G n

0.005 Mc5h7no2 0.005 -113 g mol-1 C5H7O2N Mno2-n 14 g mol-1 NO2-N

g MLVSS

The total yield values without accumulation of NO- for the growth of Nitrosomonas and Nitrobacter are 0.187 g MLVSS per g NH4-N oxidized or g NO3-N produced. Lindemann (2002) and Choi (2005) presented and compared some yield coefficients. Averaged values of YXA/NH4-N and YXA/NO2-N were calculated from values of different authors, as follows (Larsen-Vefring 1993):

Yxa/n02-n = 0.02 to 0.084 = 0.048 g MLVSS (g NO2-N)-1

1 The total parameter MLVSS (mixed liquor volutile suspended solids) includes here only the mass of Nitrosomonas or Nitrobacter, respectively.

The influence of the decay rate (death and endogenous respiration) was not considered. The yield coefficients for the growth of nitrifyers with respect to oxygen consumption are calculated as follows:

YXa/nh n 0.147 g MLVSS = Ixa/nH4-n =-= 0.043 ---(10.20)

YXA/NO2-N 0.04 g MLVSS

As the yield coefficients show, nitrification is characterized by high oxygen consumption and low biomass production. From Eqs. (10.20) and (10.21) it can be seen that almost the same amounts of oxygen are used for the cell multiplication of Nitrosomonas and Nitrobacter.

Ammonia and nitric acid are believed to be the real electron donor (substrate) of Nitrosomonas and Nitrobacter, respectively, because less energy is required for its transport into the cell compared to the transport of an ionised molecule like NH+ or NO- (Suzuki et al. 1974; Bergeron 1978; Wiesmann 1994). NH3 and HNO2 are formed by dissociation which can be described based on a dissociation equilibrium depending on pH and temperature:

The concentration of NH3 and NH+ can be expressed via the dissociation constant KD,NH3 = k2/k1 from Eq. (10.22) as follows:

with:

Introduction of Eq. (10.25) into Eq. (10.24) gives:

S = Snh4-n^ (10.27) SNH3-N= 1 + Kd,nh3 '10-ph where:

Note, that SNH+_N>2 is approximately the same as SNH+-N for 6.0 < pH < 7.8, the pH range at which wastewater is usually treated. Finally, this results in (Anthonisen et al. 1976; Wiesmann 1994):

6344

The concentration of HNO2-N is described using a similar calculation method:

with:

SHNO2_N

Fig. 10.3 Influence of the temperature and pH value on the dissociation equilibrium of NH3 and HNO2.

The dissociation equilibria of NH3/NH+ + SNH4 and HNO2/NO- + SHNO2 with regard to the influence of temperature and pH are presented in Fig. 10.3.

It is very important to recognize that SNH3 increases and SHNO2 decreases with increasing pH.

A kinetic description of nitrification is proposed based on Haldane kinetics. Both equations are valid for ammonium and nitrite-rich wastewater where both ammonium and nitrite oxidations are inhibited by substrate surplus (see Eq. 10.32 and Eq. 10.33).

KiH,NS

KiH,NB

For higher values of Snh3-n (higher pH) or Shno2-n (lower pH) the reactions are inhibited.

For lower values of Snh3-n or Shno^n, e.g. in municipal wastewater treatment plants, the inhibition according to Haldane kinetics can be neglected. Oxygen limitation can be disregarded for c 'P K'. According to these assumptions, Eqs. (10.32) and (10.33) result in simplified kinetic descriptions which are used for nitrification in the WWTP loaded with a low ammonia and nitrite concentration, respectively:

SNH 3-N

Table t0.3 presents the kinetic and yield coefficients of nitrification.

Usually, ammonium oxidation to nitrite is regarded as the bottleneck of nitrification to nitrate. However at low pH, low c' and low temperature, the oxidation rate of NO- is considerably lower than that of NH+. NO- accumulation can be observed (see Section 10.2.4). It is beneficial if NH+ is oxidized only to NO-, which is subsequently denitrified in biological nitrogen-removal systems. Nitrogen removal via the nitrite pathway is also an environmentally cleaner process which reduces the cost of aeration and carbon sources (e.g. methanol as an electron donor). Moreover, it has been reported that denitrification rates with nitrite are 1.5-2.0 times faster than with nitrate (Abeling and Seyfried 1992). The concept of nitrogen removal via nitrite accumulation will be explained in Section 10.2.4 in detail.

If CO2 is added as the carbon source in effluents with low concentrations of organics resulting in low CO2 formation, its concentration may be a rate-limiting factor, especially if high NH+ concentrations are to be oxidized in higher pH regions (Green et al. 2002; Carrera et al. 2003).

10.2 Biological Nitrogen Removal | 235 Table 10.3 Kinetic and yield coefficients of autotrophic nitrification.

NH4 oxidation

Knowles et al. (1965)

30

0.0822

0.084

-

-

-

Bergeron (1978)

25

0.0064

0.138

35

1.8

-

Nyhius (1985)

15-17

0.04

0.056

33

0.5

-

Dombrowski (1991)

20

0.0138

0.714

540

0.29

-

Wiesmann (1994)

20

0.032

0.028

540

0.3

0.147

Horn and Hempel (1996)

20-22

0.0063

0.5

-

0.5

0.062

Pirsing (1996)

25

0.038

0.03

200

0.3

0.142

Lindemann (2002)

22.5

0.0074

0.079

16.5

0.25

0.142

NO2 oxidation

Knowles et al. (1965)

30

0.058

1.9 • 10-4

-

-

-

Bergeron (1978)

25

0.005

2.5 • 10-4

35

1.4

-

Nyhius (1985)

15-17

0.016

1.7 • 10-4

0.15

0.75

-

Dombrowski (1991)

25

0.019

0.39 • 10-4

0.25

1.1

-

Wiesmann (1994)

20

0.045

0.32 • 10-4

0.26

1.1

0.042

Okabe et al. (1995)

20

0.034

0.94 • 10-4

-

0.68

0.083

Pirsing (1996)

25

0.041

0.55 • 10-4

0.1

1.3

0.048

Lindemann (2002)

22.5

0.019

3.0 • 10-4

0.26

1.27

0.048

a) For NH4 oxidation: g MLVSS (g NH4-N)-1; for NO2 oxidation: g MLVSS (g NO2-N)-1.

a) For NH4 oxidation: g MLVSS (g NH4-N)-1; for NO2 oxidation: g MLVSS (g NO2-N)-1.

10.2.2.3 Parameters Influencing Nitrification

There are several parameters which influence the ability of a population of nitrifying bacteria to perform nitrification, such as c', pH, T, tR and t^. Of all these parameters, c' and pH are the most important.

Nitrifying bacteria are strict aerobes. The nitrification rate is limited entirely if oxygen is not supplied. Equations (10.32) and (10.33) show the influence of oxygen on the nitrification rate. For example, the region of oxygen limitation can be estimated using KNb = 1.1 mg L-1 O2 (see Table 10.3). The point of limitation may be given as p = 0.9 pmax (90% of maximal growth rate) which is already reached at c' = 9.9 mg L-1 O2 (see also Eq. 6.11). This means that there is always a limiting effect of the oxygen concentration on the nitrification rate when aerating with air.

For effective nitrification, the amount of c' maintained in the aeration tank should be monitored as a control parameter to ensure permanent effluent concentrations for NH+, NO- and NO-. The practice of over-aeration is expensive and can even contribute to shearing of nitrifying bacterial flocs and/or enhance foam production.

A relationship between growth rate and pH was given by Eqs. (10.28) and (10.32) for ammonium oxidation and by Eqs. (10.30) and (10.33) for nitrite oxidation. The optimum pH for the growth of nitrifying bacteria is generally assumed to be pH 7.2-8.0, depending on SNH4 (see Eq. 10.28). If the pH of the aeration tank drops below pH 5.5 or goes above pH 9.0, a significant decrease in nitrification occurs as a result of protein damage. A low wastewater pH has the primary effect of inhibiting nitrifiers' enzymatic activity and has a secondary effect on the availability of alkalinity.

A drop in temperature results in a remarkable reduction in the growth rate of nitrifying bacteria. Some authors (Hopwood and Downing 1965; Knowles et al. 1965; Painter and Loveless 1983) described the temperature dependence of nitrification. To describe the influence of temperature on nitrification as well as denitrifi-cation, we use the Arrhenius equation for biochemical reactions (see Eq. 3.1).

The temperature dependence of the maximum growth rate during nitrification was already published by Knowles et al. (1965):

Pmax.Ns = 0.042 exp(0.0351T-2.174) Pmax.NB = 0.042 exp (0.0587T-1.13)

The nitrification rate is a function of temperature between 8 °C and 30 °C. Low wastewater temperatures in winter negatively affect the nitrification. Therefore, many regulatory agencies in temperate regions have different ammonia discharge limits according to the season.

Figure 10.4 shows the optimal range of nitrification with respect to the growth rate of nitrifying bacteria in relation to pH and temperature (Larsen-Vefring 1993).

Excursions to low temperatures, temporary and long-term drops in c' and/or extreme pH values lead to incomplete nitrification which results in operational disruptions.

Fig. 10.4 Specific growth rate of nitrifying bacteria in relation to pH and temperature (Larsen-Vefring 1993, calculated).
Table 10.4 Operational parameters influencing nitrification.

Parameters

Optimal range/value and comments

c'

2-3 mg L-1 O2, c' limits nitrification

PH

pH 7.2-8.0, pH <5.5 and >9.0 critical

Temperature

T = 28-32 °C, T <5 °C and >40 °C critical

SNH3

Inhibits Nitrosomonas >10 mg L-1, Nitrobacter >0.1 mg L-1

SHNO2

Inhibits Nitrosomonas and Nitrobacter >1.0 mg L-1

SNH4

inhibits nitrification > 400-500 mg L-1

tRX

> 4-6 days, increases with decreasing temperature

tR

> 10 h at low temperatures

X

> 2 g L-1 MLVSS

Ratio of FMa)

~ 0.5 g NH4-N (g MLVSS)-1 recommended

a) Ratio of feed to biomass.

a) Ratio of feed to biomass.

Table 10.4 summarizes the operational parameters favoring nitrification. Generally, increasing T, tR, SN and a sufficiently high t^ are beneficial for nitrifying bacteria. They result in sufficient MLVSS. A sludge age of tRX>4_6 days is needed to achieve nitrification. The presence of healthy and adequate nitrifying bacteria is the basic requirement for successful nitrification. The influences of tR and tRX on removal or removal rate are discussed in Chapter 6.

Denitrification

10.2.3.1 Denitrifying Bacteria and Stoichiometry

Denitrifying bacteria are capable of removing oxidized nitrogen from wastewater by converting it to N2 gas which escapes to the atmosphere. Most denitrifying organisms are facultative aerobic chemoorgano-heterotrophic bacteria which make up approximately 80% of the bacteria within an activated sludge environment. Under anoxic conditions nitrite and nitrate serve as electron acceptors instead of O2 and organic substrates as electron donors for ATP production at very low oxygen concentration.

Denitrifying bacteria are common soil and water microorganisms and are associated with fecal waste. They enter an activated sludge process as fecal organisms in domestic wastewater and use free molecular oxygen if it is available. The energy produced with O2 as the electron acceptor is only 7% more than with NO_ and NO_ if the same C source is used (McKinney and Conway 1957).

Besides heterotrophic denitrification, denitrification can also be performed by chemolitho-autotrophic bacteria with H2 or with reduced sulfate compounds as electron acceptors (Lompe 1992; Beller et al. 2004). Kuai and Verstraete (1998) showed the occurrence of oxygen-limited autotrophic nitrification_denitrification. The reduction of NO_ and NO_ to gases such as NO, N2O or N2 in suspended sludge or biofilm under low c' and/or anoxic condition is possible, even in the ab sence of organic carbon as endogeneous denitrification (Bernet et al. 2001). Autotrophic denitrification is used in some waterworks for treating groundwater containing NO-/NO- (Lompe 1992). We will not discuss these processes here.

There are five main nitrogenous compounds in denitrification (see Eq. 10.38). Nitrate is the initial substrate for denitrification and molecular N2 is the end-product. Other intermediates like NO and N2O can be emitted if incomplete denitrifica-tion occurs due to very high nitrate concentrations and relatively low organic substrate concentrations (Sumer et al. 1996).

The reduction of NO- is carried out by one organism in four steps. Each step can conditionally be inhibited; and intermediate products can escape by being dissolved in water and by being subsequently desorbed and transported by mass transfer into gas bubbles and then into the air. The kinetics of the intermediate steps are still not known in detail. Until now, no exact nitrogen balance has been able to show how much NO and N2O are built. It is very important to balance exactly by measurements, but it is very difficult to perform.

Nearly all denitrifiers are able to use NO- and NO-. The catabolism of denitrifi-cation that provides two growth- and energy-yielding steps is described in simplified form using methanol as the energy source (Halling-S0rensen and J0rgensen 1993; Lawrence and McCarty 1969):

6NO- + 5CH3OH ^ 3N2 + 5CO2 + 7H2O + 6OH- + AG0 (10.41)

However, this is in contrast to aerobic catabolism, during which the hydroxyl ion is not produced:

The organic substrate is completely oxidized to CO2 and H2O. The produced OH-(see Eqs. 10.40 and 10.41) is alkaline; and some of the CO2 produced is returned to the nitrification tank. The ion is compensated in part or completely depending on NH+ influent concentration and is consumed during nitrification of soft water.

In order to maintain adequate alkalinity in the activated sludge, various chemicals or alkalics can be added to the water. These chemicals include bicarbonates (HCO-), carbonates (COf-) and hydroxides (OH-) of calcium, magnesium and sodium. The following chemicals for buffering alkalinity are commonly added: sodium bicarbonate (NaHCO3), calcium carbonate (CaCO3), sodium carbonate (Na2CO3), calcium hydroxide (Ca(OH)2) and sodium hydroxide (NaOH). Sometimes this is not needed if, for example, hard water such as the water from Berlin (Beelitzhof) is being treated, which has a total hardness of15.3 °dH and a carbonate hardness of 10.8°dH (BWB 2004).

10.2.3.2 Stoichiometry and Kinetics of Denitrification

Nearly all organics can be used as substrate. For this discussion of stoichiometry for catabolism and anabolism methanol is suitable. Related to one C atom of methanol, we can write (Lawrence and McCarty 1969):

0.926 NO- + CH3OH + 0.22 H2CO3 ^ O.O5IC5H7O2N + 0.435 N2 + 0.926 HCO- + 1.56 H2O

1.49NO- + CH3OH + 0.79H2CO3 ^ 0.059C5H702N + 0.72 N2 + 1.49HCO-+ 1.84 H2O

In accordance with Eq. (10.13), the substrate utilization and denitrification rates are calculated as:

pX Yo

YXC/SC

The corresponding equations for NO2 can be obtained. From Eq. (10.43) the true yield coefficients YXC/SC and YSC/NO3-N follow:

0.051 Mxc 0.051 -12-5 g mol-1 C5H7O2N-C 1.0 ■ 12 g mol-1 CH3OH-C

gMLVSS11 gDOC

YSC/NO3-N

Msc gXC

gNO3-N

From Eq. (10.47) and Eq. (10.48), then Eq. (10.49) follows:

gNO3-N

and, respectively, for NO2 from Eq. (10.44):

0.454

gMLVSS gNO3-N

0.34

gMLVSS gNO3-N

NO3-N o

XC/NO3-N

1 It is assumed that the MLVSS consists of 50% carbon.

Table 10.5 Kinetic and yield coefficients of heterotrophic denitrification (Wiesmann 1994).

Symbol

Unit

NO3 reduction

NO2 reduction

Fmax

d-1

2.6

1.5

YX/S

g MLVSS (g DOC)-1

1

1

yx/n

g MLVSS (g NOX-N)-1

1.2

0.8

kd

d-1

0.1

0.1

ks

mg L-1 DOC

62.5

-

KNOx-N

mg L-1 NOX-N

<0.14

<0.12

Thus, 0.454 g MLVSS is produced for 1 g NO3-N removed by denitrification; and 25.5% of the CH3OH-C is used for anabolism and 74.5% for catabolism (see Eq. 10.49). However, the production of biomass depends on substrate used, resulting in different YX/N. Denitrifying bacteria can use most organic compounds commonly found in domestic wastewater. Several organic substrates such as methanol, acetic acid, ethanol, glucose, molasses or a part of the influent wastewater are often added to a denitrification tank if post-denitrification is run (see Section 10.4.2).

Table 10.5 presents some kinetic and yield coefficients of denitrification.

The specific growth rate of bacteria is influenced by both the concentration of the organic substrate and the concentration of NO_ or NO_. The kinetics of denitrification can be described by a double Monod kinetic model and an additional term to include the inhibiting effect of dissolved O2 concentration on denitrification for NO3_ (Batchelor 1982; IAWPRC 1986):

Note that all three saturation coefficients can differ if different substrates are used.

KS+ S KNO2+SNO2

10.2.3.3 Parameters Influencing Denitrification

From the kinetic observation in Eqs. (10.51) and (10.52) it can be seen that denitrification needs certain favorable conditions, such as the presence of organic substrate, very low c' (c'; 0), correct pH and T.

Sufficient organic substrate is one of the main control parameters for denitrification. From Eq. (10.48) the optimal ratio of organic carbon to nitrate is approximately YSC/NO3-N = 0.89 g DOC (g NO3-N)_1 where complete denitrification is possible. For lower ratios, the NO3 effluent concentration is increased. The value for NO2 is somewhat lower at YSC/NO2-N = 0.58 g DOC (g NO2-N)_1. This is one of the advantages of nitrification via NO2 accumulation (see Section 10.2.4). A high denitrification rate can be achieved if the concentration of readily biodegradable organic matter is controlled.

Free molecular oxygen inhibits denitrification because the oxygen suppresses the formation of the enzyme nitrate reductase (Payne 1973). Wheatland et al. (1959) found that the denitrification rate at c' = 0.2 mg L-1 was about one-half of the rate at c' = 0 mg L-1 (KiO = 0.2 mg L-1 O2).

Denitrification results in an increase in the alkalinity. The OH- produced in Eqs. (10.40) and (10.41) is used for building H2O with the H+ produced during nitrification. Denitrification can occur over a wide range of pH values. Most studies show the highest rates of denitrification occurring at pH 7.0-7.5 (Halling-S0rensen and J0rgensen 1993).

The growth rate of the organism and removal rate of nitrate are both affected by temperature. For wastewater below 5 °C, denitrification is highly limited because biological metabolism is too slow. Table 10.6 summarizes the operational factors favoring denitrification.

Within a redox potential range of +50 mV to -50 mV, oxygen is either absent or present only at a relatively small concentration. Above +50 mV, aerobic conditions dominate.

If it is possible for the carbon source for denitrification to be depleted, endogenous denitrification can occur. Adam (2004) observed a constant denitrification rate over a long time (>30 h) in a post-denitrification process without Bio-P organisms. This means that the kind of carbon source was not changed and/or depleted during this experiment. This is a typical characteristic of endogenous denitrification. Based on Eq. (10.51), rNO3 can be described to reflect endogenous respiration:

Endogenous denitrification rates are normally lower than when using external carbon sources. However, if the bacterial concentration in the anoxic zone is increased, the denitrification rate increases as a result (Adam 2004). The increase in ammonium concentration and decrease in bacterial concentration could be observed during endogenous denitrification and bacterial lysis.

Table 10.6 Operational parameters influencing denitrification.

Parameters

Optimal range/value and comments

Organic carbon

Main control parameter, ratio of 3:1 (organics as COD to NO2 and NO3)

is optimal for complete denitrification, and above 3:2 causes increase

in NO2 and NO3.

c'

Inhibits denitrification, obvious inhibition of denitrification at

c' > 0.2 mg L-1 O2.

PH

Affects enzymatic activity of denitrifying bacteria, 7.0 < pH

optimum <7.5.

Temperature

Denitrification rate increases with increasing T, until T=35 °C;

very low rate below 5 °C.

Redox potential

+50 to -50 mV, above +50 mV aerobic conditions dominate.

Nitrite Accumulation During Nitrification

Nitrite is accumulated under certain process conditions which promote the ammonium oxidation rate to a point that it exceeds the nitrite oxidation rate. Finding and optimizing these process conditions are the key points for nitrite accumulation. The following parameters are favorable for high nitrite concentrations:

• Limited dissolved oxygen concentrations due to a lower KNS = 0.3 mg L-1 O2 for Nitrosomonas compared with KNB = 1.1 mg L-1 O2 for Nitrobacter (see Table 10.3).

• Controlling the pH to obtain certain concentration levels of HNO2 and NH3 (see Section 10.2.2.2).

• Higher temperature favors Nitrosomonas (T = 28-35 °C; see Section 10.2.5).

The different K' values of Nitrosomonas and Nitrobacter (Dombrowski 1991; Wiesmann 1994; Pirsing 1996) show that nitrite oxidation to nitrate is more limited at low oxygen concentrations than ammonium oxidation. In aerobic biofilm reactors with high biomass concentrations, the conversion rate is usually limited by the oxygen transfer from liquid to biofilm (see Chapter 7). The limited oxygen transfer to a biofilm causes a very low dissolved oxygen concentration at the surface of the biofilm, so that the nitrite oxidation to nitrate is limited more effectively due to the lower K' values of Nitrosomonas compared to Nitrobacter. To take advantage of this characteristic, most research done on nitrite accumulation has centered on biofilm reactors (Abeling and Seyfried 1992; Garrido et al. 1997; Bernet et al. 2001; Antileo et al. 2003).

In some cases the oxidation of ammonia stops at the nitrite stage, even though c is high enough not to limit nitrite oxidation. This can be explained by the fact that nitrite accumulation is also linked to inhibition by ammonia. Anthonisen et al. (1976) found that ammonia inhibition of Nitrosomonas first becomes evident at concentrations of 8-124 g m-3 NH3-N (see Eq. 10.28), while the selective inhibition of Nitrobacter by HNO2 already occurs at concentrations of 0.1-1.0 g m-3 NH3-N.

By using both the characteristics of low oxygen concentration and the different ammonia inhibitions of Nitrosomonas and Nitrobacter, 74% nitrite accumulation was observed in a suspended membrane bioreactor (Choi 2005).

Figure 10.5 shows the schematic of nitrification and denitrification for achieving nitrite accumulation.

Sustained nitrite accumulation via the nitrite pathway (NH+ ^ NO- ^ N2) offers several benefits for nitrogen removal of wastewater, compared to the nitrate pathway (NH++ ^ NO- ^ NO- ^ NO- ^ N2):

• faster kinetics of the nitrification and denitrification processes,

• up to 25% energy savings during aeration,

• up to 40% savings from reduced demand for organic substrate,

• a higher rate of denitrification,

• lower biomass production (up to one third of former amount).

Fig. 10.5 Schematic for the accumulation of nitrite by nitrification and denitrification.

As disadvantages it can be mentioned:

• nitrification must be operated and controlled precisely,

• automatic measurement of NO2 concentration in effluent of the anoxic step results in increasing operating costs.

New Microbial Processes for Nitrogen Removal

The ANAMMOX process - an acronym for anaerobic ammonium oxidation - has been described as a new way for biological nitrogen removal. Certain chemolitho-autotrophic bacteria are capable of oxidizing the electron donor ammonium to nitrogen gas, with nitrite as the electron acceptor under anoxic conditions (Mulder 1992; Mulder et al. 1995; Jetten et al. 1998; Helmer et al. 2001):

NH+ + NO- ^ N2 + 2H2O + AG0 where: AG0 = -359 kJ ... -380 kJ (mol NH+)-1

The bacteria belong to the rare order of the Planctomycetes, of which Plancto-myces and Pirellula are the most important members. Current genera are Brocadia and Kuenenia (both freshwater species) and Scalindua (marine species). The bacteria catalyzing the ANAMMOX reaction are autotrophic, which means the conversion of nitrite to N2 proceeds without the use of organic carbon. The process is characterized by low sludge production and a substantial reduction in aeration energy by 60% and chemicals for neutralization. The net CO2 emissions are strongly reduced. The cost reduction compared to conventional N removal should be considerable (Van Dongen et al. 2001).

The SHARON process (acronym for single reactor system for high activity ammonia removal over nitrite) was conceived to promote biological nitrogen removal over nitrite in concentrated wastewater (Van Dongen et al. 2001) and provides several advantages (see Section 10.2.2). Its pH control is very important. Nitrite oxydation can be inhibited in regions of lower pH (higher HNO2 concentration) and limited in regions of lower oxygen concentration (Van Kempen et al. 2001). The process is operated at high temperatures (>25 °C), which selectively promote the fast-growing ammonium oxidizers, while Nitrobacter can be washed out of the system. It is characterized by a complete absence of sludge retention (tRX = tR), because the growth and washout of sludge are in equilibrium (Hellinga et al. 1998; Van Kempen et al. 2001).

Processes based on this autotrophic nitrogen removal concept have been described and investigated intensively in a sequencing batch reactor SBR (Strous et al. 1998; Fux et al. 2002), in a continuous flow moving-bed pilot plant (Helmer et al. 2001), in a fluidized-bed reactor (Van de Graaf et al. 1996) and in suspended SHARON_ANAMMOX systems (Hellinga et al. 1998; Van Dongen et al. 2001). This combined new way for nitrogen elimination can be applied technically to industrial wastewater with high ammonium concentrations but no DOC.

Cost estimates for the classic method of autotrophic nitrification/heterotrophic denitrification and for partial nitritation/autotrophic anaerobic ammonium oxidation (ANAMMOX) with anaerobic sludge digestion demonstrate that partial nitri-tation/ANAMMOX is more economical than classic nitrification/denitrification (Fux and Siegrist 2004). A full-scale cost estimation of different techniques for N removal from rejection water was carried out based on STOWA (1996) for WWTP capacity of 500 000 inh.

10.3

Biological Phosphorus Removal

Enhanced Biological Phosphorus Removal

Enhanced biological phosphorus removal in activated sludge systems was first reported in the late 1960s (Vacker et al. 1967). Acinetobacter sp. and especially the strain L. woffii were identified as the organisms responsible for accumulating excess phosphates in their cells, if they have short-chain volatile fatty acids (VFAs) available, especially acetate, as feed stock (Fuhs and Chen 1975).

Biological phosphorus removal is realized by creating conditions favorable for the growth of phosphate-accumulating organisms (PAOs). An initial anaerobic zone allows the PAOs to take up VFAs into their cells and store them as poly-p-hy-droxybuterate (PHF). The polyphosphate stored just prior to this is oxidized and used as an energy source, producing ATP; and it is thereby released into the liquid phase (Fig. 10.6). The anaerobic uptake of organic matter is inherently related to the accumulated polyphosphate.

prit

prit

anaerobic aerobic

Fig. 10.6 Mechanism of enhanced biological phosphorus removal; shown is each time the beginning of the process (Wentzel et al. 1991).

anaerobic aerobic

Fig. 10.6 Mechanism of enhanced biological phosphorus removal; shown is each time the beginning of the process (Wentzel et al. 1991).

After the mixed liquor reaches the aerobic zone, the stored PHF is used by the PAOs for cell growth and to provide energy for reforming polyphosphate from all the available orthophosphate and also for the synthesis of polyglucose (glycogen). By going through both anaerobic and aerobic conditions, PAOs are adequately established and become predominant in the biomass community after several weeks. The PAO's are the only bacteria being able to store substrate in a first anaerobic reactor and to oxidize them in a second aerobic reactor. This is only possible by enrichment of the Poly-P storage. This enrichment of the PAOs containing a high concentration of polyphosphate leads to the establishment of biological phosphorus removal. The net elimination of the process results from the bacterial cell growth and the removal of surplus sludge at the point when the phosphate is taken up to a higher level than that released in the anaerobic stage (see Fig. 10.7, below).

Kinetic Model for Phosphorus Removal 10.3.2.1 Preliminary Remarks

Obtaining kinetic and stoichiometric information requires that we make some assumptions, as follows:

• the reactors are operated as CSTRs (see Section 6.2.2),

• the process is in steady state,

• acetate is used as the substrate.

The biochemical pathway of the organic substrate metabolism is closely associated with polyphosphate storage. There is an apparent relationship between two parameters: organic substrate and polyphosphate. Substrate uptake and phosphorus release in the anaerobic phase can be described by the balances of acetate and PO4-P

246 10 Biological Nutrient Removal mixing point M

Fig. 10.7 Two-stage biological phosphorus removal in CSTR (AO process, Phoredox) with concentration profiles for phosphorus and substrate.

The process diagram is expanded compared to Fig. 6.3 by installing an anaerobic reactor in front of the aerobic one.

10.3.2.2 Anaerobic Zone

The following balances are valid for an anaerobic CSTR volume Van: for acetate S:

for PO4-P:

1 SC/PO4-P

for biomass X:

YSC/XC

with:

and:

Xan Xm

SM -San where SM is the concentration of acetate after mixing with returned sludge, San is the concentration of acetate in the anaerobic reactor, SPan is the concentration of PO4-P in the anaerobic reactor, SP,M is the concentration of PO4-P after mixing with returned sludge, rSan is the rate of acetate uptake, SPPan is the concentration of polyphosphate in bacterial cells in the anaerobic reactor and SPPM is the concentration of polyphosphate in bacterial cells after mixing with returned sludge.

In order to determine SP,an with a known reactor volume Van and flow rate Qm, it is first necessary to know the dependency of the substrate conversion rate rSAn on the concentrations of acetate San and orthophosphate-P SP,an in the anaerobic stage. The specific maximum growth rate pmax and yield coefficient YXC/SC are replaced by the rate coefficient k. The modified double-Monod kinetics could be verified by experiments (Wentzel et al. 1987; Gao 1995; Romanski 1999):

If nPP = 0, no substrate can be taken up. For nPPP KPP, the acetate uptake rate rS is only a function of San and Xan; and for Sanp KS it depends only on Xan.

10.3.2.3 Aerobic Zone

The following balances are valid for an aerobic CSTR volume Vae: for acetate S:

for PO4-P:

YSC/PO4-P

for biomass X:

YSC/XC

with:

an ae an

SP,ae SP,an SPP,ae-SPP,an and:

San Sae

In the aerobic zone, phosphorus uptake and substrate transformation rates are influenced by orthophosphate in the liquid phase and by the carbon source stored as PHB in bacterial cells. They are very closely connected with each other and it is assumed that the bacterial growth occurs based on intracellular PHB:

S PPae SP

YXC/PO4-P KPAe + SPAe KPHB + nPHB K +c r = Pmax x SP ae nPHB c (10 67)

Note that the substrate is now stored as PHB inside the cells.

Various models have been developed for the biological phosphorus removal by several authors (Wentzel et al. 1986; Tsuno et al. 1987; Ante and Vofi 1995; Gao 1995; Henze et al. 1995; Romanski 1999). But today there is no standard model to describe the kinetics of biological phosphorus removal. Its rate depends primarily on the concentration of polyphosphate-accumulating bacteria in both anaerobic and aerobic reactors and the concentrations in the Eqs. (10.59), (10.66) and (10.67). These equations have not been sufficiently validated and further investigations are needed.

Results of a Batch Experiment

Figure 10.8 shows concentration profiles of S and SP in a batch experiment, presenting a net elimination of phosphorus (Romanski 1999).

In the anaerobic period, the obligatorily aerobic poly-P bacteria (PAOs) take up substrate (e.g. acetate) and store it as lipid reserve material (PHB). Simultaneous-

Fig. 10.8 Concentration profiles of SS and SP in a batch experiment (Romanski 1999).

ly, the polyphosphate in the cells is partly utilized as an energy source and is released, resulting in an increase in SP from 20 mg L_1 to 60 mg L_1 PO4-P, which is closely correlated with the synthesis of PHB. The polyphosphate is released with a high rate as long as the acetate exists. Afterwards, other substrates being formed by lysis of bacteria are partly converted into lower fatty acids, resulting in a slower P-release.

In the following aerobic phase, the orthophosphate is taken up into the bacterial cells while the PHB is utilized for growth. The orthophosphate concentration SP decreases from the initial concentration of 20 mg L_1 at the beginning of the anaerobic batch test down to 12 mg L_1 PO4-P. The difference of 8 mg L_1 PO4-P is the net elimination of phosphorus. More phosphate is taken up aerobically than is released anaerobically because it is enriched in the biomass due to bacterial growth which is removed with the excess sludge.

Parameters Affecting Biological Phosphorus Removal

An adequate supply of VFAs is one of the key factors for successful biological phosphorus removal, due to its very strong relation to polyphosphate release or phosphate uptake. VFAs either are a part of the readily biodegradable substrate in the influent or are formed from it by fermentation in the anaerobic zone by facultative aerobic bacteria. In comparison, methanogenic bacteria are not able to grow in a system with changes from anaerobic to aerobic conditions.

If adequate dissolved oxygen is present, PAOs can grow in the aerobic zone at adequate rates. But the introduction of O2 or NO2 and NO3 to the anaerobic zone should be minimized because it is used preferentially as a terminal electron acceptor, which reduces the amount of VFAs available for uptake by the PAOs (Hascoet and Florentz 1985). As a result, phosphate uptake in the aerobic zone is reduced.

The solid retention time tRX must be adequate to allow PAOs to grow and can remarkably affect the phosphorus removal rate. Increasing the anaerobic tRX will allow increased fermentation of organic matter, resulting in increased production of VFAs and total removal rate. A low hydraulic retention time tR is beneficial in optimizing the process. The main parameters affecting biological phosphorus removal performance are summarized in Table 10.7.

Decreasing temperature in the anaerobic zone reduces the rate of fermentation. PAOs are less affected by decreasing pH than nitrifying bacteria are (US EPA 1993). Overall phosphate removal may fall with decreasing pH values because more energy is needed to take up acetates against a higher H+ concentration, because the concentration of undissociated acetate decreases.

The phosphorus content of the bacteria nPP may have a remarkable influence on the phosphorus removal rate because it is very closely linked to the capacity of PAOs for P release and uptake. The typical average nPP value is 5_7% of the bacterial mass and values as high as 12_15% are obtained in some cases, depending on the process configuration. The nPP for conventional activated sludge will typically range from 1.5% to 2.0% (Grady et al. 1999).

Table 10.7 Parameters affecting BPR process.

Parameters

Optimal range/value and comments

Concentration Adequate concentration of VFAs is beneficial. Low VFA

of VFAsa) concentration reduces the P release in anaerobic zone resulting in corresponding low P uptake in aerobic zone.

tRX tRX = 1.0-1.5 d is recommended for a growing of PAOs.

c' c' limits the formation of VFAs because VFAs are properly formed under strictly anaerobic conditions.

Temperature Low temperatures can reduce the formation of VFAs and the activity of PAOs.

pH PAOs are less sensitive to pH changes than nitrifying bacteria.

Decreasing pH adversely affects the P removal rate.

Presence of NO3 NO3 in anaerobic zone reduces P release resulting in decreasing

P uptake in aerobic zone.

P content of MLSS Very closely connected with capacity of PAOs for P-release and uptake.

a) Volatile fatty acids.

10.4

Biological Nutrient Removal Processes

Preliminary Remarks

Biological nutrient removal processes are modifications of the activated sludge process that combine anoxic and/or anaerobic zones with aerobic zones to provide nitrogen and/or phosphorus removal. Many configurations are possible, resulting in a wide range of performance capabilities and operational characteristics, which are presented in Table 10.8.

This section describes and discusses biological removal systems which provide removal of either nitrogen or phosphorus, or both components.

Nitrogen Removal Processes

The primary process for biological nitrogen removal consists of an aerobic stage for nitrification and an anoxic stage for denitrification. Figure 10.9a shows a two-stage biological nitrogen removal system (Ludzack and Ettinger 1962) called a modified Ludzak-Ettinger (MLE) process. They were the first to propose a single sludge nitrification-denitrification process using biodegradable organics in the influent wastewater.

10.4 Biological Nutrient Removal Processes | 251 Table 10.8 Characteristics of different zones in the biological nutrient removal process.

Zone

Biochemical transformation

Function

Removed component

Anaerobic

Phosphorus release

Enrichment of PAOsa)

Phosphorus

Formation of readily

Carbon

biodegradable organic matter

by fermentation

Uptake and storage of volatile

fatty acids by PAOs

Anoxic

Denitrification

Reduction of NO3-N to N2

Nitrogen

Metabolism of exogenous

Selection of denitrifying

Carbon

substrate by facultative

bacteria

heterotrophs

Production of alkalinity

Uptake of PO4b)

Phosphorus

Aerobic

Nitrification

Oxidation of NH4-N to NO2-N

Nitrogen

and/or NO3-N

Consumption of alkalinity

Nitrogen removal via gas

stripping

Phosphorus uptake

Formation of polyphosphate

Phosphorus

Metabolism of stored and

Uptake of PO4c)

exogenous substrate by PAOs

Metabolism of exogenous

Carbon

substrate by heterotrophs

a) Phosphate-accumulating organism.

b) In the presence of easily biodegradable organics, nearly all the PO4-P is taken up.

c) If all the easily biodegradable organics are used in the anoxic stages without complete PO4-P uptake, additional PO4-P is removed within the aerobic stage using organic lysis product.

a) Phosphate-accumulating organism.

b) In the presence of easily biodegradable organics, nearly all the PO4-P is taken up.

c) If all the easily biodegradable organics are used in the anoxic stages without complete PO4-P uptake, additional PO4-P is removed within the aerobic stage using organic lysis product.

Fig. 10.9 Biological nitrogen removal process for (a) pre-denitrification and (b) post-denitrification.

The anoxic stage for denitrification is located in front of the aerobic stage where NO- is formed. Both recycle streams QRN and QR have the target effluent amount of nitrate which restrains the possible amount of the denitrification. The real effluent nitrogen concentration is determined by the total nitrogen influent concentration to the process and the relation of total recycle flow QRt to the influent flow Q0 as nRN.

It is advantageous that the organic matter contained in the wastewater is consumed while no additional organic substrate is added. One drawback of this process is the remaining NO- which is discharged after formation because a typical maximum recycle flow rate is QRN;5(Q0 + QR). At higher QRN, the energy consumption for pumping is too high, resulting in high operational costs without a noticeable increase in N removal. This process enables excellent nitrification and a good degree of denitrification down to SN ; 4-8 mg L-1 Nt. In order to increase removal efficiency down to effluent levels of SN < 3.0 mg L-1 Nt, the MLE process was developed further, yielding the four-stage Bardenpho process by Bardard (1973). It involves the expansion of the process by a secondary anoxic and a small aerobic reactor.

In contrast to the MLE process (Fig. 10.9a), the aerobic zone is located in front of the anoxic zone (Fig. 10.9b). To use the biodegradable organic matter in the wastewater, a part of the influent bypasses the first stage and is introduced to the anoxic stage. Only the sludge is returned to the initial aerobic process. The energy consumption for pumping QRN is saved. If in some cases sufficient organic matter is not present in the influent or in the effluent from the aerobic nitrifying stage, a supplemental N-free carbon source, such as methanol or acetate, is added to the anoxic stage. This configuration may be useful if the price of added supplemental substrate is low and very low NH4 concentrations are required. This configuration can be expanded beyond the anoxic stage by a smaller aerobic zone to remove the remaining carbon and NH4 (in the case of a bypass of wastewater) from the anoxic stage. The addition of a supplemental N-free carbon source results in an improvement in the process efficiency, but increases chemical costs.

Chemical and Biological Phosphorus Removal

Before discussing the biological P removal process, we will briefly explain chemical P elimination by precipitation. The main part of phosphorus in domestic wastewater is orthophosphate PO4-P (Fig. 10.1). It can be separated from wastewater by precipitation with Al3+ and Fe3+ salts. Mostly two different processes are used: simultaneous precipitation occurs in the aerobic tank of an activated sludge plant, where Fe3+ is produced by the very fast oxidation of the cheaper Fe2+. If FeSO4 is applied, we write:

The insoluble FePO4 forms flocs mostly inside the activated sludge particles and can be separated as excess sludge.

In some northern countries, post-precipitation is preferred behind the secondary clarifier, using a reactor for precipitation and a settler for floc separation. The reactor is not aerated. Therefore instead of Fe2+ salts, Al3+ and Fe3+ salts are applied. If Fe2(SO4)3 is used, we write:

To obtain larger flocs with higher settling rate, polymers as flocculation aids are added.

As shown in Table 10.1, dissolved inorganic polyphosphates and organic phosphorus as well as particulate phosphorus are further components of municipal wastewater. They can only be separated partly by adsorption and co-precipitation.

The anaerobic and aerobic (or oxic) process (AO process, also called Phoredox) is a method for biological phosphorus removal (see Fig. 10.7). The placement of an anaerobic reactor in front of the conventional activated sludge process leads to the use of influent organic matter for the anaerobic formation of PHB. High rates of phosphorus removal are obtained by minimizing nitrification and maximizing the production of poly-P-storing bacteria. High solids production is beneficial if usage in agriculture is planned because the production of high phosphorus content biomass is maximized. The anaerobic zone is contained in the main process stream and is thus regarded as a mainstream biological phosphorus removal process.

Processes for Nitrogen and Phosphorus Removal 10.4.4.1 Different Levels of Performance

Many configurations have been developed as combined processes for biological nitrogen and phosphorus removal, including anaerobic, anoxic and aerobic zones. Due to the negative influence of nitrate on phosphorus removal, recycling of the nitrate into the anaerobic zone should be minimized and controlled; it is a key consideration in the selection and design of these processes.

The AAO process (Fig. 10.10a) is a combination of the anoxic and oxic MLE process (Fig. 10.9a) for nitrogen removal and the anaerobic and oxic Phoredox process (see Fig. 10.7) for phosphorus removal. The internal recycle flow rate is usually Qrn = (2-4)-(Q0 + Qr). The nitrogen removal rate is similar to that of the MLE process, but the phosphorus removal is sometimes a little lower than that of the AO Phoredox process.

Some nitrate is introduced with the return sludge into the anaerobic zone, resulting in an adverse impact on the phosphorus removal if QRN is too low. The greatest influence on the phosphorus removal is the organics content of the influent. If the organics content is high enough for both phosphorus and nitrogen removal, then the nitrate recycle will only have a slight impact on effluent quality, but if it is low then there would be serious influence on the removal rate. Denitrifica-tion for conversion of nitrate to N2 can be also carried out in part within a sludge blanket in the settler, which reduces the nitrate recycle to the anaerobic zone and leads to bacterial flocs being washed out of the system. Improper design of the

Fig. 10.10 Processes for removal of both nitrogen and phosphorus: (a) the AAO process; (b) sludge return only into the anoxic stage, partly return of O2- and AlO3-free activated sludge from the anoxic to the anaerobic stage; (c) two anoxic stages.

sludge blanket can lead to bulking, clumping and floating sludge, which reduces system effectiveness.

In order to eliminate the negative influences of the nitrate recycling to the phosphorus removal rate, a process was developed where the sludge was only returned to the anoxic stage in order to avoid the input of some oxygen into the anaerobic stage (Fig. 10.10b). Behind the anoxic stage a partial flow with NO3-free, non-thickened sludge was recycled into the anaerobic stage.

In addition to that, the anoxic zone can be divided into two (up to four) reactors (Fig. 10.10c). The first anoxic reactor receives and denitrifies the return sludge stream and the second receives and denitrifies the nitrate recirculation stream. The denitrified mixed liquor is recirculated from the effluent of the first anoxic reactor to the anaerobic zonein order to provide the influent wastewater with bacteria. The advantage is the protection of the second anoxic stage for influences of recycled nitrate with sludge return.

Many other biological nutrient removal processes for both nitrogen and phosphorus have been developed (Randall et al. 1992; Grady et al. 1999). The kind of pro cess and plant design used depends on the treatment goal, the legislation, the composition of the wastewater to be treated and the costs for operation as well as the costs for the modification

0 0

Post a comment