growth rate of nitrifying Haug, R.T., McCarty, P.L., 1972. Nitrification with submerged
bacteria. Water Res. 24, 97–101. filters. J. Water Pollut. Control Fed. 44, 2086–2102.
Beccari, M., Di Pinto, A.C., Ramadori, R., Tomei, M.C., 1992. Horn, H., 1994. Dynamics of a nitrifying bacteria population in a
Effects of dissolved oxygen and diffusion resistances on nitri- biofilm controlled by an oxygen microelectrode. Water Sci.
fication kinetics. Water Res. 26, 1099–1104. Technol. 29, 69–76.
Biesterfeld, S., Farmer, G., Russell, P., Figueroa, L., 2001. Effect of Hsu, C.L., Ouyang, C.F., Weng, H.T., 2001. Nitrogen removal from
alkalinity type and concentration on nitrifying biofilm activity. activated sludge process (ASP) treated municipal wastewater by
In: Proceedings of Water Environment Federation Conference biofilm channeling. J. Chin. Inst. Eng. 24, 595–600.
2001, Atlanta, GA, October 12–17. Huguenin, J.E., Colt, J., 1989. Design and Operating Guide for
Boller, M., Gujer, W., Tschui, M., 1994. Parameters affecting Aquaculture Seawater Systems. Developments in Aquaculture
nitrifying biofilm reactors. Water Sci. Technol. 29, 1–11. and Fisheries Science, vol. 20. Elsevier Science Publishers B.V.,
Carrera, J., Vicent, T., Lafuente, J., 2004. Effect of influent COD/N Amsterdam, The Netherlands.
ratio on biological nitrogen removal (BNR) from high-strength Knowles, G., Downing, A.L., Barrett, M.J., 1965. Determination of
ammonia industrial wastewater. Process Biochem. 39, 2035–2041. kinetic constants for nitrifying bacteria in mixed culture with the
Chen, G.H., Ozaki, H., Terashima, Y., 1989. Modelling of the aid of an electronic computer. J. Gen. Microbiol. 38, 263–278.
simultaneous removal of organic substances and nitrogen in a Kugaprasatham, S., Nagaoka, H., Ohgaki, S., 1991. Effect of short-
biofilm. Water Sci. Technol. 21, 791–804. term and long-term changes in hydraulic conditions on nitrifying
Chen, G.H., Huang, J.C., 1996. Determination of diffusion layer biofilm. Water Sci. Technol. 23, 1487–1494.
thickness on a biofilm. J. Environ. Sci. Health 31, 367–386. Lewandowski, Z., Altobelli, S.A., Majors, P.D., Fukushima, E.,
Colt, J., Armstrong, D.A., 1981. Nitrogen toxicity to crustaceans, 1992. NMR imaging of hydrodynamics near microbially colo-
fish, and mollusks. In: Allen, L.J., Kinney, E.C. (Eds.), Pro- nized surfaces. Water Sci. Technol. 26, 577–584.
ceedings of the Bio-Engineering Symposium for Fish Culture. Lewandowski, Z., Walser, G., Characklis, G., 1991. Reaction
American Fisheries Society, pp. 34–47. kinetics in biofilms. Biotechnol. Bioeng. 38, 877–882.
de Beer, D., Stoodley, P., Lewandowski, Z., 1996. Liquid flow and Ling, J., Chen, S., 2005. Impact of organic carbon on nitrification
mass transport in heterogeneous biofilms. Water Res. 30 (11), performance of different types of biofilters. Aqua Eng. 33, 150–
2761–2765. 162.
196 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197

Liu, Y., Capdeville, B., 1994. Kinetic behaviors of nitrifying biofilm Biological Systems Engineering, Washington State University,
growth in wastewater nitrification process. Environ. Technol. 15, Pullman, WA.
1001–1013. Schoberl, P., Engel, H., 1964. Dar verhalten der nitrification kakter-
Losordo, T.M., Westers, H., 1994. System carrying capacity and ien gegenuber gelostem. Archiv. Fin Mikrobiolgie , pp. 393–400
flow estimation. In: Timmons, M.B., Losordo, T.M. (Eds.), A- (cited in Stenstrom and Poduska, 1980).
quaculture Water Reuse Systems: Engineering Design and Shammas, N.K., 1986. Interactions of temperature, pH, and biomass
Management. Elsevier, pp. 9–60. on the nitrification process. J. Water Pollut. Control Fed. 58, 52–
Meade, J.W., 1985. Allowable ammonia for fish culture. Prog. Fish- 59.
Cult. 47, 135–145. Sharma, B., Ahlert, R.C., 1977. Nitrification and nitrogen removal.
Moreau, M., Liu, Y., Capdeville, B., Audic, J.M., Calvez, L., 1994. Water Res. 11897–11925.
Kinetic behavior of heterotrophic and autotrophic biofilms in Siegrist, H., Gujer, W., 1987. Demonstration of mass transfer and pH
wastewater treatment processes. Water Sci. Technol. 29, 385– effects in a nitrifying biofilm. Water Res. 21, 1481–1487.
391. Srna, R.F., Baggaley, A., 1975. Kinetic response of perturbed marine
Nijhof, M., Bovendeur, J., 1990. Fixed film nitrification character- nitrification systems. J. Water Pollut. Control Fed. 47, 472–486.
istics in sea-water recirculation fish culture systems. Aquacul- Stenstrom, M., Poduska, R., 1980. The effect of dissolved oxygen
ture 87, 133–143. concentration on nitrification. Water Res. 14, 643–649.
Nogueira, R., Lazarova, V., Manem, J., Melo, L.F., 1998. Influence Stoodley, P., Yang, S., Lappin-Scott, H., Lewandowski, Z., 1997.
of dissolved oxygen on the nitrification kinetics in a circulating Relationship between mass transfer coefficient and liquid flow
bed biofilm reactor. Bioprocess Eng. 19, 441. velocity in heterogenous biofilms using microelectrodes and
Ohashi, A., Viraj de Silva, D.G., Mobarry, B., Manem, J.A., Stahl, confocal microscopy. Biotechnol. Bioeng. 56, 681–688.
D.A., Rittmann, B.E., 1995. Influence of substrate C/N ratio on Strauss, E.A., Lamberti, G.A., 2000. Regulation of nitrification in
the structure of multi-species biofilms consisting of nitrifiers and aquatic sediments by organic carbon. Limnol. Oceanogr. 45,
heterotrophs. Water Sci. Technol. 32, 75–84. 1854–1859.
Okabe, S., Oozawa, Y., Hirata, K., Watanabe, Y., 1996. Relationship Surampalli, R., Baumann, E.R., 1989. Supplemental aeration
between population dynamics of nitrifiers in biofilms and reactor enhanced nitrification in a secondary RBC plant. J. Water Pollut.
performance at various C:N ratios. Water Res. 30, 1563–1572. Control Fed. 61, 200–207.
Okey, R.W., Albertson, O.E., 1989. Evidence for oxygen-limiting Suzuki, I., Dular, U., Kwok, S.C., 1974. Ammonia or ammonium
conditions during tertiary fixed-film nitrification. J. Water Pollut. ions as substrate for oxidation by Nitrosomonas europaea cells
Control Fed. 61, 510–519. and extracts. J. Bacteriol. 120, 556–558.
Pano, A., Middlebrooks, E.J., 1983. Kinetics of carbon and ammonia Szwerinski, H., Arvin, E., Harremoës, P., 1986. pH-decrease in
nitrogen removal in RBC’s. J. Water Pollut. Control Fed. 55, nitrifying biofilms. Water Res. 20, 971–976.
956–965. Tanaka, H., Dunn, I.J., 1982. Kinetics of biofilm nitrification.
Painter, H.A., 1986. Nitrification in the treatment of sewage and Biotechnol. Bioeng. 24, 669–689.
wastewaters. In: Prosser, J.I. (Ed.), Nitrification. IRL Press, Trussell, R.P., 1972. The percent un-ionized ammonia in aqueous
Oxford. ammonia solutions at different pH levels and temperatures. J.
Percival, S.L., Knapp, J.S., Wales, D.S., Edyvean, R.G.J., 1999. The Fish. Res. Board Can. 29, 1505–1507.
effect of turbulent flow and surface roughness on biofilm for- United States Environmental Protection Agency (US-EPA), 1984.
mation in drinking water. J. Ind. Microbiol. Biotechnol. 22 (3), Methods for the Chemical Analysis of Water and Wastewater,
152–159. EPA-600/4-79-020. U.S. Environmental Protection Agency,
Picioreanu, C., Loosdrecht, M.C.M., Heijnen, J.J., 1997. Modeling Office of Research and Development, Environmental Monitor-
the effect of oxygen concentration on nitrite accumulation in a ing and Support Laboratory, Cincinnati, OH.
biofilm airlift suspension reactor. Water Sci. Technol. 36, 147– van Benthum, W.A.J., van Loosdrecht, M.D.M., Heijnen, J.J., 1997.
156. Control of heterotrophic layer formation on nitrifying biofilms in
Rasmussen, K., Lewandowski, Z., 1998. Microelectrode measure- a biofilm airlift suspension reactor. Biotechnol. Bioeng. 53, 397–
ments of local mass transport rates in heterogeneous biofilms. 405.
Biotechnol. Bioeng. 59, 302–309. Villaverde, S., Garcia-encina, P.A., Fdz-Polanco, F., 1997. Influence
Rittmann, B.E., McCarty, P.L., 1980. Model of steady-state-biofilm of pH over nitrifying biofilm activity in submerged biofilters.
kinetics. Biotechnol. Bioeng. 22, 2343–2357. Water Res. 31, 1180–1186.
Rittmann, B.E., Manem, J.A., 1992. Development and experimental Vieira, M.J., Melo, L.F., 1999. Intrinsic kinetics of biofilms formed
evaluation of a steady-state, multispecies biofilm model. Bio- under turbulent flow and low substrate concentration. Biopro-
technol. Bioeng. 39, 914–922. cess Eng. 20 (4), 369–375.
Satoh, H., Okabe, S., Norimatsu, N., Watanabe, Y., 2000. Signifi- Watanabe, Y., Ishiguro, M., Nishido, M., 1980. Nitrification kinetics
cance of substrate C/N ratio on structure and activity of nitrify- in a rotating biological disc reactor. Water Technol. 12, 233–251.
ing biofilms determined by in situ hybridization and the use of Wezernak, C.T., Gannon, J.J., 1967. Oxygen–nitrogen relationship
microelectrodes. Water Sci. Technol. 41 (4–5), 317–321. in autotrophic nitrification. Appl. Microbiol. 15, 1211–1215.
Saucier, B., 1999. Nitrification in recirculating systems for wet Wheaton, F.W., Hochheimer, J.N., Kaiser, G.E., Krones, M.J., Libey,
storage of marine shellfish. Master’s Thesis. Department of G.S., Easter, C.C., 1994. Nitrification principles. In: Timmons,
 S. Chen et al. / Aquacultural Engineering 34 (2006) 179–197 197

M.B., Losordo, T.M. (Eds.), Aquaculture Water Reuse Systems: Zhang, T.C., Bishop, P.L., 1996. Evaluation of substrate and pH
Engineering Design and Management. Elsevier, Amsterdam, effects in a nitrifying biofilm. Water Environ. Res. 68, 1107–
pp. 101–126. 1115.
WPCF, 1983. Nutrient Control, Manual of Practice. Publication nu- Zhu, S., Chen, S., 1999. An experimental study on nitrification
mber FD-7 Water Pollution Control Federation, Washington, DC. biofilm performances using a series reactor system. Aquacult.
Wuhrman, K., 1963. Effect of oxygen tension on bio-chemical Eng. 20, 245–259.
reactions in sewage purification plants. In: Proceedings of Zhu, S., Chen, S., 2001. Impacts of Reynolds number on nitrification
the 3rd Manhattan Conference on Advances in Biological Waste biofilm kinetics. Aquacult. Eng. 24, 213–229.
Treatment. McMillan, New York. Zhu, S., Chen, S., 2002. The impact of temperature on nitrification
Zhang, T.C., Fu, Y.C., Bishop, P.L., 1994. Competition in biofilms. rate in fixed film biofilters. Aquacult. Eng. 26, 221–237.
Water Sci. Technol. 29, 263–270. Zhu, S., Chen, S., 2003. Effects of air-diffusion turbulent flow on
Zhang, T.C., Fu, Y.C., Bishop, P.L., 1995. Competition for substrate nitrification rate in fixed film biofilters: a comparison study. N.
and space in biofilms. Water Environ. Res. 67, 992–1003. Am. J. Aquacult. 65, 240–247.