JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kim, S.-H.
Right arrow Articles by Prasher, S. O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kim, S.-H.
Right arrow Articles by Prasher, S. O.
Agricola
Right arrow Articles by Kim, S.-H.
Right arrow Articles by Prasher, S. O.
Related Collections
Right arrow Water Quality
Right arrow Redox Processes
Right arrow Nitrogen
Right arrow Experiment Design
Right arrow Other Models
Journal of Environmental Quality 32:1474-1480 (2003)
© 2003 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Vadose Zone Processes and Chemical Transport

Electron Affinity Coefficients of Nitrogen Oxides and Biodegradation Kinetics in Denitrification of Contaminated Stream Water

Seung-Hyun Kim*,a, Jong-Bae Chungb, Byeong-Ryong Jeongc, Young-Deuk Leeb and Shiv O. Prasherd

a Dep. of Environmental Engineering, Yeungnam Univ., Kyongsan 712-749, Korea
b Dep. of Agricultural Chemistry, Taegu Univ., Kyongsan 712-714, Korea
c Dep. of Agronomy, Taegu Univ., Kyongsan 712-714, Korea
d Dep. of Agricultural and Biosystems Engineering, McGill University, 21111 Lakeshore Road, Ste-Anne-de-Bellevue, Quebec H9X 3V9, Canada

* Corresponding author (kimsh{at}yumail.ac.kr)

Received for publication August 7, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 THEORETICAL DEVELOPMENT
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
During the dry season in Korea, rivers become more vulnerable to contamination by biochemical oxygen demand (BOD) and nitrogen. It is hypothesized that the natural characteristics of the streams in Korea allow the contaminated water to be treated at the tributaries. Downstream river water quality in Korea may be improved by spraying the contaminated stream water from the tributaries over the surrounding floodplains. The consequent water filtration through the soil could remove the contaminants through aerobic and denitrifying reactions. In this study, the kinetics parameters of the denitrifying reaction in floodplain filtration were determined using contaminated stream water. For the electron donor the Monod kinetics was used, while the competitive Michaelis–Menten model was employed for the electron acceptors. The parameters to the competitive Michaelis–Menten model were found using continuous denitrifying reactions, instead of the batch reactions employed in previous studies, to match the conditions needed to apply the competitive Michaelis–Menten kinetics. From the result, it was found that continuous reactions as well as batch reactions could be used to determine the affinity coefficients in denitrification. The results of this study also showed that the affinity coefficient of NO-2, using continuous reactions, was similar to that of other studies in the literature found via batch reactions, whereas the affinity coefficient of N2O was much larger than that acquired with batch reactions. The parameters obtained in this study will be used in future work to simulate the contaminant behaviors during floodplain filtration using a mathematical model.

Abbreviations: BCOD, biodegradable chemical oxygen demand • BOD, biochemical oxygen demand • COD, chemical oxygen demand • MLVSS, mixed liquor volatile suspended solid • NBCOD, nonbiodegradable chemical oxygen demand


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
THEORETICAL DEVELOPMENT
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
IN KOREA, approximately two-thirds of the total annual precipitation (1274 mm) is concentrated during the monsoon season, which begins usually at the end of June and lasts for about a month. As a result, the country is subject to dry weather and diminished river water flow rates throughout the rest of the year. Industrialized and densely populated areas in Korea are producing increasingly large amounts of wastewater, which is subsequently discharged into the rivers. Although the wastewater is treated before being discharged, many contaminants, mainly BOD and nitrate, still reach the river water and deteriorate the river water quality considerably during the dry period. Nonpoint-source contaminants from agricultural areas are also discharged directly into the rivers without any treatment. The presence in the river water of nitrate causes further contaminants to be produced through eutrophication. To improve the situation, it is necessary to remove both BOD and nitrate from the river water, originating from the nonpoint sources as well as from the point sources (Kwun, 1998). However, nonpoint-source contaminants have been considered to be difficult to manage.

Although the flow rates of rivers (about 60 m3 s-1 for the Nakdong River, the second largest in South Korea) and their first tributaries (usually 1–5 m3 s-1) during the dry period are quite low, they are still too high to be treated by the conventional methods of wastewater treatment. However, if some other method is applied, it may be possible to treat the contamination of the tributaries. Plains around the rivers and streams are usually used as rice (Oryza sativa L.) paddies in Korea. The well-established waterways that flow from the paddies to streams converge as they travel downstream to become tributaries for larger rivers. Due to these hydrological characteristics, nonpoint-source contaminants tend to collect and pass through the tributaries before reaching the rivers, so the treatment of nonpoint-source contaminants may be possible at the tributary level. In Korea, flooding takes place only several times over the course of the monsoon season and lasts for several days each time. Most rivers and streams in the country have wide floodplains in between the levees and shores that become inundated during the flooding. The floodplains are composed mainly of permeable materials, mostly sandy soils, and usually remain uncultivated and weedy. The vegetation covering the floodplains supplies the topsoil with organic materials and oxygen, resulting in the development of rhizosphere, which is an excellent habitat for microbes and worms (Pierzynski et al., 1994). These facts suggest that the process of spraying contaminated stream water over the floodplains may allow the contaminants (BOD and nitrogen) to be removed by microbes and worms during filtration through the floodplain soil. The close relationship that exists between oxygen supply rate and soil water content (Collin and Rasmuson, 1988) indicates that the amount of oxygen present in the soil and the depth of the aerobic zone can be controlled by the water flow rate through the soil. A denitrifying zone develops beneath the aerobic zone due to the depletion of oxygen in this region (Cho et al., 1997b; Patrick and Jugsujinda, 1992). By using an appropriate spray rate, the contaminated stream water may be filtered through the aerobic and denitrifying zones, removing both BOD and nitrogen from the stream water. Having passed through the aerobic and denitrifying zones and having entered the ground water, the resulting treated filtrate is expected to flow into the stream again since all streams in Korea are gaining streams during the dry period (Korea Water Resources Corporation, 1996).

Much of the research to date on denitrification has studied the loss of nitrogen as a fertilizer (Burford and Stefanson, 1973; Starr et al., 1974; Colbourn and Dowell, 1984). In addition, there have been many studies on the ratio of N2O to N2 emissions in the denitrification process because N2O is a greenhouse gas (Gaskell et al., 1981; Dendooven et al., 1994). More recently, several models were developed to predict N trace gas emissions from various agricultural and natural soils (Riley and Matson, 2000). As it is an environmental concern, the N2O emission caused by the denitrification process in floodplain filtration should also be examined and predicted along with the aqueous-phase nitrogen removal before field-scale application. Since high rates of water and contaminants are loaded on the soil in floodplain filtration, a necessity for a new kinetic model that can be applied to this peculiar situation arises to analyze the behaviors of all the nitrogen oxides simultaneously. However, very limited research on the denitrification kinetics has been performed to date (Cho and Mills, 1979; Stanford et al., 1975). During the early stages of denitrification kinetics research, mass action–type models were used to describe the denitrification process (Broadbent and Clark, 1965; Focht, 1974; Starr and Parlange, 1976; Stanford et al., 1975), and later the Monod- and double Monod–type models became more popular (Leffelaar and Wessel, 1988; Widdowson et al., 1988; Lindstrom, 1992; Doussan et al., 1997; Riley and Matson, 2000). The mass action–type models were not accurate in predicting the characteristic changes in nitrogen oxides during denitrification (Cho and Mills, 1979). Although some of the Monod-type models gave predictions favorably matching with denitrification experiments (Leffelaar and Wessel, 1988; Riley and Matson, 2000), the model is basically half-empirical (Monod, 1949) and lacks the sound theoretical basis to predict the concentration of each nitrogen oxide in denitrification.

Cho et al. (1997a) derived a competitive Michaelis–Menten kinetics model regarding the electron accepting processes as a series of concurrent enzyme reactions and defined the electron affinity coefficients for the electron acceptors as the model parameters. The derivation of this model assumes that the enzyme–nitrogen oxide complexes maintain constant concentrations during the reactions (Cho et al., 1997a). In the model, the preference of electrons for an acceptor is determined by the concentration of the electron acceptor and its affinity coefficient. Also, in the model, the accumulation of an electron acceptor takes place due to the differences in the affinities among the acceptors rather than the inhibition of other acceptors. The model gave accurate predictions of both the characteristic changes in the nitrogen oxides during denitrification and the concentration profiles of the nitrogen oxides along the depth of soils (Cho et al., 1997b).

To apply the competitive Michaelis–Menten model, the electron affinity coefficients must first be known. Studies on the determination of affinity coefficients using pasture soils were performed by Dendooven et al. (1994) and Dendooven and Anderson (1995). In their studies, the concentration changes in nitrogen oxides and CO2, which took place in a batch denitrifying reactor, were measured against time, and the affinity coefficients were determined by best-fitting the model predictions with the measurements. As they used batch reactions, however, it should be noted that the concentrations of the enzyme–nitrogen oxide complexes changed along with those of other participants in the reaction. Since the concentrations of the enzyme–nitrogen oxide complexes varied with time in batch reactions, it can be seen that the conditions under which the reactions occurred conflict with the assumption of constant concentrations used in the competitive Michaelis–Menten kinetics derivation (Cho et al., 1997a). Thus, there is a need to develop a new method to determine the affinity coefficients that satisfies the assumption made in the derivation of the kinetics.

In this study, the electron affinity coefficients for nitrogen oxides in the denitrification process were determined using continuous reactions, as opposed to batch reactions. This method of determining the coefficients is more compatible with the assumption implicit in the competitive Michaelis–Menten kinetics. Along with the affinity coefficients, the biodegradation kinetic constants were also determined under denitrifying conditions using contaminated stream water as the substrate.


   THEORETICAL DEVELOPMENT
TOP
 ABSTRACT
 INTRODUCTION
 THEORETICAL DEVELOPMENT
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
When a contaminated stream water is sprayed on the floodplain soil, it is expected that the organic matter of the stream water is first removed aerobically. This aerobic removal depth depends on soil type, water content of the soil, and the contaminant concentration of the stream water, among others. Beneath the aerobic zone, the denitrifying zone develops where the residual organic matter from the aerobic zone is decomposed by denitrification.

It is assumed that denitrification proceeds according to the following reactions, with the NO between NO-2 and N2O being neglected (Cho, 1982; Cho et al., 1997a):

[1]

If the reactions in the above sequence are regarded as enzyme reactions and take place concurrently, then the concentration changes in the nitrogen oxides and the nitrogen can be described as follows:

[2a]

[2b]

[2c]

[2d]
where CN5, CN3, CN1, and CN0 denote the aqueous concentrations (mmol L-1) of NO-3, NO-2, N2O, and N2, respectively, and QN5, QN3, QN1 represent the electron affinity coefficients for NO-3, NO-2, and N2O, for which the units are L mmol-1, L2 mmol-2, and L mmol-1, respectively. The physical meaning of the electron affinity coefficient is the weighting factor of electron competition for each species of nitrogen oxide during denitrification. The term R stands for the maximum electron production rate per unit volume of reactor (mmol L-1 d-1) that can be achieved with the saturation concentration of the substrate (Cho et al., 1997a; Cho and Mills, 1979). The solutions to these equations have been known to predict effectively the characteristic changes in the nitrogen oxides during the denitrification process (Cho et al., 1997a,b; Dendooven et al., 1994; Dendooven and Anderson, 1995). The original forms of Eq. [2a–d] in Cho et al. (1997a) contain terms related to oxygen. However, the oxygen concentration is zero under the denitrifying condition, and as a result the oxygen-related terms disappear from the original forms, resulting in Eq. [2a–d]. Furthermore, this study is limited to cases in which nitrogen oxides always exist in the system, so the equations characterizing the denitrification reaction in question can be described as the "saturated" equations (Cho et al., 1997a).

The denitrification reaction rate is known to be independent of the species of the nitrogen oxides functioning as the electron acceptor (Cho, 1982; Dendooven et al., 1994; Dendooven and Anderson, 1995). Therefore, as long as the electron acceptor is any one of the nitrogen oxides, the rate of denitrification reaction does not change with the electron acceptor species. Cho et al. (1997a) regarded R as a constant based on the assumption that the organic content of the soil and its degradation rate remained unchanged in the denitrification reaction. However, denitrification is a biodegradation process in which the substrate degradation rate, otherwise known as the electron production rate, is closely related to the substrate type and concentration (Burford and Bremner, 1975). This relationship can be assumed to be linear provided that the organic content of the soil is as low as that of natural soils (Stanford et al., 1975). However, in the present case, as the concentration of the substrate is fairly high due to the introduction of external contaminants, the relationship between the electron production rate and the substrate concentration is no longer linear. Hence, the Monod model was chosen to describe this relationship (Lindstrom, 1992). The model is represented, neglecting the endogenous respiration, by the following equation:

[3]
where X denotes the biomass concentration expressed by mg MLVSS (mixed liquor volatile suspended solid) L-1, S indicates the substrate concentration represented by mg BCOD (biodegradable chemical oxygen demand) L-1, and Y, µmax, and KS stand for the conversion factor of the substrate into biomass (mg MLVSS per mg BCOD), the maximum specific growth rate (d-1), and the half-saturation constant (mg BCOD L-1), respectively. This study assumes that there exists a constant ratio between the two components of the organic contaminant, biodegradable chemical oxygen demand (BCOD) and nonbiodegradable chemical oxygen demand (NBCOD). The term aN represents the ratio of the electron production to the substrate consumption and is equal to 0.125 mmol electron produced per mg BCOD removed when the substrate consumption for biomass accumulation is neglected.

The method of using continuous reactions at steady state to determine the electron affinity coefficients agrees best with the assumption made in the derivation of the Michaelis–Menten kinetics, which states that the enzyme–nitrogen oxide complexes are kept in a constant concentration during the reactions (Sundstrom and Klei, 1979). In this study, four experimental apparatuses, as shown in Fig. 1 , were each fed with contaminated stream water in the continuous mode and operated with different hydraulic detention times to produce the steady state experimental results. The mass balance equations on the electron acceptors in a continuous reactor can be described as follows:

[4a]

[4b]

[4c]

[4d]
where q0 and qa denote the flow rates of the raw water into the reactor and the helium gas into the headspace of the reactor (L d-1), respectively, and CN5,0, CN3,0, CN1,0, and CN0,0 represent the NO-3, NO-2, N2O, and N2 concentrations of the raw water (mmol L-1), respectively. The terms CN1,a and CN0,a express the concentrations of N2O and N2 of the effluent gas (mmol L-1), respectively, and V stands for the volume of the aqueous phase in the reactor (L).



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 1. Schematization of the denitrifying reactor.

 
In Eq. [4a–d], all parameters except QN5, QN3, and QN1 are measurable in the experiment. If the corresponding measured values for steady state operation are substituted for the measurable parameters, then Eq. [4a–d] yields a homogeneous system of linear equations in which the unknowns are QN5, QN3, and QN1. Because the equations in Eq. [4a–d] were applied separately to each of the four reactors in the study, the experiments gave rise to a total of 16 equations. This homogeneous system cannot be solved directly because the equations are mutually independent. However, when any one of the unknowns is fixed as a constant, the relative solutions can be obtained from the resulting nonhomogeneous system (Kreyszig, 1999). In the present study, QN5 is fixed to the constant value of unity as in previous studies (Cho et al., 1997a; Dendooven et al., 1994; Dendooven and Anderson, 1995). Then, any two of the 16 equations can yield unique QN3 and QN1 values if the reactors operated ideally. In reality, however, combinations of any 2 out of the 16 equations yielded all different values, some of which were abnormally high or low. In this study, instead of averaging these values, another way was taken to obtain the representative values that best satisfy all the 16 equations. To make a uniform contribution of all 16 equations to the solution of QN3 and QN1, the system is rewritten as:

[5]
where A is a 16 x 2 coefficient matrix, is a vector of two unknown parameters, and denotes the known vector, containing 16 entries, that results from setting QN5 to a constant value of unity. This coefficient matrix does not have an inverse since the number of equations is larger than that of the unknowns. However, if both sides of Eq. [5] are multiplied by the transpose of the coefficient matrix, AT, the resulting system can be solved to yield the unique values of the two remaining unknowns. This system is represented by the following equation:

[6]

The solution obtained through this process is expected to be most representative of the four sets of independent experiments. Application of the same process to the subsystems yields ranges of parameter values instead of unique ones. As the number of equations decreased in the subsystem the ranges became wider. In this study, 120 different combinations of a 14-equation subsystem was chosen to see the ranges. The number 14 was chosen arbitrarily.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 THEORETICAL DEVELOPMENT
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The contaminated stream water was obtained in the lower reach of Omokcheon Stream, which passes through the outskirt of Kyongsan to enter the middle reach of the Kumho River. This river is one of the first tributaries of the Nakdong River. The stream water was taken about twice a week from 1 Mar. 2000 to 30 Nov. 2001. It was passed through 250 mesh and kept calm for 2 h so that the supernatant could be used as the raw water for the experiments. The characteristics of the raw water are shown in Table 1 . The microbes used in the experiments were obtained from the floodplain and riverbed soils by stripping them off with water. The acquired microbes were put into the denitrifying reactor, shown in Fig. 1. Then, the reactor was continuously fed for 3 wk with the municipal sewage entering the Kyongsan Wastewater Treatment Plant, of which the BOD was about 200 mg L-1, so as to increase the concentration of the biomass in the reactor to 100 mg MLVSS per L.


View this table:
[in this window]
[in a new window]
  		Table 1. Characteristics of the raw water from Omokcheon Stream.

 
A plastic vessel was employed as the raw water reservoir and kept in the refrigerator at a temperature of 4°C. The vessel was collapsible, remaining airtight by allowing its volume to shrink along with the decreasing water volume. The raw water was exposed to ultraviolet (UV) irradiation for two hours to inactivate the microbes (Jagger, 1967). This was the amount of time required for the colony forming units of the microbes to reduce to 3% of those of the raw water. Sodium sulfite (Na2SO3) and cobalt chloride (CoCl2) were added to the raw water to make their concentrations 100 and 16 mg L-1, respectively, which was needed to keep the raw water free from oxygen. These concentrations were higher than those used in the literature, 7.9 mg Na2SO3 per mg O2 and 2.3 mg CoCl2 per mg O2 (Ramalho, 1983). The NO-3–N concentration of the raw water was about 2.35 mg L-1, which was too low considering that NO-3–N was to be used as the electron acceptor. So, potassium nitrate (KNO3) was added additionally to the raw water to raise the NO-3–N concentration to about 6 mg L-1.

A peristaltic pump was used to continuously feed the reactor with raw water from the reservoir. The inner diameter of the tube connecting the pump and the reactor was 3.15 mm. Four reactors were employed in the experiment, and their operational conditions are listed in Table 2 . The aqueous phase of the reactors was continuously mixed with a magnetic stirrer. The reactor and settler were made of glass and sealed as in Fig. 1 to prevent air from entering. Helium gas was continuously introduced to pass through the headspaces of the reactor and the settler. The gas flow rates were controlled by a microvalve (Swagelok, Willoughby, OH). The outlet was submerged in water to keep the gas in the headspaces pressurized and to prevent air from entering the reactor. The helium gas flow rate was temporarily increased to about 200 mL min-1 to ensure that the system remained hermetically sealed when the sampling port of the reactor and settler was opened.


View this table:
[in this window]
[in a new window]
  		Table 2. Operational conditions of the denitrifying reactors.

 
The concentrations of chemical oxygen demand (COD), NO-3–N, and NO-2–N of both influent and effluent, the pH and Eh of the reactors, and the effluent MLVSS were measured at least once a week to check whether the reaction had reached steady state. Once steady state was achieved, the above parameters, the MLVSS concentrations of the reactors, and the aqueous and gas-phase concentrations of N2O and N2 were each measured seven times as a minimum. The averages of the seven measurements of each parameter were used to analyze the experiment. The aqueous-phase concentrations of N2O and N2 were determined indirectly in the following manner. The aqueous sample was equilibrated with a finite volume of helium gas in a sealed bottle, the gas-phase concentrations of N2O and N2 were measured directly, and then the aqueous-phase concentrations were ascertained by applying Henry's Law on the gas-phase concentrations. The N2 and N2O in the raw water were assumed to equilibrate with those in the atmosphere. During the experiment, the dissolved oxygen concentration of the reactor was measured periodically, but no oxygen entrance took place. All the experiments in this study were performed at a constant temperature of 20 ± 1°C.

The COD and the MLVSS concentrations were measured with standard methods (American Public Health Association, 1998). The Eh and pH were measured with a 692 pH/ion meter (Metrohm, Herisau, Switzerland), and NO-3–N and NO-2–N with a FIASTAR 5000 analyzer (FOSS Tecator AB, Höganäs, Sweden). The N2 and N2O in the gaseous phase were separately determined by gas–solid chromatography. For N2 analysis, a Hewlett-Packard (Palo Alto, CA) 4890 gas chromatograph (GC) equipped with a thermal conductivity detector was used. A stainless steel column (0.32-cm o.d. x 2-m length) packed with molecular sieve 5A (45–60 mesh) was employed as the analytical column. The operating parameters for N2 analysis were as follows: column temperature, 35°C; detector temperature, 200°C; inlet temperature, 150°C; carrier, helium at 30 mL min-1; sample size, 0.5 mL. The retention time and detection limit of N2 were 1.1 min and 3000 µL L-1 at signal to noise ratio (S/N) > 10, respectively. The N2 was completely resolved with late-eluting O2, Ar, and CO2 during the analysis. The N2O was analyzed by a Hewlett-Packard 5890 Series II GC equipped with a 63Ni-electron capture detector and a glass column (2-mm i.d. x 2-m length) packed with Porapak QS (100–120 mesh; Alltech Associates, Deerfield, IL). The operating parameters for N2O analysis were as follows: column temperature, 30°C; detector temperature, 250°C; inlet temperature, 100°C; carrier, argon and methane (95:5, v/v) at 35 mL min-1; sample size, 0.5 mL. The N2O was eluted at 1.5 min and was detected up to 0.04 µL L-1 at S/N > 10. The high specificity of the detector also resulted in no interference with other gaseous compounds.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 THEORETICAL DEVELOPMENT
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Experimental Results
The reactor experiment started on 21 Nov. 2000 and reached steady state in about 60 d. The experiment continued for another 60 d to confirm the attainment of steady state, after which the experimental results were collected for analysis. The COD of the raw water continued to change during the experiment mainly due to precipitation. The effluent COD, on the other hand, showed a trend of change that grew increasingly similar to that of the COD of the raw water as the system approached steady state. The removal of BCOD from the raw water achieved a rate in the range of 50 to 75%. The NO-3–N concentration of the stream water ranged from 2 to 3 mg L-1, and the adjusted NO-3–N concentration of about 6 mg L-1 decreased in the reactors to 4.1 to 4.5 mg L-1. The NO-2 and N2O concentrations increased during the reaction. The pH and the Eh did not change significantly, ranging from 7.97 to 8.44 and from 196 mV to 237 mV, respectively. The average measurements at steady state are shown in Table 3 . The mass balance on the nitrogen in the reactors showed a maximum error of 3%. This measurement error can be attributable, above all, to the indirect measurements of the aqueous-phase N2O and N2 concentrations.


View this table:
[in this window]
[in a new window]
  		Table 3. Results of denitrifying reactor operations at steady state.

 
Determination of the Biodegradation Kinetics
The data in Tables 1, 2, and 3 were used in determining all the parameters in this study. The Monod parameters of the BCOD degradation were determined by the processes of Grady and Lim (1980). The BCOD composition of the total COD was determined according to Grady and Lim (1980). The process used to determine the yield coefficient and decay rate coefficient is shown, as an example, in Fig. 2 and details of the processes for all the parameters are described by Kim (2002). The parameter values of the denitrifying reaction are Y = 0.41 mg MLVSS per mg BCOD, µmax = 1.51 d-1, KS = 29.4 mg BCOD per L, and kd = 0.037 d-1, where kd expresses the microbial decay rate. These values can be compared with those of aerobic parameters obtained by Kim (2002) using the same stream water and processes. They were Y = 0.59 mg MLVSS per mg BCOD, µmax = 5.32 d-1, KS = 35.3 mg BCOD per L, and kd = 0.042 d-1. Based on their experiment using lactate as the substrate, Doussan et al. (1997) reported that the KS values under both aerobic and denitrifying conditions were almost identical and that Y and µmax of the denitrifying reaction were about 80 and 15% of those of the aerobic reaction, respectively. It can be noted that the present study's results and the above findings of Doussan et al. (1997) exhibit a similar trend. As discussed by Kim (2002), the parameter values for the aerobic reaction fall in the range of those typical of municipal wastewater (Metcalf and Eddy, Inc., 1979). The ratio of NBCOD to the total COD of the raw water was about 30% for the denitrifying reactions, and this can be compared with 20% for the aerobic reaction (Kim, 2002).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 2. Determination of the yield coefficient and decay rate coefficient in the denitrifying reaction.

 
Determination of Electron Affinity Coefficients for Nitrogen Oxides
The experimental conditions in Tables 2 and 3 were applied to Eq. [4a–d] to yield a linear system of 16 mutually independent equations. The complete 16-equation system was solved using Eq. [6] to yield QN3 = 306 L2 mmol-2 and QN1 = 131 L mmol-1, with the assumption of QN5 = 1 L mmol-1. Also, each of 120 subsystems of 14 equations was solved independently by using Eq. [6]. The solutions of these subsystems ranged between 274 and 441 L2 mmol-2 for QN3 and 97 and 170 L mmol-1 for QN1, with the assumption of QN5 = 1 L mmol-1. Cho et al. (1997a) employed the equivalent values of 1 L mmol-1, 100 L2 mmol-2, and 0.1 L mmol-1 for QN5, QN3, and QN1, respectively. Based on batch experiments performed on pasture soils, Dendooven et al. (1994) reported their measurements as 1, 500 to 1500, and 0.75 to 2.5 for the equivalent values of QN5, QN3, and QN1, respectively, and Dendooven and Anderson (1995) reported their measurements as 1, 300 to 450, and 3 to 4, respectively. These values are summarized in Table 4 . It can be seen from the above results that the value of QN3 obtained in this study is similar to the values obtained by other researchers, while the value of QN1 is much larger than in other studies. This difference can be explained by the variations in the conditions under which the experiments were performed. In the present study, the N2O produced in the reactor was continuously removed along with the flowing gas and water since the reactor was operated in the continuous mode, whereas in other studies with the reactors operated in batch modes, the N2O was detained in the reactors. Another difference in results may be due to the fact that the initial NO-3 concentrations of other studies are much higher than that of the present study. The difference in the raw water among the different studies may also have caused different results.


View this table:
[in this window]
[in a new window]
  		Table 4. Summary of the electron affinity coefficients.{dagger}

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 THEORETICAL DEVELOPMENT
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
As a basic study contributing to BOD and nitrogen removal from stream water by floodplain filtration, the biodegradation kinetic parameters and the electron affinity coefficients for nitrogen oxides were determined in the denitrifying reaction. Experiments were conducted to determine the parameters under continuous reactions rather than the batch reaction approaches used in other studies. This experimental design choice was taken to better match the assumptions of constant concentrations of reaction participants that were made in the competitive Michaelis–Menten kinetics derivation. From the results, it was apparent that continuous reactions could be used to determine the electron affinity coefficients for nitrogen oxides in the denitrifying reaction. The results showed that the relative electron affinity coefficient ratio for NO-3, NO-2, and N2O was 1 L mmol-1 to 306 (274–441) L2 mmol-2 to 131 (97–170) L mmol-1. As is illustrated by this ratio, the affinity coefficient of NO-2 obtained in this study using continuous reactions was similar to that of the other studies in the literature found via batch reactions, whereas the affinity coefficient of N2O was much larger than that determined with batch reactions. The yield coefficient, maximum specific growth rate, half-saturation constant, and microbial decay rate coefficient of the denitrifying reaction were 0.41 mg MLVSS per mg BCOD, 1.51 d-1, 29.4 mg BCOD per L, and 0.037 d-1, respectively. The ratio of NBCOD to the total COD of the raw water was about 30% in the denitrifying reaction. The results of this study will be used in future works, where a mathematical model will be developed using the competitive Michaelis–Menten kinetics to find out the best design of the denitrifying zone in floodplain filtration and to evaluate its adverse effects. The processes and results of this study may also be used in the area where the behavior of all the nitrogen oxides should be analyzed simultaneously. In the application of the results of this study, however, it should be recognized that the results include some limitations. These are mainly caused by the deviations of conditions in the laboratory from the field, such as the controlled reactor and the alteration of the raw water using the chemicals and UV irradiation.


   ACKNOWLEDGMENTS
 
This work was supported by Korea Research Foundation Grant KRF-99-042-E00089.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 THEORETICAL DEVELOPMENT
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2003 32: 1167-1172. [Full Text]  




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kim, S.-H.
Right arrow Articles by Prasher, S. O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kim, S.-H.
Right arrow Articles by Prasher, S. O.
Agricola
Right arrow Articles by Kim, S.-H.
Right arrow Articles by Prasher, S. O.
Related Collections
Right arrow Water Quality
Right arrow Redox Processes
Right arrow Nitrogen
Right arrow Experiment Design
Right arrow Other Models

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome