PROCESS DESIGN MANUAL
FOR
SULFIDE CONTROL IN
SANITARY SEWERAGE SYSTEMS
U. S. ENVIRONMENTAL PROTECTION AGENCY
Technology Transfer
October 1974
-------
ACKNOWLEDGMENTS
This Design Manual was prepared for the Technology Transfer Office of the U. S. Environ-
mental Protection Agency by the firm of Pomeroy, Johnston and Bailey, under the direction
of Richard D. Pomeroy. Major U. S. EPA contributors and reviewers were J. M. Smith of
the U. S. EPA National Environmental Research Center, Cincinnati, Ohio, and D. J. Lussier
of Technology Transfer, Washington, D. C.
NOTICE
The mention of trade names or commercial products in this publication is for illustration
purposes and does not constitute endorsement or recommendation for use by the U. S.
Environmental Protection Agency.
u
-------
TABLE OF CONTENTS
CHAPTER PAGE
ACKNOWLEDGMENTS ii
TABLE OF CONTENTS Hi
LIST OF FIGURES v
LIST OF TABLES vii
FOREWORD ix
I INTRODUCTION
1.1 Background 1-1
1.2 Purpose 1-1
1.3 References 1-2
2 CHARACTERISTICS AND PROPERTIES OF
HYDROGEN SULFIDE
2.1 Forms of Sulfide in Waste waters 2-1
2.2 Physical-Chemical Properties of Hydrogen Sulfide 2-1
2.3 Odor of H2S 2-6
2.4 Toxicity of H2S 2-6
2.5 Analytical Methods 2-7
2.6 References 2-9
3 THE OCCURRENCE AND EFFECTS OF
SULFIDE IN SANITARY SEWERS
3.1 The Occurrence of Sulfide in Sanitary Sewers 3-1
3.2 Gains and Losses of Oxygen in Wastewater Streams 3-9
3.3 Losses of Sulfide 3-21
3.4 Other Effects of Velocity 3-24
3.5 Forecasting Sulfide Buildup 3-26
3.6 The Effects of Sulfide in Wastewaters 3-34
3.7 References 3-46
4 INVESTIGATIONS OF EXISTING SYSTEMS
4.1 Purpose 4-1
4.2 Examination of the Existing System 4-1
4.3 Flow Measurements 4-3
4.4 Character of the Wastewater 4-5
4.5 References 4-7
5 CONTROL OF SULFIDE IN EXISTING SYSTEMS
5.1 Improving the Oxygen Balance 5-1
5.2 Chemical Methods 5-30
5.3 Control of Industrial Wastes 5-41
5.4 References 5-43
-------
TABLE OF CONTENTS - Continued
CHAPTER PAGE
6 DESIGNING TO AVOID SULFIDE PROBLEMS
6.1 Basic Concept 6-1
6.2 Slopes of Small Collecting Sewers 6-1
6.3 Slopes of Larger Sewers 6-3
6.4 Sewer Sizes 6-6
6.5 Points of Turbulence 6-6
6.6 Pump Stations and Force Mains 6-8
6.7 Materials for Sewer Construction 6-11
6.8 Protection of Concrete and Asbestos-Cement
Pipe by Linings 6-15
6.9 References 6-16
iv
-------
LIST OF FIGURES
Figure No. Page
2-1 Proportions of H2S and HS~ in Dissolved Sulfide 2-5
2-2 Comparison of Total and Dissolved Sulfide
Concentrations in Trunk Sewers 2-8
3-1 Processes Occurring in Sewers with Sufficient
Oxygen to Prevent Sulfide from Entering the Stream 3-3
3-2 Processes Occurring in Sewers Under Sulfide
Buildup Conditions 3-5
3-3 Sulfide Concentrations at End of Force Main
in Long Beach 3-7
3-4 Specific Sulfide Flux Coeficients from Filled-Pipe Data 3-9
3-5 Reaeration Rates in Sewers Flowing Half Full 3-14
3-6 Relative Aeration Rates in a Sewer at Different Flows,
Compared to the Rate for the Half Filled Pipe 3-15
3-7 Oxygen Measuring Assembly 3-18
3-8 Typical Oxygen Reaction Curves in Wastewaters 3-19
3-9 Changing Oxygen Reaction Rates with Time in
Wastewater Samples Aerated in the Laboratory 3-20
3-10A Effect of Sulfide Concentration on Rate of
Sulfide Buildup or Decline 3-25
3-1 OB Sulfide Buildup Curve, Calculated from Figure 3-1OA 3-25
3-11 Solids Accumulations at Various Flow Velocities 3-27
3-12 Sulfide Occurrence in Small Sewers 3-30
3-13 Flow-Slope Relationships as Guides to Sulfide Forecasting 3-32
3-14 Unequal Distribution of Corrosion in a Concrete Sewer 3-37
3-15 Acid Attack of Sewer at Waterline 3-38
3-16 Effect of Velocity and Pipe Size on Sulfide Flux
to Pipe Wall Under Specified Conditions 3-40
-------
LIST OF FIGURES - Continued
Figure No. Page
3-17 Factor to Apply to 0SW from Figure 3-16 to
Calculate 0SW for Other than Half-Pipe Depth 3-42
3-18 Sections of Experimental Pipes After Exposure to
Septic Tank Effluent for Seven Years 3-43
4-1 Walking Through Large Sewer for Inspection 4-2
4-2 Palmer-Bowlus Slab-Type Weir for 18" Pipe 4-4
4-3 Typical 24-Hour Sulfide Test in a 60-Inch Trunk Sewer 4-6
5-1 Schematic Drawing of Force Main for Problem 5-7
5-2 H2$ Controlling Drop at the End of a Force Main 5-10
5-3 U-Tube Design, Jefferson Parish Station 5 5-11
5-4 Head Loss in Down Leg of 42-ft U-Tube 5-13
5-5 Main Trunks Relating to U-Tube Installation in
Jefferson Parish, La. 5-17
5-6 U-Tube Configuration, Pioneer Park Lift Station 5-19
5-7 Aeration Tank at Pump Station, Port Arthur, Texas 5-23
5-8 Two Types of Pressure Tanks for Dissolving Oxygen
and In-Sewer Application of Oxygen 5-26
5-9 Bayview Drive Force Main-Ft. Lauderdale System 5-39
6-1 LACSD Main Trunk Sewer System 6-5
6-2 General Profiles of Portions of Los Angeles
County Sanitation Districts System 6-7
6-3 Functional Drawing of a Pump Station with
an Air Bypass 6-9
6-4 Functional Drawing of an Air-Lift Pump Station 6-10
VI
-------
LIST OF TABLES
Table No. Page
2-1 Solubility of I^S in Water at a Pressure of
One Standard Atmosphere 2-3
2-2 Logarithmic Practical lonization Constants (pK') for
Hydrogen Sulfide 2-4
2-3 Proportions of Dissolved Sulfide Present as I-^S (j Factors) 2-6
3-1 Expected Oxygen Absorption in Waste water Falls 3-16
5-1 Suggested Oxygen Reaction Rates 5-4
5-2 Results of U-Tube Operation in Jefferson Parish, La. 5-16
5-3 Oxygen Dissolving Efficiency 5-20
5-4 Hydrogen Peroxide Treatments at Ft. Lauderdale, Florida 5-38
5-5 Effects of Hydrogen Peroxide in Bluff Cove Force Main 5-40
6-1 Sulfide Concentrations in Small Collecting Sewers 6-2
vn
-------
FOREWORD
The formation of the United States Environmental Protection Agency marked a new era
of environmental awareness in America. This Agency's goals are national in scope and
encompass broad responsibility in the area of air and water pollution, solid wastes,
pesticides, and radiation. A vital part of EPA's national water pollution control effort is
the constant development and dissemination of new technology for wastewater treatment
systems.
It is now clear that only the most effective design and operation of wastewater treatment
systems, using the latest available techniques, will be adequate to meet the future water
quality objectives and to ensure continued protection of the nation's waters. It is essential
that this new technology be incorporated into the contemporary design of waste treatment
systems to achieve maximum benefit of our pollution control expenditures.
The purpose of this manual is to provide the engineering community and related industry a
new source of information to be used in the control of corrosion and noxious conditions
resulting from hydrogen sulfide in existing sewerage systems, and in the development of
designs for new systems so as to keep them free from these problems.
Much of the information presented is based upon research sponsored by the U. S. EPA,
including investigations of the occurrence and effects of sulfide in existing sewerage systems,
and the performance of equipment and procedures in actual use for sulfide control. The
performance data given should be used as a guide and should be tempered with sound
engineering judgment based on a complete analysis of the specific application.
This manual is one of several available through the Technology Transfer Office of EPA
to describe recent technological advances and new information. Future editions of this
manual will be issued as warranted by advancing state-of-the-art to include new data as
it becomes available, and to revise design criteria as additional full-scale operational
information is generated.
ix
-------
CHAPTER 1
INTRODUCTION
1.1 Background
One characteristic by which sanitary sewage is known to the public is its potential for
creating odor nuisances. Sometimes it is the odors escaping from sewer manholes that
cause complaints; more commonly, the source is a wastewater treatment plant. Yet there
are wastewater treatment plants that are free from this stigma. Techniques to prevent
odor nuisances are available, and if there is a commitment to construct odor-free sewage
works, it can be done.
The main cause of odors in sewerage systems is hydrogen sulfide, or F^S, a gas detectable
in very low concentration. Hydrogen sulfide is also noted for its toxicity and for its ability
to cause corrosion of various materials used in sewer construction.
Much research has been done on various aspects of the sulfide problem in the last three
decades. Beside extensive studies in the United States, important contributions have been
made by engineers in Australia under the leadership of a Standing Committee of the
principal sewerage authorities. In South Africa, research has been done on the corrodability
of concrete under conditions that may develop in sewers (1). The first attempt at a
comprehensive treatise on the sulfide problem was published as a joint effort of the
engineers in Australia (2). Despite the completion of that work, there is more that can be
added on the basis of results of recent research in the United States.
1.2 Purpose
A comprehensive design manual is necessary to bring together the information now available
into a form convenient for use by engineers who are designing sewers or who are faced with
the need to apply sulfide control procedures in existing sewers. This Manual is intended to
satisfy that need. It is based principally upon sources of information in the cited references.
Some of the data, however, including that developed by research and development projects
of the U.S. EPA, have not yet been published elsewhere, and some of the deductions were
developed from published and unpublished data in the course of the preparation of the
Manual.
Sulfide control is now a well developed technology. Continuing advances in basic knowledge
and in the development of control procedures are to be expected, but application of the art
in its present state, as set forth in this Manual, will overcome sulfide-producing tendencies
in existing systems and provide an understanding of how systems can be designed to
minimize such problems in the future.
1-1
-------
1.3 References
1. Stutterheim, N., and Van Aardt, J. H. P., Corrosion of Concrete Sewers and Some
Remedies, South African Industrial Chemistry, No. 10, (1953).
2. Thistlethwayte, D. K. B., Control of Sulphides in Sewerage Systems, Butterworths
Pty. Ltd., Melbourne, Australia (1972), and Ann Arbor Science Publishers Inc.,
Ann Arbor, Michigan (1972).
1-2
-------
CHAPTER 2
CHARACTERISTICS AND PROPERTIES OF HYDROGEN SULFIDE
2.1 Forms of Sulfide in Wastewaters
The term sulfide refers to inorganic sulfur in its most reduced state, that is, with a valence
of minus two. In wastewaters, sulfide is a mixture of:
1. Insoluble metallic sulfides. There is evidence that wastewaters may contain
several iron sulfides: pyrrhotite (variable from FeS to Fe4S5), smythite ^6384),
and pyrite or marcasite (FeS^). In addition, small amounts of sulfides of zinc,
copper, lead, cadmium, and other metals may be present insofar as the wastewater
contains any such metals.
2. Dissolved sulfide. This is a mixture of ^S and HS (read HS ion) existing in
equilibrium with hydrogen ions, as shown by the following equation:
H2S :*jc: HS~ + H+
The secondary sulfide ion, S~, does not exist as a significant fraction of the
sulfide content except at very high pH. For example, less than 0.05 percent
of the dissolved sulfide is present as S~ at pH 11, and less than 0.5 percent is
present at pH 12. The proportion of ^S as a function of pH is discussed in the
next section.
There are also organic sulfur compounds in which the sulfur is arbitrarily assigned a
valence of minus two, and which may be called organic sulfides. They do not respond to
the analytical tests used to measure inorganic sulfide, and they do not have the same
significance. The volatile organic sulfur compounds, however, are very important odor
components in wastewater. They are principally of three types: the thiols, also called
mercaptans, containing an ~SH group (example, methanethiol, CH^-SH); the thioethers,
also called sulfides, in which two organic radicals are attached to sulfur (example, di-
me thylthioether, CHg-S-CHg); and the disulfides, with two organic radicals joined to a
pair of sulfur atoms (example, dimethyl disulfide, CHo-S-S-CH-j). There are also
nonvolatile sulfur compounds that cause none of the problems associated with the
volatile compounds, but that may break down by biological action to yield inorganic
sulfide. Naturally occurring compounds of this type in wastewater are principally the
albuminoid proteins.
2.2 Physical-Chemical Properties of Hydrogen Sulfide
Hydrogen sulfide is normally a gas, liquefying at -62°C under one atmosphere pressure,
2-1
-------
or at 25°C under a pressure of 20 atmospheres. At 25°C and one atmosphere pressure, a
liter of t^S weighs 1.40 grams, of which 1.31 grams are sulfur. It is moderately soluble in
water, more so than carbon dioxide and most other common gases. Table 2-1 shows the
solubility, in milligrams per liter of sulfide, at one standard atmosphere pressure, and also
the partial pressure of F^S over a solution of 1 mg/1 of HoS (as S), expressed in millionths
of an atmosphere, which is the same as parts per million of H^S by volume in the air when
the total pressure is one standard atmosphere. At a high altitude, the total pressure is
diminished, but the partial pressure of H9S over water with 1 mg/1 of HoS remains the same.
Z- ^
Consequently the proportions of H^S in the atmosphere expressed as ppm by volume varies
inversely as the atmospheric pressure.
The proportions of H^S and HS~~ in the dissolved sulfide fraction in water are primarily a
function of pH, and can be calculated most readily from the ionization constant written
in the logarithmic form:
log J-5§_1 = pH - PK'
[H2S ]
where pK' is the negative logarithm of the practical ionization constant as written in the
arithmetic form. The chemical formulas appearing in brackets are here used to signify
concentrations, which in this case mean concentrations expressed as S.
The value of pK' is influenced by temperature and by the ionic strength of the solution.
Sometimes ionic strength is estimated from the "total dissolved solids" or "filterable
residue," but in wastewaters the relationship is not very reliable. Wastewaters may contain
un-ionized solutes. Also ammonium salts may be present as ions, which are lost when the
sample is evaporated. This is especially true in the liquor of digested sludge. A better
approximation is obtained from the specific electric conductance. The specific electrical
conductance at 25°C, expressed as micromhos/cm, multiplied by 1.35x10 gives a fair
approximation of the effective ionic strength, expressed in units of moles per liter.
Table 2-2 shows values of pK' at various concentrations of salts as represented by con-
ductance, based upon the best available data (2). An all-purpose value of pK' for rough
estimates is 7.0.
2-2
-------
TABLE 2-1
SOLUBILITY OF H2S IN WATER AT A PRESSURE OF ONE STANDARD ATMOSPHERE (1)
Temp.
°C
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
2Q
21
22
23
24
25
26
27
28
29
30
35
40
Solubility
milligrams
per liter
expressed as S
6648
6434
6227
6028
5834
5646
5465
5291
5124
4964
4810
4667
4529
4398
4271
4150
4033
3922
3816
3714
3618
3523
3432
3344
3258
3175
3095
3018
2945
2874
2806
2491
2221
in atmosphere in equilibrium
with a solution of 1 mg/1 as S
partial pressure,
millionths of an
atmosphere
150
155
160
166
171
177
183
189
195
201
208
214
221
227
234
241
248
255
262
269
276
284
291
299
307
315
323
331
340
348
356
401
450
milligrams
per liter of
S in the air
0.214
0.221
0.228
0.235
0.242
0.249
0.256
0.263
0.271
0.279
0.287
0.294
0.302
0.310
0.318
0.326
0.334
0.343
0.351
0.359
0.368
0.376
0.385
0.394
0.403
0.412
0.421
0.430
0.440
0.449
0.459
0.508
0.560
2-3
-------
TABLE 2-2
LOGARITHMIC PRACTICAL IONIZATION CONSTANTS (pK') FOR HYDROGEN SULFIDE
Specific „ 0_
electrical Temperature, °C
conductance
at 25°C,
micromhos
per cm 10 15 20 25 30 35 40
0
25
100
200
400
700
1,200
2,000
3,000
4,000
5,200
7,200
10,000
14,000
22,000
50,000*
*Approximates sea water
Table 2-3 shows the proportion of dissolved sulfide existing as ^S (a ratio indicated herein
as the j factor) as a function of pH - pK', or as a function of pH if pK' = 7.0. Figure 2-1
shows the information graphically.
The rate at which ^S can escape from solution into the air under any given conditions of
exposure is proportional to the ^S concentration. Thus, at a pH of 7.0, ^S will escape
about half as fast as from a strongly acid solution having the same dissolved sulfide content.
At pH 9.0, it will escape only 1 percent as fast as from an acid solution. If part of the
H2S escapes, the remaining dissolved sulfide will be divided between HS~ and ^S in the
same ratio as before, because the equilibrium re-establishes itself almost instantly.
7.24
7.23
7.22
7.21
7.20
7.19
7.18
7.17
7.16
7.15
7.14
7.13
7.12
7.11
7.10
7.09
7.17
7.16
7.15
7.14
7.13
7.12
7.11
7.10
7.09
7.08
7.07
7.06
7.05
7.04
7.03
7.02
7.10
7.09
7.08
7.07
7.06
7.05
7.04
7.03
7.02
7.01
7.00
6.99
6.98
6.97
6.96
6.95
7.03
7.02
7.01
7.00
6.99
6.98
6.97
6.96
6.95
6.94
6.93
6.92
6.91
6.90
6.89
6.88
6.96
6.95
6.94
6.93
6.92
6.91
6.90
6.89
6.88
6.87
6.86
6.85
6.84
6.83
6.82
6.81
6.89
6.88
6.87
6.86
6.85
6.84
6.83
6.82
6.81
6.80
6.79
6.78
6.77
6.76
6.75
6.74
6.82
6.81
6.80
6.79
6.78
6.77
6.76
6.75
6.74
6.73
6.72
6.71
6.70
6.69
6.68
6.67
2-4
-------
FIGURE 2-1
PROPORTIONS OF HoS AND HS~ IN DISSOLVED SULFIDE
pH if PK = 7.0
80
vt
o
ui
o
(t
ui
Q.
p H — p K
k,«
2-5
-------
TABLE 2-3
PROPORTIONS OF DISSOLVED SULFIDE PRESENT AS H9S 0' Factors)
pH-pK
-2.0
-1.8
-1.6
-1.4
-1.2
-1.0
-0.9
-0.8
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0.0
+0.1
+0.2
+0.3
+0.4
pHif
pK = 7.0
5.0
5.2
5.4
5.6
5.8
6.0
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
7.0
7.1
7.2
7.3
7.4
Proportion
of H7S
0.99
0.98
0.975
0.96
0.94
0.91
0.89
0.86
0.83
0.80
0.76
0.72
0.67
0.61
0.56
0.50
0.44
0.39
0.33
0.28
pH - pK
+0.5
+0.6
+0.7
+0.8
+0.9
+ 1.0
+ 1.1
+ 1.2
+ 1.3
+ 1.4
+1.5
+ 1.6
+ 1.7
+1.8
+ 1.9
+2.0
+2.5
+3.0
pHif
pK = 7.0
7.5
7.6
7.7
7.8
7.9
8.0
8.1
8.2
8.3
8.4
8.5
8.6
8.7
9.0
9.5
10.0
Proportion
of H7S
0.24
0.20
0.17
0.14
0.11
0.091
0.074
0.059
0.048
0.039
0.031
0.025
0.020
0.016
0.013
0.010
0.003
0.001
2.3 Odor of H2S
t^S occurs naturally in "sulfur springs," and it may occur in decaying organic matter,
particularly in rotten eggs. Its unpleasant odor is therefore well known. It has been
shown that the threshold concentration in water for detection by humans is between 0.01
and 0.1 ng/1 (0.00001 and 0.0001 mg/1) (3).
2.4 Toxicity of H2S
Because F^S is of such common occurrence, and causes no noticeable physiological
effects when present in low concentrations, its lethal character has been largely ignored.
By comparison, hydrocyanic acid, HCN, which also occurs naturally but only in very
small amounts, is looked upon as a model of a deadly poison. No direct comparisons
2-6
-------
have been made of the toxicity of H2S and HCN, but they are certainly of the same order
of magnitude. Death appears to have resulted from an H2S concentration of 0.03 percent
(300 ppm) in the air (4).
Hydrogen sulfide is treacherous, because the ability to sense it by smell is quickly lost. If a
person ignores first notice of the gas, his senses will give him no further warning. If the
concentration is high enough, unconsciousness will come suddenly, followed by death if
there is not a prompt rescue. The deadliness of H2S is well known in industries where
it occurs, as in the petroleum industry, where high-sulfur natural gas and H2S produced in
refining operations are serious hazards. Deaths have also resulted from H2S produced in
swamps, from natural gas seeps, and from bathing in the water from hot springs in
unventilated rooms. Many workmen in sewers have died from poisoning by H2S(5).
2.5 Analytical Methods.
Analytically, total sulfide and dissolved sulfide must be distinguished. Dissolved sulfide is
determined either by separating suspended matter with the aid of flocculating chemicals and
determining sulfide in the remaining solution, or by a potentiometric titration with lead
perchlorate, using an Ag—Ag2S electrode. Estimates of dissolved sulfide can also be made
from the potential of the Ag—Ag2S electrode and the pH, although the results by this
method are of a lower order of accuracy.
The most common sulfide test made in wastewaters is for total sulfide, using the standard
methylene blue method (6). Acid is used in the test, dissolving the metallic sulfides except
Cu2S and Ag2S, which are so inert that they are not included. When FeS2 dissolves, only
one of the two sulfur atoms appears as sulfide; the other separates as free sulfur. Three of
the four sulfur atoms in Fe^S^ appear as sulfide. Knowledge of the dissolved sulfide content
of wastewater is generally of more interest than total sulfide, but if the colorimetric test is
used, it is easier to determine the total than to determine dissolved sulfide. Therefore, the
total sulfide test is most often used in routine testing.
The insoluble fraction, which is the difference between total sulfide and dissolved sulfide,
«F
varies from one wastewater to another, largely in response to differences in the amounts of
metals carried by the sewage. A collection of analytical data from 20 sewers in the
Los Angeles County Sanitation Districts' system is shown in Figure 2-2, in which dissolved
sulfide is plotted against total sulfide. In most cases, sodium sulfide had been added
upstream from the sampling points in the course of research on sulfide buildup and decline.
Many of the points in the figure came from wastewaters affected by metal-bearing waste,
most notably the points from sewer 24C, distinguished by open circles. The dashed line
represents the typical comparison of total and dissolved sulfide in domestic wastewater.
2-7
-------
FIGURE 2-2
COMPARISON OF TOTAL AND DISSOLVED SULFIDE CONCENTRATIONS
IN TRUNK SEWERS, Na2S HAD BEEN ADDED IN MOST CASES
3.0
2.0
CC
ui
cc
m
CL
05
<
CC
o
ai
a
GO
2 1-0
0
INSOLUBLE
SULFIDE
AVERAGE RELATIONSHIP
FOR DOMESTIC SEWAGES
CIRCLES REPRESENT
DATA FROM SEWER 24C
1.0 2.Q
TOTAL SULFIDE, MILLIGRAMS PER LITER
2-8
-------
It is customary, in water and wastewater technology, to express the concentrations of the
various species of inorganic sulfide in terms of the sulfur content. For example, if it is
reported that a water contains 5 mg/1 of un-ionized i^S, it is understood to mean that
5 mg/1 of sulfur is present in the water in the form of I^S. The actual amount of H^S
would be 5.3 mg/1.
2.6 References
1. Handbook of Chemistry and Physics. Cleveland, Ohio: Chemical Rubber Publishing
Company (1955).
2. Ellis, A. J., and Goldring, R. M., Spectrophotometrie Determination of the Acid
Dissociation Constants of Hydrogen Sulfide, Journal Chemical Society of London,
pp. 127-130 (1959).
3. Pomeroy, R. D., and Cruse, H., Hydrogen Sulfide Odor Threshold. Journal of the
American Water Works Association, JH, No. 12, p. 677 (1969).
4. McDonald, John M., and Mclntosh, A. P., Arch. Ind. Hyg. Occupational Med. 3,
pp. 445-447 (1951).
5. Bowlus, F. D., and Pomeroy, R. D., "Stay Alive in That Sewer," The American City,
No. 5, pp. 123-125 (May, 1959).
6. Standards Methods for the Examination of Water and Wastewater. New York:
American Public Health Association. 13th Edition (1971).
2-9
-------
CHAPTER 3
THE OCCURRENCE AND EFFECTS OF SULFIDE IN SANITARY SEWERS
3.1 The occurrence of Sulfide in Sanitary Sewers
3.1.1 Sources of Sulfide
Sulfide is sometimes present in wastewaters added to sewers, particularly in certain industrial
wastes, and in rare instances groundwaters with high sulfide concentrations have leaked into
sewers, but the commonest source of sulfide is biological activity in the sewer. To some
extent this comes about by the decomposition of sulfur-containing organic matter, particu-
larly the albuminoid proteins. In domestic wastewater, the major amount of sulfide results
from the reduction of inorganic sulfur compounds. The word reduction is here used in its
chemical sense, meaning the opposite of oxidation. An element is reduced when oxygen
bonded to it is removed, or when hydrogen is added or, in general, when the valence state
of the element becomes less positive.
The principal sulfur compound in wastewaters is sulfate, existing in solution as the SO^ ion.
Where organic matter is present and oxygen is absent, bacteria of the species Desulfovibrio
desulfuricans (also called Desulfatomaculum desulfuricans in a recent reclassification) will
reduce sulfate to sulfide, using the oxygen to oxidize organic matter. Letting C be a virtual
representative of organic matter, the reaction can be written as follows:
804 + 2 C + 2 H20 bacteria 2 HCO J + H2S
In this reaction, 96 grams of sulfate make available 64 grams of oxygen, leaving 32 grams of
sulfide. Carbon in organic matter is in a more reduced form than the free element, because
of the attachment of hydrogen. An equation showing an average formula for the organic
matter reacting would probably indicate that about 42 grams of organic matter would be
oxidized. Sulfite, thiosulfate, free sulfur, and other inorganic sulfur compounds sometimes
found in wastewaters can be similarly reduced to sulfide.
Sulfate, organic matter, and bacteria capable of bringing about the reaction that produces
sulfide are present in virtually all sewers. Despite the presence of these essential elements,
sulfide is not produced in all wastewater systems. In fact, severe sulfide conditions occur
infrequently. A major objective of sulfide research has been to discover the reasons why
sulfide appears in some sewers and not in others.
An explanation of the reasons for the observed patterns of sulfide occurrences requires an
understanding of the mechanism of sulfide generation.
3.1.2 The Mechanism of Sulfide Generation
3-1
-------
For sulfate to be reduced to sulfide requires a medium completely devoid of free oxygen of
other active oxidizing agent. The stream of wastewater in a partly filled sewer is not
completely anaerobic because it is exposed to the sewer atmosphere. Oxygen absorbed at
the surface of the stream generally reacts quite rapidly, and in large sewers its concentration
is held to a very low level, yet enough is present to prevent sulfate reduction in the stream.
The place where strictly anaerobic conditions can develop is in the slime layer that forms
on the submerged pipe wall. This layer is a matrix of filamentous microbes and gelatinous
material (zoogleae) embedding various smaller bacteria. A typical slime layer may be
considered to be 0.04-in thick, but if the velocity of the stream is high, it may be no more
than 0.01-in thick, and if abrasive material is carried by the water, the wall may be scoured
clean. At low velocities the slime may be as much as Vg -in thick, or even more.
If oxygen is present in the stream, it diffuses into the slime layer, but the aerobic bacteria
found there use it so rapidly that it advances only a very short distance. Except where
oxygen concentrations are high, the aerobic zone is less than 0.01-in thick. Beneath that
the slime layer is anaerobic, and it is there that sulfide generation occurs.
Calculations based upon diffusion coefficients and observed rates of sulfide production show
that sulfate and/or the organic nutrients available to the sulfate-reducing bacteria are used
up in a very short distance, and that the thickness of the sulfide-producing zone is generally
of the order of 0.01 in. At deeper levels the slime layer is anaerobic but largely inactive
because of the lack of a nutrient supply.
As long as the surface of the slime layer is aerobic, sulfide diffusing out of the anaerobic
zone will be oxidized there. No sulfide will be found in the stream unless it is from some
extraneous or upstream source.
Figure 3-1 illustrates the processes just described, with the slime layer depicted on a greatly
magnified scale.
It is the oxygen supply, rather than the kinds of bacteria present, that determines where the
aerobic zone ends and the anaerobic zone begins. The oxygen concentration in the stream,
as well as temperature and the concentration of organic food materials, determine how deep
the oxygen will penetrate. If the oxygen supply increases, the aerobic zone will become
thicker. Aerobes are present in the anaerobic zone, ready to oxidize the organic nutrients,
but they remain dormant as long as oxygen is absent. The anaerobes, on the other hand,
cannot carry on their processes as long as oxygen is present, but they are ready to resume
their activity when oxygen disappears. Since these determining factors are not constant, the
boundaries between the zones are not fixed. Furthermore, bits of the slime are continually
sloughing off into the stream, adding to the variability of the zone boundaries.
If the oxygen concentration in the stream drops to a low level, generally a few tenths of a
mg/1, not enough oxygen will reach the surface of the slime layer to oxidize all of the
3-2
-------
FIGURE 3-1
PROCESSES OCCURRING IN SEWERS WITH SUFFICIENT OXYGEN
TO PREVENT SULFIDE FROM ENTERING THE STREAM
AIR
OXYGEN ENTERING THE WATER
WASTEWATER
DISSOLVED OXYGEN GREATER THAN I MG/L
DISSOLVED SULFIDE ZERO OR TRACE
DIFFUSION OF 02 AND NUTRIENTS
DIFFUSION OF S04 AND NUTRIENTS,
PRODUCTION OF SULFIDE
DIFFUSION AND OXYDATION OF SULFIDE
MAGNIFIED,
TYPICALLY
0.04 INCH
v PIPE \
\ WALL X
3-3
-------
sulfide that is produced. Sulfide can then escape into the stream. The process is illustrated
in Figure 3-2.
The oxygen concentration that is critical for preventing sulfide access to the stream is
generally in the range of 0.1 to 1.0 mg/1. It depends, among other things, on temperature
and on the turbulence of the stream. If the water is stationary or moving slowly, oxygen
becomes depleted near the pipe wall and sulfide may escape from the slime layer even when
the main bulk of the wastewater contains several milligrams per liter of oxygen. If there are
organic solids slowly rolling along the bottom of the pipe, they will release sulfide even when
the stream has a high oxygen content. When the oxygen concentration is low enough so
that sulfide leaks into the stream, it may nevertheless be only a minor part of it that gets
through. Completely anaerobic conditions must be approached before all of the sulfide
produced can pass into the stream.
3.1.3 Rate of Sulfide Production by the Slime Layer
The rate at which sulfide can be produced by a slime layer is generally determined by the
rate that the reactants, that is, sulfate and organic nutrients, can reach the sulfate-reducing
bacteria. When the slime contains a maximum population of these bacteria and when their
metabolic rate is high because of favorable temperature and other conditions, the reactants
do not have far to diffuse. The rate of sulfide generation is maximal. By contrast, if the
population is sparse or the metabolic rate is lower, the reactants must diffuse farther.
Consequently sulfide generation is at a slower rate. If the slime layer is thin, the reactants
may reach the pipe wall itself without being used up.
As noted in Section 3.1.1, sulfate and organic matter will be used by the sulfate reducers in
a ratio of about 96 to 42. It is not likely that both will reach the reactive zone in the ideal
proportions. One will be used up before the other, and the relatively scarce one will then
be the constituent that limits the rate of sulfide production. If there is an abundant supply
of organic matter but sulfate is scarce, then sulfide generation will be proportional to the
sulfate concentration and will be independent of the amount of organic matter. On the
other hand, if sulfate is abundant, the rate of generation will be proportional to the organic
nutrients.
From a consideration of probable diffusion rates and relative amounts of diffusable
reactants, as well as such experimental evidence as is available (1), it seems likely that the
concentration at which sulfate will cease to be limiting in most wastewaters will be in the
range of 20 to 100 mg/1. A range of organic materials with varying diffusion rates may be
used by the sulfate-reducing bacteria, so there may be a range of conditions in which the
concentrations of both sulfate and organic nutrients influence the sulfide generation rate.
A number of reports (2) have shown the oxidation of specific organic nutrients by sulfate-
reducing bacteria in synthetic media, but little is known about the nutrients utilized during
the production of sulfide by sewer slimes. It has been assumed (1) that these nutrients
are proportional to the standard BOD in most wastewaters, and that the rate of production
3-4
-------
FIGURE 3-2
PROCESSES OCCURRING IN SEWERS UNDER SULFIDE BUILDUP CONDITIONS
TRANSFER OF H2S TO PIPE WALL,OXIDATION TO
H2S+202 -*• H2S04
AIR
H,S ENTERING THE AIR
^ t
OXYGEN ENTERING THE WATER
WASTEWATER
DISSOLVEP OXYGEN LESS THAN O.I MQ/L
DISSOLVED SULFIDE PRESENT, HS'+ H2S
OXIDATION OF SULFIDE:
2 02+ 2 HS" -*• S20| 4 H20
DEPLETION OF 02 IN THE LAMINAR LAYER
DIFFUSION OF S04 AND NUTRIENTS,
PRODUCTION OF SULFIDE
DIFFUSION OF SULFIDE INTO THE STREAM
MAGNIFIED,
TYPICALLY
0.04 INCH
PIPE
WALL \
3-5
-------
of sulfide by the slimes is proportional to the BOD if the sulfate concentration is adequate.
The assumption cannot be extended to industrial wastewaters having different spectra of
organic materials.
An alternative postulate (3) is that the rate of sulfide production varies as [BOD] -° and as
[804] . It is not likely that these relationships hold when either reactant greatly over-
balances the other.
The effect of temperature on the rate of sulfide production by a slime layer is complex. The
increased metabolic rate of the bacteria reduces the distance that the reactants need to
diffuse. At the same time, the diffusion coefficient increases. It appears that the over-all
effect is about 7 percent/deg C. The term "effective BOD" has been proposed (1) as a
convenient way to combine the temperature and BOD effects, as follows:
[EBOD] = [BOD] x(i.ov) (T-2°)
where:
[EBOD] = effective BOD, mg/1
[BOD] = standard BOD, mg/1
T = temperature, deg C
(1.07) = empirical factor
The effect of temperature on sulfide buildup is illustrated in Figure 3-3, which shows
results of sulfide tests at the end of a force main that formerly transported wastewater in
Long Beach, California. One test was made each week. With the array of 130 points, it is
seen that there is an annual cycle, corresponding to the temperature cycle. (There is also a
long-time downward trend, caused by increasing flows and consequent shorter detention
time in the force main.)
There are also daily cycles resulting from patterns of flow velocities, detention times, and
wastewater characteristics. In the minimum-size collecting sewers and the small trunks,
sulfide generation, if it does take place, is generally not a 24-hour-a-day occurrence.
During the early morning hours, when the sewage is weak, there is little likelihood of
sulfide buildup in these smaller sewers.
The rate of sulfide production by a slime layer can be measured as a sulfide flux, 0se, in
units of grams/m -hr. Assuming that the flux is proportional to the effective BOD, and
that the rate is not restricted by a scarcity of sulfate, an equation can be written to define
a "specific sulfide flux coefficient," M:
3-6
-------
FIGURE 3-3
SULFIDE CONCENTRATIONS AT END
OF FORCE MAIN IN LONG BEACH
«».u
5
* 3.0
O:
O
5 2-0
(O
< 1.0
O
K '
0
/!
«rf
•
<
V
"•
\
."
*
\
\
0
1
A S 0 N D
1946
f rV_
ml
~S>
4
A&
,
V
V
'
•j
y
• •
^>
i 4
41
•^
1
-------
Measurements in force mains and other filled pipes permit the obtaining of M values which
show the maximum sulfide-producing capability of the slime layer. For the full sulfide-
generating capability to be displayed, it is necessary that there be an ample sulfate con-
centration, no biologically inhibiting condition in the wastewater, and complete absence of
oxygen. This last condition is often not met, especially at the beginning of the force main.
Furthermore, the water must not be static for extended periods, because there may be local
depletion of nutrients near the slime layer. Where the water is completely anaerobic, a
significant amount of sulfide may be generated in the stream as well as in the slime layer.
Measurements of sulfide buildup have been made in about 40 force mains and other filled
pipes. Most of the findings have been published (4) (5) (6). A correction factor was
estimated for the effect of sulfide generation in the stream, and the following equation was
developed:
^^ = 3.28 M [EBOD] (1 + 0.48 r) r"1
where:
(1 + 0.48r) = empirical factor for sulfide production in the stream.
This equation has been published (4), although expressed in a different way. Inherent in the
form of the equation is the assumption that generation in the stream and in the slime are
both proportional to EBOD.
The distribution of observed values of M from the pressure main data is shown in Figure 3-4.
The highest 25 percent of the results are between 0.75x10~^ and 1.30x10""^ m/hr, except
for two high values found in iron pipes in which the wastewater carried a considerable
percentage of seawater, and where anaerobic corrosion of iron was in progress, producing a
very rough interior. It appears that M = 1.0x10""^ m/hr is a reasonable value to use for
forecasting sulfide buildup in a normal force main when all conditions are favorable for
sulfide buildup.
The factor most often causing low results is the presence of oxygen in the water entering the
main. Sulfate deficiency, a poorly developed slime layer, or long static periods are probable
reasons for low results in some cases. Sulfate concentrations are known to have been high in
many of the mains where the tests were made, but actual sulfate concentrations were not
recorded in most cases.
Sulfide flux coefficients have been determined in partly filled pipes by measuring sulfide
buildup in the presence of enough zinc sulfate to suppress the oxidation of sulfide and to
prevent its loss to the atmosphere as t^S. Also, apparent sulfide flux coefficients have been
determined by measuring sulfide buildup or decline rates in the presence of varying amounts
of sulfide, and extrapolating to zero sulfide concentrations (Figure 3-10A is an example).
3-8
-------
FIGURE 3-4
SPECIFIC SULFIDE FLUX COEFFICIENTS
FROM FILLED-PIPE DATA
CO
5 I0
CO
UJ
cc
UJ 5
m
2£^
O> 00 f«- O "* O
d d d - - cj
RANGE OF M VALUES X |Q3
m oo cvj oo
qo--:
d d o o
However determined, values of M in partly filled sewers are generally much lower than in
filled pipes. Quite commonly 0se is zero, especially in small sewers, because there is often
enough oxygen in the wastewater to prevent the slime layer from contributing any sulfide,
but even where sulfide buildup in a sewer is found, „„ is usually less than a third of what
St/
would be expected in filled pipes where oxygen is completely excluded.
Complete suppresion of 0se requires enough oxygen so that the entire slime layer has a
thin aerobic zone. Because of intermittent sloughing of bits of slime and because the
condition of the slime layer and the flux of oxygen to the wall vary around the submerged
surface, sulfide may be released from some parts of the slime surface while it is blocked in
other parts.
In very large trunks, oxygen concentrations may drop so low that 0se may approach the
values observed in filled pipes.
3.2 Gains and Losses of Oxygen in Wastewater Streams
It was shown in Section 3.1.1 that the oxygen concentration is a critical factor in deter-
mining whether the slime layer can release sulfide into the wastewater stream. For a
complete understanding of sulfide buildup rates, it is necessary to examine the oxygen
resources of wastewater flow in sewers.
3.2.1 Oxygen Absorption at the Surface of a Stream
3-9
-------
One physical principle is basic to the mathematical modeling of the process of gas transfer
to or from a stream of water. It may be stated thus: The rate of approach to physical
equilibrium across a phase boundary is proportional to the existing disequilibrium. As
applied to the absorption of oxygen by a stream of water, it may be stated that the flux
of oxygen through a unit of surface area is proportional to the oxygen deficit, as shown
by the following equation.
f = fD
where:
0f = flux of oxygen per unit area of stream surface, g/m -hr
f = exchange coefficient, m/hr (practically the same as K^ used by some
authors)
D = oxygen deficit, defined as the difference between the concentration of
oxygen in the aqueous phase and the concentration that would prevail under
the condition of equilibrium with the adjacent atmosphere, mg/1 or g/m
The area referred to in the definition of 0f is, for practical reasons, the superficial area of
water surface, or the area of a projection of an irregular interface on to a horizontal plane
surface.
The validity of this equation is well established. It is valid not only for streams deficient in
oxygen, but also where the stream is supersaturated, in which case the flux is in the
opposite direction, and deficit is replaced by supersaturation. The relationship has been
demonstrated where the supersaturation, due to photosynthesis, is at least 4 mg/1 (6).
Only at considerably higher degrees of supersaturation is there likely to be spontaneous
bubble formation.
The following equations are useful in aeration calculations. They are either definitive
or are derived mathematically from the above equation or others of the group.
The reaeration rate (Rj-) is defined as the rate of change of oxygen concentration due to
absorption from the atmosphere and is related to other factors by the following equation:
Rf = 3.28 0fdm] = 3.28 f IDd"1
where:
Rf - reaeration rate, mg/l-hr
3.28 = conversion factor from meters to feet
^
0f = flux of oxygen per unit of area, g/m -hr
3-10
-------
dm = mean hydraulic depth, (defined as the cross-sectional area of the stream
divided by its surface width), ft
f = exchange coefficient, m/hr
D = oxygen deficit, mg/1 or g/m
The reaeration coefficient, K2, is defined as the proportional rate of satisfying of the oxygen
deficit. The symbol K2, introduced by Streeter and Phelps (7) in 1925 (but expressed in
days), is equivalent to Kj^a as currently used by some authors, in which a is, practically,
the reciprocal of dm. K2 is related to f and to Rf by the following equations:
and
K = 3.28
R = K
where:
K2 = reaeration coefficient, hr~
If the oxygen concentration is affected only by absorption at the surface of the stream,
,-1
where:
dt
= rate of change of oxygen concentration, mg/l-hr
By integration and insertion of limits, this becomes:
2.303 log
ID
1
= K2(t2-t})
(Restriction: surface aeration
is the only factor influencing
oxygen concentration)
where:
2.303 = factor for conversion from loge to
and D2 = initial and final oxygen deficits, mg/1
= reaeration coefficient, hr
tj and t2 = initial and final times, hr
.-1
3-11
-------
It has been shown (8) that for the stream in a sewer, f can be reliably predicted by the
equation:
f = 0.64 x 0.96 CAT (su)3/«
where:
f = exchange coefficient, m/hr
0.64 = (3.28)"'8 = conversion factor from meters to feet
0.96 = empirical coefficient applicable in wastewater streams
C^ = factor representing the effect of turbulence in creating additional air-
water interface in comparison with a slow stream
T = temperature coefficient, equal to unity at 20 deg C
s = slope of the energy line of the stream
u = stream velocity, ft/sec
CA can be approximated as shown below:
0.17 u2
CA = i +
where:
0.17 = empirical coefficient
g = gravitation constant, 32.2 ft/sec
d = mean hydraulic depth, ft
••>
(u /gdm, appearing in the formula for CA, is a Froude number; it is the square of the
Froude number commonly used in the United States in respect to streams.)
It has been shown (9) that the temperature coefficient T varies with the turbulence of the
stream. In typical sewer conditions, f increases about one percent/deg C.
Substituting the empirical equation for f into the definitive equation for Rf gives the
equation:
Rf = 2.10 x 0.96 CAT (su)3/» ID"1
3-12
-------
where:
Rf = reaeration rate, mg/l-hr
2.10 = (3.28) /8 = conversion factor from meters to feet
0.96 = empirical coefficient
On the basis of the above equations, together with calculations of required slopes to give
certain velocities, Figure 3-5 has been prepared, showing reaeration coefficients in sewers
of different sizes flowing half full. Also the reaeration rates are shown for the condition
that the oxygen deficit is 7 mg/1, a reasonable figure to use when considering whether the
oxygen supply will be sufficient to maintain a residual of 1 mg/1, to provide substantial
assurance against sulfide passing from the slime layer into the stream. It should be noted
that in calculating the deficit, the figure used for the solubility of oxygen should be reduced
proportionally if the ambient pressure is less than one standard atmosphere and if the
percentage of oxygen in the sewer atmosphere, on a dry basis, is less than the normal
20.9 percent.
Figure 3-5 is calculated for a temperature of 20 deg C. As already pointed out, an increase
of temperature increases K^, and would correspondingly increase the rate of oxygen
absorption provided the oxygen deficit remained the same. The deficit is likely to be less
at higher temperature because of the lower solubility.
If the sewer flows at some depth other than half full, factors taken from Figure 3-6 can be
used as multipliers for rates obtained from Figure 3-5. The velocity input when using
Fig. 5 in conjunction with Fig. 6 must be the velocity that would prevail when the
quantity of flow is such as to half fill the pipe, not the actual velocity for the actual
quantity of flow.
3.2.2 Oxygen Absorption at Points of High Turbulence
In addition to surface aeration, oxygen is added at junctions, drops, hydraulic jumps, and
other places of intensive turbulence that mixes air with the water. A simple fall or drop can
be expected to satisfy a certain fraction of the oxygen deficit. It has been shown (10) that
in simple falls or drops the oxygen concentration approaches saturation logarithmically with
the height of the fall according to the following equation :
2.303 log — - =0.305 KR (Hj - H2)
where:
2.303 = conversion factor from loge to log JQ
3-13
-------
FIGURE 3-5
REAERATION RATES IN SEWERS
FLOWING HALF FULL
VELOCITY, f ps
E
n
K
O
CM
O
-------
FIGURE 3-6
RELATIVE AERATION RATES IN A SEWER AT DIFFERENT FLOWS,
COMPARED TO THE RATE FOR THE HALF FILLED PIPE
O
<0
-------
Dj and D2 = oxygen deficits upstream and downstream from the drop, mg/1
0.305 = (3.28)"1 = conversion factor from meters to feet
Kpj = waterfall reaeration coefficient, m
H j and t^ = elevations of the hydraulic energy lines upstream and downstream
from the fall, ft.
From the limited data available, it appears that an average value of Ku for wastewater
is probably about 0.41 m . Table 3-1 shows the percentages of the deficits expected to
be satisfied with various heights of fall.
TABLE 3-1
EXPECTED OXYGEN ABSORPTION IN WASTEWATER FALLS
(Tentative, Based upon KJJ = 0.41m )
ft
i
2
3
4
5
6
Oxygen deficit
satisfied
percent
12
22
31
39
46
53
ft
8
10
15
20
30
Oxygen deficit
satisfied
percent
63
71
86.7
91.8
97.6
In a junction where streams with different hydraulic energy lines meet, the average elevation
of the energy lines above the junction and the elevation of the line for the combined flow
below should be used to obtain a value of H to use in Table 3-1.
A large flow of water with a drop as little as one foot may entrain little or no air at the
drop, but it may produce a hydraulic jump downstream. The efficiency of the jump in
mixing air with water may not be the same as in a fall. In the absence of data, it can only
be said that the hydraulic jump would probably have a similar but somewhat smaller
aeration effect than a clear drop with the same energy dissipation.
It will be seen that if the stream is devoid of oxygen, so that the deficit is about 8 mg/1,
a drop of 1.0 ft will result in the dissolution of about one mg/1 of oxygen. This may seem
like a small amount, but it could be enough to oxidize as much sulfide as would be
accumulated in a mile or more of flow.
In large trunk sewers, the low slopes and the relatively deep flows cause surface reaeration
rates to be very slow. A drop, however, causes approximately the same increase of oxygen
concentration whether the stream is large or small. For this reason, the points of intensive
turbulence are the dominant oxygen sources in large trunks.
3-16
-------
3.2.3 Oxygen Consumption in Sewers
Short-time oxygen reaction rates in wastewaters are quite unrelated to standard BOD or even
to the one-day BOD, and they change with the age of the wastewater (11).
Figure 3-7 shows equipment suitable for field use for determining oxygen concentrations in
wastewater samples obtained from sewers, as well as for determinations of the oxygen
reaction rates (11). For determining the reaction rate, the sample is aerated to a level of
several milligrams per liter of dissolved oxygen, and then, with the disc in place, oxygen
concentrations are read at suitable intervals of time.
Figure 3-8 shows results obtained with different samples of wastewater. The slopes of the
lines are the oxygen reaction rates, which are preferably expressed as mg/l-hr. It is seen
that the rates remain essentially constant until the oxygen concentration is below 1.0 mg/1.
The slowing down of the reaction rate begins at higher oxygen concentrations in samples
containing much coarse suspended matter.
Samples of sewage taken near the upper ends of small sewers show relatively low oxygen
reaction rates, typically 2 to 3 mg/l-hr. The rate increases if a sample is kept aerated, and at
a temperature around 20°C it may reach peak rates between 5 and 10 mg/l-hr after a few
hours, after which it declines. In a sewer, the increase in the reaction rates is much faster
than in bottles. This is due to the growth and shedding of biologically active material from
the pipe wall. Sewage reaching the end of a metropolitan trunk has an ability to reach very
high oxygen reaction rates after it is aerated for an hour or two.
Figure 3-9 shows the changing oxygen reaction rates of wastewaters from various sources
when aerated in the laboratory (11). Each point on the graph represents a determination
of the oxygen reaction rate as shown by the slope of the line on a plot such as those in
Figure 3-8.
As wastewater moves from the small collecting sewers into the larger lines, the slopes of
the sewers are in general flatter and the flow depths are greater. The effect is to decrease
the rate of oxygen supply by surface aeration, but while this is happening, the rate of oxygen
demand is increasing. The wastewater in small sewers is normally aerobic if the velocity is
greater than 2 ft/sec. The same is ture in sewers as large as 18 inches if they are flowing
less than half full at a velocity of 2 ft/sec or more. In trunks 24 in or more in diameter,
flowing half full at a velocity of 2 ft/sec, the oxygen concentration will usually decline to a
range of a few tenths or a few hundredths of a mg/1 if no oxygen is supplied other than by
normal surface aeration. Oxygen consumption is then limited by oxygen impoverishment,
and the rate of usage is controlled by the rate of supply.
Oxygen is used not only in the stream, but also by the slime layer. In many sewers oxygen
usage by the slime layer is limited only by the rate that oxygen is transferred to the slime by
the motion of the water. On the basis of somewhat limited data, a provisional equation was
developed (11) to describe this oxygen flux:
3-17
-------
FIGURE 3-7
OXYGEN MEASURING ASSEMBLY
ELECTRIC CORD
TO MICROAMMETER
PLASTIC DISC
D.O.
ELECTRODE
SWITCH
BATTERY OPERATED
MiXER MOTOR
WELL TO PREVENT
WATER EXCHANGE
THERMOMETER
WATER LEVEL
RUBBER GASKET
CAST IRON BUCKET
3-18
-------
FIGURE 3-8
TYPICAL OXYGEN REACTION CURVES IN WASTEWATERS
10
TIME (minutes)
3-19
-------
FIGURE 3-9
CHANGING OXIGEN REACTION RATES WITH TIME
IN WASTEWATER SAMPLES AERATED IN THE LABORATORY
Curve Sewage
A Fresh residential
B Fresh residential
Matured residential
Metropolitan trunk
Metropolitan trunk
Metropolitan trunk
3-20
-------
e = 0.55 x5.3
where:
0e = flux of oxygen to the slime layer, g/m -hr
0.55 = (3.28) ~~/2 = conversion factor from meters to feet
5.3 = empirical coefficient
02 = oxygen concentration, mg/1
s = slope of the energy line of the stream
u = stream velocity, ft/sec
The rate of change of oxygen in the stream due to reaction at the surface of the slime layer,
designated Re, is
Rp - 1.78x5.3
C
where:
R = loss of oxygen from the stream by reaction with the slime layer, mg/l-hr
1.78 = (3.28)'2 = conversion factor from meters to ft
r = hydraulic radius of the stream, ft
In small sewers, especially where the sewage does not exhibit a rapid oxygen reaction rate
and oxygen concentrations are relatively high, a major part of the oxygen usage is at the
slime layer. In larger flows, oxygen concentrations are very low, the slope, s, is generally
small, and the hydraulic radius of the stream, r, is large. For this combination of conditions,
oxygen reaction at the surface of the slime layer is unimportant.
3.3 Losses of Sulfide
3.3.1 Reaction of Sulfide with Oxygen
Under the conditions usually found in partly filled sewers, a major part of the sulfide
passing from the slime layer into the stream is subsequently destroyed by oxidation. The
reaction of sulfide with oxygen may be either chemical or biochemical.
In wastewaters that are biologically inactive because of toxic substances, extremes of pH, or
other reasons, sulfide is oxidized chemically. The reactions are complex, converting sulfide
3-21
-------
to sulfate and various intermediate oxidation products. The rates vary with sulfide and
oxygen concentrations, being very slow where the concentrations of either sulfide or oxygen
are only a few tenths of a mg/1.
The biological reaction is more rapid. The overall product is thiosulfate, as shown below.
2 02 + 2 HS- bacteria^ g^ + H20
The weight ratio of the reactants is 0.998 part of oxygen to one part of sulfide, or,
practically, 1:1. (The oxidation of sulfide to elemental sulfur, as can be brought about by
Beggiatoa alba, does not occur to a significant extent in sewers under ordinary conditions.)
The rate of sulfide oxidation by this reaction varies with the biological activity of the
wastewater, being as low as one to two mg/l-hr in fresh wastewater, but increasing to 10 to
15 mg/l-hr in wastewater that has flowed in the sewer for a few hours. The rate is in-
dependent of oxygen and sulfide concentrations as long as their concentrations are not less
than 1 mg/1. The ability to biologically oxidize sulfide to thiosulfate does not appear to be
confined to bacteria for which sulfide is a necessary energy source. Wastewater aerated for
a few hours in the absence of sulfide and then supplied with sulfide is found to have
acquired the ability to oxidize sulfide at a high rate.
In wastewater containing sulfide and one mg/1 or more of dissolved oxygen, 20 percent to
30 percent of the oxygen reacting in a typical case will serve to oxidize sulfide; the rest is
used by organic matter. Where the oxygen concentration is low, so that the rate of
reaction is restricted by oxygen starvation, as much as 50 percent of the available oxygen
may be used by the sulfide.
3.3.2 Escape of tS to the Atmosphere
The transfer of I^S between a body of water and the atmosphere may be calculated from
an exchange coefficient, fs, analogous to the coefficient f for oxygen.
The exchange coefficients of different gases vary as a power of the molecular diffusion
coefficients. Opinions Differ as to the exponent that should be used (12) (13) (14) (15).
It appears that the exchange coefficient may vary as the first power of the diffusion
coefficient in a very slow stream, but that it may vary as the 0.5 power under extremely
turbulent conditions. The diffusion coefficient for t^S in water is 0.64 as great as for
(>). Thus, the ratio of the exchange coefficients may be between 0.64 and 0.80. If an
intermediate value of 0.72 is used, it is not likely to be in error by more than a few percent.
Thus, from the equation for the exchange coefficient for oxygen as given in Section 3.2.1,
the exchange coefficient for sulfide may be estimated as follows:
fs = 0.64 x 0.72 x 0.96 CAT (su)3/*
3-22
-------
where:
fs = exchange coefficient for H^S between a stream of wastewater and a gas
phase, m/hr
0.64 = (3.28)~ /8 = conversion factor from meters to feet
0.72 = assumed ratio of exchange coefficients of I^S and C>2
0.96 = empirical coefficient in equation for oxygen absorption in wastewater
streams
C^ = factor representing the effect of turbulence in creating additional air-water
interface in comparison with a slow stream
T = temperature coefficient, equal to unity at 20 deg C
s = slope of the energy line of the stream
u = stream velocity, ft/sec
If the atmosphere were devoid of H^S, the flux of t^S from the stream would be the
exchange coefficient multiplied by the concentration of un-ionized I^S in the water.
H2S in the air proportionally reduces the rate of escape until it is zero when the two
phases are in equilibrium. A factor, q, will be defined as the relative saturation of the air
with H2S in comparison with the equilibrium concentration. The rate of escape of t^S is
proportional to (1 — q). Ordinarily the concentration of I^S in the sewer atmosphere is
between 2 percent and 20 percent of equilibrium; 1—q is 0.97 to 0.80. Thus, the rate of
escape of I^S is usually 97 percent to 80 percent of what it would be into a sulfide-free
atmosphere.
Noting that H^S] = j jps] ; the flux of J^S from the stream into the atmosphere is given
by the equations:
0sf = fs(i-q)j [DS|
0sf = 0.64 x 0.69 CAT (su)3/» (l-q)j [ps|
where:
= flux of H2S from the stream surface, expressed as grams of sulfide per
m2-hr
0.69 = 0.72 x 0.96 from the previous equation
q = relative I^S saturation in the air
j = H2S factor from Table 2-3
IDS! = dissolved sulfide concentration in the wastewater, mg/1
3-23
-------
The decline of sulfide in the stream as a result of loss of H^S to the air is equal to the flux
from the surface of the stream divided by the mean hydraulic depth:
Rsf = 3.280sf dm1
where:
Rsf = depletion of sulfide in the stream due to escape of t^S, mg/l-hr
3.28 = conversion factor from meters to ft
dm = mean hydraulic depth, ft
3.3.3 The Sulfide Balance
The rates of oxidation and of escape into the air increase as the sulfide concentration in-
creases. Under any given set of conditions, the sulfide concentration tends to approach a
limiting value at which point the losses from the stream equal the rate of supply.
To explore this relationship, rates of change of sulfide concentrations were measured in
sewers to which various concentrations of sulfide were added at upstream manholes. Figure
3-1OA shows the results in 24 D, a 24-in sewer at a slope of 0.0072, carrying a daytime flow
averaging 3.6 cfs. Each point is the result of seven pairs of analyses at the two ends of the
test reach. The points are somewhat scattered, partly because of experimental errors and
partly because of variable behavior of the wastewater. With the exception of one high point,
however, the trend of the points can be approximated by a straight line. It appears that in
an average case during the season and time of day that the data were obtained, the initial
rate of buildup of sulfide would be 0.7 mg/l-hr, and that sulfide would neither increase nor
decrease if its concentration were 1.2 mg/1. Figure 3-1 OA is typical of results in ten sewers
ranging from 10 in to 60 in in diameter (6).
From Figure 3-10A a curve has been calculated representing the buildup of sulfide that
would be expected in sewer 24 D if the initial sulfide concentration were zero and if flow
conditions remained uniform for a sufficiently long distance. This is shown in Figure 3-1 OB.
If the initial sulfide concentration is not zero, the expected buildup will follow the same
curve from whatever point represents the initial concentration.
3.4 Other Effects of Velocity
Up to this point, the effect of velocity has been considered only as a factor influencing the
rate of absorption of oxygen and release of H^S to the atmosphere. Velocity may also
affect the thickness of the slime layer, as well as the ability of the stream to transport solids.
Observations of the effect of velocity on the slime layer are few. The slime layer varies in
thickness at different locations on the submerged surfaces (16) but the minimum thickness
for unimpaired sulfide generation is not known. It was shown (11) that a velocity of 7 ft/sec
3-24
-------
FIGURE 3-10A
EFFECT OF SULFIDE CONCENTRATION
ON RATE OF SULFIDE BUILDUP OR DECLINE
0 1.0 2.0 3.0
AVERAGE TOTAL SULFIDE
CONCENTRATION mg/I
FIGURE 3-1 OB
SULFIDE BUILDUP CURVE, CALCULATED
FROM FIGURE 3-10 A
1.2
1.0
0.8
0.6
0.4
0.2
I 2
HOURS
3-25
-------
did not impair the ability of a slime layer to use oxygen at a maximum rate. It was also
found (4) that sulfide buildup in a pressure main where the pumping velocity was 4 ft/sec
was no less than in mains where the velocity was less. In view of the above work, it would
not appear safe to rely upon velocity to keep the slime layer thin enough to effectively
control sulfide generation. However, the slime layer may be reduced to a few mils, and at
that thickness a rather small oxygen concentration might keep it aerobic or at least would
prevent the small amount of sulfide that might be formed from getting into the stream.
The effect of velocity on the transport of solids is more tangible. A small amount of
gritty matter lying in the pipe does not have much effect on sulfide generation, but when
the velocity is slower, organic solids may concentrate at the bottom, intermittently
sliding and rolling along. Because of oxygen depletion in the loosely deposited solids, and
the large interfacial area, sulfide generation is then markedly accelerated. If the velocity is
slow enough so that a deep, stable deposit of sludge forms, it does not proportionally
increase sulfide generation, because the sludge becomes starved for sulfate, but conditions do
continue to become worse with decreasing velocity. Figure 3-11 shows these relationships.
In small collecting sewers serving a few homes, velocities are generally inadequate to carry
the solids in suspension. However, the stream is very shallow and the wastewater generally
has a fairly high oxygen content when it enters the sewer. Sulfide buildup does not occur
unless solids accumulate to a degree that causes ponding.
3.5 Forecasting Sulfide Buildup
3.5.1 Attempts at Predictative Equations
Several formulas have been proposed over a 30-year period to forecast sulfide buildup. The
earlier efforts were confined to describing the marginal condition separating buildup and no
buildup. The first formula published (1), expressly tentative and expressly limited to pipes
not over half full, purported to define a marginal velocity above which buildup would not
be expected. The marginal velocity was considered to be proportional to the square root
of the effective BOD. Later the Davy equation was developed (17) which took into account
the relative depth of flow, and which used a different assumption in respect to the effect of
velocity on oxygen absorption.
Subsequently, the Davy equation was modified and combined with a hydraulic equation to
produce what came to be known as the Z formula. The Z formula was formally published
in 1970 (18), with a caution as to its provisional nature. Thisthlethwayte (3) recommends
the Z formula for screening purposes, and presents a complex equation for calculating
buildup rates where the Z formula indicates that there may be problems. The formula is
as" follows:
Z . Jf*°5L x (P/b)
S/2 Q/3
3-26
-------
FIGURE 3-11
SOLIDS ACCUMULATIONS AT VARIOUS FLOW VELOCITIES
B
o
-. o. •
Velocity 2fps, Efficient solids transport.
No sulfide buildup in small flows, up to
2 cfs. Sulfide buildup often observed in
larger flows but only at a very slow rate.
Velocity 1,4 to 2.0 fps. Inorganic grit
accumulating in the bottom. More sulfide
buildup as the velocity diminishes.
Velocity I.O'to 1.4 fps. inorganic grit
in the bottom, organic solids slowly
moving along the bottom. Strongly
enhanced sulfide buildup; severe
problems expected.
Velocity below 1.0 fps. Much organic and
inorganic solid matter accumulating,
overlain with slow-moving organic solids.
Sulfide problems worse than in C.
3-27
-------
where:
JEBODJ = effective BOD, mg/1 or g/m3
s = slope of the energy line of the stream
Q = wastewater flow, cu ft/sec
P = wetted perimeter, ft
b = surface width, ft
Sulfide buildup is common in sewers where Z exceeds 10,000, but is rare where Z is
below 5,000.
Although these equations were sometimes based upon unverified assumption, they have, in
fact, been remarkably successful. Serious sulfide problems have been avoided where any one
of the earlier formulas was followed in sewer design. It is now known, however, that sulfide
buildup frequently does occur at a slow rate in large trunks where the predictive formulas
indicate that it would not be expected, and they indicate that sulfide buildup would occur
in very small flows where in fact it does not occur.
On the basis of the more adequate understanding of the sulfide generation and oxidation
processes gained through researches in Australia and the United States in recent years, it is
theoretically possible to write an equation that will predict the rate of change of sulfide
concentration in a sewer. It could not be done on the basis of conditions in a given reach.
It would be necessary to describe the history of the wastewater for 2 to 3 hours before it
arrived at that reach, so that sulfide and oxygen concentrations entering the reach could be
estimated. It appears impractical in any ordinary case, even though theoretically possible,
to obtain all of the necessary input information. An equation can be developed to show the
maximum sulfide buildup rates in uniform reaches if oxygen concentrations are low and all
other conditions are favorable for buildup, but such a limited equation is not useful for the
practical purposes of this manual.
Sewerage systems that have severe sulfide conditions and systems that have little or none
have characteristic differences that are clearly causative. Even though a complete pre-
dictive equation is not practical, the nature of the causative relationships is now well
enough understood to permit the establishment of useful generalizations, as set forth in
the following subsections.
3.5.2 Conditions of Sulfide Buildup in Small Sewers
The conditions existing in the small collecting sewers are highly erratic. Because of the
shallow depth and the low oxygen reacting rate in the fresh wastewater, and the effects of
service laterals along the way, the stream may remain free of sulfide even though the
velocity is so slow that solids are deposited.
3-28
-------
Little is known with certainty about flows in sewers serving a few homes. The flushing of a
single toilet will produce a flow of about 0.035 cfs, which is more than 100 times the daily
average flow from a residence. The flow from a single flush would soon level out to a slower
rate in a sewer, but more than one fixture may be in use at the same time, so occasional
flows of 0.035 cfs in the sewer are not rare. A reasonable estimate is that an occasional
peak flow from two houses might be 0.05 cfs. Considering the observed hydraulic
conditions in small sewers (19), the peak flows from one or two residences in a sewer laid
at a slope of 0.006 will not produce velocities much greater than 1 ft/sec. For a part of the
time the flow will be practically zero, even from several homes. At these times, the pipe
surfaces and part of the deposited solids will become quite well oxidized if solids accumu-
lations are not excessive. This condition, and the fact that the fresh wastewater contains
dissolved oxygen, will preclude sulfide generation close to the wastewater sources unless
solids are stranded in such amounts that they produce stagnant pools.
A survey of the occurrence of sulfide in small collecting sewers was made in the 1950's,
largely under the sponsorship of the American Concrete Pipe Association. Sulfide analyses
were made in successive manholes from the upper ends of 6-in and 8-in sewers in residential
areas, continuing as far downstream as the slopes were reasonably uniform and there were
no significant junctions other than service laterals. The results do not represent sulfide
buildup as normally defined, because it was not a case of following the same body of
sewage downstream.
The tests were made in several cities in southern California. The temperature range of the
sewage was from 20 deg C to 28 deg C, with an average of 24.5 deg C. It is well known that
higher temperatures cause higher sulfide concentrations, other conditions being the same,
but the effects of temperature were overshadowed in this series by the effects of slope. No
distinction was made between the two pipe sizes (6" and 8"), since for the same flow and
slope, velocity is essentially independent of pipe size.
The sewers were divided into four slope classes, and the results averaged for each class.
Figure 3-12 shows the dominant effect of slope. The increasing flows at greater distances
down the sewers would eventually lead to a condition where concentrations would no
longer increase, but would begin to decline. The decline would start sooner at the steeper
slopes. There is not a significant indication from the data plotted in Figure 3-12 that
average sulfide concentrations have passed a peak in the approximately 1,500 to 2,000-ft
reaches represented by lines A, B, and C. The average housing density where the data
were collected was 30 residence units per 1,000 ft of sewer. It appears, therefore, that even
where the slope is 0.006 (line C),a sulfide decline will not set in until the sewer carries the
wastewater from at least 60 homes (2,000 ft), and perhaps 100. Where 100 homes are
served, the fluctuations of flow due to discharges from individual sources at the high-
flow period of the day have been largely damped out. The peak flow is the result of the
diurnal variation of average water usage. From 100 homes, a daily maximum flow rate of
0.15 to 0.20 cfs is a reasonable expectation, varying with population characteristics. The
peak velocity from 100 homes in a sewer at a slope of 0.006 will probably be about
1.6 ft/sec.
3-29
-------
FIGURE 3-12
SULFIDE OCCURRENCE IN SMALL SEWERS
o
H
(t
U
o
u
o
uZ
_i
3
CO
111
Si
K
UJ
0.6
0.4
0.2
SLOPE %
LINE RANGE AVG.
A 0.20-0.25 0.23
B 0.32-0.46 0.40
C 0.52 -0.64 0.57
D 0.72-1.20 0.9
NUMBER
OF SEWERS
18
22
16
8
AVERAGE RESULTS
B.QO. TEMR°C
253 24.7
212 24.4
178 24.7
184 24.3
400
800
1200 1600 2000
DISTANCE FROM UPPER END OF SEWER , FT.
3-30
-------
The actual velocity will, of course, be less much of the time. Organic solids settle out
during low flow periods, but are moved along quite well by the high flows.
For the same quantity of flow in clean pipes of circular cross section, velocity varies as the
0.41 power of the slope (19). Thus, at a slope of 0.004 (line B), the velocity in clean pipes
would be 0.865 as great as at s = 0.0057 (line C), but because of greater fouling as the
slope is diminished, the velocity would probably be even less. Because of this reduction of
slope and velocity, the amount of sulfide observed is about twice as great.
There are some regions where the collecting sewers have been laid at very flat slopes, yet
show no sulfide generation. This is generally because of infiltrations of groundwater, or
other factors conducive to low-strength sewage and ample dissolved oxygen. The available
data indicate that temperature differences are probably not a major factor in sulfide buildup
in the collecting sewers.
3.5.3 Sulfide Buildup in the Larger Flows
When the collected flow is great enough so that the bottom of the pipe remains essentially
free of organic deposits, the determining factor in respect to the occurrence of sulfide is the
oxygen balance in the stream. For flows up to one or two cfs, or more if the temperature
is low, and at a velocity not less than 2 ft/sec, surface aeration is generally sufficient to
maintain an oxygen residual of a few tenths of a mg/1 in the stream, and to suppress sulfide
buildup. In larger flows, a greater oxygen absorption rate at the surface of the stream would
be necessary to prevent sulfide buildup, because of the greater volume of water that must be
supplied with oxygen, but the actual absorption rate is generally less in the large sewers
because of smaller slopes. However, in these larger flows the effects of junctions and other
points of turbulence become more important than surface aeration of the normally flowing
stream.
It has been shown that the rate of absorption of oxygen in rivers can be expressed as
proportional to the rate of loss of elevation, provided a major part of the aeration occurs in
falls and rapids (15). Applying the same principle to sewers, the over-all loss of elevation
occurring over a distance equal to a flow time of an hour can be considered as a crude
parameter of oxygen supply. It is not possible to calculate an oxygen balance in the
classical sense, because oxygen concentrations in the larger sewer drop to starvation
levels, where the rate of oxygen consumption is much less than it would be if ample
oxygen were present. Nevertheless, the rate of energy dissipation does appear as a major
factor correlating with sulfide buildup.
Based upon an evaluation of available data, Figure 3-13 has been drawn as an aid in
forecasting, in a qualitative way, the possible sulfide conditions that may occur for various
slope-flow combinations in sewerage systems. (As a matter of convenience, a cube-root
scale is used.) The vertical scale is the over-all slope, that is, the elevation loss averaged over
distances equal to one-hour flow times, except for the small collecting sewers where it
3-31
-------
FIGURE 3-13
FLOW-SLOPE RELATIONSHIPS AS GUIDES
TO SULFIDE FORECASTING
0.6
CONDITIONS
EFFECTIVE B.O.D
500MG/L
.01
10 IS 20 30 40
O.I .5248
FLOW, CUBIC FEET PER SECOND (DAILY 6 HOUR HIGH)
3-32
-------
means the actual pipe slope. All substantial tributary flows should be considered in the
calculation of energy loss. Even a small tributary flow has an important effect if it drops
several feet into the larger stream.
3.5.4 Interpretation of Figure 3-13
Interpretations of Figure 3-13 will be restricted to sewers with flow depths not exceeding
two-thirds of the pipe diameter. This means that there cannot be any pressure mains unless
measures are taken to offset their oxygen-depleting and sulfide-producing potentials. Except
in small sewers, the vertical scale can be considered to be AH/AL, where H is elevation and L
is distance in the same units, calculated upstream from the point of reckoning over distances
representing flow times of about an hour. It will be referred to as "effective slope." In the
collecting sewers of minimum size, the actual pipe slope should be used.
The interpretation of Figure 3-13 must be qualified as being related to wastewater streams
of specified characteristics. The temperature of the wastewater in a sewer follows an annual
cycle, BOD and quantity of flow have diurnal cycles, and flows usually have long-time
trends. It is useful to define a climactic condition as the combination of the average BOD
and flow for the highest six -hour flow period of the day, and the average temperature for
the warmest three months of the year. Where diurnal BOD curves have not been made, it
may be assumed that the BOD for the six-hr high-flow period is 1.25 times the BOD of a
flow-proportioned 24-hr composite. The climactic EBOD (effective biochemical oxygen
demand) is defined by the equation:
CX
[EBOD] c = [BOD]
where:
[EBOD]C = climactic EBOD, mg/1
[BOD]C = climactic BOD, mg/1
TC = climactic temperature, deg C
Let it be assumed that the climactic EBOD for a sewer is 500 mg/1. Then the inter-
pretation of Figure 3-13 is as follows:
Curve A. While the climactic condition prevails, a system functioning with slope-flow
relationships as shown by Curve A may be expected to produce very little sulfide, rarely
more than 0.1 or 0.2 mg/1 of dissolved sulfide. The annual average dissolved sulfide
concentration is expected to be only a few hundredths of a mg/1.
Curve B. While the climactic condition prevails, a system functioning with slope-flow
relationships as shown by Curve B may produce dissolved sulfide at concentrations of
several tenths of a mg/1.
3-33
-------
It must be emphasized that Figure 3-13 cannot be used to accurately predict sulfide
conditions. The purpose is to indicate an apparent trend in the effect of flow-slope
relationship on sulfide buildup. Additional data are necessary to refine and fully sub-
stantiate the indicated relationships.
If the climactic EBOD is higher or lower than 500, the positions of the curves will be
altered, except that the cut-off effective slope of 0.6 percent for Curve A at small flows
will remain the same, since the determining factors are hydraulic. The relationship
between EBOD and required slope is complex. The nutrients available for the sulfide-
generating bacteria in the slime layer are presumed to be roughly proportional to the
B^OD, but the oxygen requirement to prevent sulfide buildup may not increase in the same
ratio. If slope, and hence velocity, is increased, the greater turbulence not only increases
oxygen absorption, but also provides more effective transfer of oxygen to the slime
surface. As a result, a lower oxygen concentration suffices to prevent the escape of
sulfide from the slime layer into the stream. Equally important, the length of time
that the wastewater spends in transit is reduced.
Considering all factors, it is judged that similar sulfide conditions will result if the effective
slopes are increased or decreased in proportion to the square root of the effective BOD.
Thus, Figure 3-13 shows that for the Curve A condition, a flow of 2.0 cfs requires an
effective slope of 0.175 percent. If the EBOD were 600 mg/1 instead of 500, the effective
slope for the Curve A condition at 2.0 cfs would be
0.175 percent x-y/600/5 00 = 0.19 percent
For a system in which effective slopes are substantially higher than required by Curve A,
sulfide concentrations will be negligible for practical purposes. For conditions substantially
below Curve B, higher sulfide concentrations must be anticipated, generally requiring odor
and corrosion control measures, or continuous treatment of the wastewater to prevent
sulfide buildup.
3.6 The Effects of Sulfide in Waste waters
3.6.1 Outline of the Effects
3.6.1.1 Odor
The obnoxious odor of HoS escaping from manholes, pump stations, and wastewater
treatment plants is the most potent objectionable characteristic of septic sewage. Chapter 2
discusses the odor potential of t^S.
3.6.1.2 Toxicity
Many lives have been lost from F^S poisoning in sewers. The incidence of fatal poisoning
has been less in recent years, due to greater attention to safety precautions, but a deadly
atmosphere will always be a hazard in some degree.
3-34
-------
3.6.1.3 Effects on Wastewater Treatment Processes
These effects are due to sulfide in the wastewater and not to H2S that has escaped into the
air. One effect is the adverse reaction of the activated sludge process. The growth of
filamentous organisms, especially Thiothrix, is encouraged by the presence of sulfide. The
result is a poorly-settling sludge. This is not a common occurrence, and it can be overcome
by prechlorination to destroy the sulfide, but if raw wastewater must be chlorinated, the
chlorine demand is greatly increased by the presence of sulfide.
3.6.1.4 Corrosion of Metals and of Cement Bonded Materials
H2S acts directly on silver and copper, first turning them black. The copper sulfide first
formed oxidizes slowly, under moist conditions, to copper sulfate. Prolonged exposure
causes destruction of the metal.
Cadmium is changed to yellow CdS. A similar reaction, although not properly called
corrosion, is the blackening of lead-bearing paints by the formation of PbS. Iron and
cement bonded materials are not directly attacked by H2S in the air at ordinary tem-
peratures, but they may be corroded by acid formed from H2S. Examples are given in
several publications (20) (3). The protection of existing facilities from this kind of attack",
and the design of systems to avoid damage, are mong the objectives of this manual. A
review of the mechanism of the attack and an examination of the rate-determining factors
are here presented.
3.6.2 The Corrosion of Cement Bonded Materials
3.6.2. 1 The Corrosion Mechanism
The attack of cement bonded materials in the vicinity of sulfide-containing water, and in
part the corrosion of iron, result from the intermediate formation of sulfuric acid by
oxidation of hydrogen sulfide, as follows:
H2S + 2 02 acera^ H2SO4
Bacteria of the genus Thiobacillus bring about the reaction. Under acid conditions, the
functioning species is T. concretivorus (21). These bacteria remain active at sulfuric acid
concentrations up to at least 7 percent. If the structure is made of cement bonded material,
the sulfuric acid attacks the cement, producing a pasty mass of gypsum plus the residual
inert materials. If the rate of production is slow, substantially all of the acid will react, but
if it is formed at a rapid rate it cannot diffuse through the pasty layer as fast as it is
produced, and some of it will be carried by condensed moisture back into the stream
where the alkalinity of the wastewater will reconvert it to sulfate ion.
There is no over-all effect on the pH of the stream as a result of this recycling of sulfur.
The production of sulfide does not lead to the corrosion of cement bonded surfaces
submerged by the wastewater.
3-35
-------
If there is no condensed moisture on the walls of the pipe or other structure, no sulfuric
acid is produced. If the surface is unreactive and very high acid concentrations accumulate,
sulfuric acid formation may be stopped. Under these conditions, H^S is often oxidized to
sulfur, appearing as a pale yellow deposit. If favorable conditions subsequently develop,
bacteria will oxidize the sulfur to sulfuric acid.
3.6.2.2 The Distribution of Corrosion in Sewers
Corrosion of sewer pipe made of cement-bonded' materials is not uniform. Lack of
uniformity is due in part to the air currents that control the rate of transfer to t^S to the
pipe wall. The greatest corrosion is generally observed at the soffit of the manhole outlet,
because that is where there is the greatest shear between the airstream and the pipe
material. Structures projecting into the airstream suffer more rapid corrosion than the pipe
wall. Test specimens hung in a sewer may provide information on the relative corrodability
of different materials, but they will not show how fast a pipe wall will corrode.
There is normally a flow of air down the sewer, but in addition, transverse currents are set
up by temperature differences. The pipe wall is normally cooler than the water, especially
in the summer when sulfide concentrations are maximal. The air that is cooled by the walls
moves downward, and slightly warmer air rises from the center of the stream surface. As a
result, the maximal rate of transfer of f^S to the pipe wall is at the crown.
Uneven distribution of corrosion also results from the migration of acid-containing con-
densate down the pipe wall, particularly when there is a high rate of acid production. In the
zone that is intermittently washed by the water, the pasty decomposition products are
cleaned away. As a result, the pipe wall is laid bare to the attack of the acid when the
water level is low. Deeper penetration may therefore be observed in this zone.
Figure 3-14 shows the patterns just described, and Figure 3-15 is a photograph of a portion
of a pipe showing extreme waterline corrosion, with grooves cut by the migrating acid.
The pipe served in a city on the eastern seaboard of the United States. The specimen was
taken near the discharge from a force main where high sulfide concentrations prevailed and
where the discharge produced high turbulence.
3.6.2.3 Calculation of Corrosion Rates of Cement Bonded Pipe
The corrosion of a cement bonded sewer depends upon the supply of sulfuric acid and
hence upon the rate of release of F^S from the wastewater stream. The net mass emission
of sulfur from the stream is the mass transfer to the pipe wall, except for generally
negligible amounts of t^S escaping entirely from the sewer.
Section 3.3.2 showed that the net flux of sulfide from the surface of the stream to the
air can be calculated by the following equation:
3-36
-------
FIGURE 3-14
UNEQUAL DISTRIBUTION OF CORROSION
IN A CONCRETE SEWER
3-37
-------
FIGURE 3-15
ACID ATTACK OF SEWER AT WATERLINE
3-38
-------
0sf = 0.64 x 0.69 CAT (su)% (1-q) j |bs]
where:
o
0sf = flux of H2S (as S) from the stream to the air, g/m -hr
0.64 = (3.28)~ h - conversion factor from meters to feet
0.69 = empirical coefficient
C^ = factor representing the effect of turbulence in creating additional air-
water interface in comparison with a slow stream
T = temperature coefficient, equal to unity at 20 deg C
s = slope of the energy line of the stream
u = stream velocity, ft/sec
q = relative t^S saturation in the air
j = H2S factor from Table 2-3
IpSJ = dissolved sulfide concentration in the wastewater, mg/1
Under typical sewer conditions, excluding shallow, high velocity streams or points of high
turbulence, the flux can be approximated by the following equation:
0sf = 0.64 x 0.7 (su)3/« j |bs|
If all the escaping H^S is oxidized on the pipe wall, the average flux to the wall is equal to
the flux from the stream multiplied by the ratio of stream surface area to exposed wall area,
or surface width divided by exposed perimeter:
0SW = 0.64 x 0.7 (su)3/« j IDS] (b/P')
where:
•j
0SW = flux of H2S to the pipe wall, g/m -hr
(b/P') = ratio of surface width of the stream to exposed perimeter
Figure 3-16 shows rates of sulfide flux to the pipe wall in pipes flowing half full. The strong
upward curvature of the curves for the smaller pipes is due to the increasing C^ factor at
3-39
-------
FIGURE 3-16
EFFECT OF VELOCITY AND PIPE SIZE ON SULFIDE FLUX
TO PIPE WALL UNDER SPECIFIED CONDITIONS
0.24
0.20
oc
x
(M
2
x
O
S
0.16
I
ui
5: 0.12
Q.
0.08
UJ
o
£0.04
-e-
CONDITIONS:
PIPES FLOWING HALF FULL,
UN-IONIZED HYDROGEN
SUFIDE = 1.0 MG/L.
1
12
246
FLOW VELOCITY, FPS
24
36" w
a:
UJ
48" i
72 UJ
a.
96" E
144"
8
3-40
-------
high Froude numbers. Figure 3-17 provides factors to apply to rates from Figure 3-16 to
calculate the flux at other relative flow depths. The velocity input for use in conjunction
with Figures 3-16 and 3-17 must be the velocity that would prevail if the quantity of flow
were such as to half fill the pipe, not the actual velocity for the actual quantity of flow.
The rate of corrosion of cement bonded material can be calculated from the amount of
reactive material in the pipe wall that will consume acid, and the rate of acid production.
Sulfide in the amount of 32 grams is required to produce the acid to dissolve 100 grams
of CaCOo. If all of the acid reacts, the rate that cement bonded materials can be penetrated
is as follows:
f)sw (1/2.4) (100/32) (I/A) x 10~4 = (1.3 x 10"4
JA.) cm/hr
c'=11.50sw (I/A) mm/yr
c' = 0.490sw(l/A)in/yr
where:
c' = average rate of penetration if all the acid reacts
2.4 = density of the cement bonded material
(100/32) = ratio of reacting weights of CaCO3 and S
A = alkalinity of the cement bonded material expressed as CaCOg equivalent
1C"4 = m2/cm2
Usually the acid does not all react, so a factor (k) must be inserted as a correction for the
decrease of attack due to acid running back into the stream. The equation, therefore, may
be written as follows:
c= 11.5k0sw(l/A)mm/yr
c = 0.49 k0sw( I/A) in/yr
where:
c = average rate of penetration corrected for incomplete reaction of the acid
No actual measurements of k have been made, and the choice of a value to use is a matter of
engineering judgment. It appears that k approaches unity when the rate of acid formation
is very slow, and that it may be as low as 0.3 to 0.4 if acid production is rapid and if much
condensate is formed (warm wastewater, cold pipe wall).
3-41
-------
FIGURE 3-17
FACTOR TO APPLY TO 0S FROM FIG. 3-16 TO
CALCULATE 0W FOR OTHER THAN HALF-PIPE DEPTH
s
•n
-e-
o
o
UJ
a.
a.
<
QC
O
t-
u
2
2
O
H
O
tu
e
er
o
o
1.6
—
K>
—
O
O
00
O
0)
O
O
K>
0 0.2 0.4 0.6 0.8 1.0
RELATIVE DEPTH OF FLOW
3-42
-------
The calculation is applicable only to uniform pipe reaches well removed from points of high
turbulence, and it shows only the average corrosion rate of the exposed pipe wall. It is a
matter of judgment as to how much more rapid the corrosion may be at the locations of
greatest penetration, which, in a uniform reach, will be at the crown or near the water line.
It seems likely that the most rapid attack may be typically 1.5 times the average.
The alkalinity, A, or acid reacting capability of the pipe material, can be determined by
analysis or can be estimated from the composition of the mix used in making the pipe.
Anhydrous cement has an alkalinity, expressed as calcium carbonate, equal to about
1.18 times its weight. Thus, if the proportion of cement used, expressed as a percentage of
the weight of the cured pipe, is multiplied by 1.18 the result is the alkalinity contributed
by the cement. In concrete pipes made with granitic aggregate the alkalinity is usually
between 16 and 24 percent. In asbestos cement pipe the alkalinity is usually around 50
percent.
In some areas calcareous rocks, meaning limestone or dolomite, are more available than any
other hard rock, and therefore are used as aggregate for concrete. Where calcareous rocks
are used in making concrete pipe, the alkalinity may be 100 percent or even higher. In many
places where calcareous rock is not abundant, it has nevertheless been specified as aggregate
for pipes used in sewer construction. In most of these cases a calcium carbonate equivalent
of 90 percent in the finished pipe has been specified. The difference that this makes in
durability under acid conditions has been repeatedly demonstrated (22) (23) (24). Figure
3-18 is a photograph of two sections of pipe that had been laid in tandem and served for
7 yr in an experimental sewer carrying high-sulfide effluent from a septic tank. (The water
level was high in the pipe, so only a narrow band at the top was corroded.) The pipe
marked L was made with limestone aggregate. The alkalinity of pipe L as determined by
analysis was 87.9 percent; for pipe II it was 24.4 percent.
FIGURE 3-18
SECTIONS OF EXPERIMENTAL PIPES AFTER
EXPOSURE TO SEPTIC TANK EFFLUENT FOR SEVEN YEARS
3-43
-------
CASE HISTORY
Joint Outfall B of the Los Angeles County Sanitation Districts delivers wastewater to the
Joint Water Pollution Control Plant by way of a 144-in pipe installed in 1954. The line was
inspected in 1974. In a carefully studied reach it was found that corrosion at the crown had
penetrated the concrete to a depth of 0.25 inch. Calculations were then made by the
predictive equation for the rate of corrosion.
A range of conditions needs to be considered. During a low-flow period of about 6 hr in
the forenoon the flow is 150 cfs, with a depth of 6 ft and a velocity of 2.7 ft/sec. For
about 6 hr in the evening it is 370 cfs, with a depth of 9.2 ft and a velocity of 4 ft/sec.
From 1954 to 1963, sulfide concentrations were high due to surcharging of upstream
sewers. Dissolved sulfide averaged.0.75 mg/1. With the surcharged condition largely
relieved, the average dissolved sulfide concentration now averages 0.28 mg/1.
Calculations of" corrosion rates were made for different combinations of conditions, but
for the purpose of this example only the average condition will be used. The result in this
case was very nearly the same as the average result for the conditions calculated separately.
The slope of the invert is 0.024 percent, but in the intensively studied reach the slope of
the energy line ranged from 0.0131 percent to 0.0188 percent, averaging 0.0167 percent.
Other input data are as follows:
Velocity, u 3.5 ft/sec
Water depth, d 8.0 ft
Dissolved sulfide, [bsj 0.51 mg/1
pH 7.4
pK' 6.96
j 0.275
Hydrogen sulfide, fH2sl 0.14 mg/1 as S
CA 1.01
T 1.05
1-q 0.97
b/P' 0.766
Alkalinity of the concrete, A 16 percent
3-44
-------
0SW = 0.64 x 0.69 CAT (su)°-375 (1-q) [H2SJ (b/P')
= 0.64 x 0.69 x 1.01 x 1.05 x 0.0613 x 0.97 x 0.14 x 0.766
= 0.0030 g/m2-hr
c = 0.49x0sw(l/A)k
= 0.49 x 0.0030 (1/0.16) k = 0.0092 k
where:
c = average rate of penetration corrected for incomplete reaction of the acid
It is a reasonable estimate that the rate of acid formation at the crown of this sewer at any
given time would be 1.5 times the average rate of formation on the exposed surface. It also
seems likely that in this large sewer, with a moderate rate of acid formation, not more than
5 percent would be carried away from the crown by the flow of condensate, that is k is not
iess than 0.95. Thus, the rate of crown corrosion might be expected to be as follows:
c (at crown) = 0.0092 x 1.5 x 0.95 = 0.013 in/yr
Thus, the predicted corrosion over the 20-year period would be 0.26 in. The uncertainties
in the measurement of the depth of corrosion and in the input data required for the
calculation could result in discrepancies that may well be as much as 40 percent of the
corrosion rate. Thus, the good agreement between the observed and calculated corrosion
rates, 0.0125 in/yr and 0.013 in/yr, cannot be looked upon as an indication of the
precision of the predictive equation, but it does show that it is basically sound.
A major part of the corrosion in Joint Outfall B occurred in the early years when sulfide
concentrations were high. The present rate of corrosion is believed to be about 0.7 in/
century.
3.6.3 Corrosion Rates of Ferrous Metals
The rate of acid attack of iron and steel by acid produced by the oxidation of I^S can
probably be estimated by the same general equation as is used for cement bonded materials.
For the density of iron, 7.5 would be used instead of 2.4 as for concrete, and the acid-
consuming capacity, corresponding to alkalinity, would be 1.79. The equation then
becomes
c = 0.089 0swk in/yr
The action of acid on iron may stimulate the simultaneous reaction of the metal with
oxygen, so that it is possible for the corrosion rate to be greater than would be accounted
for by the acid alone.
3-45
-------
High sulfide concentrations in water may corrode iron rapidly if oxygen is also present,
producing iron sulfides, and under some conditions, especially in the presence of high
chloride concentrations, iron may be corroded anaerobically with the simultaneous
reduction of sulfate to sulfide (25) (26).
3.7 References
1. Pomeroy, R. D., and Bowlus, F. D., Progress- Report on Sulfide Control Research,
Sewage Works Journal, 18, No. 4, pp 597-640 (1946).
2. Postgate, J. R., The Nutrition of Desulfovibrio desulfuricans, Journal General
Microbiology, Vol. 5, pp 714-724 (1951).
3. Thistlethwayte, D. K. B. (Editor), Control of Sulphides in Sewerage Systems,
Butterworths Pty. Ltd., Melbourne, Australia (1972), and Ann Arbor Science
Publishers, Ann Arbor, Michigan (1972).
4. Pomeroy, R. D., Generation and Control of Sulfide in Filled Pipes, Sewage and
Industrial Wastes, 31, No. 9, pp 1082-1095 (1959).
5. Laughlin, J. E., Studies in Force Main Aeration, Journal of the Sanitary Engineering
Division, ASCE, 90, SA6, pp 13-24 (1964).
6. Parkhurst, J. D., Pomeroy, R. D. and Livingston, J., Sulfide Occurrence and Control in
Sewage Collection Systems, Report to the U.S. Environmental Protection Agency under
Research and Development Grant No. 11010 ENX (1973).
7. Streeter, H. W., and Phelps, E. B., Public Health Service Bulletin No. 146, U. S. Public
Health Service (1925).
8. Parkhurst, J. D., and Pomeroy, R. D., Oxygen Absorption in Streams, Journal of the
Sanitary Engineering Division, ASCE, 98, SAI, Proc. Paper 8701, pp 101-124(1972).
9. Metzger, I., Effects of Temperature on Stream Aeration, Journal of the Sanitary
Engineering Division, ASCE, 94, SA6, Proc. Paper 6309, pp 1153-1159 (1968).
10 Pomeroy, R. D., and Lofy, R. J., Feasibility Study of In-sewer Treatment Methods,
Report to the U.S. Environmental Protection Agency under Contract No. 14-12-944
(1972).
11. Pomeroy, R. D., and Parkhurst, J. D., Self-purification in Sewers, Proceedings of the
6th International Conference on Water Pollution Control Research, Jerusalem (June,
1972), Pergammon Press.
3-46
-------
12. Dobbins, W. E., BOD and Oxygen Relationships in Streams, Journal of the Sanitary
Engineering Division, ASCE, 90, SA3, Proc. Paper 3949, pp 57-78 (1964).
13. Thackston, E. D., and Krenkel, P. A.., Discussion of BOD and Oxygen Relationships
in Streams, Journal of the Sanitary Engineering Division, ASCE, 91, SAI, pp 84-88
(1965).
14. Davies, V. T., Kilner, A. A., and Ratcliff, G. A., The Effect of Diffusivities and Surface
Films on Rates of Gas Absorption, Chemical Engineering Science, Vol, 19, p 583
(1964).
15. Tsivoglou, E. C., and Wallace, J. R., Characterization of Stream Reaeration Capacity,
U. S. Environmental Protection Agency, Ecological Research Series, EPA-R3-72-012
(October, 1972).
16. Parker, C. D., et al., Concrete Sewer Corrosion by Hydrogen Sulphide, Melbourne and
Metropolitan Board of Works Technical Paper No. A.8, Part 2, Generation of Sulphides
in Sewers.
17. Davy, W. J., Influence of Velocity on Sulfide Generation in Sewers, Sewage and
Industrial Wastes, 22, No. 9, pp 1132-1137 (1950).
18. Pomeroy, R. D., Sanitary Sewer Design for Hydrogen Sulfide Control, Public Works
(October, 1970).
19. Pomeroy, R. D., Flow Velocities in Small Sewers, Journal Water Pollution Control
Federation, 39, No. 9, pp 1525-1548 (1967).
20. Swab, B. H., Effect of Hydrogen Sulphide on Concrete Structures, Journal of the
Sanitary Engineering Division, ASCE (September, 1961).
21. Parker, C. D., Species of Bacteria Associated with the Corrosion of Concrete, Nature,
159, p 439 (1947).
22. Stutterheim, N., and Van Aardt, J. H. P., Corrosion of Concrete Sewers and Some
Remedies, South African Industrial Chemistry, 7, No. 10 (1953).
23. Pomeroy, R. D., Protection of Concrete Sewers in the Presence of Hydrogen Sulfide,
Water and Sewage Works (October, 1960).
24. Pomeroy, R. D., Calcareous Pipe for Sewers, Journal Water Pollution Control
Federation, 41, No. 8, Part 1, pp 1491-1493 (1969).
3-47
-------
25. Beckwith, T. D., The Bacterial Corrosion of Iron and Steel, Journal AWWA, 33,
147-164 (1941).
26. Pomeroy, R. D. Corrosion of Iron by Sulfide, Water Works and Sewerage, pp 133-138
(April, 1945).
3-48
-------
CHAPTER 4
INVESTIGATIONS OF EXISTING SYSTEMS
4.1 Purpose
Information about sulfide conditions in existing sewerage systems is important for proper
operation and maintenance, and often for the planning of new construction. An investi-
gation may therefore be undertaken to secure this information. In addition to exploring
sulfide conditions, an investigation of a system will frequently include the objective of
obtaining information on carrying capacity, physical soundness, infiltration, need for
cleaning, need for repairs, and sometimes an appraisal of value. Even where the objective
is related only to existing or possible sulfide problems, observations should be made on the
physical condition of the system, flow patterns, flow conditions as affected by solids
accumulations, and characteristics of the wastewater. Often an investigation is confined to
a limited part of a system, where there are known problems due to sulfide, infiltration, or
poor flow conditions.
4.2 Examination of the Existing System
4.2.1 Map of the Existing System
An agency operating a sewerage system should have a map or set of maps of the sewers,
showing all sewers with their sizes, slopes, types of pipe, and locations of manholes and
other pertinent structures. Usually invert elevations at manholes are also shown. A city
of moderate size can have all of the sewers shown on a single map, but for a large city it is
necessary to have a trunk sewer map plus sectional detailed maps (1) (2). Detailed maps are
necessary for whatever part of a system is the subject of an investigation, and if they do not
exist, the mapping should be a part of the project.
4.2.3 Examination of Physical Conditions
The interior of all manholes in the study area should at least be viewed from the surface of
the ground. Not less than 10 percent of the manholes, and sometimes 100 percent, should
be entered and inspected in detail. Prior to the inspections, someone familiar with the
system should locate the manholes and try the covers. In inspecting a trunk sewer, half of
the time may be wasted hunting for the manholes. Sometimes they have to be dug up from
beneath pavement.
When a man enters a manhole, it must be done with a full understanding of the potential
hazards and in accordance with all Federal, State, and local safety regulations (3).
The conditions of the manhole, especially the steps and lower sections, should be observed
and recorded. It is important to look for infiltration and to record any that is seen,
4-1
-------
estimating or qualitatively describing the flows. The inlet and outlet pipes should be in-
spected insofar as possible. Two mirrors are commonly used, one held by a man at the
top to reflect sunlight down the manhole, and the other used by the man at the bottom to
direct the beam as well as his line of sight into the pipe. Even with reflected sunlight, the
pipe cannot be effectively observed for a distance of more than 10 or 15 pipe diameters.
If artificial light must be used, it should be a high-power source. Ordinary hand flashlights
are not very effective.
If the pipe is corrodable, the most severe condition is generally found at the soffit, or
above the soffit if the pipe is of vitrified clay or other resistant material. A condition of
more severe corrosion may exist somewhere between manholes, but only if there are
irregularities or junctions that cause substantially greater release of hydrogen sulfide than
at the manholes. Partial stoppages are sometimes seen and can be removed at the time of
the inspection.
Where it is necessary to determine the condition of the pipe between manholes, use can
be made of closed-circuit TV cameras designed for the purpose. Alternatively, film
cameras can be passed through the sewer, taking frequent pictures. The cost of televising
or photographing is such that it is done only where it is necessary to see breaks, leaks, or
other suspected abnormal conditions not visible from the manholes. Sewers to be televised
generally must be cleaned first, especially if they are of the smaller sizes.
Sometimes it is necessary for men to walk through a large sewer for a detailed examination
in respect to physical conditions. Sewers 45 inches in diameter and larger have been in-
spected in this way (Figure 4-1). Occasionally large sewers have been inspected by using a
boat (4). The hazards are such that this kind of work should be done only under the direct
supervision of an experienced person. Self-contained breathing apparatus is necessary unless
satisfactory air conditions can be assured.
FIGURE 4-1
WALKING THROUGH
LARGE SEWER
FOR INSPECTION
4-2
-------
4.3 Flow Measurements
Beyond visual inspection and analysis, it is frequently desirable to measure flow velocities.
One way of accomplishing this is by adding some type of tracer to the wastewater. Often
a dye is used. The slug of dye added should be the smallest amount that can be seen at the
downstream observation point. Then averaging the "first show" and "last show" of the dye
gives a close estimate of the flow time. Ideally, the first show and last show should be, for
this purpose, the times that the dye concentration is 20 percent of the maximum (6). No
attempt should be made to reinforce the dye to compensate for attenuation as the
wastewater moves from manhole to manhole, because this leads to errors. Reflections of
sunlight into the manhole will aid in seeing the dye, but observation in manholes at depths
greater than 15 ft is difficult. It may be necessary under some conditions for a man to
descend into the manhole to take samples or to observe the dye, or samples may be
brought to the surface at frequent intervals by use of a bucket. In this case, each sample is
poured into a jar. The jar with maximum color is selected, diluted with sufficient waste-
water containing no dye to give 20 percent of the maximum color, and compared with the
other jars to estimate the times that the dye concentration was 20 percent of maximum.
Another tracer frequently used is salt, using a conductivity meter to sense it. The down-
stream conductance reading should reach a peak at least twice and preferably five times the
normal conductance. Radioactive tracers are also used. In this case the usual practice is to
pump a stream continuously from the sewer through a counter. Any of the tracer methods
will yield good determinations of velocity if properly applied.
There are several methods for measuring flow quantities. One method is velocity times
cross section area of the stream. The stream in a sewer is accessible for measurement only
at the manholes, which may not give reliable information as to the average condition for the
whole reach. Even the averaging of measurements at several manholes might not give
dependable results, because of possible consistent errors. No reliance can be placed upon
such measurements where the stream is shallow. In large sewers, the errors caused by
small irregularities are less important; the results, while not very accurate, may be acceptable
for some purposes.
The method of greatest utility is the use of the slab type Palmer-Bowlus weir (7). This is a
flat weir that yields results corresponding closely to the theoretical calculations. In
comparison with a sharp-crested weir, it can measure a greater range of flows in a given
sewer, and the broad crest does not collect stringy material as a sharp crest does. It is easy
to install, since it is merely laid in the outlet pipe. Figure 4-2 shows one of these weirs
constructed for use in an 18-in sewer. This size can be let down through the manhole
opening. Larger sizes have been split longitudinally and hinged, or let down in two pieces
and bolted together. It is possible to install split weirs of this type in pipes up to 36 in
in diameter if it is done during the early morning low-flow period.
4-3
-------
FIGURE 4-2
PALMER-BOWLUS SLAB-TYPE WEIR FOR 18" PIPE
Accurate measurements of flow can also be made by the dilution method. An illustration of
this method is the running of a small stream of concentrated salt solution at a constant rate
into the sewer, with sampling and analysis of the waste water at a downstream location. If
the chloride concentration of the salt solution and its rate of flow are known, and the
increase of chloride in the wastewater above the background level is determined, the flow
can be easily calculated. Lithium salts have been used for this purpose in wastewaters of
high salinity since the background lithium concentration is seldom more than a few tenths
of a milligram per liter.
A modification of the dilution method is to add the salt solution all at once rather than as a
steady stream. Dye is added to the salt to spot the time that it passes a sampling point.
The wastewater must be sampled at intervals of a few seconds. The area under the curve of
concentration vs. time is used to calculate the flow. Flow and velocity are determined
simultaneously. This method has been described in detail (6). Slug additions of radioactive
tracers are also used to measure flows, this method being especially valuable for very large
flows where it would be difficult to obtain an adequate response by use of salt.
Simple volumetric methods such as timing of the filling of the wet well of a pump station
can also be used to calculate flows.
4-4
-------
Installation of a Palmer-Bowlus weir (Fig. 4-2) in a small sewer allows the use of a stage
recorder to obtain a continuous flow record. In sewers too large for the installation of
weirs, the water level can be recorded, and occasional velocity and flow determinations
by some other method can allow the establishment of a correlation between water level
and flow. A valid correlation may not be obtainable if levels are variably affected by
backup or drawdown because of hydraulic conditions downstream.
Procedures and equipment for measuring flow are discussed in detail in the Handbook for
Monitoring Industrial Wastewater (5). This document should be consulted before such
work is undertaken.
4.4 Character of the Wastewater
Where an investigation is made for the purpose of identifying and locating the source of
high sulfide concentrations, the tests most often made include temperature, pH, total
sulfide, dissolved sulfide, dissolved oxygen, BOD, COD, and oxygen reaction rate. It
would not be fruitful to make all of these tests at all manholes. A reconnaissance of
most of the manholes in the study area, making only sulfide tests, will aid in the selection
of key manholes for more detailed study.
The diurnal cycle of sulfide concentrations must be taken into account in interpreting the
results of an investigation. In a sewer where sulfide occurs, the usual pattern shows
minimal amounts of sulfide at the early morning low-flow period, since this is also the time
of minimal BOD. In small sewers the minimum may be zero. A rise of sulfide concentration
begins after the start of the daily flow increase, generally within a half hour. After
reaching a peak, sulfide concentrations usually decline somewhat to a plateau lasting,
in a typical case, for six hours.
The pattern is often less distinct iji very large trunks. The daily rise of sulfide concentration
may lag two to three hours behind the rise of flow, because the wave of increased flow
travels faster than the actual velocity of the water itself. The plateau may last eight to
twelve hours. If a system contains long force mains or if sewers are subject to surcharging,
the pattern may be quite different. In some systems erratic sulfide peaks may appear
because of industrial waste discharges or other causes.
Figure 4-3 shows the results of one 24-hr test in a series of tests made in 1962 in a trunk
sewer in Memphis, Tennessee, (8) illustrating some of the characteristics just described.
The wastewater in this trunk occasionally showed sharp sulfide peaks that sometimes
exceeded 3 mg/1, probably from an industrial source. The small peak at 7:00 in Figure 4-3
may have been from that cause.
Because of the diurnal sulfide cycle, daytime tests do not represent average conditions.
However, adequate results for control purposes or for a general reconnaissance of sulfide
conditions can be obtained from tests made during the sulfide plateau. An appropriate
hour for routine testing can be determined from intensive testing over a period of hours, or
a suitable time may be estimated from the general patterns described in the preceding
paragraphs.
4-5
-------
Instrumental methods are sometimes used for continuous monitoring of sulfide con-
centration. The following devices are available:
1. A simple type for measuring r^S in air uses the principle of periodically
drawing air through a strip of paper impregnated with lead acetate or other
sensitive metallic salt. The strip moves after each exposure, so that if sulfide is
present a row of spots is produced. The lead sulfide stain oxidizes and may
fade appreciably in a few hours. A variation of this method measures the density
of the color photometrically and records this result, thus eliminating the problem
of fading. The results do not permit an estimate of the t^S concentration in the
wastewater, because the concentration of t^S in air is dependent not only on
the concentration in the water but also on the turbulence of the water and the
turbulence of the air.
FIGURE 4-3
TYPICAL 24-HOUR SULFIDE TEST
IN A 60 INCH TRUNK SEWER
DISSOLVED SULFIDE'
20
20=00 0:OO 4:00 8:00
HOUR
I2:OO
16:00
•s
K
£
8
H-l
2. A modification of the first method provides for the automatic taking of samples
of wastewater, which are then acidified and purged with an inert gas to transfer
sulfide to the strip. Even though the method is only roughly quantitative, the
results may be useful for such purposes as tracing sulfide peaks to the source.
3. H2S removed from a sample of wastewater as in Method 2 is carried by the gas
stream into a solution where it is electrolytically titrated in the presence of a
bromide salt.
4-6
-------
4. Dissolved sulfide concentrations can be calculated from automatic recordings
of the potential of the Ag— Ag2$ electrode and a pH electrode. The electrodes
must be cleaned and calibrated daily. The calibration is best done by making
standard colorimetric determinations of dissolved sulfide in the wastewater being
instrumentally monitored. The electrode system should be adjusted to give results
equaling the standard method. For this method of calibration, the wastewater
should have a dissolved sulfide concentration in excess of 0.5 mg/1.
After an intensive survey of a sewerage system to identify sulfide conditions, it may be
concluded that there are key manholes that should be watched, especially if sulfide control
procedures are being applied. These key manholes will then be sampling points for weekly
or daily sulfide tests.
If there is a pressure main in the system, the operating characteristics of the pump station
should be determined, and ample data should be obtained on sulfide concentrations into the
wet well, at the pump discharge, and at the end of the pressure main. It is usually desirable
to make at least one 24-hour test. During that test, the off and on times of the pump
should be recorded. The wet well volume between high and low water should be calculated.
From the data, calculations can be made of pumping rate and detention times in the pressure
main. The detention time in the main can be checked by use of a dye, preferably injecting
the dye solution at the pump discharge with the aid of air pressure.
4.5 References
1. Pomeroy, R. D., Sewer Maintenance in Long Beach, California, Sewage and Industrial
Wastes, 29, pp. 320-325 (March, 1957).
2. Sewer Maintenance, Water Pollution Control Federation Manual of Practice No. 7,
Washington, D. C. (1966).
3. Safety in Wastewater Works, Water Pollution Control Federation Manual of Practice
No. 1, Washington, D. C. (1969).
4. Brown, Reuben F., North Outfall Sewer Inspection, City of Los Angeles, Sewage
Works Journal,!, pp. 446-454 (1937).
5. Handbook for Monitoring Industrial Wastewater, U.S. EPA, Office of Technology
Transfer, Washington, D.C. (August 1973).
6. Pomeroy, R. D., Flow Velocities in Small Sewers, Journal Water Pollution Control
Federation, 39, No. 9, pp. 1525-1548 (1967).
7. Design of Sanitary and Storm Sewers, Water Pollution Control Federation Manual of
Practice No. 9, Washington, D. C. (1969).
8. Pomeroy, R. D., A study of Sulfide Conditions and Concrete Corrosion In Sewers of
City of Memphis, Tennessee, Report to American Concrete Pipe Association
(August 26, 1963).
4-7
-------
CHAPTER 5
CONTROL OF SULFIDE IN EXISTING SYSTEMS
5.1 Improving the Oxygen Balance
It is evident from the discussion presented in Chapter 3 that the supply and demand of
oxygen is a major determinative factor in respect to sulfide buildup. Little or no buildup
is expected if oxygen in the stream is not depleted below 1 mg/1. It is important to operate
a system in a way to avoid unnecessary oxygen depletion, and measures can be taken to
supplement the oxygen supply where it is inadequate. Relatively small oxygen supplements
may be remarkably beneficial.
5.1.1 Avoiding Unnecessary Oxygen Depletion
Pump stations should be operated in such a way that wastewater is not backed up into the
influent sewers, because any substantial amount of backing up will cause oxygen depletion
and sulfide generation. If the sewer is used for flow equalization over a period of hours,
severe sulfide problems may result. The wastewater is retained out of contact with the air,
and because of the low velocity, organic solids slowly roll along the bottom, creating ideal
conditions for sulfide generation. If inadequate pumping capacity makes surcharging
unavoidable, the installation of adequate capacity should be high on the list of priorities.
Where the influent sewer to a pump station is used in part in lieu of a wet well, results may
not be serious if the sewer is not completely filled at high water and if the retention time is
short. Nevertheless, the condition is one to be avoided if possible. It is desirable to allow
the pump to start and stop at short enough intervals to avoid backup. The frequent starting
is not harmful to properly designed equipment.
The possibility of maximizing the aeration that occurs as wastewater falls into a wet well
deserves consideration. This subject is discussed in Section 5.1.7 in connection with
methods of supplementary aeration.
5.1.2 Injection of Compressed Air into Force Mains
5.1.2.1 Basic Concept of Air Injection, Transport and Absorption
Substantial sulfide generation must be expected if wastewater is held in a completely filled
sewer for more than a half hour, or often for only a quarter of an hour. The injection of air
into such a sewer at a low point, so that the air will travel upward in the pipe, is often a
useful way to prevent sulfide buildup and to oxidize sulfide already present in the waste-
water.
Air in a completely filled sloping sewer moves principally as large, discrete bubbles. If the
sewer is nearly level, the air moves easily, the bubbles being long and flat. They produce
little turbulence since little energy is dissipated, and little oxygen is dissolved. Under this
5-1
-------
condition, the injection of air is not likely to be effective for sulfide control. If the pipe is
steep, the bubbles of air are short and fat. For the same distance traveled, they must
dissipate more energy; there is more turbulence, which creates new air-water interfaces and
dissolves more oxygen.
The velocity of the air relative to the water is little affected by slope or rate of air injection,
but increases with sewer diameter (1). An estimate of the velocity is obtainable from the
following tentative equation:
ua = 3.0 xD
a
where:
ua = velocity of the air relative to the water, ft/sec
3.0 = empirical coefficient
D = pipe diameter, ft
5.1.2.2 Basis of Design
The effect of pressure in a force main causes the rate of dissolution to increase somewhat
more than in proportion to the power input because of the greater solubility of gases at
high pressure. However, nitrogen is dissolved also, and when the pressure is reduced as the
wastewater moves up the pipeline nitrogen comes out of solution, carrying oxygen out with
it. The overall oxygen transfer efficiency does not increase very much at pressures beyond
30 psi unless the detention time of the wastewater is long enough so that there is substantial
biological utilization of the oxygen before it can regassify.
The energy dissipated by air moving up an inclined filled pipe is equal to the energy of
isothermal expansion of the air. The power of isothermal expansion of a gas is shown in the
following equation:
(hp) =0.148Q
where:
(hp)a = power of isothermal expansion in the compressed air stream, hp
0.148 = factor to convert loge to logiQ and cubic foot-atmospheres
per minute to horsepower
Q,, = flow of air measured at ambient pressure, cfm
5-2
-------
P} = ambient pressure, psi
?2 = gage pressure at the injection point, psi
14.7 = standard atmospheric pressure, psi
When there is no flow of water, this energy is dissipated in turbulence, except for some loss
of volume due to gas dissolving in the water.
The power efficiency of this method of aeration has been found to be similar to the
efficiencies of other aeration methods (1). For the case where water is not moving, it may
be expected that between 2 and 3 pounds of oxygen will be dissolved for each horsepower-
hour of potential energy in the compressed air delivered to a force main under typical
operating conditions. The efficiency will be influenced by the effect of pressure on the
solubility of oxygen, and by the biological utilization of the oxygen within the force main.
There are energy losses involved in converting the electrical energy supplied to the motor to
potential energy in the injected air. One of these losses is the effect of a compression
process that is largely adiabatic rather than isothermal. In a typical case, with a well
designed system, 30 to 40 percent of the electrical energy input appears as potential energy
of isothermal expansion in the injected air.
The required potential power of the compressed air supply to provide dissolved oxygen at a
rate that will equal the oxygen consuming capability of the wastewater is equal to the
oxygen requirement divided by the units of oxygen supplied per unit of power, as shown
below:
(62.4 x 10~6)VRr
(hp)^ = — (water column stationary)
where:
(hp)^ = potential power of the compressed air to supply the oxygen demand, hp
62.4 = density of water, Ib/cu ft
JQ — e = conversion of ppm to parts per part
V = volume of the pressure main, cu ft
Rf = oxygen reaction rate, mg/l-hr
2.5 = Ib of oxygen dissolved per hp-hr in typical case
5-3
-------
If the water is moving, the air is partly carried along by the water, and the distance that the
air moves relative to the water is thereby decreased. A part of the energy of the air goes to
help lift the water, causing a slight reduction in the pumping head. The amount of oxygen
dissolved is proportionately decreased.
For this more general case where there is wastewater movement, the total potential power
required is given by the following equations:
t:62.4xlO~6) VRr
«
where:
ua = velocity of the air relative to the wastewater, ft/sec
uw = velocity of the wastewater, ft/sec
By equating (hp)a and (hp)b, the following expression is obtained:
(17 x 10~6) VRr
This formula is a simplified model for a rather complex process involving a number of
approximations but, since it is based upon maintaining an excess of oxygen, it will generally
indicate more than enough air for sulfide control. In actual practice, sulfide control is often
accomplished with a half or third as much air as would be needed to fully supply the
oxygen-consuming capability of the wastewater.
At pressures greater than 40 psi, the effect of pressure on the dissolution of both nitrogen
and oxygen is so large that the equation is no longer useful. It may be possible to calculate
the actual dissolved oxygen curve for the pipeline at various injection rates, or the required
air injection rate may be determined by trial when the system is in operation.
Use of the equation requires insertion of a value for Rf. The best way to obtain a value is to
make several measurements using the apparatus and procedure described in Sec. 3.2.3 or
equivalent commercial apparatus. If this cannot be done, rates may be assumed using the
following table as a guide.
TABLE 5-1
SUGGESTED OXYGEN REACTION RATES
Age of Wastewater Suggested Rate, Rf (conservative)
mg/l-hr
1 hr at 20°C or Vt hr at 30°C 5
2 hr at 20°C or 1 hr at 30°C 10
Over 3 hr at 20°C orl&hr at 30°C 15
5-4
-------
Special consideration must be given to situations that are out of the ordinary. For example,
if most of the rise occurs in the early reaches of the pressure main, followed by a long, nearly
flat run, there might be oxygen depletion and sulfide buildup in the latter portion.
In some cases, prevention of sulfide buildup may not be the only objective of air injection.
In long pressure mains, substantial BOD reduction is also possible. In three force mains
operating with air injection, BOD reductions of 16 mg/1 (1), 60 mg/1 (2), and 10 mg/1 (3)
have been reported.
In a very long high-pressure main, a sufficient oxygen supply to prevent sulfide generation
for the entire length may result in initial oxygen concentrations that will support a fairly
high rate of reaction on the slime layer. An advanced degree of biological oxidation would
therefore occur by the time the end of the force main is reached. The possible reduction of
BOD can be estimated, but the factors are complex and the subject is beyond the scope of
this manual. If there is no premium on biological oxidation and the objective is only sulfide
control, the point of air injection may be located part way up the pipeline. In one currently
planned 25,000-ft force main, air will be injected at a point 8,000 ft from the discharge end.
Air injection into an existing force main may be difficult when the main has an irregular
profile that could result in air blocks. In some cases it is feasible to pump against the air
blocks. The total dynamic head when air is injected into a force main of irregular profile is
equal to the sum of the dynamic heads of the rising legs, ignoring the descending legs, since
in these legs a partly filled condition with gravity flow will prevail. (An exception should be
made for a reach that has a downward slope less than the slope of the hydraulic grade line.
In the calculations such a reach should be considered as part of a rising leg.) If the pumps
can provide a satisfactory discharge against the sum of the dynamic heads of the rising legs,
then the air blocks will not prevent a normal pumping operation.
One solution for the hydraulic problem created by an irregular profile is multiple injection
points and release of air at high points. Air release valves do not operate satisfactorily in
typical wastewater but an atmospheric break is often practical. In cases where this would
require a standpipe of excessive height, an alternative is a continuous bleed of air and water
from a high point, returning the water to a suitable point in the wastewater collecting
system. Both of these methods have been used satisfactorily (2). Another alternative for
accomplishing sulfide control in a force main of irregular profile is injection of air beyond
the last low point, provided there is enough detention time from that point to the end of the
pressure section to accomplish oxidation of sulfide produced in the unaerated portion. This
plan is used in a force main serving a small residential area in Beaumont, California. The air
injection point is only 300 ft from the end of the force main. In view of the short distance,
the last 300 ft was made of 10-in pipe, providing a retention volume of 1,200 gal. This is
twice the storage volume between high and low water in the pump station wet well, thus
assuring that the contents of the aerated section will not be displaced by a single pumping.
If dissolved sulfide concentrations are not kept at zero within a force main that has a high
point, with an air pocket, air injection may cause corrosion at that point. Even minor
5-5
-------
irregularities of grade may permit the collection of pockets of air. Bubbles entering an air
pocket cause turbulence that accelerates the release of hydrogen sulfide. Two pressure
mains have failed because of corrosion at air pockets. Another has operated satisfactorily
for 25 years with an air pocket, because the dissolved sulfide concentrations were kept
low (2).
In designing air injection systems the following guidelines should be considered:
1. Continuity of treatment. The force main produces sulfide continuously, whether
the wastewater is moving or static. Air injection should therefore continue
whether the pump is on or off, except for short off-and-on periods to modulate
the treatment.
2. Point of injection. For maximum utilization of the air, the point of injection
should be at a low elevation. If there is a vertical riser from the pump, the air may
be injected near the bottom of the riser, but it must be beyond the check valve
and in a location where air cannot get back into the pump.
3. Manner of air injection. Air diffusers have sometimes been used. In a short
distance the air collects into large bubbles, regardless of the manner of injection.
If the aeration occurs mostly in a vertical riser, the diffuser may be beneficial, but
in a sloping force main it is not of significant value. The usual practice is merely
to pipe the air to a suitable connection into the main.
4. Choice of compressor. A major consideration is to choose a compressor suitable
for long-time continuous operation with minimum maintenance. If specifications
for a competitive bid for a compressor only call for equipment to compress air at
a given rate to a given pressure, the low bidder is likely to supply a small, high-
speed unit that will soon break down. Either minimum cylinder displacement or
maximum speed should be specified. For small units the speed should not exceed
1150 rpm. For larger units the speed should be slower.
5. Provision for adjusting air rate. An air injection installation is generally designed
to supply air at a rate greater than is found to be necessary in actual operation,
or greater than is needed continuously. There is not ordinarily any objection to
an excess of air, but for economy of operation and maintenance, provisions
should be made for modifying the rate. There are several ways to do this. Often
an air receiver or reservoir is used, with the air fed to the force main by way of a
throttling valve and a rotameter or other flow indicator. A pressure switch either
starts and stops the compressor, or, more commonly, operates a cutout that
lifts the intake valves so that the compressor idles. Either method is wasteful of
power. The air is over-pressurized in the receiver, and a surprising amount of energy
is wasted in keeping the machinery running during the idling phase. If the size of
the installation is small, the power requirement may be of little consequence, but
in large units a method should be used that is less wasteful of energy. Variable
5-6
-------
drives are a possibility. Another method is to inject air intermittently by using a
program timer to start and stop the motor or to cause the compressor to idle
intermittently. If the detention time for the water in the force main is long in
comparison with the off-on cycle of the compressor, air bubbles will overtake
water that was pumped in while the compressor was not running. Results will be
similar to continuous aeration at a reduced rate. This method is used successfully
in several installations of the Los Angeles County Sanitation Districts.
6. Standby. It is not customary to install standby compressors. Temporary odor
control measures can be instituted if a compressor is out of service, or a portable
compressor can be brought in. In large installations, however, it may be advan-
tageous to have two compressors, one twice as large as the other. This gives a
selection of three air injection rates.
7. Air flow measurement. A rotameter or other visual flow indicator should be
available so that performance of the compressor can be observed. It should be
installed in a bypass through which air does not normally flow.
5.1.2.3 Example
Descriptions of several air injection installations have been published (1) (2), reporting, in
most cases, sulfide concentrations with and without air. A hypothetical air injection design
problem is presented here to illustrate the foregoing principles.
PROBLEM
An inverted siphon has a rising leg 2,000 ft long, shown schematically in Figure 5-1. It, is a
single 12-in pipe continuous with the sewer, and carries an average daytime flow of 1.2 cfs.
The total dynamic head at the low point will be 30 ft. Daytime tests made in the summer
with samples of the wastewater entering the siphon show an average oxygen reaction rate
(Rr) of 11 mg/l-hr. The location is near sea level. Design a system to inject air at the low
point to supply enough oxygen to meet the oxygen demand rate.
FIGURE 5-1
SCHEMATIC DRAWING OF FORCE MAIN FOR PROBLEM
AIR INJECTION POINT
5-7
-------
SOLUTION
(17 x 1(T6) VRr
14.7
ua + uw
1.2
1. Average daytime velocity of wastewater = uw = Q/A = ^ 7
-------
Information on the use of oxygen in a force main has been reported by the Water Pollution
Research Laboratory at Stevenage, England (4). The rise of the main is 90 feet. Oxygen is
supplied at a dosage of 75 mg/1 and complete dissolution is accomplished by feeding it into
the pump. In contrast to the practice where air is used, oxygen is supplied only while the
wastewater is being pumped into the main. In this manner all of the wastewater is given a
fixed dosage of oxygen. Residual dissolved oxygen at the end of the main averaged 16mg/l.
Of the oxygen used, 28 mg/1 reacted in the stream and 30 mg/1 reacted on the slime layer.
The Stevenage work illustrates the value of oxygen injection where maximum BOD
reduction is desired, because the rate of reaction on the pipe wall is proportional to the
oxygen concentration (5). Also a higher oxygen concentration can be carried as a reserve
in the wastewater leaving the end of the force main. Wastewater containing the normal
concentration of nitrogen can carry, in addition, 15 to 20 mg/1 of oxygen without bubble
formation or significant loss if the water does not cascade. The use of air in a force main
can seldom produce a residual of more than 2 to 4 mg/1 of oxygen.
Another method of accomplishing complete dissolution of oxygen is the use of a U-tube
through which the water would pass before entering the force main. This is discussed in
detail in Section 5.1.5.
5.1.4 Falls
Drops, or falls, and other places of high turbulence are notorious for release of odors from
wastewater, including hydrogen sulfide if any is present, and for the consequent corrosion
of susceptible structures. At the same time, falls cause dissolution of oxygen, which, given a
little time, will destroy the sulfide remaining in the stream. The amount of oxygen absorbed
by a series of falls is much larger than the amount absorbed by the smooth flow of the water
through the same loss of elevation (6). Thus, points of turbulence can be both harmful and
beneficial.
It is possible to preserve the benefits of a fall while guarding against harmful effects even
where F^S is present. This means that corrosive and odorous gases must be trapped in some
way. Sometimes this can be done quite simply and inexpensively, as illustrated by the
following discussions.
The Sacramento County Sanitation Districts (California) had a force main discharging
into a manhole at an elevation four feet above the invert. The manhole is the temporary
upper terminus of the sewer. Sulfide concentrations in the wastewater were not very high,
generally less than 1.0 mg/1, but the turbulence of the fall released so much l-^S that there
were odors in the vicinity and serious corrosion of the manhole. To remedy this condition,
the Districts installed a PVC drop pipe essentially as shown in Figure 5-2. Air is drawn
down the pipe by the falling water, and as the water leaves the pipe at high velocity it
drives the air on down the sewer. There is no recycling of air into the manhole; rather,
additional air is drawn into the sewer. The manhole is now dry and completely free of
odor. The sewer is made of vitrified clay pipe, and is unharmed by the J^S. By the time
the wastewater reaches the next manhole the sulfide has been completely oxidized.
5-9
-------
Two features of the installation should be noted: 1) the discharge of the drop pipe is
inside the sewer that carries the wastewater away, and 2) the bottom of the drop pipe is a
smooth curve, so that the wastewater conserves its momentum and discharges at high
velocity.
Pressure mains are commonly designed so that they enter the receiving manhole close to
the bottom. By minimizing the fall the problem of possible f^S release is minimized.
However, by properly trapping the released gases, and by confining them in a corrosion-
resistant structure, as shown in Figure 5-2, the drop becomes a desirable feature.
FIGURE 5-2
H2S - CONTROLLING DROP STRUCTURE
AT THE END OF A FORCEMAIN
FORCE MAIN
It may sometimes be advantageous, therefore, to purposely raise the end of a force main
several feet higher than the bottom of the manhole, so as to use this simple and inexpensive
method to dissolve oxygen and prevent sulfide problems.
5.1,5 U-Tube Aeration
5.1.5.1 Basic Concept
Basically, a U-tube consists of two vertical pipes connected by a U-bend at the bottom
(literal U-tube) or other functionally equivalent arrangement, including concentric pipes.
Air or sometimes oxygen, is dispersed into the wastewater in the descending leg, being
carried through the U-tube with the water, as shown in Figure 5-3. The relatively long
5-10
-------
FIGURE 5-3
U-TUBE DESIGN, JEFFERSON PARISH STATION 5
EXISTING FORCEMAIN
EXISTING
MANHOLE
VENTURI ASPIRATOR
THROAT I.D. 6.40"
AIR INLET
GRADE EL.43.2
EL32.7
EL 0.0
-------
contact time of the gas with the water and the increased pressure at the bottom due to the
water column make the U-tube an efficient aeration device. If sulfide is present in the
wastewater entering the U-tube, the oxygen that is absorbed will oxidize it, to some extent
in the U-tube but mostly in the downstream sewer. Where no sulfide is present, a U-tube
may nevertheless be used to provide a dissolved oxygen reserve to sustain aerobic conditions
downstream.
5.1.5.2 Variations
There are basically three ways that air can be supplied to a U-tube: 1. A venturi aspirator
may be placed in a pipeline leading to the U-tube, as shown in Fig. 5-3. This is a suitable
installation at the end of a pressure main. The venturi not only supplies air to the U-tube,
but also provides an initial intimate mixing of the air with the wastewater. The throat must
be small enough so that the negative pressure will aspirate the required amount of air at
times of minimum flow, yet not cause excessive head loss during maximum flow. These
conditions are difficult to meet where there is any substantial back pressure at the venturi
discharge (7). The venturi is a practical device to aspirate air if it can be placed higher than
the hydraulic head at the U-tube outlet. Sometimes this may mean raising the venturi out
of the ground.
If air in moderate amount is suitably diffused and the velocity in the down leg is not too
slow, the air remains dispersed. If the air input is increased, a condition is reached in which
a pocket of air is formed in the top, which may soon largely fill the column. Evidence of
this effect is seen in Figure 5-4 which shows the head loss in the down leg of the U-tube
shown in Figure 5-3. For each experimental condition, the total head loss measured
between two pressure taps in the down leg has been divided by the vertical distance to show
the average energy loss in ft of wastewater per ft. Also, the theoretical head loss has been
calculated on the assumption of perfect dispersion of the air and a 0.2 ft/sec rate of rise (slip
velocity) of the air bubbles through the wastewater. The compression of the gas was taken
into account, but no correction was made for the amount dissolving. This calculated loss
is shown as a broken line. The experimental points are scattered because of the difficulty of
getting steady pressure readings but it is clear that a major departure from the theoretical
curve for complete dispersion of the air begins when the air input exceeds 10 percent of the
wastewater flow. Soon the wastewater behaves almost as though it were in free fall, where
the head loss approaches 1.0 ft/ft of fall. In these tests, the U-tube operated with a nominal
downward water velocity of 1.5 ft/sec. If water velocity had been quite slow, it can be
assumed that an air pocket would have formed at a lower air-water ratio. Also, if the air
had been ineffectively dispersed, so that it was present as relatively large bubbles, there
would have been a greater tendency to form an air pocket.
Another way to supply air or oxygen to a U-tube is by injecting it under pressure into the
top of the descending leg. Even with a diffuser ring, a gas pocket is expected if the gas
input is more than 10 percent of the wastewater flow. The wastewater will fall through
this space, and gas will be entrained by splashing and will be carried through the U-tube.
5-12
-------
0.0
FIGURE 5-4
HEAD LOSS IN DOWN LEG OF 42 FT U-TUBE
0 O.I 0.2 0.3 0.4 0.5 0.6
AIR-.WATER INPUT RATIO
5-13
-------
By injecting gas under pressure the head loss through a venturi is reduced, but this reduction
may be more than offset by a greater head loss in the U-tube. The choice of a venturi or
compressed gas will depend partly upon whether air or oxygen will be used, and also upon
the required dissolved oxygen input and wastewater flow patterns. The compressed air
or oxygen U-tube is likely to be most suitable for use at pump stations, especially in
situations where the force main is nearly level and the discharge pressure is low. In this
case ordinary air injection would be ineffective.
A third way to operate a U-tube is for entrainment alone to supply the air input, providing
only an opening at the top to allow the air to enter. The amount of air entrained will be a
function of the available head (net elevation loss). It is not likely that entrainment alone
will meet the purposes of a U-tube .if the head is less than 10 ft. No actual installations are
operating in this mode.
In addition to the three ways discussed above for supplying air to a U-tube, there is also
the possibility of forcing air into the rising leg. If no air is entrained on the inlet side, the
device is a simple air lift pump or eductor which is discussed in Section 5.1.6. If there is air
entrainment in a descending leg as well as air injection into the rising leg, the effect is a
hybrid operation. This condition, too, will be considered under the heading of air lifts.
5.1.5.3 Design Considerations
The amount of dissolved oxygen that must be supplied to the wastewater, the available head,
the quantity of flow, and variations of flow are factors that influence the design of a U-tube.
The oxygen reaction rate of the wastewater (or the 15-minute "immediate dissolved oxygen
demand") should be determined several times by use of an amperometric probe to ascertain
average conditions and extremes. Similarly, sulfide conditions should be thoroughly
explored.
For large flows, such as 25 cfs, concentric tubes could be used for the U-tube. For small
flows, concentric tubes would probably lead to fouling with stringy material in the
wastewater, so literal U-tubes are used. The rising leg must be small enough so that the
velocity will carry away sand or small pebbles that may be in the wastewater. No difficulty
has been experienced where the upward velocity is 4 ft/sec or more. Where the flow is
variable, there will probably be no difficulty if the velocity reaches 4 ft/sec at least once
each day. In the descending leg, a lower velocity is desirable to increase the time of contact
of the air with the water.
In U-tubes presently in operation, the nominal average velocity in the descending legs, that
is, average flow of wastewater divided by cross-section area of the pipe, is about 1.5 ft/sec.
The depths of U-tubes now in use range from 17 to 54 ft.
The amount of oxygen dissolved per unit of energy dissipated increases with the height of
the U-tube. However, nitrogen dissolves in increasing amounts too. As the water moves
up the rising leg, the decreasing pressure allows these gases to come out of solution again.
5-14
-------
The same effect in aerated pressure mains is discussed in Section 5.1.2. There is a U-tube
depth beyond which there will be little further improvement in over-all efficiency. The
economical optimum when using air is probably at a depth of 50 ft or more. With pure
oxygen, the depth should be great enough so that the gas dissolves completely. A way to
calculate the required height in this case is yet to be developed. If the applied oxygen is
completely dissolved to produce a concentration of 15 to 20 mg/1 (about 1 percent of
oxygen by volume), with no supersaturation with nitrogen, practically no oxygen will be
lost in the rising leg.
A basic requirement for U-tube aeration, as well as any other aeration device, is enough
energy to mix the air with the water. The U-tube is equivalent to other aeration methods
in this respect, and under favorable condition it is probably superior. Its usefulness does not
rest upon any uniqueness in this respect, but upon simplicity and applicability where other
methods might be unsuitable. However, the U-tube or some equivalent of a U-tube will
have a major advantage where commercially pure oxygen is used in a wastewater collection
system, because it is one way to provide the pressure that needs to be used in some manner
to accomplish 100 percent utilization of the oxygen. Where oxygen is to be used, it is
better to use a U-tube or other functionally equivalent pressure device rather than the
systems usually adopted where the gas to be used is air.
A computer program has been developed for calculating the behavior of U-tubes supplied
with air (7). A rough but conservative estimate of the amount of oxygen that can be
dissolved in a U-tube under suitable operating conditions using air can be made on the
basis of the net elevation loss or, conversely, an estimate can be made of the required
elevation loss for any specified dissolved oxygen objective, using Table 3-1 as a guide.
If a U-tube is installed at the end of a pressure main, and if there was originally little or no
fall at the end of the main, an increase of pumping head will be necessary to make the device
work. Either back pressure will be produced by injected air, or the inlet to the U-tube will
need to be raised to an elevation several feet above the outlet, or even raised out of the
ground. If the pumps do not have the capability of pumping the wastewater against this
head, it may be necessary to change impellers, and perhaps to change motors. Most
commonly, pumping installations can accommodate the relatively small increase of head.
The following two case histories illustrate design and performance features of U-tubes using
aspirated air, compressed air, and oxygen.
CASE HISTORY 1
Jefferson Parish, located on the west side of the City of New Orleans, is in an area of level
terrain and high groundwater. The original sewer system was built with minimum slopes
and with many pump stations. Sulfide odors and corrosion resulted. The most severe
conditions were at force main discharges where the wastewater fell several feet into the
manholes, and at pump stations where there was also splashing and release of odorous air.
5-15
-------
Figure 5-5 shows the trunk sewers and force mains in the collection system where the
installations were made. Average dry-weather flows, pumping rates in the force mains,
locations of aeration devices, and sampling stations are indicated.
U-tubes were installed in 1970 at the ends of two force mains (Stations 5 and 7), and an air
lift aerator was placed in a gravity sewer (Station 2). The project was undertaken by
Jefferson Parish with the assistance of a research grant from the U.S. EPA (7).
Figure 5-3 shows the design of the U-tube at the end of the force main at Station 5. The
U-tube at Station 7 follows the same general design, but the depth of the tube is 54 feet
instead of 42, the U-tube legs are 12 and 8 inches in diameter instead of 19 and 13.25 inches,
and the elevation difference from inlet of the U-tube to the outlet is 7 feet instead of
9.5 feet.
A significant improvement resulted at the ends of the force mains, since the water no longer
fell several feet into the manholes. Sulfide and dissolved oxygen tests were made before
and after activating the aeration devices. Results are shown in Table 5-2.
TABLE 5-2
RESULTS OF U-TUBE OPERATION IN JEFFERSON PARISH, LA. (7)
Before U-tube Use During U-tube Use
Station No. Dissolved Dissolved No. Dissolved Dissolved
Tests Oxygen Sulfide Tests Oxygen Sulfide
mg/1 mg/1 mg/1 mg/1
7 6 0.6 0.26 22 2.1 0.06
40 34 1.9 0.42 10 2.3 0.11
6 32 0.8 0.51 9 1.6 0.05
13 24 0.0 0.30 27 0.7 0.02
The flow that passed through the U-tube at Station 5 constituted 89 percent of the flow
sampled at Station 6, and 79 percent of the flow sampled at Station 13. Station 40 is a
pump station wet well. From 40 to 85 percent of the wastewater sampled at that point had
been aerated at Station 7, the proportion varying with the operation of the pump stations.
This wastewater also received some aeration at the wet well, both before and during U-tube
operation, by virtue of a fall of 3 to 6 ft.
At stations 7, 40 and 6, some dissolved oxygen was present without operation of the U-tubes
but it declined to zero by the time the wastewater reached Station 13. With the U-tubes in
operation, there was enough oxygen to prevent sulfide buildup and sulfide was reduced to
zero at Station 13 except for a few hundredths of a milligram per liter on a few occasions,
and one reading of 0.23 mg/1.
5-16
-------
FIGURE 5-5
MAIN TRUNKS RELATING TO U-TUBE INSTALLATION
IN JEFFERSON PARISH
TREATMENT
-• PLANT
30"
• PUMP STATION
X AERATION DEVICE
O SAMPLING LOCATION
. 1000 FT. .
O STA.I; 3.0m.g.d.
STA.2; 2.8 m.g.d.
STA.I3; 1.3 m.g.d.
21'
STA.5;I.I m.g.d.
-O—X- T
STA.6 I
rV.M.,l2"x3000';
PUMPING RATE
1 I5OO g.p.m.
.7; 0.3 m.g.d.
F.M.,IO'V 2050'; ~fc
PUMPING RATE 600 g.p.m.
F.M.,8"xi520 ;
PUMPING RATE 1000 g.p.m.
5-17
-------
Station 2 was an air lift alongside the sewer, mentioned in Sec. 5.1.6. It was operated for
only a short time; performance data are not available.
The 1970 installations costs for the U-tubes were:
Station 5—Aspirated-air U-tube installed on end of 12-in force
main, depth = 42 ft, $ 16,000
Station 7—Aspirated-air U-tube installed on end of 8-in force
main, depth = 54 ft, $ 11,000
CASE HISTORY II
The City of Port Arthur, Texas, is located'on terrain similar to that of Jefferson Parish,
Louisiana. According to prevailing standards in the area, most of the sewers were laid to
flow at velocities of 1.6 ft/sec when half full. The system includes 12 principal pump
stations, and as many more smaller ones. The discharge piping is very short in many
cases, the stations serving only as lifts. One force main 5,427 ft in length has produced
sulfide at an average concentration of 15 mg/1, causing very severe problems, and three
others of lesser lengths are important contributors. Nineteen sets of sulfide tests at
various other places in the system showed concentrations averaging 3.1 mg/1. Records of
10 corrosion failures of concrete sewers indicated corrosion rates averaging from 0.08 to
0.17 in/yr. An eleventh sewer, 30 in. in diameter, failed in five years, with an indicated
corrosion rate of 0.55 in/yr.
In a project made possible by a research grant from the U.S. EPA, several different schemes
were tried out for supplying oxygen to the waste water (8).
Two U-tubes were installed in 1967 at Pioneer Park lift station and two at Lake Charles lift
station. In both of these locations the discharge piping is short; therefore the problem is not
one of sulfide buildup in pressure mains. The objective of the U-tubes in these locations is
to supply enough oxygen to oxidize the sulfide already present and to provide an excess to
maintain aerobic conditions for some distance downstream.
In both cases the U-tubes are installed in the dry wells,-which impose height limitations.
Figure 5-6 is a representation of one of the U-tubes at Pioneer Park Lift station.
Venturi aspirators are not used in these installations. At Pioneer Park compressed air is fed
to collars around the down legs of the U-tubes, whence it is diffused into the water by way
of thirty-two Vs -in holes. At Lake Charles lift station oxygen is diffused into the U-tubes
by way of perforated copper tubes inserted into the pipeline at an angle directed down-
stream.
5-18
-------
FIGURE 5-6
U-TUBE CONFIGURATION, PIONEER
PARK LIFT STATION
FROM PUMP
UJ
UJ
FORCE MAIN
-AIR INJECTION
COLLAR
-18 DIA. DOWN LEG
-12 DIA. UPLEG
5-19
-------
At Pioneer Park lift station, injection of an amount of air equal to 8 percent of the waste-
water flow resulted in the dissolving of about 4 mg/1 of oxygen. The head loss through the
system was about 6 ft greater than when no air was used. During eight test at different air
rates the sulfide concentration at the pump discharge averaged 3.4 mg/1, and at the discharge
of the U-tube it averaged 2.0 mg/1. The difference is probably a reflection of the reaction
rate between sulfide and oxygen. The reaction of sulfide with dissolved oxygen no doubt
continues in the sewer. Odor conditions at downstream stations are, in fact, much better
than formerly.
At Lake Charles lift station, where commercial oxygen is used instead of air, the dissolution
of oxygen at various injection rates was as shown in Table 5-3.
TABLE 5-3
OXYGEN DISSOLVING EFFICIENCY
Oxygen Oxygen Oxygen
applied dissolved dissolved
mg/1 mg/1 percent
6.3 4.1 65
12.5 6.0 48
18.8 7.6 40
25.1 9.6 38
Sulfide concentrations in and out of the U-tube averaged 3.2 mg/1 and 2.2 mg/1 respectively.
Variations of oxygen injection rate had no significant effect on sulfide concentrations at the
U-tube discharge. However, at the highest oxygen injection rate the amount of oxygen
remaining at the U-tube discharge was apparently 5.5 mg/1, which probably destroys the
sulfide entirely in the downstream sewer.
5.1.6 Air Lifts
A U-tube in which the outlet is at an elevation equaling or exceeding the elevation of the
inlet, with air injected into the rising leg, is an air lift. Air lifts deserve more consideration
than they have received for use in low-lift pump stations. Their power efficiency as pumps
is low, but part of the lost energy is the result of slippage of air bubbles past the water.
This slippage greatly increases small-scale turbulence, resulting in the dissolving of oxygen
from the air. If the combined effects of pumping and aeration are considered, the efficiency
looks much better.
A zero-head air lift (inlet and outlet at essentially the same elevation) can be installed as an
aeration device alongside an existing sewer. In this case the water level would be drawn
down in the descending leg, so that wastewater would cascade into it. A partial dam would
prevent recycling of the water, or perhaps permit a limited amount of recycling.
5-20
-------
An air lift was installed along a sewer in Jefferson Parish, Louisiana. It was designed to
raise the wastewater above the elevation of the top of the sewer and let it fall back into the
sewer. The operation was objectionable because of odors produced by the air exhaling
from the sewer. This condition can be avoided if the compressor takes suction from the
atmosphere of the sewer. H^S in the air supply rarely has any effect on the compressor,
because the temperature of the compressor is high enough to prevent any condensation.
Droplets of wastewater must be kept out by proper filters.
As far as is known, no zero-head air lift aerator has been installed. The amount of oxygen
that would be dissolved can probably be estimated by calculating the potential power of
the compressed air (see Sec. 5.1.2.2), converting that power to a height of fall of the waste-
water stream, and using Table 3-1 to obtain an estimate of the expected percent of the
oxygen deficit that will be satisfied.
5.1.7 Aeration in Wet Wells, Special Tanks, and Retention Basins
5.1.7.1 Basic Concept
Since sulfide in wastewater can be oxidized by the introduction of oxygen, aeration in tanks
affords another way to destroy sulfide originating from an upstream source. The addition
of even a small surplus of oxygen by this method may preserve an aerobic condition
downstream where sulfide might otherwise appear. The tank used for aeration may be a
pump station wet well.
5.1.7.2 Basis of Design
Wet wells of pump stations frequently have been the cause of odor nuisances. The splashing
of the wastewater as it enters the well releases odors, especially if sulfide is present, and the
escape of air brought into the wet well by the sewer and the air displaced as the wet well is
filling disperse the odors into the atmosphere. Such odors have given rise to the belief that
the wet well is the site of sulfide generation, but rarely is this the case. More commonly,
sulfide concentrations decline because of the aeration of the entering water. Sulfide
buildup in the unaerated wet well is not likely to occur unless the detention time is at least
two hours, and the amounts produced are small unless the detention time is many hours.
Use of pump station wet wells as aeration basins is becoming more common, but no data
on operating installations are available. The basic theory for such an operation is now
better understood, so that installations can be rationally designed.
If water containing sulfide flows into a basin where it has a detention time of only a few
minutes, and it is aerated there, objectionable amounts of sulfide will be released into the
atmosphere. If the detention time is longer, say a half hour, oxidation can generally be
carried to the point where sulfide concentrations will be close to zero. Sulfide in the
incoming stream will be diluted to a low level and held there by a rate of oxidation
equaling the rate of addition.
5-21
-------
if aeration is to be undertaken in an existing wet well, the detention time should be
maximized by setting the controls on the pumps so that the well is not pumped down very
far. Tests should be made on the influent wastewater to determine temperature, sulfide
concentrations, rate of oxygen consumption, and rate of oxidation of sulfide. These
characteristics will vary diurnally, so should be determined at different times of the day.
They also vary seasonally. It is most important to know or to estimate the characteristics of
the wastewater at the season of maximum temperature. With this information, the oxygen
requirement can be determined and used as a basis for design of the aeration facility.
If the average detention time in the wet well is not long enough to effect substantially
complete sulfide oxidation, aeration may nevertheless add enough oxygen so that the
reaction will be complete in the discharge main or in the downstream sewer. In this case,
protective measures will be needed against corrosion and odors due to H^S in the wet well.
If the wastewater is free of sulfide, detention time is not important except insofar as it is
involved in getting the desired amount of oxygen into solution.
If aeration in a wet well is undertaken, it will usually be by use of spargers or diffusers.
The installation would be designed so that the compressor takes suction from the atmos-
phere of the wet well to minimize the escape of odorous air. Impurities in the air will not
harm the compressor if particulate matter is filtered out. The possibility of excessive
depletion of oxygen in the reused air may need to be considered, but this will rarely be a
problem, especially if the water level fluctuates due to intermittent pump operation.
Another way to accomplish aeration in the wet well is to maximize the fall of the incoming
water. This may be accomplished by proper adjustment of the pump controls to keep the
water level as low as is practical. This will require more frequent starting and stopping of the
pump, but this does no harm and does not materially reduce the efficiency if the pumping
intervals are not too short. A running time of 2 to 3 minutes at each pumping is suggested
as a reasonable minimum.
The method of maximizing the fall of the water is suitable where the incoming water
contains no sulfide, or very little, and the objective is to augment the oxygen supply.
Table 3-1 shows the oxygen uptake that may be expected. The amount may not seem
large, but it can effect a substantial improvement in dowstream sulfide conditions.
An alternative to the use of a wet well as an aeration basin is the construction of a special
tank. A steel tank erected for this purpose adjacent to a pump station at Port Arthur, Texas,
is shown in Figure 5-7. Aeration of the high-sulfide sewage in the wet well was considered
undesirable because of odor release in this location. Wastewater is pumped from the wet
well into the tank, where it is aerated. Part of it then falls back into the well. The waste-
water is kept aerobic and the detention time is long enough so that it is essentially free of
sulfide. Sulfide conditions are greatly improved downstream.
5-22
-------
Basins are sometimes constructed in large sewerage systems to equalize diurnal flow
variations. Adequate aeration is normally applied to control sulfide generation and to
provide mixing energy to minimize the settling of solids. Overflow basins to retain storm
flows are sometimes aerated for the same reasons.
5.1.8 In-Line Augmentation of the Oxygen Supply in Gravity Sewers
5.1.8.1 Basic Concept
The difference between a condition of sulfide buildup and one of no buildup is often
attributable to a quite small difference in the oxygen supply. Consider a sewer 3 feet in
diameter at a slope of 0.0012 flowing half full at a velocity of 3.4 ft/sec. If the dissolved
oxygen concentration is near zero, the rate of supply by surface aeration is 1.6 mg/l-hr.
For EBOD = 350, the expected rate of sulfide buildup if no sulfide is present initially
will be 0.52 mg/l-hr. The rate of oxygen supply that would hold sulfide concentrations
to a few tenths of a milligram per liter would be roughly 2.0 to 4.0 mg/l-hr, or 1 to 2
mg/1 per mile of flow. Thus, aeration devices capable of supplying oxygen supplements
of 1.0 mg/1 at 1-mile intervals would probably prevent sulfide buildup. The importance of
such small oxygen increments is demonstrated by the evident role of junctions and
other points of turbulence in curbing sulfide generation.
FIGURE 5-7
AERATION TANK AT PUMP STATION,
PORT ARTHUR, TEXAS
5-23
-------
If in-line oxygen supplementation is to be undertaken, a decision as to the amount of oxygen
to apply must be made, and this requires that the objectives be clearly stated. The purpose
might be only to destroy the sulfide already present in the stream, or it might be desirable
to provide a surplus of oxygen to sustain aerobic conditions for a certain distance down-
stream. Another possibility that may be considered is the maximizing of BOD reduction.
This would be beyond the requirements for sulfide control, but if oxygen is to be supplied,
the chance to attain a significant reduction of BOD at the same time should not be over-
looked. In large trunk sewers, the wastewater may develop a capability to use oxygen at a
rather high rate, as previously shown by Figure 3-9. In smaller trunks, oxidation on the
slime layer may be a major factor, especially if the oxygen concentration is kept high (4)
(5). Thus, an important degree of BOD reduction is possible if an aerobic condition is
sustained by repeated oxygen feeding.
Before making design decisions for a facility to supply dissolved oxygen, it would be
desirable to test the wastewater several times at various times of the day and preferably at
different seasons, to determine sulfide concentrations, oxygen reaction rate in the waste-
water, and rates of oxidation of sulfide. Then the desired rates of oxygen input can be
determined, and the oxygen-supplying facility can be designed accordingly.
If the objective is only to destroy a certain amount of sulfide already present in the
wastewater, it generally will not require more than four pounds of dissolved oxygen for
one pound of sulfide, and sometimes only two for one. The maintaining of enough
oxygen in the water to prevent sulfide buildup in a typical case is likely to require an
oxygen supply only a half or a third as great as the oxygen reaction rate of the waste-
water in the presence of an excess of oxygen.
5.1.8.2 Possible In-Sewer Methods for Adding More Oxygen to the Stream
Various methods for in-sewer aeration have been considered, including diffused air by way
of a perforated pipe along the bottom of the sewer, aeration in the bottoms of manholes,
and mechanical aerators. As far as is known, no process of this type is in operation, and the
prospects for practical application do not look promising.
The use of commercial oxygen has also been explored. It has a number of advantages in
comparison with air for in-sewer application.
One idea for applying pure oxygen is to use it to enrich the sewer atmosphere. The effect
of a pure oxygen atmosphere over a short distance, however, would be negligible. In the
very large trunks a pure oxygen atmosphere would not prevent oxygen starvation in the
stream. The practical problems of attempting to fill the free space in a major trunk sewer
with oxygen and maintaining it there without excessive losses appear insurmountable, and
even if it could be done, the expected benefits would not be commensurate with the cost.
The bubbling of oxygen into the wastewater stream would be inefficient, since most of the
oxygen would escape unused. The enriched sewer atmosphere could be recompressed and
diffused into the water again, but it does not seem likely that this would be a practical
operation.
5-24
-------
A method that does show promise for adding pure oxygen to the stream in a free-flowing
sewer is to produce a high concentration of oxygen in a side stream under pressure, and then
release this to the main flow. An essential feature of the method is the maintenance of full
pressure all the way to a submerged nozzle where the oxygenated water escapes at high
velocity. In this way no oxygen can come out of solution in the piping, and the high
velocity of the jet causes immediate dilution with a large volume of the main flow, so that
no bubbles are formed, or only extremely small bubbles that immediately redissolve.
The method was first employed in 1970 to supply supplemental oxygen in a small privately
owned wastewater treatment plant in the City of Simi Valley, California. Preliminary
experiments with the method have been conducted recently in a sewer of the Los Angeles
County Sanitation Districts.
Figure 5-8A is a schematic drawing of a design corresponding in its main features to the
tank shown in Figure 5-7, but modified to be suitable for the production of a concentrated
oxygen solution. Figure 5-8B shows the design of the tank used for this purpose at
Simi Valley. In this design the pumping rate into the oxygen-dissolving tank is constant,
but the oxygen supply can be varied. At a low oxygen input rate, the water level is high in
the tank. If more oxygen is supplied, the water level is depressed until the increased
turbulence due to the fall of the water suffices to dissolve the oxygen as fast as it is
supplied. There is, of course, an upper limit to the amount that can be dissolved, depending
upon the pressure and the flow of water, and to some extent on the height of the tank.
Figure 5-8C shows how one of these oxygen-dissolving units could be used for applying
oxygen to the stream in a trunk sewer.
The nozzles through which the oxygen solution is released into the main wastewater
stream must be of a size matched to the characteristics of the pump for maximum
efficiency. A small flow of water at a pressure as high as several hundred psi might be used,
or a large flow at relatively low pressure. If a high-pressure oxygen solution is prepared,
use might be made of treatment plant effluent or water from any other available source,
pumped to points of application before or after charging it with oxygen. If raw wastewater
js used for surcharging with oxygen, the nozzles would have to be large enough so that they
would not clog with solids, which would favor a low-pressure, high-volume system. In the
experimental installation of the Los Angeles County Sanitation Districts above referred to,
screened wastewater from the sewer was used.
The pumping action of the jets in a large trunk sewer could add enough kinetic energy to
the main stream to significantly increase the capacity of the trunk.
5.1.9 Ventilation
Ventilation of sewers is sometimes practiced. There are six possible objectives:
1. to increase the oxygen content of the sewer atmosphere;
5-25
-------
FIGURE 5-8
TWO TYPES OF PRESSURE TANKS FOR DISSOLVING
OXYGEN AND IN-SEWER APPLICATION OF OXYGEN
FORCE MAIN
OR GRAVITY
SEWER _
OXYGEN
WASTEWATER
PRESSURE
LINE
B
OXYGEN
DRAIN
PRESSURIZED
WATER -j
BAFFLE
PLATE
OXYGEN
TRUNK SEWER
SCREEN
SUBMERGED
NOZZLES
OXYGEN
DISSOLVING
FACILITY-
-CONCENTRATED
OXYGEN SOLUTION
UNDER PRESSURE
•^—CONTROL VALVE
OXYGEN TANK
5-26
-------
2. to remove as much H2S as possible from the sewer before it is oxidized to 112804;
3. to dry exposed surfaces of the sewer structures, thereby preventing oxidation
of H2S to H2SO4;
4. to eliminate toxic atmospheres in the sewer;
5. to remove explosive atmospheres in the sewer;
6. to control odors otherwise escaping from the sewer in sensitive areas.
Only the first of these six objectives is related to improvement of the oxygen balance, but
because of the interrelationships, all aspects of ventilation will be discussed in this section.
5.1.9.1 Ventilation to Combat Oxygen Depletion in Sewer Atmosphere
The usefulness of efforts to prevent oxygen depletion in the sewer atmosphere in any
particular case must be viewed with due regard to the significance of the problem to be
solved. In reality, the depletion of oxygen in the sewer atmosphere is insignificant under
most circumstances. In some cases, such as a blockage of the air stream by a surcharged
section of sewer, as much as half of the oxygen in the sewer air must be used up, and the
concentration of course could approach zero in a completely isolated air space. In partly
filled sewers, however, air is displaced by the rise and fall of the fluid level, and there is
normally a dowstream flow of air in the sewer. In such sewers, oxygen concentrations are
rarely less than 90 percent of normal, and for the most part they are in the range of 95 to
100 percent.
Before a decision is made to ventilate a sewer, tests should be made to determine the pre-
vailing oxygen concentrations. // the concentration is above 90 percent of normal, that is,
19 percent O^ m tne sew^r aif, ventilation will not make any material difference in the
oxygen balance in the wastewater stream.
5.1.9.2 Ventilation to Reduce H2S Concentrations
The effects of ventilation on rates of production of acid on sewer wall, and hence on rate
of corrosion of susceptible materials, are more complex. There is no doubt that ventilation
can reduce H2S concentrations in the sewer air. The reduction, however, is due in large
part to the increased turbulence of the air, and the consequent increase of the coefficient of
mass transfer of H2S from the sewer atmosphere to the exposed pipe wall. Near the point
of ingress of fresh air, the H2S content of the sewer air may be greatly diluted and the rate
of corrosion reduced, but remote from the fresh air source the H2S content of the air
approaches the steady state concentration fixed by the equality of the rate of escape from
the water and the rate of oxidation on the wall. The rate of acid production on the pipe
wall is then controlled by the rate of escape of H2S from the stream, not by the concen-
tration of H2S in the air.
5-27
-------
In determining the possible effect of ventilation, it is useful to estimate the rate at which a
steady state of F^S concentration is approached in a sewer atmosphere. Consider a reach
of 48-in sewer in which the atmosphere is suddenly replaced by fresh air. It is estimated that
in a 48-in sewer flowing half full the concentration of K^S would increase to one-half of the
steady state condition in 20 minutes after the air change. If a ventilation system is moving
the air through the sewer at a velocity of 1 ft/sec, this condition would be reached in
1200 ft. If the air were completely removed from the trunk every 1200 ft and replaced
with clean air, but with no change of air velocity, the rate of corrosion at the downstream
end of each reach would be half as fast as it would be without the exchange of air. If the
rate of ventilation were increased, the distance to approach half way to the steady state
would be increased, but not proportionally. In smaller trunks the distance required to
approach half way to the steady state F^S concentration in the air would be shorter and
in larger trunks it would be longer.
The estimate is very rough, but with any other reasonable assumptions it would still appear
that complete replacement of the sewer atmosphere with fresh air at rather frequent inter-
vals would be necessary if removal of HyS by ventilation were relied upon for a major
reduction in corrosion rates.
5.1.9.3 Ventilation to Dry the Sewer Wall
Hydrogen sulfide does not cause corrosion of concrete if the surface is dry, because the
bacterial oxidation of f^S can occur only in the presence of moisture. Therefore ventilation
has sometimes been undertaken with the objective of drying the walls. Thistlethwayte (9)
estimated that the relative humidity of the air should not be higher than 85 percent if
protection is to be assured. A structure close to a point where ventilation air enters a
sewer can be dried satisfactorily, but it is not possible to continuously dry any great
length of sewer even in dry climates. Not even the usual distances between manholes can
receive year-around protection by attempted drying of the walls.
5.1.9.4 Ventilation to Prevent Lethal Atmospheres
The atmospheres in sewers are sometimes lethal, usually because of the presence of hydrogen
sulfide. Rarely, oxygen impoverishment of the air or the presence of other toxic gas may be
the causes of deadly atmospheres. Adequate ventilation of a manhole or junction chamber
by means of a blower will provide air safe to breathe, but it would be difficult to assure
safety between manholes by this means. In no case would it be feasible to keep a whole
sewerage system continuously ventilated in the hope that workmen could enter without
safety precautions.
5.1.9.5 Ventilation to Prevent Explosive Atmosphere in Sewers
Much the same considerations apply to ventilation to remove explosive atmospheres.
Explosions in sewers have resulted from leaking gas mains and from volatile hydrocarbons
discharged into sewers, but not from gases produced by the sewage, unless possibly from
5-28
-------
large accumulations of actively digesting sludge overlain by a stagnant atmosphere, a
combination of conditions that can be imagined in a sewer only under very unusual
circumstances. Occurrences of explosive atmospheres are infrequent and largely unpre-
dictable. Protection against explosions cannot be assured by any practical ventilation
program.
5.1.9.6 Ventilation as an Odor Control Measure
In some situations, the withdrawing of air from a sewer is a useful way to eliminate
nuisance conditions that would arise from the uncontrolled escape of odors at other places
(10). The air so withdrawn must be properly disposed of. The principal place that this is
practiced is at wastewater treatment plants where air is withdrawn from the end of the
sewer rather than to let it escape to the atmosphere. Odor control is the only objective
likely to be effectively served by continuous ventilation.
5.1.9.7 Objections to Ventilation
Generally ventilation is undesirable because it increases the discharge of odorous air to the
environment. Elaborate measures have been taken to deodorize air vented from sewers, but
these processes have often been quite expensive and not entirely effective. Where air is
withdrawn for the limited valid purpose of odor control, the most effective method for
deodorizing the air has proved to be biological oxidation, using activated sludge tanks,
trickling filters, or soil beds (10).
5.1.9.8 Case Histories
In the City of Los Angeles the North Outfall passes under Centinela Valley by way of a
siphon 1,860 ft long. The exhalation of air upstream from the siphon caused serious odor
problems. Closing tightly the inlet structure of the siphon only caused the air to escape
from other structures farther upstream.
In 1958 the North Outfall was paralleled by the North Central Outfall, which also has a
siphon, 8J/2 ft in diameter, under Centinela Valley. Paralleling this, however, an air by-pass
4 ft in diameter was constructed, carrying air from both of the major trunks into the
North Central Outfall downstream from the siphon. Connections from the end of the
trunk at Hyperion were then made to the suction side of the blowers for the activated
sludge part of the treatment plant. Odors upstream from the siphons are completely
eliminated, and odors at the plant are reduced to a negligible level. It had been shown
in pre-design experiments that aeration tanks reduce the odor of the air to the same level
as when pure air is used for aeration (10).
In the City of Palm Springs, California, a blower takes air from the end of the sewer and
blows it through the rock in the trickling filter. This method is equally effective, and is
now used in several places.
5-29
-------
5.2 Chemical Methods
5.2.1 Basic Concepts
Chemicals can serve to control sulfide in either of two ways:
1. By reacting with any sulfide already present in the stream to prevent the escape
of t^S into the air. A chemical applied for this purpose may function in one of
three ways. It may: a) oxidize the sulfide to sulfate or other intermediate
oxidation products; b) convert dissolved sulfide to an inert metallic sulfide; or
c) convert H2S to HS~.
2. By killing the sulfide-producing bacteria or by so altering the environment to
which the slime layer on the pipe wall is exposed that it will not produce sulfide.
The ways to do this include the adding of an oxidizing agent that raises the
oxidation-reduction potential to a level where sulfate reduction is inhibited, or
by adding a toxic substance- that either destorys the slime layer or temporarily
suppresses the activity of the sulfide-producing bacteria.
5.2.2 Chlorination
Chlorine is often applied in a trunk sewer, pump station, or headworks of a treatment
plant. It acts both to destroy any sulfide present and to prevent sulfide buildup in the
chlorinated wastewater for some time thereafter. It is sometimes used to prevent sulfide
buildup in pressure mains where air injection is impractical or ineffective because of relative
flatness of the main (11).
Calcium hypochlorite was occasionally used for the control of odors in solid and liquid
wastes well over a century ago, and elemental chlorine has been used for this purpose since
the 1920's (12). Chlorine and hypochlorite can be used interchangeably. Chlorine in dilute
aqueous solution is in fact a mixture of hypochlorite and un-ionized hypochlorous acid.
Sodium hypochlorite is used as a matter of convenience where the dosages are very small or
occasional, but elemental chlorine is chosen as a matter of economy where larger amounts
are needed. The term chlorination is to be understood as including the use of the hypo-
chlorites. Concentrations and dosages of hypochlorites are expressed as the equivalent
amount of chlorine.
Chlorine reacts not only with sulfide but also with mercaptans, which are important odor
components of wastewaters. The reaction of sulfide and mercaptans with chlorine is
immediate, whereas biological deodorization with oxygen is slow. Chlorination does not
remove odors as completely as does the slower biological oxidation, but most odor con-
ditions attributable to wastewaters (but not sludges) are adequately controlled.
If an excess of chlorine is added to a wastewater containing sulfide, the sulfide is oxidized
largely to sulfate according to the following reaction:
5-30
-------
+ 4H2O-+-SCT4 + 9H+ + 8C1"
The reaction requires 8.87 parts by weight of chlorine for each part of sulfide (13).
However, the chlorine reacts also with other constituents of the wastewater, so that 10 to
15 parts of chlorine may be consumed before the sulfur is converted completely to sulfate.
If chlorine is added slowly, with vigorous mixing, to a pure sulfide solution, the sulfide may
be destroyed by being converted to sulfur according to the equation:
HS~ + Cl2-*-S + H+ + 2C1~
Chlorine consumption in this reaction is 2.22 parts by weight per part of sulfide. Other
compounds, including thiosulfate, trithionate, and sulfite are produced when the degree of
oxidation is intermediate between sulfur and sulfate.
In laboratory tests the chlorine requirement for elimination of sulfide in the wastewaters of
the Los Angeles County Sanitation Districts was found to range from 3 to 9 times the
sulfide concentration. Observations upstream and downstream from chlorination stations
on the sewers showed somewhat larger requirements, probably because the mixing was less
efficient than in the laboratory tests (14).
The addition of chlorine to a wastewater flow depresses biological activity in the stream,
virtually stopping the consumption of oxygen until the chlorine residual has disappeared.
During this period of suppressed biological activity the stream acquires an oxygen reserve
that will delay the reappearance of sulfide downstream. The chlorine also leaves a residue
of moderately active chlorine compounds that can destroy additional sulfide. These effects
do not generally persist for more than about a half mile downstream unless excess dosages
are applied. The use of excess dosages, however, is economically inefficient from the
standpoint of over-all sulfide reduction.
Efficient use of the chlorine requires rapid and complete mixing of the chlorine solution
with the wastewater stream, just as it does when chlorine is used for disinfection. The
ideal method of application is for the wastewater and chlorine solution to come together in
a small chamber in which enough energy is dissipated to complete the mixing in the
minimum amount of time, preferably less than a second. Where a fall is available, this
objective can usually be attained quite readily. By way of illustration, consider a waste-
water stream entering the wet well of a pump station. One of the possible ways to
accomplish rapid mixing would be to install a cubical or cylindrical open-top box under
the inlet pipe, with a free fall into it and overflow out of it into the wet well. A free fall
of a foot is suitable. The water level in the wet well should not rise higher than the top of
the box. The box should provide a detention time generally in the range of 0.5 to 2 seconds.
If the chlorine solution is added in this box, efficient mixing will be attained. The fall of
one foot is satisfactory for any flow, small or large, but if the box is very large, the
chlorine solution should be dispersed by way of a perforated plastic pipe.
5-31
-------
Discharge of the chlorine solution into a hydraulic jump is another way to get good mixing.
In many cases there is not a practical way to get efficient mixing because of the variability
of the waste water composition and flow. Often there is little choice but to discharge the
chlorine solution by way of a hose or a perforated plastic pipe into the stream in a sewer.
If the stream is shallow, mixing is poor and local overdosing is likely to cause fuming. Poor
mixing also results if the chlorine solution is discharged directly into a wet well.
Even with good mixing, there is no simple way to achieve the theoretically attainable
chlorination efficiency when chlorinating a wastewater of variable flow and composition. In
adjusting the dosage for chlorination of raw wastewater, tests of relevant characteristics of
the water downstream are made. Chlorination may be carried to the point where a chlorine
residual is shown by orthotolodine or other test reagent, or to the absence of sulfide, or
perhaps to the elimination of sulfide at a relatively distant point downstream, as at the end
of a pressure main. Often the chlorine application rate is programmed according to the
time of the day after determining by trial what rates are necessary to attain the desired
condition. The diurnal changes of rates may be made manually or by program timers.
Sometimes the control is semiautomatic, wherein the chlorinator rate varies proportionally
with wastewater flow, but with the proportion subject to manual or programmed control,
or it may be fully automatic, wherein the dosage is controlled in response to signals from
a downstream sensor. The sensor may make automatic colorimetric tests for chlorine
residual or it may be an amperometric chlorine residual tester or an oxidation-reduction
potential electrode. By use of a sulfide sensing electrode, it would be possible to control
the chlorine dosage to the disappearance of dissolved sulfide, but as far as is known this
had not been done.
Fully automatic control is not uncommon for the chlorination of treated wastewater
effluents and for certain industrial wastes, notably for the destruction of cyanide, but it is
doubtful practicability for upstream wastewater chlorination.
CASE HISTORY
In 1929 the newly installed regional trunk sewer system of the Los Angeles County
Sanitation District (serving about a third of the area now served as shown in Figure 6-1)
developed high sulfide concentrations as a result of sluggish flows in large, long trunks
carrying only a few percent of their ultimate capacity. In 1931, eight chlorination stations
were established on upstream tributary trunks, whereby satisfactory sulfide control was
attained (15). In 1939, when the flow was 24 mgd, or about 2l/2 times as great as in 1931,
an assessment of the chlorination program was made. With the chlorinators turned off,
the dissolved sulfide concentrations entering the downstream treatment plant averaged
1.05 mg/1. With applied chlorine equivalent to 12 mg/1 in the total wastewater flow,
dissolved sulfide concentrations averaged 0.43 mg/1. Thus the ratio of chlorine applied to
sulfide eliminated was 19 to 1. The poor ratio was due mostly to the necessary overdosing
at the upstream locations. The wastewater flow carried by the system increased rapidly in
subsequent years, and the chlorination program was gradually phased out.
5-32
-------
5.2.3 Metallic Salts
The salts of many metals will react with dissolved sulfide to precipitate insoluble sulfides,
and thereby prevent the escape of F^S to the atmosphere. Insolubility is a matter of degree.
For this method to be fully effective, the metal sulfide formed must be highly insoluble.
The use of these salts must be considered in the light of possible effects on treatment
processes and effluent quality.
5.2.3.1 Iron Salts
One metal that has been used in this way is iron. There are several iron sulfides that may
form: pyrrhotite, varying in composition from FeS to Fe4S(j; ferric sulfide, Fe2S3;
smythite, FeoS4; and pyrite and marcasite, both having the formula FeS2- Fe$2 is found
among the corrosion products on iron that has corroded in the presence of sulfide, and it is
found in sewers where iron concentrations are high. It is probably formed by oxidation of
other iron sulfides.
Neither pyrrhotite nor ferric sulfide alone is sufficiently insoluble to lower the dissolved
sulfide content below several tenths of a milligram per liter. If a mixture of ferrous and
ferric salts is added to a sulfide-containing wastewater, better results are obtained than with
either one alone. The optimum mixture is one with a molecular ratio of one part of
ferrous to two parts of ferric (14). The dissolved sulfide concentration can be reduced
to 0.2 mg/1 with only a moderate excess of iron. Presumably this reaction is as shown
below:
Fe++ + 2Fe+++ + 4HS~-^- Fe^S,, + 4H+
The combination of a small amount of dissolved oxygen along with a ferrous salt also
reduces dissolved sulfide to low levels.
Freshly precipitated iron sulfides are readily oxidized by oxygen, producing sulfur. Thus,
while oxygen promotes the precipitation of iron sulfides, iron acts as a catalyst to promote
the oxidation of sulfide.
If an excess of iron, ferrous or ferric, is added to a sample of wastewater which is then
incubated in the absence of air so that high sulfide concentrations are produced, the
dissolved sulfide concentration will rise to about 2 mg/1. Similarly, concentrations of 1 to
2 mg/1 of dissolved sulfide were found in a sewer to which a large excess of iron had been
added (14). Presumably the iron becomes tightly bound to other radicals, which may
include organic matter, orthophosphate, polyphosphate, and other chelates added in
cleaning compounds. However, there are cases where the presence of iron seems to
hold dissolved sulfide to a few tenths of a mg/1. The difference may be related to the
amount of dissolved oxygen in the wastewater.
5-33
-------
Iron salts are occasionally added to industrial wastes of high sulfide content, provided
the reduction of dissolved sulfide to 1 mg/1 is satisfactory. Furthermore, the addition of
ferrous chloride to anaerobic sludge digestion tanks plagued with high sulfide concentrations
has caused spectacular improvement of the digestion process.
5.2.3.2 Zinc Salts
Zinc sulfide is much less soluble than iron sulfide. The addition of a zinc salt to sulfide-
containing sewage generally reduces the dissolved sulfide concentrations to less than
0.1 mg/1. No large excess of zinc is required.
Because of the strong affinity of zinc for sulfide, it was hoped that an excess of zinc added
at an upstream point would assure freedom from H^S downstream. Results have been
disappointing. As in the case of iron, the zinc appears to become bound in other com-
binations, so that the objective is not attained except by use of large excesses of zinc.
At one time zinc was applied in rather large amounts to a large trunk sewer of the City of
Los Angeles, using a solution prepared by dissolving scrap zinc in waste acids, but the
operation has been discontinued. .It is still used in some instances for treating small
industrial wastewater flows.
5.2.3.3 Other Metals
Lead sulfide, PbS, is much less soluble than ZnS, but the application of lead salts for this
purpose would be uneconomical, and the increased lead content of the sludge might be
objectional. The silver and copper sulfides, Ag2S, C^S, and CuS, are so extremely
insoluble that they do not react in the analytical tests for sulfide. Copper is present in
small amounts in ordinary wastewaters, normally in concentrations of the order of
0.1 mg/1, or more where copper plumbing is used. Thus a very small part of the sulfide
in wastewaters may be combined with copper, but there is no easy way of determining
this. Sulfide thus bound is for practical purposes nonexistent. Copper salts would probably
be very effective for sulfide control, but the cost would be very high, and if an excess of
copper were used it probably would have harmful effects on aerobic wastewater treatment
processes.
5.2.4 Nitrate
Certain bacteria can oxidize organic matter by reducing the nitrate radical. The nitrogen is
converted principally to ^. The addition of nitrate to a sewer does not directly affect the
oxygen balance, because nitrate reduction does not occur in the presence of oxygen. Sulfate
reduction does not occur in the presence of nitrate if nitrate-reducing bacteria are present.
5-34
-------
Nitrate may serve to prevent sulfide buildup by preventing sulfate reduction, and nitrate-
reducing bacteria can use nitrate to oxidize sulfide if oxygen is not available. A small
amount of nitrate in a wastewater stream may not prevent sulfate reduction in the slime
layer where nitrate as well as oxygen may be entirely depleted.
Nitrate has been used in some places for sulfide control in sewers, being added either as
calcium nitrate or sodium nitrate. It may have relatively little effect when first added,
because of a scarcity of nitrate-reducing bacteria. A population of these bacteria develops in
the slime layer, so the reaction is more rapid after a day or two of treatment. Since the
reaction occurs mostly in the slime layer, where organic matter is abundant, much of the
nitrate is consumed in oxidizing organic matter. The amount of nitrate required to destroy
sulfide in a sewer may be estimated as 10 pounds of NaNOg per pound of sulfide if sulfide
concentrations are high. The ratio may be 20 to 30 when the sulfide concentrations are low,
and will be even larger if the nitrate is added at the upper end of a trunk as a preventive.
Nitrate is more effective when used for the chemical control of sulfide in overloaded
oxidation ponds, presumably because the long residence time allows the development of a
culture of nitrate reducers in the water.
To treat an oxidation pond, the nitrate salt is first dissolved in water. It may be dispersed
into the water in the pond by adding it continually to a stream flowing into the pond or
into a recirculating stream, or by spraying the solution onto the surface of the pond.
CASE HISTORY
In 1958 the Davis Cannery at Atwater, California, was discharging 0.74 mgd of wastewater
from the canning of peaches, with a BOD concentration averaging 300 mg/1, or a total of
1,850 Ib/day of 5-day of BOD. The wastewater went to ponds with a surface area of 5.84
acres. High sulfide concentrations developed. Application of 1,000 Ib/day of NaNO^ to
the water leaving the cannery reduced total sulfide concentrations in the ponds to zero in
the daytime, with a maximum of 0.3 mg/1 in the early morning.
5.2.5 pH Control and the Use of Strong Alkalies
Since the proportion of dissolved sulfide existing as ^S diminishes rapidly as the pH is
increased, it follows that problems due to escaping ^S can be diminished by adding
alkalies. Slaked lime has been used experimentally for this purpose, but the method has
not proved useful. Rather large dosages are needed, probably 50 mg/1 or more in a typical
case, if ^S is to be reduced to acceptable levels in the presence of moderately high sulfide
levels. A dosage of 25 mg/1 might suffice for low sulfide concentrations, but chlorine might
then be more satisfactory.
The attempt to control ^S by controlling pH has the shortcoming that if the amount of
sulfide in solution remains unchanged, the problem of ^S in the air may recur from any
circumstance that lowers the pH again. The treatment is not useful, for instance, at the
5-35
-------
upper- end of a long trunk, because the pH will be lowered by tributary flows and by the
production of CC>2 and organic acids through biological action, which will not be sub-
stantially impeded as long as the pH is below 9. The attempt to apply lime continuously
at the upper end of a trunk in sufficient amount to be effective all the way down the line
would be impractical, because the upstream part of the trunk would accumulate an
incrustation of calcium carbonate.
It is desirable, from several points of view, to keep the pH of wastewater as high as is
practical by proper regulation of industrial waste discharges. The policy of the Los Angeles
County Sanitation Districts, for example, in respect to pH control has been strict exclusion
of acids, with a minimum allowable pH of 6.0 in waste waters discharged to the sewers, and
encouragement of the discharge of high pH wastes. Thus, industries that were producing
acetylene from calcium carbide were invited to run their spent lime into the sewer, and
various spent solutions of caustic soda were also welcomed if they did not have objectional
characteristics such as strong odor or high oil content.
An upper limit for the pH established in many sewer-use regulations is not based upon a
realistic appraisal of the effects. The neutralization of high-pH wastes before discharge to
the sewer is usually detrimental. The lower the pH, the less buffer there is against serious
effects of accidental acid discharges. High pH, on the other hand, diminishes corrosion and
odors, and rarely has any undesirable effects. In some instances, continuous discharges to
small sewers of high-pH wastes, as for example from bottle-washing operations, have caused
incrustations in the sewers. These problems have been resolved by requiring inpoundment
of the wastes with periodic slug discharges, perhaps once each day, or as often as once an
hour.
Excessive discharges of strong alkalis can, of course, raise the pH to unacceptable levels in
treatment plants. If the pH in a biological oxidation process is higher than 9, the efficiency
may be substantially impaired, and it may be virtually stopped at a pH of 10. The buffer
capacity of wastewater, and especially of activated sludge tanks, is high enough so that it
takes rather large amounts of strong alkalis to raise the pH to harmful levels.
Beyond serving to maintain higher pH values, alkalies have an important application as
sterilizing agents if they are added at high concentration. Sulfide generation in a sewer can
be stopped if the slime layer is inactivated, and it takes several days for it to return to
normal. This effect has been shown with strong alkalies, strong acids, and chlorine. The use
of chlorine for this purpose is quite expensive, although where a station for in-sewer chlori-
nation exists, suppression of slime activity for some distance downstream is observed.
Acid is undesirable for the reasons previously mentioned.
Both slaked lime and quicklime have proved practical for inactivating the slime. The lime
is ordinarily poured by hand into a manhole at a rate that gives a dosage between 5,000 and
10,000 mg/1 for a period of about 20 minutes. In other cases, caustic soda solutions have
been used, discharging from a tank truck directly to the sewer. The length of time for a
5-36
-------
sewer to regain its sulfide-generating capability varies with the temperature and with the
intensity of the chemical treatment. If alkalies are used in a sufficiently high concentration,
it is possible to approach the condition of 100 percent sterilization of the slime layer, but
this may not be optimal from the standpoint of economy. If the pH during in-line chemical
treatment is raised to 12, sulfide generation in the summer may return to l/2 the normal rate
within a week. In one case it reached normal levels in three days. If the pH is raised to 13,
there is likely to be very little sulfide generation for a week, but it will return to near
normal within two weeks. More intensive treatment does not extend this time any further.
The chemical dosage to reach pH 13 is very high, so the optimal treatment may be pH 12.5
or somewhat higher at weekly intervals during the summer, and at two-week intervals during
the winter. The actual schedule to be followed in any particular application will be
determined by local conditions (1).
CASE HISTORY
Sodium hydroxide is the principal chemical now used for sulfide control by the Los Angeles
County Sanitation Districts. It is used at local trouble spots, usually in flows less than 3
cfs. Waste caustic solution from industries, generally quite dilute, is used to the extent
available, and concentrated commercial caustic is purchased when necessary. In either case,
it is applied from a tank truck. The duration of a treatment is 20 minutes, and the
objective is a pH of 12.5. Required frequency of treatment is determined by a routine
testing program.
The scope of this operation is indicated by the records for August, 1973, which show that
about 70,000 Ib of NaOH were used in that month for treating 25 lines, mostly gravity
sewers but including some force mains. There is no noticeable effect on the pH at the main
treatment plant, and it is not likely that there is more than a very small effect there on the
average sulfide concentration, since none of the major trunks are treated.
The amounts used in the winter are very small.
5.2.6 Hydrogen Peroxide
Hydrogen peroxide is another chemical that has and is being used to control hydrogen sulfide
in wastewater. Data has been presented to show that hydrogen peroxide reacts with hydrogen
sulfide to form water and sulfur (17). Theoretically, this reaction requires a 1:1 ratio of hydro-
gen peroxide to hydrogen sulfide. In practice, however, a higher ratio is employed (as shown
by the following case histories). This results in a contribution of dissolved oxygen which in-
hibits the regeneration of hydrogen sulfide.
Installations where hydrogen peroxide has or is being applied for the purpose of controlling
hydrogen sulfide in wastewater include: Ocala, Florida; Ft. Lauderdale, Florida; Hollywood,
Florida; Sunrise City, Florida; Dade County, Florida; Broward County, Florida; Hampton
Roads, Virginia; Corpus Christi, Texas; Ft. Worth, Texas; and Bluff Cove, California. The fol-
lowing case histories present some of the data collected at two of these installations.
5-37
-------
CASE HISTORY I
Figure 5-9 shows schematically the Bayview Drive force main in the Ft. Lauderdale,
Florida, wastewater system. The main is about 3 miles long, made of 14", 16", and 18"
pipe. It receives four pumped flows.
It is reported (16) that applications of hydrogen peroxide at pump station B-4 gave results
as shown in Table 5-4. The dosages are expressed as ratios of peroxide added to quantity of
sulfide carried by the flow into station B-4. The tests were made over a 12-day period, but
the results are grouped by the hour of the day.
TABLE 5-4
HYDROGEN PEROXIDE TREATMENTS AT FT. LAUDERDALE, FLORIDA
Hour
9-10 AM
10-11 AM
1-2 PM
2-3 PM
3-4 PM
Ratio of
hydrogen
peroxide
to sul-
fide at
Stn. B-4
0
2:1
3:1
4:1
2:1
3:1
4:1
2:1
3:1
4:1
0
2:1
2:1
3:1
4:1
Total Sulfide Found, mg/1
Stn
B-4
3
0.1
0
0
0.2
0.1
0.1
0.2
0.2
0.1
3
0.1
0.1
0.1
0
Middle
River
4
0
0
0
0
0
0
0
0
0
3
0
0
0
0
NE
21st
St
7
0
0
0
0
0
0
0
0
0
6
0
0
0
0
Oakland
Park
Blvd
6
3
1.1
0
1.5
0.9
1.0
2
1.4
0.8
6
3
3
1.2
0
NE
37th
St
5
5
1.5
0
3
3
1.8
4
3
2
6
2
3
3
2
5-38
-------
FIGURE 5-9
BAYVIEW DRIVE FORCE MAIN
FT. LAUDERDALE SYSTEM
TREATMENT
PLANT
STA. B-7,8
STA.B-6
ADDITION OF
H202.
'SAMPLE NE 37th. ST.
•SAMPLE"OAKLAND
PARK BLVD."
STA.B-14,15
-SAMPLE NE 21st. ST.'
STA.B-3
•SAMPLE MIDDLE RIVER
STA.B-4
5-39
-------
CASE HISTORY II
Hydrogen peroxide was added at Bluff Cove pump station of Los Angeles County Sani-
tation Districts (1). The pumps discharge into a pressure main 12 inches in diameter and
1,965 feet long. Pumping is intermittent; the average retention time is about two hours.
Various dosages of hydrogen peroxide were added. The solution feed pump was electri-
cally connected to the wastewater pump, so that at any fixed feeding rate all of the waste-
water received the same dosage. Sulfide tests were made at the pump station and at the end
of the force main, with results as shown in the following table. Three qualitative tests for
peroxide at the end of the force main were negative.
TABLE 5-5
EFFECT OF HYDROGEN PEROXIDE IN BLUFF COVE FORCE MAIN
Date
16 Sept
21 Sept
22 Sept
23 Sept
25 Sept
1 Oct
2 Oct
2 Oct
7 Oct
9 Oct
feed
rate,
mg/1
0
0
27
27
32
32
0**
41
Waste-
water
Temp.,
_J?C_
27.0
26.1
26.2
26.4
26.1
26.0
25.8
25.6
24.7
Pump
No.
Tests
8
11
8
Total
Station
Avg.
mg/1
1.2
0.8
1.0
Sulfide
End Force
No.
Tests
9
12
8
8
9
4
8
8
7
8
Main
Avg.
mg/1
6.1
6.4
2.1
2.1
1.5
0.4
0.8
4.5
0.35
0.66
Dissolved
Oxygen,
End Force
Main, mg/1
2.2
1.2
8
8
16
4
The feed rate is expressed in terms of actual weight of
commercial solution.
"""Immediately after cessation of the f^C^ treatment.
***This recorded dosage is evidently erroneous. See text.
> not ^e weight of a
Comparing the results of September 16 to 25, it appears that 27 mg/1 of ^2^
about 4.4 mg/1 of sulfide but leaving a substantial sulfide residual. Comparing the results of
October 1 and 2, 32 mg/1 of H^O^. eliminated 3.9 mg/1 of sulfide, lowering the total sulfide
concentration to where little other than insoluble metallic sulfides remained. The result for
October 7 looks very good since only 15 mg/1 apparently reduced sulfide to a satisfactory
level, but the finding of 1 6 mg/1 of dissolved oxygen at the end of the force main shows that
the recorded dosage of 15 mg/1 must be in error.
An experiment was made to determine if continuous peroxide application would affect the .
regeneration of a slime layer that had been destroyed by caustic treatment. The Bluff
Cove force main was treated with NaOH and NaOCl, after which there was no sulfide
5-40
-------
generation. It required 16 days for generation to return to normal. The treatment was
repeated and then peroxide was fed continuously, except for 24 hours on the fifth and
sixth days, at a dosage rate of 12 to 22 mg/1. The slime layer returned to normal within a
day or two of the time required when no peroxide was used.
5.2.7 Other Chemicals
There are other possible oxidizing agents that could be used for sulfide control. Some of
them, such as chlorate and chromate, react biologically. Others, including permanganate
and ozone, destroy sulfide chemically.
At one time waste chromate solution was trucked from a large plating works to a point
where it was put into tanks and fed into a trunk sewer at a rate adjusted to eliminate
sulfide for a considerable distance downstream. The chromium was rendered insoluble and
harmless as a result of its reduction to the trivalent form.
Permanganate has sometimes been used for treating sulfide-containing well waters, but not
for the sole purpose of destroying sulfide. Ozone has been tried a number of times for
sulfide control in wastewaters. It is necessary to bring the ozonized air into contact with
the wastewater to dissolve ozone. In this operation, much more oxygen dissolves than
ozone, and since oxygen reacts with sulfide, the prior ozonizing of the air is unnecessary
(14).
Ozone is effective for deodorizing sewer air but an excess of ozone must be avoided, since
it can be an objectionable as the original pollutants.
5.3 Control of Industrial Wastes
The desirability of strict limitation of low-pH discharges to the sewers and the advantage
of high-pH wastes were discussed in Section 5.2.4. In addition to a prohibition against
acid discharges, secure measures must be prescribed to assure that spills do not occur.
In one treatment plant where the flow was 16 mgd, the wastewater suddenly became so acid
that the pH dropped to about 2, and I^S in the wastewater, arising from the dissolving of
iron sulfide in the slimes, reached 80 mg/1. Ordinarily f^S concentrations there were not
high enough to be of concern. Several years later there was an apparent recurrence of that
incident. Two men working in the trunk sewer at that time were killed by the high f^S
concentrations. Even where a sewer carries wastewater with little or no sulfide, iron sulfide
is generally present in the slimes in sufficient amount to produce high concentrations of
f^S when strong acid is discharged.
To assure against discharges of acids or other harmful chemicals to the sewer, it is sometimes
necessary to require a retention tank large enough to hold an amount of the industrial
wastewater equal to a 24-hr flow. The tank should be kept mixed, with the wastewater
5-41
-------
flowing through and overflowing to the sewer, with no way to dump it to the sewer.
Where large amounts of acid are stored, there must be provisions, such as diked areas, to
hold spills.
Sulfide-containing industrial wastes should be controlled or excluded from the sewers if
a low sulfide system is to be maintained. Sulfide-containing wastes may be produced by
tanneries, petroleum refineries, paper pulp mills, and a variety of other chemical industries.
Sometimes a municipality has established a policy of accepting such wastes on the basis
that the system was built for that purpose, or that a central treatment works can do the job
more economically than can a number of individual plants. This policy has the potential
disadvantage of possible lethal atmospheres and corrosion in manholes and in sewers if
they are corrodable, and odors from the sewers and at the waste water treatment plant.
The problems at the treatment plant can be dealt with by proper procedures, but in view
of the erratic nature of industrial discharges, it would be difficult to regulate a treatment
process on the sewerage system so as to maintain a low sulfide condition in the sewers.
It is usually better that the waste be treated at the source.
As an example, a small tannery that was discharging spent depilatory to the sewer was told
that the discharge was unacceptable. The procedure then adopted was to collect the wastes
in a tank equipped with a stirrer, where they were treated batchwise with an excess of
ferrous sulfate solution. Precipitation of ferrous sulfide reduced dissolved sulfide to
1.0 mg/1 or less, after which the mixture was discharged to the sewer.
The concentration of 1 mg/1 of dissolved sulfide allowed in this example was a special
action based upon a consideration of the dilution available in the trunk sewer and the
limitations of the method employed for precipitating dissolved sulfide. In cities where
sewer use ordinances include restrictions of sulfide-containing wastes, the limit is usually
placed at 0.1 mg/1 of dissolved sulfide.
It may be necessary to control industrial wastewaters that act to accelerate sulfide
generation. Two factors are involved, temperature and nutrient' concentration. High
temperature in a sewer is undesirable for a number of reasons, including thermal stresses
affecting the pipe, fog in the sewer atmosphere interfering with maintenance operations,
fog issuing from manholes, and an increase of the rate of use of oxygen and of production
of sulfide.
A regulation commonly used places an upper limit of 120 deg F (49 deg C) on the
temperature of wastes acceptable into the sewers. This does not imply that 120 deg F
would be a satisfactory condition in sewers, but recognizes that, in view of the available
dilution, it would seldom be worthwhile to require an industry to provide facilities to cool
a water below that temperature.
Where a special problem is created, as for example a large flow of water near 120 deg F is
discharged into a relatively small sewer, remedies may be called for under the general clause
of an ordinance categorically prohibiting harmful discharges.
5-42
-------
The rate at which a slime layer can produce sulfide is proportional to the concentration of
certain organic nutrients, unless sulfate is in short supply, in which case the rate may depend
only upon the sulfate concentration, or upon both the sulfate and organic nutrient con-
centrations. Rarely are specific limitations placed upon sulfate unless the concentrations
are so high that there are other problems as well, such as excessive mineralization of
effluents or interference with anaerobic digester operation.
No way is known to measure the concentrations of nutrients used by sulfide-producing
bacteria in the slime layer of a sewer, since almost nothing is known about the identity of
these nutrients. In domestic-type wastewaters, there appears to be a general proportionality
between the sulfide-active nutrients and BOD. BOD will continue to be used as a measure
of such nutrients until some new basis is found. BOD is significant as a measure of the load
on biological oxidation processes, and this is one of the reasons that limits are placed on the
BOD of wastes received into sewers, or that BOD is one of the parameters for charges for
sewer service. In this situation, the effect of BOD on sulfide generation is a minor
consideration.
Where a wastewater is not subjected to biological oxidation, BOD limitations may be
advisable for reasons of sulfide generation, but it may be necessary to establish a correlation
between sulfide generation and the BOD or COD of a waste.
In 1948-51, the City of Newport Beach, California, had to continue to use an overloaded
system, including a primary treatment plant, pending the construction of the Orange
County Sanitation Districts joint facilities. Pumping capacity in particular was overloaded,
and the wastewater backed up in the trunk sewers, leading to sulfide concentrations in the
summer averaging 6 mg/1. In the autumn, three fish canneries operated, boosting sulfide to
even higher levels. It was necessary to disinfect the effluent by chlorination, and to use
primary effluent for the chlorine solution water, which is wasteful because of its high
breakpoint chlorine demand. The required dosage was at times 120 mg/1.
When the fish canneries were operating, the total BOD load in the wastewater averaged
2,000 Ib/day, of which 1,200 Ib were due to the canneries. Sulfide buildup varied linearly
with EBOD in the domestic sewage, but the BOD of the fish cannery waste was only 25
percent as effective in stimulating sulfide generation. The increment of cost to the City
for the chlorination of sulfide due to the fish cannery operation was charged to the three
companies in proportion to the tonnage of fish that each had processed.
5.4 References
1. Parkhurst, J. D., Pomeroy, R. D., and Livingston, J., Sulfide Occurrence and Control
In Sewage Collection Systems, Report to U.S. Environmental Protection Agency under
Grant No. 11010ENX (1973).
2. Pomeroy, R. D., Generation and Control of Sulfide in Filled Pipes, Sewage and
Industrial Wastes, 31, No. 9, pp. 1082-1095 (1959).
5-43
-------
3. Laughlin, J. E., Studies in Force Main Aeration, Journal of the Sanitary Engineering
Division, ASCE, 90, SA6, pp. 13-14 (1964).
4, Boon, A. G., and Lister, A. R., Formation of Sulphide in Rising Main Sewers and Its
Prevention, Water Pollution Research Laboratory, Stevenage, England (to be published).
5. Pomeroy, R. D., and Parkhurst, J. D., Self-purification in Sewers, Proceedings of the
6th International Conference on Water Pollution Control Research, Jerusalem (1972),
Pergammon Press.
6. Pomeroy, R. D., and Lofy, R. J., Feasibility Study on In-sewer Treatment Methods,
Report to the Federal Environmental Protection Agency under Contract No. 14-12-944
(1972).
7. Mitchell, R. C., U-tube Aeration, Report to Environmental Protection Agency,
Project No. 17050 DVT, Contract 68-01-0120 (1973).
8. Sewell, R. J., Sulfide Control in Sanitary Sewers Using Air and Oxygen, U.S. Environ-
mental Protection Agency, Project No. 11010-DYO (1973).
9. Thistlethwayte, D. K. B. (Editor), Control of Sulphides in Sewerage Systems,
Butterworths, Pty. Ltd., Melbourne, Australia (1972), and Ann Arbor Science
Publishers, Ann Arbor, Michigan (1972).
10. Pomeroy, R. D., Controlling Sewage Plant Odors, Consulting Engineer, 20, pp
101-104 (Feb., 1963).
11. Design of Sanitary and Storm Sewers, Water Pollution Control Federation Manual of
Practice No. 9, Washington, D.C., p. 324 (1969).
12. Goudy, R. F., Odor Control by Chlorination, Sewage Works Journal, L, pp. 196-205
(1929).
13. Nagano, J., Oxidation of Sulfides During Sewage Chlorination, Sewage and Industrial
Wastes, J22, 884 (July, 1950).
14. Pomeroy, R. D., and Bowlus, F. D., Progress Report on Sulfide Control Research,
Sewage Works Journal, 18, No. 4, pp. 597-640 (1946).
15. Bowlus, F. D., and Banta, A. P., Control of Anaerobic Decomposition in Sewage
Transportation, Water Works and Sewerage, 79, 11, 369 (1932).
16. Duffy, W. J., E. I. du Pont de Nemours & Co., Inc., The Use of Hydrogen Peroxide to
Control Hydrogen Sulfide in the Coral Ridge Sewer System, Report to City of Ft.
Lauderdale (June 7, 1972).
5-44
-------
17. Satterfield, C. N., Reid, R. C., and Briggs, D. R., "Rate of Oxidation of Hydrogen Sulfide
by Hydrogen Peroxide," Journal Chemical Society, 76, pp. 3922-23 (August 5, 1954).
5-45
-------
CHAPTER 6
DESIGNING TO AVOID SULFIDE PROBLEMS
6.1 Basic Concept
Severe sulfide problems arise in sewerage systems where flows are sluggish or the waste-
water is contained for some time in filled pipes out of contact with air. There is practically
no buildup in partly filled small sewers flowing at a velocity of 2 ft/sec or more. In large
trunks, buildup does occur even at considerably higher velocities, but then the rate of
buildup is slow, rarely producing concentrations higher than 1 mg/1.
A gravity sewer at good velocity may, of course, show high concentrations if sulfide is
contributed by upstream pipes that are completely filled or that are carrying wastewater
at sluggish velocities.
In any case, the sulfide concentrations that are found in sewers are closely related to the
design of the system. Severe sulfide conditions, with concentrations above 1 mg/1, should
be avoided by proper design, and the lower concentrations that are common in large
trunks should be minimized or eliminated by proper treatment or should be taken into
account by planning facilities that will avoid odor problems and corrosion.
Many of the measures suggested to solve existing sulfide problems, such as air injection
into force mains, in-line aeration, and others, should be considered at the time that a
system is designed. Since they have already been discussed in Chapter 5, this section
will deal with the major design features that are of an irrevocable nature.
6.2 Slopes of Small Collecting Sewers
Figure 3-12 shows the effect of slope on sulfide buildup in collecting sewers in residential
areas.
It might appear that the rate of sulfide buildup shown by line B of that figure, for slopes
averaging 0.4 percent, would not be too bad, but the production of about 0.25 mg/1 of
sulfide in the first % mile of flow is not consistent with an intent to build a system free of
sulfide problems. Furthermore, the figure does not tell the whole story, since it shows
only average results.
Table 6-1 shows the average sulfide concentrations for the highest 5 percent and 10 percent
of the sulfide concentration results for the sewers represented in the figure. Maximum
sulfide concentrations are also shown.
6-1
-------
TABLE 6-1
SULFIDE CONCENTRATIONS IN SMALL COLLECTING SEWERS
Sulfide Concentrations
Line
(Fig. 3-12)
A
B
C
D
Average
slope
percent
0.23
0.40
0.57
0.9
Number
of
results
133
147
141
46
Average
of
highest
10%
mg/1
1.44
0.93
0.41
0.13
Average
of
highest
5%
mg/1
1.8
1.35
0.49
Maximum
mg/1
2.8
2.0
0.6
0.2
The City of Long Beach, California, has had for many years an exceptionally efficient sewer
maintenance program (1). The City has found that small sewers at a slope of 0.4 percent
should in general be cleaned annually with a ball or other hydraulic device, but that where
the slope is 0.6 percent, it is sufficient to ball the lines once in five years, and then mostly
as a precautionary measure.
The sewer ordinances in some cities now have this provision: "The standard minimum slope
shall be 0.006 (0.6 percent) except where the sewer will serve more than 400 connections.
Slopes less than the standard minimum slopes may be used only if justified in an engineering
report, approved by the City Manager, showing that it is not practical to attain the
standard slopes."
This seems like a reasonable provision, except that in the light of information now available
it might be permissible to allow a flatter slope after 100 connections instead of 400. It may
be added that when the engineering report called for in the ordinance is reviewed, any
allowed exceptions to the 0.6 percent standard should be limited to short distances, and an
exception should in no case allow a slope less than 0.4 percent.
There are areas where small sewers have been laid at slopes much flatter than the standards
here recommended, and yet they have operated with no sulfide generation. This was found
to be the case, for example, in a city near the Gulf of Mexico coast. The required slope of
the upper ends of the sewers is 0.33 percent, decreasing to flatter slopes after a short
distance. An examination of the system showed that 75 percent of the flow in the sewers
was infiltrated groundwater, which increased the oxygen supply and diluted the oxygen-
demanding components. Even pressure mains produced very little sulfide. Similar
conditions exist in many flat areas with high groundwater levels.
6-2
-------
It is now required that sewers be laid with tight joints (2). This may require revision of
previous slope or velocity standards to prevent serious sulfide problems. One city that
continued to lay small sewers at a slope of 0.2 percent, but with tighter joints than in the
past, has experienced many odor complaints, liability actions for corrosion of plumbing on
account of HoS from the sewer, serious damage to a pump station, collapse of trunk
sewers, and the loss of two lives due to hydrogen sulfide poisoning. The difficulties were
caused partly by sulfide produced in pressure mains, but were mostly due to buildup in the
flat collecting sewers.
The sewers represented in Figure 3-12 were in dry areas where there was little or no
infiltration. They indicate what may be expected in other areas where temperatures are
similar and tight joints are used. Lower temperatures might permit slightly flatter slopes.
However, velocities in the small sewers are considerably less than 2 ft/sec even at a slope
of 0.6 percent. Any lessening of the slope will augment the concentrating of organic solids
along the bottom of the channel and intermittent ponding of the water, producing sulfide
at a rate influenced much more by the deposition of solids than by the temperature.
6.3 Slopes of Larger Sewers
Figure 3-13 and the explanation following it was prepared on the basis of all of the data now
available to show what may be expected when certain slope-flow relationships prevail. Thus,
if the average rates of elevation loss are in general below Curve B of Figure 3-13, sulfide
generation will be at such a level that supplemental aeration or some chemical control
method may be considered necessary, and corrosion-resistant structures will generally be
called for. Use of average slopes that are between Curves A and B will be expected to lead
to moderate sulfide concentrations which may cause odor and excessive corrosion problems
at points of turbulence. If the slopes are above Curve A, sulfide is expected to be negligible.
The slopes shown in Figure 3-13 apply where the daytime EBOD in the summer, or
climactic EBOD, is 500 mg/1. The slopes should be increased or decreased in proportion to
the square root of the climactic EBOD.
It must be observed that the scale of rates of elevation loss does not mean the actual slopes
of uniform reaches, but the total elevation loss (or, more strictly, energy loss) divided by
distance. Distance traveled by the wastewater in one-hour intervals can be used in
determining average energy gradients.
Reasonable judgment must be used in applying Figure 3-13. Even if the desired average
energy gradient is attained, the velocity should nowhere be less than 1.5 ft/sec under daily
peak flow conditions, and preferably 2.0 ft/sec, lest sulfide be produced in amounts that
will cause trouble in subsequent steeper sections. It must also be remembered that an
upstream sulfide producer such as a pressure main can yield sulfide concentrations that may
persist for a rather long distance downstream.
6-3
-------
The slopes that should actually be used in a given situation will depend upon objectives
and upon many factors other than flow and EBOD. Consideration needs to be given to
odor problems, kind and size of the pipe, economics, effects of "septic" wastewater on
treatment plant operation, etc. If the sewage is to be pumped, a saving of pumping head
should never, by itself, be a reason for using slopes that will cause significant sulfide
buildup. The energy dissipated by the stream as it loses elevation accomplishes some degree
of oxidation. Thus, more energy spent for pumping will result in a smaller power re-
quirement at the biological oxidation plant. Where good slopes prevail, the amount of
in-sewer oxidation may be substantial (3). However, where the attainment of slopes
corresponding to Curve A would mean multiplying the number of pump stations ex-
cessively, or would require extremely expensive sewer construction deep in water-saturated
soil, it may be necessary to settle for less. Suitable procedures will then be called for to
prevent odor and corrosion problems.
CASE HISTORY
The Los Angeles County Sanitation Districts constitute a joint venture of 27 districts to
provide solid and liquid waste disposal facilities for a population of 4.0 million persons in
an area of 724 sq mi. Except for ten detached districts, the service area lies generally south
to northeast of the City of Los Angeles. A major part of the industrial area of Los Angeles
County lies in the Sanitation Districts. The Districts operate 1,050 miles of trunk sewer, to
which are connected nearly 8,000 miles of smaller sewers owned by 71 cities plus other local
agencies.
Seventeen contiguous districts, including 96 percent of the total population, operate the
main trunk sewer system shown in Figure 6-1. Many of the lines shown in the figure
represent 2 or 3 parallel trunks, necessitated by large expansions of the original service area
and intensive development. This system is connected to the Joint Water Pollution Control
Plant (JWPCP), where the collected wastewater is treated before discharge by way of
tunnels and outfalls to the ocean. However, six upstream water reclamation plants withdraw
wastewater from the trunk sewers for local treatment, with return of sludge to the trunks.
About 380 mgd, or 85 percent of the total flow generated in the main system, reaches the
JWPCP.
During the first years of operation of the system, large trunks carrying flows as small as one
percent of the design capacity were the site of severe sulfide problems (4). Eight chlorination
stations were installed on the trunk sewers and were operated for several years, until no
longer needed. On the basis of research done by the Districts (5), design standards were
adopted which in general called for slopes that would produce velocities of 3 ft/sec in the
trunk sewers. The Districts have continued to use this standard with satisfactory results.
Occasionally sewers were laid at flatter slopes; objectionable sulfide generation has often
resulted in these cases. Thus, the standard that was adopted has been successful, even
though rather crude in the light of the more extensive information now available.
6-4
-------
FIGURE 6-1
LACSD MAIN TRUNK SEWER SYSTEM
LOS ANGELES COUNTY
SANITATION DISTRICTS
MAIN TRUNK SEWER SYSTEM
A WHTTIER MARROWS WRP
A POMONA WRP
A LOS COTOTES WRP
A LONG BEACH WRP
»MTPERtO(« TREATMENT PtANT ft
OI/'Ffl^L (L. ft CIT*!
O «HIT£S PO NT PUMPING PLANT
6-5
-------
The topography in the upper parts of the system, especially in the San Gabriel Valley,
provides fall well in excess of the requirements of Curve A of Figure 3-13. Figure 6-2 shows
profile of the land surface along the routes of trunk sewers to the foothills north of
Pasadena and to the eastern end of the system at Claremont. There are local areas in the
San Gabriel Valley where inadequate slopes allow significant sulfide buildup, but the
total oxygen resources of the upper part of the system are such that dissolved sulfide
concentrations in the major trunks leaving the Valley by way of Whittier Narrows are zero.
Over much of the coastal plain the topography is less favorable, and the options for sewer
routing do not allow full benefit to be taken of the natural slopes. Approaching JWPCP,
the terrain is quite flat. The average condition of the major part of the system below
Whittier Narrows can be described as intermediate between Curves A and B of Figure 3-13,
but for a distance of about 8 mi upstream from JWPCP it is below Curve B. The annual
average total sulfide concentration reaching JWPCP is 0.66 mg/1. The figure would be a
few hundredths of a mg/1 higher if it were not for chemical sulfide control in local trouble
spots, as described in Chapter 5. Dissolved sulfide is only 0.2 to 0.3 mg/1 at JWPCP. It is
believed that the high iron content of the wastewater helps to keep dissolved sulfide to
this low level.
The City of Long Beach and other areas along the coast gain access to the system by
pumping.
The Los Angeles system, serving the City of Los Angeles and several adjacent cities, is
similar to the County system, and the physiography of the area served is similar. The
annual average dissolved sulfide concentration at Hyperion is 0.3 to 0.4 mg/1. Most of this
sulfide content is produced in the last few miles upstream from Hyperion.
6.4 Sewer Sizes
If pipe size is decreased, slope and flow remaining the same, the velocity is virtually
unchanged (6), but mean hydraulic depth (area of the cross section of the stream divided
by surface width) is increased, and hence the reaeration rate of the stream is diminished.
The effect is not large as long as the pipe is less than half full. When the pipe is nearly full,
the mean hydraulic depth increases rapidly, seriously diminishing the oxygen supply. For
the same quantity of wastewater and slope, the largest practical pipe size is the best if
sulfide generation is determined to be a problem.
6.5 Points of Turbulence
A condition of excessive turbulence that mixes air with the wastewater results in a sub-
stantial input of dissolved oxygen. At the same time, it results in a corresponding increase
in the rate of release of ^S, if any is present in the wastewater, as well as the release of
other odors. If a wastewater stream does not contain ^S, and if the other odors do not
cause a problem, turbulence is advantageous. Much more oxygen is dissolved in a fall than
6-6
-------
FIGURE 6-2
GENERAL PROFILES OF PORTIONS OF LOS ANGELES
COUNTY SANITATION DISTRICTS SYSTEM
I-
UJ
UJ
UJ
>
UJ
UJ
V)
Ul
>
o
CD
z
o
UJ
1200
1000
800
600
400
200
10 20 30 40
DISTANCE FROM JOINT WATER POLLUTION
CONTROL PLANT, MILES
50
6-7
-------
during the same loss of elevation in smooth flow. If a wastewater flow makes an abrupt fall
of one foot, it will generally dissolve more oxygen there than it will in a hlaf mile of flow at
a slope of 0.1 percent.
In any real system, points of turbulence are unavoidable, and in the large trunk sewers
sulfide concentrations of a few tenths of a mg/1 are common. Therefore there may be
corrosion of manholes and unprotected concrete pipe, and there may be significant
release of 1198 into the atmosphere.
The best way to deal with the odor problem is to design structures in such a way that the
released odors are confined. Normally air moves downstream in the sewer. If there is to be
a hydraulic jump, it should be (and usually is) in the pipe downstream from* the manhole.
A major junction may be constructed within a vault or chamber entered by a manhole
at its upstream end and so designed that air that has been exposed to the high turbulence
of the junction will be carried downstream and not exhaled from the manhole. If the
wastewater will carry sulfide, the junction structure will probably need to be protected by
a lining, and the pipe, too, will need to be protected or of non-corrodable material. The
distance downstream that abnormal sulfide concentration will prevail will depend upon
the turbulence of the air stream, its velocity, size of the pipe, and perhaps some other
factors. The F^S concentration of the air will probably be near a steady state value at a
distance downstream from the point of turbulence equal to 100 to 200 times the pipe
diameter.
6.6 Pump Stations and Force Mains
6.6.1 The Pump Station Wet Well
The wet well is seldom the site of substantial sulfide generation, even though it is often
the place of odor release. The wet well can be used as a place to aerate the wastewater.
Since this is a technique that can be applied in existing systems, it is discussed in Chapter 5,
but the possibility should be considered when a new system is being designed. Also, a
bypass for air around the pump station to control odor release from the wet well should
be included in the design if the pressure main is not so long as to make this impractical.
An air jumper or bypass around a siphon usually has a diameter half the diameter of the
sewer (7), but around a pump station it is probably better to make it larger, perhaps 2/3 of
the pipe diameter. Figure 6-3 is a schematic drawing of a pump station with an air bypass.
In stations that have no pressure main other than the riser, it may be desirable to consider
air lift pumps. The double duty of an air lift in aerating and pumping the water places it
in a more favorable light economically than when it is considered only as a pump.
An air lift pump for a wastewater pumping station would take the form of a U-tube, with
the air added near the bottom of the rising leg. The tube would probably be installed in a
6-8
-------
FIGURE 6-3
FUNCTIONAL DRAWING OF A PUMP STATION
WITH AN AIR BYPASS
TIGHTLY SEALED
WET WELL
AIR BYPASS
PRESSURE
MAIN
WASTEWATER
6-9
-------
shaft, as shown in Figure 6-4. It is not necessary that the shaft be dewatered. The com-
pressor should take suction from the air bypass connecting the wet well and the downstream
sewer, thus eliminating any net discharge of air.
FIGURE 6-4
FUNCTIONAL DRAWING OF AN AIR-LIFT PUMP STATION
TIGHTLY SEALED
WET WELL
AIR INJECTION COLLAR
6.6.2 Pressure Mains—Size and Profile
Pressure mains have sometimes been made of minimum practical diameter with the aim of
reducing sulfide buildup. The benefit of shorter detention time is partly offset by the
greater ratio of slime-supporting pipe wall to volume of water. Within the range of pipe
sizes that are practical in view of other restraints, choice of the smallest pipe size will not
materially aid in solving the sulfide problem.
Since injection of compressed air into pressure mains serves so well to control sulfide
generation in many situations, thought should be given to this procedure when the main
6-10
-------
is being designed. If possible, the slope should be continuously upward. The design of the
air injection system is discussed in Chapter 5. Where air injection would be ineffective
because of flat slopes, or impractical because of irregular profile, plans should be made for
some other way to supplement the dissolved oxygen supply.
6.7 Materials for Sewer Construction
6.7.1 Kinds of Pipe
The possibility of acid conditions resulting from oxidation of H^S is one of the factors
influencing the choice of pipe materials for sewers. If acid-proof facilities are built, then
this harmful consequence of sulfide can be overcome.
Whether to construct an acid-resistant system depends upon a number of considerations.
These considerations are:
1. What choices are there for pipe materials in the sizes required?
2. Will sulfide be present in the wastewater, and, if so, in what general range of
concentrations?
3. Are there any factors other than acid resistance that will affect the prospective
durability?
4. How do the materials compare in hydraulic smoothness?
5. What other incidental benefits and liabilities do the various kinds of pipe have?
6. What are the expected future service requirements of the sewer?
7. What are the relative costs for different kinds of conduits and how are these related
to the durability of the sewer?
A cost-benefit analysis based upon the answers to the last question would permit a choice
on the principle of lowest present worth of all present and future costs. This principle,
rigorously applied, may not serve as a satisfactory sole criterion for design in the case of
public works such as sewers having a long intended life with uncertainties in respect to
future replacement possibilities, inconveniences in case of failures, etc. Decisions will vary
in different communities under seemingly identical conditions. In any case, reasonable
judgment would dictate a choice somewhere between such extremes as saving a small amount
in calculated present worth but producing a sewer that would last only 25 years instead of
50 or 100 years, and the other extreme of spending a significant additional sum to extend
to 500 or 1,000 years the life expectancy of a sewer which at lower cost would last 100
years. Beside the economic factor, there is much uncertainty about the continuing utility
of any particular sewer for such long times.
6-11
-------
In view of these considerations, it will be evident that it is not possible to provide a fixed
set of rules that will automatically determine the choice of materials or the life expectancy
that should be built into a sewer. Rather than to attempt any such guidelines, the principal
kinds of pipe used in sewer construction will be listed and discussed, particularly in respect
to susceptibility to acid attack, but with brief mention of other characteristics affecting
their suitability. Protective coatings for corrodable pipe materials are discussed in other
sections.
6.7.2 Vitrified Clay Pipe
This material is immune to alteration by sulfuric acid or by any other materials found in
sewers. Where sulfide is expected, the use of cement mortar joints is unsatisfactory, because
the action of sulfuric acid on the cement causes expansion, which may break the bells and
even crack the pipe. A vitrified clay sewer properly laid and jointed will, if not disturbed
by external forces, remain serviceable for a very long time.
6.7.3 Steel Pipe
If the pipe flows partly filled with wastewater containing sulfide, corrosion may occur not
only by the formation of sulfuric acid, but also by the corrosiveness of H^S toward iron in
the presence of air or dissolved oxygen, producing bulky accumulations of iron sulfide. In a
pipe completely filled with wastewater there is little or no corrosion even if sulfide is
present, provided the pH is above 6.5 and chloride is less than 500 mg/1. Even where air is
injected, corrosion due to dissolved oxygen is generally at a very slow rate, and no corrosion
by acid produced from h^S is possible unless there are air pockets where the top of the pipe
is not washed by the water. If the steel pipe has a half inch or more of cement mortar lining,
then it will be protected as long as it is not exposed to enough acid to destroy the protective
lining.
6.7.4 Cast Iron Pipe
Cast iron pipes generally last longer than steel because the pipe wall is thicker. The corrosion
of cast iron exposed to water commonly proceeds by "graphitization," in which the true
iron crystals are dissolved, leaving a porous mass of carbides and silicides of iron. The
surface of the iron often appears unaltered, thus giving a false impression of the true
condition of the pipe. Like steel, cast iron gives good service when completely filled with
wastewater at a pH of 6.5 or above, and without a high chloride content. When flowing
partly filled with septic tank effluent, iron pipe is severely corroded, and may become
occluded with the bulky iron sulfide corrosion product. If there is a point of turbulence,
attack is accelerated. Photographs of sulfide attack in iron pipes have been published (8).
6.7.5 Wrought Iron, Malleable Iron Pipe
In atmospheric exposure, these forms of iron may have corrosion behavior different from
common steel, but when submerged in water or buried in the soil, all common forms of
6-12
-------
iron corrode at similar rates. If sulfuric acid is formed above the water, it will attack any
form of iron.
6.7.6 Iron or Steel Pipe with Tar Mastic Lining
Pipe made of corrugated galvanized steel, and having a tar mastic filling the corrugations to
produce a smooth internal surface has been used for sewers. Long-term durability in the
presence of H^S has not yet been established.
6.7.7 Aluminum Pipe
This metal is not resistant to corrosion under anaerobic conditions. It is not suitable for
conveying wastewaters.
6.7.8 Stainless Steel Pipe
This metal, too, can suffer severe corrosion in the absence of oxygen. It is therefore not
suitable for pipes to carry wastewater but is sometimes used for steps in manholes.
6.7.9 Concrete Pipe
The corrosive effects of H^S have potential for deterioration of concrete pipe. Examples
have been reported (9) (10). Nevertheless, concrete is an important sewer pipe material,
and for large trunk construction it leads any other material by a wide margin. It is
successfully used for small sewers as well, provided sulfide concentrations are low.
If the flow-slope relationships of sewers upstream from a given point correspond to Curve A
of Figure 3-13 (adjusted for EBOD), and there are no force mains operating without proper
sulfide control, then sulfide concentrations will be so low that the rate of corrosion of
concrete pipe will be inconsequential. Small collecting sewers so designed, made of concrete
pipe with granitic or other inert aggregate will have a life expectancy of 100 years or more.
Using calcareous aggregate, the life will be much longer. Sulfide concentrations will be
greater in large trunks, perhaps up to a few tenths of a mg/1, but seldom averaging more
than a few hundredths of a mg/1 of dissolved sulfide. The rate of corrosion will be slow
because of the small values of su (slope times velocity) in the large trunks. Considering
the thickness of the pipe wall in the very large pipes, a life of several centuries would be
expected, or longer if calcareous aggregate is used. It is possible, however, that there may
be significant attack at locations of high turbulence.
Where over-all slopes are as represented by Curve B of Figure 3-13, sulfide conditions under
some circumstances may be such that bare concrete pipe made with granitic aggregate will
be significantly corroded. Sulfide conditions become worse at flow-slope combinations
deeper in the domain below Curve B.
6-13
-------
It may be satisfactory to use concrete pipe under these conditions if it is made with
calcareous aggregate, but the probable sulfide conditions and the rates of corrosion should
be estimated. In some cases plans should be made for supplementing the oxygen sources of
the stream by methods described in Section 5.1.4, or for controlling sulfide in some other
way. Even with granitic aggregate, unprotected concrete may be suitable if the prospective
rates of corrosion, calculated as explained in Sec. 3.6.2, are shown to be tolerable.
Sometimes the wall thickness is increased so that the concrete will outlast the corrosive
forces. This is a less satisfactory strategy than to keep the same dimensions and use
calcareous aggregate, where such aggregate is available. Under severe conditions calcareous
aggregate pipe with extra wall thickness may be indicated.
Modifications of concrete pipe other than by use of calcareous aggregate have been tried in
an effort to reduce the rate of attack. Test specimens may corrode at various rates when
immersed in reservoirs of dilute acid. The principal factor determining the rates of
corrosion in a sewer, however, is the rate at which acid is produced, but with the rate also
varying inversely with the alkalinity of the material. Any one of the usual formulations of
concrete can react as fast as the acid produced on the sewer wall can get to it. Cement
type is unimportant, and the addition of a pozzolanic type of material has no effect. If a
pozzolan is substituted for part of the cement, the rate of attack is increased because the
alkalinity is decreased (11) (12). Attempts have been made to impregnate the pores of the
concrete with wax, sulfur, tar, or resin, but these methods are unsuccessful, because acid
does not reach the interior of the concrete by way of the pores, but by attacking the solid
phases. Treating the pipe with SiF^ produces a resistant surface, but it protects the
concrete in the presence of r^S only for a short time, since the surface is not impermeable.
6.7.10 Asbestos-Cement Pipe
Asbestos-cement pipe is susceptible to attack by sulfuric acid, and it therefore is not
suitable where very high sulfide concentrations will prevail. However, the cement content
of A-C pipe is higher than for concrete pipe with granitic aggregate, with a correspondingly
lower rate of corrosion. This potential benefit may be offset by the thinner pipe wall.
6.7.11 Plastic Pipes—Homogeneous
Pipes of polyvinyl chloride (PVC), acrylonitrile-butadiene-styrene (ABS), and polyethylene
(PE) have been used for sewers in smaller sizes. The materials are all resistant to sulfuric
acid attack.
6.7.12 Plastic Pipes—Composite
Polyester resin is mixed with sand and then reinforced with fiberglass to make a light-
weight "reinforced plastic mortar" pipe. The most vulnerable part of the composite is the
glass, because water has the ability to creep along the fibers. The pipe must be manufactured
so that this cannot happen. The oldest sewer made of this kind of pipe is in a trunk
installed in San Jose, California, in 1966.
6-14
-------
6.8 Protection of Concrete and Asbestos-Cement Pipe by Linings
6.8.1 Linings that Depend upon Adhesion
Since the early 1920's, experiments have been made on acid-proof linings on the inside of
concrete pipe, and later asbestos-cement pipe. More money has been expended on these
experiments than on any other aspect of sulfide research. Many cities have made com-
parative experiments with various linings, repeating fruitless efforts of others.
The reason that so many attempts have been unsuccessful is because the requirements are so
exacting. Once a lined sewer is in the ground, it is expected to perform with no defects of
material and no repairs, generally for 100 years or more. A successful lining material must
be not only immune to acid attack; it must be impermeable to the diffusion of acid. If one
installation in ten fails, even if only because of imperfect workmanship, the material is a
failure. One flaw per 1,000 ft of pipe is intolerable, because it may mean expensive repairs
by excavation. Furthermore, the joints must be as perfect as the rest of the pipeline.
An almost-perfect lining may cause earlier failure than no lining at all, because acid
migrating down an unreactive wall may cause more rapid penetration of a flaw than if the
attack were more evenly distributed.
One may find examples of linings applied to the concrete surface that have remained intact
for many years. In such cases, it is always found that F^S was virtually absent.
It may take a long time for the inherent weaknesses of a material to show up. Pores of only
molecular dimensions may allow the very slow diffusion of sulfuric acid, but as the
underlying cement is affected the adhesion is destroyed and bulges appear, stretching the
protective membrane and increasing its porosity until there is a rupture. Linings are often
tested by standing a specimen of the pipe on end and filling with 5 percent sulfuric acid.
Such lining tests should be conducted for a period in excess of one year.
6.8.2 Linings Keyed to the Concrete
Under this heading we do not include materials that have a presumed tie by virtue of
roughness of the concrete or even interpenetration of the materials. There is no evidence
that this sort of tieing of the lining to the pipe wall has been significantly better than
adhesion. Linings that are really locked in place have large-scale keys embedded in the
concrete.
6.8.2.1 Vitrified Clay Liner Plates
The first material attempted on a large scale for lining concrete sewers was vitrified clay.
Since vitrified clay itself is so durable in acid exposures, it seemed like an ideal material
for this purpose. Clay plates were formed, generally about 9 by 18 inches in size, with
lugs or keys formed on the back side. The plates were fastened to the inner forms before
pouring the concrete to form the pipe.
6-15
-------
The weakness of this system was the porosity of the clay. Acid diffused through, softening
and expanding the cement, so that the plates cracked or the lugs were broken off.
6.8.2.2 Keyed Plastic Sheet
The principal material in this class used in sewer construction is a sheet of plasticized
polyvinylchloride, about 1/16-in thick, having keys shaped like tees in cross section, running
longitudinally on the sheet. The sheets are fastened on the inside forms before the pipe is
poured, so that in the finished pipe the tees are imbedded in the concrete. The sheets are
welded at the joints by gentle heating.
With this design, an effective lining is produced. The oldest installations (Los Angeles,
California) in normal sewer construction have been in service since the early 1950's with no
evidence of significant deterioration.
Difficulties have been encountered in attempting to install this kind of lining in cast-in-place
structures, especially in tunnels, because it is difficult to pour the concrete and effectively
imbed all of the keys. Occasionally sheets have pulled loose from the wall. No such
difficulty arises in factory-made pipe. The lining may be torn by severe hydraulic stresses
at points of very high velocity or turbulence, and it can be damaged by harsh cleaning
tools. These difficulties can be circumvented by proper design and operation, and are not
considered significant deterrents to its use.
6.9 References
1. Pomeroy, R. D., Sewer Maintenance in Long Beach, California, Sewage and Industrial
Wastes, 29, pp. 320-325 (March, 1957).
2. Federal Water Pollution Control Act Amendments of 1972 (PL 92-500).
3. Pomeroy, R. D., and Parkhurst, J. D., Self-purification in Sewers, Proceedings of the
6th International Conference on Water Pollution Control Research, Jerusalem,
Pergammon Press (June 1972).
4. Bowlus, F. D., and Banta, A. P., Control of Anaerobic Decomposition in Sewage
Transportation, Water Works and Sewerage, 79, 369 (1932).
5. Pomeroy, R. D., and Bowlus, F. D., Progress Report on Sulfide Control Research,
Sewage Works Journal, 18, No. 4, pp. 597-640 (1946).
6. Pomeroy, R. D., Flow Velocities in Small Sewers, Journal Water Pollution Control
Federation, 39, No. 9, pp. 1525-1548 (1967).
6-16
-------
7. Design of Sanitary and Storm Sewers, Water Pollution Control Federation Manual
of Practice No. 9, p. 157 (1969).
8. Pomeroy, R. D., Corrosion of Iron by Sulfides, Water and Sewage, 92, pp. 133-138
(1945).
9. Swab, B. H., Effect of Hydrogen Sulphide on Concrete Structures, Journal of the
Sanitary Engineering Division, ASCE (Sept., 1961).
10. Thistlethwayte, D. K. B. (Editor), Control of Sulphides in Sewerage Systems,
Butterworths Pty. Ltd., Melbourne, Australia (1972), and Ann Arbor Science
Publishers, Ann Arbor, Michigan (1972).
11. Pomeroy, R. D., Protection of Concrete Sewers in the Presence of Hydrogen Sulflde,
Water and Sewage Works (October, 1960).
12. Pomeroy, R. D., Calcareous Pipe for Sewers, Journal Water Pollution Control
Federation, 41, No. 8, Part 1, pp. 1491-1493 (1969).
6-17
------- |