1 23

Clean Technologies andEnvironmental PolicyFocusing on Technology Research,Innovation, Demonstration, Insightsand Policy Issues for SustainableTechnologies ISSN 1618-954X Clean Techn Environ PolicyDOI 10.1007/s10098-013-0621-7

Removal of nitrate and COD fromwastewater using denitrification process:kinetic, optimization, and statistical studies

Sanak Ray, Ayusman Mohanty,S. S. Mohanty, Snehasish Mishra &G. Roy Chaudhury

1 23

Your article is protected by copyright and

all rights are held exclusively by Springer-

Verlag Berlin Heidelberg. This e-offprint is

for personal use only and shall not be self-

archived in electronic repositories. If you wish

to self-archive your article, please use the

accepted manuscript version for posting on

your own website. You may further deposit

the accepted manuscript version in any

repository, provided it is only made publicly

available 12 months after official publication

or later and provided acknowledgement is

given to the original source of publication

and a link is inserted to the published article

on Springer's website. The link must be

accompanied by the following text: "The final

publication is available at link.springer.com”.

ORIGINAL PAPER

Removal of nitrate and COD from wastewater usingdenitrification process: kinetic, optimization, and statisticalstudies

Sanak Ray • Ayusman Mohanty • S. S. Mohanty •

Snehasish Mishra • G. Roy Chaudhury

Received: 12 December 2012 / Accepted: 3 April 2013

� Springer-Verlag Berlin Heidelberg 2013

Abstract Denitrification process studies were carried out

for removal of N–NO3- and organic carbon from waste

water. Recent researches have shown the efficacy of

N–NO3- removal process from synthetic waste water;

however, they are lacking the data related to kinetics and

optimization that may aid in developing guidance for field-

scale implementation. Various incubation parameters were

varied such as nitrate and carbon source keeping biomass

concentration invariable. For nitrate variation two different

nitrates were used such as potassium nitrate (KNO3) and

calcium nitrate [Ca(NO3)2] and for organic source glucose

(C6H12O6) and sodium acetate (CH3COONa). The deni-

trifying bacteria were isolated from agricultural soil. The

dentrification rates increased with the increase of nitrate as

well as organic source. The denitrification rates were faster

in case of CH3COONa compared to C6H12O6. The regen-

eration time was also calculated which varied 1–3 days.

Analysis of variance studies indicated that both nitrate and

organic concentrations played significant role in terms of

dentrification as well as emission of nitrous oxide (N2O)

rates. Principal components analysis studies were also

carried out. Optimization studies were carried out using

response surface model to evaluate the incubation param-

eters to minimize the N2O emission and at the same time

maximize the removal efficiency of N–NO3- and COD.

Keywords ANOVA � Wastewater treatment �Denitrification � Nitrous oxide � Optimization � PCA

Introduction

Water is the source of life, and it plays an important role in

supporting the life system. Due to increase of population,

industrialization and agricultural activities the overall

quality of water is in wane. The water is gradually being

contaminated with various pollutants such as inorganic as

well as organic nitrogenous compounds originating from

agricultural and human activities. The toxic metal ions

released from metallurgical activities along with organic

compounds produced due to urbanization and industrial

activities leads to the water contamination. The wastewater

containing nitrogenous and organic compounds results in

eutrophication, biochemical oxygen demand increase and

thereby decreases the existence of water bodies.

Due to high solubility of nitrogen compounds in water,

it cannot be removed chemically by precipitation. Waste-

waters containing nitrogenous compounds are habitually

treated bio-chemically by nitrifying as well as denitrifying

bacteria (Kampschreur et al. 2009; Koren et al. 2000).

Denitrification is a primary biological process in which

nitrogenous compounds (oxidized) were reduced to atmo-

spheric nitrogen (N2) gas as a final product. In the deni-

trification process, the bacteria use NO3- or NO2

- as an

electron acceptor during respiration mechanism. The

electron donor in the overall process for the heterotrophic

microorganism are organic substrates such as acetate,

various poly as well as monosaccharide’s, organic acids

like oxalic etc. (Pant et al. 2010; Rabaey and Verstraete

2005). The waste water emerging out of domestic, food

processing, fermentation including sugar mills and poultry

S. Ray � A. Mohanty � S. S. Mohanty � G. R. Chaudhury (&)

CSIR-Institute of Minerals and Materials Technology,

Bhubaneswar 751013, Odissa, India

e-mail: gr_chaudhury@yahoo.com

S. Mishra

School of Biotechnology, KIIT University,

Bhubaneswar 751024, Odissa, India

123

Clean Techn Environ Policy

DOI 10.1007/s10098-013-0621-7

Author's personal copy

usually contaminated with nitrate as well as organic pol-

lutants as mentioned above (Pant et al. 2010; Rabaey and

Verstraete 2005). These pollutants need to be removed

before disposal order to keep ecological balance (Metcalf

and Eddy 2003). The denitrification process can be con-

sidered as a redox reaction. The process forms various

intermediates such as nitrite (NO2-), nitric oxide (NO),

N2O, and finally N2 (Wrage et al. 2001; Koren et al. 2000).

In the first step, the NO3- is reduced to NO2

- by nitrate

reductase enzyme (Nar). The NO2- is further reduced to

NO by nitrite reductase, followed by further reduction to

N2O by nitric oxide reductase. The final conversion of N2O

to nitrogen gas is carried out by nitrous oxide reductase

(Hu et al. 2012; Tavares et al. 2006).

The microbiological species involved in the total pro-

cess are termed as denitrifiers (Wrage et al. 2001). The

notable denitrifiers are Pseudomonas, Bacillus, Propioni-

bacterium, etc. (Faulwetter et al. 2009; Firestone 1982).

The biological species are facultative heterotrophs which

utilize nitrate (NO3-) as an electron acceptor instead of

oxygen (O2) during respiration. During wastewater treat-

ment method, the emission of N2O has been pragmatic

(Foley et al. 2010; Kampschreur et al. 2009), a significant

green house gas (GHG) (Dai et al. 2012; Tallec et al. 2008)

which initiates global temperature swell and causes ozone

depletion (Toma et al. 2011; Yang et al. 2011; Bergster-

mann et al. 2011). During the reduction process, N2O gas is

formed as an intermediate by microbial denitrification

route (Gogoi and Baruah 2012). Increasing atmospheric

N2O concentration can be attributed to anthropogenic

(Toma et al. 2011; IPCC 2007) as well as natural (Marinho

et al. 2004) activities. Denitrification plays an important

role in various ecosystems such as wastewater and agri-

cultural activities (Laverman et al. 2010). Various factors

such as oxygen, organic matter, nitrate redox-potential,

temperature, and pH plays an important role in determining

the denitrification kinetics (Koren et al. 2000; Yang et al.

2011; Laverman et al. 2010). The reaction is favored at low

O2 concentration as at higher concentration the denitrifiers

enzymes were deactivated (Nelson and Knowles 1978;

Terai and Mori 1975) but some of the denitrifiers can use

NO3- as well as O2 as electron acceptor (Koren et al.

2000). For continuation of metabolic activities, the denit-

rifiers require carbon source such as methanol, acetate,

ethanol, etc. (Koren et al. 2000; Park and Yoo 2009). The

denitrification rate has a positive correlation with nitrate

concentration whereas the concentration of carbon source

has a complicated correlation (Koren et al. 2000; Laverman

et al. 2010). The optimum pH for growth reported to be in

the range 7–8 and below 6.5 decreased considerably

(Koren et al. 2000). The denitrification is favored at high

temperature whereas low temperature favors the formation

of GHG.

Denitrification process is a complicated one, therefore,

an in-depth analysis is required to correlate the denitrifi-

cation as well as GHG emission rates vis-a-vis the various

parameters. However, this process of denitrification

involves emission of N2O which in turn depends on the

concentration of nutrients like COD and nitrogen source

along with microbial concentration. These parameters play

either a synergistic or an anti-synergistic effect on the

emission of N2O. Considering complicacy of the total

process to be enormous, there is an urgent need to evaluate

the mathematical intra-correlation between all these

parameters to minimize the emission of N2O and simulta-

neously maximizing the removal efficiency of N–NO3-

and COD. Moreover, the scanty relevant literature makes

the problem more difficult. Therefore, as a part of our

systemic studies on emission of GHG through biological

route (Sahoo et al. 2010; Ramulu et al. 2009, 2008), this

communication specifically aims to establish a mathemat-

ical co-relation between the parameters to minimize the

N2O emission through response surface modeling (RSM)

followed by statistical analysis which could be a way for-

ward to this situation. RSM technique is used to optimize

the response of a multivariate system by exploring the

relationships between various variables and their combined

effects on response variables (Zahed et al. 2012; Rao et al.

2012; Aber and Sheydaei 2012). This technique is also

used to reduce the number of experiments and to develop

mathematical models to determine the interactions between

process variables accurately. Batch experiment studies

were conducted to study the effect of variables such as

COD, nitrogen source (nitrate nitrogen) along with days.

The main objective of this study is to evaluate and maxi-

mize the N–NO3- along with COD removal efficiency,

simultaneously minimizing the N–N2O emission.

Materials and methods

Sampling

Soil sample was collected from a paddy field in Tangi-

banta, a rural location 20 km away from Bhubaneswar,

India at a depth of 0–15 cm using an auger (Andert et al.

2011). The typical soil analyses showed: Na: 1.5 %, Mg:

1.8 %, Al: 20.5 %, Si: 53.2 %, K: 1.2 %, and Ca: 2.4 %. It

also contained micronutrients like carbon, nitrogen, and

phosphorous. The denitrifying heterotrophic bacteria were

isolated from the agricultural soil using mineral, salt, and

nutrient (MSN) liquid media (Composition: CH3COONa—

7.86 g/L, KH2PO4—0.2 g/L, (NH4)2SO4—0.5 g/L, MgSO4.

7H2O—0.04 g/L, Ca(NO3)2—0.04 g/L). The isolated

strains were enriched at a regular basis by re-inoculating

into freshly prepared MSN media to enrich the biomass and

S. Ray et al.

123

Author's personal copy

increase their activity. The mixed culture mainly contained

species like Pseudomonas aeruginosa, Proteiniphilum

acetatigenes, and Alcaligenes faecalis. The species iden-

tification was done by 16S rDNA based method technique

at Indian Institute of Technology, Roorkee.

Experimental setup

As the denitrifiers are facultative in nature, incubation

studies were carried under anaerobic conditions which are

favorable to carryout denitrification process. Two variables

were varied such as nitrate and organic source to study the

denitrification kinetics. Under nitrate sources KNO3 and

Ca(NO3)2 and for organic CH3COONa and C6H12O6 were

used. The incubation experiments were carried out by

100 mL of solution (synthetic wastewater) containing

90 mL MSN media excluding nitrate and carbon source,

9 mL mixed culture and 1 mL of various concentrations of

nitrate source using incubation bottles (Borosil, 250 mL).

The nitrate and organic source concentrations were ranged

from 7 to 37 mg (absolute values) and 2.5–12.5 g/L

respectively. To maintain the anaerobic conditions, the

incubation bottles along with samples were initially flushed

with nitrogen gas. The bottles were fitted with air tight

rubber stopper. The head space of the bottles were vacu-

umed and again refilled with nitrogen gas. The gas and

liquid samples were drawn at regular interval (24 h) using

hypodermics syringe for N2O analysis along with various

water parameter studies. For all incubation studies, 0.07 g/L

inoculums were used. At 35 �C, the incubation studies were

carried out by putting the bottles in Julabo SW-22 shaking

incubators. The entire experiments were carried out for

7 days.

Gas sampling and analysis

Air samples were drawn through disposable syringe every

day at about 10:00 a.m. A gas chromatography (GC) of

make Shimadzu AA30 with electron capture detector

(ECD) was used to analyze N2O concentration. The GC is

equipped with auto gas sampler, semi-micro column, and

appropriate software to process the acquired data. The GC

was regularly standardized using NIST primary standard

gases. N2O in solution was determined by drawing known

amount of solution by a hypodermic needle and introduced

to a reactor under vacuum. After 5 min, the N2O stripped

was drawn and analyzed in GC as explained previously.

During N2O analysis in GC, the temperatures of column,

injector and ECD were 60�, 120�, and 350�, respectively.

The dissolved N2O in solution was estimated using Bunsen

absorption coefficient as shown below in Eq. 1 (Heincke

and Kaupenjohann 1999).

Y ¼ x� a Solution volume = head space volumeð Þ ð1Þ

where, a = 0.485, x is mass in head space, y is mass in

solution.

Each experiment was carried out in duplicate and

average value was taken for interpretation of results. The

variation of duplicate rate was within a range of ±5 %.

Water sample analysis

Parameters like nitrate nitrogen (N–NO3-), Nitrite (N–NO2

-),

ammonia (N–NH3), dissolved oxygen (DO), and COD were

analyzed following standard methods (Andert et al. 2011;

APHA, AWWA, WPCF 1998, 2005). pH of the samples was

measured using EUTECH-pH 1500 Meter.

Statistical analysis

Analysis of variance (ANOVA)

Two way ANOVA was used to determine whether signif-

icant role is played by the variables like days, nitrate

concentrations [KNO3 and Ca(NO3)2] and carbon source

(C6H12O6 and CH3COONa). Using Null hypothesis tech-

nique, the significance of the incubation parameters was

determined. Analysis of variance studies were carried out

using Microsoft office excel 2007 program. The raw

dataset were fed to the relevant excel program for carrying

out the detail statistical analyses.

Principal component analysis (PCA)

The incubation data were subjected to PCA to evaluate the

influence of various incubation parameters on the denitri-

fication rates. PCA was conducted using SPSS-10. In PCA,

Eigen values were used to determine the percentage of

variance as well as the cumulative percentage of variances.

A varimax rotation of different varifactors with factor

loading was calculated using Eigen values [1 and sorted

by the results having value [0.6 to have p \ 0.01.

Surface response modeling (optimization)

The incubation studies were carried out by statistically

designed experiments (Zahed et al. 2012; Rao et al. 2012;

Aber and Sheydaei 2012; Box et al. 1978). The principal

steps are determination of response variables, factors and

factor level, choice of experimental design, and statistically

analysis of data. For this purposes, three different sets of

experiments were carried out varying nitrate and organic

carbon source such as Set-1 [Ca(NO3)2 and CH3COONa],

Set-2 [Ca(NO3)2 and C6H12O6], and Set-3 (KNO3 and

CH3COONa). Design Expert v 8.0 was used for this work.

Removal of nitrate and COD

123

Author's personal copy

Equation’s format

All the equations written in this paper was been formatted

using Microsoft equation editor 2003, which is present in

MS-word 2003.

Results and discussions

Variation of nitrate concentration

Nitrate concentration varied from 7 to 37 mg NO3–N

(absolute values) to determine the N2O emission along with

denitrification rates. The denitrification rate for KNO3 was

observed to be more than CaNO3 may be due to more ionic

nature of the former. The decrease in nitrate concentration

in both the cases decreased progressively in time up to 4th

day, thereafter, decrease was very marginal, and the same

trend was observed for different initial nitrate concentra-

tions. The denitrification rate for different nitrate concen-

tration has been shown in Fig. 1. From Fig. 1, it can be

concluded that the rate of denitrification increased with

increase of initial nitrate concentration in both the cases.

The pH slowly increased during the incubation period. The

increase of pH is due to increase of alkalinity as shown

below in Eq. 2 (Park and Yoo 2009).

0:82CH3COOHþ NO�3 ! 0:07C5H7NO2 þ HCO�3þ 0:3CO2 þ 0:9H2Oþ 0:47N2

ð2ÞTheoretically for every 1 g of nitrogen consumed it

required 3.51 g of CH3COOH producing 0.57 and 4.36 g

Fig. 1 Evaluation of

denitrification and N2O

emission rates with respect to

variation of NO3 and carbon

source

Fig. 2 Behaviour of variables

like days, NO3–N, N2O–N,

NO2–N, COD, DO, pH, NH3

during incubation studies.

[Initial conditions Ca(NO3)2—

74.98 mg/L (NO3–N),

CH3COONa—7.5 g/L, Temp—

37 �C, pH—7)

S. Ray et al.

123

Author's personal copy

of biomass and alkalinity, respectively. The production of

alkalinity eventually increased the pH of the effluent.

Therefore, denitrification process is favored at higher pH

(Koren et al. 2000). The initial dissolved oxygen in the

media in all cases was around 8 mg/L. After the first day of

incubation, the DO was zero due to reaction of ammonia

with oxygen which may be due to the presence of aerobic

ammonium oxidizers (Sliekers et al. 2002; Feng et al.

2007) as shown in Eq. 3

NH3 þ 1:5O2 ! NO2 þ H2Oþ Hþ ð3Þ

The stoichiometry of the equation suggests that to

oxidize 1 g of NH3 it consumes 2.82 g of O2. Therefore, in

the present case, O2 availability for the oxidation reaction

is through DO as the denitrification process is carried out

under anaerobic conditions. The decrease of ammonia

concentration was commensurate with the decrease of DO.

After the first day the ammonia concentration decreased

due to oxidation to nitrate where nitrite is used as an

electron acceptor vide chemical Eq. 4 which may be due to

the presence of anaerobic ammonium oxidizers (Sliekers

et al. 2002; Feng et al. 2007)

NH3 þ 1:32NO�2 þ Hþ ! 1:02N2 þ 0:26NO�3 þ 2H2O

ð4Þ

The nitrite concentration progressively increased up to

3rd day of incubation and there after it showed a negative

trend. The N2O concentration both in the liquid and the gas

phase increased up to 4th day of incubation and there after

decreased. The initial N2O emission rate may be due to

formation of NO2- in the solution (Mathieu et al. 2006;

Wolf and Russow 2000) may be due to enzyme bound

NO2- (NO2

-)e (Wolf and Russow 2000). COD decreased

with time due to denitrification as well as formation of

biomass given by Eq. 2 (results not shown). Figure 2

shows a typical curve of all the variables (Days, N–N2O,

N–NO3-, N–NO2

-, DO, pH, COD, NH3) during

incubation period for a particular data set (Joo et al.

2007) as similar observations were observed in all other

cases. All these variables play an important role in

determining the rate kinetics in denitrification as well as

anaerobic ammonium oxidation process. N2O shown here

as a variable, was emitted as a byproduct during the

denitrification process (Wrage et al. 2001).

Since the rate of bacterial growth is directly proportional

to denitrification, the specific growth rate constant of

bacteria in lg can be calculated by plotting log N2O con-

centration (a metabolite) versus time for each dataset (Das

2002). The slope of the graph would give lg. The regen-

eration time, td, can be evaluated using Eq. 5

td ¼ 0:693=lg ð5Þ

The regeneration time for different experimental

conditions observed to vary between 1 and 3 days

depending on nitrate concentrations as well as organic

source. A faster regeneration time was observed in case of

Table 1 ANOVA for different parameters

Variables F p value F critical Remarks

CaNO3

Denitrification Days 17.59 1.02E-06 2.71 Significant

NO3 5.83 0.002816 2.87 Significant

N2O emission Days 7.64 0.000373 2.71 Significant

N–N2O 3.65 0.02163 2.87 Significant

KNO3

Denitrification Days 16.14 2.01E-06 2.71 Significant

NO3 5.12 0.005238 2.87 Significant

N2O emission Days 10.43 4.91E-05 2.71 Significant

N–N2O 10.95 7.20E-05 2.87 Significant

CH3COONa

Denitrification Days 46.72 1.28E-08 2.9 Significant

NO3 5.26 0.01117 3.29 Significant

N2O emission Days 21.87 2.12E-06 2.9 Significant

N–N2O 9.63 0.000862 3.29 Significant

C6H12O6

Denitrification Days 297.36 1.52E-19 2.49 Significant

NO3 7.91 0.001019 3.07 Significant

N2O emission Days 32.1 7.81E-10 2.49 Significant

N–N2O 16.74 8.74E-06 3.07 Significant

Removal of nitrate and COD

123

Author's personal copy

KNO3 as well as CH3COONa compared to Ca(NO3)2 and

C6H12O6 indicating that the dentrification was more in the

former.

Variation of organic source concentration

Two different types of organic carbon source such as

C6H12O6 and CH3COONa of various concentrations were

used to determine the denitrification as well as N2O

emission rate. The NO2- concentration varied in the range

0.2–5.7 mg (N–NO2-) when C6H12O6 concentration varied

in the range 2.5–12.5 g/L. For CH3COONa concentration

variation in the same range the NO2- concentration varied

from 0.85 to 9.51 mg (N–NO2-). The denitrification rate

increased with increase of organic source as shown in

Fig. 2. The denitrification rate in the presence of C6H12O6

was less compared to CH3COONa. The pH slowly

increased with time as observed in the previous case. The

DO decreases to ‘‘0’’ value after the first day (Fig. 2). The

nitrite concentration in the presence of C6H12O6 was less

than CH3COONa. The N2O emission rate in the presence

of C6H12O6 was also less due to formation of small amount

of nitrite during incubation reaction. N2O emission rate for

different organic source are shown in Fig. 1.

Statistical interpretation

Analysis of variance

Two way ANOVA studies was carried out in detail to find

out whether the variables like days of incubation, nitrate

concentration [KNO3 and Ca(NO3)2] and organic source

(C6H12O6 and CH3COONa) have any significant role or

not. The results are shown in Table 1. From Table 1, it can

be concluded that time, nutrient concentration (nitrate and

organic source) and emission of N2O had a significant

variation during incubation studies.

PCA

PCA were carried out to determine the intra correlationship

between different variables. The correlation matrix is

Table 2 Principal component analysis

Correlation matrix

Correlation Days N2O total NO3 NO2 DO pH NH3 N2O aqueous N2O gas

Days 1

N2O total 0.21 1

NO3 -0.79 -0.35 1

NO2 0.11 -0.12 0.06 1

DO -0.59 -0.43 0.53 -0.25 1

pH 0.09 0.26 -0.24 -0.81 0.09 1

NH3 -0.32 0.09 0.11 -0.59 0.37 0.67 1

N2O aqueous 0.16 -0.03 0.01 0.97 -0.27 -0.78 -0.59 1

N2O gas 0.32 0.8 -0.41 -0.16 -0.39 0.33 0.09 -0.1 1

Component matrix (a) Component

1 2 3

Days 0.77

N2O total 0.72 0.61

NO3 -0.82

NO2 -0.94

DO -0.75

pH 0.89

NH3 0.8

N2O aqueous -0.93

N2O gas 0.76

Eigen values 3.41 3.02 1.14

Cumvar (%) 37.85 71.39 84.01

S. Ray et al.

123

Author's personal copy

shown in Table 2. Day’s incubation time showed a nega-

tive correlation with nitrate and DO as the reaction

involves denitrification which was being carried out under

anaerobic conditions. It showed positive correlation with

N2O which indicates that the gas emission rate would

increase with time. The emission of N2O total (gas and

liquid) showed a negative correlation with nitrate and DO.

This is obvious as emission of N2O depends on the deni-

trification rate. Nitrate has a positive correlation with DO

which indicates that denitrification is not favored at higher

DO. pH showed a positive correlation with NH3 whereas

NH3 showed a negative correlation with N2O. This may be

due to denitrification process which involves alkalification

and NH3 is not involved in the formation of N2O,

respectively.

The PCA is shown in Table 2 contains three compo-

nents. Component-1 explained 37.85 % with Eigen value

of 3.41. It contain variable like NO2, pH, NH3, and N2O

aqueous. NO2 and N2O aqueous showed negative correla-

tion with each other which indicated that NO2 had a

positive effect on N2O emission. Since component-1 con-

sists of emission of N2O along with variables like pH, NH3,

Table 3 Design summary

Set-1: Ca(NO3)2 + CH3COONa

Study type: response surface; design type: central composite; design model: quadratic; run: 20; build time (ms): 28.09

Factor Name Units Type Subtype Minimum Maximum Mean Std. Dev.

A N–NO3- mg/L Numeric Continuous 74.32 299.68 187 55.36

B COD g/L Numeric Continuous 2.45 12.55 7.5 2.48

C Days Numeric Continuous 0.64 7.36 4 1.65

Response Name Units Obs Analysis Minimum Maximum Mean Std. Dev. Trans Model

Y1 N–NO3- % removal 20 Polynomial 0.83 99.85 78.46 34.70 None Quadratic

Y2 COD % removal 20 Polynomial 70.29 96.74 78.42 6.67 None Quadratic

Y3 N–N2O lg 20 Polynomial 0.06 2.49 0.75 0.63 Base 10 Log Quadratic

Set-2: Ca(NO3)2 + C6H12O6

Study type: response surface; design type: central composite; design model: quadratic; run: 20; build time (ms): 2.93

Factor Name Units Type Subtype Minimum Maximum Mean Std. Dev.

A N–NO3- mg/L Numeric Continuous 74.32 299.68 187 55.36

B COD g/L Numeric Continuous 2.45 12.55 7.5 2.48

C Days Numeric Continuous 0.64 7.36 4 1.65

Response Name Units Obs Analysis Minimum Maximum Mean Std. Dev. Trans Model

Y1 N–NO3- % removal 20 Polynomial 37.5 83.96 68.93 14.01 None Quadratic

Y2 COD % removal 20 Polynomial 0 58.85 16.49 13.67 Square Root Quadratic

Y3 N–N2O lg 20 Polynomial 0.1 0.70 0.34 0.18 None Linear

Set-3: KNO3 + CH3COONa

Study type: response surface; design type: central composite; design model: quadratic; run: 20; build time (ms): 2.78

Factor Name Units Type Subtype Minimum Maximum Mean Std. Dev.

A N–NO3- mg/L Numeric Continuous 75.32 374.68 225 73.54

B COD g/L Numeric Continuous 2.45 12.55 7.5 2.48

C Days Numeric Continuous 0.64 7.36 4 1.65

Response Name Units Obs Analysis Minimum Maximum Mean Std. Dev. Trans Model

Y1 N–NO3- % removal 20 Polynomial 11.11 99.68 77.01 33.04 None Quadratic

Y2 COD % removal 20 Polynomial 69.43 95.33 77.51 6.41 None Linear

Y3 N–N2O lg 20 Polynomial 0.20 3.00 1.09 0.78 Base 10 Log Quadratic

Removal of nitrate and COD

123

Author's personal copy

and NO2, therefore, this component can be termed as N2O

emission factor. Component-2 explained 71.4 % cumula-

tive variance with Eigen value of 3.02. This component

contains variables like days, incubation time, N2O (total),

nitrate, DO, and N2O gas. All the variables are positively

correlated except nitrate and DO. The positive correlation

with each other suggested that N2O emission would

increased with incubation time. Since factor-2 contains the

various variables, therefore, it can be termed as N2O var-

iable factor. Factor-3 explained 84.01 % with Eigen value

of 1.14. It contains only one variable that is N2O total.

Incubation studies using RSM

As explained earlier denitrification studies were carried out

by varying parameters i.e., nitrate and organic source.

Samples were collected at regulars to analyze the param-

eters such as N–NO3-, COD, N–NO2

-, N–N2O, N–NH3,

and DO. Among all these parameters the concentrations of

N–NO3-, COD, N–N2O and days were taken into consid-

eration for RSM studies. The main aim is to maximize the

removal efficiency of the responses like N–NO3-, COD

and simultaneously minimize N–N2O emission as it is one

of the GHG. Keeping in view of the above some efforts

have been initiated to optimize the overall process through

RSM. Coding was done to reduce the range of each factor

to a common scale, regardless of the magnitude, the typical

scheme being to set-1 as lower level, ?1 as upper level and

0 as middle level.

Fig. 3 Predicted versus actual plot for SET-1 and SET-3 responses.

Conditions SET-1 [CH3COONa and Ca(NO3)2] and SET-3 (CH3CO-

ONa and KNO3)

Fig. 4 Actual versus predicted plot for SET-2 responses. ConditionsSET-2 [C6H12O6 and Ca(NO3)2]

Table 4 Best fit model

Responses Coded factors Model

F value

p value

prob [ FR2 Adeq

precision

Set-1 N–NO3- N–NO3

- = 96.97 - 0.87A - 4.2B + 33.38C + 2.53AB -

0.73AC + 7.04BC - 4.99A2 - 2.43B2 - 19.69C219.53 \0.0001 0.95 15.53

COD COD = 79.42 + 2.21A - 5.24B + 3.39C -

0.68AB + 2.33AC - 1.54BC - 1.22A2 + 1.66B2 - 1.90C215.71 \0.0001 0.93 14.98

N–N2O Log10(N–N2O) = 0.11 + 0.26A - 0.06B - 0.17C -

0.05AB - 0.004AC - 0.16BC - 0.18A2 - 0.11B2 -

0.33C2

7.93 0.0016 0.88 7.74

Set-2 N–NO3- N–NO3

- = 79.36 + 1.21A + 0.37B + 9.43C + 2.78AB -

2.89AC - 2.45BC - 6.88A2 - 5.61B2 - 2.78C22.23 0.11 0.67 4.68

COD Sqrt(COD) = 3.81 + 1.22A -

1.45B + 0.24C + 0.21AB + 0.21AC - 0.56BC -

0.55A2 + 0.34B2 - 0.01C2

29.22 \0.0001 0.96 20.89

N–N2O N–N2O = 0.33 + 0.16A + 0.03B - 0.04C 8.51 0.0013 0.61 9.8

Set-3 N–NO3- N–NO3

- = 96.64 - 6.13A - 0.31B + 30.80C - 0.43AB -

2.57AC + 0.45BC - 12.65A2 - 0.34B2 - 15.75C211.89 0.0003 0.91 11.44

COD COD = 77.51 + 1.95A - 4.63B + 2.23C 6.011 0.01 0.53 8.24

N–N2O Log10(N–N2O) = 0.25 + 0.24A + 0.05B + 0.08C -

0.01AB + 0.003AC + 0.08BC - 0.17A2 - 0.09B2 -

0.24C2

5.16 0.01 0.82 5.75

S. Ray et al.

123

Author's personal copy

Central composite design with two labeled factorial

points such as axial and center was done (Rao et al. 2012;

Rath et al. 2012). The axial point is having all factors set to

zero, except one factor which has the value ± a. a represents

the distance from the center of the designed space to an axial

point. Since, in this study, the factors are \5, therefore,

rotatable models were chosen which correspond the alpha

value 1.68. The factorial points such as N–NO3-, COD and

days were considered as input value to analyze the responses

like N–NO3-, COD, and N–N2O. The designed summary is

shown in Table 3. The actual versus predicted value for the

responses are shown in Figs. 3 and 4. Table 4 shows the final

equation in terms of coded factors and respective F values,

Prob [ F values, R2 values and adequate precision along

with model fit. In Set-1 all the responses fits to the quadratic

model. In case of COD, the model F value was observed to be

15.71 which describe the model to be significant. The

Prob [ F\0.0001 indicates the model term to be significant.

Adequate precision were undertaken so that the measure of

signal to noise ratio, should be[4. So in conclusions, all the

statistical values are significant which shows the validity of

the model (Rath et al. 2012). Similar values were obtained for

all other responses as well as Sets. A number of solutions

were available but the maximum desirability was chosen.

Fig. 5 The contour and three dimensional plots showing the optimization of three responses such as N–NO3-, COD and N–N2O for SET-1(a–c),

SET-2(d–f) and SET-3(g–i)

Table 5 Optimum conditions obtained during the experiments

Factors Response

N–NO3- (mg/L) COD (g/L) Days N–NO3

- (% removal) COD (% removal) N–N2O (lg) Desirability

Set-1 247.39 4.50 6.00 97.6133 93.066 0.696471 0.899514

Set-2 205.61 4.50 6.00 80.6963 44.7124 0.31086 0.820904

Set-3 314.00 4.50 6.00 90.2832 86.3296 0.90384 0.807945

Removal of nitrate and COD

123

Author's personal copy

In the present studies, three responses were considered

such as N–N2O, N–NO3-, and COD along with three

factors like days, COD and N–NO3-. To evaluate the

overall design, the effect of each factor on the particular

response was considered using the contour with 3D plots.

The contour and 3D plots using the quadratic model was

developed by keeping one variable constant at a time. The

optimized contour with 3D plots for responses like

N–NO3-, N–N2O, and COD are shown in Fig. 5a–h for Set-1

(a–c), Set-2 (d–f) and Set-3 (g–i). All the input variables

like nitrate and organic carbon source concentration were

shown in Table 3. Figure 5a shows the contour with 3D

plot for two variables like N–NO3- and days keeping the

other variable COD constant at 4.50 g/L. The % of

N–NO3- removal increased with increase in days. 100 %

N–NO3- removal could be obtained at N–NO3

- concentration

and days ranged between 120–240 mg/L and 4–6 days,

respectively. Figure 5b shows COD removal increased

with increase in N–NO3- concentration and days. 90 %

COD removal could be obtained at N–NO3- concentration

and days ranged between 195–254 mg/L and 4.7–6 days,

respectively. Figure 5c shows 1.5 lg of N–N2O could be

obtained at N–NO3- concentration and days ranged

between 230–254 mg/L and 3.7–4.3 days. Similar expla-

nations could be given for two other sets namely set-2 and

set-3. In Set-1, the optimum values of the process time,

initial N–NO3- concentration and initial COD concentra-

tion in the aqueous solution were found to be 6 days,

247.39 mg/L, 4.50 g/L, respectively. During the optimal

values of the process parameters, a maximum N–NO3-

removal of 97.61 %, and COD removal of 93.07 % was

obtained. Also, the N–N2O formation was minimized to

2.96 lg. A maximum desirability of 0.89 is obtained in

Set-1 having Ca(NO3)2 as nitrate source and CH3COONa

as carbon source. Desirability is an objective function and

varies between 0 and 1 (Rao et al. 2012). Higher the

desirability value indicates the closeness between esti-

mated and experimental values. In the present case, the

desirability values were nearly 0.9 in all the three cases

which unequivocally proves the adequacy of the model.

The optimum conditions obtained during the experiments

along with desirability value is shown in Table 5.

Conclusions

The denitrification rates varied between 1.9 and 9.1 mg

N/day for nitrate as well as organic concentration. The

denitrification rate observed to be higher in the presence of

KNO3 as well as CH3COONa compared to Ca(NO3)2 and

C6H12O6. The N2O emission rate was varied from 0.12 to

0.82 lg N/day for nitrate along with carbon concentrations.

In the presence of CH3COONa and KNO3, the N2O

emission rate was observed to be higher. Detailed statistical

analysis like ANOVA and PCA were carried out. Two way

ANOVA analyses were carried out, and it was observed

that the variables like NO3-, incubation period, and

organic source played a vital role on the emission of N2O.

The PCA showed that the variables can be accommodated

in three components explaining 84 % of total cumulative

variance and the Eigen value of component 1, 2, and 3 are

3.41, 3.02, and 1.14, respectively. The optimization studies

were carried out to minimize N-N2O emission and maxi-

mize the N–NO3- along with COD removal through

denitrification process. In Set-1 under optimized conditions

(N–NO3-—247.39 mg/L, COD—4.5 g/L, time—6 days),

N–NO3- and COD removal efficiency were observed to be

97.61 and 93.07 %, respectively. N–N2O emissions were

minimized to 0.69 lg. Maximum N–NO3- and COD

removal was obtained in Set-1 in comparison with Set-2

and Set-3. The desirability factor for all the three Sets of

experiment were ranged from 0.81 to 0.9.

Acknowledgments The authors were thankful to Director, CSIR-

IMMT, for his kind permission to publish this paper. One of the

author (SR) thanks DST, New Delhi, for the award of INSPIRE

Fellowship. The authors are thankful to MoEF, New Delhi, for the

financial support without which the work could not have been done.

The authors highly acknowledge the anonymous reviewers for their

valuable suggestions, without which the manuscript may not have

come to this stage.

References

Aber S, Sheydaei M (2012) Removal of COD from industrial effluent

containing indigo dye using adsorption method by activated

carbon cloth: optimization, kinetic, and isotherm studies. Clean

Soil Air Water 40(1):87–94

Andert J, Wessen E, Borjesson G, Hallin S (2011) Temporal changes

in abundance and composition of ammonia-oxidizing bacterial

and archaeal communities in a drained peat soil in relation to

N2O emissions. J Soils Sediments 11:1399–1407

APHA (2005) Standard methods for the examination of water and

wastewater. American Public Health Association, American

Water Works Association and Water Environment Federation,

Washington

APHA, AWWA, WPCF (1998) Standard methods for the examina-

tion of water and wastewater, 20th edn. American Public Health

Association, Washington DC

Bergstermann A, Cardenas L, Bol R, Gilliam L, Goulding K, Meijide

A, Scholefield D, Vallejo A, Well R (2011) Effect of antecedent

soil moisture conditions on emissions and isotopologue distri-

bution of N2O during denitrification. Soil Biol Biochem 43:

240–250

Box GEP, Hunter JS, Hunter WG (1978) An introduction to design

data analysis and model building. Wiley, New York

Dai Z, Trettin CC, Li C, Li H, Sun G, Amatya DM (2012) Effect of

assessment scale on spatial and temporal variations in CH4, CO2

and N2O fluxes in a forested wetland. Water Soil Air Pollut

223:253–265

Das T (2002) Bio oxidation of iron using Thiobacillus ferrooxidans.

Doctoral Dissertation Thesis, Utkal University

Faulwetter JL, Gagnon V, Sundberg C, Chazarenc F, Burr MD,

Brisson J, Camper AK, Stein OR (2009) Microbial process

S. Ray et al.

123

Author's personal copy

influencing performance of treatment wetlands: a review. Ecol

Eng 35:987–1004

Feng YJ, Tseng SK, Hsia TH, Ho CM, Chou WP (2007) Partial

nitrification of ammonium-rich waste water as pretreatment for

anaerobic ammonium oxidation (anammox) using membrane

aeration bioreactor. J Biosci Bioeng 3:182–187

Firestone MK (1982) Biological denitrification. In Stevenson FJ (ed)

Nitrogen in agricultural soils. Agronomy 22:289–326

Foley J, De Hass D, Yuan Z, Lant P (2010) Nitrous oxide generation

in full-scale biological nutrient removal wastewater treatment

plants. Water Res 44:831–844

Gogoi B, Baruah KK (2012) Nitrous oxide emissions from fields with

different wheat and rice varieties. Pedosphere 22(1):112–121

Heincke M, Kaupenjohann M (1999) Effects of soil solution on the

dynamics of N2O emissions: a review. Nutr Cycl Agroecosys

55:133–157

Hu Z, Lee JW, Chandran K, Kim S, Khanal SK (2012) Nitrous oxide

(N2O) emission from aquaculture: a review. Environ Sci Technol

46:6470–6480

IPCC (2007) Climate changes: the scientific basis. Contribution of

working group 1 to the third assessment report of the intergov-

ernmental panel on climate change. Cambridge University Press,

Cambridge

Joo HS, Hirai M, Shoda M (2007) Improvement in ammonium

removal efficiency in wastewater treatment by mixed culture of

Alcaligenes faecalis No. 4 and L1. J Biosci Bioeng 103(1):66–73

Kampschreur MJ, Temmink H, Kleerebezem R, Jetten MSM, Van

Loosdrecht MCM (2009) Nitrous oxide emission during waste-

water treatment. Water Res 43:4093–4103

Koren DW, Gould WD, Bedard P (2000) Biological removal of

ammonia and nitrate from simulated mine and mine effluents.

Hydrometallurgy 56:127–144

Laverman AM, Garnier JA, Mounier EM, Roose-Amsaleg CL (2010)

Nitrous oxide production kinetics during nitrate reduction in

river sediments. Water Res 44:1753–1764

Marinho EVA, Delaune RD, Lindau CW (2004) Nitrous oxide flux

from soybeans grown on Mississippi alluvial soil. Commun Soil

Sci Plant Anal 35(1–2):1–8

Mathieu O, Leveque J, Henault C, Milloux MJ, Bizouard F, Andreux

F (2006) Emission and spatial variability of N2O, N2 and nitrous

oxide mole fraction at the field scale, revealed with 15N isotopic

techniques. Soil Biol Biochem 38:941–951

Metcalf and Eddy (2003) Wastewater engineering treatment and

reuses, 4th ed. McGraw-Hill, New York

Nelson LM, Knowles R (1978) Effect of oxygen and nitrate on

nitrogen fixation and denitrification by Azospirillum brasilensegrown in continuous culture. Can J Microbiol 24:1395–1403

Pant D, Van Bogaert G, Diels L, Vanbroekhoven K (2010) A review of

the substrates used in microbial fuel cells (MFCs) for sustainable

energy production. Bioresour Technol 101(6):1533–1543

Park JY, Yoo YJ (2009) Biological nitrate removal in industrial waste

water treatment: which electron donor we can choose: mini-

review. Appl Microbiol Biotechnol 82:415–429

Rabaey K, Verstraete W (2005) Microbial fuel cells: novel biotech-

nology for energy generation. Trends Biotechnol 23:291–298

Ramulu TS, Sahoo SK, Baral SS, Das SN, Swamy YV, Roy

Chaudhury G (2008) Estimation of nitrous oxide emission from

pulse cultivation in rain-fed uplands. Curr Sci 95(5):584–585

Ramulu TS, Sahoo SK, Mohapatra RK, Roy Chaudhury G, Das SN

(2009) Emission of N2O on pulse–rice crop rotation in upland.

Atmos Environ 43(97):2087–2090

Rao KS, Anand S, Rout K, Venkateswarlu P (2012) Response surface

optimization for removal of cadmium from aqueous solution by

waste agricultural biosorbent Psidium guvajava L. leaf powder.

Clean Soil Air Water 40(1):80–86

Rath SS, Nayak P, Mukherjee PS, Roy Chaudhury G, Mishra BK

(2012) Treatment of electronic waste to recover metal values

using thermal plasma coupled with acid leaching—a response

surface modeling approach. Waste Manag 32(3):575–583. doi:

10.1016/j.wasman.2011.11.001

Sahoo SK, Ramulu TS, Kumar M, Roy Chaudhury G, Das SN (2010)

Emission of N2O from farmer’s field during black gram

cultivation and incubation experiments in laboratory. Toxicol

Environ Chem 92:1043–1052

Sliekers AO, Derwort N, Gomez JLC, Strous M, Kuenen JG, Jetten

MSM (2002) Completely autotrophic nitrogen removal over

nitrite in one single reactor. Water Res 36:2475–2482

Tallec G, Garnier J, Billen G, Gousailles M (2008) Nitrous oxide

emissions from denitrifying activated sludge of urban waste

water treatment plants, under anoxia and low oxygenation.

Bioresour Technol 99(7):2200–2209

Tavares P, Pereira AS, Moura JJG, Moura I (2006) Mettalloenzymes

of the denitrification pathway. J Inorg Biochem 100:2087–2100

Terai H, Mori T (1975) Studies on phosphorylation coupled with

denitrification and aerobic respiration in Pseudomonas denitrif-icans. Bot Mag 38:231–244

Toma Y, Takakai F, Darung U, Kuramochi K, Limin SH, Dohong S,

Hatano R (2011) Nitrous oxide emission derived from soil organic

matter decomposition from tropical agricultural peat soil in central

Kalimantan, Indonesia. Soil Sci Plant Nutr 57(3):436–451

Wolf I, Russow R (2000) Different pathways of formation of N2O, N2

and NO in black earth soil. Soil Biol Biochem 32:229–239

Wrage N, Velthof GL, Van Beusichem ML, Oenema O (2001) Role

of nitrifier denitrification in the production of nitrous oxide. Soil

Biol Biochem 33:1723–1732

Yang WH, The YA, Silver W (2011) A test of a field-based 15N-

nitrous oxide pool dilution technique to measure gross N2O

production in soil. Glob Chang Biol 17:3577–3588

Zahed MA, Aziz HA, Isa MH, Mohajeri L (2012) Response surface

analysis to improve dispersed crude oil biodegradation. Clean

Soil Air Water 40:262–267

Removal of nitrate and COD

123

Author's personal copy