EPA 600/2-78-030
March 1978
Environmental Protection Technology Series
SOLUBLE ORGANIC NITROGEN
CHARACTERISTICS AND REMOVAL
Municipal Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental,Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-78-030
March 1978
SOLUBLE ORGANIC NITROGEN
CHARACTERISTICS AND REMOVAL
by
Stephen J. Randtke, Gene F. Parkin, John V. Keller,
James 0. Leckie, and Perry L. McCarty
Department of Civil Engineering
Stanford University
Stanford, California 94305
Grant No. R-804001
Project Officer
Charles I. Mashni
Wastewater Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does mention of
trade names or commercial products constitute endorsement or recommendation
for use.
ii
-------
FOREWORD
The Environmental Protection Agency was created because of increasing
public and government concern about the dangers of pollution to the health
and welfare of the American people. The complexity of the environment and
the interplay between its components require a concentrated and integrated
attack on the problem.
Research and development is that necessary first step in problem solu-
tion and it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Laboratory
develops new and improved technology and systems for the prevention, treat-
ment, and management of wastewater and solid and hazardous waste pollutant
discharges from municipal and community supplies, and to minimize the ad-
verse economic, social, health, and aesthetic effects of pollution. This
publication is one of the products of that research; a most vital communica-
tions link between the researcher and the user community.
Characterization of wastewater components and identification of their
sources is imperative in our continuing search for new and improved tech-
nologies for pollution abatement. This publication provides much needed
information on the sources, treatability and fate of the nitrogen containing
organic compounds present in wastewater.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
iii
-------
ABSTRACT
Soluble organic materials containing nitrogen (SON) are present in efflu-
ents from activated sludge treatment of municipal wastes. The objective of
this study was to determine the sources, concentrations, characteristics, and
methods for removal of SON to aid in the establishment of possible regulatory
criteria and in the design of treatment systems for its control.
SON ranged from 1.1 to 2.1 mg/1 in seven composite effluent samples
taken from four conventionally operated activated sludge plants which varied
in size, location, and influent waste characteristics. About two-thirds of
the SON in such effluents is estimated to be residual from the untreated
waste and the remainder is produced biologically during treatment.
About one-half of the biologically produced SON results from organism
decay, the concentration depending upon detention time and mixed liquor sus-
pended solids concentration. The other one-half represents excreted materi-
als, which approach a concentration in equilibrium with living cells that is
independent of cell concentration or aeration time. These biologically pro-
duced materials are relatively non-biodegradable. During process start-up
or stress, biologically produced SON may reach several mg/1, but this excess
material is quite biodegradable.
Ozone, chlorine (breakpoint), permanganate, and peroxide removed an
average of 14 + 7, 42+4, 28+3, and 19 ± 3 percent of the SON, respective-
ly, and 46 ±4, 18+7, 0±2, and 2+2 percent of the soluble chemical
oxygen demand (SCOD) , respectively. Ozone and chlorine removed different
fractions of the SON. Preozonation increased the fraction of SON removable
by chlorination by about 30 percent.
Chemical coagulation, ion exchange, and activated-carbon adsorption were
used singly and in combination to characterize different fractions of the SON
and SCOD. About 10 percent of the SON was removed by cation exchange and
coagulation at high pH, indicating that it was positively charged; this frac-
tion was not removed by activated carbon. About 20 percent of the SON ap-
peared to be negatively charged since it was removed by coagulation at pH 6
but not at pH 10. Activated carbon removed a relatively non-polar 70 percent
of the SON. Various coagulants specifically adsorbed 0 to 20 percent of the
SON, entrapped 5 percent with a larger size, and electrostatically adsorbed
the positively and negatively charged fractions. Ten percent of the SON re-
moved by anion exchange was also removed by activated carbon. A scheme is
provided which divides SON and SCOD into a number of such fractions, indicat-
ing which are mutually exclusive and which overlap.
iv
-------
Less than 10 percent of SON in secondary effluents consists of individual
or combinations of amino acids. Approximately 25 to 35 percent appear to be
heterocyclic compounds, probably nucleic acid degradation products having ap-
parent molecular weights of less than 780, similar to that in untreated
municipal wastewater. The excess SON produced during start-up, however,
generally had a much higher molecular weight distribution.
This report was submitted in fulfillment of Research Grant No. EPA-R-
804001 by Stanford University under the sponsorship of the U.S. Environmental
Protection Agency. This report covers a period from September 1, 1974 to
May 1, 1977, and work was completed as of June 15, 1977.
-------
CONTENTS
Foreword iii
Abstract iv
Figures viii
Tables x
List of Abbreviations and Symbols xiv
Acknowledgements xvi
1. Introduction 1
2. Conclusions 3
3. Recommendations 7
4. Sources, Occurrence, and Chemical Nature of Organic Nitrogen . . 8
5. Materials and Methods 17
6. SON in Municipal Secondary Effluents 21
7. SON Removal by Physical and Chemical Processes 33
Removal Mechanisms 33
Experimental Approach 36
Chemical Coagulation 41
Ion Exchange and Activated-Carbon Adsorption 54
Chemical Oxidation 68
Processes in Sequence 79
Discussion of Results 91
8. SON Formation and Removal by Biological Processes 103
Introduction 103
Nomenclature 103
Biologically Produced SON 104
Refractory Organics 104
SON Production and Excretion by Bacteria 110
Summary and Basis for Experimental Approach 114
Laboratory Activated Sludge Cultures 115
Batch Experiments 117
SON Removal 119
SON Production 130
Discussion of Results 169
9. SON Characterization 186
Introduction 186
Results 189
VII
-------
9. SON Characterization (continued)
Discussion of Results 206
10. Summary and Discussion 212
SON in Municipal Secondary Effluents 212
SON Characteristics 212
Control of SON 214
Ecological Significance of SON 215
References 217
Appendices
A. Statistical Calculations 234
B. Nitrate Interference with Kjeldahl SON Analysis ....... 251
viii
-------
FIGURES
Number Page
1 Diurnal Variation of SON—Palo Alto (4-14-75) 27
2 Diurnal Variation of SON—San Jose/Santa Clara (5-13-75) .... 28
3 Diurnal Variation of SON—Union City #3 (5-1-75) 29
4 Diurnal Variation of SON at Union City #3 (5-1-75) 30
5 Removal of SON by Ion Exchange as a Function of pH 61
6 Removal of SCOD by Ion Exchange as a Function of pH 62
7 Operational Fractionation of SON and SCOD 100
8 Laboratory Activated-Sludge System 116
9 Batch Study 1: Effect of Aeration Time on SON Remaining,
MLSS - 1300 mg/1 121
om
10 Batch Study 1: Effect of Aeration Time on SCOD Remaining,
MLSS - 1300 mg/1 122
om
11 Batch Study 2: Effect of Sludge Washing and Aeration Time on
SON Removal from unfiltered primary effluent, MLSS = 1200 mg/1. 123
12 Batch Study 3: Effect of MLSS on SON Removal from Filtered
Primary Effluent ?m 125
13 Batch Study 3: Effect of MLSS on SCOD Removal from Filtered
Primary Effluent ?m 126
14 Batch Study 4: Effect of Substrate on SON Remaining,
MLSS - 180 mg/1 128
om
15 Batch Study 4: Effect of Substrate on SCOD Remaining,
MLSS - 180 mg/1 129
om
16 AS Culture 1 Start-Up: SON and MLSS Concentrations vs Time
of Operation 131
17 AS Culture 2: SON and MLSS Concentrations vs Time of
Operation (All Data) 132
18 AS Culture 1, Days 20-340: SON and MLSS Concentrations vs
Time of Operation 135
19 Initial Release of SON for AS Culture 1 as a Function of MLSS
Concentration 139
20 Initial Release of SCOD for AS Culture 1 as a Function of MLSS
Concentration ,. . 140
ix
-------
Number Page
21 The Effect of Aeration Time on SON and SCOD Release during
Organism Decay at Different MLSS Values Using AS Culture 1
(Day 274) 147
22 The Effect of Aeration Time on SON Release during Organism
Decay at Different MLSS Concentrations Using Palo Alto Activated
Sludge (9-9-75) 148
23 The Effect of Aeration Time on SCOD Release During Organism
Decay at Different MLSS Concentrations Using Palo Alto
Activated Sludge (9-9-75) 149
24 The Effect of MLSS Concentration on SON and SCOD Release during
Organism Decay of AS Culture 1 152
25 Batch Study 5: Effect of SCODOC on SON Production Using AS
Culture 2, MLSS - 580 mg/1 (Day 171) 157
26 Batch Study 6: Effect of SCODOC on SON Production Using AS
Culture 1 Sludge, MLSS - 216 mg/1 (Day 61) 158
27 SCOD Removal vs Aeration Time in Batch Study 6. MLSS =
216 mg/1 (Day 61) °? 159
28 Batch Study 7: Effect of Substrate Type on SON Production
Using AS Culture 1, MLSS - 360 mg/1 (Day 98) 162
om
29 Batch Study 7: SCOD Removal vs Aeration Time,
MLSS = 360 mg/1 (Day 98) 163
om
30 Batch Study 8: Effect of MLSS on SON Production 164
om
31 Batch Study 9: Effect of NH3-N on SON Production, MLSS - 830
mg/1 (Day 176) 165
32 Conceptual Model of Changes in SONp, SONg, SONrf, and SONeq with
Batch Activated-Sludge Aeration Time 171
33 Conceptual Model of Changes in Sources of Effluent SON as a
Function of Activated-Sludge Aeration Time 179
34 Molecular Weight Distribution of Recovered SON and SCOD for
Soluble Raw Wastewater as Percent of Mass in Original Sample.
+ Figures are 95 percent confidence limits 199
35 Molecular Weight Distribution of Recovered SON and SCOD for
Soluble Primary Effluent as Percent of Mass in Original Sample.
+ Figures are 95 percent confidence limits 200
36 Molecular Weight Distribution of Recovered SON and SCOD for
Soluble Activated-Sludge Effluent as Percent of Mass in Original
Sample. + Figures are 95 percent confidence limits 201
37 Molecular Weight Distribution of Recovered SON and SCOD for
Soluble AS Culture 2 Effluent during Start-Up as Percent of Mass
in Original Sample. + Figures are 95 percent confidence limits. . 202
-------
TABLES
Number Page
1 Sources of Organic Nitrogen in Municipal Wastewater 9
2 Percentage of Total Organics in the Soluble Fraction of Raw
Wastewater 10
3 Nitrogen Forms in Two Raw Wastewaters 11
4 Non-Amino Acid Nitrogenous Constituents of Domestic Wastewater. . 11
5 Total and Soluble Organics in Secondary Effluent 13
6 Characterization of Soluble Secondary Effluent Organics 14
7 Precision of SON and SCOD Analyses 18
8 Significant Differences in SON and SCOD Analyses 19
9 Effect of Initial Ammonia Distillation pH on the SON Analysis . . 19
10 General Characteristics of the Influent and Primary Effluent
Wastewaters 23
11 General Characteristics of the Secondary Effluents 24
12 Summary of SON Concentrations at Secondary Treatment Plants ... 25
13 General Characteristics of Palo Alto Secondary Effluent 31
14 Removal Mechanisms and Factors Affecting Removal 33
15 Treatment Processes and Removal Mechanisms 35
16 Comparison between Filtered and Unfiltered Coagulated PASE
Samples 41
17 Coagulation of Filtered and Unfiltered PASE Samples 42
18 SON Removal as a Function of Mixing Time at 25 rpm 43
19 Effect of pH on Ferric Chloride Coagulation of SON 43
20 SON and SCOD Removal by Ferric Chloride Coagulation 45
21 SON and SCOD Removal by Alum Coagulation 46
22 SON and SCOD Removal by Lime Coagulation 47
23 Comparison of SON and SCOD Removal by Ferric Chloride, Lime, and
Alum Coagulation 48
24 Effect of Sequential Coagulation on SON Removal 50
25 Additional SON Removals Achieved by Sequential Coagulation ... 51
xi
-------
Number jPage
26 Removal of SON and SCOD by Addition of Ferric Chloride and C-31
Coagulant to Activated-Sludge .................. 53
27 Manufacturers' Specifications for the Ion-Exchange Resins Used . 57
28 Effect of Sequential Contacting and Dosage of Ion-Exchange Resins
on Removals from PASE ...................... 58
29 SON and SCOD of Deionized Water Samples Contacted with Ion-
Exchange Resins and Activated Carbon .............. 59
30 Removal of SON and SCOD by Sequential Contacting with Ion-
Exchange Resins ......................... 63
31 Removal of SON and SCOD by Activated Carbon and Ion- Exchange
Resins ............................. 64
32 Removal of SON and SCOD from Treated PASE at Neutral pH ..... 65
33 Removal of SON and SCOD by Activated Carbon as a Function of pH . 65
34 Summary of Activated-Carbon Adsorption Data at Neutral pH . . . . 66
35 Summary of SON and SCOD Removal by Ion Exchange ......... 67
36 Removal of SON and SCOD by Chlorine, Permanganate, and Peroxide . 74
37 Effect of Dechlorination Method of Lawrence e£ al ........ 75
38 Removal of SON and SCOD by Breakpoint Chlorination ....... 76
39 Removal of SON and SCOD by Ozonation .............. 77
40 Comparison of SON and SCOD Removals by Chlorine and Ozone .... 78
41 Removal of SON and SCOD by Combinations of Chlorine and Ozone . . 80
42 Removal of SON and SCOD by the South Tahoe P.U.D. Water Reclama-
tion Plant ........................... 82
43 Results of Laboratory Advanced Wastewater Treatment ....... 83
44 Removal of SON and SCOD by Ozone in Sequence with Chemical
Coagulation ........................... 85
45 Removal of SON and SCOD by Ozone in Sequence with Ion Exchange
and Activated-Carbon Adsorption ................. 86
46 Removal of SON and SCOD by Ozonation and Activated-Carbon
Column Adsorption ........................ 87
47 Removal of SON and SCOD by Ferric Chloride Coagulation and
Ion Exchange .......................... 88
48 Removal of SON and SCOD from Coagulated PASE by Ion Exchange at
Neutral pH ...........................
49 Removal of SON and SCOD by Coagulation, Ion Exchange, and
Activated-Carbon Adsorption ................. OQ
50 Summary of Molecular Characteristics Affecting Removal of
Organic Contaminants ...................
xii
-------
Number Page
51 Summary of Removals Achieved by Individual Processes 94
52 Summary of Results of Chemical Oxidation Experiments 101
53 Biologically Produced SON Compounds 105
54 Typical Nucleic Acid Degradation Products 109
55 Composition of Synthetic Waste 117
56 Filtration and SON Analysis Used for Batch Studies 118
57 Summary of Substrates, Filtration, and SON Analysis Used during
Removal Studies 119
58 The Effect of Growth Rate on AS Culture 2 Effluent SON during
Culture Start-Up 133
59 Summary of Steady-State Data for Activated-Sludge Cultures . . . 134
60 Characteristics of Mixed Liquor Suspended Solids for Activated-
Sludge Cultures 136
61 Effect of Culture on SON and SCOD 141
q q
62 SON Released by Sequential Dilution of Palo Alto Activated Sludge 141
63 Initial Release of SON and SCOD upon Exposure to Tap Water, AS
Culture 1 Filtrate, and Released Organics 143
64 Effect of Temperature on Initial Release of SON and SCOD .... 144
65 Initial Release of SON and SCOD: Comparison of Substrates . . . 145
66 Equilibrium SON and SCOD Values for the Cultures Studies .... 146
67 Summary of Regression Analysis Coefficients Correlating SON and
SCOD Release with Aeration Time 150
68 Linear Regression Equations Showing the Effect of MLSS on SON and
SCOD Release during Organism Decay of Palo Alto Activated Sludge. 153
69 Variables Studied and Filtration and SON Techniques Used during
Synthetic Feed Studies 154
70 Summary of SON Data for All Experiments 167
71 Summary of Data Describing SON Peaks during Aeration 168
72 Contribution of SON , SON,, and SON to SON for AS Cultures . . 176
eq d g p
73 Estimation of SON , SON + SON , SON , and SON in Palo Alto
Secondary Effluentq(MLSS = 13908mg/l)r b 182
74 Estimation of SONeq, SONd + SONg, SONr, and SONb in Palo Alto
Secondary Effluent (MLSS - 180 mg/1) 183
75 Estimation of SONeq, SONd + SONg, SONr, and SONfc in Palo Alto
Secondary Effluent at tmin 184
76 List of Samples Characterized 186
xiii
-------
Number JPage
77 Compounds Used for Molecular Weight Calibration of Sephadex
Columns l88
78 Wastewaters Evaluated for Biodegradability 189
79 Summary of Rate and Extent of Biodegradability for Treated and
Untreated Wastewaters
190
80 Biodegradability of Remaining Soluble Organics after Various
Periods of Treatment of Primary Effluents 191
81 Summary of Rate and Extent of Biodegradability for Biologically
Produced Organics 193
82 Summary of SON and SCOD Recoveries from Concentration and
Filtration 196
83 Summary of SON Elution Data for Molecular Weight Distribution
Studies 197
84 Summary of SCOD Elution Data for Molecular Weight Distribution
Studies 198
85 SON/SCOD Ratios for the Molecular Weight Fractions of the
Filtered Wastewaters Studied 204
86 Theoretical SON/SCOD for Typical Nitrogen-Containing Organic
Compounds of Differing Molecular Weights 205
87 Amino Acid Data for Palo Alto Activated-Sludge Effluent
(6-14-74) 207
88 Summary of Biodegradability Data 208
89 Comparison of Results from MW Distribution and Cationic Exchange
Studies 210
A.I Confidence Intervals for SON and SCOD Analyses 238
A.2 Significant Differences in Percent Removal 240
A. 3 Removal of SCOD be Ferric Chloride and Lime 240
A.4 Precision and Accuracy of Low-Level Kjeldahl SON Analyses .... 242
A.5 Precision and Accuracy of Technicon SON Analysis 243
A.6 Precision of Low-Level SCOD Analysis 244
A. 7 Precision of Kjeldahl SON and SCOD Analyses at High Concentrations 244
A. 8 Confidence Intervals for SON and SCOD Analysis 246
A.9 Confidence Intervals for Differences Involving SON and SCOD . 246
OC OC
B.I Effect of NO~-N Concentration on SON Recovery from a Biologically
Treated Wastewater 252
B.2 Effect of NO~-N Concentration on SON Recovery from Palo Alto
Activated-Sludge Effluent 252
B.3 Effect of Compounds Tried for Elimination of Nitrate
Interference 253
xiv
-------
LIST OF ABBREVIATIONS AND SYMBOLS
A
a
b
dSON.
pk
dSONpkm
t
dSONt(i)
dSONfc(i)
dSONt(i-j)
dSONt(i-j)
e
f
kn
MLSS
MLSSom
MLVSS
y
NH3-N
NH3-NOC
N05-N
N03-N
r
SBOD5
SCOD
SCOD,
'eq
SCOD
SCODr
SOC
oc
fitting parameter in first-order decay model
linear regression coefficient (y-axis intercept)
linear regression coefficient (slope)
minimum SCOD concentration observed during a batch experiment;
obtained by subtracting the initially calculated SCOD (SCODOC)
from the minimum SCOD measured (SCODmin) during the experiment,
mg/1
maximum SON concentration produced during a batch experiment;
obtained by subtracting the initially calculated SON (SONOC)
from the maximum SON measured (SONpea]c) during the experiment,
mg/1
difference between the maximum SON (SONpeak) and minimum SON
(SONmin) measured during batch SON production experiments, mg/1
SON produced after t hours of aeration; obtained by subtracting
the initially calculated SON (SONOC) from the measured SON at
aeration time t (SONt), mg/1
dSONt for test condition ±, mg/1
average of dSONt(i) values, mg/1
difference between dSONt for test condition i (dSONt(i)) and
control condition j (dSONt(j)), mg/1
average of dSONt(i-j) values, mg/1
standard error of estimate for linear regression analysis,±mg/1
decimal fraction of biodegradable influent SON (80%) that is
not removed by activated-sludge treatment
first-order decay rate for SON, day-1
first-order decay rate for SCOD, day-1
mixed liquor suspended solids concentration, mg/1
MLSS measured at time 0 of batch experiment, mg/1
mixed liquor volatile suspended solids concentration, mg/1
organism growth rate, day-1
ammonia nitrogen concentration, mg/1
initially calculated NH3-N, mg/1
nitrite-nitrogen concentration, mg/1
nitrate-nitrogen concentration, mg/1
correlation coefficient
soluble five-day biochemical oxygen demand, mg/1
soluble chemical oxygen demand, mg/1
SCOD concentration in "equilibrium" with activated-sludge orga-
nisms, mg/1
initially calculated SCOD, mg/1
SCOD initially released by organisms upon dilution with tap
water to reach the "equilibrium" SCOD concentration, mg/1
soluble organic carbon concentration, mg/1
xv
-------
SON
SONb
SONd
SONdb
SONdr
SON0
SON
eq
SONf
SONg
SONgb
SONgr
SONml
SON0
SONOC
SONocd)
SONom
SONp
SONpb
SON
SON
pr
SONr
SONrf
SONt
SONt(i)
SONta
SONtm
TBOD5
TCOD
TON
Vf
soluble organic nitrogen concentration, mg/1
biodegradable SON in untreated wastewater, mg/1
SON produced during activated-sludge treatment by organism decay,
mg/1
biodegradable SONd, mg/1
refractory SONd, mg/1
SON in effluents from activated-sludge treatment, mg/1
SON concentration in "equilibrium" with activated-sludge orga-
nisms , mg/1
SON of defined-substrate feed solutions, mg/1
SON produced during activated-sludge treatment as a result of
substrate oxidation, mg/1
biodegradable SON , mg/1
refractory SONg, mg/1
SON of untreated (influent) wastewater, mg/1
minimum SON measured during a batch experiment, mg/1
SON of the activated-sludge mixed liquor used in batch experi-
ments , mg/1
SON measured on day 0 of low seed biodegradation study, mg/1
initially calculated SON, mg/1
SONOC for test condition i
SON measured at time 0 of batch experiment, mg/1
SON produced during activated-sludge treatment, mg/1
biodegradable SONp, mg/1
maximum SON measured during production batch experiments, mg/1
refractory SONp, mg/1
SON initially released by organisms upon dilution with tap water
to reach the "equilibrium" SON. concentration, mg/1
refractory SON in untreated wastewater, mg/1
estimate of refractory SON; used in first-order decay model,
mg/1
SON measured after t days of low seed biodegradation, mg/1
SON measured after t hours of batch activated-sludge aeration
for test condition i, mg/1
• SON of tap water, mg/1
• SON measured after t hours of batch activated-sludge aeration,
mg/1
• batch activated-sludge aeration time, hrs, or low seed biode-
gradation incubation time, days
• unfiltered five-day biochemical oxygen demand, mg/1
• unfiltered chemical oxygen demand, mg/1
• hydraulic detention time, hours
• solids retention time, days
• unfiltered organic nitrogen, mg/1
• volume of feed solution added to batch activated-sludge reactor
liters
• volume of activated-sludge mixed liquor added to batch activated-
sludge reactor, liters
xv i
-------
ACKNOWLEDGMENTS
The cooperation of the personnel at the Palo Alto Regional Water Quality
Control Plant, the San Jose/Santa Clara Water Pollution Control Plant, the
Union District Plant #3, and the South Tahoe Public Utility District Water
Reclamation Plant in the collection of samples and use of their facilities
for portions of this study is gratefully acknowledged.
xvii
-------
SECTION 1
INTRODUCTION
BACKGROUND
In 1972 The United States Congress passed the Federal Water Pollution
Control Act Amendments [1] declaring that "it is the national goal that the
discharge of pollutants into the navigable waters be eliminated by 1985 [and]
it is the national policy that a major research and demonstration effort be
made to develop technology necessary to eliminate the discharge of pollu-
tants." As an interim step in fulfilling this national goal, secondary
treatment of all municipal wastewaters will be required by 1977. Currently
a major effort is being made to identify the pollutants present in secondary
effluent and to develop processes for their removal.
There are a large number of organic and inorganic constituents present
in secondary effluent which can be discharged into receiving waters. Of
these constituents, those containing nitrogen have received a great deal of
attention due to their ability to stimulate undesirable algal blooms. The
inorganic forms of nitrogen have also been found to act as pollutants in
other ways: nitrate can cause methemoglobinemia in infants; nitrite can com-
bine with organic compounds to form nitrosamines, which are potent carcino-
gens; and ammonia is toxic to fish and can cause lowering of the dissolved
oxygen level in receiving waters [2]. These inorganic forms of nitrogen have
been extensively studied, and many processes have been developed for their
removal. Much less is known regarding the occurrence and removal of the
organic forms of nitrogen.
The U.S. Environmental Protection Agency has been given the responsibil-
ity of setting standards for the discharge of chemical substances into the
environment. In order to set such standards intelligently a great deal of
information will be required, to insure both the protection of the public
interest and the attainability of the standards. In the case of organic
nitrogen, it is desirable to know (1) its concentration and variation in
concentration in municipal secondary effluents; (2) if the material is bio-
degradable; (3) if it can stimulate the growth of algae; (4) if it poses a
public health problem; and (5) the degree to which it can be removed by ad-
vanced waste-treatment processes.
In a strict literal sense "organic nitrogen" includes any nitrogen atom
that is part of an organic molecule. However, the term "organic nitrogen"
has been used extensively in the literature to designate organically bound
-------
nitrogen in the tri-negative state as measured by the Kjeldahl organic nitro-
gen analysis [3]. This definition will be adhered to in this study.
There is little information available concerning the concentration and
nature of the organic nitrogen present in municipal secondary effluent, al-
though it may constitute more than half of the total nitrogen in a denitrified
effluent [4]. A certain portion of this organic nitrogen is in the form of
particulate matter which has failed to settle during secondary sedimentation.
This fraction of the organic nitrogen is quite easily removed by filtration.
The remainder of the organic nitrogen, which cannot be removed by filtra-
tion, is designated as soluble organic nitrogen (SON), and is the subject of
this study. SON will be arbitrarily defined as that portion of Kjeldahl
nitrogen capable of passing through a 0.45 micron membrane filter. This is
an operational definition, and it is quite possible that a significant part
of the SON defined in this way is not in true solution, but rather in a col-
loidal state.
OBJECTIVES
Soluble organic material containing nitrogen represents a significant
portion of the nitrogen present in treated municipal wastewaters, but little
is known of the effects this material may have on receiving waters, its char-
acteristics, its sources, or methods for its removal. Proposed standards for
total nitrogen limitation in treated effluents often include this material,
but data relating to its concentration and variation in concentration and
studies on methods for its removal are lacking. Thus, information about sol-
uble organic nitrogen is needed in order to establish sound criteria and to
design treatment systems for its removal. The objectives of this study were:
1. To determine the concentration and variation in concentration of
SON in municipal secondary effluents.
2. To evaluate the potential of biological, physical, and chemical
treatment processes and combinations of processes for removal of
SON, including activated-sludge treatment, chemical coagulation,
chemical oxidation, and adsorption.
3. To determine what portion of SON in secondary effluents is residual
from the untreated wastewater and what portion is formed biologi-
cally during treatment.
4. To characterize SON chemically and operationally in order to better
understand its potential sources, nature, potential ecological
effects, and susceptibility to removal by treatment processes.
5. To determine differences and similarities in behavior between SON
and the other soluble organic materials present in secondary efflu-
ents.
-------
SECTION 2
CONCLUSIONS
SON IN SECONDARY MUNICIPAL EFFLUENTS
1. The concentration of SON in the secondary effluents ranged from 1.1
to 2.1 mg/1, with an average of 1.5 mg/1 at four activated-sludge
treatment plants, ranging in capacity from 3 to 141 mgd and varying
in influent waste characteristics.
2. Diurnal variations in the concentration of SON in the secondary ef-
fluents of the three activated-sludge treatment plants evaluated
were small, with maximum to minimum ratios ranging from 1.7 to 1.3,
and maximum to average ratios ranging from 1.1 to 1.3 mg/1. Varia-
tions in SON and SCOD (soluble COD) were very similar in all cases.
SECONDARY EFFLUENT SON REMOVAL BY PHYSICAL AND CHEMICAL PROCESSES
1. A sequence of advanced wastewater-treatment processes designed for
wastewater reclamation removed a large amount (perhaps 70-90 percent)
of the SON and SCOD from secondary effluents, the bulk of the re-
moval occurred during coagulation and activated-carbon adsorption.
2. A fractionation scheme was developed which separates the SON and
SCOD into fractions based on their removal by the various individual
processes and combinations of processes.
3. Ferric chloride, lime, and alum removed 42 ± 5, 34+5, and 27+3
percent of SON in secondary effluent, and 32 + 6, 24 + 8, and 28 + 10
percent of the SCOD, respectively. The differences in SON removal
between the three coagulants were statistically significant, but dif-
ferences in SCOD removal were not. Ferric chloride and lime removed
significantly more SON than SCOD.
4. Polyelectrolytes produced very rapidly settling floes, but were of
no benefit in removing SON.
5. Bentonite improved SON removal by about 10 percent at a dose of 200
mg/1, but no increased removal of SON was found with kaolinite.
6. Direct addition of ferric chloride to activated sludge resulted in
about 30 percent removal of SON at a dose of 200 mg/1, although a
-------
significantly higher removal of SON could be attained by separate
coagulation of the supernatant liquor. Direct addition merits con-
sideration for economic reasons.
7. Activated carbon achieved an average SON removal of 71 ± 12 percent,
and an average SCOD removal of 81 ± 8 percent, reflecting a signifi-
cantly higher removal of SCOD than of SON.
8. Cation exchange at neutral pH removed 11 ± 1 percent of the SON and
4 ± 5 percent of the SCOD. Anion exchange at neutral pH removed
12 ± 4 percent of the SON and 30 ± 4 percent of the SCOD.
9. Sequential contacting with ion-exchange resins revealed that the
fractions of SON and SCOD removed by the cationic resin at pH 2 and
by the anionic resin at pH 12 were significantly different, as were
the fractions of SON and SCOD removed by both resins at neutral pH.
However, there appeared to be some overlapping of the fractions re-
moved by the two resins at both low and high pH.
10. With a cation-exchange resin, decreasing pH sharply increased the
removal of both SON and SCOD, the percentage removed being signifi-
cantly greater for SON than for SCOD. At pH 2.0, the cationic resin
removed 42 + 7 percent of the SON and 22 + 4 percent of the SCOD.
11. At neutral pH, cation-exchangeable SON was not adsorbed by activated
carbon, anion-exchangeable SON was partially adsorbed by activated
carbon, and anion-exchangeable SCOD was almost completely adsorbed
by activated carbon.
12. The SON and SCOD remaining after activated-carbon adsorption are not
readily attacked by ozone.
13. The fraction of SON removed by cation exchange at neutral pH is also
removed by ferric chloride coagulation at pH 10, but not at pH 6.
The fractions of SON and SCOD removed by anion exchange are not re-
moved by ferric chloride coagulation at either pH 6 or 10.
14. At pH 7.0, chlorine dosages less than 100 mg/1 (below the breakpoint)
produced no significant removal of SON and very little (6 ± 2 per-
cent) removal of SCOD. A chlorine to ammonia nitrogen ratio of ap-
proximately 20:1 removed an average of 42 + 4 percent of the SON and
18 ± 7 percent of the SCOD. A slightly lower amount of oxidation of
SON was found for a 10.4:1 ratio, but SCOD removal was unaffected by
an increase above a ratio of about 10.
15. Potassium permanganate and hydrogen peroxide removed a maximum of
28 + 3 and 19 + 3 percent of the SON, respectively, at pH 10.0 and
11.0, respectively. They produced no significant removal of SCOD.
16. Ozonation removed an average of 14 + 7 percent of the SON and 46+4
percent of the SCOD, at absorbed dosages mostly in excess of 100 mg/L
-------
17. The fractions of SON removable by ozone and chlorine were found to be
mutually exclusive. Removal of SON by chlorine was increased 30 per-
cent by preozonation. Ozone removed a considerably greater amount of
SCOD than chlorine.
18. Ozonation increased the biodegradability of remaining SON and SCOD by
29 ± 4 and 17 ± 4 percent, respectively.
SON PRODUCTION AND REMOVAL BY BIOLOGICAL TREATMENT
1. From 60 to 70 percent of the SON and 70 to 80 percent of the SCOD in
untreated municipal wastewater was removed by activated-sludge treat-
ment; SCOD was in general removed at a faster rate than SON.
2. A minimum in the concentrations of effluent SON is predicted to re-
sult when municipal activated-sludge plants are operated at conven-
tional loadings, such as aeration times of six hours and solids
retention times of four to ten days. Higher or lower loadings should
result in increased SON concentrations.
3. Under the above optimal conditions and with normal operation, about
two-thirds of the effluent SON represents residual organics from the
influent wastewater and the other one-third represents materials pro-
duced biologically during activated-sludge treatment. The effluent
SON is relatively refractory to biological degradation.
4. During activated-sludge plant start-up and following plant upsets,
effluent SON concentrations can increase by several mg/1 as a result
of biological production. The incremental SON produced during these
periods is fairly biodegradable.
5. Under normal operating conditions at conventional loadings, about
one-half of the SON produced biologically during treatment results
from organism decay and the other one-half represents biological
exudates which approach an "equilibrium" concentration with the liv-
ing cells. The equilibrium concentration is independent of cell
concentration but is dependent upon culture characteristics.
6. Some evidence suggests that a separate SON fraction may be produced
biologically as a response to substrate oxidation. However, this
fraction is generally small in comparison with the fractions pro-
duced by decay and in equilibrium with the cells.
7. The quantity of SON released during organism decay is a function of
mixed liquor volatile solids concentration and time of aeration.
This material is refractory to biological oxidation.
8. The extent of biological production of SON is consistent for a given
activated-sludge culture, but varies greatly between cultures. Thus,
generalizations developed with one culture may not apply to other
cultures, even if treating a similar wastewater.
-------
CHARACTERISTICS OF EFFLUENT SON
1. Less than ten percent of effluent SON was comprised of free or com-
bined amino acids.
2. Evidence suggests that 15 to 30 percent of the SON was represented
by nucleic acid degradation products, specifically nucleic acid
bases and other heterocyclic nitrogen-containing compounds.
3. Effluent SON was much more refractory (40 to 50 percent with first-
order decay rates for the remainder of 0.014 day"1) than SON in un-
treated wastewater (18 to 38 percent with decay rates for the remain-
der of 0.08 to 0.16 day"1).
4. Fifty to sixty percent of the SON and SCOD in treated and untreated
municipal wastewater had apparent molecular weights less than 1800
as measured by Sephadex gel filtration.
5. About 25 percent of the SON had an apparent molecular weight between
165 and 340 and a high ratio of SON to SCOD of 0.31 ± 0.11, compared
with the typical ratio for other ranges of 0.045 ± 0.015. This and
other evidence suggests this material consists largely of hetero-
cyclic nitrogen-containing compounds such as nucleic acid degradation
products.
6. The increased concentration of SON and SCOD produced biologically
during activated-sludge culture start-up had a generally higher
molecular weight distribution than normal secondary effluent, with
considerably more organics in the greater-than-1800 molecular weight
range.
7. The organic materials associated with effluent SON and SCOD were
characterized operationally by various physical and chemical pro-
cesses. About 70 percent of the SON and 80 percent of the SCOD
represented relatively non-polar materials (aside from charge) re-
movable by activated-carbon adsorption, of which about one-third had
a predominantly negative charge and the remainder a neutral charge.
The remaining 30 percent of the SON and 20 percent of the SCOD was
relatively polar, one-third being characterized as positive and the
remainder as neutral.
-------
SECTION 3
RECOMMENDATIONS
1. Judgement should be exercised in the formulation of effluent standards
which may specify or otherwise include limitations on the discharge of
SON to receiving waters. The SON contained in effluents from secondary
treatment of municipal wastewaters is generally low in concentration and
also is quite refractory or slow to degrade biologically. No significant
environmental problems have yet been shown to result from discharge of
this material to receiving waters. While up to 80 percent of this mate-
rial may be removed by physical and/or chemical treatment, the construc-
tion and operational costs for the required facilities would be high.
2. A relatively large fraction of the SON-containing compounds in secondary
effluents appears to be degradation products of nucleic acids. Further
research to determine the nature of this particular material and the
effect chlorination may have on organisms which may re-incorporate this
material into their cellular components appears worthy.
3. A large fraction of SON in secondary effluents appears to be formed dur-
ing biological treatment of wastes. Perhaps a significant portion of
the SON arriving at treatment plants is also produced during biological
decomposition of human waste products during transport to the treatment
plant. More detailed characterization of such biologically produced
organics is recommended in order to obtain a better understanding of
the naturally produced refractory organic materials occurring in natural
waters and treatment plant effluents.
-------
SECTION 4
SOURCES, OCCURRENCE, AND CHEMICAL NATURE OF ORGANIC NITROGEN
INTRODUCTION
This section reviews the sources and concentration of organic nitrogen in
raw wastewater and secondary effluents. Since much of the organic nitrogen is
of natural origin, the characteristics of organic nitrogen found in soils,
sediments, surface waters, and marine environments is briefly examined. This
is followed by a discussion of the chemical nature of organic nitrogen.
SOURCES OF ORGANIC NITROGEN
Urine and feces are the two major sources of organic nitrogen in munici-
pal wastewater. Painter et al. [14] estimated the contribution of organic
nitrogen to be 1-2 g/day/adult from feces and 0.5-3 g/day/adult from urine.
These and other sources of organic nitrogen are listed in Table 1, together
with the names of compounds or classes of compounds thought to be present.
Sources of organic nitrogen other than urine and feces may also be impor-
tant. Numerous commercially available chemicals are frequently poured down
household drains and sewers. Urban runoff has been found to have an organic
nitrogen concentration of 1-9 mg/1 [2], and may be important in treatment
plants with a combined sewer system or to receiving waters where runoff is
not treated. The extra-cellular material of organisms and their cell frag-
ments also contribute to the organic nitrogen in wastewater.
ORGANIC NITROGEN IN RAW WASTEWATER
Painter et al. [14], Hunter and Heukelekian [15], Rickert and Hunter [16],
and Helfgott et al. [13] used various techniques to separate domestic sewage
into a number of different fractions. The percentages of the total organics
which they found in the soluble fraction are shown in Table 2. Hunter and
Heukelekian [15] found a smaller percentage of the organic nitrogen in the
soluble fraction than other investigators, but this may have been due to en-
vironmental factors such as temperature or travel time in the sewer, or to
differences in the separation techniques.
The class of compounds which has received perhaps the widest attention in
raw wastewater is the amino acids [5,6,13,14,15,17,18]. Kahn and Wayman [18]
found 13 amino acids in raw wastewater in concentrations of 10-15 mg/1, but
-------
TABLE 1. SOURCES OF ORGANIC NITROGEN IN MUNICIPAL WASTEWATER
Source
Nitrogenous Compounds Present
References
Jrine
Feces
Industrial Process
Effluents
Urban Runoff
animal excreta
plant matter
fertilizers
chemicals
Bacteria, Algae,
Fungi
exudates
cell walls
cell fragments
Urea, uric acid, creatine, amino acids
Amino acids, biotin, folic acid, nicotine
acid, pantothenic acid, riboflavin,
thiamine, purine bases
Aliphatic amines, arylamines, hetero-
cyclic amines, isocyanate.s, amino acids,
urea, thiourea
(See feces)
Urea
Cationic detergents (amino, quaternary
ammonium compounds)
Hydroxyamino acids, proteins, amino
acids, amides
Muramic acid, amino sugars
Nucleic acids, chitin
5,6
5
2
2
7,8,9,10
601^2,13,14
6
only 7 of these remained after primary treatment in concentrations of about
5 mg/1, indicating that the amino acids were largely present as particulate
matter. They also found approximately 41 percent of the amino acids in the
free form, although other investigators [6,17,18] have found only a few per-
cent of the proteinaceous matter in raw sewage to consist of free amino acids.
This discrepancy is attributed to environmental factors [18],
Hanson and Lee [6] found averages of 43% and 54% of the organic nitrogen
in the raw wastewater at two primarily domestic treatment plants in the form
of urea and alpha amino acids. The ranges of concentrations for various ni-
trogen forms in the two raw wastewaters are presented in Table 3.
In addition to amino acids, many other nitrogenous compounds have been
identified in raw wastewater. Hunter [5] has listed a number of these as
shown in Table 4.
ORGANIC NITROGEN IN SECONDARY EFFLUENT
It is anticipated that the organics present in secondary effluent differ
greatly from those found in raw wastewater. The differences are manifested in
the magnitude of the gross parameters used to measure the organics, in the
relative size of the soluble fraction, in the percentage of organics classi-
fied in any of the various schemes, and in the number and concentration of
specific chemical compounds present.
-------
TABLE 2. PERCENTAGE OF TOTAL ORGANICS IN THE SOLUBLE FRACTION OF RAW WASTEWATER
Organic Nitrogen
Total
mg/1
35
27
-
10.9
-
_
—
Percent
Soluble
39
37
-
53
24
20
43
Organic Carbon
Total
mg/1
412
311
-
-
-
-
—
Percent
Soluble
29
29
42
-
-
-
—
Chemical Oxygen
Demand
Total
mg/1
-
-
-
299
-
-
—
Percent
Soluble
-
-
40
43
25
23
—
Volatile Solids
Total
mg/1
-
-
116
232
-
-
—
Percent
Soluble
-
-
48
48
37
42
—
Reference
Painter et al. [14]
Painter et al. [14]
Rickert and Hunter [16J
Helfgott et al. [13]
Hunter and Heukelekian [15]
Hunter and Heukelekian [15]
Hanson and Lee [&]
-------
TABLE 3. NITROGEN FORMS IN TWO RAW WASTEWATERS [6]
Plant:
Population Services:
Number of Samples:
Nine Springs
200,000
13
Cross Plains
950
8
NH3-N, mg/1
Urea-N, mg/1
Org-N, mg/1
a-amino acid-N, mg/1
TKN, mg/1
Avg. Min. Max.
11.2 6.6 14.4
1.4 0.0 4.9
7.3 5.1 18.7
2.3 0.8 7.4
18.4 11.7 31.2
Avg. Min. Max.
25.9 16.8 30.0
6.1 3.9 10.3
16.1 10.9 21.8
3.0 2.4 4.3
42.0 27.7 49.9
TABLE 4. NON-AMINO ACID NITROGENOUS CONSTITUENTS OF DOMESTIC WASTEWATER [5]
Compound
Concentration, mg/1
Urea
Muramic acid
Amino sugars
Uric acid
Hippuric acid
Xanthine
Indole
Skatole
Aliphatic amines
Greatine-creatinine
Organic bases
Thiamine
Riboflavin
Niacin
Cobalamin
Biotin
Pantothenic acid
Folic acid
2-16
0.5
1.2-2.2
0.2-1.0
present
trace
0.00025
0.00025
0.1
0.2-7.0
3.4
0.029
0.022-0.044
0.135
0.0008
0.0003
present
present
11
-------
A typical activated-sludge plant can be expected to achieve approximately
85-95% removal of five-day biochemical oxygen demand (8005) [24], and also
significant percentages of organic nitrogen, organic carbon, chemical oxygen
demand (COD), and volatile solids. The concentrations of some of these param-
eters found by various investigators are shown in Table 5.
The information available regarding the concentration of organic nitrogen
in secondary effluent is quite sparse. Several investigators have measured
organic nitrogen in nitrified samples, but nitrate interferes in the organic
nitrogen analysis (see Appendix B). Bunch et al. [23] do not state whether or
not the plants they sampled were nitrifying. The effluent studied by Rebhun
and Manka [22] originated from a raw wastewater with a COD of 1200 mg/1 and
was treated with a trickling filter, thus producing the high concentrations of
organics shown in Table 5. The organic nitrogen concentration of 6.1 mg/1 re-
ported by Beckman et al. [19] is attributed to very cold temperatures and
failure of the organisms to hydrolyze protein.
A comparison of Table 5 with Table 2 indicates that a much larger percen-
tage of the organics in secondary effluent are soluble. Soluble generally re-
fers to materials passing a 0.45 ym filter. It is not known how much of the
soluble organic matter present in the effluent was originally present in the
raw wastewater and how much of it was produced by bacteria during the treat-
ment process. A number of investigators have shown that microorganisms are
capable of excreting extracellular organic matter [7,8,9,10] and some of the
organics in secondary effluent may be comprised of these exudates, as well as
cell walls and fragments of microorganisms [12,13].
Many soluble organic compounds identified in raw wastewater are not found
in treated effluents. Urea is rapidly hydrolized by the enzyme urease, in
some cases before it reaches the treatment plant [6], and does not appear in
treated effluent. Free amino acids identified in raw wastewater are not found
in secondary effluents [17,18]. Rudolphs and Chamberlin [25] found the nitro-
genous degradation products indole and skatole in raw wastewater, but detected
only trace amounts of them in treated effluents.
Secondary effluent organics have been characterized into known classes of
compounds as shown in Table 6. Only 35-50% of the soluble organics could be
classified in this way, the unclassified organics being generally labelled as
humic substances. Bunch et al. [23] concluded that 40% of the soluble COD and
50-70% of the organic nitrogen comprised high molecular weight compounds since
they would not dialyze within four days.
For the three biological processes studied (trickling filter, stabiliza-
tion pond, and extended-aeration activated sludge), Manka et al. [26] showed
that the distribution of the main soluble organic fractions (in secondary ef-
fluent) is similar for the various biological treatment units. Manka et al.
[26] and Rebhun and Manka [22] classified 19-25% of the soluble organics as
proteins; however, this estimate may be high since fulvic acid has been found
to interfere in the Lowry [27] test which they used to measure proteins [28].
12
-------
TABLE 5. TOTAL AND SOLUBLE ORGANICS IN SECONDARY EFFLUENT
Organic Nitrogen
Total
mg/1
3.1+
-
4.2+
6.1
2.6+
3.8+
18.0
(soluble)
1.2
(soluble)
Percent
Soluble
32+
-
26+
-
_
_
_
Organic Carbon
Total
mg/1
26.9
-
-
-
_
_
_
-
Percent
Soluble
52
69
-
-
_
_
_
-
Chemical Oxygen
Demand
Total
mg/1
_
-
81.8
-
29.9
45
185
(soluble)
48,9
(soluble)
Percent
Soluble
_
74
78
-
82
_
_
Volatile Solids
Total
mg/1
—
62
98.3
-
_
_
_
-
Percent
Soluble
_
67
87
-
_
_
_
-
Reference
Painter et al. [14]
Rickert and Hunter [16]
Helfgott et al. [13]
Beckman et al. [19]
Adrian and Hodges [20]
Esmond and Wolf [21]
Rebhun and Manka [22]
Bunch et al. [23]
*
Trickling filter effluent (others are activated-sludge effluents) .
Nitrified effluents; nitrates interfere in the organic nitrogen analysis.
A* ,
Average of 7 trickling-filter and activated-sludge samples (range of organic nitrogen was 0.34-1.82).
-------
TABLE 6. CHARACTERIZATION OF SOLUBLE SECONDARY EFFLUENT ORGANICS
Reference:
Number of Samples:
Secondary Treatment:
Sample Treatment:
Bunch et al. [23]
7
Trickling filter or
activated sludge
Whatman #5 filter,
vacuum concentrate 20X
Rebhun and Manka [22]
3
High-rate trickling
filter
Centrifuge, vacuum
concentration 20X,
0.45 y millipore
Manka et al. [26]
11
High-rate trickling
filter, stabilization
pond, extended-aeration,
activated sludge
Centrifuge, vacuum
concentration 20X,
0.45 y millipore
Soluble COD, mg/1
Ether extractables
Proteins
Carbohydrates
Tannins and Lignins
ABS
Humic substances
24-86
< 10%
< 10%
< 5%
< 5%
- 10%
65%
185 (typical)
8%
22%
12%
2%
14%
40-55%
105-167
10-20%
19-25%
4-8%
1-2%
11-21%
31-51%
All concentrations reported as equivalent COD.
-------
ORGANIC NITROGEN IN SOILS, SEDIMENTS, SURFACE WATERS, AND THE MARINE
ENVIRONMENT
The effluent from wastewater treatment plants is usually discharged into
a stream, lake, or ocean, and organic nitrogen is thus released into the en-
vironment, where it can accumulate in sediments, enter water supplies, or
eventually reach the ocean. Some of the organic nitrogen in soils may be sim-
ilar to that in secondary effluents, due to the intense activity of microorga-
nisms in the soil. While there are undoubtedly differences in the origin of
the organic nitrogen from these various sources, it is quite possible that
similarities exist in the final refractory compounds and that they share a
common ultimate fate.
Siegel and Degens [33] found dissolved amino compounds in sea water in a
combined state, and postulated that a sizable portion existed in complexes of
the phenol-quinone type.
Minear [34] concluded that a sizable percentage of the high-molecular
weight material which he isolated from lake water appeared to be deoxyribonu-
cleic acid or its fragments.
Bremner [35] and Keeney [11] reported that about 98% of the nitrogen in
soils and sediments is present as organic nitrogen. Keeney [11J studied the
distribution of amino acids and hexosamines in lake sediments. Bremner [35,
36] found that a minimum of one-third of the organic nitrogen in soils was in
the form of "protein-like combinations," and about 3-10% was present as amino
sugars. Bremner also found that 20-60% of soil organic nitrogen was not dis-
solved by acid hydrolysis (possibly part of this fraction was composed of
heterocyclic nitrogen), but the major fraction of the dissolved organic nitro-
gen was comprised of amino acids [36]. Other compounds found in soils include
amines, purine bases, pyridine and pyrimidine derivatives, and heterocyclic
compounds [35].
Stevenson and Goh [37], Otsuki and Hanya [38], and Ishiwatari [39] have
found indications of the presence of proteins in humic substances from a num-
ber of lake sediments and soils.
CHEMICAL NATURE OF ORGANIC NITROGEN IN SECONDARY EFFLUENTS
A large portion of the organic nitrogen in secondary effluent remains un-
classified. Some organic nitrogen has been classified as protein by means of
the Lowry test [22,26,27], by acid hydrolysis to yield amino acids, or by
multiplying non-dialyzable Kjeldahl organic nitrogen by a factor of 6.25 [23].
None of these methods provides conclusive evidence that the material is ac-
tually protein. Furthermore, a large fraction of the organic nitrogen does
not seem to be present in the form of combined amino acids, although some of
this nitrogen may be present in heterocyclic compounds. It is quite conceiv-
able that a number of chemical reactions or complex formations may have al-
tered the nature of the organic nitrogen considerably from what it was in the
raw wastewater. A number of investigators have postulated or observed chemi-
cal or physical phenomena which are of interest.
15
-------
The Browning (or Maillard) reaction between amino acids and carbohydrates
to form brown unsaturated polymers and co-polymers of nitrogen has been stu-
died extensively by Hodge [40] and discussed by Bremner [35], Hanson and Lee
[6], and Stevenson and Tilo [41] in their studies of organic nitrogen in
soils, domestic wastewater, and sediments, respectively.
Dugan [42] studied the extracellular polymers produced by floe forming
bacteria, and concluded that they were water soluble polysaccharides. The
polymers were shown to be capable of adsorbing dissolved organics, such as
amino acids. Dean [12] found substantial quantities of non-protein nitrogen
present in the amino sugars of cell walls and bacterial slime. Baier [43]
suggested that amino acids, amines, proteins, and polypeptides may be effec-
tive as coupling agents, promoting adhesion by chemical or physical processes
in aqueous solution.
Putnam and Neurath [44], using electrophoretic techniques, observed com-
plex formation between proteins and detergents on both the acid and the alka-
line side of the isoelectric points of the proteins.
Secondary treatment can conceivably produce a myriad of chemical reac-
tions, many of them biologically mediated, in the presence of numerous ions,
ligands, surfactants, and functional groups. The result is the production of
amorphous humic substances which defy classification by traditional methods.
16
-------
SECTION 5
MATERIALS AND METHODS
PHASES OF THE STUDY
This study on soluble organic nitrogen characteristics and removal was
conducted in four phases. The first was concerned with the variation in con-
centration of SON in secondary effluents from municipal wastewater treatment
plants. The second was an evaluation of various physical and chemical proces-
ses, individually and in combination, for the removal of SON. The third phase
evaluated the biodegradability of SON-containing organics in primary and sec-
ondary municipal wastewater effluents, and the extent of production of SON
during biological treatment. The last phase represented an attempt to charac-
terize the chemical characteristics of soluble organic nitrogen. Different
experimental methods were used for each phase and are described under each
phase of the study. However, many of the analytical procedures used were
similar for the different phases and are described below.
ANALYTICAL METHODS
Standard Methods
The following analyses were carried out as described in Standard Methods
[3]: organic nitrogen (except as noted later); chemical oxygen demand (COD),
using one-tenth strength potassium dichromate and ferrous ammonium sulfate;
total organic carbon (TOC), using a Beckman organic carbon analyzer; five-day
biochemical oxygen demand (Wn$); ammonia (distillation with Nesslerization);
nitrate (Cadmium Reduction method) and nitrite (except as noted later); pH,
using a Beckman Model 1009 Electromate pH meter equipped with a Beckman low
sodium-error electrode; alkalinity, by titration to pH 4.0; total suspended
solids (TSS); volatile suspended solids (VSS); mixed-liquor suspended solids
(MLSS); mixed-liquor volatile suspended solids (MLVSS); residual chlorine
(lodometric method); methylene-blue active substances (MBAS); hydroxylated
aromatics (Tannin and Lignin), using Hach TanniVer III reagent in place of
tannin-lignin reagent; and turbidity, using a Hach Model 2100A Turbidimeter
which gave a reading of 0.2 nephelometric turbidity units (NTU) with deionized
water. Residual ozone was measured using the iodometric method for residual
chlorine.
SON, soluble COD (SCOD), soluble 6005, and soluble TOC were measured us-
ing the same procedures as for whole samples. Analysis for these soluble
constituents was performed on samples which had been filtered through a 0.45
micron membrane filter, either Millipore HAWP or Pall DFA 3001 AXA. These
17
-------
two filters were found to produce roughly equivalent values of SON and SCOD on
samples of primary and secondary effluent. Most filtered samples were first
passed through a glass fiber filter (Reeve Angel 934AH) to prevent rapid clog-
ging of the membrane filters.
The water used for all reagents, blanks, dilutions, and rinsing of glass-
ware was deionized tap water. All chemicals used equaled or exceeded the
grades specified by Standard Methods [3]. All glassware was cleaned with a
dichromate-sulfuric acid cleaning solution. Colorimetric analyses were per-
formed with a Bausch and Lomb Spectronic 70 Spectrophotometer.
Nitrate was found to interfere in the SON analysis at concentrations
above 10 mg/1. Thus, nitrified samples could not be accurately analyzed for
their SON content. Nitrite was found not to interfere in the analysis.
The precision of the SON and SCOD analyses is of great importance in
evaluating the experimental data, since these were the two most important
constituents investigated in the majority of the experiments. The standard
deviations of SON and SCOD analyses run on replicate samples are presented, in
Table 7. Calculation of the combined standard deviation is shown in Appendix
A. Also contained in Appendix A are calculations of the confidence levels
(CI) of the SON and SCOD analyses, the estimated error for both analyses, and
a discussion of significance testing.
To be considered significant a difference between two analyses must ex-
ceed the value shown in Table 8. The term "significant" will be applied to
differences which have a 95% CI. As discussed in Appendix A, the significance
of differences in percent removal is dependent upon initial concentration.
Differences which are significant at different confidence levels will be
stated where appropriate.
TABLE 7. PRECISION OF SON AND SCOD ANALYSES
Type of Sample
Secondary Effluent
Secondary Effluent
Act. Carbon Effl.
Act. Carbon Effl.
Secondary Effluent
Act. Carbon Effl.
Anal-
ysis
SON
SON
SON
SON
SCOD
SCOD
Number of
Replicates
6
9
4
8
8
8
Average
Value
mg/1
0.95
1.08
0.19
0.37
23.5
4.8
Standard
Deviation
mg/1
0.03
0.03
0.03
0.03
0.5
0.3
Combined
Standard
Deviation
mg/1
0.03
0.4
18
-------
TABLE 8. SIGNIFICANT DIFFERENCES IN SON AND SCOD ANALYSES
Analyses
SON
SCOD
Differences to be exceeded, mg/1
95% Confidence
0.08
0.9
99% Confidence
0.11
1.3
The recovery of a standard (valine) concentration of SON (1.0 mg/1) in
Palo Alto Secondary Municipal Effluent (PASE) was checked and found to be 99%.
showing that the constituents of PASE do not appear to interfere in the anal-
ysis. An initial step in the SON analysis is removal through distillation of
ammonia, which otherwise would create a positive interference. To determine
if pH 7.4 was high enough during distillation to ensure complete ammonia re-
moval, a series of samples was distilled at various pH values as shown in
Table 9. The pH of the first distillation had no significant effect on the
value obtained for PASE samples.
TABLE 9. EFFECT OF INITIAL AMMONIA DISTILLATION pH ON THE OVERALL SON ANALYSIS
Sample
1
2
3
4
5
pH of
Distillation
7.4
9.0
10.0
11.0
12.0
SON,
mg/1
1.28
1.28
1.26
1.25
1.25
1
Methods for the Technicon AutoAnalyzer
SON and ammonia values were on occasion determined with the aid of a
Technicon AutoAnalyzer (Technicon Instruments Corporation, Terrytown, New
York) capable of analyzing a large number of samples semi-automatically. The
same instrument was frequently used to determine nitrite and nitrate.
SON was determined by distilling and digesting the samples in a Technicon
Block Digester (Technicon Instruments Corp.), and analyzing the ammonia in the
digestate. Samples were placed in digestion tubes containing a phosphate buf-
fer (Ammonia Determination: Standard Methods [3]) and heated for approximate-
ly 1.5 hours at 160°C to distill off the ammonia. Digestion reagent (Organic
Nitrogen Determination: Standard Methods [3]) was then added, and the samples
19
-------
digested for one hour at 370°C. After cooling, the digestate was diluted to a
known volume with deionized water, and the ammonia nitrogen concentration was
measured as described below. Four replicates of a 1.00 mg/1 SON standard gave
results of 1.00 ± 0.03 mg/1 when measured by this technique. When 20 mg/1 of
ammonia nitrogen was added to the SON standard, recovery of SON was 1.03 ±
0.10 mg/1, indicating no interference from ammonia nitrogen.
Ammonia nitrogen was determined by measuring the intensity of the "emer-
ald green" color developed by the reaction between ammonia, sodium salicylate,
sodium nitroprusside, and sodium hypochlorite at a pH of 12.8-13.0 (Industrial
Method No. 375-74W, Technical Instruments Corp.). Five replicates of a 1.00
mg/1 ammonia nitrogen standard gave results of 1.00 ± 0.02 mg/1 when measured
by this method.
Nitrite was determined by reaction with sulfanilamide under acidic condi-
tions to form a diazo compound which was then coupled with N-1-naphtylethyl-
enediamine dihydrochloride to form a reddish-purple azo dye measured colori-
tnetrically (Industrial Method No. 100-70W, Technicon Instruments Corp.).
Nitrate was determined by reducing the nitrate to nitrite by means of a
miniature cadmium reduction column packed with a cadmium-mercury amalgam as
described in Standard Methods [3]. The nitrite was then measured by the
method just described.
Other Methods
Ozone dosage was determined by diffusing the gas stream through 250 ml of
2% potassium iodide in a glass column. The contents were then emptied into a
flask, acidified with 2 ml of concentrated sulfuric acid, and the iodine was
titrated with IN sodium thiosulfate, using a starch indicator to determine the
endpoint.
Proteins were generally measured using the technique developed by Lowry
et al. [46]. Carbohydrates were determined using the anthrone procedure as
described by Pfeffer [47] for non-cellulose carbohydrates.
Sample Preservation
Most of the samples in this study were analyzed immediately following the
experiment in which they were collected; however, at times it was necessary to
store samples for a short period of time. Such samples were refrigerated at
4°C (Standard Methods [3]), and were found to undergo no changes in SON or
SCOD concentrations at that temperature. Hunter and Heukelekian [15] found a
temperature of 4°C to be effective in preventing losses of organic matter in
filtered raw sewage, which is much less stable than secondary effluent. (The
only exception to this procedure was that some of the samples collected during
the first phase of the study were preserved with mercuric chloride and stored
on ice, but this practice was subsequently discontinued for reasons described
in Section 6.)
20
-------
SECTION 6
SON IN MUNICIPAL SECONDARY EFFLUENTS
INTRODUCTION
This section summarizes the characteristics of wastewaters after various
stages of treatment for four different municipal wastewater-treatment plants.
Raw and primary wastewaters were analyzed to evaluate waste strength and to
provide information on the efficiency of secondary treatment in removing SON.
The diurnal variation of SON was determined after various stages of treatment
to provide data for the design of processes for its removal. The organic and
inorganic characteristics of Palo Alto secondary effluent (PASE) were examined
in greater detail, since this effluent was used for the laboratory experi-
ments .
DESCRIPTION OF THE SAMPLING SITES
Samples were collected from four activated-sludge plants, each of which
differed from the others in size and waste characteristics. These plants are:
(1) Palo Alto Regional Water Quality Control Plant; (2) San Jose/Santa Clara
Water Pollution Control Plant; (3) Union Sanitary District Plant #3; and
(4) South Tahoe Public Utility District Water Reclamation Plant.
The Palo Alto Regional Water Quality Control Plant is a typical complete-
mix activated-sludge plant located in Palo Alto next to the San Francisco Bay.
It serves the communities of Palo Alto, Mountain View, and Los Altos, and is
designed to treat an average dry-weather flow of 1.53 m^/s (35 mgd). The
plant receives a significant amount of industrial waste, and the waste re-
ceived at the plant is quite "weak" in terms of BOD and COD.
The San Jose/Santa Clara Water Pollution Control Plant is located approx-
imately 9 km north of San Jose's central business area on a site encompassing
approximately 40 x 10^ m^. The 8.3 x 10** m^ service area tributary to the
plant includes the cities of San Jose and Santa Clara, and the major portion
of the Santa Clara Valley extending southward 40 km from the San Francisco
Bay. The plant is designed to treat an average dry-weather flow of 7.0 m^/s
(160 mgd) and at the time the samples were taken was being operated as a con-
ventional activated-sludge plant. (During canning season, the Kraus Process
is employed.)
Union Sanitary District Plant #3 is located in Union City, California.
The plant is designed for a flow of 0.13 m3/s (3 mgd) and receives both
21
-------
domestic and industrial waste. The influent wastewater is screened and then
pumped into a vacuator, which provides the equivalent of primary treatment by
employing both flotation and settling. Next are four aeration tanks which
during this study were operated in parallel as in the conventional activated-
sludge process. Activated sludge is wasted to the influent wet well, which
can cause certain parameters to be quite high in the primary effluent. During
normal operation, supernatant liquor from digested-sludge storage lagoons is
recycled back to the primary settling tanks. During sampling for this study,
however, the supernatant liquor was not being returned.
The South Tahoe Public Utility District Water Reclamation Plant is a 0.33
m3/s (7.5 mgd) advanced wastewater-treatment plant located just south of Lake
Tahoe in California. The raw wastewater, comprised almost entirely of domes-
tic sewage, receives primary and conventional activated-sludge treatment prior
to advanced waste treatment consisting of lime clarification, ammonia removal,
recarbonation, chlorination, filtration, and activated-carbon adsorption.
During sampling for this study, the ammonia stripping process was not in oper-
ation.
SAMPLE COLLECTION AND STORAGE PROCEDURES
At each plant except Tahoe, samples were taken of the raw influent waste-
water and the primary, secondary, and final (chlorinated) effluents. At the
Tahoe plant samples were taken of the raw influent wastewater, secondary ef-
fluent, lime-treated effluent, chlorinated effluent, filtered effluent, and
activated-carbon effluent. Grab samples were taken from each location every
two hours for twenty-four hours. A portion of each sample was filtered
through a glass fiber filter (Reeve Angel 934AH) and both filtered and unfil-
tered portions were preserved on ice. Composite samples were prepared by mix-
ing the grab samples in volumes proportionate to flow, the remaining portions
of the grab samples being stored and analyzed individually for measuring diur-
nal variations in characteristics. After the samples reached the laboratory,
the filtered samples were passed through a 0.45 micron filter (Millipore HAWP
or Pall DFA 3001 AXA) , the filtrate being saved for analysis of the soluble
constituents.
Standard Methods [3] recommends that samples stored for nitrogen analyses
be cooled to just above freezing, and that acid or mercuric chloride be added
to the sample. EPA [48] recommends mercuric chloride plus cooling to 4°C for
preservation of nitrogen forms. Howe and Holley [49] found both mercuric
chloride and sulfuric acid to be effective in preserving the various forms of
nitrogen.
Both acid and mercuric chloride were tested as preservatives for SON in
raw sewage, and both were found to be equally acceptable. It was decided not
to use acid because (1) it would not permit measurement of pH, alkalinity,
or nitrites, and (2) it might cause a significant shift in the relative
amount of soluble and particulate matter in the sample.
Prior to April 24, 1975 the samples (except for portions stored sepa-
rately on ice only for BOD5 analysis) were preserved with 40 mg/1 of mercuric
22
-------
chloride, In addition to being cooled with ice. However, it was noted that
the addition of mercuric chloride to a wastewater sample which had been fil-
tered through a glass fiber filter and stored on ice caused the sample to be-
come cloudy; and within a day after it had been added a precipitate became
visible. If the sample was then filtered through a 0.45 micron filter, a sig-
nificant loss of organic matter resulted. For this reason the addition of mer-
curic chloride was stopped and after April 24, 1975 only ice was used as a pre-
servative. Hunter and Heukelekian [15] studied the preservation of filtrates
of raw sewage and found that no losses of organic matter occurred at 4°C.
GENERAL CHARACTERISTICS OF THE INFLUENT AND PRIMARY EFFLUENT WASTEWATERS
The general characteristics of the influent and primary effluent waste-
water, as determined from analysis of the 24-hour composite samples taken from
each of the four plants, are presented in Table 10. The plants sampled ranged
TABLE 10. GENERAL CHARACTERISTICS OF THE INFLUENT AND
PRIMARY EFFLUENT WASTEWATERSt
Parameter
SON
Ammonia-N
Organic-N
Nitrate-N
Nitrite-N
Total
Nitrogen
COD
SCOD
Soluble TOC
BOD5
Soluble BOD5
pH, units
Alkalinity
TSS
VSS
VSS, percent
Avg . Flow ,
m3/s
Palo Alto*
Influent
5.2
28.7
12.2
1.8
0.38
43.1
381
94
37
171
49
7.6
209
172
143
82
Primary
Effluent
4.4
28.4
8.3
1.3
0.40
38.4
207
86
29
125
42
7.7
209
72.5
60.5
81
1.17
San Jose/
Santa Clara*
Influent
5.5
26.7
20.0
0.0
0.03
46.7
761
234
84
306
148
7.3
284
349
267
78.5
Primary
Effluent
4.8
26.7
13.4
0.2
0.08
40.4
451
220
81
240
143
7.6
287
109
85
80
3.7
*
Union City
Influent
5.2
19.9
12.2
0.2
0.21
32.5
570
128
53
212
82
7.6
323
205
166
82
Primary
Effluent
4.2
20.8
18.9
0.0
0.07
39.8
539
107
46
203
68
7.4
328
258
220
85
0.16
Tahoe
Influent
2.9
23.9
9.3
0.2
0.02
33.4
388
123
-
186
64
160
144
138
96
0.18
Values represent mg/1 unless otherwise indicated.
Values are averages of two samples taken on different days.
23
-------
in flow from 0.16 to 3.7 m3/s (3.6 to 80 mgd), in raw wastewater SON from 2.9
to 5.5 mg/1, and in waste strength from a COD of 381 to 761 mg/1.
GENERAL CHARACTERISTICS OF SECONDARY EFFLUENT
Secondary effluent is used to designate both activated-sludge effluent
and chlorinated effluent. The general characteristics of the secondary efflu-
ents, as determined from analysis of the 24-hour composite samples taken from
each plant, are presented in Table 11. Although these plants differed in
size, raw-wastewater characteristics, loading, temperature, and MLSS, the
TABLE 11. GENERAL CHARACTERISTICS OF THE SECONDARY EFFLUENTS
t
Parameter
SON
Ammonia-N
Organic-N
Nitrate-N
Nitrite-N
Total
Nitrogen
COD
SCOD
Soluble TOC
BOD5
Soluble BOD5
pH, units
Alkalinity
TSS
VSS
VSS, percent
A
Palo Alto
Activ.-
Sludge
Effluent
1.65
25.7
3.3
0.2
0.07
29.3
53
27
9
11
3
7.4
211
22
20
88
Chlori-
nated
Effluent
1.78
25.0
3.2
0.2
0.03
28.4
52
30
10
-
-
7.4
200
19
17
89
San Jose/
Santa Clara*
Activ.-
Sludge
Effluent
1.64
21.4
3.4
1.5
1.40
27.7
59
36
15
9
2.8
7.8
268
19
17
92
Chlori-
nated
Effluent
1.79
21.8
3.3
1.7
0.45
27.2
57
39
17
-
-
7.7
261
16
15
93
Activated-Sludge System Characteristics:
Avg. Flow, m3/s 1.17
Temperature, °C 19.5
MLSS 1510
MLVSS 1188 (78%)
Loading, kgBODy
kg MLVSS/day >JO
3.7
22.8
3350
2415 (72%)
0.51
*
Union City
Activ.-
Sludge
Effluent
1.55
19.2
4.2
0.0
0.01
23.4
111
44
15
20
5
7.5
317
34
27
82
Chlori-
nated
Effluent
1.51
18.5
4.6
0.1
0.01
23.2
118
41
16
_
-
7.5
300
36
28
80
0-16
20.5
3700
2930 (79%)
0.37
Tahoe
Activ.-
Sludge
Effluent
1.17
21.4
4.2
0.1
0.31
26.0
84
25
10
41
2
7.5
166
56
48
86
0.18
16.5
2170
1840 (85%)
0.45
Values represent mg/1 unless otherwise indicated.
*
Values are averages of two 24-hour composite samples taken on different days.
24
-------
general characteristics of the effluents were similar, with the major differ-
ences apparently being due to the relative efficiencies of the secondary clar-
ifiers. The chlorinated effluents did not differ significantly from the
non-chlorinated effluents.
CONCENTRATION OF SON
The concentration of SON found in each of the filtered composite samples
is shown in Table 12. The raw wastewaters ranged from 2.9 to 6.3 mg/1 SON
with an average concentration of 4.9 mg/1. The SON of the primary effluents
averaged 4.5 mg/1, reflecting an average 15% removal. This removal may have
been due to biological oxidation, adsorption onto solids, or a combination of
both. The available data provide no information regarding these mechanisms.
The concentration of SON in secondary effluent averaged 1.5 mg/1, ranging
from 1.2 to 2.1 mg/1 and reflecting a 69% decrease through the overall treat-
ment process. The concentration of SON in the chlorinated effluents was not
significantly different from that of the non-chlorinated effluents.
The concentration of SON in the secondary effluent remained within the
rather narrow range of 1-2 mg/1 despite differences in raw wastewater strength,
TABLE 12. SUMMARY OF SON CONCENTRATIONS AND PERCENT REMOVAL
AT SECONDARY TREATMENT PLANTS
Loca-
tion
en
c1*
o >,
•H 4J
C -rl
P 0
o o
rH W
(8 rH
PH <3
« 2
^
a •
-------
waste origin, flow rate, mixed liquor suspended solids (MLSS), and loading at
the several plants. No trends have been found in the composite sample data
which might indicate the importance of a particular parameter in the removal
of SON by secondary treatment. An estimate of the relative proportion of ni-
trogen contained in the soluble organics can be obtained by taking the ratio
of SON to SCOD. SON to SCOD ratios varied widely among the treatment plants
sampled, seemingly a function of raw waste characteristics, but were quite
constant at an individual plant, as shown in Table 12.
DIURNAL VARIATION IN SON
At three of the treatment plants sampled (Palo Alto, San Jose/Santa Clara
and Union City) grab samples were taken every two hours and analyzed for SON,
SCOD, and ammonia. Figures 1 through 3 show the diurnal variations in SON
after various processes in the three treatment plants. Significant differ-
ences exist in the patterns in the influent for the three plants, but no ex-
planation for the differences can be given. Ratios of maximum to minimum
influent SON varied from a high of 4.5 at the Union City plant to a low of 2.0
at Palo Alto, and ratios of maximum to average SON varied from 1.9 at the San
Jose/Santa Clara plant to 1.2 at the Palo Alto plant. Variations in effluent
SON were much smaller, maximum to minimum values ranging from 1.7 to 1.3, and
maximum to average ratios ranging from 1.3 to 1.1.
In all cases there was a close relationship between SON and SCOD. Varia-
tions in ammonia and SON were sometimes similar, but not as similar as SON and
SCOD.
An interesting pattern was noted in the SON data from the Union City
plant. The influent SON concentration appeared to be reflected 5 to 6 hours
later in the effluent as shown in Figure 4, which is taken from Figure 3.
This pattern may indicate that a certain portion of the influent SON resisted
treatment.
PALO ALTO SECONDARY EFFLUENT
A more detailed study of the organic and inorganic constituents of Palo
Alto Secondary Effluent (PASE) was made, since this effluent was used"for
each of the experiments to determine the potential of various processes for
removal of SON. Table 13 presents the results of a six-month study of the
general characteristics of composite samples of PASE.
A sample of PASE was taken on February 6, 1976, filtered, and analyzed
for various organic constituents for comparison with the results of other in-
vestigators. The sample had an SON concentration of 1.67 mg/1 and an SCOD
concentration of 28 mg/1. The concentration of MBAS was 0.08 mg/1 as linear
alkylate sulfonate (LAS); carbohydrates were 1.7 mg/1 as glucose; and hydroxy-
lated aromatics were 1.3 mg/1 as tannic acid. MBAS was present in much smal-
ler amounts than reported for other secondary effluents [22,23,26], while
carbohydrate and hydroxylated aromatics concentrations were typical.
26
-------
o>
E
OPLANT INFLUENT
E PRIMARY EFFLUENT
A SECONDARY EFFLUENT
O FINAL EFFLUENT
J I
I I I I
I I I I I I I I
I I I I
SAM
2PM 8PM
TIME OF DAY
2AM
Figure 1. Diurnal variation of SON — Palo Alto (4-14-75).
27
-------
12
10
8
o>
E
O PLANT INFLUENT
E PRIMARY EFFLUENT
A SECONDARY EFFLUENT
O FINAL EFFLUENT
'—'—i—'—i—'—'—'
SAM 2PM 8PM 2AM
TIME OF DAY
Figure 2. Diurnal variation of SON — San Jose/Santa Clara (5-13-75)
28
-------
O PLANT INFLUENT
0 PRIMARY EFFLUENT
SECONDARY EFFLUENT
O FINAL EFFLUENT
SAM
2PM 8PM
TIME OF DAY
2AM
Figure 3. Diurnal variation of SON — Union City #3 (5-1-75).
29
-------
8
1 r 1 1
o PRIMARY EFFLUENT
D SECONDARY EFFLUENT
0800 1200 16002000240004000800
TIME OF DAY
Figure 4. Diurnal variation of SON at Union City #3 (5-1-75),
(taken from Figure 3).
30
-------
TABLE 13. GENERAL CHARACTERISTICS OF PALO ALTO SECONDARY EFFLUENT
Parameter
SON, mg/1
SCOD, .mg/1
Soluble BOD mg/1
Organic-N, mg/1
COD, mg/lT+
BOD5, mg/1
TOC, mg/1'
PH f
Suspended solids, mg/1
Nitrate-N, mg/1
Nitrite-N, mg/1
NH3-N, mg/1
Mean
Value
1.29
28
1.9
3.0
53
12
22
7.6**
16
0.4
0.3
24
Standard
Deviation
0.24
4.8
1.6
0.6
41
5.3
7.2
-
5
0.7
0.6
4.8
Minimum
Value
0.84
20
0.2
1.4
25
6
13
7.4
7
0.0
0.0
17
Maximum
Value
1.66
39
5.0
3.8
166
28
36
8.1
28
3.1
2.4
35
Number of
Samples
20
20
16
20
17
18
19
20
19
20
20
20
Based upon 24-hour composite samples taken between June 11, 1975 and
December 13, 1975.
Values obtained from Palo Alto Regional Water Quality Control Plant
Records.
Median.
The protein analysis was of particular interest, since (1) protein is a
possible constituent of SON; (2) various investigators have reported its pres-
ence in secondary effluent [22,23,26]; and (3) it is generally considered to
be highly biodegradable. The PASE sample described above was analyzed for
protein using the technique of Lowry et al. [27], using ovalbumin and gelatin
protein standards, and the protein concentration was found to be 8.0 mg/1 by
this method. The technique does not measure nucleic acid bases or urea, and
amino acids give much less color than proteins; however, most phenols will
interfere in the analysis. Close inspection of the Lowry method shows that it
is virtually the same analysis as that for hydroxylated aromatics, the impor-
tant exception being the presence of the copper catalyst. The copper has been
demonstrated to have little effect on the color produced with tyrosine and
tryptophan [27], which are hydroxylated aromatic compounds. When the copper
catalyst was omitted from the analysis, the same concentration of "protein"
in PASE was still measured by the test. This indicates that the color was
produced by hydroxylated aromatics rather than protein, and that the concen-
tration of protein was in fact negligible. It is believed that the sample did
not contain sufficient copper to serve as a catalyst.
DeWalle and Chian [28] were probably correct in stating that fulvic acids
interfere in the Lowry test by virtue of their numerous phenolic hydroxyl
groups. The Lowry method is an adaptation of an earlier method of Folin and
Ciocalteu [52] for measuring tyrosine and tryptophan, through the reaction of
31
-------
their phenolic groups with the Folin phenol reagent used in both the Lowry
test and the test for hydroxylated aromatics.
SUMMARY
1. The concentration of SON in the secondary effluents of four activated-
sludge treatment plants ranged from 1.1 to 2.1 mg/1, with an average
of 1.5 mg/1.
2. The concentration of SON in the secondary effluent of one treatment
plant ranged from 0.9 to 1.7 mg/1, with an average of 1.3 mg/1 over
a six-month period.
3. Diurnal variations in the concentration of SON in the secondary ef-
fluents of three treatment plants were small, with maximum to mini-
mum ratios ranging from 1.7 to 1.3, and maximum to average ratios
ranging from 1.3 to 1.1 mg/1.
4. Variations in SON and SCOD were very similar in all cases.
5. Data from one treatment plant indicate that the effluent concentra-
tion of SON may be a function of the influent concentration; however,
there was no other indication of the influence of any one factor on
effluent SON concentration.
6. Little or no protein was found in a filtered sample of Palo Alto
secondary effluent.
32
-------
SECTION 7
SON REMOVAL BY PHYSICAL AND CHEMICAL PROCESSES
REMOVAL MECHANISMS
Removal of SON from municipal secondary effluent can potentially be ac-
complished through two general mechanisms: (1) physical removal mechanisms,
resulting in physical removal of SON from solution; and (2) oxidative removal
mechanisms, resulting in oxidation of the nitrogen atom or of the organic
carbon, resulting in the release of ammonia.
As shown in Table 14, three mechanisms for physical removal of SON were
evaluated in this study: (1) adsorption, (2) flocculation, and (3) precipi-
tation. Adsorption is defined as the accumulation of a solute at the solid-
liquid interface, and will be used here to include ion exchange and co-
precipitation.
Flocculation is the aggregation of minute particles, whose surface
charges have been sufficiently neutralized, into large floes which can gener-
ally be removed from solution by sedimentation. Should any of the SON be in
TABLE 14. FACTORS AFFECTING REMOVAL AND REMOVAL MECHANISMS
FACTORS AFFECTING REMOVAL:
A. Charge (electrostatic interactions).
B. Polarity (dipole-dipole interactions, dipole-induced dipole inter-
actions, van der Waals forces, pi-bond interactions, hydrogen
bonding, solvation)
C. Special factors (e.g., covalent bonds, pore size, entrapment)
D. Chemical structure and reactivity.
REMOVAL MECHANISMS:
1. Adsorption (includes ion exchange and co-precipitation) —
(Factors A, B, and C)
2. Flocculation — (Factors A, B, and C)
3. Precipitation — (Factor C)
4. N-oxidation — (Factor D)
5. Deamination"— (Factor D)
33
-------
colloid form, it could be removed in this manner. Truly soluble SON can also
be removed by flocculation if sufficient quantities of the flocculant adsorb
onto an SON molecule to enable it to behave as a small colloid and flocculate
with other particles.
Precipitation occurs in the special circumstance in which a covalent bond
is formed between the surface and the solute or between an ion and the solute,
such that the solute is rendered insoluble.
There are two mechanisms by which oxidation can occur: (1) N-oxidation,
in which the nitrogen atom is oxidized to a higher oxidation state and is no
longer measured by the Kjeldahl organic nitrogen analysis; and (2) C-oxida-
tion, following which the nitrogen is released as ammonia. Removal of SON by
oxidation may potentially be accomplished by either chemical or biological
oxidation of the nitrogen atom. In either case, the nitrogen atom need not be
"physically" removed from solution, but is chemically altered to either inor-
ganic nitrogen or to a higher oxidation state of organically-bound nitrogen,
such that it is no longer measured by the Kjeldahl organic nitrogen analysis.
Nitrogen thus oxidized will be termed "removed," even though the nitrogen atom
is still present in solution. The ultimate removal, effects, and characteris-
tics of the chemically altered compounds produced by oxidation were beyond the
scope of this study.
There are a number of operating parameters such as pH, surface area, con-
tact time, and temperature, which affect the overall efficiency of a treatment
process. These parameters will be discussed along with each process. The
following discussion will focus on the specific factors which are important in
determining whether a particular molecule can potentially be removed by a
particular removal mechanism.
The factors affecting removal have been arbitrarily grouped into four
categories: (1) charge, (2) polarity, (3) chemical structure and reactivity,
and (4) special factors. Charge affects the physical removal of a solute
through attractive and repulsive electrostatic interactions. A molecule bear-
ing the same charge as a surface has less chance of being adsorbed to that
surface than a molecule of the opposite charge. However, adsorption of a mol-
ecule onto a surface of like charge is possible, provided that other factors
affecting removal are sufficiently favorable.
Polarity is the term under which the following factors will be grouped:
dipole-dipole interactions, dipole-induced dipole interactions, van der Waals
forces, pi-bond interactions, hydrogen bonding, and solvation. All, with the
exception of solvation, are generally considered to favor adsorption, but the
forces involved are usually very small relative to electrostatic forces.
Solvation relates to the energy required to displace water from the surface
and from the hydration sphere surrounding the molecule, enabling the molecule
and the surface to come in contact. Solvation always hinders adsorption.
There are a number of special factors related to the physical removal of
a substrate, such as (1) the formation of a covalent bond between surface
and substrate; (2) the pore size of the surface, which may be too small to
admit a molecule that might otherwise adsorb; or (3) entrapment of a large
34
-------
molecule by floe particles. Such special factors can prevent the removal of a
molecule that might otherwise be removed or cause a molecule to be removed
from solution despite unfavorable conditions.
The primary factors affecting the removal of SON by oxidation are the
chemical structure of the nitrogen-containing compounds and the reactivity of
the various chemical bonds.
There are a number of advanced wastewater-treatment processes which are
potentially capable of removing a significant fraction of SON from municipal
secondary effluent. Those evaluated in this study are listed in Table 15 to-
gether with the removal mechanisms pertinent to each.
Each of the processes considered to have potential for the removal of SON
has been studied extensively, and many models exist for each. However, the
majority of the work has been carried out on model systems employing perhaps a
single homogeneous substrate, a few ions of known concentration, and a well-
characterized solid. Secondary effluent poses a far more complex problem in
TABLE 15. TREATMENT PROCESSES AND REMOVAL MECHANISMS
TREATMENT PROCESSES:
1. Chemical Coagulation — (Mechanisms 1, 2, and 3)
i. Ferric Chloride
ii. Lime
iii. Alum
2. Adsorption — (Mechanisms 1, 2, and 3)
i. Activated Carbon
ii. Ion-Exchange Resins
3. Oxidation — (Mechanisms 4 and 5)
i. Chemical Oxidation
(a) ozone
(b) chlorine
(c) potassium permanganate
(d) hydrogen peroxide
ii. Biological Oxidation
REMOVAL MECHANISMS:
1. Adsorption (includes ion exchange and co-precipitation)
2. Flocculation
3. Precipitation
4. N-Oxidation
5. Deamination
35
-------
terms of (1) the substrate, and (2) the concentration and number of ions
present.
The large number of different compounds included in an analysis for a
general parameter such as SON adds complexity to interpretation of results
from both physical and oxidative removal processes. In terms of the physical
removal processes, a molecule of SON can be positively charged, negatively
charged, uncharged, polar, non-polar, colloidal, soluble, or a combination of
these. It may also possess some special characteristic which allows it to be
removed by virtue of a special factor. In addition, certain organic molecules
present may affect the removal of others by altering the properties of a sur-
face or by complex formation.
In terms of the oxidative processes, the amount of SON existing in each
particular configuration is not known, making it difficult to predict the re-
actions which will occur, the rate of individual reactions, and the end pro-
ducts of the reaction.
The numerous anions and cations present in secondary effluent add another
level of complexity to the problem. The role of each ion in catalyzing or in-
hibiting oxidation is not known, nor is the degree to which the organic mate-
rial complexes with different ions.
Due largely to the complex ionic makeup of secondary effluent, the prop-
erties of a surface introduced into the effluent may be considerably different
from those of the same surface in a simpler system. The thickness and prop-
erties of the electrical double layer surrounding a charged surface are a
function of the ionic characteristics of the surrounding liquid. Ions co-
precipitated during chemical coagulation may alter such properties of the
floes as isoelectric point or sedimentation rate.
Thus, the ionic and chemical nature of secondary effluent presents a
large number of variables, such that it is virtually impossible to examine
each of them individually in a reasonable period of time. In addition, it be-
comes very difficult to distinguish between the various removal mechanisms and
the factors affecting these mechanisms. The complexity of this system neces-
sitates an approach somewhat different than generally used to investigate
simpler systems.
EXPERIMENTAL APPROACH
The experiments for this phase of the study were conducted to determine:
1. The potential of various physical and chemical processes for the
removal of SON and the factors affecting removal.
2. The potential of various combinations of treatment processes for
removal of SON.
36
-------
Removal of SON by Various Treatment Processes
This portion of the study was of use in meeting two of the objectives of
the study: bench-scale studies allowed evaluation of the potential of ad-
vanced waste-treatment processes for SON removal, and investigation into the
various factors affecting removal provided information useful in determining
the nature and characteristics of the SON. The processes studied can be divi-
ded into three categories: (1) chemical coagulation, (2) adsorption, and
(3) chemical oxidation. (Biological oxidation is discussed in Section 8.)
Chemical coagulation was evaluated with ferric chloride, lime, and alum
as coagulants, together with various polyelectrolytes and clays as coagulant
aids. Variables studied were pH, mixing time, and coagulant concentration.
Adsorption of SON was studied using granular activated carbon and ion-
exchange resins as adsorbents and pH and adsorbent dose as variables. (Ad-
sorption is used here in the sense of adsorption "process"; the adsorption
"mechanism" is, of course, applicable to both adsorption processes and chemi-
cal coagulation.)
Chemical oxidation was studied using ozone, chlorine, hydrogen peroxide,
and potassium permanganate as oxidants, and pH and oxidant dose as variables.
Due to the complexity of the systems under investigation, it was consid-
ered unreasonable to attempt to determine the effect of each different ion on
the removal of SON by each process. Analysis of the Palo Alto effluent re-
vealed that the major inorganic constituents of the wastewater remained rela-
tively constant with time [51], and so variations in inorganic contents were
not considered a significant factor affecting noted variations in results.
Removal of SON by Combinations of Processes
It is likely that the SON removed by one treatment process differs in
some way from that removed by another, such that a combination of the two pro-
cesses would be capable of removing a greater amount of SON than either indi-
vidually. A number of bench-scale experiments were carried out on Palo Alto
secondary effluent. The processes selected for these experiments were gener-
ally those which were thought to have the greatest possibility for removing
different fractions of the SON, although at times, combinations of processes
thought to be removing the same fraction were tested to see if this was in
fact the case.
In contrast to the case with individual treatment processes, it was not
possible in these experiments to consider that the inorganic ion concentra-
tions were constant. A number of the treatment processes, such as ion
exchange or chemical coagulation, are capable of radically altering the inor-
ganic characteristics of a secondary effluent. This greatly enhanced the pos-
sibility that one treatment process might hinder or improve removal by a
second process. This necessitated very careful consideration of the possible
effects of inorganic ions on the experimental results.
37
-------
Background
The coagulants commonly used for water treatment and advanced wastewater
treatment include aluminum sulfate (alum), ferric chloride, ferric sulfate,
calcium oxide (lime), and magnesium oxide. In this study, alum, ferric chlo-
ride, and lime were evaluated for their potential to remove SON from PASE.
A number of "coagulant aids" are commercially available which can theo-
retically improve coagulation in various ways, such as bridging floe particles,
neutralizing charge, or providing nuclei for flocculation. Polyelectrolytes,
polymers containing positively and/or negatively charged functional groups;
clay minerals, such as bentonite, koalinite, and montmorillonite; activated
silica; and activated alumina are some of the materials used as coagulant aids.
The properties and behavior of polyelectrolytes have been discussed in detail
by LaMer and Healy [70], Ries and Meyers [71], Black [72], and Stumm and
O'Melia [53]. Several coagulant aids were evaluated in this study for their
potential to improve SON removal.
Chemical coagulation is a process in which the surface charge of colloidal
particles is reduced through adsorption of the coagulating ions or their hy-
drolysis products. The particles may then be flocculated and removed by sedi-
mentation. Under conditions of sufficient dose and proper pH, coagulants will
form a precipitate, which provides centers for flocculation, and which is po-
tentially capable of removing soluble organics by adsorption.
In the treatment of wastewaters, chemical coagulation has been employed
for the removal of suspended solids following secondary effluent and for the
treatment of kraft mill effluents and other industrial wastewaters. Optimiza-
tion for the efficient removal of organic materials has been largely oriented
toward removal of the particulate fraction, with relatively little attention
given to the parameters of possible importance in removing the soluble frac-
tion of the organics.
Parkin and McCarty [73] studied the removal of SON from a treated agri-
cultural wastewater with various coagulants and polyelectrolytes in combina-
tion with bentonite clay. They found that ferric chloride achieved about 67
percent removal of SON, compared with 43 percent for lime, and 35 percent for
alum. Polyelectrolytes achieved only about 16 percent removal at neutral pH,
but removal was improved at pH 3 and 11, possibly due to ionization and pre-
cipitation of SON, respectively.
Malhotra et al. [74] were able to remove 60 percent of the total organic
nitrogen and 55 percent of the COD from a secondary effluent using 250 mg/1 of
alum at pH 5, the average initial organic nitrogen concentration being 4.3
mg/1 and the COD 91 mg/1. Wolf [75] reported an average organic nitrogen re-
moval from a trickling filter effluent of about 30 percent using iron salts
and lime. Rebhun et al. [76] studied the effects of polyelectrolytes and ben-
tonite clay on removal of organic contaminants, and found that bentonite re-
moved about 50 percent of the total organic nitrogen at a dose of 500 mg/1;
however, the initial concentration of organic nitrogen was 14.0 mg/1. In most
cases for which data are available, it appears that organic nitrogen removal
38
-------
by chemical coagulation closely parallels COD removal. The soluble portion of
organic nitrogen and COD removed in these experiments was not indicated.
Assuming some portion of the soluble organics are removable by coagula-
tion, the primary mechanism would be adsorption. The factors which together
determine whether a particular molecule will adsorb can be grouped into three
categories: (1) charge, (2) polarity, and (3) special factors.
Charge has received the greatest amount of attention in the literature
for several reasons: (1) it is the most important factor in the removal of
particulate and colloidal matter; (2) it is relatively well understood; (3) a
number of models exist which deal with the nature and properties of charge and
the parameters affecting it; and (4) there are a number of simple techniques
available for studying charge including electrophoresis, titration, maximum
sedimentation rate determination, and surfactant adsorption [77,78].
When a coagulant is added to water under conditions such that a precipi-
tate forms, the hydrous oxide surface of the precipitate can be positively or
negatively charged, depending on the pH and a number of other factors. The pH
at which the net charge on the surface is zero is referred to as the point of
zero charge or PZC.
The factors controlling the sign and magnitude of the surface charge on
oxides have been discussed in detail by Parks [79,80], Hydrous oxides behave
as ion exchangers and can adsorb ions from solution which may considerably
alter the surface charge. For example, a calcium hydroxide precipitate will
be positively charged at pH 11 in deionized water, but even small amounts of
phosphate or bicarbonate can reverse this charge [81]. In general, anions
will lower the PZC if adsorbed, and cations will shift the PZC toward the PZC
of each individual cation.
The particulate matter found in secondary effluent is negatively charged
[82], and it is generally accepted that the large majority of the soluble con-
taminants are also negatively charged. If true, optimum removal of organics
would seem to call for a positively charged surface. If charge is in fact an
important factor, and this has yet to be established, then conditions which
tend to make the surface charge of a floe more positive should result in bet-
ter removal. If such conditions also make the organics positive, then the
potentially better removal may be canceled.
Polarity is a factor that is not as well understood or often modeled as
charge, and it is somewhat controversial in several respects. The hydrogen-
bonding mechanism can explain the removal of certain molecules from solution,
but water itself is a strong hydrogen-bonding solvent. Solutes can be de-
scribed as either hydrophilic or hydrophobic, but there is not a distinct
boundary between the two. In the case of chemical coagulation, the surfaces
involved are polar oxides; hence it is not likely that polarity plays a major
role in determining adsorption due to relatively small differences in polarity
between surface and solvent.
Special factors are likely to dominate in the adsorption of soluble or-
ganics during chemical coagulation. The definition of soluble used in this
39
-------
study is an operational one, and it is probable that a fraction of the "solu-
ble" organics is in fact comprised of very small colloids. These colloids, as
well as some of the larger molecules in solution, may be removed by entrapment
in the coagulant floe. Specific adsorption and precipitation are also likely
to be of importance.
Variables of importance in the chemical coagulation process are pH, tem-
perature, ionic strength, coagulant dose, and mixing time. Since ionic
strength and temperature cannot be conveniently modified under treatment plant
conditions, they were treated as constants in this study, and pH, mixing time,
and coagulant dose were the variables studied.
The pH at which samples are coagulated is an important variable, since it
will determine the charge of the organics being removed, the nature of the
coagulant metal hydrolysis products, and the surface charge of the resultant
floe particles. Malhotra et al. [74] found the optimum pH for alum coagula-
tion of phosphorus to be 5, and pH 6 to be superior to pH 8 for organic nitro-
gen removal. The optimum pH for removal of negative colloids usually falls in
the range of 5 to 6.5 [83].
Experimental Procedures
The coagulation experiments were carried out to determine (1) the poten-
tial of alum, lime, and ferric chloride for removing SON; (2) the various pa-
rameters affecting removal; (3) the effect of coagulant aids on removal; and
(4) the reasons for any differences in removal among the various coagulants.
Samples of PASE were taken between 11:00 AM and 3:00 PM and brought to
the laboratory for immediate use, and were not filtered unless otherwise indi-
cated. Filtration prior to coagulation was conducted with a Pall filter,
while filtration after coagulation was with a Millipore filter. The majority
of the samples were not filtered prior to coagulation in order to simulate
actual treatment plant conditions, and it was determined that filtration re-
sulted in only small differences in final effluent quality. For those experi-
ments involving activated sludge, the activated sludge was taken from an
aeration tank at the Palo Alto plant and used immediately without prior fil-
tration.
One to two liters of sample were placed in a beaker and stirred at 100
rpm with a 6-place stirrer (Phipps and Bird) while the coagulants and other
reagents were added, all chemicals being at least reagent grade. The ferric
chloride and alum coagulants were added from stock solutions of 100-200 mg/1,
and lime and the clays were added as slurries containing 50-200 mg/1. The
desired pH was obtained by addition of sodium hydroxide (1 to 6 N) . All ex-
periments were carried out at room temperature, approximately 20°C.
The polyelectrolytes were obtained from the Dow Chemical Company in their
solid or most concentrated form and diluted to suitable working solutions.
Those used were a cationic polymer, C-31; an anionic, A-23;and a non-ionic,
N-17.
40
-------
Mixing time was 20 minutes at 25 rpm and settling time was 30 minutes un-
less otherwise noted, after which the supernatant liquid was siphoned off and
analyzed for SON, SCOD, pH, and turbidity.
Control samples underwent the same procedures but with no coagulant added.
Values of SON and SCOD for the control samples were determined on their fil-
trate through a 0.45 micron filter (Millipore).
In the jar tests involving the clay minerals, kaolinite (J. T. Baker,
Kaolin, technical grade) or bentonite (Wyoming, 325 mesh, Wards Scientific) was
added to the sample while stirring at 100 rpm, and after 5 minutes, the coagu-
lant was added in the usual manner.
In experiments to determine the effect of coagulant dose on SON removal,
with alum and ferric chloride, the pH was adjusted to between 5 and 6.5. Fine
adjustments of the pH with these coagulants was hindered by the fact that they
"age" by adsorbing base, which causes a drift in the pH during the course of
the experiment. The pH values recorded were those of the supernatant liquor
after settling.
CHEMICAL COAGULATION
Coagulation of PASE Samples
It was desirable to know if the organics remaining after coagulation were
entirely in the soluble fraction, or if there were still particulate matter in
the samples. Therefore, three samples of secondary effluent were coagulated
with a small dose of each of three coagulants, and both filtered (0.45 micron
Millipore) and unfiltered portions of the supernatant liquid were analyzed for
organic nitrogen and COD. The results (Table 16) show that doses of 200 mg/1
or less of the three coagulants removed virtually 100 percent of the particu-
late organic matter.
Thus, coagulated samples do not require filtration prior to analysis for
soluble constituents. For coagulated samples, then, total organic nitrogen
is assumed equivalent to SON.
TABLE 16. COMPARISON BETWEEN FILTERED AND UNFILTERED COAGULATED PASE SAMPLES
Coagulant
FeCl3
CaO
A12(S04)3
Coagulant dose,
mg/1
200
200
150
Organic Nitrogen, mg/1
Total
0.75
0.89
1.26
Soluble
0.73
0.91
1.27
COD, mg/1
Total
16
17
22
Soluble
16
18
21
41
-------
In most current advanced was tewater- treatment schemes, coagulation follows
secondary clarification and precedes filtration. It would seem unreasonable
to coagulate an effluent after removing the particulate matter by filtration.
However, it is conceivable that the presence of particulates could influence
the removal of soluble material, perhaps by making the floe particles more ne-
gatively charged or acting as a coagulant aids. To test this hypothesis, both
filtered and unfiltered portions of a sample of PASE were coagulated with lime
and with ferric chloride. The results, presented in Table 17, indicate that
only very small (but significant) differences in removal of SON resulted from
filtration prior to coagulation.
TABLE 17. COAGULATION OF FILTERED AND UNFILTERED PASE SAMPLES
Sample
Control
Filtered
Unfiltered
Filtered
Unfiltered
Coagulant
_ _
600 mg/1 FeCl3
600 mg/1 FeCl3
400 mg/1 CaO
400 mg/1 CaO
pH
7.7
6.1
6.1
11.4
11.4
SON,
mg/1
1.20
0.95
0.84
0-76
0.84
Percent
SON
Removed
_
21
30
37
30
SCOD
mg/1
24
16
15
17
16
Percent
SCOD
Removed
-
36
39
32
34
Mixing time is an important variable in chemical coagulation, both eco-
nomically and operationally. In this study, it was desirable to have a mixing
time long enough to ensure maximum removal of the soluble organics. Table 18
shows that with both ferric chloride and lime, maximum removal was achieved
with a 15-minute mixing time. To ensure that mixing time was adequate in the
remainder of the experiments, 20 minutes was selected.
In those experiments involving ferric chloride and alum, the coagulant
was added first, depressing the pH sharply, and then sodium hydroxide was ad-
ded to attain the desired pH. Such additions can create very high and low
localized pH values, perhaps sufficient to alter the organics present or to
solubilize some of the particulate matter. An experiment in which a sample
was coagulated both with precombined and separately added reagents showed
identical results for both techniques. Thus, localized extreme pH concentra-
tions of the magnitude generated in these experiments should not have affected
the outcome of the experiments. directed
The coagulation of two samples with ferric chloride was studied as a
function of pH to determine the optimum pH for removal of SON These result
are presented in Table 19 and indicate that pH 5 to 8 is t-h* ^- results
SON removal by ferric chloride. The pH range 5 to 6 5 ?, n% P
of negatively charged colloids in general ?g]? ION removal "were
constant in the PH range 5 to 6.5, and this range was "eT^ 2Tt f
maining experiments. c OI
42
-------
TABLE 18. SON REMOVAL AS A FUNCTION OF MIXING TIME AT 25 RPM
Coagulant
None
FeCl3
Fed 3
FeCl3
FeCl3
CaO
CaO
CaO
CaO
Dose,
mg/1
_
600
600
600
600
400
400
400
400
Mixing
Time,
minutes
_
5
10
15
20
5
10
15
20
SON,
mg/1
1.09
0.76
0.76
0.66
0.66
0.85
0.79
0.75
0.79
Percent
SON
Removed
_
30
30
39
39
22
28
31
28
TABLE 19. EFFECT OF pH ON FERRIC CHLORIDE COAGULATION OF SON
Sample 1
SON =1.35 mg/1
A*
(> 6%, > 8%)
Sample 2
SON =0.95 mg/1
(> 8%, > 12%)**
PH
4.5
5.0
5.5
6.0
6.0
7.0
8.0
9.0
10.0
*
Coagulant dose = 600 mg/1 FeClo.
**
Differences in percent removal to be
significance level, respectively.
SON,
mg/1
0.90
0.80
0.79
0.81
0.48
0.56
0.52
0.60
0.61
Percent
SON
Removed
33
41
41
40
49
41
45
37
36
exceeded for 95% and S9%
43
-------
An optimum pH was not determined for alum coagulation, and it was assumed
to be the same as for ferric chloride coagulation. Malhotra et al. [74] found
the optimum pH for alum coagulation of phosphate to be 5.6 and that organic
nitrogen removal was better at pH 6 than at pH 8.
Series of jar tests were conducted to determine the optimum coagulant
dose for maximum SON removal by ferric chloride, alum, and lime. Results are
listed in Tables 20, 21, and 22, respectively. In all cases, a coagulant dose
of about 200-300 mg/1 effectively removed as much of the SON and SCOD as
higher doses.
Polyelectrolytes N-17, A-23, and C-31 (Dow Chemical Company) were tested
as coagulants, but due to their high solubility and the nature of secondary
effluent they failed to produce a settleable floe at concentrations up to 50
mg/1. As coagulant aids, they were tested in combination with lime and ferric
chloride, and in all cases worked extremely well and produced a very clear
supernatant liquor within minutes after stirring was stopped. However, they
showed no potential for increasing the removal of SON with ferric chloride or
lime, and actually increased the concentration of SON and SCOD (the polyelec-
trolyte contained organic nitrogen) unless used in very low doses.
Clay minerals possess a relatively large negative surface charge at neu-
tral pH, and are frequently employed as coagulant aids. Although soluble or-
ganic contaminants in general are also most likely to be charged negatively,
molecules containing organic nitrogen may have localized centers of positive
charge which would allow them to adsorb on a clay mineral surface.
Koalinite was used as a coagulant aid in ferric chloride coagulation of
PASE but did not increase SON removal. Bentonite, which has a much larger
surface area than kaolinite, was used in combination with ferric chloride and
C-31 and caused a substantially significant (99% CI) increase in SON removal
from PASE of about 10 percent with a dose of 200 mg/1 bentonite. It is pos-
sible that the bentonite merely increased the efficiency of flocculation, but
this effect was not found with kaolinite nor with any of the polyelectrolytes.
Thus, there appears to be a very small fraction of SON which is adsorbed by
bentonite, but not removed by ferric chloride. Bentonite and C-31 together
achieved an SON removal of 14 percent and a SCOD removal of 27 percent.
The differences in SON and SCOD removal, achieved by the three coagulants
shown in Table 23, were tested for significance as described in Appendix A,
Section 5. Ferric chloride removed significantly (99% CI) more SON than did
lime, which in turn removed significantly (95% CI) more SON than did alum.
This is the same order of effectiveness noted by Parkin and McCarty [73] for
removal of SON from treated agricultural wastewaters; however, there is insuf-
ficient data to generalize this finding and extrapolate it to all biologically
treated effluents.
There was not a significant difference between the SCOD removals achieved
by the three coagulants, nor did alum remove significantly more SCOD than SON
However, ferric chloride removed significantly (99% CI) more SON than SCOD
and lime also removed significantly (95% CI) more SON than SCOD '
44
-------
TABLE 20. SON AND SCOD REMOVAL BY FERRIC CHLORIDE COAGULATION
Sample 1
Organic Nitrogen = 1.4 mg/1
COD = 29 mg/1
Alkalinity = 180 mg/1
(9-24-75, 1:30 PM)
Sample 2
Organic Nitrogen =1.9 mg/1
COD = 31 mg/1
Alkalinity = 192 mg/1
(9-26-75, 1:00 PM)
A
Fe 3,
mg/1
0
100
200
250
300
350
400
500
0
100
200
300
400
500
600
700
pH
7.2
6.7
5.7
3.6
3.4
:
-
7.6
6.5
5.5
5.5
5.2
5.5
5.6
5.7
Turbidity
NTU
_
—
-
:
—
3.4
1.1
0.8
0.7
1.2
1.0
1.0
1.4
SON,
mg/1
1.06
0.87
0.75
0.70
0.72
0.81
0.95
1.17
1.11
1.03
0.91
0.80
0.86
0.78
0.82
0.78
Percent
SON
Removed
18
29
34
32
24
10
0
_
7
18
28
23
30
26
30
**
SCOD,
mg/1
23
19
16
16
16
19
22
23
22
17
14
13
14
14
12
14
Percent
SCOD
Removed
_
16
31
32
31
19
4
0
—
22
37
41
37
35
45
37
No base added in coagulation of Sample 1; adjusted with NaOH in Sample 2.
**
Two aliquots analyzed for each sample.
t_n
-------
TABLE 21. SON AND SCOD REMOVAL BY ALUM COAGULATION
Sample 1
Organic Nitrogen =2.7 mg/1
COD = 40 mg/1
Alkalinity = 210 mg/1
(9-9-75, 2:30 PM)
Sample 2
Organic Nitrogen =1.9 mg/1
COD = 27 mg/1
Alkalinity = 192 mg/1
(9-29-75, 1:00 PM)
A12(S04)3,
mg/1
0
100
150
200
250
300
350
400
450
500
0
s\ f\f\
200
300
400
500
600
800
1000
1200
PH
7.5
7.0
6.8
6.6
6.4
6.3
6.1
5.9
5.7
5.5
7.4
5f\
. 0
5.9
6.0
6.1
6.1
6.1
6.2
6.2
Turbidity,
NTU
5.0
1.2
1.0
0.80
0.80
0.80
0.65
0.70
0.75
1.2
3.2
1f\
. 3
0.60
0.50
0.45
0.40
0.45
0.40
0.35
SON,
mg/1
1.65
1.52
1.26
1.22
1.21
1.20
1.21
1.11
1.08
1.13
1.16
1^\ o
.03
0.85
0.83
0.89
0.86
0.84
0.84
0.93
Percent
SON
Removed
_
8
24
26
27
27
27
33
35
32
27
28
23
26
28
28
20
*
No base added in coagulation of Sample 1; adjusted with NaOH in Sample 2.
**
SCOD
mg/1
24
24
22
21
20
19
18
18
18
19
21
4 A
13
13
11
12
13
13
13
13
Percent
SCOD
Removed
_
2
7
11
18
21
23
23
26
21
-
39
38
46
41
37
35
39
38
**
Two aliquots analyzed for each sample.
-------
TABLE 22. SON AND SCOD REMOVAL BY LIME COAGULATION
Sample 1
Organic Nitrogen =1.4 mg/1
COD = 25 mg/1
Alkalinity = 216 mg/1
(10-8-75, 1:00 PM)
Sample 2
Organic Nitrogen =1.4 mg/1
COD = 28 mg/1
Alkalinity =154 mg/1
(10-14-75, 12 noon)
CaO,
mg/1
0
100
200
300
400
500
600
700
0
400
600
800
1000
1200
1400
1600
1800
it
PH
7.6
9.6
10.4
11.2
11.6
11.8
11.9
11.9
7.3
11.6
11.9
12.1
12.2
12.3
12.3
12.3
12.3
Turbidity
NTU
2.2
3.1
1.8
0.60
0.50
0.90
0.85
0.65
1.80
0.55
0.50
0.45
0.50
0.55
0.55
0.65
0.80
SON,
mg/1
1.05
0.97
0.89
0.87
0.79
0.82
0.73
0.73
1.12
0.78
0.73
0.81
0.76
0.70
0.68
0.68
0.68
Percent
SON
Removed
_
8
15
17
25
22
30
30
30
35
28
32
37
39
39
39
**
SCOD,
mg/1
20
19
17
16
17
15
13
15
23
18
19
16
20
18
16
16
16
Percent
SCOD
Removed
_
6
18
23
18
25
37
28
22
16
29
14
20
29
29
29
*
pH not adjusted after lime addition.
**
Two aliquots analyzed for each sample.
-------
TABLE 23. COMPARISON OF SON AND SCOD REMOVAL BY FERRIC CHLORIDE, LIME, AND ALUM COAGULATION
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24*
25
26
Average
Initial
SON,
mg/1
1.19
1.25
1.26
1.43
1.19
1.18
1.09
1.00
1.06
1.06
1.35
1.11
1.40
1.38
1.35
1.57
1.47
1.20
1.09
0.95
1.05
1.10
1.12
1.16
1.65
1.16
1.23
Initial
SCOD,
mg/1
21
—
-
25
24
23
23
—
-
23
27
22
27
23
21
30
25
31
-
-
20
22
23
24
24
21
24
FeC13
Dose,
mg/1
1000
1000
1000
600
600
600
600
600
600
250
600
500
600
600
600
600
600
600
600
600
-
-
_
-
-
-
Percent
SON
Removed
48
46
44
43
44
36
39
39
47
34
40
27
45
38
44
46
42
39
48
49
—
-
_
-
-
-
42 + 5
Percent
SCOD
Removed
19
_
-
39
29
27
31
_
-
32
39
39
29
29
38
40
32
30
-
-
—
—
_
-
-
-
32 + 6
CaO
Dose,
mg/1
1200
1000
1200
400
400
400
400
600
600
_
_
_
-
_
-
_
-
-
-
-
600
400
1000
400
-
-
Percent
SON
Removed
39
30
38
30
26
31
31
40
36
_
_
_
-
_
—
_
-
_
-
-
27
39
35
43
—
-
34 + 5
Percent
SCOD
Removed
21
—
-
36
22
29
14
_
-
_
—
_
-
_
_
_
-
_
_
-
27
13
24
33
—
—
24 + 8
Dose,
mg/1
1000
1000
1000
_
—
-
-
—
-
_
—
_
-
_
_
-
_
_
-
_
_
600
400
600
A12(S04)
Percent
SON
Removed
23
29
24
—
-
-
—
-
-
_
_
_
-
_
_
-
_
_
-
_
_
_
31
30
26
27 + 3
3
Percent
SCOD
Removed
17
-
-
—
-
—
—
—
-
_
_
_
-
_
_
_
-
_
_
-
_
_
_
33
22
39
28 + 10
*
Filtered prior to coagulation.
00
-------
It was shown earlier, in Table 19, that ferric chloride coagulation was
significantly (99% CI) more efficient in removing SON at pH 6.0 than at pH
10.0, and the results of Table 23 indicate that there are differences in the
removal efficiencies of the three coagulants. In order to more closely com-
pare the SON removals achieved by each of the three coagulants, the superna-
tant liquor from a number of jar tests was coagulated a second or third time
with a different coagulant or pH, as shown in Table 24.
Interestingly, an additional removal of SON was achieved in several in-
stances (Table 25). After an initial coagulation with ferric chloride, a sec-
ond coagulation with ferric chloride or alum produced no significant increase
in SON removal, but a second coagulation with lime increased SON removal by
17% (99% CI). After an initial coagulation with lime, a second coagulation
with lime increased SON removal by 16% (99% CI). Both ferric chloride and
lime removed a significant (99% CI) amount of SON following an initial alum
coagulation, but a second alum coagulation produced no increase in SON removal.
Other combinations of coagulations showed that: (1) there is no significant
difference in the removals achieved by lime and ferric chloride at pH 10.0;
(2) ferric chloride coagulation at pH 10.0 removes a significant (99% CI) ad-
ditional fraction of SON after an initial coagulation at pH 6.0; and (3) fer-
ric chloride coagulation at pH 6.0 removes a significant (99% CI) additional
fraction of SON after an initial coagulation at pH 10.0.
The differences in removal between the various coagulants and the addi-
tional removals achieved by a second coagulation are explainable in terms of
the factors affecting adsorption. The surface charge of the ferric chloride
floe in PASE is more positive at pH 6.0 than at pH 10.0 (Lengweiler et al.
[88] have determined the PZC of iron oxides in dilute solutions to be 6.7.)
Since coagulation with ferric chloride at pH 10.0 removed about 10-15 percent
less SON than at pH 6.0, it appears that electrostatic charge may be of some
importance. Furthermore, there is a small but significant fraction not re-
moved at each pH which is removed with a second coagulation at the other pH
(see Table 24, Samples 2 and 4).
Since the oxide surfaces involved are very polar as well as highly hy-
drated, it is unlikely that polarity plays a significant role in SON removal.
Thus, SON removal by chemical coagulation is postulated to result from elec-
trostatic attraction and special factors, i.e., entrapment, precipitation,
and specific adsorption.
Aluminum sulfate floes have shown to have a PZC of about 8.0 [54,90], and
are thus likely to be positively charged at pH 6.0, as are ferric chloride
floes at the same pH. Therefore, the differences in removal of SON between
ferric chloride and alum (see Tables 23 and 24) are considered to be due to
special factors. Ferric ion is known to form stronger complexes with carbox-
ylic acid groups, which can be present in secondary effluent in the form of
amino acids, carbohydrates, and fulvic acid. The floe formed by ferric chlo-
ride differs from that formed by alum in its size and settling characteris-
tics, such that it is conceivable that the two floes may not be of equal effi-
ciency in entrapping large molecules. Thus, the additional removal of SON
achieved by ferric chloride as compared to alum (Table 24) is due to a
49
-------
TABLE 24. EFFECT OF SEQUENTIAL COAGULATION ON SON REMOVAL
Sample 1
SON =1.25 mg/1
(12-8-75, 11:30 AM)
(> 6%, > 9%)*
Sample 2
SON =0.95 mg/1
(12-15-75, 11:00 AM)
JL
(> 8%, > 12%)
Sample 3
SON =1.00 mg/1
(12-16-75, 11:15 AM)
(> 8%, > 11%)*
Sample 4
SON = 1.06 mg/1
(12-17-75, 11:00 AM)
(> 8%, > 10%)*
First
Coagulant
FeCl3
Fed 3
FeCl3
FeCl3
CaO
CaO
A12(S04)3
A12(S04)3
A12(S04)3
A12(S04)3
FeCl3
FeCl3
FeCl3
FeClo
j
FeCl3
FeClo
CaO
CaO
FeCl3
FeCl3
FeCl,
FeCl3
FeCl3
FeCl3
Dose,
mg/1
1000
1000
1000
1000
1000
1000
1000
1000
1000
1000
600
600
600
600
600
600
600
600
600
600
600
600
600
600
pH
6.6
6.6
6.6
6.6
12.1
12.1
6.5
6.5
6.5
6.5
6.0
6.0
10.0
10.0
6.0
10.0
12.1
12.1
6.0
6.0
6.0
6.0
10.0
10.0
Second
Coagulant
FeCl3
CaO
A12(S04)3
-
CaO
_
FeCl3
CaO
A12(S04)3
—
FeCl3
FeClo
j
_
-
FeCl3
-
CaO
FeCl3
FeCl3
-
FeCl3
Dose,
mg/1
1000
1000
1000
-
1500
-
1000
1500
1000
_
600
600
_
-
600
-
600
600
600
-
600
pH
.
6.2
12.3
6.2
-
12.4
-
7.3
12.3
7.4
.
10.0
6.0
_
-
10.0
-
12.0
10.0
10.0
-
6.0
Third
Coagulant
_
—
-
-
-
-
-
-
—
_
-•
—
-
-
**
-
—
CaO
-
—
Dose,
mg/1
_
—
-
-
-
-
-
-
—
_
-
-
_
-
-
-
—
600
_
-
pH
_
_
-
-
-
-
-
—
—
_
-
—
_
-
-
-
_
12.4
-
-
Percent
SON
Removed
46
50
63
52
30
46
29
44
51
30
49
61
36
61
39
34
40
43
47
51
58
57
25
55
*
Differences in percent SON removal to be exceeded for 95% and 99% significance, respectively.
Ln
O
-------
TABLE 25. ADDITIONAL SON REMOVALS ACHIEVED BY SEQUENTIAL COAGULATION
Initial
Coagulant
Subsequent
Coagulant
Significant increase in SON removal:
Lime
Ferric chloride (pH 6.6)
Alum (pH 6.5)
Alum (pH 6.5)
Ferric chloride (pH 6.0)
Ferric chloride (pH 6.0)
Ferric chloride (pH 10.0)
Ferric chloride (pH 10.0)
Lime
Lime
Ferric chloride (pH 7.3)
Lime
Ferric chloride (pH 10.0)
Ferric chloride (pH 10.0)
Ferric chloride (pH 6.0)
Ferric chloride (pH 6.0)
No significant increase in SON removal:
Ferric chloride (pH 6.6)
Ferric chloride (pH 6.6)
Alum (pH 6.5)
Lime
Ferric chloride (pH 6.2)
Alum (pH 6.2)
Alum (pH 7.4)
Ferric chloride (pH 10.0)
Additional
SON
Removal , %
16
17
15
22
12
11
25
30
C.I.
%
99
99
99
99
95
99
99
99
combination of specific adsorption and entrapment, but the data do not indi-
cate which factor is the more important.
Lime forms a negatively charged floe in wastewater due to the presence
of phosphate and bicarbonate anions [81]. Thus, it must remove SON primarily
by virtue of special factors, since most of the organics are also negatively
charged. The calcium salts of many carboxylic acids are highly insoluble,
calcium forms strong complexes with the carboxylic acid group, and the dense,
rapidly settling nature of the lime floe undoubtedly aids entrapment. Once
again, the data do not allow distinction between these mechanisms.
A second coagulation of one sample with lime resulted in an additional
16 percent removal of SON (Sample 1, Table 24), suggesting the importance of
electrostatic charge, since the second floe would be considerably less nega-
tively charged than the first, due to the absence of phosphate and bicarbo-
nate anions.
The data presented in Table 24 suggest: (1) that two small fractions of
SON are present, one more positively charged and one more negatively charged,
and the removal of each depends upon the charge of the floe particles;
51
-------
(2) that there are fractions of SON capable of specifically adsorbing to the
floe particles of each of the three coagulants; and (3) that there may be a
fraction of SON, consisting of colloidal material, which is removed by all
coagulants either through flocculation or entrapment.
Since each of the coagulants removes SON by a combination of electro-
static charge and special factors, it is conceivable that for different waste-
waters, different coagulants may prove superior. Thus, for a particular
wastewater, a series of jar tests would be appropriate for determining the
optimum coagulant.
Direct Addition of Coagulants to Activated Sludge
Iron and aluminum salts can be added to activated sludge for phosphorus
removal without hindering the biological functions of the bacteria [85,86].
In one study [87], it was found that a weight ratio of ferric to phosphate
ions of 1.5 to 1.0 resulted in 83 percent reduction in phosphorus. This ratio
would require a ferric chloride dose of about 85 mg/1 at the Palo Alto plant.
Table 26 shows the results of an experiment in which ferric chloride and C-31
were added to activated sludge. About 30 percent of the SON was removed with
a dose of 200-300 mg/1 of ferric chloride (although a significantly, 99% CI,
larger percentage could be removed by a separate coagulation process) , and
the effluent was quite clear. An added advantage of direct addition to acti-
vated sludge is that separate coagulation and sedimentation tanks are not re-
quired, resulting in a much less expensive operation.
The effectiveness of polyelectrolytes when used directly with activated
sludge, rather than with secondary effluent, was also tested. The cationic
polymer C-31, was able to produce a small but significant (99% CI) reduction
in SON and SCOD (16 and 21 percent, respectively) when added directly to acti-
vated sludge at a concentration of 20 mg/1, while the addition of A-23, the
anionic polymer, increased COD and turbidity and had no noticeable effect on
the rate of settling. Busch and Stumm [84] found the polymeric material ex-
creted by bacteria to be much like an anionic polymer. The fact that only
the cationic polymer was able to improve settling and organics removal in
this study also indicate the anionic nature of the soluble organics present
in the activated sludge. The data indicate that the practice of adding a cat-
ionic polymer to activated sludge for improved settling may also result in
slightly increased removal of soluble organic matter; however, no experiments
were undertaken to determine the steady-state effect of such practice, with
the sludge being recycled to the aeration tanks and re-coagulated continuously.
Summary
With respect to secondary effluent:
1. Ferric chloride, lime, and alum removed an average of 42 + 5 34 + 5
and 27 ± 3 percent of the SON, respectively, from a number of PASE
samples, and an average of 32 ± 6, 24 ± 8, and 28 + 10 percent of
the SCOD, respectively. The percentage differences between coagu-
lants were statistically significant for removal of SON, but not for
removal of SCOD. Ferric chloride and lime removed significantly
more SON than SCOD.
52
-------
TABLE 26. REMOVAL OF SON AND SCOD BY ADDITION OF FERRIC CHLORIDE AND C-31 COAGULANT TO ACTIVATED SLUDGE
Sample 1
Organic Nitrogen =2.2 mg/1
COD = 32 mg/1
(12-5-75, 11:30 AM)
SON (> 6%, > 8%)**
SCOD (> 4%, > 6%)**
Sample 2
(12-11-75, 11:00 AM)
SON (> 7%, > 10%)**
FeCl3,
mg/1
_
25
50
100
300
600
900
600*
_
200
300
400
600
200
300
400
600
600*
C-31
mg/1
_
—
-
—
-
-
_
-
-
—
5
5
5
5
-
pH
7.2
6.9
6.7
6.4
6.2
6.2
6.2
6.0
7.2
5.8
5.9
6.0
6.0
5.7
5.8
5.9
5.9
5.8
Turbidity
NTU
2.6
1.6
1.4
0.9
0.5
0.7
0.3
0.4
_
-
-
—
-
-
-
-
SON,
mg/1
1.38
1.68
1.61
1.17
0.97
1.15
1.11
0.86
1.09
0.78
0.79
0.82
0.90
0.70
0.75
0.81
0.82
0.57
Percent
SON
Removed
0
0
15
30
17
20
38
_
28
28
25
17
36
31
26
25
48
SCOD,
mg/1
23
28
26
24
18
22
19
16
_
-
-
—
—
-
-
-
Percent
SCOD
Removed
_
0
0
0
22
6
8
29
_
-
-
—
-
-
-
-
Settling time = 45 minutes.
*
Added to supernatant siphoned from settled activated sludge.
**
Differences in percent removal to be exceeded for 95% and 99% significance, respectively.
Ln
UJ
-------
2. Sequential coagulation of samples, using different coagulants or pH
values, indicated differences in the fractions of organic matter re-
moved by the various coagulants and the dependence of removal of
certain fractions upon pH and surface charge.
3. Significant additional removals of SON were achieved with ferric
chloride or lime after alum coagulation, ferric chloride at pH 6.0
after ferric chloride at pH 10.0, ferric chloride at pH 10.0 after
ferric chloride at pH 6.0, and lime after ferric chloride at pH 6.0
or lime.
4. Optimum coagulant doses of ferric chloride, lime, and alum were about
200-300 mg/1 for maximum removal of SON from PASE.
5. Polyelectrolytes created rapidly settling floe, but were of no bene-
fit in removing SON.
6. Bentonite improved SON removal by about 10 percent at a dose of 200
mg/1 when added prior to ferric chloride coagulation, but no in-
creased removal was found with kaolinite.
With respect to direct addition to activated sludge:
1. Small reductions in SON and SCOD, 16 ± 3 and 21 ± 2 percent, respec-
tively, were attained by direct addition of cationic polymer to a
sample of activated sludge.
2. Direct addition of ferric chloride to activated sludge resulted in
about 30 percent removal of SON at a dose of 200 mg/1. Although a
higher removal of SON could be attained by separate coagulation of
the supernatant liquor, direct addition merits consideration for
economic reasons.
ION EXCHANGE AND ACTIVATED-CARBON ADSORPTION
Background
Ion exchange and activated-carbon adsorption are discussed together here
due to the similarity of the mechanisms by which they remove organic contami-
nants. For the purposes of this study, ion exchange is considered to be an
adsorption process in which the dominant factor affecting adsorption is elec-
trostatic charge.
Removal of organic contaminants from wastewater by adsorption is influ-
enced by three factors: (1) electrostatic charge, (2) polarity, and (3)spe-
cial factors. In the case of ion-exchange resins, adsorption is ideally de-
pendent solely on electrostatic charge; however, in the adsorption of organic
molecules by the resins, there also can be interactions between the molecules
and the resin matrix. This is particularly true of non-ionic organic mole-
cules. For ionized organic molecules, adsorption will be the result of the
superimposed effects of electrostatic charge, polarity, and special factors.
54
-------
There are a large number of anionic and cationic ion-exchange resins com-
mercially available which differ in acid/base strength, pore size, functional
group, matrix characteristics, mesh size, ion selectivity, and other parame-
ters. It is not within the scope of this study to determine which of these
resins can achieve optimum removal of SON nor to determine how the various
characteristics of the resins affect the removal of SON. Rather, a resin was
selected which was hoped would favor electrostatic charge over other factors
affecting adsorption.
Cleaver and Cassidy [91], observing the adsorption of amino acids on var-
ious resins, postulated that resins of lower equivalent weight would be more
polar, thus decreasing the non-polar surface area available for adsorption of
the aromatic and aliphatic parts of the molecules. They found Dowex-50 to
have the lowest equivalent weight of the resins investigated. In this study,
a resin equivalent to Dowex-50, Bio-Rad AG50W-X8, was used together with its
anionic counterpart, Bio-Rad AG1-X8.
Adsorption of organics from secondary effluent on ion-exchange resins has
largely been used as an analytical tool due to the more advanced state of the
art of activated-carbon adsorption, although resins are being developed with
the hope of equaling, complementing, or surpassing the performance of acti-
vated carbon.
Rebhun and Kaufman [92] investigated the removal of COD and color from
secondary effluent and found that the weak base phenolic resins, Duolite A-7
and ES-33, achieved the greatest removal of the resins tested, and were
roughly comparable to activated carbon. A non-ionic phenolic resin was also
found to achieve significant removal of COD and color, thus indicating the
importance of polarity and matrix characteristics for adsorption of organics
from secondary effluent. Little or no removals of COD and color were observed
with strong cationic resins, further supporting the postulation that secondary
effluent organics are negatively charged at neutral pH.
Parkin and McCarty [73] studied the removal of SON from a sample of
treated agricultural wastewater and found that an anionic resin achieved
greater SON removal than a cationic resin, and that either raising or lower-
ing the pH from the neutral range improved removal.
Of particular interest to this study is the work of several investigators
who have used ion exchange chromatography to separate amino acids [102,103,
104,105,106], nucleic acid degradation products [107], and ribonucleotides
[108]. These investigators used cation exchange resins and were able to ad-
sorb these nitrogen-containing organics at low pH, and elute them from ex-
change columns with buffers of increasing pH. For example, Moore and Stein
[103] adsorbed 17 amino acids onto a column of Dowex-50, and eluted 8 of them
with a buffer of pH 3.4, and 6 more with a buffer of pH 4.2.
Activated carbon has been shown to have a high affinity for adsorbing
organic molecules from solution, and granular activated carbon has been used
at several treatment plants for polishing secondary effluent, well known
plants being located at Lake Tahoe [126], Dallas [21], and Pomona [127]. The
55
-------
technology for designing and operating granular activated-carbon adsorption
units has been developed and tested and is ready for immediate use [111].
There are a number of different carbon sources and techniques used in
the manufacture of activated carbons, producing carbons with a wide range of
pore sizes, specific surface areas, acidities, and other parameters which may
affect their performance. Various investigators have discussed the catalytic
and adsorbent properties of activated carbon [119], its surface chemistry
[120,121], and the kinetics, equilibrium, and capacity for adsorption on acti-
vated carbon [122,123]. Activated carbon is known to possess ion exchange
properties [119], and has been shown by Urano [124] to adsorb heavy metals
from solution.
Activated carbon may be either acidic or basic, depending primarily on
the temperature and method of its regeneration [119], which may affect its
performance with a particular solute. The pH at which adsorption takes place
is also important, as a decrease in pH has been shown to increase adsorption
capacity [123], and an acidic activated-carbon bed in sequence after a basic
bed has been shown to accomplish an additional removal of organics from sur-
face waters [116,125]. However, the concentration of material which cannot
be adsorbed from secondary effluent by activated carbon is reportedly not
significantly affected by pH changes [73,128,129].
Kim et al. [128] have observed that the cell residence time (CRT) of the
activated sludge is an important variable in activated-carbon adsorption of
secondary effluent organics, the non-adsorbable fraction decreasing with in-
creasing CRT, but the adsorbable fraction remaining relatively constant.
Earlier studies by DeWalle and Chian [28,130] indicated that low-molecular-
weight polar compounds constituted a significant portion of the non-adsorbable
organics, and that these compounds tended to be removed by aerobic biological
treatment, suggesting agreement with and prediction of the results of Kim
et al.
Parkin and McCarty [73] contacted a sample of biologically treated agri-
cultural wastewater with 50 g/1 of activated carbon and observed an SON re-
moval of 94 percent.
Helfgott et al. [13] observed a significant positively charged organic
fraction in activated-carbon effluent, indicated both by cation exchange and
by use of gravitational electrophoresis. The resin used was in the sodium
form, so presumably the adsorption took place at neutral pH.
Experimental Procedure
The ion-exchange resins used were analytical-grade purchased from Bio-Rad
Laboratories (Richmond, California). The manufacturers' specifications are
shown in Table 27. The resins were rinsed with dilute acid or base then
rinsed with sodium chloride for conversion to the sodium and chloride forms,
and then rinsed with a large quantity of deionized water. The deionized water
was drawn from the resins by vacuum filtration through a fritted glass filter
and the resins were then stored for use. The water content of the resins was'
56
-------
TABLE 27. MANUFACTURERS' SPECIFICATIONS FOR THE ION-EXCHANGE RESINS USED
Type:
Name:
Form:
Total Capacity (meq/dry g):
Actual Wet Mesh Range (U.S.
S tandard):
Moisture Content (weight 7<>):
Functional Group:
Effective Pore Size:
Cation
AG 50W-X8
Hydrogen
5.1
40-80
50-56
Sulfonic Acid
Medium
An ion
AG1-X8
Chloride
3.2
40-80
39-45
Quaternary Ammonium
Medium
determined to be 47 percent for the cationic resin and 44 percent for the
anionic resin by drying at 103°C for several hours.
In batch studies 15 g/1 of resin were added to Pall-filtered samples of
PASE in erlenmeyer flasks, and then shaken on a "wristaction" shaker (Burrell
Corp., Pittsburgh, Pa.) for 12 hours. As shown in Table 28 this dosage was
sufficient to remove virtually all of the exchangeable (adsorbable) SON.
Resins were removed from samples by filtration through glass fiber filters
(Reeve Angel 934 AH), except for samples contacted at pH 9 or higher, which
were settled and then decanted, since filtration here was hindered by forma-
tion of a precipitate.
Hydrochloric acid and sodium hydroxide were used for pH adjustment, with
the exception that sulfuric acid was used where noted to prevent chloride in-
terference in the SCOD test. The pH was checked often if necessary and ad-
justed to maintain a constant pH in the appropriate experiments.
Levels of SON and SCOD in "blank" deionized water samples contacted with
the resins and with activated carbon were found to be satisfactorily low, as
shown in Table 29. These low-level SON values were measured by a variation
of the technique described in Standard Methods [3], in which 350 ml of the
sample was first distilled to ensure complete removal of the ammonia, fol-
lowed by digestion as usual. Then only 250 ml of deionized water was added
following digestion, and approximately 125 ml was distilled into 25 ml of
0.02 N sulfuric acid, with the final volume brought to 200 ml in a volumetric
flask. This method extended the sensitivity of the analysis to a lower level
by producing a final volume of only 200 ml, as compared with 500 ml in the
standard procedure. Using this technique, recoveries of standards were accu-
rate within ± 5 percent down to a level of 50 micrograms of organic nitrogen
per flask (0.10 mg/1 for a 500-ml sample). Below this level results were er-
ratic, and usually high. Thus, all of the SON values shown in Table 29 are
below the detection limit for the analysis.
Grab PASE samples were Pall filtered immediately upon return to the lab-
oratory, and were then either used at once or stored at 4°C until warmed to
57
-------
TABLE 28. EFFECT OF SEQUENTIAL CONTACTING AND DOSAGE OF ION-EXCHANGE RESINS ON REMOVALS FROM PASE
Samp le 1
Sample 2
First
Exchange
Cationic
Cationic
Anionic
Anionic
Cationic
Cationic
Cationic
Cationic
Dose,
8/1
15
15
15
15
0
5
15
25
50
Second
Exchange
-
Cationic
-
Anionic
-
_
-
-
Dose,
g/1
-
15
-
15
-
_
.
-
SON,
mg/1
1.27
1.16
1.14
1.12
1.13
1.27
0.88
0.68
0.68
0.64
Percent
SON
Removal0
9
10
12
11
31
47
47
50
SCOD,
mg/1
23
22
21
15
13
23
18
19
19
20
Percent
SCOD
Removal
5
9
34
41
21
18
17
14
Final
pH
7.7a
8.0a
8.2a
7.5a
7.5a
2.0b
2.0b
2.0b
2.0b
2.0b
pH not adjusted.
pH adjusted with I^SO^.
Differences in percent SON removal greater than 6% and 9% are 95% and 99% significant, respectively.
Differences in percent SCOD removal greater than 4% and 6% are 95% and 99% significant, respectively
00
-------
TABLE 29. SON AND SCOD OF DEIONIZED WATER SAMPLES CONTACTED WITH
ION-EXCHANGE RESINS AND ACTIVATED CARBON
Solid
Anionic
Anionic
Anionic
Cationic
Cationic
Cationic
Cationic
Cationic
Cationic
Cationic
Cationic
Anionic
Anionic
Anionic
Anionic
Anionic
Cationic
Cationic
Anionic
Activated Carbon
Activated Carbon
Activated Carbon
Activated Carbon
Initial
PH
2.0
7.6
12.0
2.0
7.6
12.0
3.0
4.0
5.0
10.0
11.0
3.0
4.0
5.0
10.0
11.0
2.0
12.0
12.0
7.0
7.0
12.0
2.0
Final
PH
2.0
4.4
11.9
2.2
7.4
12.0
4.3
5.9
6.8
10.2
11.1
3.1
4.1
4.3
5.6
10.7
2.3
12.1
12.0
-
-
12.0
2.2
SON,
mg/1
0.04
0.02
0.05
0.03
0.01
0.04
0.01
0.07
0.04
0.00
0.00
0.01
0.05
0.04
0.03
0.05
—
_
-
0.01
0.05
0.00
0.00
SCOD,
mg/1
*
8.9
0.0^
4.5"
*
5.5
1.6
5.1*
1.6
1.2
0.4
1.4
1.6
0.6
0.0
0.0
0.0
0.8
0.4**
1.2**
1.2**
0.0
0.0
0.0**
0.0**
*
Chloride interference in SCOD test.
pH adjusted with sulfuric acid in place of hydrochloric acid.
room temperature for use.
(approximately 20°C).
All experiments were conducted at room temperature
The activated carbon, Nuchar WV-G, 12 x 40 mesh, was washed with deion-
ized water in an upflow column to remove carbon fines, and then dried at 103°C
and stored in an airtight container until used. In the batch activated-carbon
experiments 25 g/1 of carbon (except in the case of the pH-dependence experi-
ment) were added to Pall-filtered PASE in an erlenmeyer flask, which was then
shaken for 12 hours on the "wristaction" shaker. Equilibrium was reached af-
ter 14 hours with a 10 g/1 dose, thus the selection of a 12-hour contact time
for a 25 g/1 dose appears sufficient for maximum adsorption. The carbon was
59
-------
removed from the samples by filtration through glass fiber filters (Reeve
Angel 934 AH), which were found to produce identical results to filtration
through 0.45 micron Millipore filters.
Results and Discussion
The effect of pH on removal of SON and SCOD by ion-exchange resins is
shown in Figs. 5 and 6, respectively. Below pH 5, adsorption of SON and SCOD
by the cationic resin increased sharply, while removal by the anionic resin
was relatively independent of pH. In Fig. 5, differences in percent SON re-
moval of 5% and 7% are 95% and 99% significant, respectively; and in Fig. 6,
differences in percent SCOD removal of 3% and 5% are 95% and 99% significant,
respectively.
The data for adsorption by both resins were obtained from samples which
were contacted with a mixture of the two resins (mixed resins). When removal
by mixed resins equals the sum of the removal accomplished by the resins indi-
vidually, as for SCOD above pH 5, then mutually exclusive fractions of the or-
ganics have been removed by the individual resins. The SON data at pH 2 and
3, on the other hand, indicate that at least some portion of the same fraction
was removed by both resins. Thus at least part of the removal by the cationic
resin at low pH must be due either to adsorption of largely neutral compounds
or to adsorption of compounds having both positive and negative charges.
It appears that as the pH is lowered, some of the negatively charged or-
ganics become neutral, allowing them to adsorb to either resin, while some of
the neutral or negatively charged molecules become positively charged, allow-
ing them to be removed only by the cationic resin. In the case of the anionic
resin, some molecules become neutral and adsorb, while others become positive
and do not adsorb, thus keeping the percent removal by the anionic resins re-
latively constant. The SON removed at high pH appears to be partly amphoteriq
however, some removal may also be caused by precipitation due to the high pH.
In order to determine which fractions were mutually exclusive and which
were not, a series of samples were contacted sequentially with both resins at
various pH values. The results presented in Table 30 indicate that cationic
and anionic resins remove mutually exclusive fractions of SON and SCOD at neu-
tral pH (Sample 1) , and that the SON and SCOD removed by anionic resin at both
high and low pH contain at least some different material (Sample 2). The
fraction of SON and SCOD removed by cation exchange at low pH are signifi-
cantly (99% CI) distinct from those removed by anion exchange at high pH
(Sample 3). Changes in the sequence of contacting caused no significant
changes in the results, nor did contacting with both resins simultaneously.
It has been postulated that a significant fraction of the organic matter
which is not adsorbable on activated carbon is composed of small polar com-
pounds [28]. Ion exchange should favor the removal of compounds which are
small (since they have a higher charge density if charged and easier access
to the pores) , and apparently favors the removal of compounds which can be
favorably protonated or deprotonated (which would most likely also be quite
polar since only polar functional groups can be readily protonated or
60
-------
100
80
60
o
2
Ld
PASE 3-15-76
SON = 1.53 mg/J
o ANION EXCHANGE RESIN
a CATION EXCHANGE RESIN
A MIXED RESINS
PASE 3-22-76
SON = 1.30 mg/G0
O CATION EXCHANGE
RESIN
o
CO
PH
Figure 5. Removal of SON by ion exchange as a function of pH.
-------
100
80
8 60
PASE 3-15-76
SCOD=27
i - 1
PASE 3-22-76
SCOD = 25
o ANION EXCHANGE RESIN
a CATION EXCHANGE RESIN
A MIXED RESINS
o CATION EXCHANGE
RESIN
0
8
10
pH
Figure 6. Removal of SCOD by ion exchange as a function of pH.
-------
TABLE 30. REMOVAL OF SON AND SCOP BY SEQUENTIAL CONTACTING
WITH ION-EXCHANGE RESINS
Sample 1
(2-6-76)
<
SON(5%,6%)b
SCOD@%,4%)b
Sample 2
(3-1-76)
•§
SON(6%,8%)b
SCOD(4%,5%)b
Sample 3
(3-22-76)
SON(6%,8%)b
SCOD(4%,5%)b
First
Resin
_
Cationic
Anionic
An ionic
Cationic
Mixed Resins
-
Cationic
Cationic
Cationic
Cationic
Cationic
Anionic
Anionic
Anionic
Anionic
Anionic
-
Cationic
Anionic
Anionic
Cationic
First
pH
7.8
8.0
7.5
7.5
8.0
7.4
7.6
11.0
7.9
3.2
3.2
11.0
2.8
7.4
10.5
2.8
10.5
7.7
2.0a
12. 03
12.0s
2.0a
Second
Resin
_
-
Cationic
Anionic
-
_
-
_
Cationic
Cationic
-
-
-
Anionic
Anionic
-
-
Cationic
Anionic
Second
pH
_
-
7.8
7.6
-
-
-
_
11.0
3.3
-
-
-
10.6
3.1
-
-
2.0a
12. 03
SON,
mg/1
1.70
1.50
1.53
1.23
1.25
1.25
1.32
1.23
1.16
0.74
0.73
0.79
0.96
1.06
1.02
0.75
0.87
1.30
0.81
1.21
0.65
0.68
Percent
SON
Removed
_
12
10
28
26
26
-
7
12
44
45
40
27
20
23
43
34
-
38
7
50
47
SCOD,
mg/1
29
30
20
18
20
19
24
24
21
18
19
19
16
16
17
13
15
25
19
19
13
-
Percent
SCOD
Removed
_
0
32
37
33
33
-
0
12
26
20
20
34
32
29
45
37
-
23
24
48
-
apH adjusted with sulfuric acid and sodium hydroxide.
Differences in removal to be exceeded for 95% and 99% significance,
respectively .
deprotonated as a function of pH). Thus one might expect to find differences
in the fractions of organic matter removed by ion-exchange and activated-
carbon adsorption.
The data of previous investigators support this hypothesis p-3,129], as
do the data presented in Tables 30 and 31. Cation exchange by itself removed
12% of the SON which about equals the 14% SON removal by cation exchange fol-
lowing activated-carbon adsorption (95% CI = 6%). This indicates activated
carbon and cation exchange remove mutually exclusive fractions of SON. The
high SON to SCOD ratio of this fraction indicates its small molecular weight,
since it contains a significant amount of positively charged nitrogen and al-
most no SCOD. Small amino acids could account for such results if they were
present, but as discussed earlier, this is unlikely. Nucleic acid bases
would also exhibit similar characteristics, and these may in fact be present.
63
-------
TABLE 31. REMOVAL OF SON AND SCOP BY ACTIVATED CARBON AND ION-EXCHANGE RESINS
First
Contacting
Control
Activated Carbon
Activated Carbon
Activated Carbon
Activated Carbon
Cationic resin
Anionic resin
Cationic resin
pH
7.7
8.7
8.7
8.7
8.7
8.0
7.5
8.0
Second
Contacting
_
-
Anionic resin
Cationic resin
Mixed resins
-
-
Anionic resin
pH
_
-
7.8
8.9
7.8
-
-
7.6
SON,
mg/1
1.70
0.81
0.72
0.57
0.54
1.50
1.53
1.25
Percent
SON
Remove a
_
52
58
66
68
12
10
26
SCOD,
mg/1
29
9
9
9
8
30
20
20
Percent
SCOD
Removed
_
68
70
70
73
0
32
33
/q
Differences in percent SON removal greater than 5% and 6% are 95% and
99% significant, respectively.
Differences in percent SCOD removal greater than 3% and 4% are 95% and
99% significant, respectively.
As shown in Tables 31 and 32, the SON removed by anion exchange is only
partially removed by activated carbon, but the SCOD removed by anion exchange
is almost completely removed by activated carbon. There may, however, be an-
ionic molecules present in secondary effluent which are not removable with
the anion-exchange resin used in this study.
The data of Parkin and McCarty [73] indicate that the PH at which adsorp-
tion occurs does not affect the removal of SON by activated carbon. In this
study also, the equilibrium concentration of SON in samples of PASE contacted
with 25 g/1 of carbon was found to be independent of pH (Table 33) . The ef-
Sooi°f increaslnS PH on activated-carbon adsorption is described by Weber
1123J as increasing the adsorptive capacity of the carbon, which would not be
observable at such a large dose of carbon as 25 g/1, since the capacity of
the carbon would greatly exceed the amount of material to be removed. When
cr±^S °f/ASE.were contacted with only 400 mg/1 of carbon, the effect of de-
creasing pH in increasing the adsorptive capacity of the carbon for SON was
readily apparent, but the effect was not large.
It appears that the carbon used had a maximum adsorptive capacitv for SON
and SCOD at pH 3. Since the magnitude of the fraction S nln-adsorpablf or-
games was not a function of PH, it may be concluded that "
It -n £ S£S ErjiS i^ver^ im~
carbon adsorption, since (1) polarity and elec?rostatfc e"
64
-------
TABLE 32. REMOVAL OF SON AND SCOP FROM TREATED PASE AT NEUTRAL pH
•
Sample Treatment
Prior to
Ion Exchange
Sample #1
Control
Activated Carbon Ad-
sorption (neutral pH)
Sample #2*
Control
Activated Carbon Ad-
sorption (neutral pH)
Additional
SON
Removed by
Cation
Exchange,
mg/1
0.20
0.24
Additional
SCOD
Removed by
Cation
Exchange ,
mg/1
0.0
0.0
1
Additional
SON
Removed by
Anion
Exchange ,
mg/1
0.17
0.09
0.31
0.11
Additional
SCOD
Removed by
Anion
Exchange ,
mg/1
9
1
12
1
TABLE 33. REMOVAL OF SON AND SCOD BY ACTIVATED CARBON AS A FUNCTION OF pH
25 g/1 Activated Carbon
SON (6%, 8%)b
SCOD (4%, 5%)b
400 mg/1 Activated Carbon
Sample 1
SON (6%, 8%)b
SCOD (4%, 5%)b
Sample 2
SON (5%, 7%)b
SCOD (4%, 5%)b
Contacting
pHa
Control
2.0
3.0
4.0
7.0
12.0
Control
2.0
3.0
5.0
7.8
Control
2.0
3.0
5.0
7.0
SON,
mg/1
1.45
0.39
0.39
0.39
0.38
0.39
1.32
0.90
0.78
0.97
0.97
1.52
1.14
1.09
1.18
1.21
Percent
SON
Removed
_
73
73
73
74
73
_
32
41
26
26
-
25
28
22
20
SCOD,
mg/1
26
4
4
6
6
6
25
13
15
16
18
24
14
13
16
17
Percent
SCOD
Removed
_
84
84
79
78
76
_
49
38
34
29
-
45
46
32
29
apH adjusted with sulfuric acid and sodium hydroxide.
Differences in removal to be exceeded for 95% and 99% significance,
respectively.
65
-------
Table 34 presents a summary of the activated-carbon adsorption data.
Activated carbon removed 71 ± 12 percent of the SON and 81 ± 8 percent of the
SCOD from PASE. Removal of SON was significantly (95% CI) less than removal
of SCOD.
TABLE 34. SUMMARY OF ACTIVATED-CARBON ADSORPTION DATA AT NEUTRAL pH
Activated
Carbon,
8/1
25
25
25
Column
25
10
15
15
Initial
SON
Concentration,
mg/1
1.70
1.45
1.59
0.89
1.18
0.71
1.32
1.00
Percent
SON
Removed
52
74
75
70
92
76
71
56
Avg. 71 + 12
(n-8)
Initial
SCOD
Concentration ,
mg/1
29
26
28
18
22
21
27
24
Percent
SCOD
Removed
68
78
89
85
93
73
84
79
Avg. 81 + 8
(n-8)
Table 35 presents a summary of the SON and SCOD removals achieved by ion
exchange. Cation exchange removed significantly more SON than SCOD at neutral
pH (95% CI) and at pH 2 (99% CI). Anion exchange removed more SCOD than SON
(99% CI). Both resins removed about the same amount of SON at neutral pH.
Summary
1. The PASE samples contacted in this study with activated carbon exhib-
ited an average SON removal of 71 ± 12 percent, and an average SCOD
removal of 81 ± 8 percent. Activated carbon removed significantly
more SCOD than SON.
2. The adsorptive capacity of Nuchar WV-G for adsorbable SON increased
with decreasing-pH-to pH 3.0, but the magnitude of the non-adsorbable
SON fraction was independent of pH.
3. Cation exchange at neutral pH removed 11 ± 1 percent of the SON and
4 ± 5 percent of the SCOD from samples of PASE. Anion exchange at
neutral pH removed 12 ± 4 percent of the SON and 30 ± 4 percent of
the SCOD from samples of PASE. Cation exchange removed significantly
more SON than SCOD, while anion exchange removed significantly more
SCOD than SON at neutral pH.
4. With a cation-exchange resin, decreasing pH sharply increased the
removal of both SON and SCOD, the percentage removed being signif-
icantly greater for SON than SCOD. At pH 2.0 the cationic resin
66
-------
TABLE 35. SUMMARY OF SON AND SCOD REMOVAL BY ION EXCHANGE
*
Resin
Cationic
Cationic
Cationic
Cationic
Cationic
Cationic
Anionic
Anionic
Anionic
Anionic
Anionic
Anionic
Cationic
Cationic
Cationic
Cationic
Cationic
Cationic
pH
8.0
7.0
8.0
7.9
8.0
8.0
7.5
7.0
7.5
7.4
7.6
8.0
2.0
2.0
3.2
2.0
2.0
2.0
i
Initial
SON
Concentration,
mg/1
1.27
1.53
1.70
1.32
1.25
1.21
1.27
1.53
1.70
1.32
1.25
1.21
1.27
1.30
1.32
1.59
Percent
SON
Removed
9
12
12
12
12
11
Avg. 11 ± 1
12
10
10
20
9
9
Avg. 12 ± 4
47
38
44
39
1.25 i 32
1.53
50
Avg. 42 ± 7
Initial
SCOD
Concentration ,
mg/1
23
27
29
24
22
22
23
27
29
24
22
22
23
25
24
28
22
-
Percent
SCOD
Removed
5
3
0
12
2
0
Avg. 4 ± 5
34
23
32
32
33
27
Avg. 30 ± 4
18
23
26
24
18
\
Avg. 22 + 4
A
15 g/1 of resin used for all samples.
removed 42 ± 7 percent of the SON and 22 ± 4 percent of the SCOD
from samples of PASE.
Sequential contacting of PASE with ion-exchange resins revealed that
the fractions of SON and SCOD removed by the cationic resin at pH 2
and by the anionic resin at pH 12 were significantly different (99%
CI), as were the fractions of SON and SCOD removed by the individual
resins at neutral pH. However, there appeared to be some overlap-
ping of the fractions removed by the two resins at both high pH and
low pH.
Contacting of activated-carbon effluent with ion-exchange resins at
neutral pH revealed that the cation exchangeable SON was not signif-
icantly adsorbed by activated carbon, a portion of the anion ex-
changeable SON was adsorbed, and the anion exchangeable SCOD was
almost completely adsorbed.
67
-------
CHEMICAL OXIDATION
Background
The chemical oxidants relevant to this study are those generally capable
of oxidizing organic compounds in dilute aqueous solution and those particu-
larly capable of oxidizing soluble nitrogen-containing organics. There are
two ways in which SON can be effectively removed by oxidation: (1) deamina-
tion, in which carbon-nitrogen bonds are broken, and the nitrogen measurable
by the Kjeldahl procedure is changed to inorganic nitrogen, and (2) N-oxida-
tion, in which the nitrogen is raised to a higher oxidation state not detected
by the Kjeldahl analysis for organic nitrogen. It is not within the scope of
this study to identify the end products of oxidation, either quantitatively or
qualitatively, but it is quite possible that in some cases they may be more
undesirable environmentally than the compounds from which they are formed.
The chemical oxidants evaluated were sodium hypochlorite, ozone, potas-
sium permanganate, and hydrogen peroxide, all of which are capable of oxidiz-
ing certain organic functional groups in aqueous solution. Chlorine and ozone
have been used extensively for the disinfection of drinking water for many
years, as well as for taste, odor, and color removal, generally involving dos-
ages in the milligram-per-liter range.
Chlorine has been the primary disinfectant in the United States, and is
added variously as the hypochlorite salt, as chlorine gas, or as chlorine
dioxide, depending upon availability and economics. It has long been recog-
nized that nitrogenous organic compounds interfere in this chlorination pro-
cess, due to their slow reaction with chlorine [131]. Whereas some compounds,
such as ammonia, present a rapidly satisfied chlorine demand, organic nitro-
gen imposes a slowly satisfied demand which requires a combination of an ex-
cess of chlorine and an extended contact time to ensure a stable chlorine
residual.
In recent years, chlorine has been increasingly applied in the treatment
of wastewater for control of noxious odors, to remove color, or to decrease
coliform counts. In the tertiary treatment process of breakpoint chlorina-
tion, chlorine is used to remove ammonia from secondary effluent. This gen-
erally requires a chlorine to ammonia nitrogen ratio of about 10 to 1, compared
to a theoretical ratio of 7.5 to 1. The difference is attributed to the
presence of organic and inorganic impurities which exert a chlorine demand,
and for this reason more highly treated wastewaters generally exert a de-
creased chlorine demand.
The effect of breakpoint chlorination on organic matter has received re-
latively little attention, primarily because its primary goal is ammonia re-
moval. It is generally found to accomplish a small reduction in COD, but
there is disagreement as to its effect on the organic nitrogen present in
secondary effluent. Lawrence et al. [134] found a significant decrease in
organic nitrogen below the breakpoint ratio, and complete removal at the
breakpoint. Parkin and McCarty [73] found increasing SON removal with in-
creasing chlorine concentration far in excess of the breakpoint, with 42 per-
cent removal in the presence of 130 mg/1 free residual chlorine. In both
68
-------
studies, the samples were dechlorinated by bubbling with nitrogen gas at pH 2,
a procedure which is not likely to dechlorinate organo-chloro compounds.
Pressley et al. [135] found insignificant removal of organic nitrogen by
breakpoint chlorination.
A number of investigators have studied the effects of chlorination on
specific compounds, many of which may be present in some form in secondary ef-
fluent. Wright [132,133] studied the reaction of hypochlorite with amino
acids and proteins and concluded that (1) hypochlorite can act either as an
oxidizing or as a chlorinating agent, (2) acidity favors chlorination and al-
kalinity favors oxidation, and (3) the chloramino derivatives formed by
chlorination are of variable stability.
Murphy et al. [136] and Sung [137] studied the chlorination of a number
of organic compounds thought to be representative of those present in secon-
dary effluent. Among the compounds Murphy and his co-workers found to be
easily chlorinated were amines, pyrrole, phenols, amino groups, aldehydes, and
ketones, while Sung found that amino acids, uric acid, tannic acid, and humic
acid interfered in the disinfection of wastewater by chlorine.
Taras [138,139] studied the effect of chlorine on a number of specific
nitrogenous compounds, mostly amino acids, peptides, and proteins, and meas-
ured the decrease in albumoid [3] and total nitrogen. Using a chlorine dose
of between 0.5 and 1.0 g/1, he found reductions in total nitrogen of about 50
percent for many amino acids, and 10-20 percent for proteins after a 24-hour
contact time. A contact time of one hour exhibited slightly smaller removals
of amino acid nitrogen, but almost negligible removal of protein nitrogen.
Morris [141] classified the reactions of hypochlorite with organics in
dilute aqueous solution into four categories: (1) addition to olefinic bonds;
(2) activated ionic substitution, exemplified by the chlorination of phenol
and the haloform reaction; (3) oxidation, with reduction of hypochlorite to
chloride; and (4) substitution of chlorine for hydrogen on a nitrogen atom.
He stated the chlorine which reacts to substitute on nitrogen does not lose
its oxidizing capacity; this is supported by the finding of Sung [137] that a
number of organochloro derivatives titrate as chlorine residual.
Morris [141] reported that reaction of chlorine occurs with amines,
amides, amino acids, proteins, heterocyclic compounds, and often proceeds
rapidly, especially with the more basic nitrogen atoms. Alpha amino acids
first give N-chloro derivatives, but then, in general, oxidative deamination
occurs yielding a keto acid and ammonia chloroamine. Of the heterocyclic com-
pounds, pyrimidine reacts readily, forming chloroamines and C-chlorinated de-
rivatives, purine appears to be simultaneously oxidized and chlorinated, and
pyrroles and indoles seem to undergo aromatic substitution; however, none of
these reactions has been investigated at the concentrations and conditions of
water chlorination.
Ozone has been widely used in Europe for many years to disinfect water
supplied. It is one of the strongest oxidants known, considerably stronger
than chlorine, and leaves no residual ions or taste in the water. Ozone de-
composes rapidly to oxygen in water, especially at high pH [142]. In a
69
-------
recent review of the chemistry of ozone in water, Peleg [143] concluded that
the dissociation products of ozone in water may be more powerful oxidizing
agents than ozone itself, and that the hydroxyl radical is mainly responsible
for the high oxidative potential of ozone in water.
In recent years, a number of investigators have studied the effect of
ozone on raw, treated, and industrial wastewaters. Roan et al. [144] ozonated
raw wastewater, secondary effluent, lime-clarified and filtered raw and se-
condary wastewaters, carbon-treated wastewater, and chlorinated and carbon-
treated wastewaters. In all effluents except raw wastewater, 100 mg/1 of
ozone produced at least 70 percent removal of COD. Variable amounts of or-
ganic nitrogen and ammonia were oxidized at pH 7, and organic and ammonia
oxidation increased with increased pretreatment and with increased pH.
Nebel et al. [145] treated a municipal/industrial secondary effluent with
15 mg/1 of ozone in a froth flotation process and found removals of 29 percent
of COD, 23 percent of SCOD, 79 percent of color, 70 percent of turbidity, 78
percent of phosphate, 12 percent of nitrate, and suspended solids down to 0-2
mg/1. TOC reduction was not significant and BOD reduction averaged 15 percent,
but sometimes the BOD increased during treatment, indicating that ozone can
oxidize refractory compounds to biodegradable ones.
Kirk et al. [146] studied the ozonation of secondary effluents and con-
cluded that almost all of the COD could ultimately be oxidized by ozone, but
that after 50-70 percent reduction, the reaction rates become extremely exten-
ded. The pH of the effluent following ozone treatment was found to always be
nearer neutrality than the influent pH, and increased COD removal was found
with increased initial pH for a one-hour contact time. Oxidation of ammonia
was found to occur only at high pH.
Singer and Zilli [147] investigated the oxidation of ammonia with ozone,
and found the reaction to be first order with respect to ammonia concentra-
tion, the rate increasing with increasing pH. Ozonation of secondary efflu-
ent resulted in about 30 percent removal of COD, but insignificant ammonia re-
moval at neutral pH. Raising and buffering the pH of the effluent at 9.0 re-
sulted in rapid oxidation of the ammonia, but no reduction in COD.
Bailey has written several reviews concerning the reactions of ozone with
organic compounds [154,155,156]. He reported that various organic substances
with a nucleophilic atom in their structure were readily oxidized by ozone:
tertiary anines to amine oxides, and primary and secondary amines to nitro
compounds, nitroxides, and various decomposition products. The oxidation of
certain amines proceeded faster than that of olefinic double bonds. The ma-
jority of ozone-related research has involved solvents other than water and
temperatures other than 20°C. In aqueous media, it appears that the major
reaction of ozone with amines is side-chain oxidation rather than nitroxide
formation, and in acidic solutions amines are unreactive toward ozone due to
the unavailability of the nitrogen electron pair. Various amino acids and
proteins have been ozonized in water solution, but the ozone attack appears
to occur at sulphur or aromatic or heterocyclic unsaturated carbon-carbon
bonds rather than nitrogen. The carbon-nitrogen double bond of many com-
pounds is readily cleaved by ozone.
70
-------
The ozonation of amines has been investigated by a number of researchers
[158,159,160,161,162,163,164,165]. Hewes and Davison [160] ozonated a water
containing diisopropylamine at pH 7 and found 80-85 percent removal of the
amine and 50 percent removal of the amine COD, and suggested that ozone at-
tacked the amino group preferentially.
Other nitrogen-containing compounds and functional groups have also been
examined for their reactivity toward ozone. Erickson et al. [166] found that
carbon-nitrogen double bonds were as reactive toward ozone as carbon-carbon
bonds if they were properly substituted. Wibaut [167] examined the mechanisms
of ozone attack on pyrroles. Mudd et al. [168] found the order of suscepti-
bility of amino acids in aqueous solution to ozone attack to be cysteine,
methionine, tryptophano, tyrosine, histidine, cystine, and phenylalanine; the
other amino acids were unaffected by ozone. Aryl diamines, used as antiozo-
nants in rubber products, were studied by Delman et al. [169], who found that
their protective capacity decreased as the size or number of the N-hydrogen
substituents increased.
Mosher [170] found that aromatic double bonds added ozone at about 10 per-
cent of the rate of isolated double bonds, and that some highly branched ole-
fins were oxidized readily, while others reacted at about one tenth the rate
of ordinary olefinic double bonds. Thus structural relationships are very
important in the reactivity of double bonds toward ozone.
Permanganate and hydrogen peroxide have not received much attention as
oxidants in wastewater treatment due to their inferior oxidizing ability in
water as compared to chlorine and ozone, but they are known to oxidize organic
matter under extreme conditions and to attack certain functional groups read-
ily. Medalia [171] carried out "COD-like" tests on pure compounds, comparing
permanganate, hydrogen peroxide catalyzed by iron salts, and dichromate, and
found peroxide and permanganate to perform only a fraction of the oxidation
achieved with dichromate. Felbeck [172] reported that soil humic acids were
at least partially degraded by permanganate and peroxide under extreme condi-
tions.
Although potassium permanganate is widely used for taste and odor control
and for iron and manganese removal, there is little information available re-
garding the effect of permanganate on organics under wastewater treatment con-
ditions. Spicher and Skrinde studied potassium permanganate oxidation of
organic compounds [173]. They found that permanganate oxidation of organic
refractories is more effective under alkaline than under acidic conditions,
and that the inorganic constituents of natural river water had no apparent
effect on the reaction. They also found the functional group of an organic
compound to be important in the reaction, amines, aldehydes, aromatic alco-
hols, and keto acids being readily oxidized, while carboxylic acids, ketones.
aliphatic alcohols, hydroxy acids, amino acids, and some esters were not read-
ily oxidized. The double bond of alkenes was readily oxidized, aldehydes were
oxidized to acids, and the benzene rings of phenol and aniline were apparently
broken by oxidation with permanganate.
Parkin and McCarty [73] studied the effect of permanganate and hydrogen
peroxide on the SON of treated agricultural wastewaters, and found significant
71
-------
removals of SON; however, the results are questionable since the oxidants were
not reduced prior to SON analysis. As found by Spicher and Skrinde [125],
permanganate oxidation is temperature dependent, as are most oxidation reac-
tions.
Experimental Procedures
All of the samples used in the chemical oxidation experiments were Pall-
filtered PASE, either fresh or stored at 4°C, and all experiments were per-
formed at 20°C. The pH of the samples was adjusted with 1.0 and 6.0 N solu-
tions of both hydrochloric acid and sodium hydroxide.
Chlorine was added to the samples as sodium hypochlorite. Chlorine,
hydrogen peroxide, and potassium permanganate were generally added to the
samples from 10 g/1 working solutions, although chlorine was sometimes added
from a 5.9 percent stock solution. After a contact time of one hour these
oxidants were neutralized with a quantity of 100 g/1 sodium sulfite solution
sufficient to reduce 1.33 times the initial quantity of oxidant introduced
into the sample, leaving an excess of sodium sulfite in the samples. The pH
was then adjusted to 7.5-8.0. When reducing residual permanganate, sufficient
acid and sulfite were added to allow reduction of the permanganate entirely to
the magnanous ion.
The excess sodium sulfite was not removed before the Kjeldahl nitrogen
analysis, since it was found that it produced no changes in SON analysis at
concentrations as high as 400 mg/1. The removal of excess sulfite ion prior
to SCOD analysis was accomplished by adding 10 mg/1 of cobaltous chloride,
followed by 15 minutes of vigorous aeration with cotton/fiberglass-filtered
air introduced through a porous stone diffuser.
One set of chlorinated samples was dechlorinated by the method of
Lawrence et al. [134] for comparison. After the one-hour contact time, the
pH of these samples was adjusted to 2.0 and they were then aerated for 2
hours with nitrogen gas.
Ozone was generated with a Welsbach Laboratory Ozonator (Model T-408) ,
using pure oxygen feed gas at 0.5 1/min (STP) and 5.5 x 106 pa (8.0 psi) .
Ozone was generated with 100 volts of electricity and carried in the oxygen
gas streams through Tygon tubing to the bottom of a plexiglass column where
it was released through a glass diffuser into the sample. The plexiglass
column was 50 mm in diameter, 1300 mm in overall length, and filled to a depth
of approximately 1020 mm with 2.0 liters of sample. The rubber stoppers on
the top and bottom of the column were covered with transparent tape (Scotch)
to prevent oxidation of the rubber.
The gas escaping from the sample was carried through Tygon tubing and
released into a potassium iodide trap through a glass diffuser. The trap was
made of a glass tube 29 mm in diameter, 520 mm in length, and was filled to a
depth of 380 mm with approximately 250 ml of a 2-percent potassium iodide
solution.
Ozone dose was controlled by adjusting ozonation time at constant gas
72
-------
flow rate. For calibration purposes, the flow of gas was short-circuited di-
rectly to the potassium iodide trap. The output of the ozonator was checked
frequently during use, and was quite constant on any given day. For the first
ozone experiment, calibration checks showed that the ozonator produced an av-
erage of 579 ± 4 mg of ozone during four separate 20-minute periods.
Samples contacted with ozone were not treated with sodium sulfite after
it was determined that the ozone residual was very short-lived and that the
addition of sodium sulfite, as described above, had no effect on the results
of the SON and SCOD analyses.
Results and Discussion
The results of an experiment in which the removal of SON by several oxi-
dants was studied as a function of applied dose and pH are presented in Table
36. Hypochlorite produced no significant removal of SON and only minimal, if
any, removal of SCOD at concentrations from 10-100 mg/1, well below the break-
point. Permanganate removed 28 percent of the SON at an optimum pH of 10
and with a dose of 100 mg/1. Since permanganate removed 26 percent of the SON
at a dose of 50 mg/1, it does not appear that higher doses will be effective
in removing significantly more SON. Peroxide removed 19 percent of the SON at
an optimum pH of 11 with a dose of 100 mg/1, although doses of more than 20
mg/1 did not remove significantly more SON. Neither peroxide nor permanganate
removed a significant (99% CI) amount of SCOD.
These results indicate that some nitrogenous functional groups, most
likely the amino groups, are readily and irreversibly oxidized by permanganate
and peroxide, and that the effect is greater with permanganate than with per-
oxide. Spicher and Skrinde [173] found that permanganate reacted readily with
amino groups, but not with amino acids, resulting in deamination to form am-
monia and organic acids. It is likely that this is also the major reaction
of permanganate and peroxide with SON.
If permanganate or peroxide oxidize SCOD, the reaction is probably lim-
ited to olefinic double bonds, which are probably too few in number to be
reflected in the SCOD. The presence of a high concentration of ammonia nitro-
gen interferes with the action of hypochlorite, and if organic nitrogen is
chlorinated, the reaction is reversed by the addition of sodium sulfite.
Since others [73,134] have reported removal of SON at chlorine concentra-
tions below the breakpoint, using the method of Lawrence et al. [134] for de-
chlorination, this method was investigated with the results shown in Table 37.
At chlorine doses below the breakpoint, it was found that 27 percent of the
SON could be reversibly chlorinated, which is, interestingly, about the same
percentage of SON oxidized by permanganate. Past the breakpoint, hypochlorite
oxidation of the SON appears to be irreversible, as shown by the fact that
both methods of dechlorination produced identical results.
*
The breakpoint is the chlorine dosage required for complete oxidation of am-
monia, which requires a chlorine to ammonia ratio of about 8 for PASE [175].
73
-------
TABLE 36. REMOVAL OF SON AND SCOP BY CHLORINE, PERMANGANATE, AND PEROXIDE
Oxidant
(Control)
Hypochlorite
Hypochlorite
Hypochlorite
Hypochlorite
Permanganate
Permanganate
Permanganate
Permanganate
Permanganate
Permanganate
Permanganate
Permanganate
Peroxide
Peroxide
Peroxide
Peroxide
Peroxide
Peroxide
Peroxide
_ a
Dose,
mg/1
-
10
20
50
100
10
20
50
100
100
100
100
100
10
20
50
100
100
100
100
PH
7.4
7.0
7.0
7.0
7.0
10.0
10.0
10.0
10 = 0
9.0
10.0
11.0
12.0
11.0
11.0
11.0
11.0
9.0
7.0
3.0
SON,
mg/1
1.19
1.18
1.19
1.17
1.18
1.03
0.93
0.88
0.86
1.02
0.86
0.92
0.94
1.06
0.98
0.97
0.96
0.98
1.10
1.05
Percent
SON b
Removed
-
1
0
2
1
14
22
26
28
14
28
23
21
11
17
18
19
17
7
12
SCOD,
mg/1
25
25
25
24
24
-
-
-
-
25
26
24
25
24
25
26
25
25
25
25
Percent
SCOD c
Removed
-
0
0
3
6
-
-
-
-
2
0
5
2
3
0
0
2
2
2
0
*a
As chlorine, potassium permanganate, and hydrogen peroxide.
Differences in SON removal greater than 7% and 9% are 95% and 99%
significant, respectively.
Differences in SCOD removal greater than 4% and 5% are 95% and 99%
significant, respectively.
The removal of SON by breakpoint chlorination, as a function of dose and
pH, is shown in Table 38, and indicates that just beyond the breakpoint (Cl/N
= 10.4), all of the chlorine-oxidizable SCOD has been oxidized, but that some
of the SON requires a higher dose to be irreversibly oxidized. A pH of 7 ap-
pears to be optimum for removal of organics by breakpoint chlorination, but
removal is not very dependent upon pH.
Table 39 results indicate the effect of ozone dose and pH on SON removal.
An increase in pH from 7.0 to 10.0 resulted in a significant (99% CI) increase
in SON and SCOD removal, and a large increase in ammonia removal. The in-
creases in SON and SCOD removal with increasing pH were associated with the
increase in absorbed ozone. An undetermined optimum appears to lie at some
pH below 12. Removal of SON was slight (95% CI) at doses below 100 mg/1 and
SCOD removal increased with increasing ozone dose. The SON removal achieved
74
-------
TABLE 37. EFFECT OF DECHLORINATION METHOD OF LAWRENCE ET AL. [134]
Chlorine
dose,
mg/1
&
Sample 1
Sample 2
*
Same sample as
**
Dechlorinated
0
10
20
50
100
0
472
472
SON,
mg/1
1.27
1.17
1.05
0.93
0.93
1.51
0.81
**
0.83
Percent
SON
Removed
8
17
27
27
46
**
45
in Table 36.
with sodium sulfite.
by ozonation of Sample 1, Table 39, was uncharacteristically high, as can be
seen by comparison of the removals of SON and SCOD by high dosages of chlorine
and ozone with several samples, as shown in Table 40.
On the average, chlorine achieved 42 percent SON and 18 percent SCOD re-
movals, almost the reverse of that achieved by ozone of 14 percent SON and 46
percent SCOD. It seemed likely from these results that the two were removing
different fractions of the organic matter, or attacking different functional
groups. The results of two experiments in which chlorine and ozone were used
in sequence (Table 41) shows that when ozone is preceded by chlorine, the
total removal of SON and SCOD is equal to the sum of the removals accomplished
by each oxidant individually. However, when ozonation precedes chlorination,
SON removal is greatly increased, while SCOD removal is not increased addi-
tionally by the chlorination.
Thus, it appears that the major reaction in the ozonation of SON is side-
chain oxidation, which "frees" the nitrogen to a form which can easily be
chlorinated, perhaps as amino nitrogen. Additional SCOD is not removed by
the chlorine, since ozone is a stronger oxidant and has already oxidized all
of the easily oxidizable organics, with the exception of a large part of the
readily chlorinated SON, which appears immune to attack by ozone.
Since cobalt ion catalyzes the reaction of sulfite ion with oxygen, it
was considered possible that it could act as a catalyst for ozone, but its
only effect was to increase the rate of decomposition of the ozone, with no
effect on SON or SCOD removal. Bicarbonate ion was found by Denisov et al.
75
-------
TABLE 38. REMOVAL OF SON AND SCOD BY BREAKPOINT CHLORINATION
Sample 1
SON
(5%,7%)b
SCOD
(3%, 5%)
Sample 2
SON
(6%,8%)b
SCOD
(4%,5%)b
Chlorine
Dose,
mg/1
0
118
236
354
472
0
472
472
472
472
Residual
Chlorine
mg/1
83
34
99
245
245
177
284
333
Chlorine:
Nitrogen
Ratio
5.2
10.4
15.6
20.8
22.6
22.6
22.6
22.6
PH
7.0
7.0
7.0
7.0
7.8
3.0
7.0
10.0
12.0
SON,
mg/1
1.55
1.55
1.13
0.94
0.94
1.45
1.02
0.89
0.95
Percent
SON
Removed
0
27
39
39
30
39
34
SCOD,
mg/1
28
27
23
23
23
26
23
21
21
23
Percent
SCOD
Removed
3
18
18
18
11
18
18
11
NH3-N
mg/1
22.7
17.8
0.7
1.4
1.1
21.0
4.7
1.2
0.3
3.2
N03~N1
mg/1
0.2a
0.4a
l.la
1.7a
1.8a
0.0
0.8
1.8
0.2
0.3
N02-N
mg/1
-
-
0.02
b.oo
0.00
0.01
0.43
Includes nitrite nitrogen.
'Differences in removal to be exceeded for 95% and 99% significance, respectively.
-------
TABLE 39. REMOVAL OF SON AND SCOD BY OZONATION
Sample 1
SON
(5%,6%)b
SCOD
(3%,4%)b
Sample 2
SON
(5%,7%)b
SCOD
(3%,5%)b
Initial
pH
7.7
7.7
3.0
10.0
12.0
7.4
7.4
7.4
7.4
7.4
Final
PH
7.7
3.3
9.1
12.3
7.4
7.3
7.3
7.2
Ozone
Applied,
mg/1
289
362
289
289
14.9
74.6
149
299
Ozone
Absorbed,
mg/1
114
86
193
229
13.7
50.0
79
112
SON,
mg/1
1.76
1.26
1.34
1.10
1.07
1.54
1.43
1.44
1.45
1.32
Percent
SON
Removed
28
24
37
39
7
7
6
14
SCOD,
mg/1
34
19
25
17
18
28
24
19
18
16
Percent
SCOD
Removed
45
25
50
46
14
33
36
44
mg/1
27.2
23.7
25.8
14.8
20.3
22.9
22.9
21.8
21.7
19.7
Percent
NH3-N
Removed
13
5
46
25
0
5
5
14
a
mg/1
0.6
3.7
1.5
12.3
6.5
0.4
0.4
1.3
2.1
3.4
Includes nitrite nitrogen.
Differences in removal to be exceeded for 95% and 99% significance, respectively.
-------
TABLE 40. COMPARISON OF SON AND SCOD REMOVALS BY CHLORINE AND OZONE
Oxidant
Chlorine
Chlorine
Chlorine
Chlorine*
Chlorine**
Ozone
Ozone
Ozone
Ozone
Ozone
Ozone*^
Ozone**
Dose,
mg/1
472
472
472
472
472
114
112
44
104
115
108
106
Initial
SON,
mg/1
1.51
1.55
1.45
1.54
1.54
1.76
1.54
1.16
1.08
1.25
1.54
1.54
*,** indicate sama sample.
Final
SON,
mg/1
0.83
0.94
0.89
0.80
0.91
Percent
SON
Removed
45
39
39
48
41
(Avg.) 42
1.26
1.32
1.03
1.02
1.14
1.26
1.38
28
14
11
6
9
18
11
(Avg.) 14
Initial
SCOD,
mg/1
_
28
26
24
29
34
28
24
25
22
24
29
Final
SCOD,
mg/1
_
23
21
18
26
Percent
SCOD
Removed
_
18
18
28
10
(Avg.) 18
19
16
11
14
11
13
17
45
44
53
45
51
45
41
(Avg.) 46
[174] to inhibit the oxidation of alcohols by hydrogen peroxide, but it was
found to have no effect on ozonation other than to buffer the pH of the waste-
water during ozonation.
Summary
1.
When added to a sample of PASE at pH 7.0, chlorine dosages less than
100 mg/1 (below the breakpoint) produced no significant removal of
SON and very little (6 ± 2%) removal of SCOD.
2. When added to a sample of PASE adjusted to various values of pH,
potassium permanganate and hydrogen peroxide removed a maximum of
28 ± 3 and 19 ± 3 percent of the SON, respectively, at pH 10.0 and
11.0, respectively, and produced no significant removal of SCOD.
3. A chlorine to ammonia nitrogen ratio of approximately 20:1 removed
42 ± 4 percent of the SON and 18 ± 7 percent of the SCOD from PASE
A significantly (99% CI) lower amount of oxidation of SON was found
for a 10.4:1 ratio, but SCOD removal was unaffected by an increase
above a ratio of about 10.
4.
neutral H
of the
doses mostly in excess of 100 mg/1 and at
78
-------
5. Ozone and chlorine were found to remove 14 ± 7 and 42 + 4 percent of
SON, respectively, and these fractions were found to be mutually ex-
clusive. Removal of SON by chlorine was increased about 30 percent
by preozonation.
PROCESSES IN SEQUENCE
Background
Sequences of processes not discussed previously are presented herein.
Sequences of one coagulant with another, one ion-exchange resin with another
or with activated carbon, or one oxidant with another were discussed pre-
viously.
The goals of this portion of the study were: (1) to evaluate the removal
of SON by sequences of processes used in wastewater treatment; (2) to distin-
guish differences and similarities in the fractions of SON removed by some of
the different processes; and (3) to identify any synergistic or antagonistic
effects arising from the sequential use of processes which would make it de-
sirable to arrange them in a particular order or substitute one process for
another. It is not within the scope of this study to attempt to optimize SON
removal or to evaluate all possible sequences of treatment processes. Many
of the processes are used primarily for other purposes than SON and SCOD re-
moval, and their sequence is generally dictated by other considerations,
although great flexibility exists in designing such systems.
Ozone is a relative newcomer to the wastewater treatment field. Many of
the experiments described in this chapter dealt with ozone as one of a se-
quence of treatment processes.
There is little or no information in the literature relating to the re-
moval of SON by sequences of treatment processes. Schertenleib [175] found
that overall removal of COD was unaffected by the order of breakpoint chlori-
nation and activated-carbon adsorption in a laboratory study. No evidence
was found in the literature for unexpected antagonistic or synergistic effects
in the removal of organics by a sequence of advanced waste-treatment processes.
Spicher and Skrinde [125] found that the inorganic constituents of natu-
ral river water had no apparent effect on the reaction of potassium perman-
ganate with organic refractories, but there is little or no information avail-
able on the effect of the large number and kinds of inorganic ions in secon-
dary effluent on the oxidation of organic contaminants. If inorganic ions do
catalyze or inhibit oxidation reactions, the position of ion exchange in a
wastewater treatment sequence may be highly important.
To evaluate the removal of SON by sequences of processes currently being
used or considered for advanced wastewater treatment, samples were obtained
from the South Tahoe Public Utility District Water Reclamation Plant, and
other samples were generated in a bench-scale laboratory advanced wastewater-
treatment plant, using the design criteria of the planned Palo Alto Water
Reclamation Facility [51].
79
-------
TABLE 41. REMOVAL OF SON AND SCOP BY COMBINATIONS OF CHLORINE AND OZONE
Treatment
Sample 1 Control
SON Chlorine
(5% 77)
°' Chlorine, then Ozone
SCOD
/TV /7%) Chlorine
SCOD
(L7 ci°/'\° Ozone, then Chlorine
Final
pH
7.4
7.0
6.9
7.2
7.0
7.8
7.5
7.0
7.0
Ozone
Applied,
mg/1
311
301
301
_
277
277
Ozone
Consumed,
mg/1
204
106
106
_
108
_
108
Residual
Oxidant ,
as Chlorine
.
252
11
—
264
_
_
206
241
SON,
mg/1
1.54
0.91
0.74
1.38
0.41
1.54
1.26
0.80
0.38
Percent
SON
Removed
—
41
52
11
74
_
18
48
75
SCOD,
mg/1
29
26
14
17
18
24
13
18
10
Percent
SCOD
Removed
_
10
51
41
37
_
45
28
51
n
Chlorine applied at 472 mg/1 at pH 7 for 1 hour; ozonation begun with an initial pH of 7.5.
Differences in removal to be exceeded for 95% and 99% significance, respectively.
oo
o
-------
Experimental Procedures
Samples from the Tahoe plant were collected and stored as described in
Section 5. The laboratory-scale advanced wastewater treatment consisted of
lime clarification (350 mg/1 CaO), followed by ammonia stripping (1600 m3 air
per m3 wastewater, recarbonation (single stage), mixed-media filtration (240
m-VmZ-day), breakpoint chlorination (8 mg Cl2/mg NELj-N), and finally activated-
carbon adsorption. The activated-carbon column was 50 mm in diameter, packed
to a depth of 0.6 m with Nuchar WV-G 12 x 40 mesh activated carbon, and the
flow rate was 240 m3/m2-day. A complete description of the laboratory units
and their operation is given elsewhere [51]. The samples used in the labora-
tory reclamation study were unfiltered chlorinated PASE.
Experiments involving ozone in sequence with alum, lime, ferric chloride,
activated carbon, and ion exchange were carried out on Pall-filtered PASE,
following the procedures previously outlined, except where noted.
The artificial secondary effluent used in evaluating the interference
of ozone and ferric chloride in the Kjeldahl test contained 27 mg/1 potassium
dihydrogen phosphate, 10 mg/1 potassium bicarbonate, 122 mg/1 magnesium chlo-
ride, 162 mg/1 calcium chloride, 8.4 mg/1 sodium fluoride, 60 mg/1 sodium bi-
carbonate, 128 mg/1 sodium sulfate, 181 mg/1 sodium chloride, and 54 mg/1 am-
monium chloride. This artificial effluent contained approximately the concen-
trations of the major inorganic ions present in PASE, and no organic matter.
Results and Discussion
The results of the sampling at the Tahoe plant are presented in Table 42.
The SON value for secondary effluent is an anomaly. The removal of both SON
and SCOD was much less than anticipated, and this is attributed to the fact
that plant operation was geared for inorganic nutrient removal rather than
removal of organics. Lime was added primarily to raise the pH for ammonia
stripping (which was not in operation during the sampling for this study),
only half of the wastewater was chlorinated, and the activated-carbon beds
were used primarily for dechlorination and were not being regenerated suffi-
ciently for optimum removal of organics. Thus, the results indicate that good
removal of organics by advanced treatment may not be achieved unless the plant
is operated specifically for that purpose.
The results of the laboratory wastewater treatment study, presented in
Table 43, show an average organic nitrogen removal of 88 percent and an aver-
age COD removal of 92 percent for three chlorinated PASE samples. Although
removals may be somewhat lower in the planned full-scale plant, it appears
that advanced wastewater treatment using currently available technology can
remove a large percentage of the SON and SCOD from secondary effluent. The
data also show that only a small percentage of SON is removed by breakpoint
chlorination when there is no large excess of chlorine present. Thus, the
majority of the SON removal occurs during coagulation and activated-carbon
adsorption.
It was also found during the course of the laboratory study that recarbo-
nation (either single-stage or two-stage), varying the flow rate through the
81
-------
TABLE 42. REMOVAL OF SON AND SCOD BY THE SOUTH TAHOE P.U.D. WATER RECLAMATION PLANT
Sample
Plant Influent
Secondary Effluent
Lime- Treated Effluent
Chlorinated Effluent**
Filtered Effluent
Activated- Carbon
Effluent
Organic- N,
mg/1
9.3
4.2
1.7
1.7
1.4
1.2
SON,
mg/1
2.9
1.17
1.55
1.39
1.32
1.07
Percent
SON
*
Removed
-
-
-
10
15
31
COD,
mg/1
388
84
32
29
24
20
SCOD,
mg/1
123
25
25
24
24
20
Percent
SCOD ^
Removed
-
-
-
3
4
19
Soluble
TOC,
mg/1
-
10
11
11
10
10
BOD5,
mg/1
186
41
5
-
-
1
^Assuming lime- treated value = 1007o. SON removals greater than 5% and 7% and SCOD removal
than 4% and 5% are 95% and 99% significant, respectively.
•**
50 percent of flow was breakpoint chlorinated, then recombined with non- chlorinated flow.
Soluble
BOD.,
mg/1
64
2
3
-
-
< 1
greater
00
IS3
-------
TABLE 43. RESULTS OF LABORATORY ADVANCED WASTEWATER TREATMENT
m
i— i
g
CO
m 1-1
I W
** g
CM
6 CO
CO
r-4 .
.. r-H
§ «
O !Z
2! O
CO
Parameter
Org-N, mg/1
COD, rag/1
NH3~N, mg/1
pH
Org-N, mg/1
COD, mg/1
NH3-N, mg/1
pH
Org-N, mg/1
COD, mg/1
NH3-N, mg/1
pH
PASE
2.32
38
24.9
7.5
3.12
45
24.2
7.4
1.95
35
22.0
7.7
Effluent from
Lime
Treatment
1.21
24
24.4
11.5
0.87
16
18.2
11.5
0.90
20
19.3
11.4
Ammonia
Stripping
1.10
24
13.7
11.3
0.83
17
9.3
11.2
0.85
21
11.1
11.2
Recarbo-
nation
-
-
-
8.2
-
-
7.9
0.93
22
12.9
6.9
Filtra-
tion
1.18
20
13.3
7.6
0.79
16
12.1
8.1
0.82
20
13.3
-
Breakpoint
Chlori-
nation
1.06
18
0.1
7.3
0.63
13
0.0
7.1
0.79
17
0.02
-
Activated-
Carbon
Adsorption
0.49
6
1.6
8.1
0.03
0
0.0
9.3
0.27
2
1.7
-
Percent
Remove c
79
84
94
-
99
100
100
-
86
93
92
-
00
CO
-------
mixed-media filter, and injecting 5 mg/1 of ozone before or after lime treat-
ment all had no significant effect on SON or SCOD removal.
Ozonation was studied in sequence with various chemical coagulants, with
the results presented in Table 44. Generally, SON removal was significantly
(99% CI) greater when ozonation followed coagulation than when it preceded
coagulation. The same was generally true for SCOD removal, except when ozona-
tion at neutral pH preceded lime treatment. Ozonation at high pH following
lime treatment was highly effective in removing SON and SCOD, but the absorbed
dose of ozone was unusually high.
When ferric chloride coagulation was followed by ozonation, there ini-
tially appeared to be a synergistic effect on SON removal, since removal ex-
ceeded the sum of the removals by the individual processes. However, it was
discovered that this effect was due to an interference in the Kjeldahl analy-
sis caused by ozonation of the supernatant liquid from ferric chloride coagu-
lation. This was verified with "artificial" secondary effluent, having an
ionic makeup closely resembling PASE and containing no organic matter, as
previously described. A sample of this artificial effluent was coagulated
with ferric chloride and settled, and then the supernatant liquid was split
into two portions. One portion was ozonated and adsorbed 65 mg/1 of ozone.
A 250-mg/l volume of each portion was then added to 250 ml of PASE and the
resultant 500-ml sample was analyzed for SON. Recovery was 99 percent for
the PASE sample combined with the coagulated supernatant, but was only 73
percent for the PASE sample combined with the ozonated ferric chloride super-
natant. Thus, ozonated ferric chloride supernatant interferes in the Kjeldahl
analysis.
It is possible that this interference was caused by the generation of
perchlorate ion, catalyzed by the presence of iron or iron oxide. Perchlo-
rate ion was found to strongly interfere in the Kjeldahl analysis; however,
it was not determined whether the perchlorate ion could be formed by ozona-
tion of ferric chloride supernatant liquid.
Table 45 shows the results of an experiment in which ozone was used in
various sequences with activated carbon and ion-exchange resins. At neutral
pH, removal of SON and SCOD was independent of the order of treatment, indi-
cating there are neither inhibiting/catalyzing inorganic ions present nor
synergistic/antagonistic effects between the processes. At neutral pH SON
removal was approximately additive for sequences of ozonation and adsorption,
thus indicating that the non-adsorbing fractions of SON were more readily at-
tacked by ozone. This is understandable if adsorption is biased toward
strongly ionized and hydrophobic (non-polar) molecules, while ozone preferen-
tially attacks polar functional groups.
Cation exchange by itself at pH 2 removed more SON (32%) than the se-
quence of ozonation followed by cation exchange at pH 2 (20%) , and ozonation
did not remove a significant amount of SON when following cation exchange at
pH 2. Thus, it appears that ozonation alters .some of the organics which are
removable by low pH cation exchange, and that cation exchange adsorbs a por-
tion of the SON oxidizable by ozone.
84
-------
TABLE 44. REMOVAL OP SON AND SCOP BY OZONE IN SEQUENCE WITH CHEMICAL COAGULATION
Sample
1
2
3
4
f\
Treatment
Control
98 mg/1 ozone5
600 mg/1 alum
400 mg/1 lime
1) 98 mg/1 ozoneb; 2) 600 mg/1 alum
1) 600 mg/1 alum: 2) 93 mg/1 ozoneb
1) 98 mg/1 ozone"; 2) 400 mg/1 lime
1) 400 mg/1 lime; 2) 80 mg/1 ozoneb
Control
104 mg/1 ozone
400 mg/1 lime
1) 104 mg/1 ozone; 2) 400 mg/1 lime
1) 400 mg/1 lime; 2) 202 mg/1 ozone at
pH 11.5
Control
106 mg/1 ozone0
600 mg/1 ferric chloride
1) 106 mg/1 ozonec; 2) 600 mg/1 ferric
chloride
1) 600 mg/1 ferric chloride; 2) 106 mg/1
ozone0
Control
108 mg/1 ozone0
600 mg/1 ferric chloride ,
1) 600 mg/1 ferric chlor.;2) 84 mg/Lozone
pH after
First
Treatment
7.6
7.6
6.0
11.6
7.6
6.0
7.6
11.6
7.7
7.7
11.5
7.7
11.5
7.4
7.2
6.0
7.2
6.0
7.8
7.5
6.0
6.0
pH after
Second
Treatment
_
_
-
-
6.0
7.5
11.6
4.7
_
_
_
11.5
11.3
_
_
-
6.0
7.5
_
_
-
7,5
SON,
mg/1
1.16
1.03
0.81
0.66
0.87
0.62
0.76
0.68
1.08
1.02
0.61
0.69
0.33
1.54
1.38
0.99
0.88
0.59d
1.54
1.26
0.85
O.SS4
Percent
SON e
Removed
«
11
31
43
25
47
35
41
•»
6
43
36
69
_
11
36
43
61d
„
18
45
74d
SCOD,
mg/1
24
11
16
16
8
8
9
11
25
14
16
9
5
29
17
19
15
13
24
13
14
8
Percent
SCOD e
Removed
_
53
33
33
64
66
62
54
^
45
37
62
80
41
34
47
55
45
44
65
aOzone doses are absorbed doses. Probable interference in this analysis.
JpH adjusted to 7.6 prior to ozonation. eConsult Table A-2, Appendix A for percent removals
cpH adjusted to 7.5 prior to ozonation. required for significance.
oo
-------
TABLE 45. REMOVAL OF SON AND SCOP BY OZONE IN SEQUENCE WITH ION EXCHANGE AND ACTIVATED-CARBON ADSORPTION
•
Treatment3
Control
115 mg/1 ozone
400 mg/1 activated carbon
1) 115 mg/1 ozone; 2) 400 mg/1 activated carbon
1) 400 mg/1 activated carbon; 2) 88 mg/1 ozone
15 g/1 anion exchange resin
1) 115 mg/1 ozone; 2) 15 g/1 anion exchange
resin
1) 15 g/1 anion exchange resin; 2) 82 mg/1
ozone
15 g/1 cation exchange resin
1) 115 mg/1 ozone; 2) 15 g/1 cation exchange
resin
1) 15 g/1 cation exchange resin; 2) 114 mg/1
ozone
15 g/1 cation exchange resin
1) 115 mg/1 ozone; 2) 15 g/1 cation exchange
resin
1) 15 g/1 cation exchange resin; 2) 118 mg/1
ozone
pH after
First
Treatment
7.5
7.5
8.0
7.5
8.0
7.6
7.5
7.6
8.0
7.5
8.0
2.0
7.5
2.0
a
Ozone doses are absorbed doses; initial pH for ozonation
pH after
Second
Treatment
_
_
_
7.9
7.4
_
7.6
7.3
_
7.9
7.6
_
2.0
6.8
SON,
mg/1
1.25b
1.14b
1.14
0.99
0.96
1.14
1.10
1.02
1.10
0.93
0.94
0.85
1.00
0.82
Percent
SON
Removed
-
9
9
21
23
9
12
18
12
25
25
32
20
34
SCOD,
mg/1
22b
llb
15
11
15
7
7
22
11
12
18
8
10
Percent
SCOD
«
Removed
-
51
33
58
51
33
69
67
2
51
47
18
62
55
was 7.5; applied ozone dose was 272 mg/1.
Average of 4 analyses.
differences in SON removal greater than 6% and 7%, and differences in SCOD removal greater than 3%
and 4% are 95% and 9970 significant, respectively. Differences in SON removal greater than 0.07 and
0.09 mg/1 and differences in SCOD removal greater than 0.7 and 1.0 mg/1 are 95% and 997o significant,
respectively, due to 4 observations of the control sample.
00
-------
The SON most readily adsorbed by a small dosage of activated carbon ap-
peared distinct from that removed by ozonation (Table 45). In order to eval-
uate this further, the effluent from an activated-carbon column was ozonated.
As shown in Table 46, ozone did not significantly oxidize the SON and SCOD re-
maining after activated-carbon adsorption, and preozonation did not appear to
affect removal of SCOD by activated carbon. Thus,. SCOD which was attacked by
ozone was also removed by activated carbon.
The results of the ferric chloride coagulation experiments, in which
coagulation was carried out sequentially at pH 6.0 and 10.0 (see Chemical
Coagulation) indicated that there are fractions of the organics which are
electrostatically removed, i.e., they behave as anions or cations. Since
such fractions might be expected to be closely related to the material re-
moved by the ion-exchange resins, an experiment was conducted to determine if
this was indeed the case. As shown in Tables 47 and 48, the fraction removed
TABLE 46. REMOVAL OF SON AND SCOD BY OZONATION AND
ACTIVATED-CARBON COLUMN ADSORPTION
Treatment
SON,
mg/1
Percent
SON
Removed6
SCOD,
mg/1
Percent
SCOD
Removed9
Control3
Activated-carbon adsorption
47 mg/1 ozone0
8 mg/1 ozonec
1) 47 mg/1 ozone0; 2) Activated-carbon
adsorption
1) Activated-carbon adsorption;
2) 41 mg/1 ozone
1) Activated-carbon adsorption;
2) 4.5 mg/1 ozone
0.89
0.27
0.94
0.89
0.43
0.41C
0.23
70
0
0
52
54
74
18
3
13
17
3
3
2
85
28
8
83
86
87
Control is lime-treated, air-stripped, recarbonated, and filtered PASE;
treatment parameters were those given for the laboratory advanced waste-
water treatment.
°Carbon column was 50 mm in diameter, packed with 0.6 m of Nuchar WV-G,
12 x 40 mesh; flow rate was 240 m^/m^-day.
•»
"Ozone doses are absorbed doses.
Perhaps an analytical error; value should be less than 0.27 mg/1.
^Differences in SON removal greater than 9% and 12% and differences in
SCOD removal greater than 5% and 7% are 95% and 99% significant, respec-
tively.
87
-------
TABLE 47. REMOVAL OF SON AND SCOD BY FERRIC CHLORIDE COAGULATION AND ION EXCHANGE
Sample
1 (Control)
2
3
4
5
6
7
8
9
First
Treatment3
-
Anion Exchange
Cation Exchange
Ferric Chloride
Ferric Chloride
Ferric Chloride
Ferric Chloride
Ferric Chloride
Ferric Chloride
pH after
First
Treatment
7.6
8.0
8.0
6.0
10.0
6.0
6.0
10.0
10.0
Second
Treatment3
-
-
-
-
-
Anion Exchange
Cation Exchange
Anion Exchange
Cation Exchange
pH after
Second
Treatment
-
-
-
-
-
8.0
8.0
8.0
8.0
SON,
mg/1
1.21
1.10
1.07
0.79
0.83
0.67
0.63
0.73
0.83
Percent
SON
Removed"
-
9
11
35
31
45
48
39
31
SCOD,
mg/1
22
16
23
15
17
9
14
11
17
Percent
SCOD
Removed"
-
27
0
34
24
58
35
49
24
rt
Ferric chloride = 600 mg/1 FeCl3 with settling time of 60 minutes; anion and cation exchange with
15 g/1 of resin.
Differences in SON removal greater than 7% and 9% and differences in SCOD removal greater than 4%
and 6% are 95% and 99% significant, respectively.
00
CD
-------
TABLE 48. REMOVAL OF SON AND SCOP FROM COAGULATED PASE
BY ION EXCHANGE AT NEUTRAL pH
Sample Treatment
Prior to
Ion Exchange
Control
Ferric Chloride
Coagulation, pH 6.0
Ferric Chloride
Coagulation, pH 10.0
Additional
SON
Removed by
Cation
Exchange ,
mg/1
0.14
0.16
0.00
Additional
SCOD
Removed by
Cation
Exchange ,
mg/1
0.0
0.2
0.0
Additional
SON
Removed by
Anion
Exchange ,
mg/1
0.11
0.12
0.10
Additional
SCOD
Removed by
Anion
Exchange ,
mg/1
6.0
5.4
5.6
by the cation-exchange resin was also removed by ferric chloride at pH 10.0.
However, the organics removed by anion exchange are different from those re-
moved by ferric chloride coagulation at pH 6.0. This was not anticipated
since anionic organic molecules were expected to be removed by both proces-
ses. However, anionic molecules which are small and polar are not likely to
be adsorbed on the ferric hydroxide surface, but are the preferred molecules
for removal by anion-exchange resins. Also, the lattice of the resin is an
aromatic and alkyl structure which can adsorb aromatic contaminants under the
proper conditions. Thus it is conceivable that the resin adsorbs aromatic
and small strongly charged molecules, while ferric chloride coagulation at
pH 6.0 removes larger negatively charged molecules.
It was desired to determine whether pretreatment of PASE by any of sev-
eral processes prior to activated-carbon adsorption would result in lower
concentrations of SON and SCOD than when the sample was treated only with
activated carbon. As shown in Table 49, none of the various pretreatments
resulted in lower SON and SCOD concentrations following activated-carbon ad-
sorption.
The ultraviolet spectra of a number of the samples shown in Table 49
were determined between wavelengths 200 and 320 nm on a Beckman double beam
spectrophotometer and absorption was found to decrease with increasing SCOD
removal. For Samples 7 and 10, in which SON removal exceeded SCOD removal,
the reduction in ultraviolet absorption more closely paralleled SCOD removal
than SON removal.
Summary
A sequence of advanced wastewater-treatment processes (lime coagula-
tion, ammonia stripping, recarbonation, filtration, breakpoint
chlorination, and activated-carbon adsorption) can remove a large
amount (perhaps 70-90 percent) of the SON and SCOD from secondary
89
-------
TABLE 49. REMOVAL OF SON AND SCOP BY COAGULATION, ION EXCHANGE, AND ACTIVATED-CARBON ADSORPTION
Sample
1 (Control)
2
3
4
5
6
7
8
9
10
First
Treatment
-
Fe
Fe
Fe
Fe
AC
CAT
CAT
CAT
CAT
PH
-
10.0
10.0
10.0
10.0
8.0
8.0
8.0
8.0
2.0
Second
Treatment
-
-
AC
AC
Fe
-
-
Al
Al
-
pH
-
-
8.0
8.0
6.0
-
-
6.0
6.0
-
Third
Treatment
-
-
-
Fe
-
-
-
-
AC
-
PH
-
-
-
6.0
-
-
-
-
7.9
-
SON,
mg/1
1.59
1.08
0.38
0.35
0.89
0.40
1.43
1.04
0.35
0.97
Percent
SON
Removed
-
32
76
78
44
75
10
35
78
39
SCOD,
mg/1
28
20
3
4
16
3
28
19
2
21
Percent
SCOD
Removed
-
29
89
84
41
89
0
31
91
24
Differences in SON removal of 6% and 8% and differences in SCOD removal of 4% and 6% are 95%
and 99% significant, respectively.
AC = 25 g/1 activated carbon Al = 600 mg/1 aluminum sulfate
Fe = 600 mg/1 ferric chloride CAT = 15 g/1 cation-exchange resin
VO
O
-------
effluent, the bulk of the removal occurring during coagulation and
activated-carbon adsorption.
2. Removal of SON and SCOD in filtered secondary effluent appears to be
more efficient when ozonation follows rather than precedes coagula-
tion. Pretreatment of samples by ion-exchange resins did not result
in a significant increase in SON or SCOD removal by ozone.
3. The SON and SCOD remaining after passing the effluent through a
column of activated carbon are not readily removed by ozone.
4. The fraction of SON removed by the cation-exchange resin is also re-
moved by ferric chloride coagulation at pH 10, but not at pH 6. The
fractions of SON and SCOD removed by the anion-exchange resin are
not removed by ferric chloride coagulation at either pH 6 or 10.
5, Pretreatment of PASE with several different processes prior to
activated-carbon adsorption resulted in no significant reduction of
SON and SCOD compared to the reduction achieved by activated carbon
alone.
DISCUSSION OF RESULTS
The physical removal of SON and SCOD by treatment processes and combina-
tions of treatment processes, and the oxidative removal of SON and SCOD is
discussed in the following. A scheme was developed to fractionate the SON and
SCOD based on the observed removal of different portions by the various pro-
cesses and combinations of processes. Appendix A contains the statistical
basis for discussion of the significance of the results of this portion of the
overall study.
Physical Removal Characteristics of SON
Despite the complexity of the systems under consideration and the hetero-
geneous nature of the organic material present in secondary effluent general
•conclusions may be drawn with qualification. Realizing the limitations, it is
believed that a simplified characterization scheme of the removal of organic
contaminants will facilitate this discussion. Such a scheme is developed and
employed in the following paragraphs, and is summarized in Table 50. The
scheme is useful in explaining the results of the physical removal processes,
but not the oxidative processes, the effects of which are related more to
molecular structure than to gross molecular characteristics.
Molecular Characteristics Affecting Physical Removal of SON—
The molecules comprising the organic matter in secondary effluent possess
a wide range of characteristics, i.e., molecular size, charge, polarity, and
other characteristics, such as the presence of a particular functional group
or complexing ability, which may affect their removal by the various proces-
ses. It is useful to classify organic compounds in terms of these character-
istics.
91
-------
TABLE 50. SUMMARY OF MOLECULAR CHARACTERISTICS AFFECTING REMOVAL OF ORGANIC CONTAMINANTS
Process
Chemical Coagulation
(Alum or ferric
chloride)
Chemical Coagulation
(Lime or ferric
chloride)
An ion Exchange
Cation Exchange
Activated- Carbon
Adsorption (Batch)
pH
6
10-12
All
All
All
Molecular Characteristics
Size
Small
o
0
1
0
o
o
Large
+
+
-
-
-
Polarity
Polar
-
—
-
-
-
Non-
polar
+
+
+
+
+
Charge
Positive
-
+
-
+
-
Negative
+
~
+
-
-
Neutral
o
o
o
o
o
Other*
+
+
+
+
+
^v
o Characteristic does not affect removal. Characteristics which permit the formation of a
+ Characteristic favors removal. chemical bond or result in a chemical reaction, ao-
Characteristic hinders removal. counting for specific adsorption and precipitation.
-------
Regarding size, contaminant molecules may range from molecular weights
of less than 100 to small colloids, which in this study are limited to those
smaller than 0.45 microns. Molecules are arbitrarily divided into two sizes:
small and large. Large molecules include (1) molecules of colloidal size
which can be removed by coagulation and flocculation; (2) molecules which can
be removed by entrapment during coagulation; and (3) molecules excluded from
the pores of activated carbon and ion-exchange resins. Small molecules are
those with access to the pores of the ion-exchange resins and activated carbon.
Molecules of increasing polarity tend to be less adsorbable, to be more
soluble, and to bind water more tightly. Decreasing polarity favors adsorp-
tion to a surface, such as a resin matrix or activated carbon. For simplicity,
molecules are classified as either polar or non-polar. Non-polar molecules
are those which are hydrophobic, i.e., they would prefer not to be in the
water and are thus readily adsorbed on most surfaces. Polar molecules are
less readily removed, being hydrophilic.
Molecules may be electrostatically charged, positively or negatively, or
they may be neutral. Charged molecules are attracted to surfaces of the op-
posite charge and repelled by surfaces of like charge.
Molecules may also possess other characteristics which allow them to be
removed under otherwise unfavorable conditions. For example, although some
carboxylic acid molecules are both negatively charged and polar, they may ad-
sorb to a negatively charged calcium oxide surface because they are capable
of forming strong complexes with calcium, or their calcium salts may be in-
soluble. Similarly, polar molecules may be chemisorbed on activated carbon.
Any one molecule may possess a number of different characteristics, the
combination of which will determine its behavior. Distinctive fractions of
the organic matter are therefore hard to define and seldom mutually exclusive,
leading to qualitative, rather than quantitative, descriptions of results.
Reading across Table 50 for a particular process allows one to determine
the dominant characteristics of the fractions of organics removed and not re-
moved by each process. For example, the fraction not removed by activated
carbon will contain large, rather polar, and/or charged molecules. A com-
pound possessing only one "hindering" characteristic may not be removed even
though it possesses other characteristics favoring removal. Thus, a hydro-
phobic, positively charged molecule may not be removed by cation exchange if
it is too large to penetrate the pores of the resin.
Comparison of SON and SCOD Removals—
Table 51 contains a summary of the removals of SON and SCOD achieved by
the various processes. The removals reported were achieved using large dos-
ages to insure maximum removal under batch conditions and to determine the
fractions of SON and SCOD removable by each process. Slightly higher or lower
removals may be achieved with flow-through system or lower doses, respectively.
A more detailed description of each process and the effect of various param-
eters, including dose and pH, on each process can be found in the appropriate
section dealing with that process.
93
-------
TABLE 51. SUMMARY OF REMOVALS ACHIEVED BY INDIVIDUAL PROCESSES
Process
Lime Coagulation
Alum Coagulation
Ferric Chloride Coag.
Anion Exchange
Cation Exchange
Cation Exchange
Act. -Carbon Adsorp.
Chlorination
Ozonation
Number
of
Samples
9
4
14
6
6
6
8
5
7
*
PH
> 11
6
6
N
N
2
N
N
N
SON Removed, %
Avg.
33
29
40
12
11
42
71
42
14
Max.
43
35
48
20
12
50
92
48
28
Min.
26
23
30
9
9
32
52
39
6
Std.
Dev.
6
5
5
4
1
7
12
4
7
SCOD Removed, %
Avg.
24
28
32
30
4
22
81
18
46
Max.
36
46
40
34
12
26
93
28
53
Min.
13
17
19
23
0
18
68
10
41
Std.
Dev.
8
10
6
4
5
4
8
7
4
*
N = neutral.
As shown in Table 51 there are differences between the SON and SCOD re-
movals achieved by several of the various processes. These differences are
explainable if consideration is given to the differences one might expect be-
tween molecules which contain nitrogen and those which do not. Some of these
differences are: (1) only nitrogen-containing molecules are capable of re-
taining a stable positive charge (with the exception of organo-metallic com-
plexes) since only the nitrogen atom is capable of strongly binding an
additional hydrogen atom yielding a positively charged molecule; (2) larger
molecules have been reported to be nitrogen enriched [23]; and (3) nitrogen-
containing molecules, in general, tend to be more polar than non-nitrogenous
molecules, since nitrogen-carbon bonds are more polar than carbon-carbon
bonds.
Chemical coagulation with both lime and ferric chloride was found to re-
move a significantly (95% and 99% CI, respectively) larger percentage of SON
than of SCOD. The opposite would be expected from charge and polarity con-
siderations. At pH 12 the calcium oxide surface is negatively charged as
are most organic compounds at that pH, so charge does not favor adsorption
If nitrogen-containing molecules are in fact more polar than molecules which
do not contain nitrogen, then polarity would favor the removal of SCOD over
SON. Therefore, either nitrogen enrichment of the larger molecules removed
by coagulation and/or specific adsorption of nitrogenous molecules must ac-
count for the observed difference.
Cation exchange removed significantly (95% CI) more SON than SCOD while
anion exchange removed significantly (99% CI) less SON than SCOD. This Indi-
cates the tendency of nitrogenous molecules to be more positively charged
The organic matter removed by the cation-exchange resin appears to ^nitro-
gen enrzched, while that removed by the anion-exchange resL is nitrogen de-
ficient Furthermore, a lowering of pH, which should increase the positive
charge of the organics, was found to increase the removal of nltrog^rYched
molecules by cation exchange. nitrogen-enriched
94
-------
The preference of activated carbon for SCOD adsorption relative to SON
may be due to several factors: (1) polarity, since activated carbon removes
less polar compounds more readily and nitrogenous compounds are expected to
be more polar; (2) molecular size, since the larger molecules (reportedly ni-
trogen enriched [23]) may be excluded from the pores of the activated carbon;
and/or (3) specific adsorption, if the special mechanisms by which activated
carbon removes organic molecules favor the removal of non-nitrogenous molecules.
Physical Removal of SON by Combinations of Processes
Factors Affecting Removal—
The molecular characterization scheme in Table 50 is helpful in inter-
preting the results of the experiments dealing with the physical removal of
SON by combinations of processes described in Section 7. The results of these
experiments can, in turn, be used to quantify the various fractions of organ-
ics described by the scheme.
In the removal of SON by combinations of chemical coagulation processes,
different organics fractions were removed by different coagulants or by a
particular coagulant at different pH values. Some of these differences sug-
gest that electrostatic attraction is of importance in removal and thus a
discussion of surface charge is warranted.
Lengweiler et al. [88] determined that the point of zero charge (PZC) for
iron oxides in dilute solutions is at pH 6.7. Aluminum sulfate floes have a
PZC of about 8.0 [54,90]. Various anions in secondary effluent (e.g. sul-
phate) tend to make the surface of floes more negative, while various cations
(calcium, magnesium) tend to make it more positive [79,80]. Although the
ionic makeup of PASE was determined, present knowledge does not permit calcu-
lation of a PZC for this system. It is likely that at pH 6.0 ferric chloride
and alum floes in PASE bear a positive charge (although they may conceivably
be neutral or very slightly negative) and at pH 10.0 they bear a negative
charge. Regardless of the true value of the PZC, both floes will be more neg-
atively charged at pH 10.0 than at pH 6.0. If charged molecules are present
in secondary effluent, those which are negatively charged will adsorb better
at low pH, and those which are positively charged will adsorb better at high
pH.
There are, of course, other mechanisms, in addition to electrostatic at-
traction, which are important in the removal of organic contaminants by coa-
gulation, such as co-precipitation, flocculation, and adsorption. In some
cases where a second or third coagulation is carried out, these other mecha-
nisms can sometimes be shown to be inoperative. When alum or ferric chloride
was used a second time at approximately the same pH, no significant additional
removal of SON was achieved (Table 24), nor did largely increased doses of
alum and ferric chloride remove significantly more organic material (Tables
20 and 21). Thus, when a second coagulation was performed at a different pH
than the first, it can be assumed that any additional removal was not simply
due to increased removal by a new surface, by co-precipitation, or by an im-
provement in overall efficiency.
95
-------
Removal by Combinations of Coagulants—
The effect of ferric chloride coagulation at two different pH values (6.0
and 10.0) was investigated. When the first coagulation was at pH 6.0, a sec-
ond coagulation at pH 10.0 resulted in an additional 12 percent removal of SON
(Samples 2 and 4, Table 24), reflecting the removal of a significant (99% CI)
fraction of SON not removed by the first coagulation. Similarly, when the
first coagulation was at pH 10.0 a second coagulation at pH 6.0 resulted in
an additional and significant (99% CI) 27 percent removal of SON. The total
removal of SON achieved by sequential coagulation with ferric chloride at the
two different pH values was the same, regardless of the order of coagulation.
An average of about 20 percent of the SON was removable by ferric chloride at
both pH 10.0 and 6.0. The removal of this fraction appears to be relatively
independent of pH (and therefore surface charge) and is most likely due to a
combination of co-precipitation, flocculation, specific adsorption, and en-
trapment .
That coagulation with ferric chloride at pH 6.0 and pH 10.0 appears to
remove different fractions of organic matter can be explained in terms of
electrostatic adsorption. If certain fractions of the organic matter are be-
having as anions and cations, their removal by ferric chloride will be depen-
dent upon pH. The fraction of organics removed by ferric chloride at pH 10.0
but not at pH 6.0 (comprising about 12 percent of the SON) will be termed
"positive"; and the fraction removed at pH 6.0 but not at pH 10.0 (comprising
about 27 percent of the SON) will be termed "negative." The terms "positive"
and "negative" used here are meant to describe the relative charge of the
fractions with respect to each other rather than their absolute charge. Fur-
ther evidence of the existence and nature of these fractions will be presented
later in this discussion.
The addition of lime to secondary effluent (containing carbonate anions)
results in the formation of a calcium carbonate/oxide solid,- the surface of
which is negatively charged at the values of pH used in wastewater treatment
[81]. Small amounts of phosphate and other anions present in secondary efflu-
ent tend to make the calcium carbonate surface even more negatively charged
[79,80,81]. Thus, the surface formed from lime is expected to be negatively
charged, as is the surface of the ferric chloride floe at pH 10.0.
Neither lime coagulation following ferric chloride coagulation at pHlO.O
nor ferric chloride coagulation at pH 10.0 following lime coagulation was able
to achieve any significant additional removal of SON (Samples 3 and 4, Table
21). Thus, it appears that both coagulants removed the same fractions of
organics when the floes were of the same charge. When lime coagulation fol-
lowed ferric, chloride coagulation at pH 6.0 (floes likely to be of opposite
charge) an average additional removal of SON of 10 percent was achieved
(Samples 1 and 4, Table 24), once again indicating a "positive" fraction of
organic matter of roughly 10 percent.
When lime coagulation was carried out a second time following an original
lime coagulation, an additional and significant 16 percent removal of SON was
observed (Sample 1, Table 24). This additional removal may be explained by an
expected difference in surface charge for the calcium carbonate surface In
the first coagulation, the surface is more negatively charged because of
96
-------
carbonate, phosphate, and other adsorbing anions, as explained above. These
anions are not present for the second coagulation and so the surface is more
neutral and for this reason can remove additional organics.
Alum coagulation did not achieve any significant additional removal of
SON following ferric chloride coagulation. Thus, there appears to be a frac-
tion of organic matter removable by ferric chloride and not removable by alum
coagulation, since ferric chloride was approximately 14 percent more efficient
(99% CI) in removing SON than alum (Table 23). This fraction is believed to
be specifically adsorbed on the iron hydroxide surface. Interestingly, there
was no significant difference detected in the removal of SCOD by alum and fer-
ric chloride but a small difference may have actually occurred. It is pos-
sible that the fraction of organic matter removed by ferric chloride and not
by alum is nitrogen enriched.
The addition of bentonite prior to ferric chloride coagulation at pH 6.0
increased SON removal by 10 percent (99% CI). While this effect could con-
ceivably be due to an increased efficiency of flocculation, it was not ob-
served with kaolinite clay or any of the polyelectrolytes used in conjunction
with either ferric chloride or lime. A more likely explanation for the in-
crease in removal lies in the fact that bentonite is a natural cation ex-
changer and may have been able to adsorb cationic molecules (the "positive"
fraction), which were then removed together with the bentonite upon subsequent
ferric chloride coagulation.
Removal by Combinations of Ion Exchange—
The ion-exchange resins evaluated were identical in matrix and cross-
linking, differing only in their respective functional groups. Thus, mole-
cules adsorbing on one resin and not the other should do so primarily because
they are of the proper charge, even though forces other than electrostatic
attraction may be important in the adsorption. For example, Semmens and
Gregory [183] found that only the ionized forms of carboxylate ions were ad-
sorbed by anion-exchange resins, even though there were interactions between
the carbon chains of the molecules and the resin matrices.
At neutral pH, cation exchange was found to remove 11+1 percent of the
SON and 4 ± 5 percent of the SCOD from samples of PASE. Anion exchange was
found to remove 12+4 and 30 ± 4 percent of the SON and SCOD, respectively.
The fractions of organic matter removed by the two resins at neutral pH were
found to be mutually exclusive, and it is very likely that they are also ap-
propriately charged. If so, they represent positively and negatively charged
fractions of organic matter.
Removal by Combination of Ion Exchange and Coagulation—
The positively and negatively charged fractions of organic matter defined
by ion exchange might be expected to parallel the "positive" and "negative"
fractions defined earlier for ferric chloride coagulation. The percentage of
SON found to adsorb on the cation-exchange resin agreed well with the percent-
age of SON termed "positive" as a result of the ferric chloride coagulation
experiments. However, the percentage of SON adsorbed on the anion-exchange
resin was less than half the percentage predicted to be "negative" by the
ferric chloride coagulation experiment.
97
-------
An experiment was conducted to determine if these fractions were in fact
the same. The SON removed by cation exchange was not removed by ferric chlo-
ride coagulation at pH 6.0, but it was entirely removed at pH 10.0 (Tables 47
and 48). Thus, the "positive" fraction defined by ion exchange and by ferric
chloride coagulation appear to be identical. However, the SON and SCOD ad-
sorbed by the anion-exchange resin were not removed by ferric chloride coagu-
lation at pH of either 6.0 or 10.0. Thus, the "negative" fraction defined by
ferric chloride coagulation is separate and distinct from the fraction removed
by anion exchange.
It is not clear why the "negative" fraction defined by ferric chloride is
not removed by anion exchange. Although the molecules comprising both frac-
tions might be expected to be negatively charged (or neutral), electrostatic
charge is not the only important factor affecting adsorption on the two sur-
faces involved. The ferric hydroxide surface will not adsorb certain nega-
tively charged molecules which are small and polar. The anion-exchange resin
used for this study contains quaternary ammonium functional groups, which are
probably incapable of true ion exchange for all but the smallest of anionic
molecules and are incapable of hydrogen bonding. The lattice of the resin is
an aromatic and alkyl structure which may adsorb aromatic contaminants under
the proper conditions. Thus, it is conceivable that the resin removes aro-
matic molecules. A study of the individual molecules removable by each sur-
face should help in resolving this question.
Removal by Combinations Including Activated Carbon—
It has been postulated that a significant fraction of the organic matter
which is not adsorbable on activated carbon is composed of small polar com-
pounds [28], Ion exchange should favor the removal of small charged molecules,
since they will have a higher charge density than larger molecules and easier
access to the pores of the resin. The charge on a molecule makes it more
polar, perhaps preventing adsorption on activated carbon, but may increase the
ability of the molecule to adsorb on ion-exchange resins. Thus, one might
expect to find differences in the fractions of organic matter removed by ion
exchange and activated-carbon adsorption.
Table 31 presents the results of an experiment conducted to compare the
fractions removed by ion exchange and activated carbon. These results are
summarized in Table 32 in a more readily interpretable form. The organic mat-
ter removed by the cation-exchange resin is not adsorbed on activated carbon.
Virtually all of the SCOD adsorbed on the anion-exchange resin but only a
portion of the SON so adsorbed is removed by activated carbon.
The adsorption of much of the organic matter removable by anion exchange
on the activated carbon may reflect the hydrophobic nature of these molecules,
and suggests that adsorption on the anion-exchange resin may be largely due
to interactions between the molecules and the resin matrix. The anion-
exchange resin used may not be able to remove all anions, thus, organic mate-
rial not removed by activated carbon may be anionic even though not removed
by the anion-exchange resin.
Summary of Physical Removal Process Results
Due to the large number of overlapping mechanisms for removal, the
98
-------
multiple characteristics of the molecules, and the lack of sharp distinction
between the divisions of the molecular characteristics, it is difficult to
accurately quantify the organics into a complete set of mutually exclusive
fractions. However, a few assumptions can be made which will allow the SON
to be fractionated into a useful set of categories which can be roughly quan-
tified, as presented in Fig. 7.
The fractionation scheme shown in Fig. 7 is based upon a large number of
analyses conducted on a large number of samples. It is thus a composite or
summation of the results of this study, and as such is not meant to be a pre-
cise fractionation of a single, well-characterized sample. Rather, it is
meant to depict the relative sizes of the various fractions and their inter-
relationships in a simplified and easily understandable form. The size of
each fraction has been approximated to within ± 5 percent.
The existence of a small, positively charged fraction of the SON is evi-
denced by (1) cation exchange; (2) the difference in SON removal with ferric
chloride at pH 6.0 and 10.0; (3) the additional 10 percent SON removal
achieved by adding bentonite prior to ferric chloride coagulation and (4) the
additional 10 percent removal of SON achieved by lime coagulation following
ferric chloride coagulation at pH 6.0. This fraction seems to comprise ap-
proximately 10 percent of the total SON in all cases. As shown in Table 51,
cation exchange at neutral pH removed an average of 11 ± 1 percent of the SON
from 6 samples, along with 5 ± 4 percent of the SCOD of the same samples.
This fraction is fraction "P" in the fractionation scheme of Fig. 7.
The "negative" fraction, comprising 27 ± 4 percent of the total SON (ap-
proximated as 25% in the scheme) was evidenced by the ferric chloride coagula-
tion data as discussed earlier in this section. The "neutral" fraction shown
in Fig. 7 is comprised of the organic matter not termed "positive" or "nega-
tive." While molecules in the "neutral" fraction may in fact be positively
or negatively charged, their charge character is not evident in their behavior
in the systems studied. It is quite likely that most of these molecules are
negatively charged, but that mechanisms and factors other than electrostatic
attraction are dominant in the behavior noted.
The fraction removed by activated carbon comprises 71 + 12 percent of the
SON and 81 ± 8 percent of the SCOD. Since the adsorption of SON by activated
carbon is not a function of pH (except for capacity), removal of organics on
activated carbon is not related to electrostatic charge, and in fact, charged
molecules might be expected to adsorb poorly. As discussed earlier, the pos-
itively charged fraction is not removed by activated carbon. The fraction not
removed by activated carbon and not positive should consist of molecules which
are small, very polar, and possibly negatively charged, assuming that under
batch conditions, most large molecules are adsorbed onto the outer surface of
the carbon. The fraction of SCOD removed by anion exchange is entirely re-
moved by activated carbon, indicating its hydrophobic nature.
Any given sample of secondary effluent may be fractionated by the scheme
shown in Fig. 7 simply by carrying out the treatment processes to define the
fractions of interest. This type of fractionation scheme is not limited, how-
ever, to secondary effluents or to the treatment processes shown. Any waste
99
-------
Characteristic
SON, %
SCOD, %
o
o
Size (L = large)
Charge (N = nega-
tive, P=positive)
Polarity
Process
Activated-Carbon
Adsorption
Anion Exchange
(neutral pH)
Cation Exchange
(neutral pH)
Cation Exchange
(PH 2)
Lime Coagulation
(pH > 11)
Alum Coagulation
(pH 6)
Ferric Chloride
Coagulation (pH 6)
Ferric Chloride
Coagulation (pH 10)
Small
N
Neutral
Non-Polar
(Percent)
20
40
60
25
25
25
40
30/80
20
15
15
80
100
10
10
10
10
Small
N
Neutral
Non-Polar
20
(Percent)
is—
60
25
25
25
15
10
10
50
30
15/80
100
Figure 7. Operational fractionation of SON and SCOD.
-------
may be fractionated in a similar manner, provided that care is taken to ade-
quately define the relationships between the various fractions. The important
point is that treatment processes do not all remove the same fractions of or-
ganic matter and that the differences between the fractions are predictable
and understandable from known characteristics of the processes and the mole-
cules being removed.
The fractions shown in Fig. 7 are not all mutually exclusive and a great
many assumptions were made in the definitions, descriptions, and in the as-
signed values. Nevertheless, they are useful in describing the removal of SON
from secondary effluent by a sequence of treatment processes and in suggesting
directions which future investigations might take. The methods and the prin-
ciples discussed for secondary effluent may be extended to predict the removal
of various fractions for other wastes, perhaps employing other treatment pro-
cesses.
Oxidative Removal Characteristics of SON
Table 52 contains a summary of the results of the experiments dealing
with chemical oxidation of SON.
Potassium permanganate and hydrogen peroxide removed 28 and 19 percent of
the SON, respectively, but did not remove any SCOD. Spicher and Skrinde [173]
found that permanganate reacts readily with amino groups (but not with amino
acids) resulting in deamination to form ammonia and organic acids. This is
likely to be the major reaction of permanganate with SON.
Chlorine was found to easily (at low doses) oxidize 27 percent of the
SON, but SON thus oxidized could be reduced with sodium sulfite, indicating
that the major initial reaction of chlorine with SON is substitution. This
again suggests the presence of amino groups, comprising roughly 25-30 percent
of the SON.
TABLE 52. SUMMARY OF RESULTS OF CHEMICAL OXIDATION EXPERIMENTS
Oxidant
Chlorine
Ozone
Potassium Permanganate
Hydrogen Peroxide
Dose,
mg/1
452
100
100
100
pH
7
7
10
11
Percent
SON *
Removed
42 ± 4
14 ± 7
28 ± 3
19 ± 3
Percent
SCOD ^
Removed
18 ±7
46 • ± 4
0 ± 2
2 ± 2
&
Standard deviations for chlorine and ozone are based on analysis of
a number of samples (Table 51) ; others are analytical standard
deviations .
101
-------
Ozonation resulted in very little SON removal, although a large percent-
age of the SCOD was removed. Thus SON is quite resistant to ozonation as
compared to SCOD; however, the side chains of the nitrogen atoms appear to be
altered by ozone, as evidenced by an observed 30 percent increase in SON re-
moval by chlorination after a sample had been ozonated.
In terms of physical removal characteristics, ozone does not appear to
attack the "positive" fraction, the fraction removed by anion exchange, nor
the SON removed by very small doses of activated carbon. The fraction of the
organic matter most strongly oxidized by ozone is the additional matter re-
moved by cation exchange as the pH is lowered from neutral to 2.0. Ozonation
is expected to increase the percentage of SON in the "negative" and "polar"
fractions and to break up large molecules into smaller ones.
102
-------
SECTION 8
SON FORMATION AND REMOVAL BY BIOLOGICAL PROCESSES
INTRODUCTION
The research described in this section was conducted to evaluate the
source and characteristics of SON in activated-sludge effluents. It would be
helpful to know how much SON is produced during treatment, and what factors
affect the quantity produced, as well as what fraction of influent SON re-
mains unaltered through treatment. If a large fraction of the effluent SON is
produced during biological treatment, the advisability of such treatment may
be questioned and ways to minimize the production may be sought. The same
could be said for effluent SCOD (soluble chemical oxygen demand). The effect
of process control parameters such as mixed liquor suspended solids concentra-
tion (MLSS), organic loading, and aeration time on SON production and removal
by activated sludge were evaluated in this study to help define operational
criteria for minimizing effluent SON.
SON in the effluent from activated-sludge treatment is a function of SON
in the influent to the system arid biological processes within the system,.such
as production, utilization, and biodegradation which result in SON charges.
Bacteria produce organic nitrogen compounds that are excreted, or released
during cell lysis to form SON. Some compounds are biodegradable, some refrac-
tory. Bacteria can utilize some organic nitrogen compounds for nitrogen,
carbon, and/or for energy. Factors affecting production, utilization, and
biodegradability of SON are discussed in the following.
NOMENCLATURE
Equations defining the concentration of various SON types and sources are
as follows:
SON. = SON, + SON (1)
i b r
SON = SON + SON + f(SON, ) (2)
e r p b
SON = SON + SON, (3)
P g d
SON = SON , + SON (4)
g gb gr
SON, = SON,, + SON, (5)
d db dr
103
-------
SON.^ represents the SON in the influent, and consists of a biodegradable
portion, 30%, and a refractory portion, SONr.
Activated-sludge effluent SON (SONe) consists of refractory material
originally in the influent and not removed during treatment (SONr), a fraction
of the biodegradable influent SON not removed during treatment (f(SONb)), and
SON produced during treatment (SONp). Theoretically, f can range from 0-1
depending on the efficiency of treatment.
SONp represents material produced as a result of bacterial growth (SONg)
and material produced by organism decay (SONd). SONg and SONd may contain
biodegradable SON (SONgb and SONdb) and/or refractory SON (SONgr and SONdr)•
A major objective of this research was to evaluate the contributions of
the SON types listed in Eqs. 1 to 5 to SON contained in activated-sludge ef-
fluents. The nature of and factors affecting each form of SON are discussed
in the following.
BIOLOGICALLY PRODUCED SON
The major classes of nitrogen-containing organic compounds produced by
bacteria are listed in Table 53. Materials present in activated-sludge efflu-
ents may be decomposition products, condensation products, or derivatives of
these compounds.
From 60 to 80 percent of bacterial cell dry weight is represented by pro-
tein and nucleic acids, which contain organic nitrogen [29,30,31,32]. This
percentage varies with bacterial species and phase of growth. Nucleic acids
(DNA and RNA) and proteins are usually made from amino acids, purines, and
pyrimidines that are either synthesized from inorganic nitrogen and simple
carbon compounds, or preferably found by some bacteria in the surrounding en-
vironment [30]. With the preponderance of organic nitrogen in the cell, it is
likely that two processes take place simultaneously during cellular metabo-
lism: (1) utilization of extracellular SON compounds as carbon, energy, and/
or nitrogen sources, and (2) excretion of intracellular organic nitrogen
metabolites to form biodegradable and refractory SON.
REFRACTORY ORGANICS
Effluents from activated-sludge treatment contain significant amounts of
refractory organics, including SON. A clear definition of refractory organics
is needed, as is a description of the characteristics which make organic mole-
cules refractory.
Definition
Recalcitrant, refractory, persistent, residual, and non-degradable are
terms commonly used to describe organics that are difficult, if not impossible,
for microorganisms to metabolize. Hence, such compounds can persist in the
environment for relatively long periods of time. Refractory materials may be
104
-------
TABLE 53. BIOLOGICALLY PRODUCED SON COMPOUNDS
Class
Structure
Percent of Bacterial
Cell Weight
[77,78,79,80]
Amino Acids
Proteins, Peptides
Enzyme
Amine
Pyrimidine
(heterocyclic N)
Purine
(heterocyclic N)
Nucleoside
Nucleotide
SNA
DNA
Hydroxamic Acids
(ionophores)
Amino Sugars
NH2
R-C-COOH
•Rf
large proteins
, Ar-NH
H
Purine or pyrimidine bonded
to a 5-carbon sugar
_3
Nucleoside bonded to a PO,
Polynucleotide; sugar monomer
is ribose
Polynucleotide; sugar monomer
is deoxyribose
R-C-N-R'
40-60
2-40
1-5
where R,R' = H or an aliphatic carbon structure, and
AR = an aromatic structure.
105
-------
degraded in the environment over periods measured in months, years, or centu-
ries, and by physical or chemical means, as well as biologically. For those
concerned with removal of organics from wastewaters, refractory generally re-
fers to materials that are not degraded to a significant extent by well accli-
mated and operated biological treatment systems. Such materials subsequently
degrade slowly, if at all, upon discharge to receiving waters.
Factors Affecting Recalcitrance
Alexander has reviewed and summarized reasons for the failure of micro-
organisms to degrade organic materials [55,56,57]. These are summarized, with
additional support information, in the following.
Inherent Molecular Characteristics—
Certain molecular characteristics make some molecules more resistant to
biodegradation than others, but information about specific mechanisms involved
is limited. Molecular properties that may affect biodegradability include
molecular size, length of chain, stereochemistry, and type, number, and posi-
tion of substituents in the molecule [45,55,56,57,58,59].
Unsaturated hydrocarbons are degraded more readily than corresponding
saturated compounds, most likely due to the more reactive nature of the double
bond. In general, branching of aliphatic compounds and substituent chains on
aromatics reduces biodegradability, the effect being dependent on the posi-
tion, extent, and type of branching. The presence of terminal quaternary
groups or non-alkyl substituents may markedly affect degradability, especially
if degradation proceeds by a- or B-oxidation [55,60]. Enzymes may be unable
to contact reactive moieties such as -NH2, -OH, -COOH, and others to initiate
degradation due to unfavorable molecular size, shape, or stereochemistry, thus
increasing the recalcitrance of some molecules [45,56,61].
There are data suggesting that substitution of a carbon in a chain with
nitrogen, oxygen, or sulfur (-£-*!-<£-, -(J-Q-lJ:-, -<*-S-£-) reduces degradability
[56]. Addition of halogens, nitro, or methyl groups imparts resistance to
some compounds. On the other hand, addition of -NH2, -OH, -COOH, and -CH3
to the benzene ring has been shown to render it more degradable [59]. Di-
substituted benzene compounds with substituents in the meta position are gen-
erally more refractory than ortho- and para-substituted molecules. Also,
tri-substltuted benzene derivatives degrade at rates slower than those for di-
substituted derivatives, which are in turn more refractory than the mono de-
rivatives [45,58,59]. There are notable exceptions to these last two observa-
tions.
From the preceding discussion, one could formulate the general conclusion,
with exceptions, that molecular recalcitrance increases with increasing molec-
ular complexity, that is, with increasing number and/or types of bonds.
Environmental Factors—
Common environmental factors such as pH and temperature can affect bio-
degradability if not in ranges optimum for microbial growth [55,56,57], Lack
of an appropriate terminal electron acceptor, or absence of any of a number
of essential nutrients such as nitrogen, phosphorus, iron, trace metals,
106
-------
organic growth factors, etc., will render molecules refractory [55]. Toxic
materials must be absent [55]. During activated-sludge treatment of domestic
wastewaters, these factors are not likely to have a major effect on recalci-
trance of effluent organics.
In biological waste-treatment systems, operation must insure that orga-
nisms capable of degrading a material of interest are not washed from the sys-
tem before degradation can be accomplished. Recent studies have indicated
that selection of 6C (solids retention time) , which is related to microorga-
nism growth rate, directly affects the concentration of refractory material
remaining after activated-sludge treatment [63,64,65,66].
Enzyme-Related Factors—
Since enzymes initiate and catalyze degradation sequences, enzymes cap-
able of attacking the organic molecule and subsequent degradation products
must be present, or inducible, and free of inhibition and repression. Enzyme
inhibitors may slow the rate of degradation, thus making degradable organics
appear relatively refractory. For some organics, no enzyme may exist for ini-
tiating the degradation sequence, thereby making the molecule recalcitrant
[57], Certain organics act as catabolic repressers to enzymes required for
degradation of other organics [56,67,68,69], and when present, degradation
will not take place. Enzymes from several organisms may be required for de-
gradation of a particular organic, and all must be present to insure degrada-
tion. If the enzyme system responsible for breakdown is intracellular, the
compound must be able to penetrate the cell wall or it will remain undegraded.
Apparently, any number of factors can affect enzyme capability and activity,
and in this manner contribute to the recalcitrance of organics.
Lack of Sufficient Carbon or Energy for Growth—
The absence of sufficient exogenous carbon for growth and energy may con-
tribute to the recalcitrance of otherwise degradable substances. Addition of
a utilizable carbon source to a medium containing a hitherto refractory or-
ganic caused degradation of the refractory organic concomitant with utiliza-
tion of the added carbon source [56,93], This phenomenon, called co-metabolism,
may offer a means of biologically removing otherwise refractory materials.
Low Substrate Concentration—
Degradable materials may not be broken down if they are present in con-
centrations too low to provide sufficient energy or carbon for cell growth
[55,56,57,94], Thus, if an activated-sludge effluent contained a wide range
of degradable substances, but each in concentrations too low to support growth,
the substances may be classified as "refractory." Concentration of soluble
effluent organics resulted in additional removal [62], which supports the oc-
currence of the phenomena.
Chemical Reactions that Increase Recalcitrance—
Organics can undergo chemical reactions in the environment which make
them less susceptible to degradation. Complexing with other refractory com-
pounds such as lignin, tannins, and polyaromatics reduces degradability [56,
57], as does formation of inorganic complexes with clays and various metals
[55,56], Humic substances, highly refractory organic polymers containing
some nitrogen, are formed from simple, degradable precursor molecules excreted
107
-------
by microorganisms [95,96,97]. Condensation reactions of amino acids and sim-
ple carbohydrates form refractory nitrogenous polymers (termed melanoidins),
including heterocyclic nitrogen compounds, via a mechanism called the Browning
reaction [40,41]. Reactions such as these may occur during activated-sludge
treatment since a wide variety of inorganic and organic materials are present.
Summary—
Organic molecules are refractory for a variety of reasons. Environmental
factors affecting recalcitrance can be controlled during activated-sludge
treatment to minimize adverse effects. 9C can be manipulated to minimize ef-
fluent refractory organics. Factors which cannot be controlled are most
likely to be responsible for recalcitrance of effluent organics, including
SON. Inherently refractory compounds may be present in the influent or formed
chemically or biologically during treatment. Inhibition, repression, and in-
activation of enzyme systems increase refractory concentrations. Degradable
substances may not be metabolized if sufficient carbon or energy for growth is
lacking. If concentrations of degradable materials are too low to support
growth, they may not be removed. All or some combination of these uncon-
trolled factors no doubt contribute to the recalcitrance of effluent organics.
Recalcitrance of SON Compounds
Some organic nitrogen compounds are readily degraded by many microorga-
nisms in order to obtain energy, carbon, and/or nitrogen. Urea is readily de-
graded by bacteria via the enzyme urease. Free amino acids are metabolized
by oxidative, reductive, or hydrolytic deamination reactions. Proteins are
broken down to simpler peptides, and finally to amino acids through a series
of hydrolytic cleavages catalyzed by several proteinase enzymes. Thus, efflu-
ents from activated-sludge treatment contain low concentrations of these mate-
rials.
Nucleic acids (DMA and RNA) and their precursors containing nitrogen
(nucleotides, nucleosides, purines, and pyrimidines) are degraded at varying
rates, to different extents, and accumulate at varying levels in fluids sur-
rounding bacterial cultures and in biological reactors operating under a va-
riety of conditions [61,98,99,100,101,148,149]. Nuclease enzymes aid in the
breakdown of DNA and RNA to nucleotides and nucleosides, with DNA the more
resistant of the two [61,148]. Nucleotides and nucleosides can be degraded
to purines, pyrimidines, and related heterocyclic nitrogen compounds. These
compounds may be further metabolized yielding a variety of products (Table
54), such as hypoxanthine, xanthine, uric acid, allantoin, urea, and finally,
ammonia [98]. Nucleic acids and their degradation products exhibit varying
degrees of biodegradability, making it likely that such materials will be
present in activated-sludge effluents.
There is evidence indicating that carbon-nitrogen-carbon bonds (-6-N-6 ),
both in aliphatic and heterocyclic structures, are quite resistant to degrada-
tion [56,201]. This is to be distinguished from the peptide bond
(-Ll-cJ-)
108
-------
TABLE 54. TYPICAL NUCLEIC ACID DEGRADATION PRODUCTS
o
VO
AMP
(Ribonucleotide)
Xanthine
Adenosine
(Ribonucleoside)
Adenine
(Purine base)
=0 -> —> H2N-C-NH2
H
Allantoin
Urea
Hypoxanthine
NH,
-------
which contains an oxygen molecule that adds to its reactivity. Carbon-
nitrogen-carbon bonds are contained in heterocyclic nitrogen compounds such
as nucleic acids and their breakdown products (Table 54), and in condensation
products, such as from the Browning reaction.
Bacterial cell walls are quite stable, and must first be solubilized by
means of lysozymes prior to enzymatic degradation, at rates that may be quite
slow, to less recalcitrant, nitrogen-containing precursor molecules such as
amino sugars, nucleotides, proteins, and amino acids [61,206]. The nucleo-
tides, proteins, and amino acids are degraded as described earlier. Amino
sugars are not likely to accumulate in activated-sludge effluents [5].
Aliphatic and aromatic nitriles (-C=N) are readily degradable, the nitro-
gen first being reduced to amino nitrogen and then removed by hydrolytic
cleavage [58]. Mono-substituted nitrosobenzenes are degraded in a similar
manner, but rates are considerably slower than for the nitriles or correspond-
ing amino substituents [45,58], Such compounds are not likely to be found in
domestic wastewaters and should not be present in significant quantities in
effluents from activated sludge treatment of such waters.
Apparently due to inherent recalcitrant properties, di-substituted
nitrosobenzenes, toluidines, and phenylenediamines are not degraded by micro-
organisms [45,58,59]. These compounds are generally of industrial origin and
are not likely to be major constituents in effluents from biological treat-
ment of domestic wastewater.
To summarize, no one class of organic nitrogen compounds present in do-
mestic wastewater, or produced by microorganisms (i.e., amino acids, proteins,
purines, nucleic acids, etc.) exhibits what could be termed "outstanding re-
calcitrance." However, it appears that nucleic acids and their myriad degra-
dation products may comprise a significant portion of activated-sludge efflu-
ent SON due to their ubiquitous presence in bacterial cells (up to 45 percent
by weight), and their varying rates, extents, and pathways of degradation.
Small amounts, of amino acids and proteins can be expected. If individual con-
centrations of these and other degradable SON compounds are sufficiently low,
or if there is insufficient carbon or energy for growth, these normally bio-
degradable molecules could be present in secondary effluents and classified
as refractory. In addition, refractory SON may be formed from degradable
molecules via chemical or biological reactions during treatment.
SON PRODUCTION AND EXCRETION BY BACTERIA
It is possible that SON is produced and excreted during activated-sludge
treatment of domestic wastewaters. Support for this comes from studies con-
ducted with pure cultures. While extrapolation of information obtained with
pure cultures to heterogeneous activated-sludge systems has limitations it
can be helpful in explaining observed behavior.
Mechanisms Involved
Excretion of metabolites and cellular contents occurs during cell lysis,
110
-------
or by membrane transport. Membrane transport may take place by one of three
mechanisms [150]:
1. Simple diffusion: sometimes termed leakage; requires no energy in-
put; results in equilibration of the compound across the membrane;
2. Facilitated diffusion: accomplished by an interaction with some
component of the membrane, such as a permease system; requires no
energy input; results in equilibration of the compound across the
membrane;
3. Active facilitated diffusion: carrier-mediated process; requires
energy input; may occur against or with a concentration gradient.
All three mechanisms may be involved to different degrees depending on fac-
tors discussed below.
Factors Affecting SON Production and Excretion
The following list of factors affecting SON excretion is not all-inclusive,
and individual factors are not mutually exclusive. The list contains factors
thought to be relevant for activated-sludge cultures. Under some conditions,
one factor may control excretion, but under most conditions, several factors
would no doubt be involved since activated sludge represents a complex,
heterogeneous populaton of organisms.
Concentration Gradient—
Organisms are known to excrete organics via diffusion and facilitated
diffusion to establish a concentration equilibrium across the cellular mem-
brane [150,151,152]. This type of excretion would be expected when organisms
are diluted, the dilution causing a decrease in the concentration of external
organics surrounding the cell, thus establishing a concentration gradient.
A listing of organic nitrogen compounds excreted in response to a concentra-
tion gradient could not be found.
Starvation—
Bacteria excrete organics during starvation, a condition that exists
when substrate is essentially gone and cells must obtain energy for mainte-
nance by endogenous respiration or metabolism of intracellular components
[99,153,157]. Cell lysis is not considered to be a major factor during star-
vation, and excretion is via the diffusion transport mechanisms described
earlier [109].
The major excretion products observed during starvation are the degrada-
tion products of RNA [99,100,109,153,157,177,178], RNA is an attractive en-
dogenous metabolite since it is present in large quantities in the cell and
contains a ribose (sugar) moiety that can be used as an energy source. Cells
metabolize the ribose and excrete the associated purine or pyrimidine [99,
100,177,178]. Soluble purines and pyrimidines, and, to a lesser extent,
nucleotides and nucleosides, thus accumulate in the liquid surrounding starv-
ing cells. Proteins and free amino acids are also excreted but in lower con-
centration [99,100,112,157]. Cellular DNA is not degraded, and so its
degradation products are not observed [99,153].
Ill
-------
Presence of an Energy Source—
The presence of an exogenous energy source can stimulate excretion of or-
ganics [113,114,115]. Release is energy dependent, and as such, takes place
by active facilitated transport. Appearance of the excreted metabolite in
the culture may be quite rapid, taking place within a few minutes after sub-
strate addition [113]. Excreted organic nitrogen compounds observed include
RNA, nucleotides, proteins, and amino acids [113,115,148].
Substrate-Accelerated Death—
Addition of a carbon and energy source to bacteria starved for carbon
and energy accelerates death of the bacteria, a phenomenon called "substrate-
accelerated death" [99,116,117,118,189]. Death is defined as the inability
to grow in the presence of a utilizable substrate. Cell lysis does not occur.
Concentrations of soluble excreted metabolites (RNA degradation products)
were reported greater during substrate-accelerated death than during starva-
tion [189]. In another study [116], concentrations were similar but charac-
teristics were different. From reported data, it could not be determined
what portion of the excreted metabolites resulted from the increased number
of non-viable (dead) cells, and what portion from the remaining viable cells.
Regardless, "substrate-accelerated death" results in excretion of organic
nitrogen.
Availability of Required Nutrients—
If essential nutrients are absent, or in low concentrations, organics
may be excreted [9,118,190,191]. lonophores, organic nitrogen-containing
compounds responsible for chelating various metal ions needed by the cell for
transport into the cell, are excreted by bacteria [9,190]. Larger quantities
of ionophores are excreted in metal-poor environments, and they are appar-
ently produced and excreted as required by the organism. Other nutrient de-
ficiencies (nitrogen, phosphorus) may result in the production and excretion
of organic matter [118,191]. Excretion due to nutrient deficiency is not
likely to be a major factor during activated-sludge treatment of domestic
wastewaters which are expected to contain sufficient nitrogen, phosphorus,
and trace nutrients.
Environmental Stress—
Excretion of organic nitrogen compounds due to environmental stresses
such as extreme temperature changes and osmotic shock has been observed [192,
193,194,195]. Compounds studied were enzymes and proteins.
Effect of Bacterial Growth Phase and Growth Rate—
Excretion of metabolites to form SON during bacterial growth may result
from a combination of mechanisms discussed above, and/or by factors related
specifically to growth.
The different major phases of growth are logarithmic, declining, sta-
tionary, and death, each defined as follows:
Logarithmic phase: bacteria growing exponentially; growth rate
constant and limited by regeneration time of the organisms and
their ability to process substrate;
112
-------
Declining phase: growth rate decreasing; substrate concen-
tration now limiting growth;
Stationary phase: growth rate essentially zero; starvation
conditions prevailing;
Death phase: growth rate negative; during initial stages,
cells starving and losing viability, but not lysing; during
latter stages, lysis occurs.
More organic nitrogen is produced and excreted during log growth than
during stationary growth [32,101,113,148,196,197,198]. The internal mass of
DNA, SNA, protein, and free amino acids per cell increases during logarithmic
growth. As a result, excretion of these materials and their degradation pro-
ducts by energy- independent diffusion is expected since a higher concentration
of the material is present within the cell. In addition, organisms produce
and utilize more energy during logarithmic growth [113,114]. This excess en-
ergy can be dissipated by increased excretion (energy-dependent, active, fa-
cilitated diffusion) or organic nitrogen metabolites no longer needed by the
cell [113,115], Death of cells may occur concurrently with growth, but during
log growth, production and excretion of metabolites as a result of growth far
exceeds the quantity of material released by cell lysis [101,113,148].
During the stationary phase, starvation conditions prevail, and excretion
occurs as described previously (Starvation Factor). As starvation continues,
organisms begin to die, that is, lose viability. Organics released by non-
viable cells, including nucleotides and their degradation products, can be
used as carbon and energy sources by viable cells [199]. Cell lysis occurs
in the latter stages of the death phase, releasing organic nitrogen compounds
such as nucleic acids, nucleotides, nucleosides, and proteins [32,148,149,196,
197]. Soluble material released during cell lysis is fairly degradable and
does not accumulate in large quantities in the medium [101,203,204].
Growth rate is related to phase of growth as described above, and there-
fore, may similarly affect the excretion of organic nitrogen metabolites.
Substrate concentration determines growth rate. The relationship, developed
by Monod and modified by others, is [200,205]:
- b (6)
where y = net growth rate (time ) ,
Y = growth yield coefficient (mass/mass),
k = maximum rate of substrate utilization per unit weight of microorga-
nisms (time~l) ,
S = substrate concentration surrounding the organisms,
Ks = half-velocity constant (mass/volume) , equal to the substrate con-
centration where substrate utilization is l/2k,
113
-------
and
microorganism decay rate (time ).
Examination of Eq. 6 shows that at high growth rates, the relative effect of
organism decay is small, and a higher proportion of the cells will be growing
logarithmically. In light of previous discussion, it might be postulated that
increasing growth rate may result in higher concentrations of organic nitro-
gen metabolites being excreted. This needs to be evaluated.
Summary
SON is produced by bacteria; the major classes produced are nucleic acids,
proteins, and their respective degradation products. Factors affecting pro-
duction and excretion include: (1) concentration gradients, (2) starvation,
(3) presence of an energy source, (4) substrate-accelerated death, (5) avail-
ability of nutrients, (6) environmental stress, and (7) phase and rate of
growth. None of these individual factors is likely to be the dominant factor
controlling SON production by activated sludge; all may be involved.
SUMMARY AND BASIS FOR EXPERIMENTAL APPROACH
The concentration of SON in effluent from activated-sludge treatment of
municipal wastewaters varies from about 0.8 to 2 mg/1 (Section 6). The pur-
pose of this study was to determine the characteristics of this material and
to evaluate the relative distribution of effluent SON between that formed by
the biological treatment process itself and that originally present in the
influent stream. This section has discussed factors affecting SON production
and consumption, generally by pure cultures of bacteria. Whether these fac-
tors are important in mixed activated-sludge cultures, and if so, what con-
tribution each factor makes to the total effluent SON from secondary effluent
are unknown. This experimental study was designed so that the factors which
may contribute to effluent SON and which were thought to be of significance
in normal activated-sludge operation could be evaluated.
Factors affecting SON consumption (removal) that can be controlled dur-
ing activated-sludge treatment are environmental factors; the important ones
were thought to be detention time and organism concentration. Batch studies
with domestic wastewater as the substrate, varying aeration time and activated-
sludge MLSS concentration, were used to evaluate these factors and to provide
estimates for SONe (Eq. 2).
The factors which appear of importance in effluent SON production (SON
SONg, SONd) are: (1) concentration gradients, (2) starvation, (3) presenceP
of an energy source, (4) substrate-accelerated death, (5) nutrient availabil-
ity, and (6) phase and rate of growth. Factor 1 was evaluated in batch stu-
dies to measure the SON initially released upon dilution of concentrated
activated sludge (termed Initial Release Studies) as a function of MLSS con-
centration, culture characteristics, temperature, and substrate. The effect
of starvation was evaluated with batch studies, termed Organism Decay Studies,
in which activated sludge was diluted with tap water, and aeration time and
114
-------
MLSS concentration were varied. Factors 3 to 6 are related to the presence
of a utilizable substrate and were evaluated with batch systems fed various
synthetic wastes (Synthetic Feed Studies). Aeration time, SCOD, MLSS, NH3~N,
and substrate type were the variables evaluated. To obtain additional data on
the effect of phase and rate of growth, mixed cultures were maintained by
semi-continuous feeding to simulate activated-sludge operation, and were fed
a synthetic organic waste which contained little SON so that SON production
could be better evaluated.
LABORATORY ACTIVATED-SLUDGE CULTURES
Apparatus
Semi-continuous feed, laboratory activated-sludge (AS) cultures were
grown in two 9-liter pyrex glass bottles as pictured in Figure 8, and were
operated in a walk-in constant-temperature incubator maintained at 20 ± 1°C.
The magnetic mixers used increased culture temperature to about 21°C. Pres-
sure control valves permitted close control over the flow rates of carbon
dioxide and air.
Attached microbial growth was kept at a minimum by vigorously shaking
the bottles by hand daily, and scraping excess growth off container walls
with a brush monthly. Porous diffusers were cleaned frequently with chromic
acid cleaning solution.
Feed System
The two activated-sludge cultures were grown with the synthetic sub-
strate or waste listed in Table 55. A glucose-acetate mixture was selected
to provide a relatively simple carbon source containing no SON, which would
permit development of a somewhat heterogeneous microbial population. Stock
feed solutions of glucose, acetate, NltyCl, K2HP04, and Fe-Co were made sepa-
rately in deionized water and stored at 4°C. AS Culture 1 was started on
October 15, 1975 and AS Culture 2 on March 24, 1976.
Once the systems reached steady-state operation, COD/N ratios in the
feed ranged from 24 to 30 and N/P was kept constant at 5. The quantity of
nitrogen added was varied in an attempt to keep effluent inorganic nitrogen
relatively low to minimize the chances for nitrification, since an interfer-
ence with SON analysis was observed when NO^-N exceeded 10 mg/1 (Appendix B).
Addition of Fe and Co was commenced 133 days after the start of AS Culture 1
in an effort to control sludge bulking, and were included in the feed solu-
tion of AS Culture 2 at all times. The pH was controlled in the 7.0-7.5
range by mixing pure C02 with the air flowing into the cultures (about 1-2%
C02).
AS cultures were started by adding 30 ml of one-day-old settled sewage
from the Palo Alto, California, Regional Water Quality Control Plant to 6 li-
ters of feed (Table 55). After initial start-up, the feeding procedure in-
cluded wasting one of the six liters of mixed liquor (to make 0C=6 days),
followed by one hour of settling, wasting 3 liters of supernatant liquid
115
-------
COTTON
FILTER
-
^
H20
TRAP SATURATION
BOTTLE
, COMPRESSED
ft AIR
lh- GLASS «OOL
FILTER
C02
SATURATION TRAP
BOTTLE
-J0 PYREX BOTTLE
POROUS GLASS DIFFUSER
SAMPLE PORT
TEFLON STIRRING BAR
MAGNETIC MIXER IN
SPECIAL HOUSING
Figure 8. Laboratory activated-sludge system.
-------
TABLE 55. COMPOSITION OF SYNTHETIC WASTE
Compound
Concentration
Glucose
Na Acetate
NH4C1
K2HP04
6H20
CoCl2 • 6H20
Dilution water
0.75 g/1 as COD
0.75 g/1 as COD
50-62 mg/1 as N
10-12 mg/1 as P
0.15 mg/1 as Fe
0.15 mg/1 as Co
Stanford tap water
(culture effluent) and addition of 4 liters of feed solution. The mixture
was aerated for 23 hours and the feeding procedure repeated. This semi-
continuous feeding simulates a plug flow activated-sludge operation.
Effluent or mixed liquor were periodically monitored for pH, mixed liquor
volatile and suspended solids concentration (MLSS and MLVSS), SCOD, SON, NH3~N,
NO^-N, and NO^-N. Samples for SCOD and SON analyses were first filtered
through a glass fiber filter (Reeve Angel 934 AH), and then through a 0.45U
cartridge filter (Pall DFA 3001 AXA). SON analysis was by the Kjeldahl
method [3]. Microscopic observations of the culture were made approximately
once a month and dissolved oxygen was checked infrequently. SON of the feed
solution and tap water were determined from time to time.
BATCH EXPERIMENTS
This section describes common apparatus and experimental procedures used
for all batch studies. Details on individual experiments are presented later.
Apparatus
Most batch studies were conducted in 1-liter or 4-liter aspirator bot-
tles. One study was conducted using a 9-liter purex bottle identical to that
depicted in Figure 8. All glassware was acid-washed prior to use.
The batch systems were essentially identical to the activated-sludge
system. Liquid temperatures were maintained at 21 ± 1°C during all studies
and pH control was accomplished using an air-C02 mixture. Liquid volumes
used were 750-900 ml in the 1-liter aspirator bottles, 1.0-3.5 liters in the
4-liter aspirator bottles, and 5.5-7.0 liters in the 9-liter bottles.
Procedures
Mixed cultures for the batch experiments were activated sludge taken
either from the Palo Alto complete-mix activated-sludge treatment plant, or
from one of the laboratory activated-sludge systems. The cultures were set-
tled for concentration and the supernatant liquors decanted or siphoned for
subsequent analyses. The cultures were then added to the various temperature-
equilibrated and, in some cases, oxygen-saturated substrates to be tested.
117
-------
Samples were withdrawn at time 0 (within 30 seconds after culture addition)
and at specified times for SON, SCOD, MLSS, MLVSS, and in some cases NH3-N,
N02-N, and NO^-N analyses. Temperature, pH, and dissolved oxygen were also
monitored.
Separation of the soluble fraction was accomplished by one of the fol-
lowing methods:
1. Sedimentation for 5-10 minutes followed by filtering with glass fiber
filter (Reeve Angel 934 AH) and final filtration with a 0.45y Pall
DFA AXA cartridge filter, or
2. Centrifugation of 35-ml samples in a bucket-type centrifuge (Inter-
national Chemical Centrifuge, International Equipment Co., Boston,
Mass.) for 5-10 minutes followed by filtration with 0.45y Millipore
filters in syringe cartridges (SXOO 02500 Swinnex, Millipore Corp.).
All filters were washed with deionized water prior to use. Table 56 lists
the filtration method and SON analytical technique used for the various batch
studies.
SON removal studies were conducted using Palo Alto mixed liquor, unfil-
tered and filtered (Method 1) primary effluent, and in one case, a synthetic
waste containing glucose.
SON production was evaluated using laboratory AS cultures and various
concentrations and combinations of synthetic waste in tap water. NaHC03 was
added for pH control and feed solutions were saturated with oxygen using the
air-C02 mixture prior to addition of cultures. Initial release and organism
decay studies made use both of Palo Alto and laboratory AS cultures diluted
in tap water.
To investigate the effect of extended periods of aeration on release of
SON and SCOD during organism decay, two small aerobic digesters were operated
using waste laboratory activated sludge from Culture 1. Aeration conditions
were identical to those for activated-sludge operation. One digester was
maintained at a 9C of 20 days by daily semi-continuous feeding. An infinite
9C system was developed by saving waste AS culture sludge on 4 consecutive
days and aerating with no daily wasting or additional feeding.
TABLE 56. Filtraton and SON Analysis Used for Batch Studies
Batch
Study No.
1
2-4, 6-7
all others
Bottle Type
9-liter pyrex
1-liter aspirator
4-liter aspirator
Filtration
Method
2
2
1
SON
Analysis
Technicon
Technicon
Kjeldahl
118
-------
SON REMOVAL
Obj ectives
The objectives of this phase were to determine the effect of aeration
time, sludge washing, initial MLSS concentration, and substrate type on the
SON remaining (SONe) after activated-sludge treatment of a primary treated,
municipal wastewater. This would provide insights into the changes occurring
in SON during treatment. Batch studies were used exclusively for evaluating
the different variables.
Procedures
Statistical procedures required in this study are outlined in Appendix A.
Batch studies were conducted using concentrated mixed liquor from the Palo
Alto regional treatment plant for seed, and filtered and unfiltered primary
effluent and in one case, glucose, as substrate. Concentrated mixed liquor
diluted with Stanford tap water generally served as a control. Table 57 sum-
marizes the substrates, filtration techniques, and SON techniques used. pH
was maintained between 7.0 and 7.6, and excess air was added so that dis-
solved oxygen would not be limiting.
The initial concentrations of SON and SCOD in batch experiments were
calculated as follows:
+
SONOC =
SONf(Vf)
vf
(7)
TABLE 57. SUMMARY OF SUBSTRATES, FILTRATION, AND SON ANALYSIS
USED DURING REMOVAL STUDIES
Batch Study
Number
1
2
3
4
Substrates
Unfiltered Primary,
Tap Water
Unfiltered Primary
Filtered Primary,
Tap Water
Filtered Primary,
Glucose-Nutrient-Tap ,
Tap Water
Filtration
Technique
Centrifugat ion-
Syringe Millipore
Centrifugat ion-
Syringe Millipore
Centrifugation-
Syringe Millipore
Centrifugation-
Syringe Millipore
SON
Analysis
Technicon
Technicon
Technicon
Technicon
As described in Chapter 4, using 0.45y Millipore filters.
119
-------
where SONQC = calculated initial SON in mg/1,
SONf = SON of feed solution in mg/1 (primary effluent or tap),
SONml = SON of mixed liquor in mg/1,
Vf = volume of feed solution added in liters, and
vml = volume of mixed liquor added in liters.
SCODOC was calculated in a similar manner.
Effect of Aeration Time
Batch Studies 1 and 2 were conducted to evaluate the effect of aeration
time on SON and SCOD removal. Results are presented in Figure 9, 10, and 11.
In Batch Study 2 the effect of sludge washing on SON and SCOD removal was in-
vestigated. For one of the batch units, 6 liters of Palo Alto mixed liquor
was washed four times with tap water. This included settling, decanting, and
addition of tap water to give a 6-liter volume after each wash.
In all cases, SON decreased in concentration with time until some mini-
mum was reached, after which the concentration increased. At the point of
minimum concentration, SON removal was 70 percent in Batch Study 1, and 61
and 65 percent for the unwashed and washed sludges of Batch Study 2, respec-
tively. After the minimum, the increase in SON was dramatic during Batch
Study 1 and after one day of aeration reached a concentration greater than
originally present in the influent (Figure 9), and more than 4 times the min-
imum concentration. The increase was not as great during Batch Study 2 (see
Figure 11). No explanation for this difference in behavior can be given, and
the results of Batch Study 1 could not be duplicated.
SCOD removal paralleled SON removal in both studies. When SCOD reached
a minimum value, removals were 75, 64, and 67 percent, respectively, for
Batch Study 1 and the unwashed and washed sludges of Batch Study 2. After a
minimum SCOD was reached there was a gradual increase in SCOD as aeration
time increased, an observation noted by others [62,212,213,214,215,216]. The
magnitude of this increase (140 percent) was small compared to the 400-percent
increase in SON during Batch Study 1. In Batch Study 2, the increases were
around 140 percent for both SON and SCOD. This disparity in behavior during
Batch Studies 1 and 2 cannot be explained, but shows that SON and SCOD may
not behave similarly.
SON and SCOD were released during organism decay (see control systems,
Figures 9 and 10). However, in Batch Study 1 the increases from minimum SON
and SCOD values to measured 48-hour values (3.28 and 7.2 mg/1, respectively)
cannot be entirely attributed to decay of the initial population since SON
and SCOD released from tap water controls were only 1.35 and 4.2 mg/1, re-
spectively. The initial presence of a utilizable carbon source resulted in
increased release of SON.
Examination of data presented in Figures 9 and 10 leads to questions
120
-------
3
E
O
CO
6.0
5.0
4.0
3.0
2.0
1.0
0
o UNFILTERED PRIMARY EFFLUENT
ha CONTROL
OC 0 10 20 30 40
AERATION TIME, hr
50
Figure 9. Batch Study 1: Effect of aeration time on SON
remaining, MLSSom - 1300 mg/1.
121
-------
80
70
60
50
D>
. 40
Q
O
O
10
20
10
0
1
O UNFILTERED PRIMARY
EFFLUENT
a CONTROL
-Q
OC 0 10 20 30 40
AERATION TIME, hr
50
Figure 10. Batch Study 1: Effect of aeration time on SCOD
remaining, MLSSom - 1300 mg/1.
122
-------
5!
CJ
E
O
CO
3.0
2.5
2.0
1.5
1.0
.5
0
o UNWASHED
WASHED
OCO 10 20 30 40 50
AERATION TIME, hr
Figure 11. Batch Study 2: Effect of sludge washing and aeration time on
SON removal from unfiltered primary effluent, MLSSom = 1200 mg/1.
123
-------
about differences in characteristics of the SON present after certain periods
of aeration. Samples were taken from the control and primary-fed systems at
times 0, 6, and 24 hours and subjected to low-seed biodegradation (identified
as Biodegradation Study 3) to investigate these differences. Results and
their significance are discussed later.
Data in Figure 11 indicate that sludge washing had no apparent effect on
the general shape of the SON removal curve, and did not affect SON removal
efficiency. A t-test indicated that the differences (SONunwash - SONwagh) of
equivalent aeration times were statistically significant at a 99 percent con-
fidence limit. The average SON difference between the 11 measured data points
for each sample was 0.39 ± 0.17 mg/1. Calculated initial values differed by
0.45 mg/1. Washing of sludge resulted in some removal of refractory SON,
causing the major difference between the two systems. However, since a larger
volume of primary effluent was added to the unwashed system, a portion of the
difference (37 percent) can be attributed to additional refractory SON added
that was not removed during treatment.
Effect of MLSS Concentration
The effect of MLSS on SON and SCOD removals was evaluated in Batch Study
3 by adding varying amounts of Palo Alto mixed liquor to filtered primary
effluent and to tap water controls. Results are shown in Figures 12 and 13.
Higher removal rates for SON and SCOD are obtained with higher mixed liquor
suspended solids levels, as expected.
For the 1390 mg/1 MLSS system, SON and SCOD concentrations reached mini-
mum values and then gradually increased to about 1.5 times the minimum con-
centrations. These observations agree with those from Batch Study 2.
Examination of Figures 12 and 13, and comparison of SON/SCOD ratios as
aeration times increase indicate that SON and SCOD behave differently for the
3 lower MLSS systems; SCOD was removed at a faster rate. SON was produced,
as noted by its significant increase (95 percent level of significance) after
6 hours aeration in the 14 and 1.4 mg/1 systems, while SCOD decreased.
Primary factors which could have contributed to the SON production noted
above include concentration gradient, starvation, presence of an energy
source, substrate-accelerated death, and growth rate. Since primary effluent
was the substrate, and experimental conditions were controlled to maintain a
favorable environment, nutrient availability and environmental stress should
not have been important factors. Starvation (organism decay) of a portion of
the population may have contributed to SON production because significant
quantities of SON were released from control systems. The presence of an en-
ergy source can stimulate production by energy-dependent facilitated diffu-
sion, or by substrate-accelerated death. Growth rate remained high during
the first 6 hours of aeration since substrate concentration (SCOD) decreased
slowly due to low initial MLSS values. These sustained high growth-rate
conditions may have stimulated SON excretion. Similarly, SON production was
not observed for the two higher MLSS systems because organism concentrations
were higher, and SON was removed faster than it was produced and excreted.
124
-------
to
Ul
o>
E
- r>
o 2
en
0
OCO
T I I
,MLSSom (mq/Jt)
o 1.4
A 14
n 146
o 1390
10 20 30 40 50
AERATION TIME, hr
60 70
80
Figure 12. Batch Study 3: Effect of MLSS on SON removal from filtered primary effluent.
-------
[Si
120
100
80
en
a- 60
O
o
V)
40
20
0
—i 1
MLSSom
o 1.4
A 14
D 146
o 1390
OCO
10 20 30 40 50
AERATION TIME, hr
60 70
80
Figure 13. Batch Study 3: Effect of MLSS on SCOD removal from filtered primary effluent.
-------
Effect on Substrate Type
The previous batch studies used a feed which contained SON and thus made
it difficult to ascertain what portion of the SON present after extended aera-
tion periods was attributable to the SON initially added, and what portion was
produced biologically during substrate removal. In order to gain more under-
standing of this aspect, Batch Study 4 was conducted to compare SON concentra-
tions from batch systems fed a non-SON-containing feed and primary effluent.
For the non-SON feed, a glucose-nutrient-tap water solution was used
(SCOD = 189 mg/1). Inorganic nitrogen was added in the form of NH^Cl (SCOD/N
= 20) and phosphorus was added as KoHPO^ (N/P = 5). Filtered primary effluent
and tap water (control) were used to compare SON behavior. Results are shown
in Figures 14 and 15.
Removal of primary effluent SON and SCOD was similar to results obtained
in Batch Study 3 at approximately equivalent MLSS concentrations. Comparison
of the SON concentrations measured after 48 hours aeration with glucose (0.8
mg/1) and with filtered primary (1.62 mg/1) suggests that a significant por-
tion of effluent SON may result from production during substrate oxidation.
SON released by organism decay (control system) was 0.73 mg/1 after 48 hours,
a level not significantly different than observed for the glucose system.
These data suggest that the major fraction of the produced SON may result from
organism decay. Additional data on this aspect is presented later.
A statistically significant SON peak was observed after 4 hours aeration
for both the glucose and control systems. Differences between values meas-
ured after 4 hours and the calculated initial value were greater than the 0.28
mg/1 required for a 95 percent level of significance (Appendix A). Substrate
related factors, such as presence of a utilizable energy source, substrate-
accelerated death, and growth rate could explain the SON production by the
glucose system. However, these are not the only factors involved since the
control system (starvation conditions) exhibited similar peak behavior. Peak
SON concentration (0.75 mg/1) was significantly less than for the glucose
system (1.12 mg/1), but the shape of the SON curves were similar. Additional
information on these phenomena is given later.
Summary
Batch removal studies with primary effluent gave SON and SCOD removals
of 60-70 percent, results consistent with those reported in full-scale
activated-sludge plants (Section 6). The data provide an estimate that efflu-
ent SON (SONe) equals 38+6 percent of the influent SON (SONi), or 1.40 ±
0.46 mg/1 for the 8 samples tested. There is an aeration time at which SON
concentration reaches a minimum and beyond which it increases, primarily due
to organism decay. Washing activated sludge removes some refractory SON and
thus results in generally lower SON values. Higher MLSS concentrations re-
sult in higher SON removal rates.
SON is produced during activated-sludge treatment, and a major portion of
that produced may be due to organism decay. SON production during organic ox-
idation could account for up to 50 percent of secondary-effluent SON (SONe).
127
-------
CO
o>
e
o 2
CO
0
OCO
~T I I I
o PRIMARY EFFLUENT
n GLUCOSE
A CONTROL
10 20 30 40 50
AERATION TIME, hr
60 70
Figure 14. Batch Study 4: Effect of substrate on SON remaining, MLSSom - 180 mg/1.
80
-------
200
160
0)120
E
•»
§ 80
en
40
0
OCO
~~i i I !
o PRIMARY EFFLUENT
n GLUCOSE
A CONTROL
10 20 30 40 50
AERATION TIME, hr
60 70
80
Figure 15. Batch Study 4: Effect of substrate on SCOD remaining MLSS - 180 tng/1
om
-------
SON PRODUCTION
Obj ective
The overall objective of the following experiments was to evaluate SONp,
the SON produced during activated-sludge treatment. Both semi-continuous and
batch systems were used and were fed low-SON substrates. Specific objectives
are given under each individual experimental phase. Statistical procedures
used in this study are outlined in Appendix A.
SON Production by Laboratory Activated-Sludge Cultures
Objective--
The objective of this study was to monitor the behavior of SONp during
activated sludge start-up, while the culture passes through different phases
of growth, and during steady-state operation. SCOD and other constituents,
and culture characteristics were also monitored. These data provide general
information about daily fluctuations in SONp and SCOD, and provide clues about
factors affecting SONp.
Feed Characteristics—
The low-SON feed (glucose plus acetate) described under Experimental Pro-
cedures was used as the substrate. Stanford tap water was used for dilution
and was found to contain SON with a concentration of 0.05 ± 0.03 mg/1 (range:
0.01-0.11 mg/1) for 42 samples analyzed between August, 1975 and September,
1976. During the period from April through the middle of May, the tap water
temporarily came from a different source and SON was a bit higher, ranging
from 0.11-0.23 mg/1 for 5 samples analyzed during this period.
Feed solution SON, containing all the nutrients listed in Table 55, aver-
aged 0.10 ± 0.04 for 21 samples analyzed at various times during the research.
Measured feed-solution SCOD averaged 1455 mg/1, 97 percent of the theoretical
value calculated.
Activated-Sludge Development—
Laboratory activated-sludge cultures were started as described previously.
Behavior of effluent SON and MLSS during start-up is shown in Figures 16 and
17. During start-up, the culture was not purposely wasted. After start-up,
a given percentage of the culture was removed each day to maintain a defined
solids retention time (6C).
SON reached its maximum concentration during the start-up. While still
in the start-up phase (Days 13 to 32 and 7 to 23 for AS Cultures 1 and 2, re-
spectively) , SON concentration sharply decreased, and continued to decrease
as the rate of MLSS increase began to decrease. SON decreased further when
culture was wasted for control of 6C until steady-state values were reached.
Values for effluent SCOD ranged from 36-145 mg/1 for AS Culture 1 and 26-60
mg/1 for AS Culture 2 during the period of sharp SON increase and decrease,
and from 10-30 mg/1 for both cultures when SON and MLSS values leveled off.
The SON released during culture start-up was primarily the result of ex-
cretion of metabolites since the feed contained little SON by comparison.
130
-------
3000
- 2500
BOETON
OSON
MLSS
0C=10 days—*
o»
E
1500 co
CO
10 20 30 40 50 60 70 80
TIME, days
Figure 16. AS Culture 1 start-up: SON and MLSS concentrations vs time of operation.
-------
LO
NJ
3
a
£
o
CO
i—:—i 1
START 0cs6days
oSON
QMLSS
20 40
60 80 100 120 140 160
TIME, days
3000
- 2500
- 2000
1500
CO
- 1000
- 500
180
0
Figure 17. AS Culture 2: SON and MLSS concentrations vs time of operation (all data).
-------
Data in Table 58 indicate that phase and rate of growth did not control the
production of SON. SON decreased significantly while still in the logarithmic
start-up phase. Comparison of steady-state SON concentration (y = 0.17 day"1)
with SON concentrations measured during culture start-up (y = 0.15 day"1) sug-
gests that growth rate does not control SON production.
Apparently, factors other than rate and phase of growth have a signifi-
cant influence on SON production. Concentration gradients are not likely to
be a factor since produced SON levels were so high. Nutrient availability
and environmental stress should not affect production since feed solutions
contained most of the required nutrients, and culture environment (tempera-
ture, pH, etc.) was controlled. The effect of starvation and substrate accel-
erated death are not known because it was impossible to determine if starva-
tion conditions existed during the logarithmic start-up phase. The presence
of an energy source probably enhanced energy-dependent excretion of SON meta-
bolites. Factors controlling SON production by heterogeneous activated-sludge
cultures seem to be complex and interrelated, and definitive statements about
the effect of individual factors is difficult. Nevertheless, the observed be-
havior remains; SON was produced in large quantities during culture start-up.
As MLSS concentrations increased to levels greater than 1700 mg/1, SON
decreased, most likely due to utilization of produced SON by the increased
concentration of organisms, and perhaps, due to decreased production and ex-
cretion of SON.
TABLE 58. THE EFFECT OF GROWTH RATE ON AS CULTURE 2
EFFLUENT SON DURING CULTURE START-UPa
Day of
Operation
7
10
12
13
15
16
17
18
19
20
23
27
steady-state
MLSS
(mg/1)
232
412
468
680
1050
1220
1350
1480
1630
1730
2100
2010
2210
Effluent
SON
(mg/1)
2.47
1.99
2.32
2.40
2.87
2.11
1.83
1.26
1.02
0.94
1.43
2.11
0.26
Effluent
SCOD
(mg/1)
43
40
44
44
49
42
44
38
31
29
49
60
18
y .
(day X)
0.14
-
0.17
Logarithmic growth occurred from Days 7 to 23 (see text) .
Average values for Days 104 to 176.
133
-------
The second, smaller SON peak observed during AS Culture 2 start-up may
have been a response to a temperature increase from the normal of 20°C to a
maximum of 29°C on Day 27 because of an incubator failure. Culture color
changed from a dark, golden brown to a much lighter brown between Days 23 and
30. Such color changes have been noted to signify subtle population shifts
which affect effluent quality [217]. Other explanations are of course pos-
sible.
Steady-State Operation—
Steady-state operation was considered to be attained when MLSS concentra-
tion leveled off and remained relatively constant with time (± 250 mg/1).
Table 59 contains a summary of steady-state data for periods meeting this cri-
teria. AS Culture 1 initially reached steady-state operation after about 70
days (Figures 16 and 18), and AS Culture 2 after 80 days (Figure 17). Between
Days 130 and 199, severe sludge bulking occurred in AS Culture 1 and several
techniques (discussed later) were tried to alleviate the problem.
A t-test indicated the difference between the mean feed SON (0.10 ± 0.04
mg/1) and the mean culture SON values listed in Table 59 were statistically
significant at the 99 percent level of confidence; this indicates SON was in-
deed produced during steady-state operation.
Effluent NH3-N ranged from 0.2 to 19.8 mg/1 at steady-state operation for
the two cultures, and nitrate-nitrogen never exceeded 0.41 mg/1, thus elimi-
nating any concern over NOo-N interference with the Kjeldahl procedure for SON.
Dissolved oxygen ranged between 5-6 mg/1 one hour after feeding to near 8 mg/1
at the end of the 23-hour aeration cycle. Routine microscopic observations
of the cultures during steady-state operation showed predominance of large
floes of bacteria with free-swimming protozoa and an occasional rotifer. More
TABLE 59. SUMMARY OF STEADY-STATE DATA FOR ACTIVATED-SLUDGE CULTURES
Data Set
Number
I
2
3
AS Culture Number:
Days of Operation
1: Days 69-124
1: Days 204-293
2: Days 104-176
SON*
0.33 ± 0.04
(0.28-0.39)
n = 14
0.63 ± 0.19
(0.40-1.11)
n = 16
0.26 ± 0.05
(0.19-0.37)
n = 14
SCOD*
12.0 ± 3.1
(7.2-19.4)
n = 15
15.9 ± 5.0
(9.3-28.2)
n = 16
18.1 + 6.0
(7.7-33.6)
n = 14
MLSS*
2060 ± 121
(1870-2300)
n = 22
2270 ± 153
(2040-2540)
n = 15
2210 ± 72
(2110-2330)
n = 14
*
All concentrations in mg/1; data are mean, standard deviation, and
range (Brackets), n = number of samples analyzed.
134
-------
CO
Ui
O>
E
O
en
5
4
0
BULKING
1
o SON
n MLSS
3500
3000
2500
2000
1500
1000
o>
e
en"
en
20 60 100 140 180 220 260 300 340
TIME SINCE START UP, days
500
Figure 18. AS Culture 1, Days 20-340: SON and MLSS concentrations vs time of operation.
Squares represent MLSS and circles represent SON.
-------
filamentous organisms were observed during the period from Days 204-293 than
the period from Days 69-124 for Culture 1. Culture color during steady-state
periods ranged from light brown and grey to dark brown for AS Culture 1, and
was predominantly light grey for AS Culture 2.
Effect of Population Changes—
Population shifts were suggested by obvious changes in culture color,
settling characteristics, and microscopic examination (Table 60).
A t-test indicated that respective effluent SON and SCOD mean values for
Data Sets 1 and 2 (Table 59), representing different time periods for Culture
1, were significantly different at a 99 percent confidence level. Thus, ob-
served culture differences affected effluent characteristics, since all other
experimental variables were held constant over those time periods. Other re-
ports support this observation [207,217].
SON and SCOD data from AS Culture 2 (Data Set 3) were compared with Data
Sets 1 and 2 (Seed Culture 1) using the t-test. Sample means differed
TABLE 60. CHARACTERISTICS OF MIXED LIQUOR SUSPENDED SOLIDS
FOR ACTIVATED-SLUDGE CULTURES
Data Set
Number
1
2
3
AS Culture Number:
Days of Operation
1: Days 69-124
1: Days 204-293
2: Days 104-176
Culture
Color
Light brown
Dark brown to
light brown
to whitish
grey to dark
grey. Some
foaming near
end of period
Light grey
Settling Char-
acteristics of
the Mixed Liquor
Suspended Solids
after 1 hour
5 liters settled
to about 500 ml
5 liters settled
to 200-900 ml
Sliters settled
to 500-800 ml
General
Microscopic
Examinations
Many large
bacterial
floes; free-
swimming pro-
tozoa; some
rotifers
Some bacterial
floes ; some
filamentous
forms present;
few protozoa;
no rotifers
Large bacterial
floes; free-
swimming pro-
tozoa; no
rotifers
— — -
136
-------
significantly at the 99 percent level for the SON data. SCOD means for Data
Sets 1 and 3 were significantly different, while for Data Sets 2 and 3, they
were not. Culture characteristics were different (Table 60) which may explain
the differences in effluent SON and SCOD data.
Inspection of Figures 16, 17, and 18, and data listed in Table 59 show
larger fluctuations in SON, SCOD, and MLSS for Data Set 2 than for Data Sets
1 and 3. In light of the previous discussion, and noting culture characteris-
tics listed in Table 60, these larger fluctuations may be due to culture popu-
lation shifts.
Data on the effect of population changes on SON production also empha-
sizes that growth rate did not control SON production. Growth rate was con-
stant (0.17 day"1) at all times (Tables 59 and 60), and yet significant
changes in effluent SON levels occurred.
Sludge bulking was observed in AS Culture 1 approximately 130 days after
initial start-up. After the daily 1-hour settling period the 5 liters of
mixed liquor settled to levels ranging from 1.8 to 4 liters. Several poten-
tial remedies were tried including the addition of a Fe-Co solution. Effluent
SON began to rise slightly as the degree of bulking increased. Commencing on
Day 132, mixed liquor suspended solids which did not settle below the 1.8-
liter level were wasted, resulting in the drop in MLSS noted in Figure 18.
MLSS levels later began to increase until levels greater than 2000 mg/1 were
reached. An increase in SON was observed as MLSS increased rapidly, similar
to that noted during culture start-up. SON then decreased rapidly as MLSS
concentration leveled off. However, SON again rose sharply after the initial
decrease as depicted in a second peak near Day 180 in Figure 18. No notice-
able culture changes were observed during this period, and no suitable explan-
ation for this second peak can be given. After the bulking problem, from
Days 204 to 293, SON, SCOD, and MLSS varied more than during the period from
Days 69 to 124, and were significantly different for the two periods as al-
ready noted.
Summary—
Activated-sludge cultures were developed to monitor SON production dur-
ing start-up and during steady-state, semi-continuous operation. During the
start-up phase, a maximum of 5 mg/1 SON was produced. When cultures were
operating at steady state, produced SON levels were 0.2-0.8 mg/1. Although
growth rate and phase may affect the levels of SON produced by activated-
sludge treatment, they were not the major factors involved here. Other fac-
tors such as starvation, substrate-accelerated death, energy-dependent excre-
tion in response to the presence of a substrate, and bacterial utilization
of produced SON were most likely involved. Significant differences in pro-
duced SON concentrations during steady-state operation were associated with
different culture characteristics such as color, settling characteristics,
and microscopic properties.
Initial SON Release
Objectives—
During batch removal experiments, SON was released from Palo Alto
137
-------
activated sludge upon dilution with tap water (see difference between calcu-
lated and observed time zero SON for controls, Figures 9 and 10). Subsequent
dilution of activated-sludge solids resulted in release of additional SON.
These data agree with observations that release of SON may be affected by con-
centration gradients [150,151,152]. A set of experiments was conducted to
determine if this release of SON was real and reproducible, and what factors,
if any, affected magnitude of the release.
Procedures—
Samples were taken within 30 seconds after dilution of culture with water
and filtered as described under Batch Experiments to prevent further SON re-
lease or consumption. Laboratory AS cultures were used, experiments with Cul-
ture 1 being conducted 220 to 302 days after culture start-up, and with Cul-
ture 2, 124 to 173 days after culture start-up. The effects of MLSS concen-
tration, culture source, temperature, and dilution water characteristics were
studied.
The increase in SON (SONq, mg/1) during those experiments was calculated
as follows:
SON = SON - SON (8)
q om oc
where SONom is the initial measured SON (mg/1) and SONOC is the calculated
initial SON defined by Eq. 7 (mg/1). SCODq is defined in a similar manner and
is a measure of the additional SCOD released upon dilution of concentrated
activated-sludge culture.
Effect of MLSS—
The effect of MLSS concentration on SON- was evaluated using AS Culture
1. Experiments were conducted on 10 different days, and each day from one to
five different MLSS concentrations were evaluated. Results are shown in
Figures 19 and 20.
Average values of SONq and SCODq were 0.11 ± 0.04 and 4.0 ± 2.9 mg/1,
respectively, and were both significantly greater than zero at the 99 percent
level of confidence. Initially released SON and SCOD were independent of
MLSS concentration, linear correlation coefficients (r) were only 0.02 and
0.37, respectively, for 23 MLSS concentrations ranging from 7 to 2390 mg/1.
The fact that initially released SON concentration was independent of
MLSS concentration led to a hypothesis that certain nitrogen-containing orga-
nics diffuse through cell walls of microorganisms to establish an equilibrium
concentration between the surrounding fluid and the microorganisms. This
equilibrium concentration is independent of microorganism concentration. The
hypothesis is consistent with discussion of the effect of concentration gra-
dients [150,151,152] given previously.
Effect of Culture-
Initial SON release was evaluated for activated-sludge cultures from
Palo Alto and from AS Cultures 1 and 2 as a function of MLSS concentration.
Results are summarized in Table 61.
138
-------
1-0
.t
5! '3
cn
1 .2
CT
~7
^_
o
<" .1
I I 1 1 1 1 1 1 1
/SONq = 0.11 ±0.04
? o J2>-- 9— / - -- - -- -
o° 0 I / ^ o
o o0 o -
_ r\ _ _ o. — — — — _(^\_ _ _ — . — — _^s_ .
^-^ (J U- - -
-------
12
10 -
8
i r
i 1 1 r
o
o
o
o
O
O
O
j I
o
o
SCODq = 4.0 ± 2.9 mg/j0
-er
O
O
O o
O
O
J I
J I
0 250 500 750 1000 1250 1500 1750 2000 2250 2500
MLSS,
Figure 20. Initial release of SCOD for Seed Culture 1 as a function of MLSS concentration.
-------
TABLE 61. EFFECT OF CULTURE ON SONn AND SCOD,,
Culture
Palo
Alto AS
AS Cul-
ture 1
AS Cul-
ture 2
MLSS
Range
(mg/1)
2-1400
7-2390
120-1910
SONq
(mg/1)
0.20±0.09
(0.09)*
0.11±0.04
(0.02)*
0.05 + 0.01
(0,04)*
SCOD
(mg/1)
2.0 + 2.0
(1.0)*
4.0±2.9
(0.6)*
1.2+0.4
(1.2)*
Number of
sludge
samples used
5
10
3
Number of MLSS
values studied
(n)
8
23
6
( )* SONq and SCODq needed to be statistically greater than zero at a 95
percent level of confidence for n samples.
The individual SON means in Table 61 are significantly greater than zero
at a 99 percent level of confidence, indicating that SON is indeed released
when activated sludge is diluted. Release of SCOD was significantly greater
than zero at a 99 percent confidence level for AS cultures, but only at a 95
percent level for Palo Alto AS.
Sample means, were compared using the "one-factor analysis of variance"
method described by Crow, Davis, and Maxfield [220], which states that dif-
ferences in sample means (when comparing more than 2 means) are not considered
significant unless those differences are large when compared with variations
within samples. The method involves calculation of F values. Mean values of
the three cultures for SONq were significantly different at the 99.5 level of
confidence, while mean SCOD values were significantly different at the 95
percent level. Thus, it can be concluded that the concentration of SON ini-
tially released is different for different cultures.
A second dilution of Palo Alto AS resulted in the release of additional
SON (Table 62), providing support for the "equilibrium" level hypothesis. If
the effect is truly an equilibrium effect, bacteria upon dilution should al-
ways excrete SON until an equilibrium concentration is reached, at least un-
til the internal level of excretable SON drops below some critical level.
TABLE 62. SON RELEASED BY SEQUENTIAL DILUTION OF PALO ALTO ACTIVATED SLUDGE
Dilution
Number
1
2
SONq
(mg/1)
0.28
0.45
SCODq
(mg/lT
3.2
3.1
MLSS
(mg/1)
1400
1400
141
-------
Testing the Equilibrium Hypothesis—
Three experiments were conducted to evaluate the equilibrium release
hypothesis. In the first experiment, AS Culture 1 was added to both tap water
and filtered AS Culture 1 supernatant. For the second and third experiments,
AS Culture 1 was first added to tap water, the solids were settled, and the
resulting supernatant liquor was filtered; then, additional AS Culture 1 was
added. If the equilibrium hypothesis were correct, no additional SON should
be released in the filtered supernatant samples, since they should already
contain the equilibrium level SON. Results (Table 63) indicate that no addi-
tional SON was released as hypothesized. The values of SONq and SCODq for
the controls are significantly greater than for the filtrate at a 95 percent
level of confidence.
Effect of Temperature—
The effect of three different temperatures on initial SON release was
evaluated. AS Culture 1 was equilibrated at the given temperature for at
least one hour before use. Results in Table 64 indicate all values for SONq
and SCODq were significantly greater than zero, but exhibited no dependence
on temperature over the range studied as determined by correlation analysis.
Effect of Substrate—
The effect on initial SON release of mixing concentrated activated sludge
with different substrates was evaluated using AS Culture 2. The substrates
tested were as follows:
1. Tap water + NaHC03 (100 mg/1 as CaC03, as for all previous studies),
2. Deionized water,
3. Tap water + NaHC03 + 15 mg/1 NH^Cl-N + 3 mg/1 K2HP04-P,
4. Tap water + NaHC03 + 15 mg/1 NItyCl-N + 3 mg/1 K2HP04-P + 150 mg/1
glucose-COD + 150 mg/1 NaAcetate-COD, and
5. Seed Culture 2 filtrate.
All solutions were air saturated prior to culture addition. Results are
presented in Table 65.
Although there were differences in SONq for the various systems, the dif-
ferences were not significant at a 95 percent level of confidence. Because of
the low SON concentrations, analysis of several samples would be required to
statistically evaluate whether the differences noted are significant. Addi-
tional information about the effect of substrate on the magnitude of'SON is
presented in the section on Significance of SONq and dSONt Peaks.
Estimation of the True Equilibrium SON—
SONq, the SON initially released upon dilution of activated sludge is
an apparent measure of an equilibrium SON concentration established between
the organisms and their surrounding environment. The true equilibrium SON
termed SONeq, will be the sum of SONq plus the equilibrium SON remainine in
the activated-sludge culture to be diluted, that is, a portion of SONOC (Eq.7).
142
-------
TABLE 63. INITIAL RELEASE OF SON AND SCOP UPON EXPOSURE TO TAP WATER,
AS CULTURE 1 FILTRATE, AND RELEASED ORGANICS
^\^
Parameter0 ^^\^^
SONOC
SCODOC
MLSSom
SONom
SCODom
SON,,
q
SCOpq
Experiment 1 (255)a
Culture SON =0.56 mg/1,
SCOD =13.3 mg/1
Tap Water
(Control)
0.12
3.1
1776
0.25
6.9
+ 0.13
+ 3.8
AS Culture
Filtrate
0.56
13.3
1776
0.57
12.9
+ 0.01
- 0.4
Experiment 2 (256)a
Culture SON =0.43 mg/1,
SCOD =9.3 mg/1
Tap Water
(Control)
0.10
3.1
904
0.19
5.2
+ 0.09
+ 2.1
Filtered Organics
Released
from Control
0.20
5.4
904
0.19
4.8
- 0.01
- 0.6
Experiment 3 (302)a
Culture SON = 0.79 mg/1,
SCOD =25.8 mg/1
Tap Water
(Control)
0.04
2.4
252
0.15
4.4
+ 0.11
+ 2.0
Filtered Organics
Released
from Control
0.16
4.8
330
0.16
6.4
0
+ 1.6
a( ) Day of AS Culture Operation.
bAll in mg/1.
SONq and SCODq required to be significantly greater than zero at a 95 percent level of confidence
are 0.06 and 1.7 mg/1, respectively, for n = 3.
CO
-------
TABLE 64. EFFECT OF TEMPERATURE ON INITIAL RELEASE OF SON AND SCOD£
Temp.
10°C
20°C
30°C
SON b
(mg/D
+ 0.09
+ 0.08
+ 0.08
SCOD c
(mg/1)
+ 2.6
+ 2.2
+ 2.6
MLSS
(mg/1)
570
520
560
aSeed Culture 1 sludge used; SON = 0.73 mg/1, SCOD = 20.1 mg/1.
bSON of 0.
confidence
CSCQDq of 1
confidence
07 mg/1 required for statistical significance at 95 percent
limit for n = 3.
.7 mg/1 required for statistical significance at 95 percent
limit for n = 3.
This sum can be written as
(SON
SON = SON + (SON - SON ) k—p
eq q oc tap SON
(9)
Rearranging terms and solving for SON ,
SON
SONeq = (SONe) I _ SQN
f SON
oc tap_
(10)
in which SONe is the SON of the activated-sludge culture, SONnr is the calcu-
lated initial SON, and SONtap is the SON of the dilution water. A similar
equation can be developed for SCODe . Mean values calculated for SON and
SCODeq are listed in Table 66. q
All SONeq and SCODeq means were significantly greater than zero at a 99
percent level of confidence, and are independent of MLSS concentration (r val-
ues for SONeq ranged from 0.05-0.24, and from 0.14-0.59 for SCOD). SONe and
SONq means (Table 61) are not significantly different for any of the cultures;
the same was true for SCOD means.
Summary—
SON and SCOD are released upon dilution of activated sludge with tap wa-
ter. The magnitude of the initially released SON is independent of MLSS con-
centration and temperature, but dependent on culture characteristics.
An equilibrium hypothesis was proposed, tested, and found to be consis-
tent with the data and with reports by others [150,151,152]. Organisms
144
-------
TABLE 65. INITIAL RELEASE OF SON AND SCOD: COMPARISON OF SUBSTRATES*
^^\Subs trate
Parameter ^^\^^
MLSS (mg/1)
SONQC (mg/1)
SCODoc (mg/1)
SONom (mg/1)
SCODom (mg/1)
SONq (mg/l)b
SCODq (mg/l)C
Tap Water
(Control)
570
0.03
3.1
0.08
3.9
+ 0.05
+ 0.8
Deionized
Water
590
0.01
1.0
0.08
4.3
+ 0.07
+ 3.3
Tap Water plus
Inorganics
590
0.04
3.1
0.13
5.1
+ 0.09
+ 2.0
Tap Water plus
Inorganics plus
Organics
600
0.05
270
0.21
238
+ 0.16
-32.1
AS Culture 2
Filtrate
620
0.37
23.8
0.35
22.5
- 0.02
- 1.3
aAS Culture 2; SON = 0.37 mg/1, SCOD = 24 mg/1.
± 0.10 mg/1 change required for 95 percent confidence limit (n = 1) .
C± 2.8 mg/1 change required for 95 percent confidence limit (n = 1).
Definitions given in text and in Appendix A.
-------
TABLE 66. EQUILIBRIUM SON AND SCOP VALUES FOR THE CULTURES STUDIES
Culture
Palo Alto AS
AS Culture 1
AS Culture 2
MLSS Range
(mg/1)
2 - 1400
7 - 2390
120 - 1910
SONeq
(tng/D
0.22 ± 0.10
(n = 8)
0.12 + 0.04
(n = 23)
0.06 ± 0.01
(n = 6)
SCODeq
(mg/D
2.3 ± 2.5
(n = 8)
4.5 ± 3.6
(n = 23)
1.3 + 0.4
(n = 6)
excrete certain SON compounds in order to reach an equilibrium concentration
with these particular compounds. These compounds may be different from the
majority of organics comprising effluent SON, since effluent SON concentra-
tions are generally much higher. Regardless, "equilibrium" SON may comprise
a significant portion of the SON produced during treatment. Mean SON val-
ues for the three sludges examined were 0.22, 0.12, and 0.06 mg/1; correspond-
ing mean SCODeq values were 2.3, 4.5, and 1.3 mg/1.
SON Release during Organism Decay
Obj ective—
SON released during organism decay (starvation conditions) is one pos-
sible source of activated-sludge effluent SON (see SON Production and Excre-
tion by Bacteria). During initial studies on SON removal, significant quan-
tities of SON and SCOD were released in tap water control systems, presumably
due mostly to organism decay. The objective of the following studies was to
better define the magnitude of the release during decay (SONd and SCODd), and
to evaluate factors affecting this release.
Procedures—
Organism decay during starvation was studied by aerating activated-sludge
cultures in tap water. Data from previous batch studies using Palo Alto acti-
vated sludge were compared with similar batch studies using AS Cultures 1 from
Days 268 to 293 and AS Culture 2 from Days 124 and 173. Tap water withNaHCO^
(100 mg/1 as CaCOs) was aerated with an air-C02 mixture for 15-30 minutes
prior to culture addition. Variables investigated were aeration time and
MLSS concentration. '
Effect of Aeration Time—
QP™ f gUye%21' 22' and 23 illustrate the effect of aeration time on SON and
SCOD levels for concentrated Palo Alto AS (Batch Study 3) and AS Culture 1
Additional data on tap water systems for Palo Alto AS are found in Figures
y 9 JLO j JL^- y cincl 15.
146
-------
OCO 5 10 15 20
AERATION TIME, hr
25
Figure 21. The effect of aeration time on SON and SCOD release during organism
decay at different MLSS values using AS Culture 1 (Day 274).
147
-------
00
oco
20 30 40 50
AERATION TIME, hr
60 70
80
Figure 22. The effect of aeration time on SON release during organism decay at different
MLSS concentrations using Palo Alto activated sludge (9-9-75).
-------
-e-
vo
a
o
o
30
25
20
- l5
10
0
MLSSom (mq/Jt)
OCO 10 20 30 40 50
AERATION TIME, hr
60 70
80
Figure 23. The effect of aeration time on SCOD release during organism decay at different
MLSS concentrations using Palo Alto activated sludge (9-9-75).
-------
SON released during aeration (dSONt) is defined as
dSONt =
(ID
where SONtm is the measured SON (mg/1) at aeration time t (hours), and SONQC
is the calculated initial SON (mg/1). Table 67 lists coefficients (a, b)
obtained from regression analysis of data from Figures 21, 22, and 23 using
the following equation:
dSON,
a + bt
(12)
Similar expressions for dSCODt were evaluated. SON released during organism
decay was significantly correlated to aeration time at higher MLSS levels
where analytical errors were not predominant. A similar observation was made
for dSCOD. Peak concentrations in SON were noted to occur within the first
few hours of aeration. This effect is explored in more detail in a later
section.
The effect of long aeration periods on SON and SCOD released by organism
decay was studied using aerobic digestors (see Batch Experiments). Two sys-
tems were operated. One with an "infinite" 8C had increases from initial SON
and SCOD values of 0.36 and 10.6 mg/1, respectively, to 1.11 and 17.3 mg/1.
Nitrate nitrogen after 59 days was less than 10 mg/1, sufficiently low to
avoid interference with SON analysis. During this period MLSS decreased from
2700 to 2000 mg/1 and total organic nitrogen (TON) dropped from 242 to 196
mg/1. Since the increase in SON was only 0.75 mg/1 (1.11 - 0.36), at least
98 percent of the 46 mg/1 decrease in particulate TON was decomposed to
TABLE 67. SUMMARY OF REGRESSION ANALYSIS COEFFICIENTS CORRELATING
SON AND SCOD RELEASE WITH AERATION TIME
Sludge Source
AS Culture 1
it
it
Palo Alto AS
II
MLSS
516
68
7
1340
140
9
1.4
dSONt*
a
0.23
0.13
0.08
0.31
0.23
0.27
0.23
b
0.01
- 0.003
-0.001
0.01
0.01
0.002
0.001
Correlation
Coefficient
(r)
0.63
0.66
0.21
0.90
0.57
0.37
0.35
dSCODt'f"
a'
1.5
1.7
0.9
1.8
1.3
-0.03
1.1
b'
0.33
-0.07
-0.03
0.31
0.06
0.05
0.01
*
Correlation
Coefficient
(r')
0.99
0.54
0.55
0.96
0.63
0.82
0.26
dSONt = a + bt.
tdSCODt = a1 + b't.
150
-------
inorganic nitrogen. After 31 days of operating the 20-day 6C system, SON and
SCOD increased from 0.33 to 0.60 mg/1 and 11.3 to 31.5 mg/1, respectively,
MLSS and TON decreased from 2200 to 1700 mg/1 and 190 to 123 mg/1, respec-
tively, and NO^-N remained at zero. For this system, over 99 percent of the
particulate TON lost was degraded to NH3. Long periods of aeration do not re-
sult in large concentrations of released SON; organisms efficiently oxidize
most of the cellular nitrogen-containing organics lost during organism decay.
The same was found for SCOD.
Effect of MLSS—
The effect at different AS Culture 1 MLSS levels on the quantity of SON
released during organism decay was evaluated by measuring the change in SON
concentration over 23 hours of aeration, dSON23, determined as follows:
dSON23 = SON - SON (13)
is the SON (mg/1) measured after 23 hours of aeration and SONOC is the
calculated initial SON concentration. Calculations for dSCOD23 were made in
a similar way. Experiments were conducted on four different days (between
Days 283 and 293), and from two to four MLSS values were tested each time.
Results are presented in Figure 24.
Coefficients calculated from linear regression analyses using the follow
ing equations are listed in Figure 24:
dSON23 = a + b MLSS (14)
dSCOD = a + b MLSS (15)
23
Correlation coefficients of 0.96 and 0.86 were obtained for Eqs. 14 and 15,
respectively, both indicating significant correlation of SON and SCOD release
with MLSS of a 99 percent level of confidence (method of Crow et al. [220].)
Standard errors of estimate for the two equations were 0.02 mg/1 and 0.7 mg/1,
respectively [218]. Values for SON and SCOD for AS Culture 1 during these
experiments were 0.09 ± 0.03 mg/1 and 4.0 ± 2.0 mg/1, respectively.
These data support the equilibrium level hypothesis proposed earlier. If
there is an equilibrium concentration of SON excreted by organisms, this mate-
rial should always be present and should not be degraded during organism de-
cay, i.e., dSON23 should not decrease below SONeq for any MLSS concentration.
Extrapolation of linear regression equations to MLSS = 0 should yield SONeq
and SCODeq. Although SONeq was somewhat larger than the MLSS = 0 intercept
(a), SONeq and a were not significantly different, implying that dSON23 did
not decrease below SONeq. Similar observations were made about dSCOD23? how-
ever the considerable scatter in the SCODeq data makes detection of signifi-
cant differences difficult.
Regression analyses of five MLSS concentrations (range: 110 - 1910 mg/1)
of AS Culture 2 gave the following equations with significant correlation co-
efficients of 0.94 and 0.91, respectively:
151
-------
Q
O
O
8
4
dSCOD23 = 1.9 + 0.0016 MLSS
o
i
ro
cvi
I
•o
0
r 1 1 1
dSON23 = 0.05+ 0.00008 MLSS
500 1000 1500 2000 2500
MLSS, mg/J
Figure 24. The effect of MLSS concentration on SON and SCOD
release during organism decay of AS Culture 1.
152
-------
dSON23 = 0.02 + 0.00008 MLSS
dSCOD23 = 0.9 + 0.0021 MLSS
Standard errors of estimate were 0.02 and 0.7 mg/1, respectively, SONeq and
SCOD were 0.06 ± 0.02 and 1.3 ± 0.4 mg/1, respectively, values not signifi-
cantly different than dSON23 and dSCOD23, for MLSS = 0. These data provide
additional support for the equilibrium hypothesis, and for the dependence of
SON release during organism decay on MLSS concentration.
Additional data about the effect of MLSS can be obtained from experiments
with Palo Alto Activated Sludge (Figures 22 and 23). Linear regression analy-
sis was used to correlate dSONt (Eq. 11) and dSCODt with MLSS. Aeration times
of 12, 24, and 48 hours were selected for correlation of dSONt and dSCODt
since they should give representative values of the SON and SCOD released
during organism decay.
Correlation coefficients listed in Table 68 show that SON released during
organism decay is significantly correlated to MLSS concentration (Crow et al.
[220]), except for the 24 hours aeration case; here, analytical error may have
hindered the correlation. SCOD released during organism decay was most
strongly correlated to MLSS, correlation coefficients were all in excess of
0.96. SONeq a"d SCODeq (0.22 ± 0.10 mg/1 and 1.0 ± 1.1 mg/1, respectively)
were not significantly different (95 percent confidence level) from dSON^ and
dSCODt values for MLSS = 0 (Table 68), providing additional support for the
equilibrium hypothesis. SON and SCOD concentrations were equal to or in ex-
cess of SONeq abd SCODeq under all aeration conditions, regardless of MLSS
concentration.
TABLE 68. LINEAR REGRESSION EQUATIONS SHOWING THE EFFECT OF MLSS ON SON AND
SCOD RELEASE DURING ORGANISM DECAY OF PALO ALTO ACTIVATED SLUDGE
dSON = 0
dSON9, = 0
dSON/0 = 0
48
dSCOD12 - 1
dSCOD0/ = 1
dSCOD. 0 = 3
4o
.26 +
.50 +
.34 +
.6 + 0
.0 + 0
.4 + 0
0.00015
0.00020
0.00030
.00169
.00764
MLSS
MLSS
MLSS
MLSS
MLSS
.00934 MLSS
r = correlation coefficient, e =
MLSS values
tried were 1340, 140,
(r =
(r =
(r =
(r =
(r =
(r =
standard
9, and
0
0
0
0
1
1
.92, e =
.45, e =
.95, e =
.96, e =
.00, e =
.00, e. =
error of
1.
4 mg/1.
0.
0.
0.
0.
0.
0.
04)
13)
05)
4)
4)
3)
estimate.
153
-------
Summary—
Experiments were conducted to study the release of SON and SCOD during
organism decay. Most of the cellular organic nitrogen and COD lost during de-
cay is converted to inorganic end products; only a small percentage accumu-
lates in the surrounding fluid as SON and SCOD. The SON and SCOD produced
during organism decay is strongly correlated with MLSS, increased SON and SCOD
concentrations are associated with higher MLSS concentration. Extrapolation
of linear regression equations for released SON and SCOD vs MLSS to MLSS = 0
resulted in residual SON and SCOD concentrations equivalent to values found
for SONeq and SCODeq, lending further support to the equilibrium hypothesis.
SON Release with Synthetic Feed
Obj ectives—
The objectives of these studies were to better define SONp and SONg (the
SON produced specifically as a result of substrate oxidation) by determining
the effect of initial SCOD, MLSS, and NH3-N, and substrate type on production
of SON. Substrates were chosen which contained little SON, and laboratory AS
Cultures were used since they were grown on substrates with little SON.
Procedures—
Batch studies using synthetic feed and AS Cultures 1 and 2 were conduc-
ted. For some studies quantities of culture were required that were greater
than the one liter available from routine wasting. Increased culture harvest
was accomplished by not wasting on days prior to experiments, and adjusting
feed volumes, waste supernatant volumes, and mass of COD fed to maintain a
constant growth rate. In this manner the required quantity of sludge was
produced without changing seed culture growth characteristics. Table 69
lists variables studied, and filtration and SON techniques used.
Batch Studies 6 and 7 were conducted with AS Culture 1 from Days 61 and
98, respectively, after start-up, and the other studies were conducted with
AS Culture 2, 140 and 180 days after start-up. Culture growth was stable
during these periods. Dissolved oxygen was greater than 1.5 mg/1, and pH was
between 7.0 and 7.5 during all experiments. During Batch Studies 6 and 7,
TABLE 69. VARIABLES STUDIED AND FILTRATION AND SON TECHNIQUES USED
DURING SYNTHETIC FEED STUDIES
Batch
Study No.
5
6
7
8
9
Variable
SCOD
SCOD
Substrate
Type
MLSS
NH3-N
Filtration Technique*
Glass Fiber - Pall Filtration
Centrifugation- Syringe Millipore
Centrifugation- Syringe Millipore
Glass Fiber - Pall Filtration
Glass Fiber - Pall Filtration
As described in Chapter 4.
'
SON Analytical
Technique
Kjeldahl
Technicon
Technicon
Kjeldahl
Kjeldahl
"" ' *• ""• - " ~* • • • . ^
154
-------
NH3~N was maintained above 5 mg/1 so that nitrogen would not be growth-
limiting, and below 20 mg/1 to minimize analytical problems in the Technicon
procedure.
Values of dSONt (Eq. 11) vs aeration time were calculated since they
should give the most reasonable estimate of the SON produced in excess of that
present in the feed solution and seed sludge at the beginning of the experi-
ment. Use of dSONt also helps correct for possible NH3-N carry-over during
SON analysis by the Technicon method.
Statistical Analysis—
Values for SON obtained during these studies were low, necessitating sta-
tistical analysis for data evaluation. Answers to two questions were sought:
(1) Was SON produced during the treatment, and (2) Did the variable studied
affect SON production?
The first question was answered using the following form of Eq. 11:
dSONt(i) = SONtm(i) - SONocd) (16)
where SONtm(i) is the SON concentration measured at time t under test condi-
tion i, and SONoc(i) is the calculated initial SON for test condition i.
Values of dSONt(i) for all aeration times, including time 0, were combined
using Eq. 17 to yield a mean value and standard deviation (n samples) for each
test condition in a given study (i.e., one for SCOD = 1000, one for SCOD =
500, etc.):
dSON (i) = I I dSON (i)|/n (17)
lt=0 J
A t-test was used to determine if the mean was significantly greater than
zero at a 95 percent level of confidence.
For the second question, the effect of aeration time was first ascer-
tained by a linear regression analysis of the following:
dSONt(i) = a + bt (18)
If aeration time was not significantly correlated with dSONt (95 percent level
of confidence as determined with the method described by Crow et al. [220]),
question 2 posed above was answered using a linear correlation analysis of
Eq. 19:
dSONt(i) = a + bC (19)
in which dSONt(i) is defined by Eq. 17, and C is the concentration (mg/1) of
the variable being tested. The strength of this correlation was measured to
a correlation coefficient, a 95 percent or higher level of confidence was
considered necessary for the correlation to be significant [220]. A second
method employed to answer question 2 involved comparing experimental test
155
-------
conditions using the following equations:
dSON (i-j) = dSONt(i) - dSONt(J) (20)
dSON (i-j) = \l dSON (i-j)]/n (21)
t Lt=0 * J
where dSONt(i-j) is the difference in dSONt values (Eq. 11) for test condition
i and control condition j, dSONt(i-j) is the mean difference averaged over all
aeration times, and n is the number of samples analyzed. A t-test indicated
whether dSONt(i-j) differed significantly from zero at a 95 percent level of
confidence. This method was used to evaluate data from Batch Studies 6 and 7
in which SON varied significantly with aeration time, but not in a linear
fashion.
Effect of Initial SCOD—
In Batch Studies 5 and 6 the effect of SCODOC (calculated initial SCOD)
on SON production was evaluated using AS Cultures 2 and 1, respectively. Feed
solutions contained equal concentrations of glucose and acetate COD. A tap
water control was used during Batch Study 6 but not during Batch Study 5.
Initial SCOD:N ratios were 25:1 for Batch Study 5. Ammonia-N was added in
steps (described earlier) to the 954-and 477-mg/l systems of Batch Study 6.
The ratio of added N:P was 5:1 for all systems. Results are presented in
Figures 25, 26, and 27.
SON was produced during oxidation of the synthetic substrates at observed
levels of significance of 98 percent for Batch Study 5 and 99 percent for
Batch Study 6, but dSONt was not significantly correlated with aeration time
for either batch study. There was no significant correlation of dSONt with
SCODoc for Batch Study 5; in other words, influent SCOD concentration could
not be shown to affect the concentration of SON produced.
The concentration of SON released after 23 hours aeration with glucose-
acetate was not statistically different from that released after 23 hours of
aeration with no feed (organism decay for AS Culture 2). This observation
was made by comparing results from Batch Study 5 with the following linear re-
gression equation developed from decay studies with the same cultures and
over the same time period (see preceding section):
dSON23 = 0.02 + 0.00008 MLSS
An approximate standard error of estimate for the above equation is ± 0.02mg/L
as calculated by the method of Spiegel [218]. For an MLSS of 580 mg/1, the
average initial MLSS concentration in Batch Study 5, a value of 0.07 ± 0.02
mg/1 is predicted for the SON released during organism decay. Values of*
dSON23 for the systems evaluated in Batch Study were 0.13, 0.08, 0.06, and
0.07 mg/1, which are not significantly different from the predicted value.
Thus, the concentration of SON produced after 23 hours aeration in the pres-
ence of an exogenous carbon source and that produced after 23 hours of orga-
nism decay are apparently equal. The nature of the SON may be different
but concentrations are similar, and very low. '
156
-------
E
#*
Q
V)
1000
800
600
400
200
0
O
CO
T3
T - 1
SCOD
o
n
A
o
OC
929
468
227
98
OCO 5 10 15 20
AERATION TIME, hr
Figure 25. Batch Study 5: Effect of SCODOc on SON production using AS Cul-
ture 2, MLSSom ~ 580 mg/1 (Day 171).
157
-------
Ln
00
o CONTROL (3.1)
0
OC 0
20 30 40 50
AERATION TIME, hr
60 70
80
Figure 26. Batch Study 6: Effect of SCODOC on SON production using AS Culture 1 sludge,
MLSSom -216 mg/1 (Day 61).
-------
1,000
800
600
g 400
C/)
200
0
I
SCOD
=3=
oc
O 954
a 477
A 98
OC 0 10 20 30 40 50 60 70 80
AERATION TIME, hr
Figure 27. SCOD removal vs aeration time in Batch Study 6, MLSSom - 216 mg/1 (Day 61)
-------
However, in Batch Study 6 using AS Culture 1, the presence of a utiliz-
able substrate resulted in a significantly increased level of produced SON, as
determined by a comparison of the three systems fed glucose-acetate with the
tap control system using Eq. 21. This increase, an average of 0.10 + 0.13 mg/1
for the 27 samples analyzed, was significant at a 99 percent level of confi-
dence.
Differences in SON production behavior have already been linked to dif-
ferences in cultural characteristics, and these differences seem further exem-
plified by the results from Batch Studies 5 and 6. However, SON concentra-
tions were so low during Batch Study 5 with AS Culture 2 that limitations in
the sensitivity of the SON analysis may have prevented observance of substrate
effects on SON and SCOD production. A similar limitation was not present with
Batch Study 6.
Comparison of the SON production by the three fed systems for Batch Study
6 using the 98-mg/l SCODOc system as the control condition (Eq. 21) indicated
that the concentration of SON produced was independent of glucose-acetate
SCODOC. Evidently, the presence of a utilizable substrate, but not its abso-
lute concentration, was important in SON production.
Although there was no significant linear correlation between dSONt and
aeration time, inspection of Figures 25 and 26 suggests there were significant
differences in SON concentrations over the aeration period. The effect, a
noticeable dSONt peak within the first two hours of aeration, was most pro-
nounced in Batch Study 6. This effect is addressed in a later section.
SCOD removal was linear in the two studies, as expected, and minimum ef-
fluent SCOD was a function of influent SCOD, as indicated by the following
correlations for Batch Studies 5 and 6, respectively:
0 + 0.0079 SCOD^ , (r = 0.99, e = 0.4 mg/1)
dSCODmin(6) = 1.1 + 0.0142 SCOD c> (r = 0.99, e = 0.6 mg/1)
where dSCODmin is the difference between the minimum SCOD measured during the
experiment (SCODmin) and the SCOD contributed by the tap water and AS seed.
SCODOC is the initially calculated SCOD. Similar linear dependency has been
reported by others [50,62,214].
Extrapolation of the above equations to SCODOC = 0 should yield values
approaching SCODeq for the two cultures. SCODeq for Batch Study 5 was 1.3 ±
0.4 mg/1, based on samples analyzed during the same time period, and SCODeq
for Batch Study 6 was 3.0 mg/1, the value calculated from the tap water con-
trol (Figure 27). These SCODeq values are not significantly different from
extrapolated values. While limitations in analytical technique at such low
SCOD levels may have prevented significant differences from being observed,
when viewed in combination with previous data regarding equilibrium SON and
SCOD concentrations, data from Batch Studies 5 and 6 lend further support to
the equilibrium hypothesis.
160
-------
Effect of Substrate Type—
The objective of this experiment (Batch Study 7) was to determine if sub-
strate type affects SON production. Solutions used were tap water and tap
water with approximately 860 mg/1 COD of glucose, acetate, and a glucose-
acetate mixture. NH3~N was added to maintain a range of 5-20 mg/1 N, and ad-
ded N:P ratio was 5:1. AS Culture 1 was used, and results are depicted on
Figures 28 and 29.
Significant quantities of SON were produced by all systems, but there was
no linear correlation between produced SON and aeration time. Also, no sig-
nificant difference could be shown between the concentration of SON produced
by the three substrate-fed systems and the control system.
Noticeable differences in dSONt over the aeration period were again ob-
served, the effect being most dramatic within the first hour of aeration when
large dSONt peaks were observed. This will be discussed later. SCOD removal
was again linear.
Effect of MLSS—
In Batch Study 8 the effect of initial MLSS concentration (MLSSOm) was
evaluated on three different occasions with AS Culture 2. MLSS0m values
ranged from 150 to 2000 mg/1, SCODOc was kept relatively constant near 270
mg/1, and initial SCODrN and N:P ratios were 25:1 and 5:1, respectively. Re-
sults are summarized in Figure 30.
SON production was statistically significant, although sometimes very
low, was not correlated to aeration time, but was correlated linearly to
MLSSOC (r = 0.89 [220]). Concentrations of produced SON appeared similar to
those resulting from organism decay, and linear regression equations for aera-
tion times of 23 hours indicated the differences were not statistically sig-
nificant :
dSON (organism decay) = 0.02 + 0.00008 MLSS
dSON23 (SCODQc - 270 mg/1) = -0.01 + 0.00009 MLSS
AS Culture 2 data for the above equations were taken over the same time period
(Days 156 to 173). Standard errors of estimate are 0.02 and 0.03 mg/1, re-
spectively [218]. The observation here agrees with Batch Study 5: it is dif-
ficult to distinguish between SON produced in the presence of utilizable
carbon and that produced during organism decay.
Effect of NH3-N—
Batch Study 9 was conducted to evaluate whether initial NH3-N concentra-
tion (NH3~NOC) affected SON production. Feed solution SCODOC was maintained
constant at around 500 mg/1, and NH3-N was initially added in concentrations
varying from 0 to 60 mg/1. Results are shown in Figure 31. SON production
was statistically significant for all three systems, and was independent of
aeration time and initial NH3-N concentration.
161
-------
1.0
.8
o»
e -
o
co •
•o
.2
o GLUCOSE-ACETATE
a GLUCOSE
A ACETATE
o CONTROL
OC 0 10 20 30 40 50
AERATION TIME, hr
Figure 28. Batch Study 7: Effect of substrate type on SON production
using AS Culture 1, MLSS = 360 mg/1 (Day 98).
162
-------
1000
800
600
8 400
c/j
200
0
I I I
o GLUCOSE-ACETATE
D GLUCOSE
A ACETATE
OCO
10 20 30
TIME, hr
40
Figure 29. Batch Study 7: SCOD removal vs aeration time,
MLSS = 360 mg/1 (Day 98).
om
163
-------
5
1"
400
300
100
0
.20
.15 -
D>
o
CO
.05 -
0
OCO 5 10 15 20
AERATION TIME, hr
Figure 30. Batch Study 8: Effect of MLSS on SON production.
25
164
-------
500
400
300
200
100
0
E
•
Q
8
NH3-Noc(mg/^)
o 59.5
a 20.2
A 0.9
OCO 5 10 15 20
AERATION TIME, hr
Figure 31. Batch Study 9: Effect of NH3-N on SON production,
MLSS - 830 mg/1 (Day 176).
Offl
165
-------
Summary—
The objective of this series of experiments was to determine which vari-
ables and factors affect SONp and SONg (SON produced as a result of the pres-
ence of a utilizable carbon source). Although one study showed statistically
that more SON was produced by systems fed a utilizable substrate than by con-
trol systems (organism decay), the majority of the data indicated that SON
levels produced by fed and control systems were not significantly different.
This observation implies that SONg is zero; or at least very small. SON was
produced during all experiments, but only initial MLSS concentration could be
shown to significantly affect the amount of SON produced. Initial SCOD,
NH3-N, and substrate type did not affect produced SON levels, but SON concen-
trations were so low that analytical limitations may have hindered the detec-
tion of significant differences, especially for AS Culture 2.
Significance of SONq and dSONt Peaks
Two general types of SON peaks were noted during SON production experi-
ments: initially released peaks (SONq, defined by Eq. 8) and peaks occurring
during aeration (dSONt). The SON initially released upon dilution of culture
with tap water (see Initial SON Release) appeared to establish a concentra-
tion which was in equilibrium with an organic nitrogen pool within the cells.
This equilibrium concentration, SONeq, was independent of MLSS concentration
(10-2400 mg/1) and temperature (10-30°C), but dependent on cultural charac-
teristics.
SONq Peaks—
Table 70 is a summary of initial SON release for the various studies con-
ducted and indicates that the presence of a utilizable substrate affected the
quantity of initially released SON (SONq). Mean values of SONq (SONq) for
systems fed utilizable carbon were 0.48 ± 0.17 mg/1 (n = 7) and 0.07 ± 0.04
mg/1 (n = 10) for AS Cultures 1 and 2, respectively. Comparison of these
means with respective values for SONSq (Table 70) showed a significant dif-
ference at a 99 percent level of confidence for AS Culture 1, but not for Cul-
ture 2. The presence of an exogenous substrate caused release of SON concen-
trations in excess of SONeq for Culture 1, but not for Culture 2. However,
SON concentrations were so low for Culture 2 that significant differences may
not have been detectable. Both cultures released SON initially; Culture 1
released significantly more (99 percent confidence limit) SON than Culture 2.
Others [207,217] have also shown that differences in cultural characteristics
affect process performance.
The excess initial SON released by AS Culture 1 was due to the presence
of a utilizable substrate and was therefore part of SONg. This excess SON
may be a function of growth rate or growth phase, may result from substrate-
accelerated death, and/or may involve energy-dependent facilitated transport.
Which of these factors, if any, controls initially released SON cannot be
determined since excretion is a complex phenomenon, and not well understood
even in pure cultures. Extrapolation to heterogeneous activated-sludge cul-
tures is useful but difficult. The major observation is that presence of a
utilizable carbon source, in some cases at least, stimulates initial release
of SON in excess of SONeq.
166
-------
TABLE 70. SUMMARY OF SONq DATA FOR ALL EXPERIMENTS3
Day or
Dateb
8/75-9/75
9-23-75
220-302
221
98
it
it
ii
61
ii
it
ii
124-173
171
ii
it
ii
168
ii
165
it
M
162
M
156
153
Culture :
MLSS (mg/1)
Palo Alto
(PA)
PA: 178
AS Culture
1 (Cl)
Cl: 2040
Cl: 362
Cl: 370
Cl: 352
Cl: 350
Cl: 218
Cl: 212
Cl: 222
Cl: 212
AS Culture
2 (C2)
C2: 615
C2: 575
C2: 565
C2: 565
C2: 2040
C2: 60
C2: 1080
C2: 585
C2: 150
C2: 1450
C2: 760
C2: 2000
C2: 590
Feed Condition
clnitial release studies
(SON = 0.22 ±0.10 mg/1,
n = 8)
200 mg/1 glucose COD
clnitial release studies
(SONeq = 0.12 ±0.04 mg/1,
n = 23)
1000 mg/1 glucose-acetate
Control
869 mg/1 glucose-acetate COD
854 mg/1 glucose-acetate COD
869 mg/1 glucose-acetate COD
Control
954 mg/1 glucose-acetate COD
477 mg/1 glucose-acetate COD
98 mg/1 glucose-acetate COD
Control
°Initial release studies
(SONeq = 0.06+ 0.01 mg/1,
n = 6)
929 mg/1 glucose-acetate COD
468 mg/1 glucose-acetate COD
227 mg/1 glucose-acetate COD
78 mg/1 glucose-acetate COD
670 mg/1 glucose-acetate COD
670 mg/1 glucose-acetate COD
270 mg/1 glucose-acetate COD
270 mg/1 glucose-acetate COD
270 mg/1 glucose-acetate COD
270 mg/1 glucose-acetate COD
270 mg/1 glucose-acetate COD
270 mg/1 glucose-acetate COD
300 mg/1 glucose-acetate COD
SONq
(mg/D
0.20±0.09
0.38
0.11±0.04
0.42
0.14
0.41
0.84
0.52
0.14
0.30
0.48
0.41
0.09
0.05±0.01
0.10
0.07
0.07
0.07
0.04
0.05
0.06
0.08
0.04
0.09
0.03
0.11
0.16
SON
Technique
Technicon
&Kjeldahl
Kjeldahl
Kjeldahl
Kjeldahl
n
Technicon
ii
n
n
Technicon
n
n
M
Kjeldahl
Kjeldahl
n
M
ii
Kjeldahl
M
Kjeldahl
n
n
n
n
M
Kjeldahl
3SON = SON - SON (Eq. 8).
, q om oc M
Day of Operation reported for AS Cultures; date reported for Palo Alto
Culture.
°Fed tap water; SON reported (see Initial SON Release); n = number of
samples.
167
-------
dSONt Peaks--
Review of Figures 14, 25, 26, 28, and 30 indicates a peak in produced SON
occurred within zero to two hours after mixing of feed solutions and activated
sludge. The peaks generally occurred prior to the point of complete substrate
removal. To test for statistical significance of observed SON peaks, the
following equations were used:
dSON . = SON . - SON
pk peak oc
dSON . = SON , - SON .
pkm peak mxn
(22)
(23)
where SONpeak is the maximum SON concentration observed during the experiment,
SONoc is the initially calculated SON, and SONmin is the minimum SON measured
after the peak SON had occurred. Differences required for a 95 percent confi-
dence level are listed in Table 71 along with the summary of SON peak data.
TABLE 71. SUMMARY OF DATA DESCRIBING SON PEAKS DURING AERATION
Culture
AS Culture 1
AS Culture 1
n
"
n
AS Culture 1
"
"
"
AS Culture 2
"
n
n
AS Culture 2
it
it
Palo Alto AS
SCOD
oc
(mg/1)
1000
954
477
98
(Tap) 3
869
854
869
(Tap) 2
929
468
227
976
270
270
270
186
(Tap) 4
MLSS
om
(mg/1)
2040
218
212
222
212
362
370
352
350
615
575
565
565
2000
1450
150
178
178
dSON .
pk
(mg/1)
0.42
0.62
0.67
0.61
0.69
0.54
0.84
0.52
0.47
0.15
0.07
0.05
0.11
0.15
0.09
0.10
0.99
0.41
dSON ,
pkm
(mg/1)
0.22
0.62
0.58
0.54
0.67
0.47
0.84
0.46
0.33
0.12
0.03
0.01
0.08
0.07
0.05
0.08
0.62
0.19
Time from
Hour 0 to
Peak
(hrs.)
0
1
1
1
1
2
0
0
1
1
0
1
0
3
0
1
4
4
Time from
Peak to
Minimum
(hrs.)
0.5
35
7
3
3
10
2
8
7
2
1
2
6
3
1
2
8
20
SON
Technique
Kj eldahl
Technicon
"
"
11
Technicon
n
"
n
Kj eldahl
n
n
"
Kj eldahl
n
n
Technicon
n
dSON required for 95 percent significance level = 0.10 mg/1 (Kieldahl)
and 0.28 mg/1 (Technicon).
dSONpkm rectuired for 95 Percent significance level =0.08 mg/1 (Kieldahl)
and 0.23 mg/1 (Technicon).
168
-------
Of the 18 samples listed in Table 71, 15 values for dSONp^ and 14 values
for dSONpkm were high enough to be statistically significant, implying that
SON was in fact produced in excess of SONOC during initial stages of aeration
and then removed to varying degrees as aeration continued. The magnitude of
the peak height for the different samples varied, as did peak width. For
some systems, the initially measured SON was the maximum observed. The mean
dSONpk for AS Culture 1 (0.60 ± 0.13 mg/1) was greater than SONeq (0.12 ± 0.04
mg/1) at a 99 percent confidence level. A similar difference could not be
demonstrated statistically for AS Culture 2 which had measured SON values
near the analytical limit.
The production of SON peaks during substrate oxidation might be the re-
sult of a particular phase of growth or rate of growth. However, tap water
control systems exhibited similar peaking behavior, and no statistically sig-
nificant difference between levels of produced SON for these systems and those
fed exogenous carbon was found. Thus, presence of external carbon sources
and resulting differences in growth rate or phase of growth cannot alone ex-
plain the produced SON; other factors must also be involved.
The only significant difference found between the behavior of systems
fed a utilizable carbon source and those fed tap water was that initially re-
leased SON from carbon-fed systems with some cultures was greater. This in-
crease is directly attributable to SONg. The major observation is that SON
is released rapidly by cells immediately after dilution with tap water or a
new substrate, a peak concentration of SON is reached within a few hours,
after which the concentration decreases. This behavior is consistent and is
affected by factors other than growth rate.
DISCUSSION OF RESULTS
Sources of Effluent SON
The SON contained in activated-sludge effluents (SONe) has two general
sources: (1) influent SON (S0%) that is not removed during treatment, and
(2) SON produced during treatment (SONp). SON-[ contains both biodegradable
(SONb) and refractory material (SONr), as does produced SON (SONpb and SONpr,
respectively). SONp includes an equilibrium concentration (SONeq), SON pro-
duced by organism decay (SON
-------
26, 28, and 30), and supported by fundamental theory (see "SON Production and
Excretion by Bacteria"). This SON production function is meant to provide a
framework for discussing changes in produced SON sources during activated-
sludge treatment, and is not meant to be quantitative or absolute. It is a
composite of observed experimental behavior. Reference to the model, along
with presentation of relevant supporting data and fundamental theory, will be
made when discussing SONp, SONeq, SONd, and SONg.
SONp—
Although growth rate (y) could not be mathematically correlated to SON
production during start-up of semi-continuous fed laboratory Activated Sludge
(AS) Cultures (Figures 16, 17, and 18), y probably exerted an effect on SONp
levels. The peak SON concentration (up to 5 mg/1) was observed when MLSS was
less than 1500 mg/1. At these lower MLSS levels, longer aeration periods were
required to remove added glucose-acetate substrate during the 23-hour daily
feed cycle. Growth rate thus remained high for a longer period and resulted
in continued production and excretion of SON metabolites. As MLSS levels in-
creased to near 1500 mg/1, starvation conditions were reached earlier in the
feed cycle and degradable SONp excreted during substrate unlimited conditions
was utilized as expected [101,148,199]. Experiments utilizing chemostats with
varying 6C (1/y) over a wide range (0.5 to 20 days) would be needed to reli-
ably determine the effect of y on SONp. However, the semi-continuous data do
indicate that operation under unlimited substrate conditions may result in
high concentrations of produced SON.
Significant differences in SONp concentration were associated with
changes in AS Culture characteristics. These differences occurred during
steady-state operation (y = 0.17 day~l). This is an important observation
because it implies that even if growth rate can be controlled to minimize
SONp for one set of conditions, significant changes in SONp may still occur
as a result of natural culture population shifts. The optimal growth rate
may also shift.
Fluctuation in SONp levels during activated-sludge treatment may primar-
ily be due to differences in production of biodegradable SON. Estimates of
refractory SON were similar for AS Culture 2 during start-up, AS Culture 2 at
steady-state, and AS Culture 1 at steady-state (0.40, 0.26, and 0.14 mg/1, re-
spectively) . These differences were statistically significant but small when
compared to differences in SONpb. Values of SONpb for the three samples were
1.63, 0.62, and 0 mg/1, respectively. AS Culture start-up corresponds to
operation at aeration times less than tmin (Figure 32), while steady-state
operation corresponds to aeration times greater than t^n. Operation to the
left of tmin will likely result in production of higher concentrations of de-
gradable SON than aeration times greater than tmin. Variability in SONpb lev-
els for systems operating to the right of tmin (steady-state samples) again
illustrates perturbations in SON production under identical operating condi-
tions.
Peak SONp concentrations (Figure 32) observed during batch studies were
real, but variable for different activated-sludge cultures (0.10 to 0.60 mg/1).
Variability in peak SONp was probably due to differences in cultural charac- *
teristics. Peak heights were independent of SCOD, MLSS, and NH3-N
170
-------
O
CO
0
OCO
'min
AERATION TIME
Figure 32. Conceptual model of changes in SONp, SONg, SON
-------
concentrations, and substrate type. The SON produced in peak concentration
was degradable since it was removed to a significant degree as aeration con-
tinued. Peaking behavior is explained by bacterial production and utilization.
Bacteria produce and excrete SON in response to concentration gradients, star-
vation conditions, addition of exogenous substrate, and changes in phase and
rate of growth. Produced SON is utilized by other organisms in the hetero-
geneous culture as aeration continues. The utilization response is fairly
rapid (within 2 to 4 hours) and a general equilibrium between production and
utilization seems to be reached near tmin (~ 4-8 hours for the systems stu-
died). As aeration time exceeds tmin, substrate becomes limiting and starva-
tion conditions prevail. A gradual release of SON under such conditions is
well documented [99,100,109,153,157,177,178].
Based on discussion presented above, hydraulic detention time (9) is an
important process variable affecting the levels of SON produced by activated-
sludge treatment. Variations in influent flow (Q), which change 9, may be
responsible for fluctuations in produced SON levels. As Q increases, 9 de-
creases and is shifted to the left in Figure 32. If the decrease is large
enough, the peak region of the SON production function may be encountered,
resulting in a significant increase in produced SON (largely degradable).
SONeq—
Activated-sludge organisms excrete SON compounds to establish an equilib-
rium concentration (SONeq) with these particular compounds, an observation
consistent with information given by others [150,151,152]. The discovery of
SONeq was a result of the experimental procedure used. Concentrated activated-
sludge culture was diluted with tap water, and SONeq was defined as the SON
initially released. The magnitude of SONeq was independent of MLSS (10-2400
mg/1) and temperature (10-30°C). Cultural characteristics were once again
linked to significant differences in SON concentrations since the value of
SONeq was dependent on cultural characteristics. Average values of SONeq for
the three sludges studied are:
AS Culture 1 0.12 ± 0.04 mg/1
AS Culture 2 0.06 ± 0.01 mg/1
Palo Alto AS 0.22 ± 0.10 mg/1
SONeq is assumed to be present at all times during activated-sludge
treatment because: (1) SONeq is refractory (Section 9), and (2) SONp levels
measured during organism decay and synthetic feed batch experiments did not
decrease below the SONeq level for any aeration time. There is some question
as to the appropriateness of including SONeq as part of SONp, since during
steady-state operation of continuously fed activated-sludge systems, SON may
not be "produced" as such. However, it will always be present. As conditions
in the aeration basin change, for example, when population shifts occur, SO
will be adjusted, increasing or decreasing depending on the situation. The
major point is that SONeq is part of SONe. For purposes of this discussion,
SONeq is considered to be part of the fraction of SONe represented by SON
172
-------
SONeq may be considered refractory with respect to removal by activated
sludge, but may be degradable under some circumstances. Since its magnitude
is dependent on cultural characteristics, SON compounds contained in SONeq
for one organism may not be included in SONeq for other organisms. A portion
of SONeq from one organism may be removed when population shifts occur and
other organisms become predominant. If for some reason organisms were exposed
to higher concentrations of the nitrogenous components comprising SONeq, the
additional SON should be removed until SONeq is reached. A limited experiment
showed that exposure of AS Culture to concentrated SONeq (~ 2 SONeq) resulted
in some SON removal. However, additional experiments would be needed to reach
a firm conclusion. For purposes of discussing SONe, SONeq will be considered
refractory, since Biodegradation Study 3 (Section 9) showed it to be 100 per-
cent refractory after 30 days degradation.
SONd-
SON produced due to organism decay (SONj) was defined to be the SON re-
leased in excess of SONeq during batch aeration of activated-sludge cultures
with tap water. SON^ was calculated from Eq. 25:
SONd = dSONt - SON (25)
in which dSONt is the SON released at a given aeration period (t). Relation-
ships for dSON23 were determined directly for AS Cultures 1 and 2, and can be
estimated for Palo Alto culture using the equations listed in Table 68. The
method for deriving the Palo Alto culture relationship is described later.
The resulting relationships are:
AS Culture 1 dSON23 = 0.05 + 0.00008 MLSS
AS Culture 2 dSON23 = 0.02 + 0.00008 MLSS
PA Culture dSON23 = 0.22 + 0.00020 MLSS
It is apparent that the contribution of SON
-------
SONd for aeration times greater than tmin was strongly correlated with
MLSS concentration; increased SON concentrations were associated with higher
MLSS concentrations. An important implication of this observation is that
operation of activated-sludge systems at high MLSS values and a given waste
detention time should result in production of larger quantities of refractory
SON than will lower MLSS systems. The dependence of SONd on MLSS can be ex-
plained by differences in SON production and utilization rates as discussed
above. Assuming the net SONd production rate (production - utilization) to
be constant for a culture of organisms, larger concentrations of organisms
will result in more net production of SONd.
Whether the SONd peaks described by Figure 32 would be observed with
continuously fed activated-sludge systems is open to speculation. During
normal activated-sludge operation where solids recycle is practiced, orga-
nisms are starved while in the secondary sedimentation basin and the recycle
line. When these starved organisms are suddenly diluted upon re-entry to the
aeration basin, a portion of the organisms will remain starved, and excrete
SON as described above. Other viable organisms may rapidly remove the ex-
creted SON such that a peak may not be observed. Appropriate chemostat ex-
periments would have to be conducted to adequately determine if SONd peaks
would appear at lower values of 8. Reliable estimates for the contribution
of SONd to SONp are difficult because production and excretion of SON by
heterogeneous activated sludge is complex. The approach used during this
research, although not without limitation, did point out that the contribu-
tion of SONd may be quite significant at short aeration times.
SONg—
SONg, estimated from Eq. 26, was not significantly greater than zero for
any batch experiment except Batch Study 6:
SONg = dSON (substrate) - SONd - SONg (26)
dSON (substrate represents the SON produced by an activated-sludge system fed
a utilizable substrate. The only consistently significant difference between
SONg and SONd + SONg was measured at time zero for AS Culture 1. SONp values
for AS Culture 2 were so near the analytical limit that significant differen-
ces could not be detected. Two interpretations are possible from these ob-
servations: (1) SONg is in fact zero for aeration times greater than time
zero, and (2) estimation of SONd is erroneously high and SONg is not really
zero.
If SONg is zero, it means growth rate had no effect on SON production
and most of the SON produced during the batch experiment was the result of
organism decay. Data from the literature suggest that this should not be
the case; substrate unlimiting conditions should result in excretion of
larger concentrations of SON than starvation conditions [101,113,148]. The
fact that initially released SON for substrate-fed systems was significantly
higher than' for tap control systems (for AS Culture 1) implies that substrate
has an effect. Perhaps organisms utilized excreted SON at a faster rate in
the substrate-fed systems thus depressing the SONp peak height to a level
experimentally indistinguishable from the SONd peak.
174
-------
It could not be determined from experimental data whether the method
used to estimated SONd resulted in erroneously high values. A satisfactory
alternative method for estimating SONd could not be found. The major observa-
tion was that SON was produced rapidly by organisms when added to tap water or
Ifap water plus substrate, and then removed to a significant degree. In gene-
ral, no significant difference between SON levels produced by these two sys-
tems was discernible, that is, SONg may be zero as measured experimentally.
Contribution of SONeq, SONd, and SONg to SONp—
The contribution of the individual sources
Methods used in making this estimation are presented below.
The contribution of the individual sources of SON to SONp was estimated.
1. SONp—Steady-state, AS Culture effluent SON concentrations (Table 59)
were used to estimate SONp. Two different estimates with different assump-
tions were made. In the first estimate, feed solution SON (SONfj 0.10 ± 0.04
mg/1) was assumed to be completely refractory and SONp was calculated as fol-
lows:
SON = SONe - SONf
where SONe represents the steady-state effluent SON at the end of the 23-hour
daily feed cycle. For the second estimate, SONf was assumed to be completely
degradable, making SONp equal to SONe. The true value for SONp is probably
somewhere between these two estimates. Table 72 lists the values of SONp
which resulted.
2. SONeq—Values of SONeq used were discussed earlier and are listed in
Table 72.
3. SONd—Since 23-hour measurements for SONe were used, the relation-
ships developed previously for 23-hour estimates of SONd are appropriate.
These relationships are:
AS Culture 1 SONd = 0.05 + 0.00008 MLSS - SONeq, e = 0.02
AS Culture 2 SONd = 0.02 + 0.00008 MLSS - SONeq, e = 0.02
where e is the standard error of estimate. The MLSS values listed in Table
59 were used to estimate SONd- These estimates are contained in Table 72.
4. SONg—Estimates for SONg (Table 72) were calculated using the rear-
ranged form of Eq. 24 listed below:
SONg = SONp - SONeq - SONd
Use of this method for estimating SONg was both necessary and reasonable,
since experimentally determined values of SONg were not significantly differ-
ent from zero.
Although the proposed method is not eithout limitations, it should yield
175
-------
TABLE 72. CONTRIBUTION OF SON . SON,, AND SON TO SON FOR AS CULTURES
eq Q g p
Sample
(see Table 5.3)
Assume SONf is refractory:
AS Culture 1 - Data Set 1
AS Culture 1 - Data Set 2
AS Culture 2
Assume SON^ Is degradable:
AS Culture 1 - Data Set 1
AS Culture 1 - Data Set 2
AS Culture 2
SON
P
(mg/l)a
0.23 ±0.06
0.53 ±0.19
0.16 ±0.06
0.33 ±0.06
0.63 ±0.19
0.26 ±0.06
SON
eq
mg/la
0.12 ±0.04
0.12 ±0.04
0.06 ±0.01
0.12 ±0.04
0.12 ±0.04
0.06 ±0.01
Percentb
of SONp
52 ±22
23 ±11
38 ±16
36 ±14
19 ±9
23 ±7
SON,
d
mg/la
0.09 ±0.04
0.11 ±0.04
0.13 ±0.02
0.09 ±0.04
0.11 ±0.04
0.13 ±0.04
Percent*3
of SON
39 ±20
21 ±11
81 ±33
27 ±13
17 ±8
50 ±19
± values are standard deviations.
SON
g.
mg/la
0.02 ±0.07
0.30 ±0.19
-0.03 ±0.07
0.12 ±0.07
0.40 ±0.19
0.07 ±0.07
Percent"
of SONp
9 ±32
57 ±41
-
36 ±22
63 ±31
27 ±28
± values are standard deviations for percent as calculated by method similar to that described in
Appendix B, item 8-e.
-------
reasonable ranges for percent of SON- contained in the SONeq, SONj, and SON
fractions. Results are summarized in Table 72.
Caution must be used in interpreting the results listed in Table 72 be-
cause SON concentrations were, for the most part, low and near the analytical
limit. However, some general comments can be made. SONeq constitutes a sig-
nificant fraction of produced SON, and since it is expected to be present at
all times during treatment, cannot be effectively controlled or minimized.
The magnitude of SON
-------
SON (SONe) of 1.40 ± 0.46 mg/1 for the eight samples tested, results similar
to those found for the full-scale Palo Alto plant (Section 6). SON and SCOD
removal rates increased as MLSS was increased.
SCOD concentration in general decreased more rapidly than SON concentra-
tion. Differences in the nature and source of these materials can in part ex-
plain removal rate differences. Characterization studies showed that (1) SCOD
decreased at a faster rate (relative rates were about 0.26 day"1 for SCOD and
0.13 dayl for SON), (2) apparent molecular weight distributions for SON and
SCOD were different, and (3) ion-exchange behavior of the SON and SCOD were
different. In addition, relative production of SON during batch aeration may
be greater than relative production of SCOD, resulting in depression of the
net SON removal rate relative to that of SCOD. This effect was most dramatic
for the 1.4-and 14-mg/l MLSS systems for Batch Study 3 in which SON actually
increased during initial stages of aeration while SCOD decreased.
Conceptual Model of SON Removal—
It is helpful to construct a conceptual model of net SON removal during
activated-sludge treatment to better understand SON behavior. The model
should include the production function described under SON production, along
with a description of changes in influent degradable (50%) and refractory
(SONr) fractions during treatment. Such a model is shown in Figure 33, and
serves as a basis for discussing the sources of effluent SON. The shape of
the SONe curve is a composite of experimental behavior observed during batch
removal studies with Palo Alto activated sludge.
The model is not without limitations. Experiments were with batch sys-
tems in which primary effluent was mixed with Palo Alto activated sludge.
Measured time zero SON and SCOD concentrations for systems with MLSS levels
near 1000 mg/1 were 10 to 20 percent lower than in solutions without MLSS as
uptake occurred during processing (15 minutes) of the time zero sample. Re-
lease of SONg and release of SON to reach SONeq (see Figure 32) also took
place during this 15-minute period. After time zero, SON behavior was meas-
urable and consistent. Figure 33 and subsequent discussion of SONe are based
on observed behavior for time periods after initial mixing.
As drawn, Figure 33 represents results from a relatively high MLSS system
(> 1000 mg/1). For lower MLSS values, the decrease in SONe would be slower
(net removal rate less), and if MLSS were low enough, SONe may actually in-
crease during initial stages of aeration due to SON production (see 1.4- and
14-mg/l MLSS systems, Batch Study 3). This aspect will be discussed in detail
under SONd 4- SONB.
o
Individual Sources of SONe—
Sources contributing to SONe include SONi (80% + SONr), SONeq, SONd,
and SONg.
1. SONj—SONj was shown to be highly degradable; more than 80 percent of
SONi may be represented by 50%. S0% will comprise the major fraction of
SONe during early stages of aeration, but may be removed almost completely as
aeration continues (see Biodegradation Study 3). Thus, it is not likely that
50% will comprise a significant fraction of SONe if sufficient detention
178
-------
0
AERATION TIME
Figure 33. Conceptual model of changes in sources of effluent SON
as a function of activated-sludge aeration time.
179
-------
time is provided. The rate at which SON^ is removed during treatment in-
creases as MLSS concentrations increase.
Estimates for SONr ranged from 18 to 38 percent of 80%. Reliable ap-
proximations of the "true" SONr are difficult to obtain from low seed biode-
gradation experiments .because the quantity of refractory SON produced during
the time period is unknown. Estimates for the refractory portion of SONe
ranged from 40 percent (Biodegradation Studies 1 and 2) to 100 percent (Bio-
degradation Study 3 and [202]). Predictions for the concentration of SONr
were sometimes significantly greater than corresponding predictions for re-
fractory SONe concentration (Biodegradation Studies 2 and 3). This apparent
dichotomy emphasizes the complexity in separating SON sources. Perhaps some
SONr was removed during activated-sludge treatment, or more likely, additional
SONr was produced during the biodegradation experiment used to predict SONr.
Regardless of these limitations in estimating SONr, it is apparent that SONr
will comprise a significant fraction of SONe. For simplification, SONr is
assumed to be constant during aeration.
2. SONeq — SONeq was described in detail under SON Production. Primary
wastewater already contains at least a portion of the compounds comprising
SONeq. As such, they would be "measured" as part of SONr because SONeg was
found to be 100 percent refractory. However, if present in concentrations
greater than SONeq , the increment above SONeq would be removed and hence
measured as part of SON^. Since it could not be determined what fraction of
SONeq, if any, was contained in SONr or SONb, and since reasonably reliable
estimates for SONeq were easily obtained, SONeq was considered a separate
source of SONe.
3. SONd + SONg — For reasons alluded to under SON Production, reliable
separation of SONd and SONg was difficult. Data from Batch Study 4 with Palo
Alto activated sludge support this observation by showing that the only sig-
nificant differences between SON produced by a tap control system and a
glucose-fed system occurred at aeration times of 2 and 4 hours (corresponding
to peak SON concentrations) , and even then, differences were relatively small.
Therefore, SONd and SONg will be discussed together rather than individually.
+ SONg include a biodegradable fraction, most prevalent at aeration
times less than tmin, and a refractory fraction. The refractory fraction
ranges from 20-100 percent of total SONd + SONg (Biodegradation Studies 4 and
5 with AS culture); SONd ma7 be 100 percent refractory. SONd + SONg exhibited
a linear dependence on MLSS, and SONd was approximately linear with aeration
time for aeration times greater than tmin.
Direct evidence for the contribution of SONd + SONg to SONe comes from
Batch Study 3 data (MLSS = 1.4, 14 mg/1) . SONe increased by up to 0.6 mg/1
while SCOD was being removed (10-20 percent) during the first six hours of
aeration of the primary effluent with Palo Alto activated sludge. Acclima-
tion or lag phenomena do not explain the increase in SON since SCOD was being
removed. The magnitude of the increase (0,4 to 0.6 mg/1) was similar to the
average peak SON produced by AS Culture 1 systems.
180
-------
At some time during the first 4 hours of treatment, SONe increased, or
at least remained constant, for all batch systems studied. If no SON were
produced during treatment, the SON removal curve should be a relatively
smooth, continuously decreasing curve similar to that predicted by classical
Monod kinetics for organic removal [200], but it was not.
The contribution of SON
-------
0.22 + [0.0001 + (4.17)(10~6)(t)] MLSS
(28)
The estimated concentration of SONd + SONg then becomes dSONfc minus SONeq.
The standard error of estimate for this equation, 0.15 mg/1, was high due to
possible analytical error in the 24-hour SON measurement (Table 68). SONr
was estimated as 20 percent of the initially calculated SON for the primary-
fed system. SONb was estimated by subtracting SONeq, SONd + SONg, and SONr
from measured SON-. Results for this estimate are summarized in Table 73.
The release of SON due to organism decay becomes more significant (up
to 21 percent) as aeration time was increased, an expected occurrence. SON
and SCOD produced were expected to be refractory (see Biodegradation Study 3).
A limitation of this estimate is the inability to accurately determine SONr
and SONfe. BY definition, SONb cannot increase since it was the biodegradable
portion of the influent. Limitations in the estimate of SONr have already
been discussed. Perhaps the increase in SONb after tm^n was in fact due to
increases in SONd + SON from substrate oxidation that were not observed in
the tap water control system used to estimate SONd + SONg. The data show
that the contribution of SONd + SON was 13 to 21 percent for the listed
aeration times.
The second estimate was obtained from Batch Study 4 data; MLSS was about
180 mg/1. This experiment represented an attempt to estimate the contribu-
tion of SONd + SONg to Palo Alto SONe by feeding a glucose substrate to Palo
Alto culture and comparing this system with one fed primary effluent. SONd +
SONg was calculated using the glucose-fed system and correcting for the con-
tribution of SONe and SONOC (initially calculated SON). SON ,
SON,., and
TABLE 73. ESTIMATION OF _SONeq, j30Nd + SONg, SONr. AND SONh IN
PALO ALTO SECONDARY EFFLUENT
(MLSS = 1390 mg/1)
Aeration
Time
(hours)
0
6 (tmin)
12
24
48
"A ~~
Measured
SON *(mg/l)
3.16 ± 0.16
1.30 ± 0.16
1.66 ± 0.16
1.81 ± 0.16
1.98 ± 0.16
SONb
(percent
of SON )
e
7 ± 3
17 ± 8
13 ± 8
12 ± 6
11 ± 5
SONd + SONg
(percent ..
of SON )
e
4 ± 4
13 ± 11
13 ± 9
15 ± 8
21 ± 8
SONr
(percent ^
of SON )
e
23 ± 5
57 ± 14
45 ± 10
41 ± 9
37 ± 8
SONb
(percent A
of SON )
e
66 ± 9
13 ± 22
29 ± 17
32 ± 15
31 ± 14
± values are 95 percent confidence limits.
± values are estimated derivations based on standard error of estimate
(Eq. 26) and 95 percent CI of SON .
e
182
-------
TABLE 74. ESTIMATION OF SONgg. SONd + SONg, SONr, AND SONfa IN PALO ALTO SECONDARY EFFLUENT
(MLSS = 180 mg/1)
Aeration
Time
(Hrs.)
0
. 1
2
4
8
12
24
48
72
Measured
SONe Ofc 0.16)1"
(mg/1)
3.93
3.90
3.15
2.89
2.46
1.91
1.85
1.62
1.71
SONeq (± 0.08)f
mg/1
0.22
0
.22
Percent
of SONe*
6 ± 2
6 ± 2
7 ± 3
8 ± 3
9 ± 3
12 ± 4
12 ± 4
14 ± 5
13 ± 5
SONd + SONg (± 0.29)t
mg/1
0.16
0.03
0.40
0.77
0.21
0.15
0.24
0.46
0.67
Percent
of SONe*
4 ± 7
1 + 10
13 ±9
27 ± 10
9 ± 12
8 ± 15
13 ± 16
28 ± 18
39 ± 17
SONr (± 0.16)
mg/1
0.96
0
.96
Percent
of SONe*
24 ± 4
24 ± 4
30 ± 5
33 + 6
37 ± 7
50 ± 9
51 ± 10
59 ± 11
56 ± 11
SONb (± 0.38)1"
mg/1
2.61
2.71
1.59
0.96
1.07
0.60
0.45
-0.02
-0.14
Percent
of SONe*
66 ± 10
69 ± 10
50 ± 12
33 ± 13
44 ± 16
31 ± 20
24 ± 21
(_!)**
**
(-8)
± values are 95 percent confidence limits (mg/1) .
± values are 95 percent confidence limits for percent contribution.
**
See text for explanation.
oo
OJ
-------
SON]., were estimated as in the previous example. Results are summarized in
Table 74, along with 95 percent confidence limits for estimated values. Con-
fidence limits for SONd + SON and SONb were high since subtraction of several
SON measurements were used to obtain these estimates.
Despite the relatively high uncertainties for the estimates listed in
Table 74, the general shape of the conceptual SONe behavior model was approxi-
mated. The contribution of SONd + SONg is maximum near the peak of the pro-
duction function and at long aeration times. SONd + SONg present after long
periods of aeration will likely be refractory. The contribution of SONb,
although somewhat erratic, is highest during initial stages of aeration.
SONr constitutes the major fraction after long periods of aeration.
As previously discussed, inability to accurately determine SONr limits
the estimation of SON source contribution. Higher estimates for SONr would
have resulted in SONb values much less than zero (in this case, for 48- and
72-hour data, SONb values were slightly negative) , while lower estimates
would yield high values for 50% at long aeration times. Changes in the
estimate of SONa + SONg would also change SONb estimates. Separation of SON
production from SON^ removal is clearly a complex undertaking. This particu-
lar attempt, one employing reasonable assumptions, showed that 20 to 50 per-
cent of SONe may be produced during treatment.
For the third estimate, the contribution of the individual SON sources
to Palo Alto SONe at the aeration time corresponding to minimum SON was esti-
mated by combining data from all Batch Studies using Palo Alto activated
sludge. SONd + SONg was estimated from tapwater control systems and the
glucose-fed system of Batch Study 4, and SONr, SON^, and SON were determined
as described earlier. Estimations are summarized in Table 75.
The estimate of SONb may again be erroneously high due to the inability
to separate SONg from SONj. Within the recognized limitations, it is still
reasonable to state that (1) SONeq was a significant source of minimum SONe,
(2) SON produced due to organism decay (SONd) is significant even at minimum
SON , and (3) SON represents the major fraction of minimum SON .
e r e
TABLE 75. ESTIMATION OF SONeq, SONd + SONg, SONr, AND SONb
IN PALO ALTO SECONDARY EFFLUENT AT tmin
SON
e
mg/1
1.40±0.46
(n=8)
SONeq
mg/1
0.22±0.10
(n=8)
% of
SONe
16 ±9
SONd + SON
o
mg/1
0.26+0.16
(n=7)
% of
SONe
19 ±13
SONr
mg/1
0.73±0.17
(n=8)
% of
SONe
52 ±2
SONb
mg/1
0.19±0.53
% of
SONe
14 ±39
i
n = number of samples used for estimation.
± are standard deviations for combined data.
184
-------
Although it was difficult to experimentally separate and determine pre-
cisely the contribution of individual SON sources to Palo Alto activated-
sludge effluent SON, reasonable estimates were obtained. Aeration times of
about 6 hours and MLSS values of 1200-1500 mg/1 are typical for this plant.
For these conditions, using batch study data as a basis for estimation, SONeq
may account for around 14 to 17 percent of SONe; SONd + SONg may account for
13 to 19 percent of SONe; SONr may account for 50 to 60 percent of SONe; and
the remainder is SON, .
b
185
-------
SECTION 9
SON CHARACTERIZATION
INTRODUCTION
The objective of the characterization studies was to determine the nature
of the nitrogen-containing soluble organics in selected wastewaters in order
to help evaluate the source of SON in activated-sludge effluents. Character-
istics evaluated included biodegradability, molecular weight distribution,
and free and combined amino acid content.
Wastewaters Characterized
Several treated and untreated municipal wastewater samples from studies
described in Section 8 were characterized as listed in Table 76.
Biodegradability
Biodegradation studies were used to estimate the refractory fractions
of SONi, SONe, and SONp, (see Section 8) and to determine the rate and extent
of degradation of the wastewater samples listed in Table 76. Samples were
filtered by Method 1 (Table 56) and placed in acid-washed bottles. Phos-
phorus (0.2 mg/1 P from K2HP04) plus 0.5 ml/liter of one-day-old settled
sewage was added as a bacterial seed. The biodegradation bottles with pre-
pared samples were placed in the dark at 20 °C and aerated with an air-C02
mixture to maintain an aerobic environment, to provide gentle mixing, and to
control pH (7.0-7.5). Air and C02 were provided in a manner identical to
that depicted in Figure 8 except that porous glass diffusers were not used.
Sample evaporation was prevented by bubbling the gases through deionized
water for humidification.
TABLE 76. LIST OF SAMPLES CHARACTERIZED
Filtered and unfiltered Palo Alto untreated wastewater.
Filtered and unfiltered Palo Alto primary effluent.
Filtered Palo Alto activated-sludge effluent.
Filtered laboratory AS culture effluent prior to steady-state operation.
Filtered laboratory AS culture effluent during steady-state operation.
186
-------
Samples were taken on Day 0 and at frequent intervals thereafter, fil-
tered, and subjected to the following analysis: SON,SCOD, NH3-N, NO^-N, and
NO^-N. Filtration was by Method 2 and SON analyses by the Technicon method
for Biodegradation Study 3 (Section 8). All other studies used Filtration
Method 1 and Kjeldahl SON analysis. Unfiltered organic nitrogen and COD
analyses were conducted during Biodegratation Study 3. The pH was monitored
frequently during all studies to note the onset of nitrification (significant
decrease in pH). As pH decreased, NaHC03 was added to maintain pH between
7.0 and 7.5.
Molecular Weight Distribution Using Sephadex
Molecular weight distribution was estimated with Sephadex G-15 gel
(Pharmacia Fine Chemicals, Piscataway, N.J.). This gel has a molecular
weight separation range of 0-1500, and was selected since our preliminary
studies and data in the literature [129,209,210,211] indicated that the major
fraction of wastewater organics is contained in this range. Data from Sepha-
dex studies helped to determine the nature of activated-sludge effluent SON,
and was used to compare the distribution for different SON sources.
Preparatory sample treatment included concentrating with an evaporator
concentrator (CALAB, Emeryville, CA). Pall-filtered 500-ml samples were
placed in a 1-liter concentration flask and attached to the evaporator-
concentrator. The sample flask was partially immersed in a 40°C water bath.
Samples were concentrated to approximately 50-100 ml , and if a greater mass
of organic material was desired, additional 500-ml aliquots were added and
concentrated to give the desired concentration of organics. A precipitate
was formed during the concentration step and was removed by a 0.45y Millipore
filter. Recoveries of SON and SCOD were measured.
Sephadex gels were prepared as described in the manual Sephadex: Gel
Filtration in Theory and Practice [219]. Gels were allowed to swell in excess
eluant on a boiling water bath and were then carefully poured into a K26/100
chromatographic column (Pharmacia Fine Chemicals). Inner diameter of the
column was 26 mm, column length was 100 cm, and column capacity was 530 ml.
Once the column was filled with swollen gel to the desired volume, a R25/26
eluant reservoir (Pharmacia Fine Chemicals) was attached and several bed
volumes of eluant were passed through the column to allow for equilibration
of the gel. Void volume was determined by passing high molecular weight
Blue Dextran (MW = 2 x 10^) through the column. Molecular weight estimations
were made by comparing elution properties of a series of standard compounds
(Table 77) of known molecular weight with those of the unknown sample [219].
Five-mi aliquots of the concentrate were applied to the prepared columns
and eluted with 0.04N Na2S04 in deionized water (pH = 7.6-7.8) at flow rates
of 0.5-0.7 ml/min. Column effluents were collected in 10-ml aliquots using
an automatic fraction collector (SMI 1205, Emeryville, CA). The aliquots
were combined to form five larger fractions and duplicate SON (Technicon
method) and SCOD analyses were conducted on each fraction.
187
-------
TABLE 77. COMPOUNDS USED FOR MOLECULAR WEIGHT CALIBRATION
OF SEPHADEX COLUMNS
Compound
Molecular Weight
Blue Dextran (Standard Reference
Compound)
Poly-DL-Alanine
Streptomycin sulfate
Phenylphenylalanine
1-phenylalanine
x 10
1,800
565
312
165
6
Free and Combined Amino Acid Analysis
Free and combined acid concentrations in Palo Alto activated-sludge
effluent were estimated using a Beckman automatic amino acid analyzer made
available by the Chemical Evolution Branch of the NASA Ames Research Labora-
tory in Mountain View, California. The techniques used were developed by
personnel at the Ames Laboratory.
Samples for analysis were concentrated by procedures described under
Molecular Weight Distribution Using Sephadex. Free amino acids were separa-
ted from the concentrated sample by elution with four bed volumes (-50 ml)
of distilled water through a 12-mm-diameter Chromaflex chromatographic
column (No. K42251-2512, Kontes Glass Company, Berkeley, CA) packed with
about 25 gm of Dowex AG50W-8 cationic-exchange resin in the hydrogen form
(Bio-Rad Laboratories, Richmond, CA). Elution of the amino acids from the
column was accomplished with four bed volumes of 2N NttyOH. Flow rates were
approximately 0.5 ml/min. Combined amino acids were hydrolyzed in 6N HC1
at 110°C for 18 hours, evaporated to dryness, resolubilized in distilled
water, pH adjusted to 6.0 with NaOH, and separated as described above.
Organic recoveries through those procedures were calculated by monitoring SOC.
After amino acids were eluted, NIfyOH was collected in a 6N HC1 trap.
Evaporated samples were then analyzed on the Beckman analyzer by trained
personnel at the Ames Research Center. The automated technique for amino
acid separation and detection was similar to that developed by Spackman,
Stein, and Moore [221,222]. Concentrations in the unknown sample were
calculated using a chromatogram obtained from a standard amino acid mixture
containing known concentrations of amino acids.
188
-------
RESULTS
Biodegradability
Objectives—
The objective of these studies was to determine the biodegradability of
SON contained in different treated and untreated wastewaters. These data
would provide information about the relative nature of the SON, the effects
of SON on the environment, and the refractory fractions of 80%, SONe, and
SONp (Eqs. 1 to 3).
Procedures—
Table 78 lists the wastewaters and treated wastewaters evaluated.
Nitrate nitrogen was minitored during biodegradation as concentrations ex-
ceeding 10 mg/1 interfere with SON analysis.
Decay Rate—
One method of comparing the SON biodegradability of different waste-
waters is to compare their decay rates. The reduction in biodegradable SON
was expected to follow first-order kinetics [73,207], and a first-order decay
rate '(kn, day~l) was calculated from the following after Parkin and McCarty
[73]:
SON.
(SONQ - SON f) (Ae"
+ SON
rf
(27)
where SONt is the SON remaining (mg/1) after t days of degradation, and SONrf
is the refractory SON (mg/1). A is a fitting parameter included since the
TABLE 78. WASTEWATERS EVALUATED FOR BIODEGRADABILITY
Biodegradation Study
Number
Sample
BS 1
BS 2
BS 3
BS 4
BS 5
Filtered Palo Alto AS Effluent
Filtered Palo Alto Primary Effluent
Filtered Palo Alto AS Effluent
Filtered and Unfiltered Palo Alto Primary Effluent
Filtered and Unfiltered Palo Alto Raw Wastewater
Batch Study 1 (SON Removal Studies): Primary-Fed
and Control Systems after 0, 6, and 24 Hours
Aeration
Filtered AS Culture 1 Effluent during Logarithmic
Growth of Culture Start-Up
Filtered AS Culture 1 and 2 Effluents during
Steady-State Operation
Filtration as described in Chapter 4
189
-------
initial SON measurement, SONO, is no more accurate than SON measurements at
any other time. SONrf was estimated by a trial and error procedure in which
an initial value for SONrf was assumed. A and kn were evaluated using a
least-squares experimental curve fit, and a correlation coefficient (r) was
calculated. The procedure was repeated until a maximum value of r was ob-
tained for the experimental data and assumed SONrf. It was hoped that this
procedure would provide a reasonable estimate for refractory SON and decay
rate. Decay rates for SCOD (ks) were calculated in a similar manner. The
percent refractory SON contained in the various wastewater samples was
assumed to equal the calculated SON remaining after 60 days degradation using
the A and k values obtained as described above. The assumption is somewhat
arbitrary, but was felt to be satisfactory for comparative purposes.
Treated and Untreated Wastewaters—
The biodegradability of SON and SCOD contained in activated-sludge
effluent, primary and raw wastewaters, and primary wastewater was evaluated.
Results are summarized in Tables 79 and 80.
TABLE 79. SUMMARY OF RATE AND EXTENT OF BIODEGRADABILITY
FOR TREATED AND UNTREATED WASTEWATERS
Sample
(Biodegradation
Study No.)
Filtered raw wastewater
(BS 2)
Unfiltered raw waste-
water (BS 2)
Filtered primary
effluent (BS 1)
Filtered primary
effluent (BS 2)
SON
Percent
Refrac-
tory3
27
19
21
38
Unfiltered primary
effluent (BS 2) 18
Filtered AS effluent
(BS 1)
Filtered AS effluent
(BS 2)
40
50
k
n-l
(day X)
0.075
0.16
0.15
0.12
0.14
0.016
0.012
rb
0.96
0.97
0.98
0.95
0.97
0.82
0.?7
SCOD
Percent
Refrac-
torya
27
26
24
29
28
100
100
k
n-l
(day L)
0.21
0.16
0.30
0.40
0.24
—
-
r
r
0.93
0.97
0.98
0.98
0.98
_
-
Percent refractory = concentration remaining after 60 days degradation
(predicted) divided by original concentration.
r = correlation coefficient for first-order decay model fit.
190
-------
TABLE 80. BIODEGRADABILITY OF REMAINING SOLUBLE ORGANICS AFTER VARIOUS PERIODS
OF TREATMENT OF PRIMARY EFFLUENTS.
Samples Taken during Batch Study 1 (BS 3: Day 0 = 8-18-75)
Day of
Incubation
0
5
10
15
20
30
Time 0-Hr.
Primary
Effluent3
SON*
2.37
1.85
1.50
1.48
1.81
1.72
SCOD*
60
24
23
22
22
-
Time 0-Hr.
Tap
Water
SON
0.35
0.54
0.43
0.61
0.66
0.54
SCOD
14
9
11
11
12
14
Time 6-Hr.
Primary
Effluent
SON
0.90
1.09
1.14
1.26
1.23
1.43
SCOD
20
18
19
-
25
-
Time 6-Hr.
Tap
Water
SON
0.41
0.72
0.72
0.71
0.54
0.41
SCOD
10
10
10
10
11
12
Time 24-Hr.
Primary
Effluent
SON
2.68
2.84
2.71
2.68
2.46
(2.12)
SCOD
23
21
20
23
27
29
Time 24-Hr.
Tap
Water
SON
1.17
1.25
1.06
1.11
1.17
1.07
SCOD
13
10
12
13
14
14
Percent refractory SON = 62 (Day 15) ; percent refractory SCOD = 36 (Day 15) ; all other sample
SON and SCOD values increased during degradation.
*
Values in parentheses represent suspected nitrate interference with Kjeldahl SON analyses.
-------
SON and SCOD in untreated wastewaters were from 18-38 and 24-29 percent
refractory, respectively. Treated wastewaters were more refractory, with SON
and SCOD being 40-50 and 100 percent refractory, respectively. Fluctuations
in the SCOD data prevented obtaining a more reliable estimate for refractory
SCOD.
Decay rates for the biodegradable portion of untreated wastewater SON
were much higher (0.075-0.17 day"1) than for treated wastewater SON (0.012-
0.016 day"1). Untreated wastewater SON and SCOD (0.16-0.40 day-1) decay
rates were similar to those reported for BOD in domestic sewage [24],
Variability in SCOD data prevented the calculation of reliable SCOD decay
rates for activated-sludge effluent samples.
Examination of rate constants (Table 79) and changes in SON/SCOD ratios
during degradation of primary and raw wastewaters shows that SON behaved
differently than SCOD; SCOD was in general degraded at a faster rate. This
agrees with data presented under SON removal, and emphasizes that the portion
of total organics represented by SON acts differently from the organics in
general as measured by SCOD.
The SON remaining after 30 days of biodegradation of a partially treated
primary effluent was significantly higher than that remaining after 6 hours
of high MLSS activated-sludge aeration of the same effluent (Table 80) at a
99 percent level of confidence. This suggests that a short aeration time
with high MLSS activated-sludge treatment may give a better estimate of
untreated wastewater refractory SON (SONr, Eq. 1) than long-term, microbial
seed degradation.
SON and SCOD released during organism decay was not degradable (Table 80,
Tap Water Control Systems). SON and SCOD in general increased during degra-
dation, perhaps due to production by the microbial seed from oxidation of
utilizable organics in the sample, and/or from seed organism decay. Seed
concentrations were the same (0.5 ml/1 of day-old settled sewage) as for
other Biodegradation Studies, and seem too low for seed organism decay to
entirely explain the increase in SON and SCOD. Regardless, data from the
Time 0 control sample indicates that the "equilibrium level" SON (SONeq)
released upon dilution of activated sludge is refractory, an important
observation because it implies that SON will not be removed during treatment
and may persist in the environment upon discharge.
Biologically Produced Organics—
The biodegradability of SON produced during start-up and during steady-
state operation of AS Cultures was evaluated (Table 81). For BS 4, a com-
posite of AS Culture 2 effluent from three consecutive days (Days 15-17)
was taken near the time when the start-up peak concentration of SON occurred
(Figure 17). For BS 5, samples were taken during steady-state operation,
on Days 302-304 for AS Culture 1 and Days 141-143 for Culture 2. Samples
for the first two days were filtered and stored at 4°C prior to combining
with effluent from the third day.
The SON produced near the peak concentration during start-up was 20
percent refractory, a value lower than for treated wastewater SON, but nearly
192
-------
TABLE 81. SUMMARY OF RATE AND EXTENT OF BIODEGRADABILITY
FOR BIOLOGICALLY PRODUCED ORGANICS
Sample
(Biodegradation
Study No.)
Filtered AS Culture 2
during start-up
(BS 4)
Filtered AS Culture 1
during steady-state
(BS 5)
Filtered AS Culture 2
during steady-state
(BS 5)
SON
Percent
Refrac.
torya
20
18
100
kn
(day'1)
0.027
0.029
"
rb
0.94
0.94
"
SCOD
Percent
Refrac-
tory
40
31
39
ks
(day'1)
0.075
0.14
0.090
r
0.87
0.95
0.90
Percent refractory = concentration remaining after 60 days degradation
(predicted) divided by initial concentration.
r = correlation coefficient for first-order decay model fit.
equal that for untreated wastewater SON (Table 79). SON decay rates were low
(about 0.028 day ), about the same as those reported in Table 79 for
activated-sludge effluent samples. SCOD decay rates were somewhat higher
than SON decay rates.
There was a significant difference in the degradability of SON produced
during steady-state operation by AS Cultures 1 and 2: AS Culture 2 SON was
100 percent refractory while AS Culture 1 was about 20 percent refractory.
The initial SON for AS Culture was significantly higher (0.76 mg/1) than that
of Culture 2 (0.24 mg/1), and Culture 1 had undergone a noticeable change in
cultural characteristics 10 days prior to taking samples for evaluation.
During this cultural change, MLSS dropped from around 2000 mg/1 to 1600 mg/1
and SON increased from about 0.4 mg/1 to near 0.8 mg/1. Culture 2 exhibited
stable operation (no cultural changes) for more than 20 days prior to sample
testing. These differences in cultural history provide a partial explanation
for differences in SON degradability, since throughout this report, cultural
characteristics have been linked to effluent characteristics and behavior.
SCOD degradabilities and decay rates were similar for the two cultures.
Changes in SON/SCOD ratios during degradation of steady-state AS Culture
1 and 2 samples again emphasize that the SON subset of organics acts differ-
ently from SCOD in general. The effect was most pronounced for AS Culture 2,
as SON remained constant while SCOD decreased. Significantly more degradable
193
-------
SON was produced during start-up of AS Culture 2 (1.63 mg/1) than during
steady-state operation (0 mg/1). A higher concentration of refractory SON
(0.4 mg/1) was produced during initial start-up than during steady-state
operation (0.26 mg/1); differences were significant at a 99 percent level of
confidence, but were not great.
Summary—
Biodegradation studies have given estimates of the refractory fractions
of 80% (untreated domestic wastewaters), SONe (activated-sludge treated
domestic wastewaters), and SON (SON produced by AS Cultures). 80% is 18
to 38 percent refractory, with the biodegradable portion degrading at rates
of 0.075-0.17 day'1. SONe was from 40 to 50 percent refractory; decay rates
were 0.012 to 0.016 day'1, values much less than those for untreated waste-
waters. SON produced biologically by AS Cultures was 18 to 100 percent
refractory, depending on culture characteristics, and decayed at rates of
about 0.030 day1.
SON released during organism decay was refractory, as was SONeq. SON
and SCOD exhibited different degradabilities; SCOD was degraded at a faster
rate. Cultural characteristic changes affected the quantity of refractory
SON produced by activated-sludge treatment of a low SON synthetic waste. More
degradable SON was produced during AS Culture start-up than during steady-
state operation.
Molecular Weight Distribution
Objective—
The objective of these experiments was to determine the molecular weight
distribution, as defined by Sephadex gel filtration, for selected wastewaters.
This information was used to compare molecular weight characteristics of
different SON sources and to give clues as to the nature of the soluble
nitrogen-containing organics.
Procedures—
Samples analyzed included filtered Palo Alto raw waste-water, primary
and AS effluents and AS Culture 2 effluent taken near peak SON concentration
during culture start-up. The raw, primary, and AS Culture 2 samples were
the same as evaluated in Biodegradation Studies 2 and 4, respectively.
Concentrated samples were eluted through Sephadex G-15 gel which has a
nominal molecular weight exclusion limit of 1500. Eluted fractions were com-
bined in the following molecular weight (MW) ranges: < 165, 165-340, 340-780,
780-1800, and > 1800, as defined by passing organic compounds of known
molecular weight (Table 77) through the column. A set of molecular weight
standards was eluted before and after each set of wastewater samples. Elution
patterns were identical for all sets of standards checked.
For the raw wastewater and primary effluent samples, replicate 5-ml
aliquots were eluted to check reproducibility of the elution procedure.
Duplicate SCOD and Technicon SON analyses were run for each MW fraction. The
SON and SCOD of sample-free eluant were also determined.
194
-------
Statistical methods used to evaluate experimental data are presented in
Appendix A under Statistical Methods Used for Molecular Weight Distribution
Studies.
Sample Concentration—
Table 82 contains a summary of SON and SCOD recoveries obtained during
sample concentration and filtration. Recoveries during concentration were
79 ± 9 percent for SON and 66 ± 9 percent for SCOD. Losses were associated
with a precipitate formed during concentration and removed by filtration
with a 0.45y Millipore filter.
During the vacuum concentration step, an average of 90 percent of the
NH3~N was lost from the samples, reducing potential interference in the SON
analysis.
Molecular Weight Distribution Results—
Tables 83 and 84 respectively summarize SON and SCOD data obtained
from Sephadex separation. All SON and SCOD mass values listed are averages
of duplicate analyses. Experimental uncertainties indicated represent values
for a 95 percent level of confidence.
Reproducibility of the elution procedure was very good as indicated by
results for replicate elutions of primary effluent and raw wastewater (Tables
83 and 84). SON recoveries through the column were 86 +_ 8 percent for the 6
samples tested, and this represented 70+4 percent of the SON in the original
concentrated sample. Corresponding SCOD recoveries were 92 ± 11 percent and
56 + 11 percent.
Some caution must be exercised when interpreting results from these MW
experiments since experimental uncertainty was fairly high, and potential
changes in molecular characteristics, and hence molecular weight distribution,
caused by the concentration procedure are unknown. However, trends can be
noted, and in certain instances, specific conclusions about the MW distribu-
tions can be made.
Molecular weight distributions for the four samples studies, expressed
as percent of SON or SCOD present in the original sample, are summarized in
Figures 34 to 37. Of the material recovered and measured by gel filtration,
50 to 60 percent of the SON and SCOD in raw, primary, and activated-sludge
samples had molecular weights less than 1800. These results agree with data
presented by others [28,129,209,210,211]. MW distributions for filtered raw,
primary effluent, and activated-sludge effluents were very similar. The
activated-sludge effluent contained a slightly higher > 1800 MW fraction and
slightly lower 165-340 MW fraction than did other samples, but differences
were not statistically significant at a 95 percent confidence level. These
data contradict results by Zuckerman and Molof [211] who reported that
activated-sludge treatment selectively removes the 400 MW fraction, leaving
behind the 1200 MW fraction.
The MW distribution pattern for AS Culture 2 effluent during start-up,
however, differed markedly from the wastewater samples just described.
Twenty-three percent of the SON and 41 percent of the SCOD had measured
195
-------
TABLE 82. SUMMARY OF SON AND SCOD RECOVERIES FROM CONCENTRATION AND FILTRATION
Sample
Secondary
Effluent
Secondary
Effluent
Secondary
Effluent3
Secondary
j Effluent^
Primary
Effluent
Raw Waste-
water"
AS Culture
2 Effluentb
Volume
Concen-
trated
(liter)
1.71
1.71
1.0
SON Mass (mg)
Concen-
tration
Factor
46
17
28
2.0 47
l
1.0
1.0
2.0
20
20
36
Before Con-
centration
1.74 ± 0.07
1.74 ± 0.07
1.16 ± 0.03
2.22 ± 0.06
5.50 ± 0.17
5.31 ± 0.20
4.06 ± 0.08
After Concen-
tration &
Filtration
1.21 ± 0.05
1.43 ± 0.05
1.00 ± 0.04
1.56 ± 0.07
4.66 ± 0.19
4.78 ± 0.15
2.88 ± 0.10
Percent •
SON
Recovery
70
82
SCOD Mass (mg)
Before Con-
centration
37.0 ± 2.1
37.0 ± 2.1
j
i
86 20.6 + 0.8
. i
i
J
70 I 47.2 + 1.6
j
i
i
85 i 92.7 ± 2.3
90
71
89.5 ± 2.9
89.0 ± 1.4
After Concen-
tration &
Filtration
28.4 ± 1.5
29.2 ± 1.5
13. 4 ± 0.2
27.6 ± 1.8
50.8 + 0.4
53.5 ± 4.9
63.4 ± 1.0
Percent
SCOD
Recovery
77
79
65
58
55
60
71
aSanrole used to check concentration recoveries. SON and SCOD mass values are averages + standard
i deviations of replicate samples.
I Samples used for molecular weight characterization; SON and SCOD mass values are averages + standard
' deviations of replicate samples.
-------
TABLE 83. SUMMARY OF SON ELUTION DATA FOR MOLECULAR WEIGHT DISTRIBUTION STUDIES
Raw waste-
water:
Elution 1
Elution 2
Primary
Effluent:
Elution 1
Elution 2
AS Effluent
AS Culture
2 during
start-up
phase
Ug SON
Added to
Column
467 ± 23
467 + 23
478 ± 22
478 ±22
178 ± 10
262 ± 16
yg SON Recovered in each Molecular Weight Range*
< 165
11. 6 ± 23.1
8.7±23.1
24.8±23.1
27.7±23.1
16. 3± 23.2
7.7 + 24.6
165-340
189.3±14.5
185.1 ±14. 5
149.6 + 14.6
158. 5± 14.6
59.4 ±14.8
54.8 + 15.6
340-780
120.4 + 14.4
102. 9 ±14. 3
95.5+14.6
107.9 ± 14.7
47. 9 ±14. 7
17.6 + 15.6
780-1800
81.5 + 14.4
72.6 + 14.5
81.8±14.8
77.8 + 14.7
28.4+ 14.6
75.6±14.7
>1800
18.1 + 14.4
10.0 + 14.5
15.6 + 14.6
17. 4± 14.6
19.2 + 14.6
85.6 + 14.7
Percent
Recovery
through
Column
90 ±0
81 ±9
77 ±9
79 ±9
92 + 22
92 ± 16
Percent
Recovery
through
Column
77+7
69 ± 7
69 + 8
71 ± 8
67 + 15
65 ± 11
•* "
± values represent 95 percent confidence limits for analyses.
-------
TABLE 84. SUMMARY OF SCOD ELUTION DATA FOR MOLECULAR WEIGHT DISTRIBUTION STUDIES
Sample
Raw Waste-
water:
Elution 1
Elution 2
Primary
Effluent :
Elution 1
Elution 2
AS Effluent
AS Culture
2 during
start-up
phase
mg SCOD
Added to
Column
5.08±0.41
5.08± 0.41
5.35± 0.40
5.35± 0.40
3.14± 0.27
5.76 + 0.34
mg SCOD Recovered in each Molecular Weight Range*
< 165
0.22± 0.24
0.26±0.24
0.44± 0.24
0.33± 0.24
0.34± 0.25
0.15±0.26
165-340
0.43± 0.15
0.43±0.15
0.46 ±0.15
0.51± 0.15
0.36 ±0.15
0.18± 0.16
340-780
1.86±0.15
1.83± 0.15
2.21±0.15
2. 19 ±0.15
1.19 ±0.15
0.93+0.16
780-1800
1.21± 0.15
1.31+ 0.15
1.15± 0.15
1.03±0.15
0.88± 0.15
1.57 + 0.15
>1800
0.52±0.15
0.47+0.15
0.44 + 0.15
0.38±0.15
0.48± 0.15
3.32 + 0.15
Percent
Recovery
through
Column
83 ± 10
83 ± 10
83 ± 10
83 ± 10
104 + 15
107 ± 9
Percent
Recovery
through
Column
46 ± 5
46 ± 5
53 ±5
53 ±5
60+7
76 ±6
± values represent 95 percent confidence limits for analyses.
VO
OO
-------
co
CO
40
30
20
SCOD RECOVERY=467<
10
o
a:
LJ
30
20
10
0
±4%
\\X\\v\
±3% SON
RECOVERY
= 73%
±3%
±3%
±3%
I800
MOLECULAR WEIGHT RANGE
Figure 34. Molecular weight distribution of recovered SON and SCOD for
soluble raw wastewater as percent of mass in original sample.
± figures are 95 percent confidence limits.
199
-------
SCOD RECOVERY=5I%
o
en
UJ
20
10
0
SON RECOVERY=7I%
±3%
±3%
±3%
±3%
^S^
I800
MOLECULAR WEIGHT RANGE
Figure 35. Molecular weight distribution of recovered SON and SCOD for
soluble primary effluent as percent of mass in original sample.
+ figures are 95 percent confidence limits.
200
-------
40
30
SCOD RECOVERY=60°/<
SON RECOVERY=67%
I800
MOLECULAR WEIGHT RANGE
Figure 36. Molecular weight distribution of recovered SON and SCOD for
soluble activated-sludge effluent as percent of mass in original
sample. + figures are 95 percent confidence limits.
201
-------
50
40 -
30
20
o
E 10
o
0
u 30
o
cr
UJ
°- 20
10
0
Figure 37.
SCOD RECOVERY = 76%
±3%
±2%
±2%
±3% ±2%
\\X\\\\k\WX-CI
SON RECOVERY=65%
±4%
MOLECULAR WEIGHT RANGE
Molecular weight distribution of recovered SON and SCOD for
soluble AS Culture 2 effluent during start-up as percent of mass
in original sample. + figures are 95 percent confidence limits.
202
-------
molecular weights greater than 1800. During culture start-up, bacteria pro-
duce and excrete high molecular weight compounds such as RNA, enzymes, and
proteins, partially explaining the increased proportion of higher MW, SON and
SCOD compounds observed [32,101,113,148].
There is an apparent difference in the nature of the materials in the
various MW fractions as indicated by the SON/SCOD ratios (Table 85). Uncer-
tainties (95 percent confidence limits) were very high for ratios in the
< 165 and 165-340 ranges due to the low SCOD values. The ratios for all MW
fractions except the 165-340 range were very similar, averaging 0.045 ± 0.015.
SON/SCOD values for the 165-340 fraction, however, were significantly higher
(95 percent confidence level), ranging from 0.165 to 0.435 for the 4 samples
studied, and averaging 0.306 + 0.111.
Theoretical SON/SCOD ratios for typical nitrogen-containing organics of
varying molecular weights are listed in Table 86. Nucleic acid bases, their
derivatives, and other heterocyclic nitrogen compounds are the most likely
candidates for the 165-340 MW fraction. Creatinine and hypoxanthine are
examples of these compounds found in wastewaters and treated wastewaters [5].
Most other classes of nitrogen-containing organics have similar SON/SCOD
ratios.
An explanation for the high SON/SCOD ratio of the 165-340 MW fraction
might be that the organics present are resistant to oxidation by the COD
procedure, thus resulting in high measured SON/SCOD ratios. Organics known
to be incompletely oxidized include pyridine and its derivatives, aromatic
hydrocarbons, methylamines, and ethylamines [3,140]. Purine, pyrimidines,
pyroles, and their derivatives are completely oxidized [140]. The former
group of compounds are not likely to be present in large quantities in domes-
tic wastewater nor produced during biological treatment, while the latter
group of SON compounds are known to be produced biologically and to be pres-
ent in treated and untreated wastewaters [5,32,101,148]. Therefore, it seems
more reasonable that the high SON/SCOD ratio for the 165-340 MW fraction was
caused by the presence of nucleic acid bases and/or other heterocyclic nitro-
gen compounds.
Comparison of SON/SCOD ratios for primary and activated-sludge effluents
showed that significantly more SCOD than SON was removed from the 165-340
and 780-1800 MW fractions during activated-sludge treatment (95 percent
confidence level).
Summary—
Molecular weight distribution studies using Sephadex gel filtration has
provided useful information about the nature and source of SON. At least
50 to 60 percent of the SON and SCOD in filtered raw wastewater, primary
effluent, and activated-sludge effluent represented small molecules having
molecular weights of less than 1800. Only three to nine percent of the
measured organics were contained in the > 1800 fraction. However, approxi-
mately 30 and 50 percent of the SON and SCOD, respectively, were not recov-
ered and hence were not measured by the Sephadex procedure. The MW distri-
bution of this fraction in unknown.
203
-------
TABLE 85. SON/SCOP RATIOS FOR THE MOLECULAR WEIGHT FRACTIONS OF THE FILTERED ¥ASTEWATERS STUDIED
Sample
Primary Effluent
Raw Wastewater
Activated- Sludge Effluent
AS Culture 2 Effluent
during -start-up phase
Molecular Weight Range*
< 165
0.068 ±0.052
0.042 ±0.073
0.048 ±0.077
0.051 ±0.177
165-340
0.318 ±0.071
0.435 ±0.178
0.165 ±0.079
0.304 ±0.300
340-780
0.046 ±0.005
0.061 ±0.006
0.040 ±0.013
0.020 ±0.017
780-1800
0.073 ±0.012
0.061 ±0.009
0.032 ±0.017
0.048 ±0.010
> 1800
0.040 ±0.027
0.028 ±0.021
0.040 ±0.024
0.026 ±0.011
&
± values are 95 percent confidence limits for SON/SCOD ratios (Appendix B, item 9).
-------
TABLE 86. THEORETICAL SON/SCOP FOR TYPICAL NITROGEN-CONTAINING
ORGANIC COMPOUNDS OF DIFFERING MOLECULAR WEIGHTS
MW Range
< 165
165-340
340-780
780-1800
> 1800
Class of Compound
Amino acids
Urea
Small nucleic acid bases
Nucleic acid bases
Heterocyclic nitrogen compounds
Small peptides
Small nucleotides and nucleosides
Amino sugars
Hydroxaminic acids
Peptides and amines
Small proteins
lonophores
Nucleotides and nucleosides
Proteins and amines
lonophores
Nucleotides and nucleosides
Coenzymes
Vitamins
Large proteins
Enzymes
RNA
DNA
Humic acids
Theoretical
SON/SCOD
ratios
0.02-0.15
CO
0.20-0.70
0.20-0.40
0.20-0.40
0.05-0.15
0.10-0.20
< 0.10
0.05-0.15
0.05-0.15
0.05-0.15
0.05-0.15
0.10-0.15
0.05-0.15
0.05-0.15
0.10-0.15
0.05-0.15
0.05-0.15
0.05-0.10
0.05-0.10
0.10-0.15
0.10-0.15
< 0.10
MW distributions for the three wastewaters from various stages of treat-
ment were not significantly different. However, activated-sludge treatment
did remove significantly more SCOD than SON from the 165-340 and 780-1800 MW
fractions. SON and SCOD produced during start-up of AS Culture 2 exhibited
a significantly different MW distribution, with 24 and 41 percent of the SON
and SCOD, respectively, contained in the > 1800 fraction.
The nature of the organic material in the 165-340 MW range with an
average SON/SCOD ratio of 0.306, was significantly higher than that of all
other MW ranges (average SON/SCOD - 0.045). This high SON/SCOD ratio sug-
gests that heterocyclic nitrogen compounds and nucleic acid bases are the
major constituents in the 165-340 MW fraction.
205
-------
Free and Combined Amino Acids
Analysis of free and combined amino acids was conducted to estimate the
contribution of these constituents to activated-sludge effluent SON. Filtered
Palo Alto activated-sludge effluent was analyzed and results are summarized
in Table 87.
Amino acid recovery was not determined directly, but was estimated by
monitoring soluble organic carbon (SOC) recovery from each step of the pro-
cedure. Recoveries for the free and combined amino acid procedures were 50
and 25 percent, respectively. Calculating amino acid recoveries using SOC
recovery data may yield high estimates for amino acid concentrations since
the desalting procedure used was designed to select for amino compounds, and
this should give higher SON than SOC recoveries. An additional potential
error is introduced by the concentration procedure in which SON and SOC
recoveries were 70 and 75 percent, respectively. Any amino acids lost during
the concentration step would not be measured. Even with these limitations,
observed data used in conjunction with supporting literature data provided
useful information about activated-sludge effluent SON.
A relatively small fraction of activated-sludge effluent SON and SOC
was comprised of free and combined amino acids (Table 87). These values are
within the percentages of SOC for free and combined amino acids of 0.1 to 4.6
and 1.7 to 10 percent, respectively, reported by others [5,23,213,223,224],
Although none of these previous reports indicate amino acid concentration as
a fraction of SON, it can be stated with reasonable assurance that these
materials comprise a relatively small fraction of activated-sludge effluent
SON.
DISCUSSION OF RESULTS
Summary of Experimental Results
Biodegradability—
Table 88 contains a summary of percent refractory SON and SCOD and
decay rate data for the different types of wastewaters studied.
Treated wastewaters were much more refractory than untreated wastewaters.
SON decay rates for the treated wastewater were very low, as were SON decay
rates for SON produced by laboratory AS Cultures. SCOD in general degraded
at a faster rate than SON, emphasizing differences in the nature of the
nitrogen-containing organics and the general organic fraction as measured
by SCOD.
An important observation is the variability in the degradability of the
two AS Cultures operating under steady-state conditions. Cultures had
similar refractory SON concentrations, 0.14 and 0.26 mg/1 for AS Cultures
1 and 2, respectively, but AS Culture 1 contained 0.62 mg/1 biodegradable
SON while AS Culture 2 contained no degradable SON.
206
-------
TABLE 87. AMINO ACID DATA FOR PALO ALTO ACTIVATED-SLUDGE EFFLUENT (6-14-74)
Amino Acid
Aspartic Acid
Threonine
Serine
Proline
Glut ami c Acid
Glycine
Alanine
Valine
Methionine
Isoleucine
Leucine
Tyrocine
Phenylalanine
Lysine
Histidine
Total
Percent of Original
Effluent SON* or SOC*
Free Amino Acids
yg/i
as SON
1.2
1.3
5.5
-
0.7
3.5
1.6
-
-
0.4
0.3
OJ2
0.3
4.4
3.1
22.5
3.1
yg/i
as SOC
4.1
4.5
14.1
-
3.0
6.0
4.1
_
-
2.1
1.5
1.5
2.3
11.3
5.3
59.8
0.9
Combined Amino Acids
Vig/1
as SON
2.3
0.8
-
-
3.3
5.7
2.3
1.5
0.2
-
-
2.3
0.2
0.5
0.2
19.3
5.3
ug/1
as SOC
7.9
2.7
-
—
14.1
9.8
5.9
6.4
0.9
-
-
17.7
1.5
1.3
0.3
68.5
2.2
Total Amino Acids
Ug/1
as SON
3.5
2.1
5.5
_
4.0
9.2
3.9
1.5
0.2
0.4
0.3
2.5
0.5
4.9
3.3
41.8
8.4
yg/i
as SOC
12.0
7.2
14.1
—
17.1
15.8
10.0
6.4
0.9
2.1
1.5
19.2
3.8
12.6
5.6
128.3
3.1
A
Corrected for estimated recoveries through the analytical procedure (see text) .
SON = 1.45 mg/1; SOC =12.5 mg/1.
ISJ
O
-------
TABLE 88. SUMMARY OF BIODEGRADABILITY DATA
Sample
Untreated wastewaters
AS treated wastewaters
AS Culture 2 during
start-up
AS Cultures during
steady-state
SON
Percent
Refractory
18-38
40-100
20
20-100
Decay Rate
(day-1)
0.075-0.16
0.014-0.016
0.027
0.029
SCOD
Percent
Refractory
24-29
100
40
, 39-40
Decay Rate
(day-1)
0.16-0.40
-
0.075
0.090-0.14
The low-seed microbial procedure for determining biodegradabilities has
limitations because it is not known how much refractory SON is produced
during the procedure. Results from Biodegradation Study 3 showed that high
MLSS batch aeration gave a significantly lower estimate for refractory SON
concentration (0.90 mg/1) than did the low microbial seed procedure (1.48
mg/1). Refractory SON may have been produced during the low microbial seed
procedure. Regardless of these limitations, and since experimental proce-
dures and methods for data analysis were identical for all biodegradation
studies, the percent refractory and decay rate values listed in Table 88 were
considered acceptable for comparison of waste-water degradabilities.
Chemical Nature—
Molecular weight distribution studies using Sephadex gel filtration
showed that from 50 to 60 percent of the SON and SCOD in treated and untreated
wastewaters had apparent molecular weights less than 1800. Three to nine
percent had apparent MW greater than 1800. Thirty percent of the SON and 50
percent of the SCOD were not recovered during the experimental procedure and
hence the MW distribution of this material was not determined. SON produced
during biological oxidation of a glucose-acetate substrate (AS Culture 2)
start-up was significantly different in nature from the treated and untreated
wastewaters; 24 percent of the produced SON and 41 percent of the SCOD had
molecular weights in excess of 1800.
The nature of the SON in the 165 to 340 MW range (containing 15-34
percent of the SON) was significantly different than that in all other MW
fractions. The high SON/SCOD ratio of the 165-340 MW fraction (0.306 versus
0.045 for all other fractions) suggested the presence of heterocyclic nitro-
gen compounds such as nucleic acid bases.
Analysis for free and combined amino acids (proteins) showed that these
materials probably comprise less than 10 percent (3.1 and 5.3 percent,
208
-------
respectively) of the SON contained in activated-sludge effluent. This obser-
vation agrees well with data available in the literature [5,23,212,223,224].
Implications
Biodegradability—
Organic compounds are refractory for many reasons (Section 8). One of
the major reasons for the refractory nature of activated-sludge effluent SON
(0 to 60 percent degradable; decay rates of about 0.014 day'-'-) may be the
low concentration of individual SON compounds. Molecular weight distribution
studies showed that SON compounds of wide-ranging molecular weights were
present; the major fraction, the 165-340 MW range, contained approximately
20 percent of the total SON. If all the SON were a single compound, the
concentration would be about 0.2 to 0.3 mg/1, a level which may approach a
limit for supporting growth and providing maintenance energy. It is not
likely that a single SON compound represents the SON in any MW fraction
[5,13,210,224,226]. Degradation of this wide range of SON compounds may
thus proceed at very slow rates, if at all. Support for this concentration
effect was given by Chudoba [62], who showed that concentration of activated-
sludge effluents resulted in additional organic removal (measured as SCOD).
The low decay rates for produced SON (0.027 and 0.029 day"-'-) , and low
biodegradability (100 percent in some cases), indicates that this SON will
persist for long periods of time upon discharge to receiving waters. The
refractory SON is not likely to support algal growth [73,227,228], but the
slow release of NH3~N from the biodegradable portion will provide some nitro-
gen for growth stimulation.
The biodegradable fraction of activated-sludge effluent SON was found to
be variable, ranging from 0 to 60 percent, 0.010-0.045 day"1. The degradable
fraction of produced SON ranged from 0 to 80 percent, and was related to
changes in culture characteristics (Biodegradation Studies 4 and 5).
The SON in untreated wastewaters was highly degradable, probably due to
the presence of readily degradable substances such as urea and free and com-
bined amino acids, the major components of untreated wastewaters [223,229,
230,231]. Activated-sludge treatment efficiently removes these materials
[223,224]; in this research, free and combined amino acids represented less
than 10 percent of activated-sludge effluent SON.
Chemical Nature—
Both MW distribution studies and earlier reported studies (Section 7)
with cationic exchange at pH 2 suggested the presence of nucleic acid bases
or heterocyclic nitrogen compounds. Similar cation-exchange studies were
conducted on samples analyzed for MW distribution. A comparison of the
results of these studies was made to estimate the fraction of SON likely to
be comprised of these types of materials. Table 89 summarizes the results of
this comparison
Based on apparent molecular weight data nucleic acid bases and hetero-
cyclic nitrogen compounds may comprise up to 23 + 6 percent of activated-
sludge effluent SON. The difference between the SON removed by cationic
209
-------
TABLE 89. COMPARISON OF RESULTS FROM MW DISTRIBUTION AND
CATIONIC-EXCHANGE STUDIES
Percent SON
in 165-340
MW Fraction
Percent SON
removed by
cationic
resins at
pH 2.0
Raw
Waste-
water
34 ±3
Primary
Effluent
29 ± 3
50 ± 7
Acti-
vated-
Sludge
Effluent
23 ±6
35 ± 7
AS Culture 2
during
start-up
15 ±4
33 ±9
AS Culture 1
during
steady-
state
56 ± 38
kS Culture 2
during
steady-
state
.
42 ±12
± values are 95 percent confidence limits.
resins at pH 2 (35+7 percent) and that contained in the 165-340 MW range
may be explained by the presence of free and combined amino acids or the
presence of nucleic acid degradation products (MW > 165-340). Free and
combined amino acids were found to account for approximately 10 percent of
Palo Alto activated-sludge effluent SON. Combining MW distribution, ion
exchange, and amino acid data suggests that a combination of free and com-
bined amino acids (proteins), nucleic acid degradation products, nucleic
acid bases, and heterocyclic nitrogen compounds may account for up to 35
percent of activated-sludge effluent SON.
Heterocyclic nitrogen and nucleic acid bases may comprise up to 30 per-
cent of the SON in untreated wastewaters. Free and combined amino acids or
nucleic acid breakdown products may account for the additional 20 percent
removed by cationic exchange at pH 2.0. Hanson and Lee [229] found that an
average of 36 percent of untreated wastewater SON was free and combined amino
acids. They also estimated the uncharacterized SON fraction at 40-50 percent;
some of this uncharacterized fraction was most likely material similar to
nucleic acid degradation products.
Fifteen percent of the SON produced from oxidation of a glucose-acetate
substrate may be comprised of nucleic acid bases and heterocyclic nitrogen
compounds. . Whether the additional 18 percent SON removal observed with pH
2.0 cationic exchange was due to removal of amino acids and degradation
products of nucleic acids is open to speculation.
The presence of significant quantities of nucleic acid degradation
products, nucleic acid bases, and heterocyclic nitrogen compounds might
210
-------
have been expected based on reports in the literature [100,101,109,112,148].
Organisms excrete and partially degrade RNA during logarithmic growth and
during starvation. RNA degradation products may accumulate to varying degrees
while proteins generally do not [112].
The precise chemical nature of SON contained in other MW fractions could
not be determined. Possible constituents were listed in Table 86. A popular
"catch-all" category for unidentified organic material is humic substances
[22,23,26,224]; from 30 to 65 percent of the organic matter in biological
treatment plant effluents may be contained in this fraction. Analysis for
percent SON associated with such humic substances was not conducted in this
research because of the ambiguous nature of this classification method.
211
-------
SECTION 10
SUMMARY AND DISCUSSION
SON IN MUNICIPAL SECONDARY EFFLUENTS
Analysis of secondary effluents from four activated-sludge plants treat-
ing municipal wastewaters indicated that the soluble organic nitrogen (SON)
concentrations varied from 1.1 to 2.1 mg/1, with an average of 1.5 mg/1.
Analysis of twenty 24-hour composite samples from one treatment plant over a
six-month period indicated a range in SON concentration from 0.8 mg/1 to 1.7
mg/1, with an average of 1.3 mg/1. The standard deviation was 0.24 mg/1.
Thus, the range in concentrations of effluent SON between plants and within
plants appears quite small.
This same survey indicated that soluble chemical oxygen demand (SCOD)
of the secondary effluents was somewhat proportional to the SON. Ratios of
SON to SCOD for the three treatment plants ranged from 0.031 to 0.067 with an
average of 0.047.
SON CHARACTERISTICS
A detailed study of the biological removal and production of SON at one
treatment plant, together with laboratory studies with synthetic wastewaters
indicated that about one-third of the SON in secondary effluents in produced
biologically during treatment, and the other two-thirds represents the
remainder of SON originally present in the untreated wastewater. The latter
material is composed mainly of refractory organic materials which are not
degraded after an additional 30 to 60 days of incubation with a bacterial
seed. About one-half of the portion of SON formed biologically during treat-
ment represents exudate or cellular material released during organism decay.
The other half is released from the cells when they are diluted with new
wastewater. This latter portion of SON rapidly establishes an "equilibrium"
concentration between the cells and the surrounding fluid which is a function
of the bacterial cultural characteristics, and independent of the nature of
the water in which they are diluted.
The concentration of biologically produced SON is highly dependent upon
the characteristics of the microorganisms comprising activated-sludge. How-
ever, for a given culture of microorganisms the extent of production is
consistent. In all cases, the concentration of biologically produced SON
increases significantly up to several mg/1 if there is a disturbance in the
system and during start-up of an activated-sludge system. The SON produced
during this period appears to be quite readily biodegradable. However, the
212
-------
biologically produced SON resulting during normal treatment plant operation
near steady-state is quite resistant to biodegradation.
The SON in untreated wastewater and secondary effluents was character-
ized in various ways in order to evaluate its nature and source. An evalua-
tion of the distribution of molecular weights of the materials comprising
SON indicated that about one-third of the SON organics in secondary effluent
had a molecular weight of less than 340, while only about one-eighth of the
SCOD was in this molecular weight fraction. Most of the SON and SCOD had
molecular weights of less than 1800. However, 30 to 40 percent of the SON
and SCOD were not recovered by the analytical procedure. This portion may
have been dominated by materials with higher molecular weights. Molecular
weight distributions were similar with untreated and primary treated waste-
waters, except that the quantities of SON in the molecular weight range below
340 were greater.
The molecular weight distribution of the relatively large quantity of
soluble organic materials produced biologically during culture start-up gave
a different picture than the above. Here, the fraction of SON with molecular
weight less than 340 was only about 20 percent of the total, and the SCOD only
about 5 percent of the total.
The molecular weight of almost half of the SON was greater than 780, and
of almost half of the SCOD greater than 1800. Thus, in addition to being
highly biodegradable, the material produced during culture start-up was
comprised of larger molecules.
Various lines of evidence suggest that about one-third of the SON, in
particular that with molecular weight less than 340, is comprised of nucleic
acid degradation products. Other studies with pure cultures indicate excre-
tion of such products is common among bacteria. The low COD of this material
is consistent with this hypothesis. In addition, the ability to remove a
significant fraction of the SON by cation exchange at low pH also suggests
the presence of such products. It would be worth while to direct some future
studies specifically towards measurement for such degradation products.
Analysis specifically for amino acids and proteins indicated that they
comprised less than 10 percent of the SON in secondary treatment plant
effluents. This finding is similar to that reported by others.
About 50 percent of the SON could not be characterized by the analytical
procedures used. However, through use of physical and chemical techniques,
the materials could be operationally classified in a manner which was useful
for predicting removal of these materials by physical and chemical processes.
Figure 7, contained in Section 7, presents this fractionation in pictorial
form. About 10 percent of the SON appears to be dominated by a positive
charge and is somewhat polar. This material can be removed by cation ex-
change or chemical coagulation at high pH with either lime or ferric salts.
About 60 percent of the SON is predominantly neutral in charge. One-third
of this appears to be polar and difficult to remove by physical processes.
The other two-thirds is non-polar and can be removed by activated-carbon
adsorption, or partially by coagulation with lime or ferric salts. About 25
213
-------
percent of the SON is dominated by a negative charge, and except for this,
is relatively non-polar. This material can be removed by activated-carbon
adsorption or coagulation with alum or ferric salts at a neutral or slightly
acidic pH. The remainder of the SON appears to represent larger materials,
perhaps colloidal, which are removed by all coagulants and by activated-
carbon adsoprtion.
A similar operational analysis of the fractions comprising SCOD was
carried out. There was considerable similarity between the distribution of
SON and SCOD fractions. The major exception was that a larger fraction of
SON than SCOD was represented by the positively charged materials, as might
be expected since amino groups tend to carry a positive charge. Separation
of soluble organics into fractions in the above way is useful when consider-
ing the potential of different treatment processes. Although analysis for
specific compounds may be of more value when considering potential health
implications of the organic materials, such analysis for a large portion of
the SON materials would be difficult with currently available procedures.
CONTROL OF SON
Physical, chemical, and biological methods were evaluated to determine
their potential for reducing the concentration of SON in treated effluents.
The biological methods involved an evaluation of activated-sludge operational
procedures to maximize removal of SON in the original wastewater and to
minimize the biological production of SON. It was determined that a deten-
tion time and operation representative of conventional design and operation
satisfied this goal. With minicipal wastewaters, a detention time of about
6 hours which corresponds to a solids retention time or sludge age of 4 to
10 days would result in minimal effluent SON. SON production is greatest
during system start-up and following upsets, and least during good steady-
state operation. Long solids retention times also result in more SON pro-
duction through organism decay. Thus, good treatment plant operation with
normal conventional loadings tends to minimize the concentration of effluent
SON. Such operation also minimizes the concentration of effluent SCOD.
The physical and chemical methods evaluated included chemical coagula-
tion, activated-carbon adsorption, ion exchange, chlorination, and ozonation.
In general it was found that the different processes removed different SON
fractions. Thus, a combination of processes was generally capable of re-
moving more SON than a single process alone. These same observations apply
to SCOD removal.
As already discussed under SON Characteristics, SON was operationally
divided into various fractions which are particularly useful when considering
removal by chemical coagulation, activated-carbon adsorption, and ion ex-
change. These results are summarized in Figure 7 which should be reviewed
for reference. Activated-carbon adsorption is capable of removing the non-
polar fraction which constitutes about 70 percent of the SON and 80 percent
of the SCOD. Cation exchange is particularly useful for selectively removing
the positively charged fraction of SON. At a neutral pH it removes about 10
214
-------
percent of the SON and 5 percent of the SCOD. If wastewater pH is lowered
to about 2, nitrogenous materials tend to become further protonated and more
positive in charge so that cation exchange can then remove about 40 percent
of the SON and 20 percent of the SCOD. The materials removed appear largely
to be nucleic acid degradation products and perhaps some amino acids. Cation
exchange in general removes a larger fraction of SON than of SCOD. Anion ex-
change, on the other hand, is not very effective for SON removal, although it
is reasonably good for SCOD removal. The resin used appeared to be most
effective on a predominantly neutral fraction of organics which was partially
polar and partially non-polar. Only about 10 percent of the SON was removed
by anion exchange, while about 30 percent of the SCOD was removed. This
removal was independent of pH.
Chemical coagulation at neutral or slightly acidic pH was effective in
removal of a portion of the soluble organic material which was also removed
by activated-carbon adsorption. Both alum and ferric chloride under these
conditions removed a larger molecular weight fraction and a negatively
charged fraction which comprised about 30 percent of both the SON and SCOD.
In addition, ferric chloride removed 15 percent of the non-polar SON. At
high pH ferric chloride also removed the positive fraction but not the
negative fraction. The fractions removed by lime and ferric chloride under
these conditions were about the same.
These results indicate that a combination of activated-carbon adsorp-
tion, and either lime coagulation or cation exchange were capable of removing
about 85 percent of the SON and SCOD. The non-removed remainder represented
a polar fraction of organics which was neutral in charge and not removed by
any of these particular processes. In order to obtain the maximum removal
noted, large dosages of chemical coagulants or activated carbon would be
required. Use of these particular processes specifically to remove SON would
be expensive, and should only be attempted if removal is absolutely neces-
sary and justified.
Several strong oxidizing chemicals were evaluated. Large dosages of
potassium permanganate and hydrogen peroxide at pH of 10 or higher were
capable of removing 20 to 30 percent of the SON, but none of the SCOD.
Thus they appear to act simply by deamination to form ammonia and nitrogen-
free organics. Large concentrations of free chlorine residual were effective
in removing about 40 percent of the SON, but no more than 20 percent of the
SCOD. On the other hand, ozonation removed only 14 percent of the SON, but
almost 50 percent of the SCOD. Thus, chlorination appears to oxidize the
nitrogenous portion of the organics, while ozone is more effective in oxida-
tion or the carbonaceous portion of the organic molecules. Interestingly,
preozonation of a sample made it much more susceptible to chlorination for
SON removal. Ozonation also tends to make the organics more susceptible to
biological degradation.
ECOLOGICAL SIGNIFICANCE OF SON
This study was initiated to determine the nature, effects, and potential
methods for control of SON in effluents from municipal treatment plants. One
215
-------
primary concern was proposals to develop effluent standards to limit the
total concentration of nitrogen which could be discharged to receiving waters.
Questions have been raised about the desirability of including organic nitro-
gen in a total nitrogen requirement. This study has provided information on
the concentrations of SON which can be expected in secondary effluents, the
nature of this material, and the effectiveness of various processes for its
control. This study did not consider all of the possible ecological ramifica-
tions of SON discharge to receiving waters and does not provide sufficient
new knowledge about the nature of the materials comprising SON which could be
used to evaluate overall ecological effects. The information can be used,
however, to determine the implications of effluent standards in process de-
sign, operation, and cost.
In general, the organics comprising SON in effluents from conventionally
designed and operated activated-sludge plants are only about 50 percent
biodegradable. The biodegradable fraction is oxidized very slowly at rates
of one to two percent per day. Thus, after 30 days only about two-thirds of
the biodegradable material would be oxidized under normal stream conditions.
In this time period, about one-third of the SON would be degraded and re-
leased as ammonia, which can then be used by algae. Previous studies suggest
that little if any of the secondary effluent SON itself is available for
algal growth, although the small portion of SON in the form of amino acids
could presumably be used for this purpose. A concern which is worth further
research is the ecological effect that chlorination of SON materials may have.
Amino acids and nucleic acid degradation products, if chlorinated and then
reincorporated into living forms could possibly cause undesirable effects.
Evaluation of such potential problems, however, was beyond the scope of this
study.
The results of this study indicate that SON is not removed to a signifi-
cant extent by processes used to control or remove the inorganic forms of
nitrogen. In general, the adaptation of specific processes for reduction of
SON would be expensive. Thus, in the formulation of standards for nitrogen
control in wastewater effluents, it is essential that careful consideration
be given to the cost of removal and the potential benefits to be gained,
before soluble organic forms of nitrogen are included.
216
-------
REFERENCES
1. U.S. PublicLaw 92-500; 92nd Congress, S.2770; Oct. 18, 1972.
2. Nitrogenous Compounds in the Environment. E. M. Mrak, Chairman,
Hazardous Materials Advisory Committee, United Stated Environmental
Protection Agency, Wash., B.C., Report No. EPA-SAB-73-001 (Dec. 1973).
3. Standard Methods for Examination of Water and Wastewater (1971), 13th
edition, American Public Health Association, Inc., New York.
4. Horstkotte, G. A., D. G. Niles, D. S. Parker, and D. H. Caldwell,
"Full-scale testing of & water reclamation system," Jour. Water Pollu-
tion Control Federation. 4(3(1), 181 (1974).
5. Hunter, J. V., "Origin of organics from artificial contamination," in
Organic Compounds in Aquatic Environments, S. D. Faust and J. V. Hunter,
eds., Marcel Dekker, Inc., New York, 1971.
6. Hanson, A. M., and G. F- Lee, "Forms of organic nitrogen in domestic
wastewater," Jour. Water Pollution Control Federation, 43(11), 2271
(1971).
7. Fogg, G. E., "Production of extracellular nitrogenous substances by a
blue green algae," Proc. Royal Society (London), B139. 372 (1952).
8. Fogg, G. E., "The extracellular products of algae," Oceanographical
and Marine Biology Annual Review, t\_t 195 (1966).
9. Neilands, J. B., "Hydroxamic acids in nature," Science, 156 1443 (1967).
10. Hunter, S. H., "Inorganic nutrition." Annual Reviews of Microbiology,
_26, 313 (1972).
11. Keeney, D. R., J. G. Konrad, and G. Chesters, "Nitrogen distribution
in some Wisconsin lake sediments," Jour. Water Pollution Control Federa-
tion, ^42(3), 411 (1970).
12. Dean, R. B., "Colloids complicate treatment processes," Environmental
Science and Technology, _3(9), 820 (1969).
13. Helfgott, T., J. V. Hunter, and D. Rickert, "Analytic and process
classification of effluents," Jour. Sanitary Engineering Div., ASCE,
96, SA3, 779 (June 1970).
217
-------
14. Painter, H. A., M. Viney, and A. Bywaters, "Composition of sewage and
sewage effluents," Jour, and Proc. Institute of Sewage Purification
(Great Britain), Part 4, 302 (1961).
15. Hunter, J. V., and H. Heukelekian, "The composition of domestic sewage
fractions," Jour. Water Pollution Control Federation, _37(8) , 1142
(1965).
16. Rickert, D. A., and J. V. Hunter, "General nature of soluble and
particulate organics in sewage and secondary effluent," Water Research,
.5, 421 (1971).
17. Sastry, C. A., P. V. R. Subrahmanyam, and S. C. Pillai, "Amino acids
in treated sewage in India," Sewage and Industrial Wastes, _30(10, 1241
(1958).
18. Kahn, L., and C. Wayman, "Amino acids in raw sewage and sewage efflu-
ents," Jour. Water Pollution Control Federation, ^6(11), 1368 (1964).
19. Beckman, W. J. , R. J. Avendt, T. J. Mulligan, and G. J. Kehrberger,
"Combined carbon oxidation-nitrification," Jour. Water Pollution Con-
trol Federation, 44(10), 1916 (1972).
20. Adrian, G. W., and W. W. Hodges, Jr., "Pilot scale activated carbon
advanced treatment and plans for future recycling at Los Angeles,"
Proc. 15th Water Quality Conference, Univ. of Illinois, Urbana, 111.,
(1973).
21. Esmond, S. E. , and H. W. Wolf, "The status of organics—the Dallas
Water Reclamation Pilot Plant," Proc. 15th Water Quality Conference,
Univ. of Illinois, Urbana, 111., (1973).
22. Rebhun, M., and J. Manka, "Classification of organics in secondary
effluents," Environmental Science and Technology, 5, 606 (1971).
23. Bunch, R. L. , E. F. Earth, and M. B. Ettinger, "Organic materials in
secondary effluents," Jour. Water Pollution Control Federation, 33(2),
122(1961).
24. Wastewater Engineering, Metcalf and Eddy, Inc., McGraw-Hill, Inc.,
New York (1972).
25. Rudolph, W. , and N. S. Chamberlin, "Indole and skatole in sewage,"
Industrial and Engineering Chemistry, 24, 111 (1932).
26. Manka, J. , M. Rebhun, A. Mandelbaum, and A. Bertinger, "Characteri-
zation of organics in secondary effluents," Environmental Science and
Technology, j?, 1017 (1974). "
27. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall, "Protein
measurement with the Folin phenol reagent," Jour. Biological Chemistry.
193, 265 (1951). ~—
218
-------
28. DeWalle, F. B., and E. S. K. Chian, "Removal of organic matter by
activated carbon columns," Jour. Environmental Engineering Div., ASCE,
100. EE5, 1089 (Oct. 1974).
29. Gunsalus, I. C., and C. W. Shuster, "Energy-yielding metabolism in
bacteria," in The Bacteria. Volume II: Metabolism, I. C. Gunsalus and
R. Y. Stanier, eds., 1-58, Academic Press, New York, 1961.
30. Gale, E. F., "The synthesis of proteins and nucleic acids," in The
Bacteria. Volume III: Biosynthesis, I. C. Gunsalus and R. Y. Stanier,
eds., 471-576, Academic Press, New York, 1961.
31. Magasanik, B., "Biosynthesis of purine and pyrimidine nucleotides,"
in The Bacteria. Volume III; Biosynthesis, I. C. Gunsalus and R. Y.
Stanier, eds., 295-334, Academic Press, New York, 1961.
32. Herbert, H., "The chemical composition of micro-organisms as a function
of their environment," Symposium for Society of General Microbiology,
11, 391-416 (1961).
33. Siegel, A., and E. T. Degens, "Concentration of dissolved amino acids
from saline waters by ligand-exchange chromatography," Science, 151,
1098 (March 1966).
34. Minear, R. A., "Characterization of naturally occurring dissolved
organophosphorus compounds," Environmental Science and Technology, j>
(5), 431 (1972).
35. Bremner, J. M. , "Studies on soil organic matter. I. The chemical
nature of soil organic nitrogen," Jour. Agricultural Science, 39, 183
(1949).
36. Bremner, J. M., "Studies on soil humic acids. I. The chemical nature
of humic nitrogen," Jour. Agricultural Science, 46, 247 (1955).
37. Stevenson, F. J., and K. M. Goh, "Infrared spectra of humic acids and
related substances," Geochimica et Cosmochimica Acta, 35, 471 (1971).
38. Otsuki, A., and T. Hanya, "Some precursors of humic acid in recent
lake sediment suggested by infrared spectra," Geochimica et Cosmo-
chimica Acta. _31, 1505 (1967).
39. Ishiwatari, R., "Structural characteristics of humic substances in
recent lake sediments," in Advances in Organic Geochemistry, G. D.
Hobson and G. C. Speers, eds., Pergammon Press, Inc., New York, 1970.
40. Hodge, J. E., "Chemistry of browning reactions in model systems,"
Jour. Agricultural and Food Chemistry, I, 928 (1953).
41. Stevenson, F. J., and S. N. Tilo, "Nitrogenous constituents of deep-
sea sediments," in Advances in Organic Geochemistry, G. D. Hobson and
G. C. Speers, eds., Pergammon Press, Inc., New York, 1970.
219
-------
42. Dugan, P. R., "Bioflocculation and the accumulation of chemicals by
floe-forming organisms," U.S.E.P.A. Report No. 600/2-75-032 (1975).
43. Baier, R. E., E. G. Shafrin, and W. A. Zisman, "Adhesion: mechanisms
that assist or impede it," Science, 162, 1360 (1968).
44. Putnam, F. W., and H. Neurath, "Complex formation between synthetic
detergents and proteins," Jour. Biological Chemistry, 150, 263 (1943).
45. Fitter, P., "Determination of biological degradability of organic
substances." Water Research, 10, 231-235 (1976).
46. Lowry, 0. H., N. J. Rosebraugh, A. L. Farr, and R. J. Randall, "Protein
measurement with the Folin phenol reagent," Jour. Biological Chemistry,
193, 256 (1951).
47. Pfeffer, J. T., "Reclamation of energy from organic refuse," Progress
report on work at Univ. of Illinois, Dept. of Civil Engineering on
EPA-R-800776, April 1973.
48. Methods for Chemical Analysis of Water and Wastewater, U.S.E.P.A.
Report No. 16020-07/71 (1971).
49. Howe, L. H., and C. W. Holley, "Comparisons of mercury(II) chloride
and sulfuric acid as preservatives for nitrogen forms in water samples,"
Environmental Science and Technology, J3, 478 (1969).
50. Eckhoff, D. W., and D. Jenkins, "Transient loading effects in the
activated sludge process," Proceedings of the 3rd International Con-
ference of Water Pollution Research, Munich, 1966.
51. McCarty, P. L., R. Schertenleib, and S. Niku, "Preproject water quality
evaluation for the Palo Alto Water Reclamation Facility," Technical
Report No. 206, Dept. of Civil Engineering, Stanford Univ., April, 1976.
52. Folin, 0., and V. Ciocalteu, "On tyrosine and tryptophane determination
in proteins," Jour. Biological Chemistry? 73, 627 (1927).
53. Stumm, W., and C. R. O'Melia, "Stoichiometry of coagulation," Jour.
American Water Works Association, j60(5), 514 (1968).
54. Black, A. P., and S. A. Hannah, "Electrophoretic studies of turbidity
removal by coagulation with aluminum sulfate," Jour. American Water
Works Association, 13(4), 438 (1961). ~~
55. Alexander, M., "Biodegradation: problems of molecular recalcitrance
and microbial fallibility," Advances in Applied Microbiology 7 35-80
(1965).
56. Alexander, M., "Nonbiodegradable and other recalcitrant molecules,"
Biotechnology and Bioengineering. 15, 611-647 (July 1973).
220
-------
57. Alexander, M., "Environmental and microbiological problems arising from
recalcitrant molecules," Microbial Ecology, _2_, 17-27 (1975).
58. Marion, C. V., and G. W. Malaney, "Ability of activated sludge micro-
organisms to oxidize aromatic compounds," Proc. 18th Ind. Waste Con-
ference, Purdue Univ., Lafayette, Indiana, 1963.
59. Symons, J. M., R. E. McKinney, R. M. Smith, and E. J. Donovan, Jr.,
"Degradation of nitrogen-containing compounds by activated sludge,"
in Advances in Biological Water Treatment, W. W. Eckenfelder and
Brother J. McCabe, eds., Pergamon Press, New York, 1963.
60. Dugan, P. R., Biochemical Ecology of Water Pollution, Plenum Press,
New York, 1972.
61. Rogers, H. J., "The dissimilation of high molecular weight substances,"
in The Bacteria. Volume II; Metabolism, I. C. Gunsalus and R. Y.
Stanier, eds., 257-318, Academic Press, New York, 1961.
62. Chudoba, J., J. Cervenka, and M. Zima, "Residual organic matter in
activated sludge process effluents. IV. Treatment of municipal sewage,"
Scientific Papers of the Institute of Chemical Technology (Czech.),
Technology of Water, F15, 5-19 (1969).
63. Saunders, F. M. , "Effect of cell residence time on the extracellular
organic compounds in activated sludge suspensions," Ph.D. Thesis,
Dept. of Civil Engineering, Univ. of Illinois, Urbana, Illinois, 1975.
64. Grady, C. P. L., Jr., L. J. Harlow, and R. R. Riesing, "Effect of
growth rate and influent substrate concentration on effluent quality
from chemostats containing bacteria in pure and mixed culture,"
Biotechnology and Bioengineering, 14, 391-410 (1972).
65. Bisogni, J. J., Jr., and A. W. Lawrence, "Relationships between
biological solids retention time and settling characteristics of
activated sludge," Water Research, _5, 753-763 (1971).
66. Sherrard, J. H., and E. D. Schroeder, "Relationship between the ob-
served cell yield coefficient and mean cell residence time in the
completely mixed activated sludge process," Water Research, _6,
1039-1049 (1972).
67. Grady, C. P. L., Jr., A. F. Gaudy, Jr., and E. T. Gaudy, "Control
mechanisms operative in a natural microbial population selected for
its ability to degrade £-Lysine. I. Effect of glucose in batch sys-
tems ," .A£pJ^d_Mcrobiologj_, _18, 776-784 (Nov. 1969).
68. Grady, C. P. L., Jr., and A. F. Gaudy, Jr., "Control mechanisms
operative in a natural microbial population selected for its ability
to degrade Jl-lycine. II. Effects of fructose and ribose in batch
systems." Applied Microbiology, 18, 785-789 (Nov. 1969).
221
-------
69. Grady, C. P. L., Jr., and Gaudy, A. F., Jr., "Control mechanisms opera-
tive in a natural microbial population selected for its ability to
degrade £-lycine. III. Effects of carbohydrates in continuous-flow
systems under shock load conditions," Applied Microbiology, 18,
790-797 (Nov. 1969).
70. LaMer, V. K., and T. W. Healy, "Adsorption-flocculation reactions of
macromolecules at the solid-liquid interface," Reviews of Pure and
Applied Chemistry, 13, 112 (1963).
71. Ries, H. E., and B. L. Mayers, "Flocculation mechanism: charge neu-
tralization and bridging," Science, 160, 1449 (1968).
72. Black, A. P., "Basic mechanisms of coagulation," Jour. American Water
Works Association, j>2(4), 492 (1960).
73. Parkin, G. F., and P. L. McCarty, "The nature, ecological significance,
and removal of soluble organic nitrogen in treated agricultural waste-
waters," Technical report No. 180, Stanford Univ., Dept. of Civil
Engineering (Sept. 1973).
74. Malhotra, S. K., G. F. Lee, G. A. Rohlich, "Nutrient removal from
secondary effluent by alum flocculation and lime precipitation,"
International Journal of Air and Water Pollution, 8^, 487 (1964).
75. Wolf, H., "Water-borne nitrogenous materials from municipal areas,"
in Nitrogenous Compounds in the Environment, E. M. Mrak, Chairman,
Hazardous Materials Advisory Committee, U.S.E.P.A., Washington, B.C.,
Report No. EPA-SAB-73-001 (Dec. 1973).
76. Rebhun, M., N. Narkis, and A. M. Wachs, "Effect of polyelectrolytes
in conjunction with bentonitic clay on contaminants removal from
secondary effluents," Water Research, 3, 345 (1969).
77. Yopps, J. A., and D. W. Fuerstenau, "The zero point of charge of
alpha-alumina," Jour. Colloid Science, 19, 61 (1964).
78. Healy, T. W. "Selective adsorption of organics on inorganic surfaces,"
in Organic Compounds in Aquatic Environments, S. D. Faust and J. V.
Hunter, eds., Marcel Dekker, Inc., New York, 1971.
79. Parks, G. A., "The isoelectric points of solid oxides, solid hydroxides,
and aqueous hydroxo complex systems," Chemical Reviews, 65, 177 (1965).
80. Parks, G. A., "Aqueous surface chemistry of oxides and complex oxide
minerals," in Equilibrium Concepts in Natural Water Systems, Advances
in Chemistry Series, No. 67, American Chemical Society, Washington,
B.C., 1967.
81. Black, A. P., and R. F. Christman, "Electrophoretic studies of sludge
Particles produced in lime-soda softening," Jour. American Water Works
Association, _53, 737 (1961). ~
222
-------
82. Faust, S. D., and M. C. Manger, "Distribution of electromobility values
of particulate matter in domestic waste water," Air and Water Pollution,
£, 565 (1965).
83. Sawyer, C. N., and P. L. McCarty, Chemistry for Sanitary Engineering,
2nd edition, McGraw-Hill, Inc., New York, 1967.
84. Busch, P. L., and W. Stumm, "Chemical interactions in the aggregation
of bacteria—bioflocculation in waste treatment," Environmental Science
and Technology, 2, 49 (1968).
85. Process Design Manual for Phosphorus Removal, U.S.E.P.A., Report No.
EPA-625/l-76-001a, April 1976.
86. Green, 0., D. VanDan, B. LaBeau, T. L. Campbell, and S. L. Daniels,
"Ferris chloride and organic polyelectrolytes for the removal of
phosphorus," EPA Report No. 670/2-73-103 (1974).
87. Boggia, C., and G. L. Herriman, "Pilot plant operation at Warren,
Michigan," Proc. of the 43rd Annual Conference of the Michigan Pollu-
tion Control Association, 1968.
88. Lengweiller, H., W. Buser, and W. Feitknecht, "The determination of
the solubility of iron(III) hydroxides with FE5^," Helvitica Chimica
Acta, 44, 796 (1961).
89. Stumm, W., and J. J. Morgan, "Chemical aspects of coagulation," Jour.
American Water Works Association, 54, 971 (1962). Discussion, p. 992,
A. P. Black.
90. Collins, T. T., H. L. Davis, and B. W. Rowland, "A study of the behav-
ior of alumina in relation to sizing," Paper Trade Journal, 113, 94
(1941).
91. Cleaver, C. S., and H. G. Cassidy, "Chromatographic adsorption of
amino acids on organic exchange resins," Jour. American Chemical
Society, _7_2, 1147 (1950).
92. Rebhun, M., and W. J. Kaufman, "Removal of organic contaminants.
Sorption of organics by synthetic resins and activated carbon," SERL
Report No. 67-9, U.S. Berkeley, Calif., Dec. 1967.
93. Horvath, R. S., "Microbial co-metabolism and the degradation of organic
compounds in nature," Bacteriological Reviews, 36, 146-155 (June 1972).
94. McCarty, P. L., "Energetics of organic matter degradation," in Water
Pollution Microbiology, R. Mitchell, ed., Wiley-Interscience, New York,
1972.
95. Nissenbaum, A., and I. R. Kaplan, "Chemical and isotropic evidence for
the in-situ origin of marine humic substances," Limnology and Ocean-
ography, 17.* 570-582 (July 1972).
223
-------
96. Nissenbaum, A., "The organic geochemistry of marine and terrestrial
humic substances: Implications of carbon and hydrogen isotope studies,"
in Advances in Organic Geochemistry 1973, B. Tissot and F. Bienner,
eds., Editions Technip, Paris, 1974.
97. Welte, D., "Recent advances in organic geochemistry of humic substances
and kerogen. A review," in Advances in Organic Geochemistry 1973,
B. Tissot and F. Bienner, eds., Editions Technip, Paris, 1974.
98. Vogels, G. D., and C. van der Drift, "Degradation of purines and
pyrimidines by microorganisms," Bacteriological Reviews, 40, 403-468
(June 1976).
99. Dawes, E. A., and D. W. Ribbons, "Some aspects of the endogenous metab-
olism of bacteria," Bacteriological Reviews, 28, 126-149 (June 1964).
100. Gronlund, A. F., and J. J. R. Campbell, "Nitrogenous substrates of
endogenous respiration in Pseudomonas aeruginosa," Journal of Bac-
teriology, 86, 58-66 (July 1963).
101. McGregor, W. C., and R. K. Finn, "Factors affecting the flocculation
of bacteria by chemical additives," Biotechnology and Bioengineering,
11, 127-138 (March 1969).
102. Moore, S. , D. H. Spackman, and W. H. Stein, "Chromatography of amino
acids on sulfonated polystyrene resins," Analytical Chemistry, 30(7),
1185 (1958).
103. Moore, S., and W. H. Stein, "Chromatography of amino acids on sulfon-
ated polystyrene resins," Jour. Biological Chemistry, 192 663 (1951).
104. Hamilton, P. B., "Ion exchange Chromatography of amino acids," Analyti-
cal Chemistry, J30(5), 914 (1958).
105. Buchanan, D. L., "Desalting amino acid solutions by displacement with
piperidine," Analytical Chemistry, 29, 1877 (1957).
106. Cleaver, C. S., and H. G. Cassidy, "Chromatographic adsorption of amino
acids on organic exchange resins," Jour. American Chemical Society, 72,
1147 (1950).
107. Brown, D. M., "The separation of nucleic acid degradation products,"
in Ion Exchange and Its Applications, Society of Chemical Industry,
London, 1955.
108. Cohn, W. E., "The anion-exchange separation of ribonucleotides," Jour.
American Chemical Society, 72, 1471 (1950).
109. Thomas, T. 0., and R. D. Batt, "Degradation of cell constituents by
starved Streptococcus lactis in relation to survival," Journal General
Microbiology, 58, 347-362 (Nov. 1969).
224
-------
110. Giusti, D. M., R. A. Conway, and C. T. Larson, "Activated carbon
adsorption of petrochemicals," Jour. Water Pollution Control Federation,
46(5), 947 (1974). ~
111. Gulp, R. L., and G. L. Gulp, Advanced Wastewater Treatment, Van
Nostrand Reinhold, New York (1971).
112. Strange, R. E., H. E. Wade, and A. G. Ness, "The catabolism of proteins
and nucleic acids in starved Aerobacter aerogenes," Biochemical Journal,
J36, 197-203 (1963).
113. Saier, M. H., Jr., B. U. Feucht, M. T. McCaman, "Regulation of intro-
cellular adenosine cyclic 3*:5'-monophosphate levels in Escherichia
coli and Salmonella typhimurium," Journal of Biological Chemistry, 250,
7593-7661 (Oct. 10, 1975).
114. Neijssel, 0. M., and D. W. Tempest, "The role of energy-spilling reac-
tions in the growth of Klebsiella aerogenes NCTC 418 in aerobic chemo-
stat culture," Ach. Microbiol., 110, 305-311 (1976).
115. Thompson, J., "Characteristics and energy requirements of an a-
aminoisobutyric acid transport system in Streptococcus lactis," Journal
of Bacteriology, 127. 719-730 (Aug. 1976).
116. Postgate, J. R., and J. R. Hunter, "Accelerated death of Aerobacter
aerogenes starved in the presence of growth-limiting substrates,"
Journal of General Microbiology. 34, 459-473 (May 1964).
117. Calcott, P.H., and J. R. Postgate, "On substrate-accelerated death in
Klebsiella aerogenes," Journal of General Microbiology, 70, 115-122
(March 1972).
118. Pirt, S. John, Principles of Microbe and Cell Cultivation. John Wiley
and Sons, New York, 1975.
119. Garten, V. A., and D. E. Weiss, "The ion- and electron-exchange proper-
ties of activated carbon in relation to its behavior as a catalyst and
adsorbent," Reviews of Pure and Applied Chemistry, ]_, 69 (1957).
120. Snoeyink, V. L., and W. J. Weber, Jr., "The surface chemistry of
active carbon. A discussion of structure and surface functional
groups," Environmental Science and Technology, _!, 228 (1967).
121. Coughlin, R. W., and F- S. Ezra,."Role of surface acidity in adsorp-
tion of organic pollutants on the surface of carbon," Environmental
Science and Technology, ^, 291 (1968).
122. Weber, W. J., Jr., and J. C. Morris, "Kinetics of adsorption on carbon
from solution," Jour. Sanitary Engineering Div., ASCE, 89, SA2, 31
(1963).
225
-------
123. Weber, W. J., Jr., and J. C. Morris, "Equilibria and capacities for
adsorption on carbon," Jour. Sanitary Engineering Div., ASCE, 90, SA3,
79 (1964).
124. Urano, K., and K. Nakamura, "The behavior of heavy metals in the regen-
eration process of sewage treatment activated carbon," Gesuido Kyokai-
Shi. 115(10), 19 (1973), Japan.
125. Spicher, R. G., and R. T. Skrinde, "Potassium permanganate oxidation of
organic contaminants in water supplies," Jour. American Water Works
Association. .55(11), 1174 (1963).
126. Slechta, A. F., and G. L. Gulp, "Water reclamation studies at the South
Tahoe PUD," Jour. Water Pollution Control Federation, _39_, 787 (1967).
127. Parkhurst, J. D., et al., "Pomona activated carbon pilot plant," Jour.
Water Pollution Control Federation, 39, R70 (1967).
128. Kim, B. R., V. L. Snoeyink, and F. M. Saunders, "Influence of activated
sludge CRT on adsorption," Jour. Environmental Engineering Div., ASCE,
EE1, J55 (1976).
129. Keller, J. V., "Activated carbon adsorption of soluble organic nitrogen
compounds in municipal secondary effluent," Engineer Thesis, Stanford
University, Feb. 1976.
130. DeWalle, F. B., and E. S. K. Chian, "The kinetics of formation of humic
substances in activated sludge systems and their effect on floccula-
tion," Biotechnology and Bioengineering, JL6_(6), 739 (1974).
'131. Williams, D. B. , "The organic nitrogen problem," Jour. American Water
Works Association. 4_3, 837 (1951).
132. Wright, N. C., "The action of hypochlorite on amino acids and proteins,"
Biochemistry Journal. J20, 524 (1926).
133. Wright, N. C., "The action of hypochlorite on amino acids and proteins.
The effect of acidity and alkalinity," Biochemistry Journal, 30, 1661
(1936). ~
134. Lawrence, A. W., W. S. Howard, and K. A. Rubin, "Ammonia nitrogen
removal from wastewater effluents by chlorination," Presented at 48th
Mid-Atlantic Industrial Waste Conference, Univ. of Delaware, Nov. 19
1970.
135. Pressley, T. A., D. F. Bishop, and S. G. Roan, "Nitrogen removal by
breakpoint chlorination," Presented at 160th national meeting of the
American Chemical Society, Chicago, 111., Sept., 1970.
136. Murphy, K. L., R. Zaloum, and D. Fulford, "Effect of chlorination
practice on soluble organics," Water Research, 9, 389 (1975).
226
-------
137. Sung, R. D., "Effects of organic constituents in wastewater on the
chlorination precess," Ph.D. Thesis, Univ. of California, Davis, 1975.
138. Taras, M. J., "Preliminary studies on the chlorine demand of specific
compounds," Jour. American Water Works Association. _4_2(5), 462 (1950).
139. Taras, M. J., "Effect of free residual chlorination on nitrogen com-
pounds in water," Jour. American Water Works Association, 45(1), 47
(1953). ~~
140. Chudoba, J., and J. Dalesicky, "Chemical oxygen demand of some nitro-
genous heterocyclic compounds," Water Research. _7, (1973).
141. Morris, J. C., "Formation of halogenated organics by chlorination of
water supplies (a review), "National Technical Information Service
Report No. PB-241 511 (March 1975).
142. Hewes, C. G., and R. R. Davison, "Kinetics of ozone decomposition and
reaction with organics in water," Jour. American Institute of Chemical
Engineering, 1.7(1), 141 (1971).
143. Peleg, M., "Chemistry of ozone in treatment of water," Water Research,
10(5), 361 (1976).
144. Roan, S. G., D. F. Bishop, and T. A. Pressley, "Laboratory ozonation
of municipal wastewaters," EPA Report No. 670/2-73-075 (1973).
145. Nebel, C., et al., "Ozone disinfection of industrial-municipal secon-
dary effluents," Jour. Water Pollution Control Federation, 45, 2493
(1973).
146. Kirk, B. S., R. McNabney, and C. S. Wynn, "Pilot plant studies of
tertiary treatment with ozone," in Ozone in Water and Wastewater Treat-
ment, F. L. Evans, III, ed., Ann Arbor Science Publishers, Inc. Ann
Arbor, Michigan (1972).
147. Singer, P. C., and W. B. Zilli, "Ozonation of ammonia in wastewater,"
Water Research. _9, 127 (1975).
148. Demain, A. L., R. W. Burg, and D. Hendlin, "Excretion and degradation
of ribonucleic acids by Bacillus subtilis," Journal of Bacteriology,
89_ (March 1965).
149. Nichikowa, S., and M. Kurivama, "Nucleic acid as a component of muci-
lage in activated sludge," Water Research, 2, 811 (1968)
150. Harold, F. M., "Conservation and transformation of energy by bacterial
membranes," Bacteriological Reviews, 36, 172 (1972).
151. Payne, J. W., "Peptides and micro-organisms," Advances in Microbial
Physiology. 13, 56-113 (1976).
227
-------
152. Christensen, H. N., Biological Transport, W. A. Benjamin, Inc.,
Reading, Mass., 1975.
153. Boylen, C. W., and J. C. Ensign, "Intracellular substrates for endo-
genous metabolism during long-term starvation of rod and spherical
cells of Arthrobacter crystallopoietes," Journal of Bacteriology, 103,
578-587 (Sept. 1970).
154. Bailey, P. S., "The reactions of ozone with organic compounds,"
Chemical Reviews. _58, 925 (1958).
155. Bailey, P. S., "Organic groupings reactive toward ozone-mechanisms in
aqueous media," in Ozone in Water and Wastewater Treatment, F. L. Evans,
III, ed., Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan, 1972.
156. Bailey, P. S., "Reactivity of ozone with various organic functional
groups important to water purification," in First International Sym-
posium on Ozone for Water and Wastewater Treatment, R. G. Rice and
M. E. Browning, eds., International Ozone Institute, Waterbury, Conn.,
1975.
157. Burleigh, I. G., and E. A. Dawes, "Studies on the endogenous metabol-
ism and senescence of starved Sarcina lutea," Biochemical Journal,
102, 236-250 (1967).
158. Maggiolo, A., and S. J. Niegowski, "Preparation of tertiary amine
oxides by ozonation," Advances in Chemistry Series, 21, 202 (1959).
159. Razumovskii, S. D. , A. L. Buchachenko, A. B. Shapiro, E. G. Rozantsev,
and G. E. Zaikov, "The formation of nitroxyl radicals in the reaction
of amines with ozone," Doklady Chemistry, 183, 1086 (1968).
160. Hewes, C. G., and R. R. Davison, "Kinetics of ozone decomposition and
reaction with organics in water," Jour. American Institute of Chemical
Engineering, r7(l), 141 (1971).
161. Bailey, P. S., J. E. Keller, D. A. Mitchard, and H. M. White,
"Ozonation of amines," Advances in Chemistry Series, 77, 58 (1968).
162. Bailey, P. S., D. A. Mitchard, and A. Y. Khashab, "Ozonation of amines.
II. The competition between amine oxide formation and side-chain oxi-
dation," _J^u£._Org^ni.c_aiemi^tr^, J33(7), 2675 (1968).
163. Bailey, P. S., and J. E. Keller, "Ozonation of amines. III. t-Butyl-
amine," Jour. Organic Chemistry, _33(7)» 2680 (1968).
164. Bailey, P. S., J. E. Keller, and T. P. Carter, Jr., "Ozonation of
amines. IV. Di-ti-bytylamine," Jour. Organic Chemistry. 35, 2777
(1970).
165. Bailey, P. S., and J. E. Keller, "Ozonation of amines. V. Di-t-butyl
nitroxide," Jour. Organic Chemistry, 35, 2782 (1970).
228
-------
166. Erickson, R. E., P. J. Andrulis, Jr., J. C. Collins, M. L. Lungle, and
G. D. Mercer, "Mechanism of ozonation reactions. IV. Carbon-nitrogen
double bonds," Jour. Organic Chemistry. _34, 2961 (1960).
167. Wibaut, J. P., "Ozonolysis of pyrroles, furans, and y-py^ones in connec-
tion with reactivity of bonds in ring system," Advances in Chemistry
Series. 21, 153 (1959).
168. Mudd, J. B., R. Leavitt, A. Ongren, and T. T. McManus, "Reaction of
ozone with amino acids and proteins," Atmospheric Environment, 3, 669
(1969).
169- Delman, A. D., A. E. Ruff, B. B. Simms, and A. R. Allison, "Influence
of Molecular structure and substituents on antiozonant properties of
aryl diamine compounds," Advances in Chemistry Series. 21, 176 (1959).
170. Mosher, W. A., "Structural relationships in addition of ozone to double
bonds," Advances in Chemistry Series, 21, 140 (1959).
171. Medalia, A. I., "Test for traces of organic matter in water," Analyti-
cal Chemistry. .23(9), 1318 (1951).
172. Felbeck, G. T., Jr., "Structural chemistry of soil humic substances,"
Advances in Agronomy, 17, 327 (1965).
173. Spicher, R. G., and R. T. Skrinde, "Effects of potassium permanganate
on pure organic compounds," Jour. American Water Works Association,
_57(4), 472 (1965).
174. Denisov, E. T., V. M. Solyanikov, and A. L. Alexandrov, "Ionic cataly-
sis in chain oxidation of alcohols," Advances in Chemistry Series,
^5, 112 (1968).
175. Schertenleib, R. , "Laboratory investigations for the Palo Alto Water
Reclamation Plant," Engineer's Thesis, Stanford University, 1977.
176. Williams, R. T., in Biological Oxidation of Nitrogen in Organic Mole-
cules, J. W. Bridges, J. W. Gorrod, and D. V. Parke, eds., Halsted
Press, New York, 1972.
177. Postgate, J. R., and J. R. Hunter, "The survival of starved bacteria,"
Journal of General Microbiology, J29, 233 (1962).
178. Radar, R. L., and J. Hochstadt, "Regulation of purine utilization in
bacteria. VII. Involvement of membrane-associated nucleoside phos-
phorylases in the uptake and the base-mediated loss of the ribose
moiety of nucleosides by Salmonella typhimurium membrane vesicles,"
Journal of Bacteriology, 128, 290 (1976).
229
-------
179- Sontheimer, H., "New experiences with ozone in Germany," in First
International Symposium on Ozone for Water and Wastewater Treatment,
R. G. Rice and M. E. Browning, eds., International Ozone Institute,
Waterbury, Conn., 1975.
180. Stern, A. M., and Gasner, L. L., "Degradation of lignin by combined
chemical and biological treatment," Biotechnology and Bioengineering,
16, 789 (1974).
181. CRC Handbook of Biochemistry. Selected Data for Molecular Biology,
H. A. Sober, ed., CRC Press, Cleveland, Ohio, 1970.
182. Lehninger, A. L., Biochemistry, Worth Publishers, Inc., New York, 1970.
183. Semmens, M., and J. Gregory, "Selectivity of strongly basic anion
exchange resins for organic ions," Environmental Science and Technology,
j}(9), 834 (1974).
184. Pavoni, J. L., M. W. Tenney, and W. F. Echelberger, Jr., "Bacterial
exocellular polymers and biological flocculation," Jour. Water Pollu-
tion Control Federation. 44(3), 414 (1972).
185. Moroney, M. J., Facts from Figures, Penguin Books, Inc., Baltimore,
Md., 1951.
186. Laitinen, H. A., and W. E. Harris, Chemical Analysis, 2nd ed. , McGraw-
Hill, Inc., New York, 1975.
187. Analytical Quality Control, U.S.E.P.A., Technology Transfer, June 1972.
188. Edwards, Allen L., Statistical Analysis, 3rd ed., Holt, Rinehart, and
Winston, Inc., New York (1969).
189. Strange, R. E., and F. A. Dark, "'Substrate-accelerated death1 of
Aerobacter aerogenes," Journal of General Microbiology, 39, 215 (1965).
190. Hunter, S. H., "Inorganic nutrition," Annual Reviews of Microbiology,
^6, 313-346 (1972).
191. Kinoshita, S., et al., "Inorganic nutrition," Annual Reviews of Micro-
biology. 2£, 313 (1972). ~~
192. Smeaton, J. R., and W. H. Elliot, "Selective release of ribonuclease-
inhibitor from B. subtilis cells by cold shock treatment," Biochemical
and Biophysical Research Communications, 26, 75 (1967).
193. Rogers, D., "Osmotic pools in Escherichia celi," Science. 159, 531
(1968).
194. Heppel, L., "Selective release of enzymes from bacteria," Science 156
1451 (1967).
230
-------
195. Nossal, N. G., and L. A. Heppel, "The release of enzymes by osmotic
shock from E. coli in exponential phase," Journal of Biological Chem-
istry, 241. 3055 (1966).
196. Lampen, J. 0., "Secretion of enzymes by microorganisms," in Function
and Structure of Microorganisms. Proc. 15th Symposium of Society for
General Microbiology, London, 1965.
197. Clark, V. L., D. E. Peterson, and R. W. Bernlohr, "Changes in free
amino acid production and intracellular amino acid pools of Bacillus
licheniformis as a function of culture age and growth media," Journal
of Bacteriology. 112. 715-725 (1972).
198. Kjeldaard, N. 0., "Regulation of nucleic acid and protein formation in
bacteria," Advances in Microbial Physiology, J., 39-95 (1967).
199. Strange, R. E., F. A. Dark, and A. G. Ness, "The survival of stationary
phase Aerobacter aerogenes stored in aqueous suspension," Journal of
General Microbiology. 25, 61 (1961).
200. Lawrence, A. W., and P. L. McCarty, "Unified basis for biological
treatment design and operation," Jour. Sanitary Engineering Div.,
ASCE, 96, 757-778 (June 1970).
201. Wuhrman, K., "Microbial aspects of water pollution control," Advances
in Applied Micrpbiology, 6>, 119-151 (1964).
202. Randtke, S. J./'Occurrence and removal characteristics of soluble
organic nitrogen in municipal secondary effluent," Ph.D. Thesis,
Stanford Univ., Stanford, California, 1977.
203. Gaudy, A. F., Jr., P. Y. Yang, and A. W. Obayashi, "Studies on the
total oxidation of activated sludge with and without hydrolytic pre-
treatment," Jour. Water Pollution Control Federation, 43, 40 (1971).
204. Gaudy, A. F., Jr., M. Ramanathan, P. Y. Yang, and T. V. DeGeare,
"Studies on the operational stability of the extended aeration process,"
Jour. Water Pollution Control Federation, 42, 165 (1970).
205. Monod, J., "The growth of bacterial cultures," Annual Reviews of
Microbiology. J3» 371-394 (1949).
206. Strominger, J. L., "Biosynthesis of bacterial cell walls," in The
Bacteria. Volume III. Biosynthesis, I. C. Gunsalus and R. Y. Stanier,
eds., 413-470, Academic Press, New York, 1961.
207. Pike, E. B., and C. R. Curds, "The microbial ecology of the activated
sludge process," in Microbial Aspects of Pollution, G. Sykes and F. A.
Skinner, eds., Academic Press, New York, 1971.
208. Jewell, W. J., and P. L. McCarty, "Aerobic decomposition of algae,"
Environmental Science and Technology, _5_, 1023 (Oct. 1971).
231
-------
209. Bender, M. E., W. R. Matson, and R. A. Jordan, "On the significance of
metal complexing agents in secondary sewage effluents," Environmental
Science and Technology, 4-, 520 (1970).
210. "Characterization and separation of secondary effluent components by
molecular weight," 16020 FEN 03/71, Water Pollution Control Research
Series, Arthur D. Little, Inc., Environmental Protection Agency,
Washington, D.C. (1971).
211. Zuckerman, M. M., and A. H. Molof, "High quality reuse by chemical-
physical wastewater treatment," Jour. Water Pollution Control Federa-
tion, 42_, 437 (1970).
212. Irvine, R. L. , and D. J. Schaezler, "Kinetic analysis of data from
biological systems," Jour. Sanitary Engineering Div., ASCE, 97, 409
(1971).
213. Murphy, K. L., M. J. Sedivy, and G. A. Vigers, "Residual organics
following biological treatment," Proc., 6th Canadian Symposium on
Water Pollution Research, Univ. of Toronto, Toronto, Canada (1971).
214. Chudoba, J., "Residual organic matter in activated sludge process
effluents. I. Degradation of saccharides, fatty acids, and amino
acids under batch conditions," Scientific Papers of the Institute of
Chemical Technology (Czech.), Technology of Water, F12, 39 (1967).
215. Gaudy, A. F., Jr., and R. S. Englebrecht, "Basic biochemical considera-
tions during metabolism in growing vs. respiring systems," in Advances
in Biological Waste Treatment, W. W. Eckenfelder and Brother J.
McCabe, eds., Pergamon Press, New York, 1963.
216. Komolrit, K., K. C. Goel, and A. F. Gaudy, Jr., "Regulation of
exogeneous nitrogen supply and its possible applications to the
activated sludge process," Jour. Water Pollution Control Federation,
39_, 251 (1967).
217. Rao, B. S., and A. F. Gaudy, Jr., "Effect of sludge concentration on
various aspects of biological activity in activated sludge," Jour.
Water Pollution Control Federation, 38, 794 (1966).
218. Spiegel, M. R., Schaums's Outline of Theory and Problems of Statistics,
McGraw-Hill Book Co., New York, 1961.
219. Sephadex: Gel Filtration in Theory and Practice, Pharmacia Fine
Chemicals, Inc., Piscataway, New Jersey, 1975.
220. Crow, E. L., F. A. Davis, and M. W. Maxfield, Statistics Manual, Dover
Publications, Inc., New York, 1960.
232
-------
221. Spackman, D. H., W. H. Stein, and S. Moore, "Automatic recording
apparatus for use in the chromatography of amino acids," Analytical
Chemistry. 30, 1190 (1958).
222. Spackman, D. H., W. H. Stein, and S. Moore, "Chromatography of amino
acids on sulfonated polystyrene resins," Analytical Chemistry, 30,
1185 (1958).
223. Kahn, L., and C. Wayman, "Amino acids in sewage and sewage effluents,"
Jour. Water Pollution Control Federation, 36, 1368 (1964).
224. Notes on Water Pollution, No. 70, Water Research Centre, Stevenage
Laboratory, Stevenage, Herts. (Sept. 1975).
225. A Textbook of Clinical Pathology, S. E. Miller, ed., Williams and
Wilkins, Baltimore, Md., 1955.
226. O'Shea, J. , and R. L. Bunch, "Uric acid as a pollution indicator,"
Jour. Water Pollution Control Federation, 37, 1444 (1965).
227. Dunstan, W. M., and D. W. Menzel, "Continuous cultures of natural
populations of phytoplankton in dilute treated sewage effluent,"
Limnology and Oceanography, 16, 623 (1971).
228. Middlebrooks, E. J., D. B. Porcella, E. A. Pearson, P. H. McGauhey,
and G. A. Rohlich, "Biostimulation and algal growth kinetics of waste-
water," Jour. Water Pollution Control Federation, 43, 454 (1971).
229. Hanson, A. M. , and G. F. Lee, "Forms of organic nitrogen in domestic
wastewater," Jour. Water Pollution Control Federation, 43, 2271 (1971).
230. Hanson, A. M., and T. F. Flynn, Jr., "Nitrogen compounds in sewage,"
Proc. 19th Ind. Waste Conference, Purdue Univ., Lafayette, Indiana,
1965.
231. Painter, H. A., and M. Viney, "Composition of a domestic sewage,"
Jour, of Biochemical and Microbiological Technology and Engineering, ^,
143 (1959).
232. Schlucter, A., "Nitrate interference in total Kjeldahl nitrogen deter-
minations and its removal by anion exchange," U.S.E.P.A. Report 600/
7-77-017 (1977).
233
-------
APPENDIX A
STATISTICAL CALCULATIONS
CALCULATION OF THE COMBINED STANDARD DEVIATIONS SHOWN IN TABLE 7
Before combining the standard deviations from different sets of samples,
it must first be determined that all samples are from the same sample popula-
tion. This can be done by using the variance ratio test (or F-test) describ-
ed by Moroney [185] at the 5-percent level of significance. For the SCOD
analysis :
F=
F(7,7) = 3.79
Since 2.44 < 3.79, the two sets of samples are from the same sample
population and their standard deviations, s , may be combined by use of the
following equation: „
B
2
Q — - --- .
(n -1) S + (n-l) 8 + ... (n-l) a
nl + n2 + ' " "i ~ ±
where n represents the number of replicate analyses of a sample, i repre-
sents the number of sample sets, and s. represents the standard deviation
of a given sample set.
Thus, for the SON analysis:
«2 = (6-D(0.03)2 + (9-1K0.03)2 + (4-1H0.03)2 + (8-l)(0.03)2
6 + 9 + 4 + 8-4
s2 = 0.0009
s = 0.03
For the SCOD analysis:
2 = (8-l)(0.50)2 + (8-l)(0.32)2 •
8 + 8-2
s2 = 0.1762
s = 0.42
234
-------
ESTIMATED ERROR OF THE SON AND SCOP ANALYSES
The estimated error of an analysis may be calculated by determining the
inherent error in each step of the analysis and then propagating the errors.
The errors associated with the SON analysis result from the individual in-
herent errors in the use of glassware and instruments. The errors listed for
SON and SCOD analyses were taken from the manufacturers' specifications, the
error printed on the glassware, or from Ref. 187. For the Kjeldahl SON
analysis the errors are:
Error
Error
A. spectorphotometer reading 0.1% T + 0.010
B. volumetric flask (500 ml) + 0.2 ml + 0.00040
C. graduated cylinder (350 ml) ± 1 ml ± 0.00286
D. Nessler tube (50 ml) + 0.05 ml + 0.0010
E. Nessler tube + Nessler 's
reagent (52 ml) + 0.07 ml + 0.0014
The error of the spectrophotometer given by the manufacturer (Bausch
and Lomg) is 0.1% T. The minimum detectable change in concentration (MDC)
for such an instrument is given by:
MDC =
T In T
The transmittance (T) for the least concentrated SON samples was always
less than 0.851 (absorbance = 0.070), giving
MDC -
(0.851) ln(0.851)
= 0.010
The estimated error (E) in the Kjeldahl SON analysis resulting from the
inherent errors in the measurements above is calculated as follows:
E2 - (0.010)2 + (0.0004)2 + (0.00286)2 + (0.0014)2
E2 = 0.00011
E = 0.0105
E = 0.05%
For the Technicon SON analysis, the inherent errors were:
235
-------
Error Relative
Error
Digestion tubes (75 ml) + 0.08 ml _+ 0.0011
Pipetting of sample (20 ml) ± 0.02 ml + 0.0010
Instrument error + 1% full- + 0.01
scale
Detection of the colorimeter-recorder system of the technicon was given by
the manufacturer as ± 1% of full-scale deflection. For low-level SON deter-
minations, full scale was adjusted to equal 1 mg/1 SON, making the relative
error + 0.01. Since calculation of the SON concentration involved subtract-
ing both a baseline reading and a sample blank reading, propagation of in-
dividual inherent errors results in
E2 = f (.Ol)2 + (.Ol)2 + (O.I)2 I + (.001)2 + (.0011)2
E = 1.74%
For the SCOD analysis the inherent errors are:
„ Relative
Error „
Error
A. Pipetting of sample (20 ml) + 0.02 ml 0.001
B. Pipetting of
dichromate (10 ml) + 0.02 ml 0.002
C. Titration of sample + 0.05 ml
D. Titration of blank + 0.05 ml
E. Titration of
standard (25 ml) + 0.05 ml 0.002
Since calculation of SCOD involves subtration of the sample titration
volume from the blank titration volume, the resultant difference has an
error of
2 21/2
[(0.05) + (0.05) ]' = ± 0.071 ml
The relative error of this difference is dependent upon the SCOD of the
sample. For a sample with a SCOD of 20 mg/1, the difference is 5 ml, and
E2 = [-^Pj + (.001)2 + (.002)2 + (.002)2
E2 = .000211
E = .0145
E = (.0145)(20) = 0.29 mg/1
236
-------
For duplicate SCOD analyses, as run during this study,
" - °*29 0.21 mg/1
The estimated errors of the SON and SCOD analyses are less than the
standard deviations of these analyses, indicating that the deviations shown
are reasonable.
CALCULATION OF CONFIDENCE LEVELS
The confidence level (CI) for a given measurement is equal to the
product of the standard deviation and students' t divided by the square
root of the number of observations [186].
Students' t is a function of the degree of confidence (such as 68, 95,
or 99 percent confidence) and the degrees of freedom in calculating the
standard deviation. The number of degrees of freedom is given by the total
number of samples minus the number of sample sets. For the SON and SCOD
analyses, as shown in Table 7, the degrees of freedom are 23 and 14, respec-
tively.
For the SON analysis, in the case of a single observation,
,1/2
CI(95%) =
= 2.064(0.03)
= 0.06
CI(99%) = 2.797(0.03)
= 0.08
For the SCOD analysis, in the case of a duplicate observation:
1/2
CI(95%) = ts/(2)J-/Z
_ 2.145(0.42)
1.414
= 0.6
2.977(0.42)
= 0.9
The discussion of the results of the experiments described in Section 7
is largely concerned with the arithmetic difference between two measurements,
237
-------
and in many cases this difference is quite small. The standard deviation of
the difference between two numbers equals the square root of the sum of the
squares of the individual standard deviations. Thus the confidence level for
the difference between two numbers equals the confidence level for a single
observation multiplied by the square root of 2. The confidence levels for
the difference between two analyses are presented in Table A-l, along with
those for single analyses.
The results of many of the experiments are expressed in terms of percent
removal. Calculation of percent removal involves division of the difference
between two measurements by the concentration present in the untreated sample.
The standard deviation of percent removal depends upon the relative standard
deviations of two numbers, and thus cannot be calculated precisely without
knowing the particular numbers involved. An example of such a calculation is
as follows:
Consider a case in which the initial SON concentration is 1.00 mg/1 and
two treated samples contain 0.40 and 0.25 mg/1 of SON. The difference be-
tween these samples is 0.15 + .04 mg/1 (or 0.15 + 0.12 mg/1 with 99% con-
fidence) .
The difference in percent removal (R) is given by:
0.40 - 0.25
R =
1.00
(100) = 15%
The relative standard deviation of R , S , is given by
K
0.03 2
1.00
= 0.0793
SD = 0.282
K.
TABLE A-l. CONFIDENCE LEVELS FOR SON AND SCOD ANALYSES
Value for a single analysis ,
mg/1:
Value for the difference
between two analyses , mg/1:
Analysis
SON
± 0.03
± 0.06
± 0.08
± 0.04
± 0.08
± 0.11
SCOD*
±0.3
± 0.6
± 0.9
±0.4
± 0.9
± 1.3
Confidence
Level,
Percent
67
95
99
67
95
99
*
Duplicate samples.
238
-------
In this example the relative standard deviation of the initial concentra-
tion is negligible and S is approximately equal to the relative standard
deviation of the difference:
0.280 * 0.282
This will be true whenever the difference is small compared to the initial
concentration, i.e., for small values of percent removal.
SIGNIFICANCE TESTING FOR DIFFERENCES BETWEEN TWO ANALYSES
Students' t test may be used to test the hypothesis that two means do
not differ significantly. The quantity t is defined as the difference be-
tween two means divided by its standard deviation [186]. To be significant,
a difference must exceed the produce of Students' t and its standard devia-
tion, the confidence level being determined by the confidence level of t .
The confidence levels shown in Table A-l for the difference between two
analyses are equal to the product of Students' t and the standard deviation;
thus, differences must exceed the values shown in Table A-l to be considered
significant at a given confidence level. For example, the difference between
two SON analyses is significant with 95% confidence if it exceeds 0.08 mg/1.
In this study, the data are frequently expressed in terms of percent
removal. The difference in percent removal which is significant is depen-
dent upon the initial concentration of SON or SCOD of the sample, since it
equals the appropriate value from Table A-l multiplied by 100 and divided
by the initial concentration. Significant differences in percent removal as
a function of initial SON or SCOD concentration and confidence level are
presented in Table A-2.
SIGNIFICANCE TESTING FOR DIFFERENCES BETWEEN SETS OF ANALYSES
In some instances, it is desirable to compare the difference between two
sets of analyses run on a number of different samples. The sample variation
may be quite large in some cases, concealing differences in the data. This
necessitates a pair-wise comparison of the data points, and the application
of Students' t test to the difference between means of correlated observa-
tions as described by Edwards [188].
For example, in Table 23, one might wish to know whether ferric chloride
removes significantly uore SCOD than does lime. Since there is considerable
sample variation, only those samples may be considered for which both coagu-
lants were used (in this case, Samples 1 and 4 through 7). The data to be
used for the sample calculation is shown in Table A-3, and was taken from
Table 23.
The significance test is as follows [188):
n = the number of paired observations
239
-------
TABLE A-2. SIGNIFICANT DIFFERENCES IN PERCENT REMOVAL
Concentration, mg/1
Initial SON:
0.90
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
Initial SCOD:
20.0
22.5
25.0
27.5
30.0
Difference in Percent Removal to Be Exceeded
to Constitute a Significant Difference
677= CI
4
4
4
3
3
3
3
3
2
2
2
2
1
1
95% CI
9
8
7
7
6
6
5
5
5
5
4
4
3
3
99% CI
12
11
10
9
' 8
8
7
7
6
7
6
5
5
4
TABLE A-3. REMOVAL OF SCOD BY FERRIC CHLORIDE AND LIME
Sample
1
2
3
4
5
n = 5
Percent SCOD
Removed by
FeCl3
19
39
29
27
31
xx = 29.0
Percent SCOD
Removed by
CaO
21
36
22
29
14
x2 = 24.4
D
-2
3
7
-2
17
£ = 23
D2
4
9
49
4
289
2 = 355
-—• -- " *
240
-------
S(D) - the standard deviation of the differences between the
paired observations
K.J, = the average of the first set of observations
X2 = t'ie avera8e °f tne second set of observations
D = the difference between paired observations
~x2) = the standard error of the difference between two means when
observations are paired.
S(D) = p(D2) - (Sd)2/n1 1/2 m P355 - (23)2/5l
= 7.89
7.89 _
~ 3'53
°VA1 ~2' nl/2 ~ 5l/2
Xl " X2 29.0 - 24.4 . _n
75T)—^
For a table of Students' t , we find
t_95>4 - 2.776
Since 2.776 > 1.19, the difference between the means is not significant (at
a 95% level of confidence).
STATISTICAL METHODS USED IN SECTION 8
Precision and Accuracy of SON and SCOD Analyses
Data in Tables A-4 to A-7 show the precision and accuracy of the SON
and SCOD analyses used for the studies described in Section 8.
Confidence Levels for SON and SCOD Analyses
Confidence levels for low-level SON (< 1.3 mg/1) and SCOD (< 25 mg/1)
analyses were calculated as described under Calculation of Confidence Levels.
Confidence levels for high-level SON and SCOD values were estimated by
noting that relative deviations, sample set standard deviations divided by
the average concentration of the set, were essentially constant. For example,
from Table A-7:
241
-------
TABLE A-4. PRECISION AND ACCURACY OF LOW-LEVEL KJELDAHL SON ANALYSES
Type of Sample
0.05 mg/1 Norleucine standard
0.10 mg/1 Norleucine standard
plus 40 mg/1 NH3-N
40 mg/1 NH3-N
0.91 mg/1 E.P.A. standardb
1.06 mg/1 Norleucine standard
plus 20 mg/1 NH3-N
Stanford Tap Water
Stanford Tap Water
Palo Alto Secondary Effluent
Palo Alto Secondary Effluent
Palo Alto Secondary Effluent
AS Culture 2 Effluent
Palo Alto Secondary Effluent
Palo Alto Secondary Effluent
Activated Carbon Effluent
Activated Carbon Effluent
Number
of
Replicates
4
5
4
5
5
5
4
3
3
3
3
6
9
4
8
Average
SON
mg/1
0.05
0.12
0.01
0.89
1.07
0.07
0.04
1.16
1.11
1.17
0.24
0.96
1.08
0.19
0.38
Standard3
Deviation
mg/1
0.01
0.02
0.02
0.03
0.03
0.03
0.02
0.03
0.03
0.04
0.02
0.04
0.03
0.04
0.03
Reference
This study
This study
This study
This study
This study
This study
This study
This study
This study
This study
This study
Keller0
Keller
Keller
Keller
All samples were double-distilled.
Combined standard deviation = +0.03 mg/1 as determined by the method
described in Appendix B, items 3 and 4.
Prepared standard provided by E.P.A. Analytical Quality Control
Laboratory, Cincinnati, Ohio.
Data provided by John V. Keller, Research Assistant, Stanford Uni-
versity.
242
-------
TABLE A-5. PRECISION AND ACCURACY OF TECHNICON SON ANALYSIS
Type of Sample
3.0 mg/1 Norleucine standard
1.0 mg/1 Norleucine standard
0.2 mg/1 Norleucine standard
4.0 mg/1 Norleucine standard
plus 20 mg/1 NH3-N
3.0 mg/1 Norleucine standard
plus 20 mg/1 NH3-N
2.0 mg/1 Norleucine standard
plus 20 mg/1 NH3-N
1.0 mg/1 Norleucine standard
plus 20 mg/1 NH3-N
0.2 mg/1 Norleucine standard
plus 20 mg/1 NH3-N
20 mg/1 NH3-N
20 mg/1 NH3-N
Palo Alto AS Effluent
Palo Alto AS Effluent
Palo Alto AS Effluent
San Jose/Santa Clara AS
Effluent
AS Culture 1 Effluent
Palo Alto Primary Effluent
Palo Alto Primary Effluent
Number
of
Replicates
8
4
4
8
8
8
4
4
4
8
3
3
4
2
10
8
8
Average
SON
mg/1
2.97
1.00
0.20
4.14
3.00
2.06
1.03
0.34
0.08
0.10
1.08
1.22
1.17
1.69
0.35
4.12
4.14
Standard3
Deviation
mg/1
0.06
0.04
0.05
0.12
0.06
0.06
0.10
0.08
0.05
0.06
0.08
0.08
0.07
0.06
0.07
0.11
0.12
Reference
This study
Tl
II
II
11
II
II
1?
11
II
II
[129]
[129]
This study
ii
11
ii
aCombined standard deviation = ±0.08 mg/1 .
243
-------
TABLE A-6. PRECISION OF LOW-LEVEL SCOP ANALYSIS
Type of Sample
Stanford Tap Water
Stanford Tap Water
Stanford Tap Water
Palo Alto Secondary Effluent
Palo Alto Secondary Effluent
Palo Alto Secondary Effluent
Palo Alto Secondary Effluent
AS Culture 1 Effluent
AS Culture 2 Effluent
Number
of
Replicates
3
6
3
3
3
3
3
3
3
Average
SCOD
mg/1
2.8
2.5
2.4
23.4
20.6
23.6
25.9
24.3
16.8
Standard3
Deviation
mg/1
1.3
0-4
0.4
1.1
0.8
1.1
0.4
1.1
0.8
All data from this study.
aComblned standard deviation = ±0.8 mg/1.
TABLE A-7. PRECISION OF KJELDAHL SON AND SCOD ANALYSES
AT HIGH CONCENTRATIONS
Type of Sample
Palo Alto Primary Effluent
11
ii
Palo Alto Raw Wastewater
AS Culture 2 Effluent
Palo Alto Primary Effluent
ii
it
Palo Alto Raw Wastewater
AS Culture 2 Effluent
Anal-
ysis
SON
1!
II
II
SON
SCOD
it
ii
M
SCOD
Number
of
Repli-
cates
3
3
3
3
6
3
3
3
3
6
Average
Concen-
tration
(mg/1)
5.31
4.19
4.47
5.50
2.03
89.5
97.6
103.3
92.7
44.5
Standard
Devia-
tion
(mg/1)
0.22
0.13
0.16
0.17
0.06
2.9
1.5
5.7
2.2
1.8
Combined
Standard
Deviation
(mg/1)
0.17
0.06
3.5
1.8
244
-------
Average Standard Relative
SON Deviation Deviation
Sample (mg/1) (mg/1) (mg/1)
Primary effluent 5.31 0.22 0.04
Primary effluent 4.19 0.13 0.03
Primary effluent 4.47 0.16 0.04
Raw wastevater 5.50 0.17 0.03
AS Culture 1 2.03 0.06 0.03
The average relative deviation is 0.03 + 0.0006, compared to the combined
standard deviation of + 0.03 mg/1 for low-level SON (< 1.3 mg/1). Since
high-level SON determination involves multiplying the concentration actually
measured by the dilution factor used to bring the sample into the
0.8-1.3 mg/1 range, it was felt that estimating the high-level standard
deviation as shown below was reasonable:
s(SON > 1.3 mg/1) = SON(0.03)
Similarly, the standard deviation of high level SCOD values (> 25 mg/1)
was found to vary as SCOD increased, the average relative deviation for the
samples listed in Table A-7 was 0.03 + 0.02. For purposes of this report,
high-level SCOD standard deviation was taken to be
s(SCOD > 25 mg/1) = SCOD(0.03)
Results from these studies in many cases were compared by noting
arithmic differences between two analyses. The standard deviation of such
a difference is the square root of the sum of the squares of the individual
standard deviations. Confidence levels for differences can then be calculated
by multiplying the confidence level for a single analysis by the square root
of 2. Confidence levels for single analyses and for differences between two
analyses are presented in Table A-8.
Confidence Levels for Differences Involving SONOC and SCODOC
For some results in Section 8, a statistical comparison between measured
SON and initially calculated SON (SONOC, Eq. 7) was required. SONOC is the
weighted sum of two measured SON values, making the standard deviation of
differences involving SON
oc
s =/(.03)2 + (.03)2 + (.03)2
Differences involving SCODOC are similarly defined. Appropriate confidence
levels for these differences are listed in Table A-9.
245
-------
TABLE A-8. CONFIDENCE LEVELS FOR SON AND SCOD ANALYSES
Kjeldahl SON
(< 1.30 mg/1)
Kjeldahl SON
(> 1.3 mg/1)
Technicon SON
SCOD (< 25 mg/1)
SCOD (> 25 mg/1)
Value for single
analyses
(mg/1)
95% CI
±0.06
±0.06(SCOD)
±0.16
±1.6
±0.06 (SCOD)
99% CI
±0.08
±0.08(SCOD)
±0.21
±2.3
±0.08(SCOD)
Value for difference
between two analyses
(mg/1)
95% CI
±0.08
±.008(SCOD)
±0.23
±2.4
±0.09(SCOD)
99% CI
±0.11
±0.11(SCOD)
±0.30
±3.3
±0.12(SCOD)
TABLE A-9. CONFIDENCE LEVELS FOR DIFFERENCES INVOLVING SONOC AND SCOD
'oc
Analysis
95% CI
99% CI
Kjeldahl SON (< 1.3 mg/1)
Technicon SON
SCOD (< 25 mg/1)
±0.10 mg/1
±0.28 mg
±2.9 mg/1
±0.14 mg/1
±0.37 mg/1
±3.9 mg/1
STATISTICAL METHODS USED FOR MOLECULAR WEIGHT DISTRIBUTION STUDIES
1. Negligible error in measuring volumes was assumed.
2. Values for organic mass standard deviations for concentrated samples
were calculated from:
m
(s) (V_)
V
where
s = mass standard deviation, pg or mg;
V = volume of original sample concentrated, liters;
V = volume applied to Sephadex column, ml;
3.
246
-------
V = volume of concentrate, ml; and
s = standard deviation of analysis, mg/1.
Values used for s are listed below (from Statistical Methods used
in Section 8).
Kjeldahl SON -• raw wastewater s = ± 0.17 mg/1
primary effluent s = ± 0.16 mg/1
AS effluent s = ± 0.03 mg/1
AS Culture 2 eff. s = ± 0.06 mg/1
SCOD - raw wastewater s = ± 2.8 mg/1
primary effluent s = ± 2.7 mg/1
AS effluent s = ± 0.8 mg/1
AS Culture 2 eff. s = ± 1.3 mg/1
Using these values for s , appropriate values for s are:
Sample s (SON) s (SCOD)
m m
Raw wastewater ± 17 yg ± .28 mg
Primary effluent ± 16 yg ± .27 mg
AS effluent ± 7 yg ± .18 mg
AS Culture 2 effluent ± 11 yg ± .23 mg
Estimates of 95 percent confidence limits for mass measurements were
made from the following:
y = X + t^S
m — /-
vn
where
y = estimated population mean, yg or mg;
X = mass average of replicate samples, yg or mg;
m
s = as determined in 2;
m
n = 2; and
t = t statistic.
For each wastewater studied:
247
-------
4.
5.
Sample
Raw wastewater
Primary effluent
AS effluent
AS Culture 2 effluent
SON (yg)
X ± 23
X ± 22
X ± 10
X ± 16
SCOP (mg)
X ± 0.41
X ± 0.40
X ± 0.34
X ± 0.34
These confidence limits were reported in Table 82 and were used to
calculate confidence limits for percent recoveries.
The following procedure was used to calculate 95 percent confidence
limits for SON and SCOD mass contained in the five MW fractions
(Tables 83 and 84).
The SON and SCOD of the eluant were subtracted from the meas-
ured SON and SCOD of each MW fraction. The standard deviation of
such a difference is the square root of the sum of the squares of
the individual standard deviations [218], and for the Technicon
SON and low-level SCOD analyses are 0.11 and 1.1 mg/1, respectively.
Values for 95 percent confidence limits were then estimated from
the following
In which
y
estimated population mean, yg or mg;
Vf = volume collected for a particular MW fraction, liter;
average SON or SCOD concentration for a particular MW
fraction, mg/1;
t = t statistic for SON or SCOD analysis;
s, = standard deviation as described above, mg/1; and
n = 2 (replicate analyses) .
For Technicon SON and low-level SCOD analyses,
A i¥* i L1W
Confidence limits (95 percent level) for percent recoveries were
estimated using [186]:
IR
R
a_B
B
where
R = A/B;
3™ = standard deviation of the ratio A/B;
248
-------
s.,8^ = standard deviations for A and B, respectively
A. J5
B = initial concentration (SON or SCOD, mg/1); and
A = final concentration (SON or SCOD, mg/1).
To estimate percent recoveries from original sample or from organic
mass applied to the Sephadex column, sg was calculated as in 3 for
a 95 percent level of significance. When estimating total recovery
through the Sephadex column, s. was calculated by summing values
from 4 as shown below [218]:
SA =±
t!a
2
2
These values are reported in Tables 83 and 84.
For calculating 95 percent confidence limits for data presented
gures 34 and 37, values for s were those calculated in (4).
&
The estimated 95 percent confidence limits were then reported
as
A/B ± sp
K.
for the cases described above.
6. For differences in percent of original SON or SCOD mass contained
in two MW fractions (Figures 34 to 37) to be significant at a 95
percent confidence level, they must be greater than [218]
where s,j is the standard deviation of the difference, t is the
t statistic, and n is one. s
-------
SON tSON • SSON 2 fcSCOD • SSCOD 2
*
R - SCOD ^ . SQN £- . SCQD
where
tQAM,t = t-statistics for SON and SCOD;
n = number of analyses;
SOAM,S A = standard deviations for SON and SCOD analyses
SON SCOD
SON = measured SON;
SCOD = measured SCOD; and
SD = 95% CI.
K
b. For differences in SON/SCOD ratios to be significant, they must
be greater than:
2 .L 2
± SR1 +SR2
where SRI and s _ are the 95% CI values for the two ratios
being compared.
250
-------
APPENDIX B
NITRATE INTERFERENCE WITH KJELDAHL SON ANALYSIS
High levels of nitrate nitrogen were found to interfere with Kjeldahl
analysis for SON. The problem was discovered during a biodegradation study
with Palo Alto activated-sludge effluent in which SON decreased from 1.45
mg/1 to a value calculated to be less than zero after 50 days of degradation,
an unreasonable result based on previous biodegradation results. Preliminary
experiments indicated high nitrate concentrations caused the abnormal SON
decrease. The following is a brief discussion of factors found to affect the
interference, the mechanisms involved, and methods attempted to eliminate the
interference.
FACTORS AFFECTING THE INTERFERENCE
Tables B.I and B.2 contain data describing the magnitude in SON reduc-
tion caused by various concentrations of NO^-N. It appears the critical
N03~N concentration for interference is between 6 and 10 mg/1, and may vary
for different samples. The magnitude of the interference was non-reproducible
at high NO^-N values; SON recoveries ranged from 11-57 percent (5 samples) at
40 mg/1 NO^j-N. The presence and concentration of N02-N (0-40 mg/1), alkalin-
ity (0-111 mg/1 as CaC03), and Cl~ (0-2000 mg/1) did not produce nor prevent
the interference. Addition of reducing organics such as glucose was found
to "produce" SON by reducing NO^ to NH3 during digestion, the NH^-N is then
measured as SON; thus creation of a positive interference by certain organics
was also found to be possible.
MECHANISMS INVOLVED
The interference occurred during the digestion step, and involved a
reaction between NH3 and N03 resulting in the disappearance of all, or a
portion of the 1^3. A recent study by Schlueter[232] which was instigated
as a result of our finding has shown that the probable mechanism is as
follows:
NH+ + NO^ = N20 + 2H20
The nitrous oxide excapes in the digestion gas as confirmed by infrared
analysis of the collected gases [232].
251
-------
TABLE B.I. EFFECT OF NO^-N CONCENTRATION ON SON RECOVERY FROM A
BIOLOGICALLY TREATED WASTEWATER*
Added
NO^-N
(mg/1)
1.4
6.4
11.4
21.4
31.4
41.4
Measured
SON
(mg/D
0.55
0.53
0.42
0.41
0.23
0.19
Percent
SON
Recovery
100 (assumed)
96
76
75
51
35
*
Effluent from a symbiotic biological process (grass-bacteria)
treating agricultural return water; see REf . 76 for description.
TABLE B.2. EFFECT OF NOo-N CONCENTRATION ON SON RECOVERY FROM
PALO ALTO ACTIVATED-SLUDGE EFFLUENT
NOo-N
(mg/D
0 (assumed)
5
10
15
20
SON
(mg/1)
1.23
1.23
1.22
0.86
1.14
Percent
Recovery
100 (assumed)
100
99
70
93
METHODS ATTEMPTED TO ELIMINATE THE INTERFERENCE
A simple, rapid chemical method was sought, one that would quantita-
tively reduce the N03 to non-interfering NO^ or NH3, or, alternatively, a
modified digestion technique that would not produce the interference. Diges-
tion with catalysts other than HgO [3], with permaganate, hydrogen peroxide,
and mixtures of these compounds with sulfuric acid did not eliminate the
interference. This observation should have been expected since the inter-
ference reaction involved (conversion to ^0) is catalyzed by high-
temperature acid conditions similar to those used for the standard and
252
-------
modified Kjeldahl digestions [232]. A wide range of reducing compounds
(Table B.3) were tried and none eliminated the interference.
Schlueter [232] reported that removal of NO^ by anionic exchange resulted
in 100 percent recovery of the SON compound alanine. However, results pre-
sented in this research showed anionic resins remove a significant portion
of treated and untreated wastewater SON, and thus may have limited applica-
tion in eliminating the NO^ interference with these types of samples. Per-
haps a selective resin, removing NO^ but no wastewater SON, can be found.
This warrants fv" >.r study.
CONCLUSIONS
Nitrate nitrogen interferes with SON analysis by the Kjeldahl method,
resulting in decreased SON recovery. During this research, NO^-N was moni-
tored at all times, and SON values for samples containing NO~-N in excess
of 7 mg/1 were not reported, or reported as questionable.
TABLE B.3. EFFECT OF COMPOUNDS TRIED FOR ELIMINATION OF
NITRATE INTERFERENCE
Glucose +
Na Acetate o
Pthalic Acid +
Parafin +
Cd S +
FeS +
SeS2 +
S° +
Na2S203 +
Na2S06 +
Na2S03
NaAs02
Fe° +
Cd° +
Cu
FeCl2
CrCl3
+ increased SON recovery in presence of 40 mg/1 N03-N.
decreased SON recovery in presence of 40 mg/1 N03-N.
o no effect on SON recovery in presence of 40 mg/1 NOrj-N.
253
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO.
EPA-600/2-78-030
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
SOLUBLE ORGANIC NITROGEN CHARACTERISTICS AND REMOVAL
5. REPORT DATE
March 1978 (Issuing Date}.
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Stephen J. Randtke, Gene F. Parkin, John V. Keller,
James 0. Leckie, and Perry L. McCarty
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Stanford University
Stanford, California 94305
10. PROGRAM ELEMENT NO.
1BC611 - SOS #3 - Task 32
11. CONTRACT/GRANT NO.
R-804001
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory--Cin.,OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final - Sept. '74 - Feb. '77
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES
C. Mashni, Project Officer
U.S. EPA - Cincinnati, Ohio 45268
513/68^7684
16. ABSTRACT
This report discusses sources, concentrations, characteristics and methods
for removal of Soluble Organic Nitrogen (SON) in wastewater. Removal by various
physical, chemical and biological processes are described and molecular weight
distribution is characterized. A significant portion of the SON in secondary
effluent is produced biologically during treatment.
Chemical coagulation, ion exchange and activated carbon were used singly and
in combination to characterize different fractions of the SON and the Soluble
Chemical Oxygen Demand (SCOD).
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
wastewater
activated sludge
nitrification
chemical removal (sewage treatment)
nitrogen
b.lDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
soluble organic nitrogen,
soluble chemical oxygen
demand, molecular weight
distribution, effluent
standards, sorption
coagulation, biodegrada-
tion
13B
3. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
272
20.SECURITY.CLASS /Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
254
U. S. GOVERNMENT PRINTING OFFICE: 1978-757-140/6807 Region No. 5-11
------- |