USF Libraries
USF Digital Collections

Investigation of the optimal dissolved co2 concentration and ph combination for the growth of nitrifying bacteria

MISSING IMAGE

Material Information

Title:
Investigation of the optimal dissolved co2 concentration and ph combination for the growth of nitrifying bacteria
Physical Description:
Book
Language:
English
Creator:
Morris, Raymond Anthony
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Ammonium
Aob
Carbon Dioxide
Fish
Nob
Dissertations, Academic -- Environmental Engineering Microbiology -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Ammonium (NH4+) is a biological nutrient that is transformed in a wastewater treatment plant (WWTP) in a process called activated sludge. This is accomplished in an aerobic environment using microorganisms and inorganic carbon that convert the ammonium to nitrate (NO3-). This process is termed nitrification. Removal of ammonium is necessary due to its oxygen demand and toxicity to the environment. Nitrification is considered a slow process due to the slow growth rate of the nitrifying bacteria. Ammonia oxidizing bacteria (AOB) first covert the ammonium (NH4+) to nitrite (NO2-) followed by conversion to nitrate (NO3-) by nitrite oxidizing bacteria (NOB). These slow rates limit the treatment capacity of the WWTP. The initial hypothesis suggested that these slow rates were due to limited carbon in the aeration basin of a WWTP. A series of designed experiments and observational studies revealed substantial dissolved CO2 exists throughout a WWTP. Based on these findings, the central research focused on determining if an optimum dissolved CO2 concentration/ pH combination exists that maximizes nitrification. Experimentation conducted at a pH of 7.0 and varying concentrations of dissolved CO2 concentration revealed inhibition at low (<5 mg/l) and high (>30 mg/l) dissolved CO2 concentration levels. Further research found that optimum nitrification can be attained in a dissolved CO2 concentration range of 10 - 15 mg/l and a pH range of 7.5 - 8.0. A maximum specific growth rate of 1.05 - 1.15 days-1 was achieved. A partitioning of the sums of squares from these designed experiments found that pH accounts for approximately 83 percent of the sums of squares due to treatment with the dissolved CO2 concentration accounting for 17 percent. This suggests that pH is the dominant factor affecting nitrification when dissolved CO2 concentration is optimized. Analysis of the growth kinetics for two of the designed experiments was conducted. However, a set of parameters could not be found that described growth conditions for all operating conditions. Evaluating the results from these two experiments may suggest that a microbial population shift occurred between 16 and 19 mg/l of dissolved CO2 concentration. These dissolved CO2 concentrations represent pH values of 7.1 and 7.0, respectively, and were compared to experimentation conducted at a pH of 7.0. Though the pH difference is minor, in combination with the elevated dissolved CO2 concentration, a microbial shift was hypothesized. Microbial samples were collected from the designed experiment that optimized dissolved CO2 concentration (5, 10 and 15 mg/l) and pH (6.5, 7.0, 7.5 and 8.0). These samples were evaluated using Fluorescence in situ hybridizations (FISH) to determine the population density of common ammonium oxidizing bacteria (AOB) (Nitrosomonas and Nitrosospira) and nitrite oxidizing bacteria (NOB) Nitrobacter and Nitrospirae). The dominant AOB and NOB microbes were found to be Nitrosomonas and Nitrospirae. These results suggest that increased nitrification rates can be achieved by incorporating appropriate controls in a wastewater treatment plant (WWTP). With higher nitrification rates, lower nitrogen values can be obtained which will reduce the WWTP effluent nitrogen concentration. Conversely, these increased nitrification rates can also reduce the volume of an aeration basin given similar effluent nitrogen concentrations.
Thesis:
Disseration (Ph.D.)--University of South Florida, 2011.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Raymond Anthony Morris.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 182 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0004964
usfldc handle - e14.4964
System ID:
SFS0028201:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2011 flu ob 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004964
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Morris, Raymond Anthony.
0 245
Investigation of the optimal dissolved co2 concentration and ph combination for the growth of nitrifying bacteria
h [electronic resource] /
by Raymond Anthony Morris.
260
[Tampa, Fla] :
b University of South Florida,
2011.
500
Title from PDF of title page.
Document formatted into pages; contains 182 pages.
Includes vita.
502
Disseration
(Ph.D.)--University of South Florida, 2011.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
520
ABSTRACT: Ammonium (NH4+) is a biological nutrient that is transformed in a wastewater treatment plant (WWTP) in a process called activated sludge. This is accomplished in an aerobic environment using microorganisms and inorganic carbon that convert the ammonium to nitrate (NO3-). This process is termed nitrification. Removal of ammonium is necessary due to its oxygen demand and toxicity to the environment. Nitrification is considered a slow process due to the slow growth rate of the nitrifying bacteria. Ammonia oxidizing bacteria (AOB) first covert the ammonium (NH4+) to nitrite (NO2-) followed by conversion to nitrate (NO3-) by nitrite oxidizing bacteria (NOB). These slow rates limit the treatment capacity of the WWTP. The initial hypothesis suggested that these slow rates were due to limited carbon in the aeration basin of a WWTP. A series of designed experiments and observational studies revealed substantial dissolved CO2 exists throughout a WWTP. Based on these findings, the central research focused on determining if an optimum dissolved CO2 concentration/ pH combination exists that maximizes nitrification. Experimentation conducted at a pH of 7.0 and varying concentrations of dissolved CO2 concentration revealed inhibition at low (<5 mg/l) and high (>30 mg/l) dissolved CO2 concentration levels. Further research found that optimum nitrification can be attained in a dissolved CO2 concentration range of 10 15 mg/l and a pH range of 7.5 8.0. A maximum specific growth rate of 1.05 1.15 days-1 was achieved. A partitioning of the sums of squares from these designed experiments found that pH accounts for approximately 83 percent of the sums of squares due to treatment with the dissolved CO2 concentration accounting for 17 percent. This suggests that pH is the dominant factor affecting nitrification when dissolved CO2 concentration is optimized. Analysis of the growth kinetics for two of the designed experiments was conducted. However, a set of parameters could not be found that described growth conditions for all operating conditions. Evaluating the results from these two experiments may suggest that a microbial population shift occurred between 16 and 19 mg/l of dissolved CO2 concentration. These dissolved CO2 concentrations represent pH values of 7.1 and 7.0, respectively, and were compared to experimentation conducted at a pH of 7.0. Though the pH difference is minor, in combination with the elevated dissolved CO2 concentration, a microbial shift was hypothesized. Microbial samples were collected from the designed experiment that optimized dissolved CO2 concentration (5, 10 and 15 mg/l) and pH (6.5, 7.0, 7.5 and 8.0). These samples were evaluated using Fluorescence in situ hybridizations (FISH) to determine the population density of common ammonium oxidizing bacteria (AOB) (Nitrosomonas and Nitrosospira) and nitrite oxidizing bacteria (NOB) Nitrobacter and Nitrospirae). The dominant AOB and NOB microbes were found to be Nitrosomonas and Nitrospirae. These results suggest that increased nitrification rates can be achieved by incorporating appropriate controls in a wastewater treatment plant (WWTP). With higher nitrification rates, lower nitrogen values can be obtained which will reduce the WWTP effluent nitrogen concentration. Conversely, these increased nitrification rates can also reduce the volume of an aeration basin given similar effluent nitrogen concentrations.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor:
Stroot, Peter G.
653
Ammonium
Aob
Carbon Dioxide
Fish
Nob
690
Dissertations, Academic
z USF
x Environmental Engineering Microbiology
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4964



PAGE 1

Investigation of the Optimal Dissolved CO2 Concentration and pH Combination for the Grow th of Nitrifying Bacteria by Raymond A. Morris A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Civil and Environmental Engineering College of Engineering University of South Florida Major Professor: Peter G. Stroot, Ph.D. James R. Mihelcic, Ph.D. James R. Garey, Ph.D. Scott W. Campbell, Ph.D. Daniel H. Yeh, Ph.D. Date of Approval: April 7, 2011 Keywords: Carbon Dioxide, AOB, NOB, Ammonium, FISH Copyright 2011, Raymond A. Morris

PAGE 2

Dedication This dissertation is dedicated to my wif e, Sandra Kay Morris, who without her support and understanding, this research and document would not have become a reality.

PAGE 3

Acknowledgements I would like to my major professor, Dr. Peter Stroot, for research guidance and support while conducting this research. Mathew Cutter, Samuel DuPont, Kathry n Bailey, Andrea Rocha for general lab support and assistance when conducting FISH studies. A special thank you for Steve Heppler and Micah Smith for their reactor studies support. Their assistance allowed me to occasionally receive some sleep. Dr. Jeffrey Cunningham & Dr. Daniel Yeh for allowing me to participate in their classes and receive enhanced learning opportunities. The Northside and Glendale wa stewater treatment plant s in Lakel and, FL and the South Cross Bayou wast ewater treatment plant in St. Petersburg, FL for activated sludge samples used during my research.

PAGE 4

A very special thank you for Dr. Fr ank Young, chairman of the engineering department at the University of Sout h Florida – Polytechnic campus, for employment support and car eer opportunities while conducting this research.

PAGE 5

i Table of Contents List of Tables iv List of Figures vi Abstract x Chapter 1: Research Objective 1 1.1 Main Objective 1 1.2 Research Goals 1 1.3 Hypothesis and Approach 2 Chapter 2: Wastewater Treatment Industry, Literature Review and Preliminary Research 4 2.1 Wastewater Treatment in the Un ited States 4 2.2 Biological Nutrient Removal (BNR) Systems 5 2.2.1 BNR Wastewater Treatment Processes 6 2.2.1.1 Total Nitrogen Removal Only 6 2.2.1.2 Total Nitrogen and Total Phosph orus Removal 8 2.2.1.3 Wastewater Treat ment Plant Configurati ons 10 2.3 Nitrifying Bacteria and Nitri fication 15 2.4 Heterotrophic Bacteria, C hemical Oxygen Demand (COD) and Ammonium Removal 27 2.5 Carbon Dioxide and Wastewater Tr eatment Plants 29 2.6 Substrate Utilization in Wastewater Treatment Plants 35 2.7 Estimation of the Maximum Specific Growth Rate, max, from NOx Generation Rate in Batch R eactor 37 2.8 Carbon Dioxide and Nitrificati on 39 2.9 Preliminary Research 41 Chapter 3: Stimulation of Nitri fication by Carbon Dioxide in Lab-Scale Activated Sludge Reac tors 43 3.1 Abstract 43 3.2 Keywords 44 3.3 Introduction 44 3.4 Materials and Methods 46 3.4.1 Experiment 1 46 3.4.2 Experiments 2-4 48 3.4.3 Data Collection and Sample Analyses 51

PAGE 6

ii 3.5 Results 54 3.5.1 Experiment 1 54 3.5.2 Experiment 2 56 3.5.3 Experiment 3 57 3.5.4 Experiment 4 59 3.6 Discussion 64 3.6.1 Effect of pH on Nitrifica tion 64 3.6.2 Nitrification in Activated Sludge Systems 64 3.7 Conclusions 66 Chapter 4: Evaluation of Nitrifying Bacteria Specific Growth Rate Sensitivity to Carbon Di oxide for Full-Scale Activated Sludge and Municipal Wa stewater 67 4.1 Abstract 67 4.2 Keywords 68 4.3 Introduction 68 4.4 Methodology 69 4.4.1 Field Evaluation of Nitrification in Three BNR Systems 69 4.4.2 pH vs. Dissolved CO2 70 4.4.3 Specific Growth Rate Measurement in Lab-Scale Bioreactors 70 4.4.4 Estimation of Specific Growth Rate of Nitrifying Bacter ia 71 4.4.5 Evaluation of Nitrify ing Bacteria Abundance by Fluorescence in situ Hybridization 73 4.5 Results 75 4.5.1 Field Evaluation of Three BNR Systems 75 4.5.2 Estimation of Specific Growth Rate of Nitrifying Bacter ia 79 4.5.3 Evaluation of the Specific Growth Rate of Nitrifying Bacteria Sensitivity to Dissolved CO2 Concentration using Lab-Scale Bi oreactors 80 4.5.4 Evaluation of Nitrifyi ng Bacteria by Fluorescence in situ Hybridization 82 4.6 Discussion 86 4.7 Conclusions 90 Chapter 5: Determination of t he Relationship of Dissolved CO2 Concentration and pH and a Design Space for Optimum Nitrification 91 5.1 Methodology and Materials 91 5.1.1 Data Collection and Sample Analyses 96 5.2 Results 98 5.2.1 Experiment 1 98 5.2.2 Experiment 2 100 5.2.3 Growth Kinetics 102 5.2.4 Experiment 3 104 5.3 Discussion 107

PAGE 7

iii 5.3.1 Experiment 1 107 5.3.2 Experiment 2 108 5.3.2.1 Effect of pH on Nitri fication 110 5.3.2.2 Growth Kinetics 111 5.3.2.3 Nitrification in Activa ted Sludge Systems 112 5.3.3 Experiment 3 113 5.4 Conclusions 116 Chapter 6: FISH Analysis of Mi crobial Samples Co llected from Batch Reactors Operated at Different Dissolved CO2 Concentrations and pH 118 6.1 Introduction 118 6.2 Methods and Materials 119 6.3 Results 121 6.3.1 AOB Results 124 6.3.1.1 Nitrosomonas 127 6.3.1.2 Nitrosospira 129 6.3.2 NOB Results 130 6.3.2.1 Nitrobacter 133 6.3.2.2 Nitrospirae 135 6.3.3 Validation Study of FISH Results 136 6.3.4 Biomass Growth Determination 144 6.4 Discussion 146 6.5 Conclusions 151 Chapter 7: Conclusions 154 References: 157

PAGE 8

iv List of Tables Table 2-1: Comparison of Common BNR Proc ess Configurations 9 Table 2-2: Effluent TN Components and Ac hievable Limits 20 Table 2-3: Properties of Predominant Nit rifying Bacteria 22 Table 2-4: Specific Growth Rate at Selected pCO2 and pH 42 Table 3-1: Description of Experiments 2 th rough 4 Conducted in a SBR 50 Table 3-2: Average Solids Concentration – Experiment 4 62 Table 4-1: Constants Used to Ca lculate the Optimal Specific Growth Rate for Nitrifying Ba cteria 72 Table 4-2: FISH Probe Informa tion 74 Table 4-3: Dissolved CO2 Concentration and pH of Influent, Unit Processes, and Effluent of Five Wastewater Treatment Plants 76 Table 4-4: Influent Properties and Activated Sludge Operating Conditions for Five Wastewater Treatment Plants 78 Table 4-5: Optimum Specific Growth Rate of Nitrifying Bacteria for Optimal Dissolved CO2 Concentration of 5 mg/l and Corresponding pH 80 Table 4-6: Estimated Specific Growth Rate of Nitrifying Bacteria and 95% Confidence Interval of the Activated Sludge from the WWTP with Extended Aera tion for Two Defined Dissolved CO2 Concentrations 82 Table 4-7: FISH Analysis of Five WWTP and Lab-Scale Reactors Operated at Extr eme Dissolved CO2 Concentrations 86 Table 5-1: Experiment 2 pH vs. Dissolved CO2 Concentration 100

PAGE 9

v Table 5-2: Combined Growth Parame ters for Experiment 1 102 Table 5-3: Combined Growth Parame ters for Experiment 2 102 Table 5-4: Experiment 2 Results of Completely Randomized Design of max at Selected Dissolved CO2 Concentrations 109 Table 5-5: Experiment 3 ANO VA of Main Effects of max 113 Table 5-6: Multiple Comparisons of Factor Effects Using Tukey Method with a 95.0% Confidence Level 114 Table 6-1: Nitrosomonas Percent Abundance Results 126 Table 6-2: Nitrosospira Percent Abundance Results 126 Table 6-3: ANOVA of Percent Abundanc e of AOB Bacteria 127 Table 6-4: ANOVA of Percent Abundance of Nitrosomonas Bacteria 128 Table 6-5: ANOVA of Percent Abundance of Nitrosospira Bacteria 129 Table 6-6: Nitrobacter Percent Abundance Results 132 Table 6-7: Nitrospirae Percent Abundance Results 132 Table 6-8: ANOVA of Percent Abundanc e of NOB Bacteria 133 Table 6-9: ANOVA of Percent Abundance of Nitrobacter Bacteria 133 Table 6-10: ANOVA of Percent Abundance of Nitrospirae Bacteria 135 Table 6-11: Slide Preparation for Validation Study 137 Table 6-12: Partitioned Treatm ent Sums of Squares by Tr eatment Effect 150

PAGE 10

vi List of Figures Figure 2-1: Extended Aerati on (shown in Oxidation Ditch Configuration) 11 Figure 2-2: Modified Ludzack-Ettinger (MLE) 12 Figure 2-3: Bardenpho 4 Stage 13 Figure 2-4: Bardenpho 5 Stage 14 Figure 2-5: Fraction of Dissolved Carbon Dioxide in Species Form as Function of pH in a Closed System 31 Figure 2-6: 1% CO2 Air Mixture and 60 mg/l of NH4Cl in an Open System 34 Figure 3-1: The Experimental SB R System that Features pCO2 Control in the Experimental Reactor (left) and the Control Reactor (right) 49 Figure 3-2: Ammonium, Nitrite, Nit rate, Total Nitrogen, pH, and DO for the Control Reactor in Experiment 1 55 Figure 3-3: Ammonium, Nitrite, Nitra te, Total Nitrogen, pH, and DO for the Experimental Reactor in Ex periment 1 55 Figure 3-4: Nitrate Formation Rates fo r Experiment 3 59 Figure 3-5: Nitrate Formation Rates fo r Experiment 4 60 Figure 4-1: FISH Analysis Scale R epresentation 75 Figure 4-2: Effect of pH at Varying Dissolved CO2 Concentrations 79 Figure 4-3: Evaluation of Specific Growth Rate of Nitrifying Bacteria Using Air (Control) or 7 mg /l (Experimental) Dissolved CO2 Concentration 81

PAGE 11

vii Figure 4-4: Representative FISH Images for Nitrifying Bacteria in MLE #1 including (A) Nitrosomonas spp ., Nitrosococcus mobilis; (B) Nitrosospira spp.; (C) Nitrobacter spp. and (D) most members of the phylum Nitrospirae 83 Figure 4-5: Representative FI SH Images for 4-Stage Bardenpho including (A) Nitrosomonas spp ., Nitrosococcus mobilis; (B) Nitrosospira spp.; (C) Nitrobacter spp. and (D) most members of the phylum Nitrospirae 84 Figure 5-1: Estimated max at 12 mg/l of Dissolved CO2 Concentration and a pH of 7.0 99 Figure 5-2: Experiment 1 Specific Growth Rate of Nitrifying Bacteria at Varying Levels of Dissolved CO2 Concentration at pH 7.0 99 Figure 5-3: Experiment 2 Growth Curve at Selected Dissolved CO2 Concentrations with 95% Confidence Levels 101 Figure 5-4: Composite Biomass Describing max from Experiment 1 with 95% Confidence Levels 103 Figure 5-5: Composite Biomass Describing max from Experiment 2 with 95% Confidence Levels 104 Figure 5-6: Experiment 3 Results of max at Selected pH and Dissolved CO2 Concentration with 95% Confidence Levels 105 Figure 5-7: Experiment 3 Results of Main Effects Plot for max 105 Figure 5-8: Experiment 3 Results of AOB at Selected pH and Dissolved CO2 Concentration with 95% Confidence Levels 106 Figure 5-9: Experiment 3 Results of NOB at Selected pH and Dissolved CO2 Concentration with 95% Confidence Levels 106 Figure 5-10: Experiment 2 Results of Boxplot of Completely Randomized Design Showing max at Selected Dissolved CO2 Concentrations 110

PAGE 12

viii Figure 5-11: Interaction Plot of max as a Function of Dissolved CO2 Concentration and pH 115 Figure 6-1: Typical Digital Images of (A) DAPI Stain, (B) Cy3 Stain, and (C) Merged Image of DAPI and Cy3 Stain 121 Figure 6-2: AOB Bacteria, R epresentative FISH Results Showing Percent Abundance 122 Figure 6-3: NOB Bacteria, R epresentative FISH Results Showing Percent Abundance 123 Figure 6-4: Percent Abundance of AOB Bacteria from Activated Sludge 125 Figure 6-5: Interaction Effect of Nitrosomonas Bacteria 128 Figure 6-6: Interaction Effect of Nitrosospira Bacteria 129 Figure 6-7: Percent Abundance of NOB Bacteria from Activated Sludge 131 Figure 6-8: Interaction Effect of Nitrobacter Bacteria 134 Figure 6-9: Interaction Effect of Nitrospirae Bacteria 135 Figure 6-10: Slide 1 No Probe with (A) Nitrosomonas (B) Nitrosospira (C) Nitrobacter (D) Nitrospirae (E) Seed Material and (F) E. coli 138 Figure 6-11: Slide 2 NSM156 Probe with (A) Nitrosomonas (B) Seed Material, and (C) E. coli using a Multiplicative Factor of 2.6 139 Figure 6-12: Slide 3 Nsv433 Probe with (A) Nitrosospira (B) Seed Material, and (C) E. coli using a Multiplicative Factor of 2.3 140 Figure 6-13: Slide 4 NIT3 Probe with (A) Nitrobacter (B) Seed Material, and (C) E. coli using a Multiplicative Fact or of 1.6 141 Figure 6-14: Slide 5 Ntspa0712 Probe with (A) Nitrospirae (B) Seed Material, and (C) E. coli using a Multiplicative Factor of 1.9 142

PAGE 13

ix Figure 6-15: Slide 6 LGC353b Probe with (A) Nitrosomonas (B) Nitrosospira (C) Nitrobacter (D) Nitrospirae (E) Seed Material, (F) E. coli, and (G) Bacillus subtilis 143 Figure 6-16: Nitrospirae Shown with High Accumulation of Cells (100X Objective) 144 Figure 6-17: Autotrophic Biomass as Percent of Total Biomass to Confirm High Percent Abundance Measurements 145

PAGE 14

x Abstract Ammonium (NH4 +) is a biological nutrient that is transformed in a wastewater treatment plant (WWTP) in a process called acti vated sludge. This is accomplished in an aerobic environment using microorganisms and inorganic carbon that convert the ammonium to nitrate (NO3 -). This process is termed nitrification. Removal of ammonium is necessary due to its oxygen demand and toxicity to the environment. Nitrification is considered a slow proce ss due to the slow growth rate of the nitrifying bacteria. Ammonia oxidizing bac teria (AOB) first co vert the ammonium (NH4 +) to nitrite (NO2 -) followed by conversion to nitrate (NO3 -) by nitrite oxidizing bacteria (NOB). These slow rates limit the treatment capacit y of the WWTP. The initial hypothesis suggest ed that these slow rates we re due to limited carbon in the aeration basin of a WWTP. A series of designed experiments and observational studies reveal ed substantial dissolved CO2 exists throughout a WWTP. Based on these findings, the central research focused on determining if an optimum dissolved CO2 concentration/ pH combination exists that maximizes nitrification.

PAGE 15

xi Experimentation conducted at a pH of 7.0 and varying concentrations of dissolved CO2 concentration revealed inhibiti on at low (<5 mg/l) and high (>30 mg/l) dissolved CO2 concentration levels. Further research found that optimum nitrification can be attained in a dissolved CO2 concentration range of 10 15 mg/l and a pH range of 7.5 – 8.0. A maximum specific growth rate of 1.05 – 1.15 days-1 was achieved. A partitioning of the sums of squares from these designed experiments found that pH accounts for appr oximately 83 percent of the sums of squares due to treatment with the dissolved CO2 concentration accounting for 17 percent. This suggests that pH is the dom inant factor affecti ng nitrification when dissolved CO2 concentration is optimized. Analysis of the growth kinetics for two of the designed experiments was conducted. However, a set of param eters could not be found that described growth conditions for all operating conditions. Evaluat ing the results from these two experiments may suggest that a micr obial population shift occurred between 16 and 19 mg/l of dissolved CO2 concentration. These dissolved CO2 concentrations represent pH values of 7.1 and 7.0, respectively, and were compared to experimentati on conducted at a pH of 7.0. Though the pH difference is minor, in combinat ion with the elevated dissolved CO2 concentration, a microbial shift was hypothesized. Microbial samples were collected from the designed experiment that optimized dissolved CO2 concentration (5, 10 and 15 mg/l) and pH (6.5, 7.0, 7.5 and 8.0).

PAGE 16

xii These samples were evaluated using Fluorescence in situ hybridizations (FISH) to determine the population density of common ammonium oxidizing bacteria (AOB) ( Nitrosomonas and Nitrosospira ) and nitrite oxidizing bacteria (NOB) Nitrobacter and Nitrospirae ). The dominant AOB and NOB microbes were found to be Nitrosomonas and Nitrospirae. These results suggest that increased ni trification rates can be achieved by incorporating appropriate c ontrols in a wastewater treatment plant (WWTP). With higher nitrification rates, lowe r nitrogen values can be obtained which will reduce the WWTP effluent nitrogen concentra tion. Conversely, these increased nitrification rates can also reduce the vo lume of an aeration basin given similar effluent nitrogen concentrations.

PAGE 17

1 Chapter 1 Research Objective 1.1 Main Objective The main objective of this research is to determine if an opt imum pH/ dissolved CO2 concentration exists that will minimize the time required for nitrification in an activated sludge wastewater treatment facility. 1.2 Research Goals This research will focus on answering the following questions: Do ammonia oxidizing bacteria (AOB) and nitrite oxidizing bacteria (NOB) exhibit reduced growth due to carbon limitation? Is there a preferred dissolved CO2 concentration that provides for optimum nitrifier growth?

PAGE 18

2 Is there a preferred pH val ue in combination with dissolved CO2 concentration that provides fo r optimum nitrifier growth? Can the microbes most abundant in the nitrification process be quantified at varying pH/dissolved CO2 concentrations that bracket this optimum combination? 1.3 Hypothesis and Approach It is hypothesized that t he autotrophic nitrifying bact eria in activated sludge systems grow slowly due to CO2 limitation. Elevated levels of dissolved CO2 concentrations above atmospheric concentrati ons will improve the nitrifier growth rate and thus reduce the nitrification time In order to answer this research question, a series of designed experiments we re conducted.. Testing protocol is outlined as follows: Conduct a series of preliminary ex periments to determine if elevated dissolved CO2 concentration at specified pH levels using synthetic feed as well as influent from a wastewater treatment facility exhibit increased nitrifier growth as compared to air systems.

PAGE 19

3 Determine operating conditions, dissolved CO2 concentrations and pH at several wastewater treatment facilitie s. Evaluate these conditions as compared to preliminary experiments discussed above. Based on results from previous ex perimentation and assessment of field studies, determine a range of dissolved CO2 concentration/ pH combinations that encompass the opt imum combination of these two variables to achieve maximum nitrification growth. Quantify the microbial percent abundance of the most common nitrifiers at the optimum dissolved CO2 concentration/ pH combination.

PAGE 20

4 Chapter 2 Wastewater Treatment Industry, Literature Review and Preliminary Research 2.1 Wastewater Treatment in the United States There are 16,024 publicly-o wned wastewater treatme nt processes (WWTP) currently in operation in t he United States, serving a population of approximately 190 million people (approximat ely 72 percent of the U. S. population). Their treatment capacity represents a wastewater flow of ap proximately 32,175 million gallons per day. Of these plants, 9,388 facilities provide secondary treatment, 4,428 facilities provide advanced treatm ent, and 2,032 facilities do not discharge to surface waters. In addition, there ar e 176 facilities that provide a treatment level that is less than secondary (these include facilities with ocean discharge waivers and treatment facilit ies discharging to other fa cilities meeting secondary treatment or better [1]. There are several types of wastewater treat ment facilities currently in operation in the U.S. The most prevalent type utilizes an aeration basin to treat and remove biological matter (secondary treatment ). The removal of nitrogen and phosphorus are considered advanced treatm ent methods and in many facilities

PAGE 21

5 are dealt with separately from the secondary treatment. In recent years, several waste treatment designs have been develop ed that incorporate these advanced removal processes into the aeration basins [2, 3]. The energy impact of the water industry is considerable. Most wastewater treatment systems require a high level of energy to operate, especially advanced treatment systems [4]. It is estimated that more than 5 percent of all global electricity is used to treat wastewater [5 ] and approximately 3 percent of electrical usage in the United States [6]. In addi tion, energy costs can account for 30 percent of the total operational and mainte nance costs of a wastewater facility [6] with 50 percent of the energy co sts for the aeration system [7]. 2.2 Biological Nutrient Removal (BNR) Systems Biological nutrient removal (BNR) is defined as the removal of total nitrogen (TN) and total phosphorus (TP) from wastewat er through the use of microorganisms under different environmental conditions in the treatment process [2]. This activated sludge process dates back to the 1880's but was not officially described until 1914 by Arden a nd Lockett. During experimenta tion, they discovered that aerating a mass of microorganisms provi ded for stable organic material in wastewater. The aeration process was termed activated sludge and gave rise to the modern wastewater tr eatment processes (WWTP) we have today [2].

PAGE 22

6 2.2.1 BNR Wastewater Treatment Processes There are a number of BNR process configurations available. Some BNR systems are designed to remove only tota l nitrogen (TN), or both (TN) and total phosphorus (TP). The configuration most appropriate for any particular system depends on the target effluent quality, operat or experience, influent quality, and existing treatment processes. BNR conf igurations vary based on the sequencing of environmental conditions (i.e., aer obic, anaerobic, and anoxi c)and timing [3]. Some common BNR system configurations based on their biological nutrient removal focus are discussed below [2, 8]. 2.2.1.1 Total Nitrogen Removal Only Modified Ludzack-Ettinger (MLE) Pr ocess – continuous-flow suspendedgrowth process with an in itial anoxic stage followed by an aerobic stage Step Feed Process – alternati ng anoxic and aerobic stages; however, influent flow is split to severa l feed locations and the recycle sludge stream is sent to the beginning of the process Bardenpho Process (Four-Stage) – continuous-flow suspended-growth process with alternating anoxic /aerobic/anoxic/aerobic stages

PAGE 23

7 Sequencing Batch Reactor (SBR) Process – suspended-growth batch process sequenced to simulate the fou r-stage process; used to remove TN (TP removal is inconsistent) Extended Aeration Process (EAAS or usually called EA) [2] a process used on wastewaters that have not been tr eated in a physica l operation to remove suspended organic matter (primary clarifier). In this case, the insoluble organic matter becomes trapped in the biofloc and undergoes some oxidation and stabilization. Mo st other activated sludge systems are used on wastewaters from which settl eable solids have been removed [9]. EA processes utilize long solid ret ention times (SRT) to stabilize the biosolids resulting from the remova l of biodegradable organic matter. SRTs of 20 to 30 days ar e typical, which means hy draulic retention times (HRT) around 24 hours are required to maintain reasonable mix liquor suspended solids (MLSS) concentrations Long SRT's offer two benefits: reduced quantities of solids to be disposed of and greater process stability. These benefits are obtai ned at the expense of the large bioreactors required to achieve the long SRT's, but for many small installations the benefits outweigh t he drawbacks [9]. It has good capacity for nitrogen removal; less than 10 mg/l effluent TN is possible. However, nitrogen removal capability is related to skills of operating staff and control methods. (The Extended Aeration proces s identified in this research was

PAGE 24

8 originally built as an Oxi dation Ditch. Due to its operation, it is classified as an Extended Aeration process.) 2.2.1.2 Total Nitrogen a nd Total Phosphorus Removal A2/O Process – MLE process preceded by an initial anaerobic stage Modified Bardenpho Process (Five Stage) – Bardenpho process with addition of an init ial anaerobic zone Modified University of Cape Town (UCT) Process – A2/O Process with a second anoxic stage where the inte rnal nitrate recycle is returned Oxidation Ditch – continuous-flo w process using looped channels to create time sequenced anoxic, aerobic, and anaerobic zones A comparison of the TN and TP re moval capabilities of common BNR configurations is provided (Table 2-1) This table provides only a general comparison of treatment performance am ong the various BNR configurations; site-specific conditions dictate th e performance of each process [3].

PAGE 25

9 Table 2-1: Comparison of Co mmon BNR Process Configurations Process Nitrogen Removal Phosphorus Removal MLE Good None Four-Stage Bardenpho Excellent None Step Feed Moderate None SBR Moderate Inconsistent A2/O Good Good Modified UCT Good Excellent Five Stage Bardenpho Excellent Good Oxidation Ditch Excellent Good Although the exact configur ations of each system differ, BNR systems designed to remove TN must have an aerobic z one for nitrification and generally incorporate an anoxic zone for denitrificat ion. BNR systems designed to remove TP must have an anaerobic zone free of dissolved oxygen and nitrate. Often, sand or other media filtrati on is used as a polishing st ep to remove particulate matter when low TN and TP effluent concen trations are requir ed. Sand filtration can also be combined with attached grow th denitrification filters to further reduce soluble nitrates and effluent TN levels [10]. Choosing which system is most appropriate for a particular facility primarily depends on the target effluent concentra tions (usually permit driven), and whether the facility will be constructed as new or retr ofit with BNR to achieve more stringent effluent limits. New plants have more flexibility and options when deciding which BNR conf iguration to implement because they are not constrained by existing treatment uni ts and sludge handling procedures [8].

PAGE 26

10 2.2.1.3 Wastewater Treatme nt Plant Configurations The four WWTP's used in this research included the Extended Aeration (EA), a Modified Ludzack-Ettinger (MLE), a Ba rdenpho 4 stage and a Bardenpho 5 stage facility. A schematic of each plant c onfiguration is provided on the following pages [2]:

PAGE 27

11 Figure 2-1: Extended Aeration (shown in Oxidation Ditch Configuration) Return Activated Sludge Secondary Clarifier Effluent Waste Activated Sludge Influent Internal Recycle

PAGE 28

12 Figure 2-2: Modified Ludzack-Ettinger (MLE) Return Activated Sludge Secondary Clarifier Effluent Waste Activated Sludge Anoxic Aerobic Internal Recycle Influent

PAGE 29

13 Figure 2-3: Bardenpho 4 stage Return Activated Sludge Secondary Clarifier Effluent Waste Activated Sludge Anoxic Aerobic Internal Recycle Influent Anoxic Aerobic

PAGE 30

14 Figure 2-4: Bardenpho 5 stage Return Activated Sludge Secondary Clarifier Effluent Waste Activated Sludge (Contains Phosphorus) Anaerobic Anoxic Internal Recycle Aerobic Anoxic Aerobic

PAGE 31

15 2.3 Nitrifying Bacteria and Nitrification Bacteria found in the aeration basin of a wastewater treatment system are defined as either heterotrophs or autot rophs. Heterotrophs use organic carbon for formation of biomass and are primar ily responsible for the reduction of organic matter (BOD). Autotrophic bacter ia derive cell carbon from carbon dioxide and are responsible for converting ammonium (NH4 +) to nitrite (NO2 -) and then to nitrate (NO3 -) [2]. Concentrations of the types of bacteria found in wa stewater vary depending on operating conditions (SRT, in fluent qualities, domestic/ industrial percentages, activated sludge operating tem perature, etc.) and results vary widely. One study that evaluated the waste activated sludge (WAS) from a membrane bioreactor found bacteria percentages in the following ranges [11]: Heterotrophs 15 50 percent with an average percentage of 35 percent. Autotrophs 2 8 percent with an average of 3 percent. Another study evaluated the effect of t he mixed liquor suspended solids (MLSS) and heterotrophic and autotrophi c biomass as a function of solids retention time (SRT). At a 12 day SRT, the MLSS concentration was 3000 mg/l with the heterotrophic and autotrophic concentrations at 1300 and 85 mg/l, respectively. (All values reported as mg /l as COD (carbonaceous oxygen demand)). Thus, heterotrophs represent approx imately 43 percent of th e biomass with autotrophs

PAGE 32

16 representing approximately 3 percent. Additionally, this represents approximately a 15:1 ratio of het erotrophs to autotrophs [9]. Protozoa are also found in wastewater and may contribut e as much as 5 percent of the biomass [9]. They are t he main predators in suspended growth bioreactors that feed on bacteria. Cilia tes are usually the dominant protozoa, both numerically and on a mass basis. Almost all are known to feed on bacteria and the most important are either attach ed to or crawl over the surface of biomass flocs. Viruses and polyphosphat e accumulating organisms comprise other microbes found in wastewater [9]. Total effluent nitrogen comprises ammonia, nitrate, particulate organic nitrogen, and soluble organic nitrogen. The biological processes that primarily remove nitrogen are nitrification and denitrification [3]. In BNR systems, nitrification is the controlling reaction because ammonia oxidizing bacteria lack functional diversity, have stringent growth r equirements, and are sensitive to environmental conditions [3]. Nitrification by itself does not actually remove nitrogen from wastewater. Rather, deni trification is needed to convert the oxidized form of nitrogen (nitrate) to nitr ogen gas. Nitrification occurs in the presence of oxygen under aer obic conditions, and denitri fication occurs in the absence of oxygen under anoxic conditions.

PAGE 33

17 Microorganisms use an electron donor substr ate to meet their growth needs, cell synthesis (fs), and their cell maintenance needs (fe) [12]. These two values, fs and fe, add to one and are expressed in te rms of electron equivalents (e-eq). The fraction fs can be converted into mass units such as g cell produced/ g COD consumed. When expressed in mass units it is termed the true yield and given the symbol Y. The conversion from fs to Y is given as: Y = fs (Mc g cells/ mol cells)/ [ne -eq/ mol cells)(8 g COD/ e-eq donor) Where: Mc = the empirical formula weight of cells ne = the number of electron equivalent s in an empirical mole of cells When cells are represented by C5H7O2N and ammonium is the nitrogen source, Mc = 113 g cells/ mol cells, ne = 20 e-eq/ mol cells. This conversion gives Y = 0.706 fs and Y is in g cells/ g COD. The numbers used in t he conversion change if the cell formula differs or if the ce lls use oxidized nitrogen sources, such as NO3 [12]. From a practical viewpoint, low fs values translate into slow cell growth as they have high maintenance needs. As a comparison, ammonium oxidizers have a fs value of 0.14, nitr ite oxidizers have a fs value of 0.10, and aerobic heterotrophs have typical fs values of 0.6 0.7. These low fs values for the ammonium and nitrite oxidizer s translate into low autotro phic biomass growth. In characterizing a biochemical process, inve stigators can use substrate removal or

PAGE 34

18 biomass growth to describe this activity [9]. This relationship is given by the formula: Where: = maximum specific growth rate = maximum specific rate of substrate utilization Y = yield for cell synthesis Each of these parameters is related but describes different aspects of the biochemical process. is influenced by variation in Y as well as variation in Like Y is influenced by the substrate bei ng consumed and the microorganisms performing the consumption. However, Y is a reflection of the energy available in a substrate whereas is a reflection of how rapidly a microorganism can process that energy and grow. Becaus e they represent different characteristics, there is no correlation between the two parameters. For example, some substrates that are consumed very slowly (low ) provide more energy to the degrading organism (higher Y) than do substrates that are degr aded rapidly [13]. This suggests that inferences about the variability of cannot be made on alone, and vice versa. Knowledge of the true growth yield is also important in assessing these relationships [9].

PAGE 35

19 Nitrification is a two-step process utilizi ng aerobic, autotrophic, nitrifying bacteria to complete the conversion process. Ammonium (NH4 +) is first converted to nitrite (NO2 -) according to the energy yielding equation [12]: 1/6 NH4 + + 1/4 O2 = 1/6 NO2 + 1/3 H+ + 1/6 H2O Nitrosomonas an ammonia oxidizing bacteria (AOB), is considered the predominant bacteria species for this conversion [2]. The nitrite is further oxidized to nitrate (NO3 -) according to the energy yielding equation [12]: 1/2 NO2 + 1/4 O2 = 1/2 NO2 Of the nitrite oxidizing bacteria (NOB), Nitrobacter has been considered the predominant microbe, but in recent years Nitrospirae bacteria has been found to play a more significant role. Both AOB and NOB are thought to have slow growth rates and are sensitive to pH and temperature swings, making nitrification difficult to maintain in activated sludge systems [14, 15]. Although autotrophic bacteria are the dominant microbe in nitr ification, ammonium oxidation can be performed by archaea [16, 17]. Ammonium-oxidizing archaea were found to occur in WWTP’s that were operated at low dissolved oxygen levels and long solid retention times [18].

PAGE 36

20 A complete reaction for the conversion of NH4 + to NO3 (fs = 0.1) is written as follows [12]: NH4 + + 1.73 O2 + 0.154 CO2 + 0.038 HCO3 0.038 C5H7O2N +0.962 NO3 + 1.92 H+ + 0.923 H2O Denitrification involves the biological reduc tion of nitrate to nitric oxide, nitrous oxide, and nitrogen gas [2]. Both het erotrophic and autot rophic bacteria are capable of denitrification. The most common and widely distributed denitrifying bacteria are Pseudomonas species, which can use hydrogen, methanol, carbohydrates, organic acids, alcohol s, benzoates, and other aromatic compounds for denitrification [2]. Table 22 provides a review of the different forms of nitrogen and removal cap ability from wastewater [3]. Table 2-2: Effluent TN Co mponents and Achievable Limits Form of Nitrogen Common Removal Mechanism Technology Limit (mg/l) Ammonia-N Nitrification <0.5 Nitrate-N Denitrification 1 – 2 Particulate organic-N Solids separation <1.0 Soluble organic-N None 0.5 – 1.5 (Organic nitrogen is not removed biologica lly. Only the parti culate fraction can be removed through solids separation via sedimentation or filtration [8].) Nitrosomonas and Nitrobacter have been considered the predominant AOB and NOB bacteria involved in nitrification and have been investigated extensively [1922]. In recent years, Nitrosospira and Nitrospirae have been identified as important microbes involved in nitrification (Table 2-3 and [12, 20, 23, 24]). And

PAGE 37

21 in one study, Nitrospirae was found to be the most ab undant nitrite oxidizer in wastewater treatm ent systems [24]. The properties of these predominant AOB and NOB bacteria are provided in Table 2-3. Some properties are similar among the bacteria types but differences do exist. Nitrosomonas and Nitrospirae have similar optimum pH ranges but differ from the optimum pH for Nitrobacter A WWTP optimized for pH may not obtain optimum nitrification if Nitrosomonas and Nitrobacter are the predominant AOB and NOB bacteria due to their optimum pH ranges. Additionally, Nitrosospira growth may be enhanced at low temperature and Nitrospirae may dominate under low c oncentrations of NH4 + and NO2 [24, 25]. Information is limited as evidenced by several missing cells in the table. This may be due to the limited availability of pure cultures of nitrifying bacteria to study.

PAGE 38

22 Table 2-3: Properties of Pr edominant Nitrifying Bacteria Genus Morphology pH Optimum pH Oxygen (mg/l) Temp Range (C) Optimum Temp (C) Nutrients Maximum Specific Growth Rate, days-1 (Optimum @ 20oC) Comments AOB Nitrosomonas gram negative, rod shaped or pear shaped 6-9 7.9 8.2 [26, 27] >2.0 2530 0.76 Inhibited at pH of 6.5, Nitrification ceases at pH 6.0. Can grow in low salinity environments [28] Nitrosospira spiral 7.5 -8.0 Enhanced at low temp NOB Nitrobacter gram negative; rod shaped, pear-shaped or pleomorphic 7.2-7.6 [26, 27, 29] >2.0, More strongly affected by low DO than Nitrosomonas 049 2530 Needs phosphates 0.81 More likely to dominate nitrite oxidation under conditions with low ammonium and nitrite concentrations Nitrospirae Long, Slender rods 8.0-8.3 [22]

PAGE 39

23 Although all of the ni trifiers were once included in the same family because of their activities, it is now recognized that they are phylogenetically diverse [23]. With improvements in genetic techniques, other species have been identified that are not necessarily the most common or mo st active in the env ironment. Hence, nitrifying activity should not be assigned to these genera unless they are actually identified [30]. The AOB include genera with in the Proteobacteria: Nitrosomonas ( ), Nitrosospira ( ), and Nitrosococcus ( ). Nitrosococcus bacteria is considered to dominate in marine environments [30]. Nitrosococcus mobilis was originally isolated from brackish water [31] but has been identified as a major contributor in the nitrification process of sewage treatment [32]. T he NOB also includes genera within the Proteobacteria: Nitrobacter ( ), Nitrococcus ( ), and Nitrospina ( ) [33, 34]. In addition, Nitrospirae a member of the Xenobacteria has been identified as a NOB. Recently, Nitrosomonas has been found in low salinity environments In recent years it has been found that wastewater tr eatment plants are highly diverse microbial systems and are usually not represented by one nitrifying bacteria [19, 20, 23, 24, 32, 35-37]. The co existence of different nitrifiers implies functional redundancy which may allow comm unities to maintain physiological capabilities when conditions change. Thus, a high level of nitrifier diversity is thought to confer perform ance stability [23].

PAGE 40

24 Juretschko [20] evaluated waste from a WWTP in Germany that received high ammonia concentrations (5,000 mg/l) fr om a high protein-rich animal waste processing facility. Analysis re vealed the predominate AOB to be Nitrosococcus mobilislike bacteria [20]. It should be noted that Nitrosococcus mobilis is now considered to be a member of the genus Nitrosomonas [20]. This animal waste could have influenced the selection of this AOB. The major NOB genus was Nitrospirae Dionisi [37] investigated two WWTP's. The first was a 40 million gallon/day WWTP (6 hour HRT) treating primarily muni cipal waste with so me industrial and hospital discharges. Nitrosomonas (AOB) and Nitrospirae (NOB) were identified as the predominant microbes. The second was a 27 million gallon/day industrial WWTP treating fibers, plastics and chemical s. Its waste consisted mainly of acetic acid, propionic acid, n-butyric ac id, ethylene glycol, ethanol, methanol, isopropanol, and acetone and no municipal waste. Nitrosomonas (AOB) and Nitrospirae (NOB) were identified as the predominant nitrif ying bacteria, but the AOB were different species of Nitrosomonas between the WWTP's. Using a fluidized bed reactor, Schramm (1998) used low concentrations of NH4 + (40 M) and identified the predominate AOB as Nitrosospira and the NOB as Nitrospirae [19]. No mem bers of the genus Nitrosomonas (AOB) or Nitrobacter (NOB) could be detected. This is agr eement with other studies conducted in natural systems in which the ammoni um concentration was low [38-40].

PAGE 41

25 Green et. al [41] conducted a similar st udy using a fluidized bed reactor with chalk (solid calcium carbonate). In this study, the pH established in the reactor varied between 4.5 and 5.5 with higher nitr ification rates obtained at the lower pH. In spite of the low pH, a high ni trification rate was observed and found similar to nitrification rates observed in a biological reactor operated at a pH>7.0 [41]. Nitrosomonas (AOB) and Nitrospirae (NOB) were identified as the predominant microbes. Over time thes e microbes may have become acclimated to these environmental conditions or may represent new species. The pH of a WWTP does have an effect on the nitrification rate. A pH of 7.5 8.0 is considered optimum with rates declining below a pH of 6.8 [2]. Studies conducted by researchers confirm these pH ranges [22, 42, 43]. However, a study conducted by Tarre and Green found that nitrification could be achieved at a low pH [44]. When using a biofilm reactor, a specific nitrific ation rate of 0.55 days-1 were achieved at a pH of 4.3+0.1. This is similar to values reported for nitrifying reactors at optimum pH. When conducted using a suspended-biomass reactor, a specific nitrif ication rate of 0.24 days-1 was achieved at a pH of 3.8+0.3. Nitrosomonas (AOB) and Nitrospirae (NOB) were identified as the predominant microbes in both systems. (Note: The suspended-biomass study was repeated in the USF Stroot lab us ing equipment to conduct the elevated dissolved CO2 concentration study. Nitrification could not be achieved below a pH of 6.0.)

PAGE 42

26 Temperature impacts nitrifier growth rates with lower temperatures producing lower nitrification rates [2]. In on e study, a temperatur e difference of 10oC (30oC versus 20oC) showed a three-fold increase in maximum growth rates [45]. Studies conducted by Siripong a nd Rittman [23] showed that Nitrosomonas has the potential to grow twice as fast as Nitrosospira in the optimum temperature range. This growth advant age favors detection of Nitrosomonas rather than Nitrosospira with culture based methods. W hen investigating WWTP's during summer and winter conditions, which had 6.7-13.4oC lower temperatures and 1349% higher solids retention ti me (SRT), higher levels of Nitrosospira were detected during t he winter [23]. Other research has suggested that AOB and NOB are quite versatile in their ability to adapt [46]. Under anaerobic conditions, Nitrosomonas (AOB) was found to be capable of nitrite denitri fication with molecular hydrogen, hydroxylamine or organic matter (pyruvate, formate) as electron donors resulting in production of N2O and N2 [47-50]. It has been s uggested that this is a protection mechanism against the negative effects of hi gh nitrite concentration [51, 52]. Alternatively, it has been recognized as a pr ocess of high importance for anaerobic growth [51, 52] as well as for the supply of NO necessary for ammonium oxidation [53, 54]. Under oxygen-limited or anoxic conditions, ammonium could act as an electron donor that is oxidized with nitrite instead of oxygen as the electron acceptor [50, 55].

PAGE 43

27 Several strains of Nitrobacter are capable of heterotroph ic growth under oxic as well as anoxic condition [33, 56, 57]. Some strains of Nitrobacter were shown to be denitrifying organisms as well. Under anoxic conditions, nitrite can be used as an acceptor for electrons derived fr om organic compounds to promote anoxic growth [58]. Since the oxidation of nitr ite is a reversible process, the nitrite oxidase-reductase can reduce nitrate to nitrite in the absence of oxygen [59]. The aeration basin of a WWTP is a complex microbial community probably containing several different gene ra of microbes capable of nitrification [23]. Their food source and operating conditions (t emperature, pH, DO) undoubtedly have a significant effect as to which species dominates. 2.4 Heterotrophic Bacteria, Chemical Oxygen Demand (COD) and Ammonium Removal Heterotrophic bacteria consume COD and ni trogen in order to produce biomass. Ammonium (NH4 +) and organic nitrogen compounds are the preferred nitrogen sources but nitrate will also be utiliz ed in the absence of ammonia [9].

PAGE 44

28 A complete mass based stoichiometric equation for the consumption of carbohydrate COD removed using a mmonia as the nitrogen source (fs = 0.71) is written as follows [9]: CH2O + 0.309 O2 + 0.085 NH4 + + 0.289 HCO3 0.535 C5H7O2N + 0.633 CO2 + 0.515 H2O Based on this stoichiometric equation, one mg of NH4 + is required to convert approximately 19.6 grams of the carbohydr ate COD to the COD biomass. For 300 mg/l of influent COD, approximately 15 .3 mg/l of ammonia is necessary to convert the COD into biom ass. The ammonium not consumed will be converted to nitrate (NO3 -) through nitrification utilizing autotrophic bacteria. Though uncommon, processes with limited influent nitrogen sources (high COD:N ratio) will experience difficulties in converting all of the COD. During anoxic conditions, heterotrophic bac teria will use nitrate as an electron acceptor (instead of oxyg en) and ammonium as a nitrogen source. A complete mass based stoichiometric equation (fs = 0.71) is provided [9]: CH2O + 0.479 NO3 + 0.085 NH4 + + 0.289 HCO3 + 0.008 H+ 0.535 C5H7O2N + 0.634 CO2 + 0.108 N2 + 0.584 H2O

PAGE 45

29 The above reaction, known as denitrificati on, occurs in wastewater treatment plants that incorpor ate an anoxic zone to convert nitrate (NO3 -) to nitrogen gas (N2). The denitrification rate (g NO3 --N reduced/g MLVSS d), which determines the amount of nitrate deni trified, is primarily a function of availability of rapidly biodegradable organic matter (R BOM) and temperature [3]. Denitrifiers, typically heterotrophs but certain autotrophs are capable of denitrification[60], use organic matter as the energy and carbon source. As a first approximation, a minimum BOD:TKN rati o of approximately 3:1 is required in the bioreactor influent for reliable denitrific ation. The actual ratio will depend on operating conditions and substr ate biodegradability. Wi thin limits, higher F/M ratios in the anoxic zone achieve higher denitrification rates due to the presence of increased RBOM. Likewise, the ty pe of substrate also impacts the denitrification rate. Signif icantly higher denitrificat ion rates are possible with methanol and fermentation end-products, such as volatile fatty acids (VFAs) present in the influent wastewater. Denitrification suppor ted by endogenous decay is associated with slow denitrification rates [3]. 2.5 Carbon Dioxide and Wastewater Treatment Plants Aquatic systems can be modeled with dissolved CO2 in open or closed systems. With rare exceptions, wastewater treat ment facilities are open systems as they are exposed to the atmosphere and liquid is entering and existing continuously.

PAGE 46

30 Additionally, depending on the pH, the CO2 will dissociate into three species within the aquatic systems, H2CO3 *, HCO3 -, CO3 2(Figure 2-5). As most wastewater treatment fac ilities operate in the pH range of 6.8 – 7.3, HCO3 is the predominant carbon dioxide species. Th is is true for both open and closed CO2 systems. At a pH of 7. 0 in the closed system, approximately 81% of the carbon dioxide exists as bicarbonate (HCO3 -) with the remainder as H2CO3 *. It should also be noted that 99% of carbon dioxide in solution exits in the form of dissolved carbon dioxide [61]. Speciation is governed by the following equations: CO2(g) CO2(aq) KH = 10-1.48 H2CO3 CO2(aq) + H2CO3 H2CO3 H+ + HCO3 pKa1 = 6.35 @ 25oC HCO3 H+ + CO3 2pKa2 = 10.33 @ 25oC Where: g = gas aq = aqueous KH = Henry’s constant pKa = acid dissociation constant

PAGE 47

31 Figure 2-5: Fraction of Dissolved Carbon Dioxide in Species Form as Function of pH in a Closed System Closed and open systems do have some differences [62]: In open systems, H2CO3 remains constant. The total carbonate concentration, [H2CO3 *] + [HCO3 -] + [CO3 2-], is constant in a closed system but varies with pH in the open system (Figure 2.6). Nitrification results in the destructi on of 7.1 mg of alkalinity (CaCO3) per mg of NH4 +-N oxidized. As ammonium is oxid ized, it produces two strong acid 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 01234567891011121314Fraction of CtpH H2CO3* HCO3 CO3 2-

PAGE 48

32 equivalents per mole of NH4 + removed [12]. If the influent contains inadequate alkalinity, nitrification would be compromis ed. As alkalinity is destroyed, pH is decreased and this could potentially reduce the nitrification rate as the alkalinity is needed to buffer the system. Most WW TPs operate in a pH range of 6.8 to 7.3. Denitrification results in the recovery of 3.6 mg of alkalinity as CaCO3 and 2.9 mg of oxygen per mg of NO3 --N reduced. This oxygen equi valent is a useful factor when calculating the total oxygen required fo r nitrification-denitrification biological treatment systems [2]. Ther efore, by combining nitr ification (aerobic) and denitrification (anoxic), partial alkalini ty recovery and oxygen credit can be attained. An additional benef it of incorporating an anox ic selector is improved sludge settleability [3]. Carbon dioxide concentrations in a WWTP will vary depending on the unit operation. Dissolved CO2 in the influent is usually lo w (10 mg/l or less) but can be high if anaerobic conditions exist in t he sewer system. In the aeration basin, carbon dioxide is produced in the consumption of ca rbohydrate COD and during denitrification. (See t he section 2.4, Heterotr ophic Bacteria and COD and Ammonia Removal, for a review of the stoichiometric equations.) However, during nitrification CO2 is consumed at the ra te of 0.085 moles of CO2 for every mole of NH4 + consumed. From secondary cl arification to discharge, the CO2 concentrations will decrease. Typical discharge concentrations (effluent) of 12

PAGE 49

m S S c A c N s r e o o c t o A N [ H T m g/l or les s S pecific Gr o S ludge and oncentrati o A pC-pH di a oncentrati o N concentr a hown as i e search. T f pH is s h perate, bi c onstant ac r o a pH of a p A mmonium N H4 + H +] = [NH3( a T he governi s are com m o wth Rate Municipal o ns at seve r a gram sho w o n) CO2 air a tion is p r i t was sel e T he varying h own. A t c arbonate r oss pH le v p proximat e speciation H+ + NH3 a q)] + [OH-] ng equatio n m on. (Se e Sensitivit y Wastewa t r al WWTP’ s w ing conce n mixture in r ovided (Fi e cted in s species c o a neutral (HCO3 -) i s v els and C O ly 10 wher e is governe d n s for CO233 e chapter y to Carb o t er,” for a s by unit o p n trations f o an open c a gure 2-6). ubsequent o ncentratio pH, which s the pre d O 3 2is low. e the conc e d by the fol were listed 4, “Evalu a o n Dioxid e complete p eration.) o r a 1 perc e a rbonate s y This a m experime n n of the c a h most wa s d ominate s The bica r e ntration of lowing equ pKa = 9 Proton C d previousl y a tion of Ni t e for FullS review of e nt (17 mg / y stem with a m monium c n tation co n a rbon dioxi d s tewater t r s pecies, H r bonate re m f the CO3 2a u ations: .3 C ondition y t rifying Ba c S cale Acti v measured / l dissolved a 60 mg/l N c oncentrati o n ducted in d e as a fun r eatment p H 2CO3 re m m ains dom a ttains equ c teria v ated CO2 CO2 N H4 +o n is this ction p lants m ains inant ality.

PAGE 50

34 NH4 + remains at a constant conc entration until it reaches its pKa value. Upon reaching this value, it transitions to NH3. The decrease in concentration as pH increases is due to NH3 being a base. Figure 2-6: 1% CO2 Air Mixture and 60 mg/l of NH4Cl in an Open System pH2CO3 *pHCO3 -pCO3 2pOH pNH4 +pNH3 pH 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 01234567891011121314pCpH

PAGE 51

35 2.6 Substrate Utilization in W astewater Treatment Plants At steady state, the mass bal ance equation for substrate in an activated sludge system may be written as: Substrate in influent Substrate consumed = Substrate in effluent Substrate in WAS The change in substrate c oncentration with time can be determined by starting with the substrate mass balance for a comple tely stirred tank reactor (CSTR) [2, 63]: dS/dtV = Q So – Q S + rsu V 2-1 where: rsu = substrate utilization rate = (max S X) / [ Y (Ks + S)] V= volume of wastewater in the aeration tank Q = flow rate of wastewater So = substrate concentration in influent, t = 0, mg/l [substrate for growth of heter otrophs (aerobic) and nitrifiers] S = substrate concentration in effluent at time t, mg/l Y = fraction of substrate mass converted to biomass Ks = half saturation constant, mg /l = concentration of limiting substrate when = 0.5 max max = maximum specific growth rate, days-1 X = concentration of biomass, mg/l

PAGE 52

36 Experimentation conducted in this resear ch used a batch reactor. Since Q is equal to zero for a batch reactor and volume is constant, equation 2-1 can be simplified to: dS/dt = (max S X) / [ Y (Ks + S)] 2-2 Integration of equation 2-2 wit h respect to time yields: Ks ln(So/St) + (So – St) = X (max / Y) t 2-3 Where: So = substrate concentration in influent, time = 0, mg/l S = substrate concentration at time t, mg/l t = time, days For nitrification, the Monod kinetic coeffici ents are substituted in equation 2-3 to yield: Ks ln(No/Nt) + (No – Nt) = Xn (max/Y) [ DO / (Ko + DO) ] t 2-4 Where: Xn = concentration of nitr ifier biomass, mg/l DO = dissolved oxygen concentration, mg/l Ko = half saturation const ant for oxygen, mg/l No = substrate concentration (ammoniu m) in influent, time = 0, mg/l Nt = substrate concentration (a mmonium) at time t, mg/l

PAGE 53

37 2.7 Estimation of the Maximum Specific Growth Rate, max, from NOx Generation Rate in Batch Reactor The rate of generation of the NOX concentration (nitrite + nitrate) is equal to the disappearance of ammonium utilized for nitr ification. Its relationship is given by: dSNOx/dt = dSNH4+/dt 2-5 Initial reactor conditions must provide a hi gh ammonium concentration, relative to the half velocity constant from Monod kinetics, Ks to ensure that the nitrification rate is at a maximum [21]. From Monod kinetics, = max (SNH4+/(Ks + SNH4+). With high concentrations of ammonium the specific growth rate, , will essentially equal the maximum specific growth rate, max. Its relationship is given by: dSNOx/dt = max (XAUT/YAUT) 2-6 And, rearranging the right si de of equation 2.6 results in, dSNOx/dt = (max/YAUT) XAUT 2-7 The change in nitrifier biomass concentration, XAUT, is determined by growth and decay.

PAGE 54

38 dXAUT/dt = (maxXAUT) – (bAUTXAUT) = (max bAUT) XAUT 2-8 Integrating this equation from ti me zero to ti me t yields: XAUT,t = XAUT,0e(XAUT – bAUT) t 2-9 where: SNOx = oxidized nitr ogen concentration AUT = maximum specific nitrifier growth rate XAUT,t = nitrifier concentration at time t XAUT,0 = nitrifier concentration at time zero bAUT = nitrifier decay rate YAUT = nitrifier yield coefficient Substituting equation 2-9 into equation 2-7and integrating fr om time zero to time t yields: SNOx,t= SNOx,0 + [(AUT XAUT,0) / (YAUT (AUT-bAUT))] [(e((AUT-bAUT)t) 1)] 2-10

PAGE 55

39 where: SNOx,t = oxidized nitrogen concentration at time t SNOx,0 = oxidized nitrogen conc entration at time zero YAUT = nitrifier yield coefficient For estimating (AUT bAUT), non-linear regression is used to fit equation 2-10 using the measured NOx data versus time [21]. In high F/M (food to microorganisms) experiment ation, which was used in this research, bAUT values from 0.14 – 0.17 were recommended. Thes e range of decay rates were selected based on a series of experiments conducte d using various methods, testing and temperature conditions as communicated by various authors [9, 21, 64-68]. Based on these conditions a value of 0.15 days-1 was selected for the nitrifier decay rate, bAUT, used in this research. 2.8 Carbon Dioxide and Nitrification The slow growth rate and associated nitr ification rate requires a lengthy solids retention time (SRT), as much as 20 days. Previous work has demonstrated that the growth of some autotrophic bacteria is carbon limited [69-71]. Inorganic carbon was found to be a limiting factor in biological nutrient removal (BNR) systems due to the low partial pr essure of carbon dioxide (pCO2) of the atmospheric air introduced, and the loss of CO2 by stripping [72]. These factors were reported to limit the bulk concentration of CO2 in wastewater and

PAGE 56

40 consequently affect nitrification. Wett and Rauch suggest that pH is not a limiting factor per se but instead the limiting factor is the low bicarbonate concentration resulting from the low pH [72]. Addi tional evidence of the influence of CO2 on the specific growth rate of ni trifying bacteria has been de monstrated in a lab-scale, ideal mixed aerat ed reactor with CO2 concentrations of up to 17% in air [71]. These preliminary results suggest ed a strong influence of dissolved CO2 concentration on nitrification rates. Gr een et al. found a correlation between the concentration of CO2 and the ammonium oxidation rate on a nitrifying chalk reactor [70]. In this experiment, the oxid ation rate of ammonium increased as the pCO2 increased. The authors r eported that increasing pCO2 improved the rate of nitrification up to 1% CO2. Beyond wastewater treatment, elevated pCO2 was also reported to stimulate nitrification in the soil and is usua lly measured at a pCO2 of 10-2 (1% CO2). Kinsbursky and Salt zman reported that CO2 was a possible limiting substrate for nitr ifying bacteria in the soil [73]. Research conducted by these authors suggests that prov iding elevated pCO2 to the activated sludge system should incr ease nitrification rates. Based on published literature, a one percent CO2 mixture in air (17 mg/l dissolved CO2 concentration) was chosen as an initial condition for this study.

PAGE 57

41 2.9 Preliminary Research A series of experiments were conducted utilizing two 3-liter beakers set up as sequential batch reactors. One reactor re ceived air while the other received a one percent CO2 mixture in air (17 mg/l dissolved CO2 concentration). All other parameters were consistent between reactors. Results indicate that a significant increase in NH4 + conversion (three to five fold) occurred in the reactor supplied with a one percent CO2 mixture. These reactors were not pH controlled so some loss of NH4 + probably occurred as the air suppli ed reactor reached pH values as high as 8.57, thus affecting the conver sion rate. However, the loss of NH4 + could not fully account for the differences obser ved. (See chapter 3, “Stimulation of Nitrification by Carbon Dioxide in Lab-Sc ale Activated Sludge Reactors,” for a complete review of this study.) Based on results from the previous rese arch and the fact that most aeration basins are open systems with minimal pCO2 available, it was hypothesized that these organisms maybe carbon limited. Optimization of the nitrification process could be achieved by understanding the relationship of dissolved CO2 concentration on nitrifier growth rates. Based on this initial research, a seri es of experiments were conducted to determine nitrifier growth rates at contro lled pH comparing varying level of pCO2 versus an air system. Synthetic feed and influent were incorporated into the

PAGE 58

42 study and a phosphate buffer was used for pH control. (See chapter 3, “Stimulation of Nitrification by Carbon Dioxide in Lab-Scale Activated Sludge Reactors,” for a complete review of me thods used to conduct this study.) A partial list of the experiments is provided (Table 2-4). Table 2-4: Specific Growth Rate at Selected pCO2 and pH WWTP Feed pH Source (days1 ) % Improvement MLE 1 Synthetic Not Controlled Air 0.41 56 1% 0.64 MLE 1 Synthetic 7 Air 0.29 107 1% 0.6 MLE 2 Influent 7 Air 0.56 50 1% 0.84 EA Synthetic 7 Air 0.45 33 1% 0.6 EA Influent 7 Air 0.22 91 1% 0.42 EA Influent 7 Air 0.5 78 2% 0.89 EA Influent 7.5 Air 0.74 37 0.1% 1.013 As can be observed from the study, in all cases the elevated levels of pCO2 provided enhanced ni trification rates. The varying pCO2 concentrations above atmospheric levels provided enhanced nitrification, and thus may not be limited to specific dissolved CO2 concentrations.

PAGE 59

43 Chapter 3 Stimulation of Nitrification by Ca rbon Dioxide in Lab-Scale Activated Sludge Reactors 3.1 Abstract It is hypothesized that the autotrophic, ni trifying bacteria in activated sludge systems grow slowly due to CO2 limitation. To test this hypothesis, four experiments were conducted with two lab-scale reactors fed synthetic wastewater or influent from a wastewater treatment fac ility. The control reactor was supplied with air (0.03% CO2), while the experimental reactor was supplied with air containing elevated pCO2 (1%). The first exper iment was conducted with a small inoculum, no carbon source, and phosphate buffer used to maintain pH 7. A 6.9 fold increase in the rate of nitrate formation was observed in the reactor with elevated pCO2. The last three experiment s operated both reactors as sequencing batch reactors fed with synt hetic wastewater with acetate as a carbon source. The second experiment demonstrated that providing elevated pCO2 for the entire react cycle improved the nitrate formation rate, but severely degraded the solids settling performanc e. The last two experiments demonstrated a five-fold increase by providing elevated levels of pCO2 for the

PAGE 60

44 final five hours of the 7-hour react cycle without affecting so lids settling or COD removal performance. 3.2 Keywords Activated Sludge, Autotrophic, Carbon Dioxide, Nitrificati on, Nitrifying Bacteria 3.3 Introduction Nitrification is the first step for the re moval of nitrogen from wastewater, where ammonium (NH4 +) is oxidized to nitrate (NO3 -) by aerobic, autotrophic, nitrifying bacteria. These bacteria are thought to have slow growth rates and are sensitive to pH and temperature swings, making nitr ification difficult to maintain in activated sludge systems [14, 15]. T he slow growth rate and associated nitrification rate requires a lengthy solids retention time (SRT), as much as 20 days. Previous work has demonstrated that the growth of some autotrophic bacteria is carbon limited [69-71]. Inorgani c carbon was found to be a limiting factor in biological nut rient removal (BNR) system s due to the low partial pressure of carbon dioxide (pCO2) of the atmospheric air introduced, and the loss of CO2 by stripping [72]. These factor s were reported to limit the bulk concentration of CO2 in wastewater and consequently affect nitrification. Moreover, Wett and Rauch [72] suggest t hat pH is not a limiting factor per se Instead, the limiting factor is the low bi carbonate concentration resulting from the low pH. Additional evi dence of the influence of CO2 on the specific growth rate of nitrifying bacteria has been demonstrated in a labscale, ideal mixed aerated

PAGE 61

45 reactor with CO2 concentrations of up to 17% [ 71]. These preliminary results suggested a strong influence of pCO2 on nitrification rates. Green et al. [70] found a correlation between t he concentration of CO2 and the ammonium oxidation rate on a nitrifying chalk reactor. In this experimen t, the oxidation rate of ammonium increased as the pCO2 increased. They reported that increasing pCO2 improved the rate of nitrification up to 1% CO2. Beyond wastewater treatment, elevated pCO2 was also reported to stimulat e nitrification in the soil. Carbon dioxide is usually m easured in the soil at a pCO2 of 10-2 (1% CO2). Kinsbursky and Saltzman [ 73] reported that CO2 was a possible limiting substrate for nitrifying bacteria in the soil. These results suggest that providing elevated pCO2 to the activated sludge system should increase nitrific ation rates, however, addit ional research is needed to answer three fundamental questions: Does elevated pCO2 or pH depression increase the nitrification rate in activated sludge systems? When an activated sludge system is c hallenged with a lo wer target SRT, does nitrification persist with elevated pCO2?

PAGE 62

46 Does elevated pCO2 negatively impact the general performance (i.e., chemical oxygen demand removal and adequate solids settling) of the activated sludge system? Experimentation was conducted using lab-sca le reactors to investigate these three research questions. 3.4 Materials and Methods 3.4.1 Experiment 1 This experiment was conducted to determine whether elevated pCO2 or pH depression caused by elevated pCO2 was the principal cause of higher nitrification rates in bench-scale acti vated sludge systems. In addition, the conversion rate of NH4 +–N to NO3 -–N and a complete nitrogen mass balance was determined. The experiment was conduc ted based upon prev iously published guidelines [21]. Two 3 liter beakers were used for the reactors. The control reactor was fed air, while the experiment al reactor was fed a mixture of air and 1% CO2. Both reactors were fed a synthetic wa stewater with the following composition (per L): 3.33 mL of nutrient soluti on consisting of (per L): 22.65g NaH2PO4 2H2O, 27.00 g MgSO4 7H2O 10.80 g KCl, 4.20 g CaCl2 2H2O, 0.90 g

PAGE 63

47 EDTA, 0.30 g Yeast Extract, and 90 mL of trace metal solution. The trace metal solution consisted of (per L): 5.00 g FeSO4 7H2O, 0.05 g H3BO3, 1.60 g CuSO4 5H2O, 0.01g KI, 5.00 g MnCl2 4H2O, 1.10 g (NH4)6Mo7O24 4H2O, 2.20 g ZnSO4 7H2O, 0.05 g CoCl2 6H2O, and 50.0g EDTA. The synthetic wastewater and stock solutions were prepared with de ionized water from a reverse osmosis system. A series of pr eliminary experiments were conducted to establish appropriate operating conditions. Base d on these results, 58 mg/l of NH4 +-N was used as the sole nitrogen source. The di ssolved oxygen was relatively constant at 7.3 mg/l as O2, which ensured that oxygen was not limiting. Each reactor had an initial addition of 0.5 grams of sodi um bicarbonate with 0.5 gram additions at 49 and 94 hours for a total of 1.5 grams. This approa ch prevented interference with the nitrite probe, whil e providing adequate bicarb onate for nitrification. The pH was maintained between 6.95 and 7.05 through the addition of a phosphate buffer. Three phosphate buffers wit h pH values of 9.1, 7.0, and 4.4 were prepared with Na2HPO4 7H2O (pH = 9.1) and NaH2PO4 2H2O (pH = 4.4). The pH 7.0 buffer was prepared by mixing 57.7 ml of the Na2HPO4 7H2O solution and 42.3 ml of the NaH2PO4 2H2O solution. Each reac tor received identical phosphate buffer additions. The pH 7 buffer was used to equilibrate the total addition. For example, if the control reac tor required 8 ml of the pH 4.4 buffer to reach pH 7.0 and the exper imental reactor only required 5 ml of the same phosphate buffer, then an additional 3 ml of the pH 7.0 buffer was added to the experimental reactor to maintain the phos phate concentration. A total of 0.042

PAGE 64

48 moles of phosphate buffer was added to each reactor during the course of the experiment. Each reactor was inoculated with 35 ml of mixed liquor suspended solids (MLSS), that was collected from the nitrif ication basin of a full-scale activated sludge system (Glendale Wastew ater Reclamation Plant of the City of Lakeland, FL) on the same day that the experim ent was initiated. Throughout the experiment, NH4 +, NO2 -, NO3 -, pH, and dissolved oxygen (DO) were periodically measured. Experiments were discont inued when ammonium was less than 20 mg/l NH4 +-N in either the control or experimental reactor. 3.4.2 Experiments 2-4 The reactors were operated with a workin g volume of 3 liters and were seeded with 1 liter of MLSS from t he nitrification basin of a full-scale activated sludge system (Northside Wastewater Reclamation Plant of the City of Lakeland, FL), which was operated at an SRT of 22 days (Fi gure 3). For three cycles per day, both reactors were fed every cycle with 2 liters of synthetic wastewater as described for experiment 1 with the following modifications (per liter): 0.168 g of NaHCO3 and 0.850 g of C2H3O2Na 3H2O were added directly to the solution; and 32.10 g of NH4Cl was added to the nutrient solution. For Experiments 2 and 3, the synthetic wastewater and stock solutions were prepared with deionized water provided by Culligan Water (Lakel and, FL). For Experiment 4, the

PAGE 65

49 synthetic wastewater and stock soluti ons were prepared with deionized water from a reverse osmosis system. Synthet ic wastewater for experiments 2-4 had the following characteristics: Alkalinity of 100 mg/l as CaCO3, chemical oxygen demand (COD) of 400 mg/l as O2, ammonium concentration of 28 mg/l NH4 +-N, and pH of 7.6. Figure 3-1: The Experimental SBR System that Features pCO2 Control in the Experimental Reactor (left) and the Control Reactor (right) The target hydraulic retention time (HRT) for both reactors was 0.5 days, which is similar to common values for municipal activated sludge systems [2]. The cycles were automatically operat ed with a Chrontrol XT-4 (ChronTrol Corporation, San Diego, CA), that controlle d the feed pump (Masterflex L/S Pump Drive, Model 1.0% CO2 Air stones CO2 Sensor Chambe r AirPump CO2 CO2 from Gas cylinder CO2 Controller Air stones Air Pump Feed Tk Waste Pump Feed Feed Meters (Electrodes immersed in reactor) Meters Waste Tank

PAGE 66

50 7518-10, Cole-Parmer Instrument Com pany, Vernon Hills, IL), waste pump (Masterflex L/S Fixed Flow Drive, Model 7531-01, Cole-Parmer Instrument Company), and air supply system. Each sequence of cycles was 8 hours with three distinct cycles: Fill for 10 minutes at the beginning of the React cycle; React cycle for 7 hours; and Settling and Decanting for 45 and 15 minutes, respectively. The reactors were operat ed at room temperat ure (20-22C). Information regarding target SRT, length of experiment, and CO2 addition for experiments 2-4 is provided (Tab le 3-1). For Experiment 2, CO2 was supplied during the entire React cycle, w hereas for Experiments 3 and 4, CO2 was added during the last 5 hours of the React cycle. For these experiments, the activated sludge biomass was challenged by decreas ing the SRT from 8 days sequentially to 6, 4, and 2 days. Experiment 4 was designed to operate the reactors for a period equal to three times each target SRT, in order to evaluate the impact of pCO2 on nitrification for ext ended operation and performance. Table 3-1: Description of Experi ments 2 through 4 Conducted in a SBR Experiment SRT (days) Days Tested per SRT Total Days Tested Hours 1% CO2 was supplied during React cycle 2 8 6 8 3 11 Entire 7 hours 3 8 6 4 2 8 6 4 2 20 Last 5 hours 4 8 6 4 2 24 18 12 6 60 Last 5 hours

PAGE 67

51 3.4.3 Data Collection and Sample Analyses For experiment 1, measurements were ta ken at least 4 times per day with a 4 hour time interval between measuremen ts. Instruments used for chemical measurements included: ion selective electrodes (Ammonium combination glass body electrode, Cole-Parmer 27502-03 and Nitrate combination glass body electrode, Cole-Parmer 275 02-31, Cole-Parmer Instru ment Company), (Nitrite glass body electrode (Orion 9700BNWP, Thermo-Electron Corporation), Dissolved Oxygen Meter (Traceable* Port able Dissolved Oxygen Meter, Fisher Scientific), pH meter (pHTestr3+, Oakt on Instruments) and ion meters (Oakton Benchtop Ion 510 Meter and Oakton Ion 6 Meters, Cole-Parmer Instrument Company). All instruments were calibra ted daily before use. The ammonium electrode used a 0.1M NaCl filling solution (Cole Parmer 27503-78 reference filling solution, Cole-Parmer Instrum ent Company) and was calibrated with a 1,000 mg/l NH4 +-N standard solution (prepared in the laboratory with reagentgrade NH4Cl) and a 5M NaCl Ionic Str ength Adjuster (ISA). The nitrate electrode used a 0.1M (NH4)2SO4 filling solution (Cole Parmer 27503-79 reference filling solution, Cole -Parmer Instrument Company) and was calibrated with a 1,000 mg/l NO3 --N standard solution (prepared in the laboratory with reagent-grade NaNO3) and a 1M NaSO4 ISA prepared in th e laboratory.

PAGE 68

52 The nitrite electrode used an Optimum Results Type F filling solution and was calibrated with a 1,000 mg/l NO2 --N standard solution (prepared in the laboratory with reagent-grade NaNO2). A nitrite interference suppressor solution (NISS) was used for the nitrite probe measurement s to negate any bicarbonate or nitrate interference. In experiments 2-4, samples were collect ed daily during the entire React cycle to determine NO3 formation rates, pH, and DO. Sa mples of MLSS were collected daily at the end of the React cycle for settling evaluation and biomass analysis. Nitrate concentration, expressed as NO3 --N, was measured every 30 minutes during the React cycle to deter mine nitrification rates. Samples for total suspended solids (TSS), volatile suspended solids (VSS), and COD analysis were collected once per day from the mixed liquor during the last 15 minutes of the React cycl e. For the solids sample s, 45 mL of MLSS was collected and transferred to 50 mL conical tubes and stored at 4C. The sludge settling performance was eval uated by allowing 100 mL of MLSS collected at the end of the React cycle to settle in a graduated cylinder for 30 minutes and recording the sludge blanket volume. The TSS and settled sludge blanket volume measurements were then used to calculate the sludge volume index (SVI). The TSS and VSS were measured in triplicate according to Standard Methods for the Examination of Water and Wastewater Analysis [74] sections 2540D and 2540E respectively. Samples fo r COD analysis were withdrawn from

PAGE 69

53 both reactors (10 mL of MLSS) at the end of the React cycle and settled for 30 minutes. Next, the supernatant was filter ed by a syringe filter with a 25 mm diameter and 0.2 m pore size (Fisher Scie ntific). Filtered samples were stored in 15 mL conical tubes at -20C. Later, determination of COD was performed using the Reactor Digestion Method 8000 [75] for the COD range of 3 150 mg/l as O2. The vials used for this procedure (Digestion solution for COD 0-150 mg/l as O2 range, HACH Company, Loveland, CO) we re mixed with 2 ml of sample as indicated in the Method 8000 and digested for 2 hours at 150C in a digital reactor block DRB 200 (HACH Company). Vials were placed in a rack for cooling to room temperature (~21C). A portable spectrophotometer DR/2400 (HACH Company, Loveland, CO ) adjusted to a wavelengt h of 420 nm (program 430 COD LR) as indicated by the Me thod 8000 was used to read the COD concentrations of the samples. A vial mixed with 2 mL of deionized water was used as a blank. Additional vial s each mixed with 300 mg/l as O2 standard solution at different dilutions were digest ed to check the calibration curve of the spectrophotometer with defined COD c oncentrations. The effluent COD concentration was compared to the initia l COD concentration of 267 mg/l as O2 corresponding to two thirds of the COD in the synthetic wastewater (400 mg/l as O2) to obtain the COD removal efficiency.

PAGE 70

54 3.5 Results 3.5.1 Experiment 1 The results from the experiment that compared the effect of elevated pCO2 on nitrification rates at constant pH 7.0 are presented (Figures 3-2 and 3-3). The nitrate formation rate for t he control reactor was 1.50 x 10-6 mg NO3 --N/l-min, which remained relatively constant th roughout the experiment. By contrast, the experimental reactor showed an over all conversion rate of 10.3 x 10-6 mg NO3 -N/l-min, which represents a 6.9 fold in crease. The conversion rate in the experimental reactor increased throughout the experiment. During the first 42 hours, the conversion rate was 5.90 x 10-6 mg NO3 --N/l-min, while the conversion rate for the remaining 101 hours more than doubled to 12.2 x 10-6 mg NO3 --N/lmin. A loss of ammonium was observed in both reactors, but was pronounced in the control. The experimental reactor lost 6.9 mg/l of NH4 +-N or 12 % of the initial ammonium, while the control reactor lost 23 mg/l of NH4 +-N or 40% of the initial ammonium. Nitrite was not detected in the control reactor, while nitrite was present in the experimental reactor at low concentrations with a maximum concentration of 1.3 mg/l NO2 --N.

PAGE 71

55 Figure 3-2: Ammonium, Nitrite, Nitrat e, Total Nitrogen, pH, and DO for the Control Reactor in Experiment 1 Figure 3-3: Ammonium, Nitrite, Nitrat e, Total Nitrogen, pH, and DO for the Experimental Reactor in Experiment 1 6 6.5 7 7.5 8 0 10 20 30 40 50 60 0255075100125150pH Nitrogen, mg/lTime (Hours) NH4+ NO2NO3Total N DO pH 6 6.5 7 7.5 8 0 10 20 30 40 50 60 0255075100125150pH Nitrogen, mg/lTime (Hours) NH4+ NO2NO3Total N DO pH

PAGE 72

56 3.5.2 Experiment 2 An experiment was performed to determine the effect of providing elevated pCO2 with aeration throughout the React cycle in bench-scale activated sludge reactors operated as sequencing batch reactors. The positive impact of adding 1% CO2 during aeration was evident, where nitrat e formation rates in the experimental reactor were more than five times great er than the control (data not shown). Maximum nitrate formation ra tes were 0.0140 and 0.0040 mg NO3 --N/l-min for the experimental and control reactors, re spectively, while the average nitrate formation rates were 0.0080 and 0.0020 mg NO3 --N/l-min for the experimental and control reactors, respectively. Slud ge blanket volumes were greater than 40 ml/100mL and washout of biomass was only observed in the experimental reactor, whereas the control reac tor demonstrated adequate solids settling performance. For both reactors, the COD re moval efficiencies were greater than 90%. The pH in both reactors was cons istent with an averag e pH of 7.59 and 8.45 in the experimental and control reac tors, respectively, which constituted a difference in the average pH of 0.86. Upon completion of the React cycle, a difference in the pH of 0.77 was observed between the reactors with an average pH of 7.91 in the experiment al reactor and pH of 8.68 in the control reactor. The significant reduction in the pH of the experimental reactor was due to the elevated pCO2. In summary, when CO2 was supplied throughout the 7-hour React cycle, the nitrate formation rates were significantly greater and the COD

PAGE 73

57 removal efficiency was unaffected, but the solids settling performance was impacted severely. 3.5.3 Experiment 3 Based on the results of Ex periment 2, the operational c onditions were altered to reduce the impact on solids settlin g by supplying elevated pCO2 to the experimental reactor after the first two hours of every 7-hour React cycle. With this change in strategy, it was assumed that 2 hours would be ample time for the heterotrophic bacteria to consume the bul k of the COD (i.e. acetate) without being impacted by elevated CO2 levels. The remaining five hours of the React cycle would provide sufficient time for nitr ification. In order to challenge the biomass in both reactors with washout pressure, the target SRT was decreased consecutively from 8 days to 6, 4, and 2 days. The nitrate formation rates in both reac tors during Experiment 3 are provided (Figure 3-4). As can be seen from the gr aphic, the daily nitrate formation rate was always greater in the experimental re actor compared to the control reactor. Nitrate formation rates were much higher in the experimental reactor (maximum: 0.0160 mg NO3 --N/l-min; average: 0.0070 mg NO3 --N/l-min) compared to the control reactor (maximum: 0.0040mg NO3 --N/l-min; average: 0.0020 mg NO3 --N/lmin). For operation at lower SRT, the nitr ate formation rates were lower in both reactors, which may indicate washout of the nitrifying biomass. Due to

PAGE 74

58 equipment failure, the rates of NH4 + oxidation were not measured. Peak sludge blanket volumes greater than 40 ml/100 mL were observed twic e in the control reactor whereas the experimental reac tor showed adequate settling performance ( 33 ml/100 ml). This significant im provement in solids settling in the experimental reactor contra sts sharply with the results from Experiment 2. The COD removal efficiencies were greater than 90% throughout the experiment in both reactors. Similar to Experiment 2, the average pH at the beginning of the React cycle were 7.32 and 8.40 in the experimental and control reactors respectively. By the end of the React cycle, the average pH values were 8.07 and 8.78 in the experimental and control reactors, respectively, which were consistent with the result s from Experiment 2. In summary, the results for Experiment 3 suggest that the nitrifying bacteria grew fa ster when provided 1% CO2 and were able to maintain nitrificati on at a lower SRT wit hout affecting the general performance of the system (i .e. solids settling and COD removal efficiency). At a very low SRT of 2 day s, nitrification rates were much lower compared to operation at an SRT of 4 days, which may be due to washout of nitrifying bacteria.

PAGE 75

59 Figure 3-4: Nitrate Forma tion Rates for Experiment 3. The Start of Each SRT Period is Indicated by an Arrow 3.5.4 Experiment 4 To confirm the results from Experiment 3, a final experim ent was designed with the same operational paramet ers, but the operational period for a target SRT was extended for a period equal to three times each target SRT value. This experimental approach provided sufficient time for the biomass to acclimate to the conditions for each target SRT. Simi lar to Experiments 2 and 3, the average pH was 8.45 for both reactors at the begi nning of the React cycle. By the end of the React cycle, the pH va lues were 7.85 and 8.66 in the experimental and control reactors, respectively, which we re consistent with the results from Experiments 2 and 3. The nitrate formation rates for both reactors are presented in Figure 3-5. Similar to Experiment 3, the daily nitrate formation rate in t he experimental reactor was 0.000 0.005 0.010 0.015 0.020 02468101214161820NO3 --NTime (days) Control Experimental 8-day 6-day 4-day2-day

PAGE 76

60 greater than the control r eactor. Maximum nitrate fo rmation rates were 0.0120 and 0.0050 mg NO3 --N/l-min for the experimental and control reactors, respectively, which were slightly lower t han Experiment 3. For both experiments, the maximum nitrate formation rates were observed during operation at an 8-day SRT, which can be attributed to high levels of nitrifying bacteria in the inoculum. The average nitrate formation rates over the course of the entire experiment were 0.0050 and 0.0010 mg NO3 --N/l-min for the experi mental and control reactors, respectively. This five-fold in crease in the average nitrification rate is greater than the three and a hal f-fold increase from Experiment 3. In addition, the results provide evidence of high ra te nitrification at a lower SRT when elevated pCO2 is provided during aeration and the biomass is allowed to acclimate to the lower SRT operation. Figure 3-5: Nitrate Forma tion Rates for Experiment 4. The Start of Each SRT Period is Indicated by an Arrow 0.000 0.002 0.004 0.006 0.008 0.010 0.012 0.014 051015202530354045505560NO3 --N (mg/l)Time (days) Control Experimental 8-day 6-day 4-day 2-day

PAGE 77

61 Nitrate concentrations in the samples collected at the end of the React cycle for both reactors were low throughout the ex periment. These low levels of nitrate and the high SVI values (presented below ) may indicate that denitrification occurred during the settling period. Howe ver, the average concentration in the experimental reactor was twice the aver age concentration in the control reactor (data not shown). Nitrate concentrations in the supernatant did not exceed 10 mg NO3 --N/l in the experimental reactor. In both reactors, no significant impact of elevated pCO2 and low SRT operation was observed in COD removal efficiencies. Both reactors showed the same trends and had comparable values meeting the required removal efficiency of COD for secondary treatment (90%), and the supernatant concentrations were always below 30 mg/l as O2, indicating adequate perfo rmance of the system. Even though the experimental reactor exhi bited slightly higher COD removal efficiencies, they were not significant. Total suspended solids (TSS) and volatile suspended solids (VSS) were measured during Experiment 4 (Table 3-2) Both reactors had similar solids values and operating perfo rmance during the 8-day-SRT period. During the 6day-SRT period, a significant difference was observed between the two reactors. The experimental reactor showed sign ificantly lower, but stable solids concentrations. The control reactor was significantly impacted by poor solids settling performance and unintentional wa sting of biomass was observed during

PAGE 78

62 the Decant cycle. However, the solids concentration in the control reactor was much higher relative to the experimental r eactor, which is difficult to explain. During the 4-day SRT period, no discernable differences were observed in the solids concentration or operating conditions for both reactors. During the 2-day SRT period, a reduction in the solids concentration and poor settling performance was observed in both reactors. Table 3-2: Average Solids C oncentration – Experiment 4 SRT-Reactor TSS, mg/l %VSS VSS, mg/l 8-Experimental 1,803 92 1,659 8-Control 1,696 92 1,560 6-Experimental 963 95 915 6-Control 1,456 87 1,267 4-Experimental 1,170 92 1,076 4-Control 1,350 87 1,175 2-Experimental 931 92 857 2-Control 864 92 795 The solids settling performance during Expe riment 4 was evaluated by the use of the sludge volume index (SVI). Alt hough the SVI measuremen t is associated with the evaluation of clarif ier performance in full-scale activated sludge systems, it was utilized in this study to provide so me guidance on the impact of elevated pCO2 on solids settling [2]. A comparison of the SVI for both reactors during Experiment 4 revealed better overall se ttling performance in the experimental reactor, as well as better ability to re cover from the reduction of the SRT (data not shown). An SVI value greater than 150 ml/g TSS indicates poor settling and the possible proliferation of filamentous bacteria in fu ll-scale systems. For the experimental reactor, the maximum SVI was 446 ml/g TSS which was less than

PAGE 79

63 the control reactor maximum SVI of 636 ml/g TSS. Similarly, the daily average SVI for the experimental reactor was 210 ml/g TSS whereas the control reactor maintained an average daily value of 254 ml/g TSS. The settling performance of the experimental reactor was acceptable throughout the experiment except fo r the 6-day-SRT period ( days 25 to 42), when poor settling and bulking problems were observed in both reactors. Foaming was only observed during the poor settling period (d ays 30 to 40). The reduction of the SRT from 6 days to 4 days on day 42 and the subsequent absence of foaming may indicate that the foaming was due to the slow growth of foam-causing microorganisms such as Nocardia and Microthrix [76]. Poor settling was observed in the control reactor from day 6 50. Approximately 100 ml of MLSS per 8-hour cycle was unintentionally wast ed on days 27 through 29 and days 32 through 38 with corresponding SVI values great er than 300 ml/g TSS. This value is twice the value reported for biomass washout, which highli ghts the limitations of using an SBR system to fully represent full-scale systems [2]. During these periods of poor settling, bubbles were observed in t he rising sludge blanket and may indicate denitrification. Additionally viscous bulking, as suggested by the jelly-like appearance of the MLSS, was associated with the high SVI values and washout of biomass. Overall, the exper imental reactor exhibited improved solids settling performance compar ed to the control reacto r. These results are consistent with Experiment 3 and sugges t that providing elevated pCO2 for the latter portion of the React cycle reduces the negative impact on solids settling.

PAGE 80

64 3.6 Discussion 3.6.1 Effect of pH on Nitrification Results from Experiment 1 clearly demons trate increased nitrification as shown by the generation of NOx (NO2 -+ NO3 -) as a result of elevated pCO2 while pH is held constant. However, significant amm onium loss was observed in the control reactor. Some of the loss may have resulted from stripping; however, the ammonia concentration at pH 7 only constitutes 0.8% or 0.46 mg/l NH3-N of the initial ammonium concentration. An alte rnative and perhaps better explanation of the ammonium loss may be attributed to uptake of ammonium by ammonia oxidizing bacteria without subsequent nitrit e formation. Schmidt et al. reported that starving Nitrosomonas cells rapidly take up and accumulate ammonium/ammonia without simultaneous nitr ite formation [77]. Based on the results of Experiment 1, it appears that the autotrophic nitrifying bacteria in the experimental reactor were converting the ammonium due to the elevated pCO2. In the control reactor, the autotrophic nitrifying bacteria were able to accumulate ammonium in the cell, but were unable to convert the ammonium to nitrite because of carbon limitation. 3.6.2 Nitrification in Activated Sludge Systems These experimental results are consistent with the findings of other researchers, which have found a positive effect of elevated pCO2 on nitrification rates and in

PAGE 81

65 the specific growth rate of nitrifiers [71, 72, 78-80]. Alt hough nitrate formation rates were not reported by these resear chers, observed growth rates based on the increase of NOx-N concentration were reported to be approximately three times higher (1.5% CO2 vs. 0% CO2) after two hours of operation, which is similar to results from Experiments 3 and 4 [71]. Additionally, Denecke and Liebig [71] reported that the specific growth rate ( obs) of mixed autotrophic and heterotrophic sludge increased by 20% when the pCO2 was elevated to approximately 1%. Other author s also suggested a positi ve impact of elevated pCO2 on the specific growth rates of nitrifying bacteria [78, 81]. The role of pH was not evaluated on the nitrate formation rate in Experiments 24, however, it is important to consider The average pH for the experimental and control reactors were 8.03 (s.d. 0.24) and 8.57 (s.d. 0.02), respectively, which are slightly higher than the opt imal range of 7.5 – 8.0 [2 ]. Although the specific growth rate of microorganisms is sensitive to pH, it is difficult to attribute the substantial increase in nitrate formation rates to a half-unit difference in pH especially when considering the re sults from Experiment 1. The results of all four ex periments demonstrate a positive effect of elevated pCO2 on nitrate formation rates. The results from Experiments 3 and 4 suggest that CO2-sensitive nitrifying bacteria require adequate acclimation periods for low, target SRT operation, which will result in consistently higher rates of nitrification. Furthermore, the rapid improvement in the nitrate formation rates at the

PAGE 82

66 beginning of Experiments 2 4 suggest that the CO2-sensitive nitrifying bacteria are not exotic, but are commonly found in full-scale activated sludge systems. Molecular biology based methods, such as fluorescence in situ hybridizations (FISH), may be useful in identifying these CO2-sensitive nitrifyi ng bacteria. Finally, it is unknown whether elevated pCO2 may increase the specific growth rates of other autotrophic bacteria that are of importanc e in wastewater treatment, such as the ANAMMOX bacteria [82, 83]. These results suggest that the investigation of the effect of elevated pCO2 on these autotroph ic bacteria may prove to be beneficial. 3.7 Conclusions The experimental results sugges t that supplying elevated pCO2 in the aeration basin of an activated sludge system may significantly increase the nitrification rate. The primary cause of the higher nitr ification rates was determined to be the elevated pCO2 and not the pH de pression caused by increasing the pCO2. These findings also challenge the notion that nitrificat ion is a slow process and the recommendations of a lengthy SRT fo r adequate nitrification in activated sludge systems. This is significant, since it suggests that nitrif ication in full-scale activated sludge systems may be impr oved by providing elevated pCO2 to a portion of the aeration basin. In addition, this strategy may provide additional flexibility for operation with respect to the SRT.

PAGE 83

67 Chapter 4 Evaluation of Nitrifying Bacteria Specifi c Growth Rate Sensitivity to Carbon Dioxide for Full-Scale Activated Sludge and Municipal Wastewater Raymond A. Morris1, Micah J. Smith2, and Peter G. Stroot1* 1Department of Civil and Envir onmental Engineering, Univer sity of South Florida, Tampa, FL 33620 USA 2Department of Physics, Kalamazoo Co llege, Kalamazoo, MI 49006 USA Phone: (813) 396-9323 email: pstroot@eng.usf.edu October, 2009, Water Environment Federation, High-Rate Nitrification by CO2Sensitive Nitrifying Bacteria WEFTEC 2009, Orlando, FL 4.1 Abstract Biological ammonia removal in wastewater treatment plants is a slow process. It has been theorized that the dissolved CO2 concentration and pH are important parameters in optimizi ng the specific growth rate of nitrifying bacteria. Five wastewater treatment plants (WWTP) representing the three major plant configurations, extended aeration (EA), Modified L udzack-Ettinger (MLE), and

PAGE 84

68 Bardenpho, were evaluated based upon their operating conditions and activated sludge properties. The specific growth rates of the nitrifying bacteria were calculated for field and optimal conditions for pH and dissolved CO2 concentrations and suggest potential for im provement. Evaluation of nitrification in activated sludge at defined dissolved CO2 concentrations and constant pH 7 verified these findings. Fluorescence in situ hybridizations (FISH) were used to determine the abundance of nitr ifying bacteria populations in the activated sludge from each WWTP and lab-scale reactors. Changes in the community structure of the nitrifying bacteria suggest sensitivity to dissolved CO2. 4.2 Keywords Nitrification, CO2, pH, Wastewater, FISH 4.3 Introduction Nitrification is the first step for the re moval of nitrogen from wastewater, where ammonium (NH4 +) is oxidized to nitrate (NO3 -) by aerobic, autotrophic, nitrifying bacteria. These bacteria are thought to have slow growth rates and are sensitive to pH and temperature swings, making nitr ification difficult to maintain in activated sludge systems [14, 15]. T he slow growth rate and associated nitrification rate requires a lengthy solids retention time (SRT), as much as 20 days. Previous work has demonstrated that the growth of some autotrophic bacteria is carbon limited [69-71]. Inorgani c carbon was found to be a limiting

PAGE 85

69 factor in biological nut rient removal (BNR) system s due to the low partial pressure of carbon dioxide (pCO2) of the atmospheric air introduced, and the loss of CO2by stripping [72]. These factor s were reported to limit the bulk concentration of CO2 in wastewater and consequently affect nitrification. This paper evaluates the effe ct of elevated pCO2 on the specific growth rate of nitrifying bacteria using activated slud ge from three different types of BNR processes: extended-aerat ion, Modified Ludzack-Etti nger (MLE), and Bardenpho [3]. 4.4 Methodology 4.4.1 Field Evaluati on of Nitrification in Three BNR Systems Five wastewater treatment plants (W WTP) representing the three major biological nutrients removal (B NR) configurations, were evaluated in this study that include an Extended Aeration, tw o MLE, 4-stage Barden pho, and 5-stage Bardenpho. Dissolved CO2 and pH were measured in each unit operation where dissolved CO2 would be present. Dissolved CO2 measurements were collected with the OxyGuard CO2 meter. All pH values in the field were measured with an Oakton pH Tester 10. Field measurement s were collected during June and July 2009. All pH values in the laboratory were measur ed with an Oakton model 510 pH meter.

PAGE 86

70 4.4.2 pH vs. Dissolved CO2 An activated sludge sample was collec ted from the aerat ion basin of each WWTP evaluated. Within one hour of colle ction, the sample was evaluated in the laboratory to determine the pH at varying dissolved CO2 concentrations. The sample was placed in a one liter beaker in a sealed desi ccant cabinet and air or an air/CO2 mixture was introduced into the cabinet. An air pump inside the cabinet subsequently introduced the at mosphere into the beaker. The atmosphere was maintained for a minimum of 15 minutes at which time dissolved CO2 and pH were measured. 4.4.3 Specific Growth Rate Measu rement in Lab-Scale Bioreactors The experiments were conducted based upon previously published guidelines [21]. Two 3 liter beakers were used for t he reactors. The control reactor utilized air, while the experimental reactor was aerated with a mixture of air and pure CO2 to produce dissolved CO2 concentrations of 12 and 103 mg/l. The pH was maintained between 7.0+0.05 through the addition of a phosphate buffer. Each reactor received identical phosphate buffer additions. Both reactors were fed influent from t he MLE #1 WWTP. A series of preliminary experiments were conducted to establis h appropriate operating conditions. Based on these resu lts, 60mg/l of NH4 +-N was added to the influent wastewater

PAGE 87

71 which contained, on average, 25 mg/l of NH4 +-N. The dissolved oxygen was held constant at 8.3 mg/l as O2, which ensured that oxygen was not limiting. Each reactor had an initial addition of 0.5 gram s of sodium bicar bonate with 0.5 gram additions during the reaction sequence based on NH4 +-N conversion. Each reactor was inoculated with activat ed sludge that was collected from the aeration basin of the MLE #1’s activated sludge system on the same day that the experiment was initiated. A MLVSS tar get value of 35 mg/l was specified in these experiments. Thr oughout the experiment, NH4 +, NO2 -, NO3 -, pH, and dissolved oxygen (DO) were routinely m easured. A non-linear regression model was used to regresses the NOx concentration levels (NO2 + NO3 -) versus time. An estimate the maximum specific growth rate, , of the nitrifying bacteria was calculated using a non-li near regression software package (Oakdale Engineering, Oak dale, PA.). 4.4.4 Estimation of Specific Grow th Rate of Nitrifying Bacteria Growth rate optimization wa s based on Monod kinetics. An Andrew’s equation was used to determine the ef fect of the dissolved CO2 concentration on the specific growth rate [30]. The pH sens itivity of the specific growth rate was calculated by using an optimal pH of 8 as reported optimum values range from 7.5 to 8.5 [2]. Specific growth rate opt imization was based on results previously reported [71]. The paramet ers and coefficients are provided in Table 4-1.

PAGE 88

72 Table 4-1: Constants Used to Calcul ate the Optimal Specific Growth Rate for Nitrifying Bacteria Constant Value KCO2, mg/l 0.5 Ki, mg/l 42 K1for pH 1.58E-07 K2for pH 6.31E-10 max 0.75 b 0.1 pH Term Max 0.88 CO2 Term Max 0.82 The formula to determine the field and op timum specific growth rate of the nitrifying bacteria is provided: The CO2 term max is the value ob tained at a dissolved CO2 value of 5 mg/l. The pH term max is the value obt ained at a pH of 8. These values are used to normalize the formula by using the maxi mum specific growth rate for ideal dissolved CO2 concentration and pH. Denecke reported that a 5 mg/l dissolved CO2 concentration is equi valent to 0.4% CO2. When calculated using Henry’s constant, 0.4% equates to 6.89 mg/l. Fo r purposes of this study, 5 mg/l was used as the optimum CO2 concentration. Field pH measurements used in this study were calculated from activated sl udge evaluated at varying levels of CO2 concentrations in the laboratory. Al though actual field measurements are

PAGE 89

73 reported later in this paper, there was co ncern as to how well they represented actual pH values at the specified dissolved CO2 concentrations. 4.4.5 Evaluation of Nitrifying Bact eria Abundance by Fluorescence in situ Hybridization Four fluorescently-labeled ol igonucleotide hybridizati on probes, that target two ammonia oxidizing bacteria (AOB) and nitrifying oxidizi ng bacteria (NOB) groups were used in this study (Table 4-2) and were synthesized and conjugated with the cyanine dye, Cy3, before purificati on with oligonucleotide probe purification cartridges. Fluorescently labele d probes were diluted to 50 ng/ l with RNasefree water and stored at -20C in the dark. Samples (1 ml) were collected from the aeration basin from each WWTP and fix ed with 1 ml of 4% PFA for 12-24 hours. The samples were centri fuged and supernatant decanted, and suspended in 2 mL of ethanol PBS (EtOHPBS). The samples were stored at 20C until further analysis. Fixed samples were applied to a sample well on a 10 well Heavy Teflon Coated microscope slide (Cel-Line Associates, New Field, NJ) and air-dried. After dehydration with an in creasing ethanol series (50, 80, 95% [vol/vol] ethanol, 1 min each), each sample well was covered with a mixture of 18 l of hybridization buffer (20 % [vol/vol ] formamide, 0.9 M NaCl, 100 mM TrisHCl [pH 7.0], 0.1% SDS) [84] and 2 l of the stock fluorescently labeled oligonucleotide probe. The hy bridizations were conduct ed in a moisture chamber containing excess hybridization buffer (to prevent dehydration of buffer on

PAGE 90

74 sample wells) for 1.5 h, in the dark, at 46C. The slides we re washed for 30 min at 48C with 50 ml of pre-warmed washing buffer solution (215 mM NaCl, 20 mM TrisHCl [pH 7.0], 0.1% SDS, and 5 mM EDTA ) [84]. Fixed, hybridized cells were mounted with Type FF immersion oil (Carg ille, Cedar Grove, NJ) and a cover slip. Cells were stained with 4',6-d iamidino-2-phenylindo le (DAPI) at a concentration of 1 g/ml for 1 minute and rinsed with DI water. Table 4-2: FISH Probe Information Probe Targeted bacteria Reference AOB NSM156 Nitrosomonas spp ., Nitrosococcus mobilis [14] Nsv433 Nitrosospira spp. [14] NOB NIT3 Nitrobacter spp. [15] Ntspa0712 most members of the phylum Nitrospirae [85] Whole cell fluorescence was visualiz ed with an upright epiflourescence microscope (Leitz DiaPl an, Heerbrugg, Switzerland), and digital images were captured using a Spot-FLEX charge coup led device (CCD) camera (Diagnostic Instruments, Inc., Sterling Heights, MI). Images were collect ed using a 100X oil objective and constant exposure time of 1.2 sec and gain of 2. For each FISH probe, ten images were collected for each sample and analyzed based on the relative abundance of Cy3 fluorescent ce lls. Direct measurement of abundance was difficult due to the background fluores cence of the samples, thus a simple scale (Figure 4-1) was used to estimate the abundance. The value of each set of images was totaled and averaged.

PAGE 91

F 4 4 A c u c p t h d t h f a w r e u s F igure 4-1: 4 .5 Res u 4 .5.1 Field A n analysi s onfiguratio n nit proces s onditions arameters h e measu r etermined h e particul a a cility provi w here the p e ported as ses magn e ulfide whic h FISH An a u lts Evaluatio s of the th r n s was ev a s es, and e (Table 44 of primary r ements f o by obtaini n a r unit pro c ded result s robe was i n 7.35, we e e sium hydr o h causes o d a lysis Scal e n of Thre e r ee major a luated ba s ffluent (Ta 4 ). Diss o interest. T o r the diff e n g surface c ess. As s ranging f r n serted 8-1 xpected a o xide at th e d or proble m 75 e Represe n e BNR Sys t types of w s ed on the d ble 4-3) a n o lved CO2 T he dissol v e rent unit samples, w an examp l r om 26 to 5 0 feet bel o pH of 6.77 e ir lift stati o m s. This w n tation. S c t ems w astewater d issolved C n d influent concentr a v ed CO2 v a processes w hich may l e, the an o 5 8 mg/l of d o w the surf a 7 .0. The 4 o ns to neg a w ould acco u cale bars e treatment C O2 and p H propertie s a tion and a lues are r e The p H not be re p o xic zone f d issolved C a ce. Altho u stage Bar d a te the effe u nt for the e e qual 10 plant (W W H of the infl u s and oper pH were e presentati v H values w p resentativ f or the ML E C O2 in its b u gh the pH d enpho pr o cts of hydr o e levated in f m W TP) u ent, ating the v e of w ere v e for E #1 asin, was o cess o gen f luent

PAGE 92

76 pH at this facility. Only the MLE #1 WWTP received anaerobic sludge brought in from other sources. Table 4-3: Dissolved CO2 Concentration and pH of Influent, Unit Processes, and Effluent of Five Wast ewater Treatment Plants Extended Aeration MLE #1 MLE #2 4-Stage Bardenpho 5-Stage Bardenpho CO2 mg/l pH CO2 mg/l pH CO2 mg/l pH CO2 mg/l pH CO2 mg/l pH Influent 31 6.5 17 7.4 12 7.4 6 7.9 20 7.1 % Domestic 100 95 81 100 95 1 Clarifier N/A N/A 9 7.6 29 6.9 N/A N/A N/A N/A Zone 1 N/A N/A N/A N/A N/A N/A N/A N/A 31 7.0 Zone 2 (ANX) 24 6.7 26-587.35a 23-247.2 11 7.3 20 7.1 Aeration 13.5a 6.8 34 6.9 15-247.3a12 7.3 16 7.1 Zone 4 N/A N/A N/A N/A N/A N/A N/A N/A 22 7.0 2nd Aeration N/A N/A N/A N/A N/A N/A N/A N/A 23 7.0 2 Clarifier 12 6.9 23 7.1 23 7.3 11 7.4 19 7.2 PostFiltration N/A N/A 16 7.3 N/A N/A 6 7.5 12 7.4 Effluent 9 7.0 16 7.3 12 7.4 6 7.7 10 6.9 a the average of several measurements N/A: unit processes are not part of the configuration or were not in use. Large differences in t he influent dissolved CO2 concentrations were observed among the WWTP. The infl uent of the ex tended aeration plant had a high dissolved CO2 level but receives its influent through a large collection system where anaerobic conditions are quite proba ble and lead to these high readings. The 4-stage Bardenpho process, which has a low dissolved CO2 concentration, is located in a residential community wit h a limited collection system. Little time is afforded for the influent to reach anaerobic conditions.

PAGE 93

77 The MLE #2 exhibited a lower influent dissolved CO2 concentration than observed in the primary clarifier. This WWTP is fed by a large underground piping system which suggests that anaerobi c conditions are possible. On the day of the plant visit, a thunderstorm was in-progress and had increased the influent rate by 30 percent duri ng the last hour. A diluted CO2 influent concentration was recorded, while the pr imary clarifier had probably not seen the full effect of this dilution. In addition, the primary clarifier is a covered and sealed tank, which may promote anaerobic activity. The influence of the WWTP configuration is readily seen in the dissolved CO2 concentration of the aeration basins. The dissolved CO2 concentration in the anoxic basin is influenced by the mixt ure of the influent, internal recycled wastewater, and RAS combined wit h generation of dissolved CO2 by denitrification. The 5-st age Bardenpho system has the add itional contribution of dissolved CO2 from the anaerobic treatment bas in. This treated wastewater enters the aeration basin wi th an elevated dissolved CO2 concentration that ranges from 11 to 58 mg/l. In the aeration basin, dissolved CO2 is produced through the metabolism of the carbonaceou s BOD by the heter otrophic bacteria, but dissolved CO2 is also removed by stripping due to the intensive aeration. The dissolved CO2 concentration and pH were measured in unit processes beyond the activated sludge system. All WWTP are discharging final effluent with elevated dissolved CO2 concentrations when compared to the dissolved CO2

PAGE 94

78 concentration of water in equilibrium wit h the atmosphere (0.6 mg/l). The elevated level of dissolved CO2 is not surprising since the terminal unit processes do not provide adequate stripping. Table 4-4: Influent Pr operties and Activated Sl udge Operating Conditions for Five Wastewater Treatment Plants Property units Extended Aeration MLE #1 MLE #2 4-Stage Bardenpho 5StageBard enpho BOD mg/l 300 200 550 207 200 NH4 +-N mg/l 25 28 25 35 31 COD mg/l 587 N/A 1,250 N/A N/A MLSS mg/l 3,190 2,900 4,092 2,815 3,200 MLVSS mg/l 2,490 2,320 3,384 2,252 2,240 SRT days 17 12 9 25.9 15 Aeration DO mg/l 1-3 2-5 1.5-3 0.8-1.2 0.4 N/A: Not available. MLE #2 has the lowest domestic wast ewater percentage of all the plants evaluated. It services major food proce ssing industries as indicated by its high influent BOD and COD, which requires an elevated solids concentration (MLSS) to ensure proper treatment. The dissolved oxygen (DO) concentrati ons are markedly different among the WWTPs. The extended aeration and the MLE plants show expected DO levels typically encountered at wastewater fac ilities. The Bardenpho processes utilize reduced DO levels to achieve their BOD and ammonia conversions as higher DO concentrations interfere with conversion in their anoxic and anaerobic zones.

PAGE 95

79 4.5.2 Estimation of Specific Grow th Rate of Nitrifying Bacteria A sample of activated sludge from t he aeration basin of each process was obtained and evaluated at different dissolved CO2 concentrations (Figure 4-2). The numbers in the figure represent the dissolved CO2 concentrations in the aeration basin for the WWTP. Figure 4-2: Effect of pH at Varying Dissolved CO2 Concentrations Results show a general downward trend (low er pH) with increasing levels of CO2. Although different configurat ion types appear to segregate, this difference maybe more related to their MLVSS concentrations. Each WWTP was further evaluated to det ermine the potential for increasing the specific growth rate of the nitrifyi ng bacteria by optimizing the dissolved CO2 concentration and allowing for pH adjus tment. Our results suggest that improvements are possible for each WWTP evaluated in this study with the MLE 6.5 7.0 7.5 8.0 8.5 01020304050pHDissolved CO2Concentration, mg/l EA MLE #1 MLE #2 4-Bardenpho 5-Bardenpho 12 34 20 14 16

PAGE 96

80 facilities offering the greatest potential (T able 4-5). The Bardenpho processes offer less potential for improvement due to the low dissolved CO2 concentrations and higher operating pH values, whic h are near the optimum levels. Table 4-5: Optimum Specifi c Growth Rate of Nitrifying Bacteria for Optimal Dissolved CO2 Concentration of 5 mg /l and Corresponding pH Properties Extended Aeration MLE #1 MLE #2 4-Stage Bardenpho 5-Stage Bardenpho CO2, field 14 34 20 12 16 pH, field 7.17 6.92 7.01 7.57 7.26 pH, corresponding to optimal CO2 7.54 7.56 7.51 7.89 7.7 observed 0.4238 0.22 0.3226 0.5501 0.4368 optimum 0.6016 0.6058 0.595 0.6473 0.6297 % Improvement 42% 175% 84% 18% 44% 4.5.3 Evaluation of the Specific Grow th Rate of Nitrifying Bacteria Sensitivity to Dissolved CO2 Concentration using Lab-Scale Bioreactors An initial study of t he effect of dissolved CO2 concentration on the specific growth rate of nitrifying bacteria was conducted using activated sludge from the extended aeration facility. The resu lts of an analysis with pH 7.0 and CO2 concentration at 7 mg/l versus air are prov ided (Figure 4-3). The selection of the 7 mg/l dissolved CO2 (0.4%) concentration was based on previous research [71].

PAGE 97

81 Figure 4-3: Evaluation of Specific Growth Rate of Nitrifying Bacteria Using Air (Control) or 7 mg/l (E xperimental) Dissolved CO2 Concentration Both reactors display a buildup of NOX concentration (NO2 and NO3 -) over a 10 day period. However, it is evident that the rate of NOX concentration buildup is significantly higher in the experimental reac tor. The specific growth rate of the nitrifying bacteria was estimated by fitting the non-linear response. The maximum specific growth rate, max for both conditions and the associated 95% confidence interval are provided (Table 4-6). The regression analysis was conducted to NOX values of approximately 20 mg /l. Inhibition effects were observed at values greater than this concentration (data not shown). 0 10 20 30 40 50 60 0246810NOxConcentration, mg/lTime (days) Experimental Control

PAGE 98

82 Table 4-6: Estimated Specific Growth Rate of Nitrifying Bacteria and 95% Confidence Interval of the Activated Sludge from the WWTP with Extended Aeration for Two Defined Dissolved CO2 Concentrations Reactor Dissolved CO2 (mg/l) days Lower Limit Upper Limit Control 0.6 0.578 0.479 0.677 Experimental 7 1.011 0.802 1.219 Further research was conducted using activated sludge from the WWTP with MLE #1. The sludge was evaluat ed at varying levels of pCO2 from 7 to 17 mg/l at a constant pH of 7. An optimum specific growth rate of 0.84 days-1 was achieved at a dissolved CO2 of 12 mg/l. 4.5.4 Evaluation of Nitrifyi ng Bacteria by Fluorescence in situ Hybridization Representative FISH images for the samples collected from the MLE #1 and the 4-stage Bardenpho are provided in Figures 4-4 and 4-5. Individual cells and small clusters of cells are present in the flocs for each of the major ammonia oxidizing bacteria (AOB) and nitrite oxid izing bacteria (NOB). Frequent background fluorescence made enumerati on difficult, which required a more qualitative approach that utilized a relative abundance scale (Figure 4-1).

PAGE 99

F i n s F igure 4-4: n cluding ( A pp.; (C) N i Represe A ) Nitros o i trobacter s ntative FI S o monas s p s pp. and ( D 83 S H Image s p p ., Nitros o D ) most m e s for Nitrif y o c occus m embers o f y ing Bact e m obilis; ( B f the phylu e ria in ML B ) Nitroso s m Nitrosp i E #1 s pira i rae

PAGE 100

F ( A N A p a l o a F igure 4-5: A ) Nitros o N itrobacte r A nalysis of rovided (T a nd NOB i n o west abu n ppears to Represe n o monas s p r spp. and ( the digit a a ble 4-7). T n each W W n dance of A have a si m n tative FI S p p ., Nitros o ( D) most m a l FISH i m T hese val u W TP. The e A OB and N m ilar comm u 84 S H Image s o coccus m m embers o m ages usin u es show a e xtended a OB comp a u nity struc t s for 4-Sta m obilis; ( B o f the phyl u n g the rel a strong pr e a eration sy s a red to the t ure to the ge Barde n B ) Nitroso s u m Nitros p a tive abun d e sence of e s tem appe a other WW T 5-stage B a n pho inclu s pira spp. p irae d ance sca ach major a rs to hav e T Ps, altho u a rdenpho. ding ; (C) le is AOB e the u gh it The

PAGE 101

85 two MLE samples have similar NOB co mmunity structure; however the AOB appear to have some differences. Our attempts to alter the specific grow th rate of the nitrifying bacteria by operation at extreme dissolved CO2 concentrations of 12 and 103 mg/l produced interesting results. For optimal dissolved CO2 concentration (12 mg/l), the AOB populations appear to be even, while the Nitrospirae spp. appears to dominate the Nitrobacter spp. amongst the NOB. For th e extreme suboptimal dissolved CO2 concentration (103 mg/l), the Nitrosomonas spp. dominate the Nitrosospira spp. for the AOB and the NOB populations are higher but more even compared to the field sample. When compar ed to each other, the abundance of the Nitrosomonas spp. and Nitrospirae spp. appear to be similar, while Nitrosospira spp. are much higher for the reactor operating under optimal CO2 concentration and the Nitrobacter spp. are much higher for the reactor operating under suboptimal CO2 concentration. A careful review of the dissolved CO2 and pH values suggest that the 4-stage Bardenpho system should be operat ing at near optimal condit ions for nitrification. In this system, the dominant AOB appears to be the Nitrosospira spp. and the dominant NOB appe ars to be the Nitrospirae phylum. In contrast, the 5-stage Bardenpho system has a higher abundance of Nitrosomonas spp., but the Nitrosospira spp. is still dominant amongst t he NOB. The members of the

PAGE 102

86 phylum Nitrospirae are much lower relative to the 4-stage Bardenpho, while the Nitrobacter spp. is similar. Table 4-7: FISH Analysis of Five WWTP and Lab-Scale Reactors Operated at Extreme Dissolved CO2 Concentrations EA MLE #1 MLE #2 4Stage BP 5Stage BP Field 12 mg/l CO2* 103 mg/l CO2* AOB NSM156 Nitrosomonas spp., Nitrosococcus mobilis 2.40 3.30 4.90 5.00 5.60 1.00 3.50 Nsv433 Nitrosospira spp. 4.50 5.60 4.80 3.30 4.50 5.00 5.40 NOB NIT3 Nitrobacter spp. 2.73 7.20 4.00 6.20 6.73 2.80 2.44 Ntspa717 most members of the phylum Nitrospirae 1.90 5.80 6.00 6.80 5.50 5.00 1.70 pH 7 4.6 Discussion One important finding in th is study is the high concentration of dissolved CO2 in the aeration basins and other unit processes. Significant differences are evident and upon investigation are qui te plausible. As an exam ple, the aeration system on an MLE process uses three anoxic a nd four aerobic zones in a carousel arrangement to convert BOD and ammonia. A mixture of in fluent, RAS, and internal recycle from the aeration basin enter the ano xic basin, where

PAGE 103

87 denitrification generates additional dissolved CO2 as a by-product. This treated wastewater with a high level of dissolved CO2 then flows into the aeration basin where additional dissolved CO2 is generated with minimal stripping. Evidence of the impact of anoxic tr eatment and minimal CO2 stripping are observed in the MLE and Bardenpho systems. Plant influ ent also impacts the dissolved CO2 concentration in the aeration basin and a ppears to be a functi on of the influent quality and collectio n system. Finally, the dissolved CO2 concentration in the effluent is much higher than expected, when you consider that water in equilibrium with the atmosphere has a CO2 concentration of 0.6 mg/l. It is unknown whether this elevated dissolved CO2 concentration negatively impacts receiving water by providing a carbon source for the growth of algae and cyanobacteria. Evaluation of the activated sludge from the WWTPs with Ex tended Aeration and MLE #1 showed differences in the specific growth rates of the nitrifying bacteria when the dissolved CO2 concentration was optimized. The EA facility achieved a maximum growth rate at 7 mg/l CO2 while the MLE #1 facility achieved a maximum growth rate at 12 mg/l, whic h are both near the optimal dissolved CO2 concentration reported previously [71]. Th e community structure of the nitrifying bacteria in the activated sludge is expect ed to have a significant influence on the optimal dissolved CO2 concentration. It should be noted that pH was held constant at 7 and optimiz ation of the dissolved CO2 concentration will increase the pH (Figure 4-2).

PAGE 104

88 The FISH results indicate differences in the community structur e of the nitrifying bacteria amongst the WWTPs. Each facili ty appears to have its own established community of nitrifying bacteria. Thes e results show that several AOB and NOB bacteria coexist in the same system, which is similar to a previous study [23]. The four stage Bardenpho process, whic h operates near the ideal dissolved CO2 concentration, shows a dominance of one AOB ( Nitrosospira spp.) and NOB (phylum Nitrospirae ). Due to its long SRT of nearly 26 days, the presence of other microbes is not unexpec ted. This suggests that as a process approaches the ideal dissolved CO2 concentration for the growth of nitrifying bacteria, the community structure ma y become less diverse. The differences in the observed pres ence of microbes among the WWTPs as seen in the FISH analysis have one distinct possible cause (Table 4-7). The community structure of the nitrifying bacteria may simply be different due to the influent variability. This is evident in obs erving the differences in the contribution of domestic wastewater in the influent between the plants. MLE #1 and MLE #2 have distinct variability in their AO B and NOB concentrations despite having essentially the same conf iguration and operational pa rameters. MLE #1 has a very low contribution of indus trial wastewater, but is more diverse in the type of industrial wastewater it receives. MLE #2 has a large contribut ion of industrial wastewater, but consists mainly of wastew ater from food processors as indicated by the high average BO D concentration.

PAGE 105

89 FISH was used to investigate the nitrif ying bacteria in lab-scale bioreactor experiments, which were conducted at dissolved CO2 concentrations of 12 and 103 mg/l at a pH of 7.0. Compared to the seed material (MLE #1), the community structure of the nitrifying bacteria changed dramatically in unanticipated ways. Surprisi ngly, similar levels of Nitrosomonas spp. and Nitrospirae members were observed for both extreme dissolved CO2 concentrations. However, levels of Nitrosospira spp. were much greater for the optimal dissolved CO2 concentration and levels of Nitrobacter spp. were much greater for the subo ptimal dissolved CO2 concentration. In our attempts to provide optimal conditions for nitrification for the MLE #1 sludge, we were unable to produce a community structure of the ni trifying bacteria that was similar to the 4-stage Bardenpho. There may be several explanations for this failure. First, failure may be attributed to vastly different nitrifying bacteria in both samples, which would make it impossible to achieve this dominance of AOB and NOB populations present in the 4-stage Bardenpho. Second, it may be due to a lack of a wasting operation, wh ich would remove slow-gro wing nitrifying bacteria. Third, we may be underestimating the diffe rence in the effect of the influent wastewater properties. Fourth, we may be experiencing a pH effect, since the ideal dissolved CO2 concentration increases the pH of the activated sludge to 7.56, which is more than half a pH unit above the lab-scale bioreactor experiment.

PAGE 106

90 4.7 Conclusions The dissolved CO2 concentration in the influent, unit processes, and effluent of the five WWTPs evaluated in this study proved to be quite different. The dissolved CO2 concentration in the aeration basin was a function of the influent dissolved CO2 concentration, generation of dissolved CO2 through denitrification in the anoxic basin and fermentation in the anaerobic basin, dissolved CO2 concentration of both internal recycl ed wastewater and RAS, heterotrophic conversion of carbonaceous BOD to CO2 in the aeration basin, and limited CO2 stripping in the aeration basin. The microbial ecology of the nitrifying bacteria of the plants appears to be plant specific, but commonalities are evident. Further research is planned to optimize the conditions for nitrification for each type of process and to evaluate the microbial ecol ogy of the nitrifying bacteria for those conditions.

PAGE 107

91 Chapter 5 Determination of the Rela tionship of Dissolved CO2 Concentration and pH and a Design Space for Optimum Nitrification Based on the field study results as report ed in chapter 4, a series of designed experiments were conducted to asce rtain if an optimum dissolved CO2 concentration/pH condition exists that maximizes specific growth rate. Experiment one was conducted to determine t he effect of varying concentrations of dissolved CO2 at a constant pH of 7.0. Experiment 2 was conducted at varying concentrations of dissolved CO2 at specific pH levels that coincide with sludge from a WWTP. Experiment three was conducted at varying concentrations of dissolved CO2 and pH to determine a design space for optimum nitrification. 5.1 Methodology and Materials Three experiments were conducted to det ermine the maximum specific growth rate of the microbes at vary ing levels of dissolved CO2 concentrations. The experiments were conducted based upon previously published guidelines [21]. In experiments 1 and 2, six one liter beak ers were used for the batch reactors and were filled to 800 ml using influent fr om a commercial wastewater facility.

PAGE 108

92 The dissolved CO2 concentrations fed to the batch reactors in the first experiment were 134, 61, 29, 12, 7 and 2 mg/l, respecti vely. In the seco nd experiment, the dissolved CO2 concentrations were 34, 25, 19, 16, 12 and 8, respectively. In the third experiment, 12 one liter beakers were used for the batch reactors and were filled to 800 ml using influent from a commercial wastewater facility with dissolved CO2 concentrations maintained at 5, 10 and 15 mg/l, respectively. Deionized water from a reverse osmosis system was used to replenish water in the reactors during the experiment. Establishment of the CO2 percentages was conduct ed using a dissolved CO2 meter (OxyGuard CO2 Portable Analyzer). T he measured dissolved CO2 was compared to the theoretical dissolved va lue based on Henry’s constant. An R2 of 0.9978 was achieved. A series of preliminary experiments we re conducted to establish appropriate operating conditions. Based on these results, 60 mg/l of NH4 +-N was used as the sole nitrogen source and added to influent waste water from a commercial waste water treatment facility. A MLE proce ss was selected. The dissolved oxygen concentration was constant at 8.2 mg/l as O2, which ensured that oxygen was not limiting. Alkalinity was maintained in all experiments at approximately 250 mg/l as CaCO3.

PAGE 109

93 In experiment 1, each reac tor had an initial addition of 0.2 grams of sodium bicarbonate with 0.1 gram additi ons at 94 hours for a total of 0.3 grams. The pH was maintained between 6.95 and 7.05 th rough the addition of a phosphate buffer. Three phosphate buffers with pH values of 9.1, 7.0, and 4.4 were prepared with Na2HPO4 7H2O and NaH2PO4 2H2O. Each reactor received identical phosphate buffer additions. The pH 7 buffer was used to equilibrate the total addition. For example, if the cont rol reactor required 8 ml of the pH 4.4 buffer to reach pH 7.0 and the experimen tal reactor only required 5 ml of the same phosphate buffer, then an additional 3 ml of the pH 7.0 buffer was added to the experimental reactor to maintain t he phosphate concentration. A total of 0.019 moles of phosphate buffer was added to each reactor during the course of the experiment. In experiment 2, each reac tor had an initial addition of 0.3 grams of sodium bicarbonate with no further additions. The pH was maintained at values appropriate for the dissolved CO2 concentration. These values were determined by aerating a sample of activated sludge from the treatment facility used in this study and recording the pH value at varying levels of dissolved CO2 concentration. A total of 0.011 mole s of phosphate buffer was added to each reactor during the course of the exper iment. Additional measurements were taken to minimize variation in the growth rate parameter.

PAGE 110

94 In experiment 3, each reac tor had an initial addition of 0.3 grams of sodium bicarbonate with further additions to maintain pH. The pH was maintained at 8.0, 7.5, 7.0 and 6.5. A total of 0.02 mo les of phosphate buffer was added to each reactor during the course of the exper iment. Additional measurements were taken to minimize variation in the growth rate parameter. Six sealed desiccant cabinets were used to maintain the appropriate atmospheres. PVC tubing was used to c onnect the cabinets in series. Rena air pumps (Air 50, 2.0 watts) were used to introduce air into the cabinets. An Optima air pump (4.5 watts ) was used in the first c abinet to ensure an adequate system air flow. A carbon dioxide sensor (COY laboratory products) was installed in the first cabinet to establish the initial pCO2 atmosphere. Each reactor was inoculated with an a ppropriate volume of mixed liquor suspended solids (MLSS), that was collected from the nitrificat ion basin of a fullscale activated sludge system on the same day that the experiment was initiated (South Cross Bayou Water Reclamation Faci lity of the City of St. Petersburg, FL). These volumes were 14, 11 and 12 ml for experiment 1, 2 and 3, respectively. (The volumes were based upon the MLSS concentration on the day the sample was obtained.) Throughout the experiment, NH4 +, NO2 -, NO3 -, pH, and dissolved oxygen (DO) were perio dically measured. Experiments were discontinued when the combined NO2 and NO3 concentrations totaled 30 mg/l or greater. This was done to negate inhibition effects.

PAGE 111

95 Experimentation was conduct ed with an ammonia concentra tion in the reactor high enough (relative to the half velo city constant from Monod kinetics, Ks) to ensure that the nitrification rate is at a maximum [21]. Sixty mg/l of NH4 + was significantly greater than the Ks values reported for the ammonia oxidizers, 1.0, and the nitrite oxidiz ers, 1.3, at 20oC [12]. The growth rates were modeled using a non-linear regression equ ation as described in Methods for Wastewater Characterization in Activated Sludge Modeling [21]. The growth rate expression is provided: The parameters SNOx,0, (oxidized nitrogen concentr ation at time zero), umax (maximum specific nitrifier growth rate) and XAUT,0 (initial nitrifier concentration) were calculated using this equation. YAUT (nitrifier yield coefficient) and bAUT (nitrifier decay rate) were give n values of 0.15 mg VSS/mg NH4 + and 0.15 days-1, respectively. Software from Oakdale Engineering (Oakdale, PA) was used to conduct the non-linear regression modeling. Software from Minitab, Inc. (State College, PA) was used to analyze the experimental design and generate other statistics. Microsoft Excel was used to estimate gr owth kinetics for both experiments and compared to findings published by Denecke The following e quation, which is based upon an Andrews’s model [30], wa s used to estimate the kinetics:

PAGE 112

96 The parameters max, Ks (saturation constant fo r substrate), Ki (inhibition constant), K1 and K2 are ca lculated using this equation. Specific growth rate ( obs), dissolved CO2 concentration ([CO2]), and proton concentration ([H+]) were measured or specified dur ing experimentation. 5.1.1 Data Collection and Sample Analyses Measurements for experiment 1 were taken at least 3 times per day with a 4 hour time interval between measurements. Measurements for experiment 2 were taken approximately every one and one-hal f hour over 10-12 hours per day. Measurements for experiment 3 were ta ken every 1.5 hours over 18-20 hours per day. Holes were drilled into the top of the cabi nets where rubber stoppers were installed. During measurement ta king, electrodes were lowered through the holes and placed into the r eactors. This was done to mi nimize atmospheric loss. The electrode wires were encapsulated in a rubber stopper cut to facilitate the wire. Stoppers were replac ed after taking measurements. Instruments used for chemical measurements in cluded: ion selective electrodes (Ammonium combination glass body electrode, Cole-Parmer 27502-03 and Nitrate combination glass body electrode, Cole-Parmer 27502-31, Cole-Parmer

PAGE 113

97 Instrument Company), (Nitrite comb ination electrode ( 4230-A94, Thomas Scientific), Dissolved Oxygen Meter (Traceable* Portable Dissolved Oxygen Meter, Fisher Scientific), pH meter (pHTestr3+, Oakton In struments) and ion meters (Oakton Benchtop Ion 510 Meter and Oakton Ion 6 Meters, ColeParmer Instrument Company). All instrume nts were calibrated daily before use. The ammonium electrode used a 0.1M NaCl filling solu tion (Cole Parmer 27503-78 reference filling solution, Cole -Parmer Instrument Company) and was calibrated with a 1,000 mg/l NH4 +-N standard solution (prepared in the laboratory with reagent-grade NH4Cl) and a 5M NaCl Ionic Str ength Adjuster (ISA). The nitrate electrode used a 0.1M (NH4)2SO4 filling solution (Cole Parmer 27503-79 reference filling solution, Cole -Parmer Instrument Company) and was calibrated with a 1,000 mg/l NO3 --N standard solution (prepared in the laboratory with reagent-grade NaNO3) and a 1M NaSO4 ISA prepared in t he laboratory. The nitrite electrode was ca librated with a 1,000 mg/l NO2 --N standard solution (prepared in the laborat ory with reagent-grade NaNO2). The nitrite combination electrode was found to be pH sensitive. Using a pH of 7.0 as the reference pH, a pH of 6.5 exhibi ted a 10 percent higher nitrite reading. A pH of 7.5 and 8.0 exhibited lower readings of 91 perc ent and 77 percent, respectively. Adjustments were made to the maximum specific growth rates results as these were based on NOx (NO2 + NO3 -) concentration. However, these adjustments had minor to no effect if the results were not corrected.

PAGE 114

98 5.2 Results 5.2.1 Experiment 1 The results from experiment 1 comp ared the effect of dissolved CO2 concentrations on nitrification rates at const ant pH 7.0. In analyzing the results, the nitrite (NO2 -) and nitrate (NO3 -) concentrations were added together and then regressed against time in order to estimate the specific growth rate. The growth curve for a dissolved CO2 concentration of 12 mg/l an d at a pH of 7.0 using the Oakdale Engineering software is provided (Figure 5-1). Concentrations of the NO2 ranged from 0.5 0.7 mg /l in the 7 103 mg/l dissolved CO2 concentrations. The NO2 in the 2 mg/l dissolved CO2 concentration ranged from 0.8 1.2 mg/l. These represented small concentrations compared to the nitrate, re mained relatively constant, and did not accumulate over time. The concentrations of the NO2 and NO3 were regressed separately to ascertain differences from their combination (data not shown). No differences were noted.

PAGE 115

F p F V A c r e F igure 5-1: p H of 7.0 F igure 5-2: V arying Le v A maximu m oncentrati o e sults of 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0max, d-1 Estimate d Experi m v els of Di s specific g r o n of 12 m g .8 day-1 fo 20 d max at 1 2 m ent 1 Sp e solved C O r owth rate g /l (Figure 5 r ammoni u 40 Dissolv — 99 2 mg/l of D e cific Gro w O 2 Concen t of 0.84 da y 5 -1). This i s u m-oxidizin g 60 ed CO2C o — Growth R D issolved C w th Rate t ration at p y -1 was ac h s in agree m g and nitr a 80 o ncentrati o R ate C O2 Conce of Nitrifyi n p H 7.0 h ieved at a m ent with t y a te-oxidizin g 10 0 o n, mg/l ntration a n n g Bacter i a dissolved y pical publi g bacteria 0 12 0 n d a ia at CO2 shed [30]. 0

PAGE 116

100 The shape of the curve in figure 5-2, which appears similar to a log normal distribution, is indicative of inhibiti on and can be described by means of an Andrews model [30]. The Andrews model is based on a modification of the Monod equation and incorporates an inhibitory coefficient. Extremely low and elevated dissolved CO2 concentrations produce unfav orable growth conditions. Good model fits were achieved for each dissolved CO2 concentration with regression model R2 values ranging from 0.92 0.98. 5.2.2 Experiment 2 Based on the results from experiment 1, a second experiment (replicated) was conducted to determine an optimum specific growth rate based upon a dissolved CO2 concentration and its associated pH val ue (Table 5-1).The pH values were determined using activated sludge from a WWTP (Figur e 4-2). (See chapter 4 for a complete review of this study.) Table 5-1: Experiment 2 pH vs. Dissolved CO2 Concentration Dissolved CO2 Concentration, mg/l pH 8 7.32 12 7.23 16 7.16 19 7.1 25 6.98 34 6.86 An optimum specific growth rate of 1.05 days-1 was achieved at a dissolved CO2 concentration of 8 mg/l (Figure 5-3).

PAGE 117

101 Figure 5-3: Experiment 2 Growth Curve at Selected Dissolved CO2 Concentrations with 95% Confidence Levels Except for the 8 mg/l dissolved CO2 concentration in reactor 1 (R2 = 0.81), good model fits were achieved for each dissolved CO2 concentration with regression model R2 values ranging from 0. 92 0.98. (The nitrate electrode used for the 8 mg/l dissolved CO2 concentration in reactor 1 failed six days into the experiment and was replaced. This was the caus e of the increased variation. The experiment was conducted for ten days ensuring adequate observations were taken.) The decrease in specific grow th rate appears linear but both reactors due show a marked decrease when the dissolved CO2 concentration increases from 16 19 mg/l. From 19 34 mg/l, the s pecific growth rate remained relatively constant. 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 010203040max, d-1Dissolved CO2Concentration, mg/l Reactor #1 Reactor #2

PAGE 118

102 5.2.3 Growth Kinetics Experiments 1 and 2 were evaluated and their grow th parameters were compared to those reported by D enecke (Tables 5-2 & 5-3). Table 5-2: Combined Growth Parameters for Experiment 1 Table 5-3: Combined Growth Parameters for Experiment 2 (The values reported by Denecke were for mixed sludge at 0.99 percent CO2.) A set of parameters could not be found that described both sets of operating conditions describing experi ments 1 and 2. Evaluating the curves generated by experiment 2 (Figure 5-3), it was hypot hesized that a microbial population shift occurred between 16 and 19 mg/l dissolved CO2 concentration. These dissolved CO2 concentrations represent pH values of 7.1 and 6.98, respectively, and experiment 1 was conducted at a pH of 7. Though the pH difference is minor, in combination with the elevated dissolved CO2 concentration, a microbial shift is Parameter Units Biomass 1 Biomass 2 Denecke Values [71] Ks mg CO2/l 1.5 0.45 0.5 Ki mg CO2/l 50 25 42 max days1 2.5 0.9 0.75 K1 2E-7 9E-6 6.99E-7 K2 1E-9 5E-8 1.25E-10 Parameter Units Biomass 1 Biomass 2 Denecke Values [71] Ks mg CO2/l 1.1 1.1 0.5 Ki mg CO2/l 44 70 42 max days1 2.2 1.7 0.75 K1 2E-7 9E-6 6.99E-7 K2 1E-9 5E-8 1.25E-10

PAGE 119

103 possible. Therefore, a pr oportion of the microbial pop ulations was hypothesized and given values of 0.3, 0.35, 0.45, 0.75, 0.80, and 0.83. These values represent the 2 103 mg/l dissolved CO2 concentrations in experiment 1. Their complement, 1 proportion, repres ents the 8 34 mg/l dissolved CO2 concentration in experiment 2. These percentages were developed by first evaluating experiment 2 so an appropriate set of parameters could be developed that adequately described the inflecti on from 16 to 19 mg/l dissolved CO2 concentration (Figure 5-3). Parameters developed produced a good fit for the specific growth rate at its dissolved CO2 concentration (Tables 5-2 and 5-3 and Figures 5-4 and 5-5). T he values reported by Dene cke were used as starting point values and fits were calculated by minimizing the difference sums of squares for the model. Figure 5-4: Composite Biomass Describing max from Experiment 1 with 95% Confidence Levels 0 0.2 0.4 0.6 0.8 1 1.2 1.4 020406080100120max, d-1Dissolved CO2Concentration, mg/l Composite Experiment 1

PAGE 120

104 Figure 5-5: Composite Biomass Describing max from Experiment 2 with 95% Confidence Levels 5.2.4 Experiment 3 Based on the results from experiment 2, a third experiment was conducted to determine an optimum specific growth rate based upon an observed optimum dissolved CO2 concentration (8-12 mg/l) from previous experimentation Results indicate that a combination of dissolved CO2 concentration and pH produce significant growth rate differ ences (Figures 5-6 to 5-9). 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 010203040max, d-1Dissolved CO2Concentration, mg/l Composite Experiment 2

PAGE 121

105 Figure 5-6: Experiment 3 Results of max at Selected pH and Dissolved CO2 Concentration with 95% Confidence Levels Figure 5-7: Experiment 3 Result s of Main Effects Plot for max 0 0.2 0.4 0.6 0.8 1 1.2 1.4 357911131517MAX, d-1Dissolved CO2Concentration, mg/l pH=8.0 pH=7.5 pH=7.0 pH=6.5 15 10 5 1.0 0.9 0.8 0.7 0.6 0.5 8.0 7.5 7.0 6.5 CO2Mean pH

PAGE 122

106 Figure 5-8: Experiment 3 Results of AOB at Selected pH and Dissolved CO2 Concentration with 95% Confidence Levels Figure 5-9: Experiment 3 Results of NOB at Selected pH and Dissolved CO2 Concentration with 95% Confidence Levels 0 0.2 0.4 0.6 0.8 1 1.2 1.4 357911131517AOB, d-1Dissolved CO2Concentration, mg/l pH= 8 pH=7.5 pH=7 pH=6.5 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 357911131517NOB, d-1Dissolved CO2Concentration, mg/l pH=8.0 pH=7.5 pH=7.0 pH=6.5

PAGE 123

107 Growth rates at a pH of 8 are approximately twice the va lues of those at pH of 6.5 and 7. It is also evident that lo wer growth rates ar e observed at low dissolved CO2 concentrations. This relations hip was observed in previous experimentation (Figure 5-2). The data from the dissolved CO2 concentration of 5 mg/l at a pH of 7.0 is not displayed as this reactor received twice the activated sludge aliquot, thereby skewi ng the results. A main effects plot for max clearly shows the effect of dissolved CO2 concentration and pH on specific growth rate (Figure 5-7). The growth rates for the AOB and NO B microbes were evaluated at each CO2/pH combination using non-linear regressi on as previously reported. It is clearly evident t hat at dissolved CO2 concentrations of 10 and 15 mg/l at a pH of 8, significant AOB growth rates occur. This relationship is not observed at lower pH. Also included are the 95% CI for each microbe at the specific dissolved CO2 concentration/pH combination. With additional measurements, low standard errors were achieved. 5.3 Discussion 5.3.1 Experiment 1 Experiment 1 was conducted as an un-repl icated completely randomized design (CRD). The NO2 and NO3 concentrations were summed and regressed against

PAGE 124

108 time to determine the maximum specific growth rate (max). As this experiment was non-replicated, degrees of freedom are not available to calculate an error term and thus generate an appropriate analysis of variance (ANOVA). Evaluation and comparison of the specific growth rate curves was possible by conducting a two-sample Kolmogorov-Smirnov (KS) test. As a non-parametric test, no assumptions about the parameters of a distribution nor is its underlying distribution are made. The null hypothesis for this test is that the two samples have the same distribution. Evaluatio n of the 7 and 12 mg/l dissolved CO2 curves ( = 0.76 days-1 and 0.84 days-1, respectively) provided a p-value of 0.829 indicating the underlying distributions are ve ry similar. Evaluating the 103 and 12 mg/l dissolved CO2 curves, which had the largest di fferences in m values (0.16 days-1 and 0.84 days-1 respectively), provided a pvalue of 0.147. Although not statistically significant, results indica te a marked departure in their underlying distributions. This is not unexpect ed given the large differences in max values. Evaluation of the 103 mg/l dissolved CO2 curve showed the results to be linear (R2 = 0.951) indicative of a normal dist ribution instead of an expected exponential distribution. Indicating gr owth inhibition is evident. 5.3.2 Experiment 2 Experiment 2 was conducted as a replicated CRD. Anal ysis indicated significant differences among the varying dissolved CO2 concentration (Figure 5-3). An

PAGE 125

109 average optimum specific gr owth rate of 1.0 days-1 was achieved at a dissolved CO2 concentration of 8 mg/l. The dissolved CO2 concentrations from 8 16 mg/l show a downward trend to a constant grow th rate from 19 34 mg/l. Growth rates in reactor 2 were always lower than reactor 1. The significantly lower growth rate in reactor 2 at a dissolved CO2 concentration of 12 mg/l cannot be explained. Based on the results, a predi cted specific growth rate of 0.91 days-1 should have been observed. It is believed that the reactor was inadvertently contaminated during the experiment. The ANOVA conducted for this experiment shows significant differences at the specified dissolved CO2 concentrations (Table 5-4). Table 5-4: Experiment 2 Results of Completely Randomized Design of max at Selected Dissolved CO2 Concentrations Source DF SS MS F P-Value Dissolved CO2, mg/l 5 0.16167 0.03233 4.92 0.039 Error 6 0.03940 0.00657 Total 11 0.20107 S = 0.08103 R-Sq = 80.40% R-Sq (ad) = 64.07% Pooled Standard Devi ation = 0.0810 Individual 95% CIs For Mean Ba sed on Pooled Standard Deviation Level N Mean Std Dev -----+---------+--------+---------+-8 2 1.0050 0.0354 (-------*--------) 12 2 0.8350 0.1768 (-------*--------) 16 2 0.8550 0.0495 (------*--------) 19 2 0.7000 0.0141 (--------*--------) 25 2 0.7200 0.0424 (--------*--------) 34 2 0.6650 0.0495 (--------*-------) -----+---------+--------+---------+-0.64 0. 80 0.96 1.12

PAGE 126

110 As this is a balanced experimental desi gn, a Tukey multiple comparison test was selected and shows only the 34 and 8 mg/l dissolved CO2 concentrations to be statistically different (p-value = 0.039). It is evident that ma jor differences do exist but are masked by the large variat ion occurring at the 12 mg/l dissolved CO2 concentration (Figure 5-10). Figure 5-10: Experiment 2 Results of Boxplot of Completely Randomized Design Showing max at Selected Dissolved CO2 Concentrations 5.3.2.1 Effect of pH on Nitrification A study was conducted to determine the pH of an activated sludge sample at varying levels of dissolved CO2 concentration (Table 5-1) Measurement taken in the aeration basin of two Modified Ludza ck-Ettinger (MLE) wastewater plants had 34 25 19 16 12 8 1.0 0.9 0.8 0.7 0.6 C1C2 Dissolved CO2Concentration, mg/l max, d-1

PAGE 127

111 dissolved CO2 concentrations of 35 and 26 mg/l. This would equate to pH values of 6.86 and 6.98, respectively. Achievi ng a pH of 7.5 would require that the dissolved CO2 concentration be reduced to a value less than 6 mg/l. However, this “optimum pH value” is moderately higher than the pH va lue of 7.32 which is achieved at the optimum growth rate va lue of 8 mg/l found during experiment 2. Although experimental error does exist, the e ffect of pH cannot be understated. This is evidenced in the different specif ic growth rates at similar dissolved CO2 concentrations for the two experim ents. The 7 mg/l dissolved CO2 in the first experiment had a grow th rate of 0.76 days-1 while the 8 mg/l dissolved CO2 concentration in the second experiment had a growth rate of 1.0 days-1. The increased growth rate was achieved with a pH difference of only +0.3 units (pH 7.0 versus pH 7.32). 5.3.2.2 Growth Kinetics Good fits of the model parameters were achieved. However, these are not optimum as other conditions may satisfy and achieve good model fits. Values were selected that had reasonable agreem ent with those reported by Denecke. These results indicate that determini ng specific growth rates may be more complicated than previously reported. It appears that different microbial populations can exist at different pH and experimental condit ions will dictate which set of AOB/ NOB organisms are mo st in abundance. In evaluating these

PAGE 128

112 models, different maximum specific grow th rates were obtained when evaluating these two hypothesized sludges. Thus, suggesting that maximum specific growth rates of some AOB/ NOB organism s may be much higher than previously reported. 5.3.2.3 Nitrification in Activated Sludge Systems These experimental results are consistent with the findings of other researchers, which have found a positive effect of elevated pCO2 on nitrification rates and in the specific growth rate of nitrifiers [71, 72, 78-80]. Alt hough nitrate formation rates were not reported by these research ers, observed growth rates based on the increase of NOx-N concentration were reported to be approximately three times higher (1.5% CO2 vs. 0% CO2) after two hours of operat ion. Evaluation of nitrification using air was not conducted in this study as observations at various treatment plants show much higher dissolved CO2 concentrations in their processes. Additionally, Denecke and Liebi g [71] reported t hat the specific growth rate ( obs) of mixed autotrophic and heterotrophic sludge increased by 20% when the pCO2 was elevated to approximatel y 1% (17 mg/l dissolved CO2). Other authors also suggested a positive im pact of elevated pCO2 on the specific growth rates of nitrifying bacteria [78, 81].

PAGE 129

113 5.3.3 Experiment 3 Experiment 3 was conducted as a non-replicated CRD due to limited laboratory equipment. The ANOVA conducted for th is experiment shows significant differences for pH levels but dissolved CO2 concentrations were not significant at the = 0.05 level (Figure 5-6 to 5-9 and Ta ble 5-5). However, a p-value of 0.058 does indicate that dissolved CO2 concentration is influencing the growth rate of the nitrifiers. A partitioning of the sums of squares of the tr eatment effects shows that dissolved CO2 contributes 17.4% to m odel understanding while pH contributes 82.6%. This is shown graphically in Fi gure 5-7. These results indicate that pH is the dom inant factor that affects t he nitrification rate at the specified dissolved CO2 concentrations and pH levels used in this experiment. Table 5-5: Experiment 3 ANO VA of Main Effects of max Factor Type Levels Values CO2 fixed 3 5, 10, 15 pH fixed 4 6.5, 7.0, 7.5, 8.0 Analysis of Variance for max Source DF SS MS F P-Value CO2 2 0.092056 0.046028 5.33 0.058 pH 3 0.438156 0. 146052 16.91 0.005 Error 5 0.043194 0.008639 Total 10 0.573406 S = 0.0929456 R-Sq = 92.47% R-Sq (adj) = 84.42%

PAGE 130

114 A multiple comparison test for the main effects indicates differences (Table 5-6). Dissolved CO2 concentration does not i ndicate differences at the = 0.05 level. However, pH does show differences wher e a pH of 8.0 is different from pH values of 7.0 and 6.5. At a pH of 7.5, results show that this level exists in both groups. Table 5-6: Multiple Comparisons of Factor Effects Using Tukey Method with a 95.0% Confidence Level Factor Level N Mean (max)Grouping* Dissolved CO2 Concentration 15 4 0.8 A 10 4 0.7 A 5 4 0.5 A pH 8.0 3 1.0 A 7.5 3 0.7 A & B 7.0 2 0.5 B 6.5 3 0.5 B *Means that do not share a lette r are significantly different. An interaction plot was generated to gr aphically assess the relationship between the main effects (Figure 5-11) These results indicate that as pH increases the maximum specific growth increases. Some effect of the dissolved CO2 concentration on this increased growth ra te can be observed. The effect of dissolved CO2 concentration shows similar growth effects at different pH values indicating that a significant interaction e ffect is probably not evident. It must be emphasized that these results are not a ccompanied by a statis tical analysis and therefore a p-value cannot be generated to confirm the pr esence of an interaction effect.

PAGE 131

115 8.0 7.5 7.0 6.5 1.2 1.0 0.8 0.6 0.4 15 10 5 1.2 1.0 0.8 0.6 0.4 5 10 15 6.5 7.0 7.5 8.0 Figure 5-11: Interaction Plot of max as a Function of Dissolved CO2 Concentration and pH Increased maximum specific growth rates were observed at each dissolved CO2 concentration as pH increased from 6.5-8.0. At a dissolved CO2 concentration of 10 and 15 mg/l, the maximum specific gr owth more than doubled when the pH increased from 6.5 to 8.0 (Figures 5-6 and 5-7). The results from the dissolved CO2 concentration at 10 and 15 mg/l and at a pH of 8.0 were surprising but not unexpected. Nitrosomonas, identified as the major AOB bacteria found in wastewater, has an ideal gr owth condition at pH 7.88.0. As conditions were selected that promote the growth of this microbe, high growth rates were observed. AOB bacteria are considered limi ting in the conversion of ammonia to nitrate, but this comb ination of dissolved CO2 concentration and pH favored elevated AOB concentrations (Figure 5-8). CO2 pH max max

PAGE 132

116 Nitrobacter, identified as the major NO B bacteria found in wastewater, has an ideal growth condition at pH 7.3-7. 5. Many WWTP's oper ate at these pH conditions favoring the growth of this microbe. Unfortunat ely, at this pH the NOB does not display the elevated growth rate ob served at pH 8.0. Growth rates for the NOB bacteria were mix ed depending on the dissolved CO2 concentration and pH (Figure 5-9). Observations from experiments 1 and 3 do suggest that there is a lower dissolved CO2 concentration limit. Experim ent 1 had a lower dissolved CO2 concentration of 2 mg/l and a growth rate of 0.43 days-1. Experiment 3 had a lower dissolved CO2 concentration of 5 mg/l and a gr owth rate of 0.41 – 0.53 days-1 depending on the pH. The experim ental results suggest a minimum dissolved CO2 concentration between 5-10 mg/l is needed to obtain satisfactory nitrification rates. 5.4 Conclusions Low and high levels of dissolved CO2 concentration result in inhibition and reduce the nitrification ra te. Though experimentation was only conducted at one pH level (7.0), similar results are ex pected at other pH levels. Further experimentation with adjusted pH leve ls based on activated sludge from a WWTP (Table 5-1), showed reduced nitrific ation rates at these higher dissolved CO2 concentrations. It was hypothesized t hat these were due to shifts in

PAGE 133

117 microbial ecology that were less conducive to nitrification. Even so, whether the effect was due to a lower pH level or increased dissolved CO2 concentration is unknown requiring further experimentation. Incorporating results from previous expe rimentation, an optim ization experiment found an optimum dissolved CO2 concentration range of 10 -15 mg/l and a pH range of 7.5 – 8.0. A partition of the su ms of squares treatment show that pH contributes to 83 percent of model understanding. Dissolved CO2 concentration does contribute to nitrification (17 percent ) but is minor when pH is optimized.

PAGE 134

118 Chapter 6 FISH Analysis of Microbial Sampl es Collected from Batch Reactors Operated at Different Dissolved CO2 Concentrations and pH 6.1 Introduction The optimization studies conducted in this research (Chapter 5) showed that maximum specific growth rates vary depending on the dissolved CO2 concentration and pH. Dissolved CO2 concentration was established at 5, 10 and 15 mg/l with pH values maintained at 6.5, 7.0, 7.5 and 8.0. These values were selected based on previous resear ch and literature recommendations. Upon completion of each CO2/pH combination reactor experiment, a biomass sample was obtained for microbial assess ment. It was theorized that certain ammonia oxidizing bacteria (AOB) and ni trite oxidizing bact eria (NOB) would predominate at these reactor conditions. The AOB bacteria Nitrosomonas and Nitrosospira and the NOB bacteria Nitrobacter and Nitrospirae were selected as they are the most frequently mentioned bacteria specie s found in literature as related to wastewater.

PAGE 135

119 6.2 Methods and Materials A total of 52 samples (10 digital images each) were prepared for this study. These samples were obtained from experiment three reviewed in chapter 5. The samples were the combination of dissolved CO2 concentration and pH by bacteria microbe (AOB and NOB) as well as the seed activated sludge by microbe. The large sample size wa s obtained to minimize variation. Fluorescence in situ hybridizations (FISH) was used to evaluate the population abundance of the common AOB and NOB bacteria listed above. For a complete review of this molecular bi ological technique, see the section entitled, “Evaluation of nitrifying bacteria abundance by fluorescence in situ hybridization”, from Chapter 4. FISH probe and additional information on the AOB and NOB bacteria studied in this research can be found in Table 4-2. From Chapter 4, see the section entitled, “Evaluat ion of nitrifying bacteria abundance by fluorescence in situ hybridization”, for a complete review of this molecular biological technique. Also se e Table 4-2 for FISH probe and additional information on the AOB and NOB bacteria studied in this research. The digital images from the FISH analysis were analyzed using the software daime (digital image analysis in microbial ecology) [86]. Each biomass was initially stained with DAPI, a blue-fluorescent nucleic acid stain that preferentially

PAGE 136

120 stains dsDNA but will also bind to RNA though it is not as strongly fluorescent [87]. Next, the sample was hybridized wi th the specific Cy3 probe which targets a specific sequence DNA presence associat ed with the microbe of interest. The Cy3 probe is a reactive water-soluble fluo rescent dye of the cyanine dye family. The Cy3 will appear as a r ed fluorescent color when bonded with the appropriate sequence. After hybridization, its abund ance was compared to the total biomass contained within the microbi al image. Digital image s of DAPI, Cy3 and their merged images were conducted for each samp le (Figure 6-1). Blue fluorescence can be seen in the mer ged image (image C) indicati ng areas where the bacteria of interest in not present. (The lengt h measure shown in each image represents 10m). A two dimensional automatic se gmentation with custom thresholding was used to determine these concentrations. Items appearing smaller than 10 pixels were ignored. A total bio-volu me fraction was calculated based on these image concentrations. These ar e reported as perc ent (percent of total biomass). Ten observations were measured for eac h combination and are reported as percent (percent of total bi omass). Additional statisti cs were also generated. Statistical and graphic a ssessment was conducted us ing Minitab statistical software (State College, PA.).

PAGE 137

F M 6 D s t h t o g N c F igure 6-1: M erged Im a .3 Res u D igital mer g ludge sam p h e initial a b o samples o rowth at n e N OB and d e ompares t h A Typical D a ge of DA P u lts g ed image s p le (Figure b undance o o btained a f e ar optimu m e picts the h em to sa m C D igital Ima g P I and Cy3 s for eac h 6-2 and 6 o f Nitrosom o f ter experi m m conditio n initial abu n m ples obtai n 121 g es of (A) Stain h microbe 3). Figure o na s and N m entation. T n s is clearl y n dance of N n ed after e x B DAPI Staiwere gen e 6-2 repre s N itrosospir a The abund y evident. N itrobacter x perimenta t B n, (B) Cy3 e rated fro m s ents the A a and then ance of th e Figure 63 and Nitro s t ion. A s wi Stain, an d m the acti v OB and d e compares t e microbes 3 represent s s pirae and th the AO B d (C) v ated e picts t hem after s the then B the

PAGE 138

a e F A S p a bundance vident. Th e F igure 6-2: A bundanc e S ample, (C p H = 6.5, a nd pH = 7. A C of the mi c e measure m AOB Ba e (A) Nitro s ) Nitroso m a nd (D) Ni t 5 c robes afte r m ent bar i n cteria, Re p s omonas m onas at D t rosospira 122 r growth a t n each digit a p resentati v – WWTP S D issolved C at Dissol v B D t near opti a l image r e v e FISH R S ample, ( B CO2 Conc v ed CO2 C mum cond e presents 1 R esults Sh o B ) Nitroso s entration = C oncentra t itions is cl 0 m. o wing Pe r s pira – W W = 5.0 mg/l t ion = 5.0 early r cent W TP and mg/l

PAGE 139

F A S = = D a d o F igure 6-3: A bundanc e S ample, (C ) = 7.0, and ( = 7.0 D escriptive nd NOB ( T eviation, r a f the spe c C A NOB Ba e (A) Nitr o ) Nitrobac t D) Nitros p statistics f o T ables 6-6 a a nge and t h c ified bact e cteria, Re p o bacter – t er at Diss p irae at Di s o r each mi c a nd 6-7). T h e number e ria are pr e 123 p resentati v WWTP S olved CO2 s solved C O c robe are p T he averag e of digital i m e sented in D B v e FISH R S ample, ( B 2 Concentr O 2 Concen t p rovided: A e microbe c m ages dep these ta b R esults Sh o B ) Nitros p r ation = 10 t ration = 5 A OB (Tabl e c oncentrati icting a hi g b les. The o wing Pe r p irae – W W .0 mg/l an d .0 mg/l an d e s 6-1 and on, its sta n g h accumul results fo r r cent W TP d pH d pH 6-2) n dard ation r the

PAGE 140

124 activated sludge sample obtained from t he WWTP are presented in these tables as well and are listed first. The high accumulation ratio represents t he number of slides out of a total of 10 that exhibit this phenomenon. Microbe col onies exhibiting a size greater than 10um are considered high accumulation. Digital images depicting this phenomenon are presented with digital images of the seed material obtained from the wastewater treatm ent facility (Figures 6-2 and 6-3). These results are not unexpected as ideal growth conditions, based upon previous designed experiments, could account for these high microbe accumulations. Analysis of the dissolved CO2 concentration/ pH combinations were evaluated by bacteria type (AOB and NOB) as a r andomized block design with a factorial arrangement. Each bacteria ty pe will be presented separately. 6.3.1 AOB Results An initial analysis was conducted to determine if differences in the AOB abundance exist between the bacteria obtai ned for use as the seed material. Operating conditions at the wastewater tr eatment plant from which this sample was received exhibited a dissolved CO2 concentration of 34 mg /l and pH of 6.86. A two sample t-test indicated a statistica lly significant difference exists for the

PAGE 141

125 AOB bacteria (p-value = 0.036) and shows Nitrosospira to predominate (Figure 6-4). Nitrosospira Nitrosomonas 5 4 3 2 1 Percent Abundance Figure 6-4: Percent Abundance of AO B Bacteria from Activated Sludge

PAGE 142

126 Descriptive statistics for the experimen tal results of the AOB bacteria are presented in Tables 6-1 and 6-2. Table 6-1: Nitrosomonas Percent Abundance Results Dissolved CO2Conc pH max Average Microbe Conc Standard Deviation Range Coefficient of Variation Images w/ High Accum 34 6.86 2.2 0.74 2.1 0.34 0/10 5 6.5 0.55 15.6 5.8 20.5 0.37 7/10 5 7.0 19.7 15.72 38.9 0.80 7/10 5 7.5 0.56 17.8 9.7 32.1 0.54 10/10 5 8.0 0.78 54.1 31.2 84.1 0.58 8/10 10 6.5 0.58 18.3 6.3 18.3 0.34 10/10 10 7.0 0.56 22.4 10.2 31 0.46 9/10 10 7.5 0.62 15.5 4.7 16.5 0.30 10/10 10 8.0 1.15 38.9 23.4 57.5 0.60 10/10 15 6.5 0.53 15.1 6.5 15.8 0.43 4/10 15 7.0 0.63 17.5 4.6 16.6 0.26 7/10 15 7.5 0.7 14.4 7.1 20 0.49 2/10 15 8.0 1.04 18.1 17.4 59.9 0.96 7/10 Table 6-2: Nitrosospira Percent Abundance Results Dissolved CO2Conc pH max Average Microbe Conc Standard Deviation Range Coefficient of Variation Images w/ High Accum 34 6.86 3.0 0.91 2.7 0.30 0/10 5 6.5 0.55 11.7 5.5 14.6 0.47 1/10 5 7.0 16.7 11.8 41.5 0.71 3/10 5 7.5 0.56 21.1 11.2 36.3 0.53 6/10 5 8.0 0.78 4.8 2.1 5.5 0.44 1/10 10 6.5 0.58 16.4 5.0 15.7 0.30 1/10 10 7.0 0.56 13.9 3.3 9.5 0.24 2/10 10 7.5 0.62 23.2 13.5 37.7 0.58 7/10 10 8.0 1.15 7.4 4.1 10.4 0.55 3/10 15 6.5 0.53 20.8 6.5 20 0.31 4/10 15 7.0 0.63 29.5 14.1 43.2 0.48 9/10 15 7.5 0.7 18.5 6.1 18 0.33 3/10 15 8.0 1.04 6.9 2.6 8.3 0.38 0/10

PAGE 143

127 Analysis of the AOB bacteria type showed the blocking variable, AOB Type, to be significant with Nitrosomonas being predominant (Table 6-3). The Nitrosomonas had a percent abundance across all treatment combinations of 22.3 percent compared to 15.9 percent for the Nitrosospira The main effects, dissolved CO2 concentration and pH, were not significant but their interaction effect is significant. As a significant interaction is present, analysis of the main effects is not appropriate. Due to the complexity of the interaction, treatment effects will be evaluated for each AOB type microbe separately. Table 6-3: ANOVA of Percen t Abundance of AOB Bacteria Source DF SS MS F P-Value AOB Type 1 2453.8 2453.8 11.9 0.001 CO2 2 285.2 142.6 0.69 0.502 pH 3 944.8 314.9 1.53 0.208 CO2*pH 6 3368.1 561.3 2.72 0.014 Error 227 46820.1 206.3 Total 239 53871.9 S = 14.3616 R-Sq = 13.09% R-Sq (adj) = 8.5% 6.3.1.1 Nitrosomonas Analysis of the Nitrosomonas bacteria shows all tr eatment effects to be statistically significant (Table 6-4). As t he interaction is significant, analysis of the main effects is not appropriate. A graphica l display and review of the interaction is provided (Figure 6-5).

PAGE 144

128 Table 6-4: ANOVA of Percent Abundance of Nitrosomonas Bacteria Source DF SS MS F P-Value CO2 2 2343.6 1171.8 5.68 0.005 pH 3 9021.2 3007.1 14.57 0.000 CO2*pH 6 4429.8 738.3 3.58 0.003 Error 108 22286.2 206.4 Total 119 38080.8 S = 14.3650 R-Sq = 41.48% R-Sq (adj) = 35.52% 8.0 7.5 7.0 6.5 50 40 30 20 10 15 10 5 50 40 30 20 10 5 10 15 6.5 7.0 7.5 8.0 Figure 6-5: Interaction Effect of Nitrosomonas Bacteria Results indicate that a higher percent abund ance exists at a pH of 8.0 at the 5 and 10 mg/l dissolved CO2 concentrations. This is not unexpected as Nitrosomonas has a preferred optimum pH range of 7.9 to 8.2. However, all appear to converge to a lower percent abundance as dissolved CO2 concentrations increase for all pH levels. CO2 pH Percent Abundance Percent Abundance

PAGE 145

129 6.3.1.2 Nitrosospira Analysis of the Nitrosospira bacteria shows all treatment effects to be significant (Table 6-5). As the interaction is signifi cant, analysis of the main effects is not appropriate. A graphical displa y and review of the intera ction is provided (Figure 6-6). Table 6-5: ANOVA of Percent Abundance of Nitrosospira Bacteria Source DF SS MS F P-Value CO2 2 602.42 301.21 4.42 0.014 pH 3 4012.61 1337. 54 19.62 0.000 CO2*pH 6 1360.44 226.74 3.33 0.005 Error 108 7361.93 68.17 Total 119 13337.41 S = 8.25627 R-Sq = 44.80% R-Sq (adj) = 39.18% 8.0 7.5 7.0 6.5 30 20 10 15 10 5 30 20 10 5 10 15 6.5 7.0 7.5 8.0Figure 6-6: Interaction Effect of Nitrosospira Bacteria CO2 pH Percent Abundance Percent Abundance

PAGE 146

130 Percent abundance results for the Nitrosospira are mixed. For two of the pH levels, 7.5 and 8.0, the per cent abundances are relative ly equal with the other two pH levels showing trends that increase with dissolved CO2 concentration. However, all three dissolved CO2 concentrations converge at a pH of 7.5 which may not be unexpected as the optimum pH fo r Nitrosospira is 7.5-8.0. At a pH of 8.0, they again converge to low percent abundances. Nitrosomonas also showed similar low percent abundances at a pH of 8.0 although its percent abundance was still greater than the Nitrosospira (18.1 versus 6.9). This may suggest another AOB may be present. 6.3.2 NOB Results An initial analysis was conducted to det ermine if differences exist between the NOB bacteria obtained for use as the seed material. Operating conditions at the wastewater treatment plant from which this sample was received showed a dissolved CO2 concentration of 34 mg/l and a pH of 6.86. A statistical difference of the abundance of the NOB was not evi dent (p-value = 0.152, Figure 6-7).

PAGE 147

131 Nitrospirae Nitrobacter 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Percent Abundance Figure 6-7: Percent Abundance of NO B Bacteria from Activated Sludge Descriptive statistics for the NOB experim ental results are presented (Tables 6-6 and 6-7).

PAGE 148

132 Table 6-6: Nitrobacter Percent Abundance Results Dissolved CO2Conc pH max Average Microbe Conc Standard Deviation Range Coefficient of Variation Images w/ High Accum 34 6.86 2.1 0.87 2.7 0.41 0/10 5 6.5 0.42 16.9 6.1 20.9 0.36 5/10 5 7.0 17.3 6.8 20.1 0.39 5/10 5 7.5 0.77 24.5 11.0 31.7 0.45 8/10 5 8.0 0.59 5.7 1.2 3.9 0.21 1/10 10 6.5 0.55 9.6 4.8 9.5 0.50 1/10 10 7.0 0.52 13.9 5.1 17.6 0.37 3/10 10 7.5 0.52 16.6 8.7 20.8 0.52 2/10 10 8.0 0.77 13.9 7.2 18.6 0.52 3/10 15 6.5 0.58 23.9 8.4 26 0.35 8/10 15 7.0 0.57 22.2 8.4 27.4 0.38 8/10 15 7.5 0.57 14.7 3.9 11.3 0.27 2/10 15 8.0 0.5 7.0 2.2 7.8 0.31 3/10 Table 6-7: Nitrospirae Percent Abundance Results Dissolved CO2Conc pH max Average Microbe Conc Standard Deviation Range Coefficient of Variation Images w/ High Accum 34 6.86 1.6 0.83 2.6 0.52 0/10 5 6.5 0.42 20.5 6.6 19.3 0.32 3/10 5 7.0 32.7 7.6 25.4 0.23 5/10 5 7.5 0.77 22.8 7.5 24.7 0.33 5/10 5 8.0 0.59 7.7 2.5 8.4 0.32 2/10 10 6.5 0.55 17.8 3.6 10.2 0.20 3/10 10 7.0 0.52 19.9 5.8 14.0 0.29 3/10 10 7.5 0.52 21.6 3.8 13.0 0.18 4/10 10 8.0 0.77 19.4 6.3 18.6 0.32 5/10 15 6.5 0.58 15.4 6.8 18.4 0.44 3/10 15 7.0 0.57 19.1 4.0 13.2 0.21 4/10 15 7.5 0.57 13.8 4.0 13.1 0.29 3/10 15 8.0 0.5 9.8 3.3 11.4 0.34 2/10 Results show the blocking vari able, NOB Type, as well as all treat ment effects to be significant with Nitrospirae being predominant (Table 6-8). The Nitrospirae had a percent abundance across all treatment combinations of 18.4 percent compared to 15.5 percent for the Nitrobacter As a significant interaction is present, analysis of the main effects is not appropriate. Due to the complexity of

PAGE 149

133 the interaction, treatment effects will be evaluated for each NOB type microbe separately. Table 6-8: ANOVA of Percen t Abundance of NOB Bacteria Source DF SS MS F P-Value NOB Type 1 491.06 491.06 11.56 0.001 CO2 2 319.44 159.72 3.76 0.025 pH 3 3610.68 203. 56 28.34 0.000 CO2*pH 6 2767.74 461.29 10.86 0.000 Error 227 9641.51 42.47 Total 239 16830.44 S = 6.51718 R-Sq = 42.71% R-Sq (adj) = 39.69% 6.3.2.1 Nitrobacter Analysis of the Nitrobacter bacteria shows all treatment e ffects to be statistically significant (Table 6-9). As the interact ion is significant, analysis of the main effects is not appropriate. A graphical display of the interaction is provided (Figure 6-8). Table 6-9: ANOVA of Percent Abundance of Nitrobacter Bacteria Source DF SS MS F P-Value CO2 2 259.04 129. 52 3.16 0.046 pH 3 1825.30 608.43 14.85 0.000 CO2*pH 6 2051.02 341. 84 8.35 0.000 Error 108 4423.92 40.96 Total 119 8559.28 S = 6.40017 R-Sq = 48.31% R-Sq (adj) = 43.05%

PAGE 150

134 8.0 7.5 7.0 6.5 25 20 15 10 5 15 10 5 25 20 15 10 5 5 10 15 6.5 7.0 7.5 8.0 Figure 6-8: Interaction Effect of Nitrobacter Bacteria Results are mixed depend ing on the dissolved CO2 concentration and the pH level. The general trend shows that as pH increases, the percent abundance of Nitrobacter decreases. Results indicate two species of Nitrobacter exist. This is observed by reviewing fi gure 6-10. Dissolved CO2 concentrations at 5 and 15 mg/l and at pH levels of 7.5 and 6.5, respectively, provide the greatest percent abundance. The high abundance level at a pH of 7.5 can be explained as this is the preferred pH range of th is microbe (Table 2-3). The pH level of 6.5 but with a higher dissolved CO2 concentration suggests t he interrelationship of CO2 with pH on microbe growth of another species of Nitrobacter CO2 pH Percent Abundance Percent Abundance

PAGE 151

135 6.3.2.2 Nitrospirae Analysis of the Nitrospirae bacteria shows all treatment effects to be statistically significant (Table 6-10). As the interact ion is significant, analysis of the main effects is not appropriate. A graphical di splay of the interaction is provided (Figure 6-9). Table 6-10: ANOVA of Percent Abundance of Nitrospirae Bacteria Source DF SS MS F P-Value CO2 2 915.27 457.64 15.57 0.000 pH 3 2059.90 686.63 23.36 0.000 CO2*pH 6 1630.17 271.70 9.24 0.000 Error 108 3174.75 29.40 Total 119 7780.10 S = 5.42179 R-Sq = 59.19% R-Sq (adj) = 55.04% 8.0 7.5 7.0 6.5 30 25 20 15 10 15 10 5 30 25 20 15 10 5 10 15 6.5 7.0 7.5 8.0 Figure 6-9: Interaction Effect of Nitrospirae Bacteria Percent Abundance Percent Abundance CO2 pH

PAGE 152

136 Results for the Nitrospirae microbes are mixed. T he microbial concentration appears to increase and then decr ease as pH goes from 6. 5 to 8.0. The percent abundance at specific dissolved CO2 concentration is dependent on pH but at 10 mg/l their concentrations appea r similar. The greatest Nitrospirae abundance occurs at a pH of 7.0 and a dissolved CO2 concentration of 5 mg/l. This is not in agreement with published data for this micr obe showing an optimum pH of 8.0 8.3 (Table 2.3). This elevated per cent abundance at this dissolved CO2 concentration/pH combinatio n may suggest a species of Nitrospirae not previously identified. 6.3.3 Validation Study of FISH Results As many of the result s of the dissolved CO2 concentration/ pH combinations exhibited high percent abundances, a study was conducted to determine if the digital image results reflected high micr obial concentrations ( percent abundance). Some concern existed in the digital imag e areas identified as high accumulation that these could be phosphate crystals or perhaps some other contaminant fluorescing and giving false readings. Si x slides were prepared that included representative AOB and NO B that exhibited high abundance (Table 6-11). E. coli was used as a negative control and Bacillus subtilis was used as a positive control.

PAGE 153

137 Table 6-11: Slide Preparation for Validation Study Slide 1 2 3 4 5 6 Comment Samples CO2 = 5 mg/l pH = 8.0 X X X High levels of Nitrosomonas CO2 = 15 mg/l pH = 7.0 X X X High levels of Nitrosospira CO2 = 5 mg/l pH = 7.5 X X X High levels of Nitrobacter CO2 = 5 mg/l pH = 7.0 X X X High levels of Nitrospirae SCBWWTP X X X X X X Seed material E. coli X X X X X X Negative control Bacillus subtilis X Positive control Fish Probes No probe X Hybridization Buffer NSM156 X Nitrosomonas probe Nsv433 X Nitrosospira probe NIT3 X Nitrobacter probe Ntspa0712 X Nitrospirae probe LGC353b [88] X Bacillus probe Evaluation of the digital images (40X obj ective) for each slide revealed that the high percent abundances do re flect high AOB or NOB (F igure 6-10 to 6-15). Except for the Bacillus subtilis which fluoresces as it is the positive control, only the AOB and NOB nitrifying probes display fluorescence for the nitrifying bacteria.

PAGE 154

F N e N w T w e r a w i m l e i m T F igure 6-1 0 N itrobacte r qual 10 m N o auto flu o w ith the hyb T he digital i w ith the re s valuated t o a tio was c a w as used t o m age (blu e e gends. F o m age has b T he high a b A D 0 : Slide 1 r (D) Nitr o m o rescence i s ridization b mages in f s ults from o determin e a lculated, t e o adjust th e e ) was adj o r the bior e b een mag n b undance o No Probe o spirae (E ) s observed b uffe r f igures 6-1 the daim e e their brig e rmed mul t e images o usted and e actor and n ified to sh o f cells for t B E 138 with (A) N ) Seed M a in Figure 6 1 through 6 e analysis. htness by t iplicative b o n each sli d its multip l seed mat e ow the cel t he bioreac N itrosomo n a terial, an d 6 -10 indica t 6-14 were The DA P evaluating b rightness a d e. On ea c l icative eff e e rial merg e l abundan c c tor images C F n as (B) N i d (F) E. c o t ing no flu o adjusted t o P I and Cy 3 the pixel v a djustment ch set of s e ct is liste e d images, c e for eac h (Image A C F i trosospir a o l i Scale o rescence e o be consi 3 images w v alue rang e factor, an d s lides, the D d in the f i a region o h AOB or N of Figures a (C) bars e ffect stent w ere e A d this D API i gure o f the N OB. 6-11

PAGE 155

139 through 6-14) is evident with reduced cell abundance for the seed material image (Image B of Figures 6-11 th rough 6-14). The same expo sure times were used on all images to maintain consistency. Figure 6-11: Slide 2 NSM156 Probe with (A) Nitrosomonas (B) Seed Material, and (C) E. coli using a Multiplicative Factor of 2.6. Scale bars equal 10 m Some non-specific binding fluorescence is observed from the images (A and B) but this is not unexpected as FISH pr obes do bind to extracellular polymeric substances, which gives low level fluorescenc e in parts of the flocs. The small, discrete objects are cells and appear to be pear shaped indicative of the Nitrosomonas morphology (Table 2-3). E. coli is a negative control and does not exhibit any fluorescence due to this FISH probe. A magnified section can be observed in image C. Micro-colonies of ce lls are observed in higher levels in the bioreactor compared to the seed material. Furthermore, the micr o-colonies in the bioreactors have higher levels of Nitrosomonas and are brighter, which indicates higher ribosome content and therefore, a higher specific growth rate [89]. A B C

PAGE 156

140 Figure 6-12: Slide 3 Nsv433 Probe with (A) Nitrosospira (B) Seed Material, and (C) E. coli using a Multiplicative Factor of 2.3. Scale bars equal 10 m Some non-specific binding fluorescence is observed from the images (A and B) but this is not unexpected as FISH pr obes do bind to extracellular polymeric substances, which gives low level fluorescenc e in parts of the fl ocs. The small, discrete objects observed in these images should hav e a spiral appearance indicative of Nitrosospira morphology but are too small at this magnification to be positively identified (Table 2-3). E. coli is a negative control and does not exhibit any fluorescence due to this FISH probe. Micro-colonies of cells are observed in higher levels in the bioreac tor compared to the seed ma terial. Furthermore, the micro-colonies in the bioreac tors have higher levels of Nitrosospira and are brighter, which indicates higher ribosome content and therefore, a higher specific growth rate [89]. A B C

PAGE 157

141 Figure 6-13: Slide 4 NIT3 Probe with (A) Nitrobacter (B) Seed Material, and (C) E. coli using a Multiplicative Factor of 1.6. Scale bars equal 10 m Some non-specific binding fluorescence is observed from the images (A and B) but this is not unexpected as FISH pr obes do bind to extracellular polymeric substances, which gives low level fluorescenc e in parts of the fl ocs. The small, discrete objects observed in these images should have a rod shaped appearance indicative of Nitrobacter morphology but are too small at this magnification to be positively identified (Table 2-3). E. coli is a negative control and does not exhibit any fluorescence due to this FISH probe. Micro-colonies of cells are observed in higher levels in the bioreac tor compared to the seed ma terial. Furthermore, the micro-colonies in the bioreactors have higher levels of Nitrobacter and are brighter, which indicates higher ribosome content and therefore, a higher specific growth rate [89]. A B C

PAGE 158

142 Figure 6-14: Slide 5 Ntspa0712 Probe with (A) Nitrospirae (B) Seed Material, and (C) E. coli using a Multiplicative Fact or of 1.9. Scale bars equal 10 m Some non-specific binding fluorescence is observed from the images (A and B) but this is not unexpected as FISH pr obes do bind to extracellular polymeric substances, which gives low level fluorescenc e in parts of the fl ocs. The small, discrete objects observed in these im ages should have a long slender rod appearance indicative of Nitrospirae morphology but are too small at this magnification to be positivel y identified (Table 2-3). E. coli is a negative control and does not exhibit any fluorescence due to this FISH probe. Micro-colonies of cells are observed in higher levels in the bioreactor compared to the seed material. Furthermore, the mi cro-colonies in the bioreactors have higher levels of Nitrospirae and are brighter, which indicates higher ribosome content and therefore, a higher specif ic growth rate [89]. A B C

PAGE 159

F N a M e n d F igure 6-1 N itrosospi r nd (G) Ba c M icro-colon i xpected a s ot show a n uring the d A D G 5: Slid e r a (C) Nit r c illus subt i i es of cells s this micro b n y fluoresc e aime anal y e 6 LGC 3 r obacter ( D i l i s. Scale are observ b e was us e e nce indica t y sis do repr e B E 143 3 53b Pro D ) Nitrosp bars equ a ed in the B e d as a pos t ing that th e e sent high be with p irae (E) S a l 10 m acillus sub t s itive contr o e observed accumulat i (A) Nitro S eed Mate r t illis image o l. The oth e high abun d i ons of AO B C F somonas r ial, (F) E. This resu e r images d d ance rep o B and NO B (B) coli, lt is d o o rted B

PAGE 160

D i n p F O 6 A c D uring this s n dividual c rovided as F igure 6-1 6 O bjective) .3.4 Bio m A model w onclusion o s tudy, an i m ells and h an exampl 6 : Nitros p m ass Grow t as develo p o f a react o m age of Ni t h igh accu m e of nitrifie r p irae Sho w t h Determ i p ed to as c o r study. T 144 t rospirae w m ulation (F r s under c o w n with H i nation c ertain the T his was d as taken s h igure 6-1 6 o nditions hi g H igh A ccu m abundan c d one as a h owing hig 6 ). This d g hly condu c mulation o c e of the check to e h abundan c d igital ima g c ive to gro w o f Cells ( 1 nitrifiers a t e nsure tha c e of g e is w th. 1 00X t the t the

PAGE 161

145 reported percent abundance could be achi eved. Two simultaneous equations, the substrate utilization rate [ ] and the biomass growth rate [ ], were evaluated using an it erative approach. Both heterotrophic and autotrophic bacteria were evaluated using MATLAB (Natick, MA) and a ratio of their biomass concent rations was calculated. Growth was based on Monod kinetics and standard kinetic coefficients were utilized [2, 12, 63]. After 10 days of reaction, the aut otrophs are approximately 85% of the biomass (Figure 6-17). The biomass conc entrations, depicted on the left vertical axis, show the microbe growth and decay ov er the 10 day period. The percent of autotrophic biomass to total biomass is shown on the right vertical axis. Figure 6-17: Autotrophic Biomass as Pe rcent of Total Biomass to Confirm High Percent Abundance Measurements 0 10 20 30 40 50 60 70 80 90 100 0 20 40 60 80 100 120 140 0246810Percent of Total Biomass Concentration, mg/lTime, days Autotrophic Biomass Heterotrophic Biomass Autotrophic Biomass as Percent of Total

PAGE 162

146 6.4 Discussion Before reviewing and commenting on the resu lts of this study, a proper microbial assessment could not have been undertak en without knowledge and expertise on the use of the daime software. Appropriate time was spent (six hours) in developing the expertise to ensure pr oper segmentation was conducted. Interpreting the digital images prior to segmentation can be considered “artsy” and therefore care was taken to ensure consistent results were attained. Statistical model assumptions were gene rally satisfied but some departures in normality of error terms and differences in dissolved CO2 concentrations/ pH combination variances were observed. Th is is not unexpected given the range of microbe percentages seen during the evaluati on. As these deviations were not severe and statistical models are r obust, data transformations were not undertaken. In addition, outliers were not removed as t hey added to model understanding. The activated seed sludge was obtained from the discharge side of the aeration basin of a modified Ludzack-Ettinger wast ewater facility. Process measurements reported in this study, 34 mg/l dissolved CO2 concentration and a pH of 6.86, remained constant from samples obtained over several years from this plant location. The four microbes evaluated in this study we re all identified in this activated sludge. Nitrosospira was the dominate AOB and statistically greater

PAGE 163

147 than Nitrosomonas while the NOB microbes, Nitrobacter and Nitrospirae exhibited similar abundances. These abundanc es are invariably due to influent and plant operating conditions. Analysis of the dissolved CO2 concentration/ pH combinations does present challenges. Those combinations with the higher microbe abu ndances could be said to predominate and thus be the preferred set of operating conditions. This interpretation is too simple Interpretation of the results requires that the maximum specific growth rate as well as optimum growth conditions suggested from literature be used in c onjunction with the dissolved CO2 concentration/ pH combinations. Achieving a high abundance of a particular microbe at a high growth rate may not coincide. In fact, this occurred during this study. The AOB microbes exhibit thei r greatest maximum specific growth at a pH of 8.0 for each dissolved CO2 concentration (figure 5-6). This is in agreement with literature for optimum pH growth conditions for Nitrosomonas (7.9 – 8.2). Nitrosospira exhibited its greatest microbe perc entage at a pH of approximately 7.0 and a dissolved CO2 concentration of 15 mg/l. Other high abundance occurred at a pH of 7.5 fo r the 5 and 10mg/l dissolved CO2 concentration. This may suggest a shift in ideal growth conditions as dissolved CO2 concentrations change or the optimum growth conditions may exist in the pH range of 7.0 to 7.5. Literature provides an optimum pH range for Nitrosospira of 7.5 to 8.0 but cited references are not available.

PAGE 164

148 Evaluation of the seed material showed the predominance of the Nitrosospira microbe. However, at a pH of 8.0 and across all dissolved CO2 concentrations, Nitrosomonas was dominant. Based on ratios when compared to Nitrosospira this was 11.3, 5.3 and 2.6 times greater at 5, 10 and 15 mg/l, respectively. Nitrosospira was dominant at a pH of 7.5; and when compared to Nitrosomonas, averaged approximately 23 pe rcent greater microbe abundanc e. However, this concentration difference does not compare to the predominance of Nitrosomonas at a pH of 8.0. Even t hough the highest abundance of Nitrosomonas was observed at a dissolved CO2 concentration of 5 mg/l, the highest maximum specific growth rate occurred at a dissolved CO2 concentration of 10 mg/l. Thus, suggesting the synergistic effect of dissolved CO2 concentration with pH on microbe growth and a combination of AOB at these conditions. Although the NOB microbes were found to be statis tically different, their numerical differences in percent abunda nce were not pronounced (18.4 versus 15.5 for Nitrospirae and Nitrobacter respectively). At a pH of 7.0 and a dissolved CO2 concentration of 5 mg/l, the Nitrospirae which has an optimum growth pH of 8.0 – 8.3, was approxim ately twice the perc ent abundance as the Nitrobacter Thus, suggesting that another Nitrospirae species may exist not previously identified. At a pH of 7.5, where Nitrobacter has an optimum growth pH of 7.2 – 7.6, the percent abundance of the Nitrobacter and Nitrospirae microbes were approximately equal. And at a pH of 8.0, where Nitrospirae has an optimum

PAGE 165

149 growth pH of 8.0 – 8.3, Nitrospirae exhibits approximately a forty percent greater abundance than Nitrobacter Although this result is to be expected, the percent abundance of the Nitrospirae at the pH 8.0 level across all dissolved CO2 concentrations was still lower than at other pH levels. Interestingly, some NOB results were expected but others did not match expected optimum pH based on microbial percent abundances. The highest maximum specific growth ra tes for the NOB microbes occurred at dissolved CO2 concentrations and pH values of 5 mg/l and a pH of 7.5; and 10 mg/l and a pH of 8.0, respectively (Figur e 5-8). Each shows a maximum specific growth rate of 0.77 with approximately th e same percent microbial abundance. This suggests that neither NOB microbe is predominant; with both contributing to conversion of nitr ite to nitrate. Figures 5-6 to 5-9 in chapter 5 show the effect of max at the dissolved CO2 concentration/ pH level combinations. It is evident that differences due exist but identifying specific microbes that are predominant at these dissolved CO2 concentration/ pH level combinations has not been successful in all cases. As the aeration basin of a WWTP has been said to be a complex microbial community probably containing several diffe rent genera of microbes capable of nitrification [23]; a combination of different AOB/ NOB appears a more likely scenario in assessing the relationship between dissolved CO2 concentration/ pH level combinations, microbe abundance and nitrification growth rates.

PAGE 166

150 Of particular interest is the effect of dissolved CO2 concentration on the abundance of the Nitrosomonas microbe. The percent abundance values are much greater than for the Nitrosospira, Nitrobacter and Nitrospirae microbes which exhibit equivalent abundances. This may suggest that these microbes are more sensitive to the e ffects of carbon dioxide. It has been shown that max is affected by dissolved CO2 concentration and pH but which effect, if any, dominates at t hese experimental conditions (Figures 5-6 to 5-9 in chapter 5). A partitioning of the treatm ent sums of squares was undertaken (Table 6-11). The AOB and NOB were compared to the designed experiment that optimized the dissolved CO2 concentration/ pH combination (experiment three from c hapter 5). Percent contri bution for the dissolved CO2 concentration, pH, and the dissolved CO2 concentration / pH interaction were calculated. These are compared to the ma in effects previously reported as well as the R2 generated from their ANOVA analysis (Figures 5-6, 6-4, and 6-10). Table 6-12: Partitioned Treatment Sums of Squares by Treatment Effect Microbe R2 CO2 pH CO2/pH Interaction Chapter 5 Experiment 3 92.5 17.4 82.6 NA Nitrosomonas 42.7 26.1 49.2 24.7 Nitrosospira 48.9 4.3 72.2 23.5 Nitrobacter 48.3 6.3 44.1 49.6 Nitrospirae 59.2 28.2 63.4 8.4 With one exception, pH dominated the ma ximum specific growth rate of the microbes. Only for the Nitrobacter microbe did the CO2*pH interaction show a

PAGE 167

151 higher percent contribution. This is due to increased variation seen across treatment combinations. Even so, it was still significantly greater than the percent contribution of the dissolved CO2 contribution. The validation study showed that t he high percent abundan ce measurements were genuine when using these experiment al reactor operating conditions. The initial concerns of phosphat e crystals generated during t he experiment or other material fluorescing were not warranted. Calculation of autot rophic biomass as a percent of total biomass confirmed that high levels of percent abundance measurements are possible when reactor o perating conditions select for AOB or NOB. 6.5 Conclusions The maximum specific growth rate of ni trifying bacteria is influenced by dissolved CO2 concentration and pH. Though each contributes to enhancing the nitrification rate, pH has a more pronounced influence. This is evidenced by larger F values for the pH effect fr om ANOVA source t ables, the percent influences of the treatment sums of s quares, and the largest nitrification rates occurring at pH values of 7.5 and 8.0, depending on the microbe. The AOB bacteria appear to have the most influence on nitrification rates with dissolved CO2 concentrations of 10 mg/l or 15 mg /l at a pH of 8.0 providing the

PAGE 168

152 highest nitrification rates. Based on lit erature and statistical analysis of this research, Nitrosomonas is the predominant AOB mi crobe at these dissolved CO2 concentrations and pH combinations al though its percent microbial abundance was not pronounced at 15 mg/l. As both Nitrosomonas and Nitrosospira had low percent microbial abundance at this dissolved CO2 concentration, a third unidentified AOB may be present. Optimi zing conditions for the growth of AOB microbes is necessary if maximum specif ic growth rates are to be realized. The NOB microbes are statistically different based on their percent concentrations across the dissolved CO2 concentration and pH combinations with Nitrospirae being dominant. However, at many combinations to include those conditions that provide the maximum specific growth rate, max = 0.77 (Figure 5-9), their concentrations are equivalent. Thus, suggesting that both NOB microbes, Nitrobacter and Nitrospirae, contribute to the nitrification rate. This study was based on dissolved CO2 concentrations ranging from 5 to 15 mg/l, pH levels from 6.5 to 8.0, and non-limiting substr ate (ammonium) and dissolved oxygen levels. These combin ations provided for optimum growth conditions based on many previously cond ucted experiments. As it has been suggested that activated slu dge is comprised of a divers e microbial ecology [23], similar results should be achieved using seed material from other wastewater treatment processes. However, whet her similar percent abundance results will be achieved using experimental conditions specific to WWTP’s other than a

PAGE 169

153 Modified Ludzack-Ettinger process, whic h this experiment is based upon, is unknown. Appropriate experimentat ion would need to be conducted to determine optimum dissolved CO2 concentration and pH levels for process condition typical of other treatment processes.

PAGE 170

154 Chapter 7 Conclusions The original hypothesis stat ed that nitrification was li mited due to reduced levels of carbon dioxide in the aerat ion basin of a wastewater treatment facility. This was based on the premise t hat the aeration basin is in equilibrium with the atmosphere. Subsequent field testing re vealed this assumption to be incorrect with elevated levels of ca rbon dioxide found throughout a wastewater treatment facility. Research focused on understanding the effects of carbon dioxide and pH on nitrification and determining if an optimum dissolved CO2 concentration/ pH combination exists that maximizes nitrific ation. Experimentatio n revealed that at low (< 5 mg/l) and high (> 30 mg/l) dissolved CO2 concentrations inhibition effects are apparent. Further research found a dissolved CO2 concentration of 10-15 mg/l and a pH of 8.0 to prov ide for optimum nitrification. Microbial studies were conducted on t he designed experiment that determined the optimum dissolved CO2 concentration/ pH combinat ions using the two most common AOB and NOB. Results were mixed depending on the dissolved CO2

PAGE 171

155 concentration/ pH combinati ons, but across all levels Nitrosomonas was the dominant AOB with Nitrospirae being the dominant NOB. Additionally, high abundance measurements for some dissolved CO2 concentration/pH combinations that were not at optimum pH sugges t that these genera have multiple members (i.e., species) with different growth sensitivities. Based on these results, future research should focus on the following items: Pilot Plant or Full-Scale Demonstration: Evaluatio n of the optimal dissolved CO2 concentration/pH on the rate of nitr ification at a WWTP will validate this research. Elevated pH Operating Protocol: Establishing an effective pH control methodology that adjusts and maintain s an appropriate pH is not without challenges. Treatment of the influent and sequential metering locations would need to be established. In addition, the effluent pH may need adjustment to a lower pH in order to comply with an existing permit. As there are many different WWTP configurations, a custom ized approach for each facility would probably be necessary. This should be conducted prior to a pilot plant or full-scale demonstration. Treatment of High Ammonium Levels in Anaerobic Digester Supernatant: Several WWTPs that operate anaerobic digesters were found to contain

PAGE 172

156 ammonium levels from 6001,000 mg/l and had dissolved CO2 concentrations of 100 mg/l. The treated solids of these digesters were disposed by application onto agricultural land, remova l of the solids for disposal in a sanitary landfill or drying of the solids fo r subsequent sale as a fertilizer. In processes where the solids are removed le aving a liquid supernatant high in ammonia and dissolved CO2 concentration, the supernatant is returned to the head works of the WWTP for treatm ent. For WWTP’s treating their supernatant, this liquid is mixed with the influent at approximately 15 percent of the influent flow rate. A strategy to treat the supernatant would benefit the WWTP by removing this nitrogen sour ce and improve their treatment capabilities. Low F/M Experimentation: Research in this study focused on high F/M in experimentation. Evaluati on of the low F/M should pr oduce similar results as found in this study, but confirmation is needed. The current regulations for nitrate concent rations in drinking water have been set to 10 mg/l in the USA, Japan and Korea an d 11.3 mg/l for the European Union. Levels may be permitted lower at the state or local level [90-92]. With the recent proposals by the Environment al Protection agency (EPA) to establish nutrient criteria for the State of Florida which in many ca ses are much lower than currently permitted, this research coul d prove very beneficial in meeting these proposed standards [93].

PAGE 173

157 References 1. EPA, Primer for Municipal Wast ewater Treatment Systems in EPA 832-R04-001 2004. 2. Metcalf & Eddy, Wastewater Engineering: Treatment and Reuse. 4 ed. McGraw-Hill series in civil and env ironmental engineering. 2003, New York, NY: McGraw-Hill. 1819. 3. Jeyanayagam, S., True confessions of the biological nutrient removal process. Florida Water Resources Journal: January, 2005. 4. Ahmed M., Life cycle analysis in Wastewater : a sustainability perspective. Waste Water Treatment and Reuse in the Mediterranean Region: The Handbook of Environmental Ch emistry, 2011. 14: p. 125-154. 5. Deslauriers, S., et al., CSS Factsheets, US Wastewater Treatment System. University of Michigan, Ann Arbor, 2004. 1-2(CSS): p. 4-14. 6. Carns K., Bringing energy efficiency to t he water and wastewater industry: how do we get there? WEFTEC Proceedings, 2005. 7. Ningwei, Z., et. al., Effectiveness of pH as a cont rol parameter for nitrogen removal via SND in a se quencing batch reactor. Bioinformatics and Biomedical Engineering ICBBE 2009. 3: p. 1-5. 8. EPA, Biological Nutrient Remo val Processes and Costs in EPA 823-R-07002 2007. 9. Grady, C., G. Daigger, and H. Lim, Biological Wastewater Treatment 1999: CRC Press. 10. WEF, Biological Nutrient Removal (B NR) Operation in Wastewater Treatment Plants 2006: McGraw-Hill: New York.

PAGE 174

158 11. Johnson, B.R., Daigger, G.T., Crawford, G.V., Method of Treating Wastewater in Patent 2007. 12. Rittmann, B.E., and P.L. McCarty, Environmental Biotechnology: Principles and Applicarions 2001: McGraw Hill New York. 754. 13. Grady, C.P.L., Jr., Aichinger, G., Cooper, S.F., & Naziruddin, M., Biodegradation kinetics for selected toxic/hazardous organic compounds. Hazardous Waste Treatment: Biosystems for Pollution Control Pittsburgh, Pennsylvania: Air and Waste Managemen t Association, 1989: p. 141153. 14. Mobarry, B.K., et al., Phylogenetic probes for analyzing abundance and spatial organization of nitrifying bacteria. Appl Environ Microbiol, 1996. 62(6): p. 2156-62. 15. Wagner, M., et al., In situ analysis of nitr ifying bacteria in sewage treatment plants. Water Science and Technology, 1996. 34(1-2): p. 237244. 16. Konneke, M., et al., Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature, 2005. 437(7058): p. 543-546. 17. Kampschreur, M., et al., Nitrous oxide emissi on during wastewater treatment. Water Research, 2009. 43: p. 4093-4103. 18. Park, H.D., et al., Occurrence of ammoniaoxidizing Archaea in wastewater treatment plant bioreactors. Applied and Environmental Microbiology, 2006. 72(8): p. 5643-5647. 19. Schramm, A., et al., Identification and acti vities in situ of Nitrosospira and Nitrospira spp. as dominant populati ons in a nitrifying fluidized bed reactor. Applied and Environmental Micr obiology, 1998. 64(9): p. 3480. 20. Juretschko, S., et al., Combined molecular and conventional analyses of nitrifying bacterium diversity in ac tivated sludge: Nitrosococcus mobilis and Nitrospira-like bacteria as dominant populations. Applied and Environmental Microbiology 1998. 64(8): p. 3042.

PAGE 175

159 21. Melcer, H., et al., Methods for Wastewater Char acterization in Activated Sludge Modeling 2003: Water Environment Research Foundation and IWA Publishing. 22. Blackburne, R., et al., Kinetic characterisation of an enriched Nitrospira culture with comparison to Nitrobacter. Water Research, 2007. 41(14): p. 3033-3042. 23. Siripong, S. and B. Rittmann, Diversity study of nitrifying bacteria in fullscale municipal wastewat er treatment plants. Water Research, 2007. 41(5): p. 1110-1120. 24. Daims, H., et al., In situ characterization of Nitrospira-like nitrite-oxidizing bacteria active in wastewater treatment plants. Applied and Environmental Microbiology, 2001. 67(11): p. 5273. 25. Ehrich S., B.D., Lebedeva E., Ludwig W., Bock E., A new obligatory chemolithoautotrophic, nitr ite-oxidizing bacterium. Nitrospira moscoviensis sp. nov. and its phylogenetic relatioship. Arch Microbial 19 95. 164: p. 1623. 26. Alleman J.E., Elevated nitrite occurrence in biological wastewater treatment systems. Water Science Technology, 1984. 17: p. 409-419. 27. Sinha, B. and A. Annachhatre, Partial nitrification operational parameters and microorganisms involved. Environmental Science Biotechnology, 2007. 6: p. 285-313. 28. Kumar, V., et al., Mass production of nitrifying ba cterial consortia for the rapid establishment of ni trification in saline re circulating aquaculture systems. Microbiol Biotechnology 2009. 25: p. 407-414. 29. Villaverde, S., et. al., Influence of pH over nitrif ying biofilm activity in submerged biofilters. Water Research, 1997. 31: p. 1180-1186. 30. Vaccari, D., et al., Environmental biology fo r engineers and scientists 2006: Wiley-Interscience.

PAGE 176

160 31. Koops, H., H. Harms, and H. Wehrmann, Isolation of a moderate halophilic ammonia-oxidiz ing bacterium, Nitrosoc occus mobilis nov. sp. Archives of Microbiology, 1976. 107(3): p. 277-282. 32. Wagner, M., et al., Combining fluorescent in situ hybridization (FISH) with cultivation and mathematical modeli ng to study population structure and function of ammonia-oxidizing bacteria in activated sludge. Water Science and Technology, 1998. 37(4): p. 441-450. 33. Schmidt, I., et al., New concepts of microbial tr eatment processes for the nitrogen removal in wastewater. FEMS Microbiology Reviews, 2003. 27(4): p. 481-492. 34. Koops, H.P., and A. Pommerening-Roser, Distribution and ecophysiology of the nitrifying bacteria emphasing cultured species. FEMS Microbilogical Ecology, 2001. 37: p. 1-9. 35. Biesterfeld, S., et al., Quantification of nitrifying bacterial populations in a full-scale nitrifying trickling filter us ing fluorescent in situ hybridization. Water Environment Research, 2001: p. 329-338. 36. Daims, H., et al., Novel NitrospiraNitrospira-like bacteria as dominant nitrite-oxidizers in biofilms from wast ewater treatment plants: diversity and in situin situ physiology. Water Sci. Technol, 2000. 41: p. 85-90. 37. Dionisi, H., et al., Quantification of Nitro somonas oligotropha-like ammonia-oxidizing ba cteria and Nitrospira spp. fr om full-scale wastewater treatment plants by competitive PCR. Applied and Environmental Microbiology, 2002. 68(1): p. 245. 38. Hiorns, W., et al., Amplification of 16S ribosomal RNA genes of autotrophic ammonia-oxidizing bacteri a demonstrates the ubiquity of nitrosospiras in the environment. Microbiology, 1995. 141(11): p. 27932800. 39. Kowalchuk, G., et al., Analysis of ammonia-oxidiz ing bacteria of the beta subdivision of the class Proteobac teria in coastal sand dunes by denaturing gradient gel el ectrophoresis and sequenc ing of PCR-amplified 16S ribosomal DNA fragments. Applied and Environmental Microbiology, 1997. 63(4): p. 1489.

PAGE 177

161 40. Stephen, J., et al., Molecular diversity of soil and marine I6S rDNA sequences related to b subgr oup ammonia-oxidizing bacteria. Appi Environ. Microbiol., 62, 4147, 1996. 4154. 41. Green, M., et al., High nitrification rate at low pH in a fluidi zed bed reactor with either chalk or sintered glass as the biofilm carrier. Israel Journal of Chemistry, 2006. 46(1): p. 53-58. 42. Grunditz, C. and G. Dalhammar, Development of nitrification inhibition assays using pure cultures of Nitrosomonas and Nitrobacter. Water Research, 2001. 35(2): p. 433-440. 43. Antoniou, P., et al., Effect of temperature and pH on the effective maximum specific growth ra te of nitrifying bacteria. Water Research, 1990. 24(1): p. 97-101. 44. Tarre, S. and M. Green, High-rate nitrification at low pH in suspended-and attached-biomass reactors. Applied and Environmental Microbiology, 2004. 70(11): p. 6481. 45. EPA, Process design manual for nitrogen control in ED162870 E.P. Agency, Editor. 1975: Washington, DC. 46. Geets, J., N. Boon, and W. Verstraete, Strategies of aerobic ammoniaoxidizing bacteria for coping with nutrient and oxygen fluctuations. FEMS microbiology ecology, 2006. 58(1): p. 1-13. 47. Ritchie, G. and D. Nicholas, Identification of the s ources of nitrous oxide produced by oxidative and reducti ve processes in Nitrosomonas europaea. Biochemical Journal, 1972. 126(5): p. 1181. 48. Abeliovich, A. and A. Vonshak, Anaerobic metabolism of Nitrosomonas europaea. Archives of Microbiology, 1992. 158(4): p. 267-270. 49. Stven, R., M. Vollmer, and E. Bock, The impact of organic matter on nitric oxide formation by Nitrosomonas europaea. Archives of Microbiology, 1992. 158(6): p. 439-443.

PAGE 178

162 50. Bock, E., et al., Nitrogen loss caused by denitrif ying Nitrosomonas cells using ammonium or hydrogen as elec tron donors and nitrite as electron acceptor. Archives of Microbiology, 1995. 163(1): p. 16-20. 51. Poth, M. and D. Focht, N Kinetic Analysis of N (2) O Production by Nitrosomonas europaea: an Examinatio n of Nitrifier Denitrification. Applied and Environmental Microbiol ogy, 1985. 49(5): p. 1134. 52. Stein, L. and D. Arp, Loss of ammonia monooxygenase activity in Nitrosomonas europaea upon exposure to nitrite. Applied and Environmental Microbiology 1998. 64(10): p. 4098. 53. Schmidt, I., R. van Spanning, and M. Jetten, Denitrification and ammonia oxidation by Nitrosomonas europaea wild-type, and NirK-and NorBdeficient mutants. Microbiology, 2004. 150(12): p. 4107. 54. Schmidt, I. and P. Steenbakkers, Physiologic and proteomic evidence for a role of nitric oxide in biofilm formation by Nitrosomonas europaea and other ammonia oxidizers. Journal of bacteriology, 2004. 186(9): p. 2781. 55. Philips, S., et al., Oxygen-limited autotrophic nitr ification/denitrification by ammonia oxidisers enables upward motion towards more favourable conditions. Applied microbiology and biot echnology, 2002. 59(4): p. 557566. 56. Steinmller, W. and E. Bock, Growth of Nitrobacter in the presence of organic matter. Archives of Microbiol ogy, 1976. 108(3): p. 299-304. 57. Steinmller, W. and E. Bock, Enzymatic studies on autotrophically, mixotrophically and heterot rophically grown Nitrobacte r agilis with special reference to nitrite oxidase. Archives of Microbiology, 1977. 115(1): p. 5154. 58. Bock, E., P. Wil derer, and A. Freitag, Growth of Nitrobacter in the absence of dissolved oxygen. Water Research WATRAG, 1988. 22(2). 59. Sundermeyer-Klinger, H., et al., Membrane-bound nitrite oxidoreductase of Nitrobacter: evidence for a nitrate reductase system. Archives of Microbiology, 1984. 140(2): p. 153-158.

PAGE 179

163 60. Zhang T. C., L.D.G., Sulfur:limestone autotrophic denitrification processes for treatment of nitrate-contam inated water: batch experiments Water Research, 1999. 33(3): p. 599-608. 61. Umbreit, W., Burris, R.H., Stauffer, J.F., Manometric techniques Chapter 2 1957, Minneapolis, Minnesota: Burgess Publishing Company. 62. Snoeyink, V. and D. Jenkins, Water Chemistry 1980: John Wiley and Sons. 463. 63. Davis, M.L., Water and Wastewater Engineeri ng: Design Principles and Practice 2011: McGraw Hill. 64. Copp, J.B., and K.L. Murphy, Estimation of the active nitrifying biomass in activated sludge. Water Research, 1995. 29: p. 1855-1862. 65. Katehis, D., J. Fi llos, and L.A. Carrio, Comparison of bench scale testing methods for nitrifier gr owth rate measurement. Water Science Technology, 2002. 46( 1-2): p. 289-295. 66. Nowak, P., Schweig hofer, and K. Svardal, Nitrification in hibition a method for estimation of actual maximu m growth rates in activated sludge systems. Water Science Technology, 1994. 30(6): p. 9-19. 67. Siegrist, H., et al., Reduction of biomass decay rate under anoxic and anerobic conditions. Water Science Technology, 1999. 39(1): p. 129-137. 68. Lee, Y., and J.A. Oleszkiewicz, Evaluation of maximu m growth and decay rates of autotrophs under different ph ysical and environmental conditions. in 75th Annual Conference of the Water Environment Federation. 2002: Chicago, IL. 69. Dagley, S. and C.N. Hinshelwood, Physicochemical Aspects of bacterial Growth. Part II. Quantitative dependenc e of the Growth Rate of Bact. Lactis aerogenes on the Carbon Dioxi de Content of the Gas Atmosphere. J. Chem. Soc., 1938. 70. Green, M., et al., The effect of CO2 concentration on a nitrifying chalk reactor. Water Res, 2002. 36(8): p. 2147-51.

PAGE 180

164 71. Denecke, M. and T. Liebig, Effect of carbon dioxide on nitrification rates. Bioprocess Biosyst Eng, 2003. 25(4): p. 249-53. 72. Wett, B. and W. Rauch, The role of inorganic car bon limitation in biological nitrogen removal of extremely a mmonia concentrated wastewater. Water Res, 2003. 37(5): p. 1100-10. 73. Kinsbursky, R.S. and S. Saltzman, CO2-Nitrification Relationships in Closed Soil Incubation Vessels. Soil Biol Biochem, 1990. 22(4): p. 571– 572. 74. American Public Health Association, Standard Methods for the Examination of Water and Waste Water 20 ed. 1998, Baltimore: United Book Press, Inc. 75. Jirka, A.M. and M.J. Carter, Micro semi-automated ana lysis of surface and wastewaters for chemical oxygen demand. Anal Chem, 1975. 47(8): p. 1397-1402. 76. Pitt, P. and D. Jenkins, Causes and control of Nocardia in activated sludge. Res. J. Water Polln. Contro l. Fedn., 1990. 37(2): p. 151-62. 77. Schmidt, I., et al., Ammonium and hydroxylamine uptake and accumulation in Nitrosomonas. Microbiology, 2004. 150(Pt 5): p. 1405-12. 78. Gordon, L. and A. Paskins, Influence of High Partia l Pressure of Carbon Dioxide and/or Oxygen on Nitrification. J. Chem Tech 1982. 32: p. 213-23. 79. Sakairi, M.A., K. Ya suda, and M. Matsumura, Nitrogen removal in seawater using nitrifying and denitrif ying bacteria immobilized in porous cellulose carrier. Water Science and Technology, 1996. 34(7-8): p. 267274. 80. Byong-Hee, J., T. Yasunori, and U. Hajime, Stimulating Accumulation of Nitrifying Bacteria in Por ous Carrier by Addition of Inorganic Carbon in a Continuous-Flow Fluidized Bed Wastewater Treatment Reactor. J. Biosci Bioeng, 2000. 89(4): p. 334-39.

PAGE 181

165 81. Bringmann, G., Vollsta¨ndige biologische St ickstoff-Elimination aus Kla¨rwa¨ssern im Anschlu an ein Ho chleistungs-Nitrifi kations-Verfahren. Gesund Ing Heft, 1961. 8: p. 233-5. 82. Jetten, M.S., et al., 1994-2004: 10 years of resear ch on the anaerobic oxidation of ammonium. Biochem Soc Trans, 2005 33(Pt 1): p. 119-23. 83. Jetten, M.S., et al., The anaerobic oxidat ion of ammonium. FEMS Microbiol Rev, 1998. 22(5): p. 421-37. 84. DelosReyes, F.L., W. Ritter, and L. Raskin, Group-specific small-subunit rRNA hybridization probes to char acterize filament ous foaming in activated sludge systems. Applied and Envi ronmental Microbiology, 1997. 63(3): p. 1107-1117. 85. Daims, H., et al., In situ characterization of Nit rospira-like nitrite oxidizing bacteria active in wastewater treatment plants. Applied and Environmental Microbiology, 2001. 67(11): p. 5273-5284. 86. Daims, H., Lucker, S., and Wagner, M, daime, a novel image analysis program for microbial ecol ogy and biofilm research. Environmental Microbiology, 2006(8): p. 200-213 87. Kapuscinski, J., DAPI: a DNA-specific fluorescent probe. Biotech Histochem, 1995. 70(5): p. 220-233. 88. Felske A., A.A.D.L.a.D.V.W.M., In situ detection of an uncultured predominant Bacillus in Dutch grassland soils. Applied Environmental Microbiology, 1998. 64: p. 4588-4590. 89. Pace, N.R., Structure and synthesis of the ribosomal ribonucleic acid of prokaryotes. Bacterioogical Review, 1973. 37: p. 562-603. 90. Jang, A., et. al., Assessment of characterist ics of biolfilm formed on autotrophic denitrification. Microbiol Biotechnology, 2005. 15(3): p. 455460. 91. McAdam, E. and S. Judd, A review of membrane bioreactor potential for nitrate removal from drinking water. Desalination, 2006. 196: p. 135-148.

PAGE 182

166 92. Park, J. and Y. Yoo, Biological nitrate removal in industrial wastewater treatment: which electr on donor we can choose. Applied Microbiol Biotechnology, 2009. 82: p. 415-429. 93. State of Florida Numeric Nutrie nt Criteria Development Plan in Florida 2009, Florida Department of Envi ronmental Protection: USA.