Uniad Stales
EnvwonmarrtaJ Protacaon
Agere/
Otlice CI Water
(WH-5471
430/09-91-010
September 1991
Hydrogen Sulfide Corrosion
In Wastewater Collection And
Treatment Systems
Report To Congress
Technical Report
Printed on Recycled Paper
-------
TECHNICAL REPORT
HYDROGEN SULFIDE CORROSION IN WASTEWATER
COLLECTION AND TREATMENT SYSTEMS
U.S. Environmental Protection Agency
Office of Water (WH-595)
Washington, DC 20460
May, 1991
-------
ACKNOWLEDGEMENT
This document was prepared by the U.S. Environmental Protection Agency (EPA)
with the assistance of J.M. Smith & Associates, PSC, Consulting Engineers (JMS)
under subcontract to HydroQual, Inc. (EPA Contract No. 68-C8-0023). JMS
employees who made major contributions to the document included
Robert P.G. Bowker, John M. Smith, Hemang J. Shah, and Peter A. Flaherty.
Previous reports were prepared in 1988 by EPA with the assistance of E.C. Jordan
Co. under EPA Contract No. 68-03-3412. Under subcontract to E.C. Jordan Co.,
Brown and Caldwell staff, including Peiry Schafer, Robert Witzgall, Roy Fedotoff,
and Walt Meyer, prepared five case studies of corrosion.
Many people provided valuable assistance in the preparation of this study.
However, special acknowledgment is appropriate for the staff of the County
Sanitation Districts of Los Angeles County, especially the support from
John Redner and Calvin Jin.
Ms. Irene M. Suzukida [Horner] was the Work Assignment Manager for this report
This report is dedicated to ail the individuals who work to preserve the
wastewater systems of this country, whose contributions are too
numerous to identify.
-------
DISCLAIMER
This report contains discussions of several proprietary products and processes used
for the control and prevention of corrosion induced by hydrogen sulfide. Mention
of trade names or commercial products does not constitute endorsement by EPA or
recommendation for use.
For this report, information was not collected for all products and processes, and
omission of products or trade names from this report does not reflect a position of
EPA regarding product effectiveness or applicability.
ii
-------
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENT i
DISCLAIMER ii
1.0 BACKGROUND AND OVERVIEW 1-1
1.1 Legislative Charge 1-1
1.2 Los Angeles County System History 1-1
1.3 Consequences of Corrosion 1-4
1.4 Mechanism of Hydrogen Sulfide Corrosion 1-5
2.0 NATIONAL ASSESSMENT OF CORROSION 2-1
2.1 Introduction 2-1
2.2 Site Visits to Assess Hydrogen Sulfide
CoiTosion in Sewers 2-2
2.3 Site Visits to Assess Hydrogen Sulfide Corrosion at
Wastewater Treatment Plants and Pump Stations 2-20
2.4 Site Visits to Investigate Corrosion Mechanism 2-29
2.5 Other Cities Reporting Hydrogen Sulfide Corrosion 2-32
2.6 Case Studies 2-46
2.7 Hydrogen Sulfide Corrosion in Other Countries 2-60
2.8 Conclusions 2-61
3.0 EFFECTS OF INDUSTRIAL PRETREATMENT 3-1
3.1 Overview 3-1
3.2 Theoretical Impacts of Sulfide Precipitation
by Metals 3-2
3.3 Biological Inhibition by Metals and Toxic Compounds 3-5
3.4 Comparison of Metals at LA County with Other Cities
Before Pretreatment 3-15
3.5 Site Visits to Industrialized Cities 3-20
3.6 Beneficial Effects of Local Industrial Pretreatment Programs 3-22
3.7 Conclusions 3-22
iii
-------
TABLE OF CONTENTS (cont)
Page
4.0 DETECTION, PREVENTION AND REPAIR OF HYDROGEN SULFIDE
CORROSION DAMAGE 4-1
4.1 Detection and Monitoring of Hydrogen
Sulfide Corrosion 4-1
4.2 Prevention of Hydrogen Sulfide Corrosion
in Existing Systems 4-2
4.3 Prevention of Hydrogen Sulfide Corrosion
in the Design of New Systems 4-7
4.4 Repair of Damage Caused by Hydrogen Sulfide Corrosion 4-14
4.5 Conclusions 4-19
APPENDIX A ANNUAL AVERAGE WASTEWATER CHARACTERISTICS FOR LA
COUNTY 1971-1986
iv
-------
LIST OF TABLES
Page
1-1 Factors Affecting Sulfide Generation and Corrosion
in Sewers 1-9
2-1 Selected Information from Site Visits 2-4
2-2 Summary of CSDLAC Survey Data 2-33
2-3 AMSA Survey Summary 2-37
2-4 Summary of Responses to WPCF Survey - Corrosion of
Wastewater Treatment Systems 2-39
2-5 Selected Information from Thirty-Four Cities 2-41
2-6 Summary of Information from Selected Associations,
Manufacturers and Contractors 2-43
3-1 Probable Metal - Sulfide Precipitation Reactions in
Wastewater Devoid of Oxygen 3-3
3-2 Theoretical Increase in Dissolved Sulfide Based on Metal
Precipitation; LA County 3-4
3-3 Toxicity of Wastewater Constituents on Sulfate Reducing
Bacteria 3-7
3-4 Concentration of Agents Added to Upflow Packed
Columns 3-10
3-5 Average Influent Sulfide, Total COD, Suspended Solids
and Effluent Sulfide; Upflow Packed Columns 3-11
3-6 Comparison of Control and Test Columns' Sulfide
Generation Upflow Packed Columns 3-12
3-7 Average Influent and Effluent Sulfide; Pipeline
Pilot Plant 3-16
v
-------
LIST OF TABLES (cont)
Page
3-8 Comparison of Control and Test Pipeline Sulfide
Generation: Pipeline Pilot Plant 3-16
3-9 Ranking of Cities by Levels of Metals and Cyanide
in Wastewater . 3-18
3-10 Metals and Cyanide Concentrations in Wastewater
from 51 Cities . 3-19
3-11 Beneficial Impacts of Controlling Industrial
Discharges on Hydrogen Sulfide Corrosion 3-23
4-1 Summary of Sulfide Control Techniques 4-3
4-2 Approaches to Prevent Hydrogen Sulfide Corrosion During Design 4-8
4-3 Principal Methods for Pipeline Rehabilitation 4-15
vi
-------
LIST OF FIGURES
Page
1-1 Processes Occurring in Sewers Under Differing Conditions 1-7
2-1 States Having Severe Corrosion Problems in Wastewater
Systems of Four or More Municipalities 2-44
2-2 Use of Proprietary PVC Lining to Prevent Corrosion
^ of Concrete Pipe 2-45
3-1 Sulfide Generation Pilot Plant 3-8
3-2 Percent Change in Sulfide Generated Due to Metals
and Cyanide; Upflow Packed Column Pilot Plant 3-13
3-3 Sulfide Generation Pilot Plant 3-14
3-4 Percent Change in Sulfide Generated Due to Metals and
Cyanide 3-17
4-1 Guide for Estimating Sulfide Generation Potential 4-9
vii
-------
-------
1.0 BACKGROUND AND OVERVIEW
1.1 Legislative Charge
In the mid 1980's, the County Sanitation Districts of Los Angeles County
(CSDLAC) observed that the rate of corrosion of concrete sewer pipe in their system
had increased dramatically since the early 1970's. Subsequent studies showed a high
correlation between the reduction in the levels of metals and other wastewater
constituents and the increase in levels of hydrogen sulfide responsible for the corrosion.
Metals and other constituents of industrial origin had been reduced through
implementation of industrial pretreatment standards in 1975-1977 (ocean discharge
requirements) and in 1983 (EPA categorical pretreatment standards). This raised the
question of whether implementation of industrial pretreatment standards had resulted in
an increase in corrosion rate, which would have significant economic implications.
Section 522 of the Water Quality Act of 1987 requires the U.S. Environmental
Protection Agency (EPA) to conduct a study and prepare a report on corrosion in
wastewater collection and treatment systems and to coordinate its activities with the City
and County of Los Angeles. Section 522 of the Act specified that:
The Administrator shall conduct a study of the corrosive effects of sulfides [sic]
in collection and treatment systems, the extent to which the uniform imposition
of categorical pretreatment standards will exacerbate such effects, and the range
of available options to deal with such effects (1).
The study concentrated on the three areas mandated by the Act Many factors
influence corrosion besides the implementation of pretreatment requirements, such as
solids deposition, turbulence, temperature, and so on. The lack of an accurate
corrosion-measuring technique and the limited data base on hydrogen sulfide corrosion
would have limited the ability of EPA to ascertain the effects of these factors. In
addition, the study did not explore the impacts of transporting sewage further to
regional treatment plants, constructing separate sewers for sanitary wastewater and
storm water, or implementing water conservation programs. This report does not
discuss the problems caused by the toxicity of hydrogen sulfide gas or the odor nuisance
associated with its presence, as these issues were not mandated by the Act
1.2 Los Angeles County System History
The County Sanitation Districts of Los Angeles (CSDLAC) provide wastewater
collection, treatment, and disposal services for approximately four million residents of
Los Angeles County. The service area covered by CSDLAC, 640 square miles in size,
includes most of metropolitan Los Angeles County with the exception of the City of Los
Angeles. Wastewaters from residential, commercial, and industrial sources, totalling
over 500 million gallons per day (gpd), are conveyed through 9,000 miles of collection
l-l
-------
sewers to six wastewater treatment facilities. Approximately 1,000 miles of these sewers
are owned and maintained by CSDLAC, while the remaining 8,000 miles are owned and
maintained by local cities or Los Angeles County (2).
Most of the CSDLAC collection sewers, especially the large-diameter lines in the
lower reaches of the tributary system, are constructed of reinforced concrete pipe with
no protective coatings or liners. These large sewers generally range in size from 54
inches in diameter up to 144 inches in diameter. The oldest of these sewers have been
in service for approximately 65 years.
At the time these sewers were designed, concerns existed about the possibility of
conosion. To guard against this possibility, the earliest of the large sewers had vitrified
clay liner plates installed on the interior sides and crowns. However, sulfuric acid easily
penetrated the joints between the tiles and destroyed the grouting and cementitious
materials underneath. By the late 1930's after approximately 10 years of service, enough
of the tiles had fallen off into the bottom of the pipes to create flow obstructions and
necessitate cleaning of the debris from these pipes.
Because of the problems experienced with the tile liners, CSDLAC looked for
another method to prevent corrosion damage. The Districts chose to design sewers to
induce sufficient wastewater velocities so that natural reaeration would minimize the
growth of the anaerobic slime layers on the submerged pipe walls where the sulfide-
geoerating bacteria grow. Such natural reaeration forces would also help oxidize any
sulfide that did form in the wastewater, preventing its release to the sewer headspace as
hydrogen sulfide gas.
In the early 1950's, concrete pipe manufacturers began to market internally lined
pipes to protect against hydrogen sulfide corrosion. However, at that time, little
information was available to document how well these plastic liners would remain
securely bonded to the concrete and provide effective protection. The lined pipe was
expensive when compared to regular, unlined pipe, and CSDLAC decided to rely on
high design velocities to control conosion, rather than lined pipe. Consequently, during
the 1950's and 1960's, as the size of the collection system increased dramatically,
CSDLAC continued to install unprotected, reinforced concrete pipe for much of the
sewer system. Current County standards require lined concrete pipe in all new
installations to prevent corrosion.
By the mid-1960's, sulfide generation was increasing within CSDLAC major trunk
sewers, especially at locations where depletion of available dissolved oxygen (DO)
occurred. To protect its substantial capital investment in unprotected concrete pipe
sewers, CSDLAC undertook a three-year research program in 1968. The objectives of
the research program were to better understand the processes by which sulfide is
generated by the Desulfovibrio bacteria and to develop methods to control these
bacteria. This research was partially funded by federal agencies that would later be
1-2
-------
merged to form EPA.
The research was conducted by Dr. Richard D. Pomeroy, who in the 1940's had
studied sulfide generation in the CSDLAC system. Through measurements made at a
number of monitoring stations throughout the CSDLAC sewerage system, he refined an
empirical formula which predicts sulfide generation rates and resulting concrete
corrosion rates. A final report of the study, entitled "Sulfide Occurrence and Control In
Sewage Collection Systems" was published in 1973 (3). The predictive formula is also
included in more recent design manuals and guidance documents (4)(5)(6)(7).
In the early to mid-1970's, CSDLAC conducted an inspection of the wastewater
collection system and concluded that actual corrosion matched closely the coiTosion
predicted by Pomeroy's formula. Based on the estimated rates of corrosion, CSDLAC
calculated that the remaining structural lives of most of the sewer pipes ranged from at
least several decades for the oldest of the sewers, up to hundreds of years for most of
the post-World War II sewers.
In the early 1980's, a second inspection of these same sewers was made, with very
different results. In less than 10 years, reinforcing steel had become exposed in many
sewers. Based on measurements taken during the inspections, CSDLAC calculated that
corrosion rates increased from 0.01 inches per year to 0.25 inches per year in some
instances. A decrease in pipe surface pH levels from 3 to 4 in the 1970's to 1 to 2 in
1980's accompanied the increase in rate of corrosion. Total sulfide levels entering the
main wastewater treatment plant increased from an average of 0.4 mg/1 in 1971 to 3.0
mg/1 in 1986.
The inspections suggested that the rate of corrosion had increased markedly and
could no longer be predicted with the existing empirical formula. The causes of the
apparent increase of the rate of conosion are not understood. However, CSDLAC has
data that show a strong correlation between an increase in wastewater sulfide levels and
a decrease in levels of cyanide and certain priority pollutant metals regulated by EPA
categorical pretreatment standards. CSDLAC measurements of wastewater constituent
concentrations over the period 1971 through 1986, along with the results of their
statistical correlation analyses, are included in Appendix A.
CSDLAC believes that cyanide and these heavy metals in their system may have
played an important role in inhibiting the biological reduction of wastewater sulfate to
sulfide. In addition, the metals form insoluble metal-sulfide precipitates that would
reduce the amount of hydrogen sulfide released to the sewer headspace.
In the past few years, CSDLAC has implemented an intensive program to control
hydrogen sulfide corrosion by attempting to reduce the growth of the Desulfovibrio
bacteria or to chemically bind the sulfide which is generated using established control
techniques. Adding hydrogen peroxide to the sewage to oxidize the sulfide was tried,
1-3
-------
but the required dosage of hydrogen peroxide was found to be too high to be cost-
effective. Both ferrous chloride and liquid sodium hydroxide solutions are now being
routinely added to selected sewer lines at a cost of almost $2 million per year. The
ferrous chloride is added continuously to form an insoluble iron sulfide precipitate that
is carried in suspended form to the main treatment plant. The sodium hydroxide is
added at a weekly frequency to certain other sewers to provide a 30-minute, high pH
shock to the Desulfovibrio bacteria to inhibit their activity. In addition, pure oxygen has
been added to wastewater in one of the large gravity sewers on an experimental basis.
CSDLAC monitoring of hydrogen sulfide concentrations in the headspaces of the
sewers has in general shown only modest reductions (i.e., 50 to 60 percent) as a result
of these treatments, even though significant (i.e., 75 to 95 percent) dissolved sulfide
reductions have been obtained in the wastewater. Measurements taken of the surface
pH on the crowns of corroding sewers which have received these treatments have risen
one half to two pH units compared to untreated conditions. This is expected to slow
the rate of corrosion and extend the life of the pipes.
CSDLAC has estimated that at least $130 million will be needed to replace or
repair approximately 25 miles of sewers that are severely corroded. An additional 16
miles will likely require repair or replacement within five years.
Approximately 500 miles of sewers show some evidence of sulfide generation but
exhibit no corrosion damage according to CSDLAC. The goal of CSDLAC is to
understand the causes and control of corrosion to prevent damage to these vulnerable
parts of their system.
1.3 Consequences of Corrosion
Conosion of wastewater and treatment systems induced by the presence of
hydrogen sulfide can cause rapid and extensive damage to concrete and metal sewer
pipe, equipment used in the transport and treatment of wastewater, and electrical
controls and instrumentation systems. Such problems are rarely brought to the
attention of the public until a catastrophic failure occurs such as with street collapses
resulting from sewer pipe failure. However, sewer systems suffering from hydrogen
sulfide corrosion generally require costly, premature replacement or rehabilitation of
pipes, manholes, lift stations, and pump stations.
Equipment used in treatment of wastewater is often subject to hydrogen sulfide
corrosion, resulting in equipment malfunctions, poor reliability, increased maintenance,
and premature replacement Electrical components (e.g. brushes, switches, relays)
process instrumentation, air conditioning and ventilation units, and computer systems
are particularly vulnerable to attack by hydrogen sulfide at pumping stations, lift
stations, and treatment plants. This can cause poor reliability of control systems,
increased maintenance requirements, and often premature replacement of costly
1-4
-------
electrical components and computer equipment
Hydrogen sulfide corrosion can also compromise structural integrity by corroding
equipment (bar screens, conveyors, etc.), pipe and equipment supports, wastewater and
sludge storage tanks, and guard rails, walkways, and grating at the treatment plant
1.4 Mechanism of Hydrogen Sulfide Corrosion
Hydrogen sulfide corrosion may result from two mechanisms: 1) acid attack
resulting from the biological conversion of hydrogen sulfide gas to sulfuric acid in the
presence of moisture and 2) the direct attack of metals such as copper, iron, and steel
by hydrogen sulfide gas. The first mechanism is responsible for corrosion of sewers and
concrete structures used in the conveyance and treatment of sewage. The second
mechanism is generally responsible for corrosion of electrical contacts, copper pipe, and
metal components in pumping stations and treatment plants.
First, for hydrogen sulfide to be formed, the wastewater must be anaerobic
(devoid of oxygen). Oxygen is depleted due to the activity of microorganisms. In
properly designed gravity sewers the velocity of the sewage is such that natural
reaeration occurs from the atmosphere in the sewer, helping to replenish any losses of
oxygen due to microbial activity. Certain structures and flow conditions often create
turbulence of the wastewater, increasing the rate of reaeration and helping to maintain
aerobic (oxygenated) conditions. Sources of turbulence include manholes with flows
dropping in from the side, manholes with flows colliding, metering flumes, drops in the
line, sections with steep slopes, and force main discharges.
Under certain conditions oxygen is depleted faster than it is supplied, causing a
change from aerobic to anaerobic conditions. Such conditions can occur in gravity
sewers with low sewage velocities or long detention times, force mains which convey
wastewater through a full pipe under pressure with no opportunity for reaeration, wet
wells of pumping stations having detention times sufficiently long as to cause oxygen
depletion due to uptake by bacteria, and other structures or processes where wastewater
is detained under near-stagnant conditions with insufficient opportunity for reaeration.
Under anaerobic conditions, the microbial community shifts to organisms that can
flourish without oxygen. These may be strict anaerobes that cannot utilize oxygen and
may be sensitive to its presence, or facultative anaerobes which can utilize either free
oxygen or other compounds in their metabolic cycle. The process of sulfide generation
and sulfuric acid coiTosion is as follows (4)(5)(6)(7)(8):
1. Under anaerobic conditions, strict anaerobic bacteria of the genus Desulfovibrio
colonize the wastewater and attach to the slime layer that coats the submerged
surfaces of pipes. The bacteria reduce sulfate (SO*2'), one of the most common
anions in water and wastewater, to sulfide (S2 ).
1-5
-------
2. The sulfide ion combines with hydrogen ions to form dissolved hydrogen sulfide
gas (HZS) and hydrosulfide ion (HS~), depending on pH. At neutral pH of 7, the
distribution of species is approximately 50 percent H2S and 50 percent HS\ At
pH 6, the distribution is approximately 90 percent dissolved H2S gas, and 10
percent HS".
3. Hydrogen sulfide gas is released from the wastewater to the sewer atmosphere.
The dissolved gas (HjS) is the only form of sulfide which can be released. The
release of H2S from solution is accelerated under turbulent conditions and at
higher temperatures. Thus, turbulence may be beneficial in maintaining
wastewater in an aerobic state, but if the wastewater is anaerobic and dissolved
sulfide is present, this same turbulence can cause rapid release of the H2S to the
sewer atmosphere. The H2S produces the "rotten egg" odor characteristic of
stagnating sewage. Since equilibrium conditions are rarely observed, it is virtually
impossible to predict atmospheric H2S concentrations based on Henry's Law.
4. The released H2S combines with moisture on the non-submerged surfaces of the
pipe and is oxidized to sulfuric acid by aerobic bacteria of the genus Thiobacillus.
which colonize the pipe surfaces above the water level. There are many species
of this bacteria which successively colonize the slime layer as sulfuric acid is
produced and the pH drops. More acid-tolerant species then predominate.
While new pipe has an alkaline surface pH, weathered pipes have a surface pH
of about 6, and pipes which are subject to active sulfuric acid corrosion may have
a surface pH of 3 to 1.
5. The hydrogen ions of the acid attack the calcium hydroxide in the hydrated
Portland cement of the concrete sewer pipes, while the sulfate combines with the
calcium ions to form gypsum (CaS04), a soft corrosion product In addition,
calcium sulfoaluminate (3CaO Al203CaS04 31H20), also known as ettringite,
may form. Goth gypsum and ettringite occupy considerably greater volume than
the compounds they replace. This leads to expansion and disruption of the
concrete, and loss of aggregate. Both products are easily washed away by
wastewater, thus exposing fresh material to sulfuric acid. In early stages of
corrosion, the pipe wall swells, making it difficult to measure concrete loss due to
corrosion.
Figure 1-1 summarizes the processes which occur in sewers under aerobic and
anaerobic conditions.
H2S directly attacks metals including iron, copper, and silver. H2S can cause
blistering and embrittlement of ductile iron pipe. Even at low concentrations in the
atmosphere (<1 ppm), H2S can cause extensive damage to electrical contacts and
circuits present in controls, switchgear, and computer equipment
1-6
-------
Pmciiiii occulting In tawaii wlili lulliclaul
oiygan lo piavml aullida tium anluilny Him
tllaailk
Slims
liypically
I nun llntkl
Uiygan Eulailiig ihe Waiuf
Dittolwad Oiyguii >1 ing/|
Diifolvad Sullida 2a 10 ui Tiaca
DiIIumoii ol O, and Nuiiianii
Uilluiiun ol SO, miilHuiilnui
Diffusion andOmuanon ol Sullida j j j~y f
I //I
Pfocaitaa occulting la hmiii undar aullida
buildup condition*
HjS flui lo Plpa Wall —|
Oxidation lu ll|SO,
Alt
IIiS Euiailng ilia All
Oiygan Emailnu ilia Walar
Simla*
tayai
(typically
I HUH
Pipa $
waiiawaiai
Wall v;
OiSkulvoiJ Oxygon <0 I ing/l
DiktulvaJ SiilliiJa Pdklnl. IIS" > II4S
/It!/i
Iviji
I I ~ bc
I r 3 f '
Oiipliilion ol 0| In ilia 1&HIU14I l*yei I ",/l i
Uilluitiuii ol SO« and Nuiiiuius ^
UiIImsiuii ol Sullida Into ilia Suaain-<—
Figure l-l Processes occurilng In scuurs under differing conditions
-------
The rate of sulfide-induced corrosion is affected generally by the characteristics
of the wastewater and the collection system. Many variables directly or indirectly affect
sulfide generation, H2S release, and sulfuric acid corrosion. These variables are
summarized in Table 1-1.
1-8
-------
TABLE 1-1
FACTORS AFFECTING SULFIDE GENERATION
AND CORROSION IN SEWERS
FACTOR
EFFECT
Wastewater Characteristics
Dissolved oxygen
Biochemical oxygen demand
(organic strength)
Temperature
pH
Presence of sulfur compounds
Sewer System Characteristics
Slope and velocity
Turbulence
Surcharging
Presence of force mains and inverted
siphons
Sewer pipe materials
Concrete alkalinity
Accumulated grit and debris
Low DO favors proliferation of anaerobic
bacteria and subsequent sulfide
generation
High soluble BOD encourages microbial
growth and DO depletion
High temperatures increase microbial
growth rate and lowers DO solubility
Low pH favors shift to dissolved H2S gas
Sulfur compounds required for sulfide
generation
Affects degree of reaeration, solids
deposition, HZS release, thickness of slime
layer
Same effect as slope/velocity
Reduces oxygen transfer and promotes
sulfide generation, will not corrode while
surcharged
Same effect as surcharging, releases H2S
at the turbulent discharge end
Corrosion resistance of pipe materials
varies widely
Higher alkalinity reduces corrosion rate
Slows wastewater flow, traps organic
solids
1-9
-------
REFERENCES
1. Water Quality Act of 1987, Public Law 100-4-Feb.4, 1987, Sec. 522. Sulfide
Corrosion Study.
2. Stahl, J.S., Redner, J., and R. Caballero, "Sulfide Corrosion in the Sewer System
of Los Angeles County," presented at 11th U.S./Japan Conference on Sewage
Treatment Technology, the Public Works Research Institute, Tsukuba, Japan,
October, 1987; and ASCE Conference on Sulfide Control in Wastewater
Collection and Treatment Systems, Tucson, AZ, February, 1989.
3. Pomeroy, R.D., Parkhurst, J.D., Livingston, J., and H.H. Bailey, "Sulfide
Occurrence and Control in Sewage Collection Systems," EPA 600/x-85-052,
Cincinnati, OH, 1973.
4. "Process Design Manual for Sulfide Control in Sanitary Sewerage Systems,"
USEPA, Cincinnati, OH, 1974.
5. "Odor and Corrosion Control in Sanitary Sewerage Systems and Treatment
Plants," EPA/625/1-85/018 USEPA, Cincinnati, OH 1985.
6. 'Sulfide and Corrosion Prediction and Control," American Concrete Pipe
Association, Vienna, VA, 1984
7. "Sulfide in Wastewater Collection and Treatment Systems, " ASCE Manual
No.69, ASCE, New York, NY, 1989.
8. Thistletbwayte, D.K.B., "The Control of Sulphides in Sewerage Systems," Ann
Arbor Science, Ann Arbor, MI, 1972.
1-10
-------
2.0 NATIONAL ASSESSMENT OF CORROSION
2.1 Introduction
In the spring and again in the fall of 1987, EPA met with representatives of the
County Sanitation Districts of Los Angeles County (CSDLAC) and the City of Los
Angeles to learn about the nature and severity of the corrosion problem in the
wastewater collection system. CSDLAC presented slides and video recordings that
documented the degree of corrosion and its increase over time. In addition, a field
inspection of a severely corroded sewer line was conducted. It was concluded from the
visits that a national assessment of corrosion was warranted to document the severity of
corrosion problems in other cities and to determine if other municipalities had
experienced increases in corrosion rate upon implementation of industrial pretreatment
standards.
The national assessment of corrosion consisted of the following major elements:
• site visits in early 1988 to six cities with documented corrosion problems to
determine if severe hydrogen sulfide corrosion problems were unique to
CSDLAC.
• site visits in 1989 and 1990 to three cities with hydrogen sulfide corrosion
problems at wastewater treatment plants and lift stations to document the
extent of problems and prepare case histories.
• site visits in 1988 to three cities with pretreatment programs to assess the
effects of pretreatment
• compilation of detailed case histories in 1988 for several wastewater
collection systems to document the history of corrosion in those systems.
• collection of samples from two wastewater collection systems in 1988 for
physical, chemical, and microbiological analyses to gain a better
understanding of the mechanisms of corrosion.
• telephone discussions with officials of various cities reported to have
corrosion problems.
• contacts with engineers, sewer rehabilitation contractors, and
manufacturers of materials used in sewer rehabilitation or replacement
• evaluation of information collected in surveys conducted by CSDLAC,
municipal associations, and pollution control organizations.
2-1
-------
2.2 Site Visits to Assess Hydrogen Sulfide Corrosion in Sewers
One of the first efforts of the project was to determine whether severe sulfide-
induced corrosion was unique to the CSDLAC sewer system. To answer this question,
site visits were made to six cities with reported corrosion problems. To identify
potential cities for site visits, information was reviewed from several sources, including
surveys conducted by the Association of Metropolitan Sewerage Authorities (AMSA) in
1987, CSDLAC in 1984, and the Water Pollution Control Federation (WPCF) in 1984.
In addition, regional EPA offices and literature articles provided supplemental
information. Based on this information, a list of 131 candidate cities was compiled. Of
these 131 cities, 66 were reported to have problems with sewer corrosion. Further
review resulted in preliminary selection of 34 cities expected to yield the most valuable
corrosion data during site visits.
Six sewer systems were selected for site visits from the list of 34: Albuquerque,
New Mexico; Baton Rouge, Louisiana; Boise, Idaho; Casper, Wyoming; Forth Worth,
Texas; and Seattle, Washington. Each city had reported severe corrosion problems.
In addition, two other cities, Charlotte, North Carolina and Milwaukee,
Wisconsin, were selected for further study. These two differed from the others in that
they did not have known corrosion problems, and they had certain sewer segments that
carried a large proportion of industrial flow, including pretreated metal finishing wastes,
while other sewer segments carried primarily residential wastewater. It was postulated
that this may allow observation of whether the industrial sewers had been "protected" by
the presence of metals or other industrial waste constituents compared to those sewer
segments conveying strictly residential waste. All eight cities had industrial pretreatment
programs in effect
The primary purpose of the site visit program was to determine if severe and
high-rate corrosion is unique to the CSDLAC wastewater collection and treatment
system. Each site visit typically included one and one-half to two days in the visited city.
On the afternoon of the first day, the field team met with representatives of the city or
agency to review the characteristics of the local collection and treatment system and to
select six to 10 locations for observation the following day.
Field observations and measurements included monitoring gaseous hydrogen
sulfide in manhole and sewer atmospheres (Industrial Scientific Devices, Model HS
267), measuring sewage pH with a portable pH meter (Nester Instruments, Model
34100-403), measuring total sulfide in sewage with a portable test kit (HACH Co., Kit
No. HS-6), and using a screwdriver to probe manhole and sewer walls to evaluate
depth of corrosion and integrity of concrete. Photographs were taken and pH paper
(Color pHast, 0-14) was used to measure surface pH at several locations on sewer and
manhole walls. Additional observations were made of smoothness of sewage flow,
sewage velocity, and the presence of sewer laterals, bends, and drops. Background
2-2
-------
information (e.g., pipe age, slope, diameter, approximate sewage age, and the presence
of coatings or linings) was also recorded when available.
To distinguish levels of severity and rates of corrosion, the following arbitrary
definitions were developed for this study:
Severe corrosion - loss of one inch or more of concrete, loose or missing
aggregate, exposed reinforcing steel.
High-rate coirosion - rate of corrosion which would cause a loss of at least one
inch of concrete in twenty years. This rate is significant since reinforcing steel is
generally about one inch below the interior concrete surface of large pipes
constructed according to industry standards. Exposure of reinforcing steel to
conosion can lead to structural impairment
Accelerated corrosion - an increase in the rate of corrosion with time.
Corrosion was observed in all eight cities, and was considered severe in at least
one location in each city except Charlotte. High-rate corrosion was observed in at least
one location in each city except Charlotte and Milwaukee. During EPA's site
investigations, the estimated depth of corrosion was divided by the age of the pipe to
yield a lifetime average corrosion rate. However, it is impossible to determine from
these data whether the corrosion rate has changed with time. Such inspections merely
offer a "snapshot" of the corrosion processes and provide no information on the history
of corrosion, i.e., whether accelerated corrosion had occurred. A summary of conditions
in each city is presented in the following paragraphs. A summary of pertinent
information collected during the site visits is shown in Table 2-1.
2.2.1 Albuquerque, New Mexico
The City of Albuquerque maintains approximately 1,400 miles of sewer which
serve approximately 450,000 people and transport an average of 49 million gallons per
day (mgd) of wastewater to the city's treatment facility. Separate storm sewers are used
throughout most of the city, but some combined systems do exist
Albuquerque experiences 90 to 100 collapses per year that are attributed to
hydrogen sulfide corrosion in its approximately 400 miles of 8-inch-diameter concrete
pipe. These collapses are mostly in residential areas, and each typically involves two to
four pipe sections (20 feet). The problem of pipe collapse is widespread in the city, but
seems concentrated in North Valley, an older part of town that has the most concrete
pipe, and in pipe 40 to 60 years old. The rest of the collectors are mostly clay pipe.
Corrosion seems to be worst at locations where a force main discharges to a
manhole, at lift stations in gravity sewers (the city is beginning to use polyvinyl chloride
2-3
-------
TABLK 1-1
Reproduced Irom
best available copy.
City and Stale
Sewer
*»«
(ycara)
Seer
Diaaetcr
(intbea)
Albuquerque, HII
Arno
Marquette Ave. anil Editb St.
Iron Ave. and 14th St.
Coora Blvd. and Churchill Rd.
Atrlrco Dr.
Roaaaioor Rd.
25
1
32
SELECTED INFORMATION FROM SUE VISITS
Surlace pll
U|> 111 c a ¦ _ DowuH rea» Hanhole
Croun Vail Ciowo Vail Wall Roof
(a.u.) (a.u) (a.u.) (a.u.) (a u.) (a.u.)
12
54
60
30 to, 36 out
54 rvc
JB
2.5
4
2
2.J
5.0
2.0
1.0
4.0
3.0
1.0
2.}
Eat iaated
Concrete
Lota
(iuchea)
Eatiaated
Corroaloa
Rate
(lochea
/*'»*) _
(nil)
0 0.25
0.04
0.14
<0.01
Statue
Severe
Severe
Severe
Severe
Negligible
Baton Rouge, LA
PS. No. 59
27
NA
NA
NA
NA
NA
6.0
6.0
0.25-0.50
0.01-0.02
Shallow
front St. and North St.
230
30
4.5
6.0
4.0
6.0
4 5
--
0.25-0.50
<0.02
Shallow
Uevall St. and Blount St.
230
20
in, 42 out
2 5
2.5
5.0
5.5
2 0
--
1.0-1.5
<0.05
Severe
Harding St. and Georgia St.
230
30
in, 36 out
6.0
6 0
5.5
6.0
6.0
0.25-0.50
<0.02
Shallow
Wtnbourne Ave. and
230
36
50
5.0
6.0
--
6.0
--
1.0-1.5
<0.05
E. Brookalone l)r .
E. Contour and S. Contour
230
54
3 5
--
2.5
--
--
4.0
0.25-0 50
<0.02
Shallow
Starring Lane
230
60
2.5
--
5.0
--
3.0
IS
0.5-1.0
<0.03
Seveie
>i»e, ID
Prolcat and Federal
14
21
--
--
--
--
--
--
1.0-2.0
0.14-0.28
Severe
Wars Spring* and Els
>30
10 VC
4.0
--
3.0
2.0
--
1.0-2.0 (tlH)
<0.06
Severe
Vara Springs and N- Straugliaa
>30
io rvc
4.0
--
3 0
--
5.0
--
0 (Mil)
Brick Mil
NA
Rrurr and Jef(eraon/Bannock
- -
—
6.5
--
6 5
--
--
--
NA
Lined HII
NA
Horlli Gary and Baron
12
--
6.0
--
6.0
--
1.0
--
0.5-1.0
0.04-0.08
Severe
Bluebird and Gary
12
--
3.0
--
3 0
--
1.0
--
0.5-1 0
0.04-0.08
Severe
Gtenvood and State
12
--
--
--
--
--
3.0
—
--
--
Severe
Glenwood and CliinJoo
12
--
--
--
--
--
--
--
1.0-1.5
0.05-0.1
Severe
-------
TABLE 2~1 (coalluued)
i
tn
Surlice pH
Eti iiileif'
[¦lUiltd
Cor rot loo
Sewer
Srurr
Djwuilf**¦
htoJioll
CoiickIc
Bile
Age
Dlwtlti
Crouo
Will
Crown
IMI
Utll
Root
lug*
(incbci
Corroaton
City and Slate
<*«•««)
(lucbe*)
•:» )
(« » > (
;^L_
uc be* ^
/*«•«)
Statua
Casper, WY
K Street
29
36
J0
SO
4.0
S.O
3.0
--
I.O-I.S
0.04-0.01
Severe
C and Centcr fit.
1
48
--
--
--
--
2.0
--
Shallow
101 u. Vclluwatoue
1
48
4.0
SO
3.0
--
2.0
--
0 0.25
0.010.0*
Shallow
Falr|roundi Read
1
36
SO
SO
s.o
SO
6.0
--
0 0.25
0.01-0.04
filial low
Hidway Di1vc
1
10
to
6.0
6.0
6.0
1.0
--
0 0 21
0.01-0.04
SbilUw
hills Cut-Off Road
1
30
10
SO
SO
S.O
S.O
..
0-0.12
0.01-0.02
fihallow
Alter aud Oafiod11
1
21 lo, 24 out
6.0
--
6.0
--
6.0
--
0-0.12
0.01-0.02
Shallow
Charlotte, NC
Clautou fit. at 1. Creek
—
36
6.0
6.1
6.0
6.0
6.0
--
--
--
Abaent
Rcaouot fit.
--
36
6.0
6 0
6.0
6.0
6.0
--
0-0.12
Shallow
Frccdoa Dr. and Thrift ltd.
--
36
6.0
6 0
6.0
6.0
6.0
--
--
—
Abacut
E. I2tli Si. aud Hcycia
--
30
6.0
6 0
6.0
6.0
6.0
--
--
—
Abacot
Davidson and E. 22ud
--
30
6.0
6.0
6.0
6.0
6.0
--
--
Could
eot obaer
Old Providence Rd. near
Sliaroiivlcw
--
42
SO
SO
4.S
SO
6 0
--
0-0.25
__
filial low
Afborw»y near Sedley Road
--
24
6 0
6.0
6.0
6 0
6.0
--
--
Abacot
Paik Road and lloncure
--
S4
6 0
6 0
6.0
6.0
6.0
--
--
--
Abacot
Old Natluna Foid near Etvln Laiir
--
14
6 0
6 0
6.0
6 0
6.0
--
--
Abaent
GranJte Si. near Cunliucnlal Blvd
. --
21
6 0
6.0
6.0
6.0
6.0
--
0-0.12
--
Slial low
I'ort Vorlh, TM
Roiedile St.
>30
36
6 0
6.0
6.0
6.0
--
--
2.0-30
-------
Cily and Slate
6cwcc
Age
iJIEiLil
Sewer
Dlaaelcr
(inchcaj
Dlluiukct, Wl
H20 2»30 In, 96 out
IS la, 1ft aul
14 S4
15
I
TABIJS J 1 (continued)
Up»t m»
Clown Utll
li:S: L.
Surlace {U
Hioliiilt
Crown Vail Wall lool
iiJiJ
Elllaaltd
Concrete
Loaa
(lucbci)
iallaaltd
Corrnaioo
Hall
(inchel
ItmX—
Slalui
iS 6 5
6.S 6.1
6 5 6.1
6. S
6 1
6 1
6.S
6 4 6 1
6.S 6.S
6.S 6 1
6 1
6. S
6.1
6.0
6.S
6S
6 1
1 S
6 5
6S
6 0
6.S
0.02
Abaenl
Abllal
Abaenl
Abaenl
Abieal
Cuuld not obidve
Could oal uLitrvt
Abaenl
Severe
Aliienl
10 3.0 2.0 2.0 2.0 1.0 I 0.04 6e»rrc
2 1 1.0 IS <0.0» Severe
1.0 -- -- -- 6.0 1.0 O.2S-0.S0 — Shallow
3 0 — 0 0.25 0.01-0.02 Shallow
0.21 0 50 0.01-0.04 Shallow (HII)
-------
[PVC] liners at those locations), near interceptors where hydrogen sulfide moves back
into laterals, and at locations midway between manholes. Albuquerque does not have
much industrial discharge, and pipe failures are not related to the presence of industrial
discharges. There are only about three small electroplaters in the area.
In the past, Albuquerque has not had a formal program to identify corrosion.
The city now has a television inspection program for small-diameter sewers (8 inches).
The city replaces about 18,000 feet per year of 8- to 10-inch pipe. The goal is to
replace 30,000 feet per year. The city has replaced up to 12-inch concrete lines with
clay or PVC to prevent further corrosion. A total of about 40 miles of mostly 8-inch-
diameter pipe has been sliplined since 1978.
Albuquerque experiences a summertime odor problem, and injects chlorine gas
and hydrogen peroxide at several locations for odor control during the summer.
Untreated wastewater has had total sulfide concentration of up to 4.3 mg/1. The city
will be switching some of the chlorine units to hydrogen peroxide in the future, because
of longer lasting effects and safety concerns.
Sewers 24 inches or less in diameter are cleaned at intervals ranging from three
months to two years. Larger sewers are not cleaned.
Corrosion at the wastewater treatment facility is limited primarily to metal
components. Ventilation is used to help control corrosion inside the treatment
buildings.
The city identified six sites scattered throughout the area to exemplify the
corrosion problem in Albuquerque: Arno and Wesmeco streets; Marquette Avenue and
Edith Street; Iron and 14th streets; Coors Boulevard and Churchill Road; Atrisco
Street; and Rossmoor Street
An example of high-rate corrosion exists at the Atrisco Street site. The site is
seven years old and near the upstream end of the system. The slope in this reach of
pipe is very flat, and sewage velocity was estimated to be 0.5 feet per second (fps). The
manhole at this site was installed with a bituminous coating that has separated almost
entirely from the concrete. In seven years, the concrete on the walls and soffit of the
manhole has conoded up to an estimated depth of 1.0 inch. The inlet and outlet pipes
at this manhole are PVC-lined and in good condition. Measurements of pH on the
manhole and pipe walls ranged from 1 to 5.
Three other sites (i.e., Arno and Wesmeco streets. Iron and 14th streets, and
Coors and Churchill streets) have experienced severe corrosion. Measurements of pH
on the walls of manholes and pipe ranged from 2 to 5 at Arno Street, and from 1 to 2
at Coors Street, and were 4 at Iron Street Depth of friable concrete or corrosion
product ranged from 0.50 to 2 inches. However, the pipe and manholes at these sites
2-7
-------
are considerably older than the Atrisco Street site, reflecting a lower rate of corrosion
over the life of the installation. The current rate of corrosion at these sites cannot be
determined from available information.
The manhole at Marquette Avenue and Edith Street has experienced some cor-
rosion; however, manhole access problems prevented quantification. The flow in this
manhole is turbulent The Rossmoor Street site is not corroded badly, although pH
ranged from 2J to 4 at this site.
Except for the Marquette Avenue and Edith Street site, release of hydrogen
sulfide gas is not believed to be accelerated by turbulence or drops at the Albuquerque
sites. Long detention times, flat slopes, and warm sewage temperatures are thought to
promote hydrogen sulfide corrosion of concrete system-wide in Albuquerque, as
reflected by low pH readings at all sites.
2.2.2 Baton Rouge, Louisiana
The City of Baton Rouge maintains approximately 250 miles of sewer which
transport an average of 36 mgd of wastewater to the city's three treatment facilities.
The sewer system serves the entire East Baton Rouge Parish except for two small
communities. The system serves 375,000 people. Baton Rouge officials estimate that
they have approximately 75 miles of unlined reinforced concrete pipe larger than 24
inches in diameter.
Industry contributes less than 5 percent of the total sewered flow. The major
industries, including a large oil refinery, treat their own waste and do not discharge
industrial effluent to the sewers. Those industries that do discharge to the Baton Rouge
system are generally in compliance with the established pretreatment program. Industry
is not concentrated in any one area of the system, and city engineers do not correlate
corrosion in their system with industrial discharge.
The Baton Rouge sewer system is completely separate. Corrosion of concrete
pipe is system-wide. Baton Rouge experienced its first sulfide-related pipe collapse
about five years ago. This collapse was the first indication to the city of the severity of
its corrosion problem. A consultant's report to the city on preventative maintenance of
the system made reference to odor control, but did not focus on corrosion. The city did
try chlorine addition in the mid-1970s, but abandoned the program in less than one year
because of high costs. The city does some television inspection of the system, but does
not have a system-wide hydrogen sulfide corrosion prevention program.
The city has experienced multiple problems in some pipe reaches. Repairs made
with fiberglass or plastic pipe appear to be holding up well; however, one repair done
with concrete pipe experienced corrosion and needed subsequent replacement Baton
Rouge acknowledges that turbulent flow conditions due to changes in grade or
2-8
-------
direction, pump station discharges, or drop connections are usually prevalent at problem
areas.
Baton Rouge selected eight sites for EPA to observe: a pump station, the
Central Treatment Plant, and six manholes located throughout the sewer system (two in
the north subsystem, one in the central subsystem, and three in the south subsystem).
The sites ranged in location from one within approximately 1 mile of a treatment plant
and 10 miles from the upstream end of a reach, to one located near the upstream end
of a reach. Corrosion was observed at each site, with varying degrees of severity. All
sites visited were constructed in the early 1960s. Pipe slopes ranged from 0.003 ft/ft to
0.00015 ft/ft
Pump Station No. 59, a 27-year-old structure, which is located 1 mile upstream
from the Central Treatment Plant and collects wastewater from about 10 miles
upstream, was the first site visited. A pH of 6 was measured on the wet well walls;
shallow corrosion, 0.25 to 0.5 inches deep, was observed. The wet well often surcharges,
washing the walls.
The Central Treatment Plant had shallow corrosion of some concrete structures.
The force main discharge structure at the plant headworks was corroded, and aggregate
was exposed in both the primary clarifier influent and effluent channels. Some
corrosion of metal had also occurred at the plant headworks. A 0.6 parts per million
(ppm) total sulfide content was measured in wastewater at the plant headworks. Plant
influent pH was 6.
The first manhole visited is located at Front and North streets in the Central
District, about 0.75 miles downstream of a pump station and within 1 mile of the
Central Treatment Plant Measurements of pH in the manhole and the 30-inch-
diameter pipe ranged from 4 to 6. Large aggregate, indicating up to 0.5 inches of pipe
loss, was visible in the pipes above normal water line. Flow at this location was
turbulent due to the pump station upstream, a change in slope about 100 feet
downstream, and a 12-inch-diameter inlet with to 2- to 3-foot drop. The wastewater
had a trace of sulfide and pH of 6.
The next two sites are in the North District: Devall Lane off Blount Road, and
Georgia Street at Harding Boulevard. The Devall Lane site is directly downstream of a
pump station, and flow is made more turbulent by a 1.5-foot drop across the manhole.
The site is located at the midpoint of a 12-mile-long drainage area. A total sulfide
concentration of 0.05 mg/1 and a pH of 6 were measured in the wastewater. Pipe
surface pH measurements ranged from 2 to 5.5. The pipe at this location was severely
corroded above the normal water surface (during pump discharge). Some mortar is
missing between the bricks in the manhole and some bricks were observed on the floor
of the downstream pipe. Observations revealed that as much as 1.5 inches of concrete
may be corroded.
2-9
-------
The Georgia Street site is in the upstream third of the same drainage area. The
wastewater had a sulfide concentration of 0.08 mg/1 and pH of 6.5. Pipe pH
measurements ranged from 5.5 to 6. Although this site is also less than 0.25 miles
downstream of a pump station, pipe corrosion was estimated to be minor. Only small
aggregate was exposed, indicating 0.25 to 0 JO inches of concrete loss. Two drop pipes
enter this manhole, and the pipe changes direction about 100 feet downstream.
The final three sites visited are in the South District Winbourne Street at East
Brookstown Drive, East Contour Drive, and Staring Lane. The Winbourne Street site
had a wastewater pH of 7 and sulfide of 0.17 ppm. Pipe surface pH measurements
ranged from 5 to 6. The 36-inch pipe at this location is corroded severely and
corrugations were visible at reinforcing steel locations. There is a 2-foot drop across the
manhole. Winbourne Street is located near the upstream end of a 15-mile-long
drainage reach.
The Contour Drive and Staring Lane sites are on the same 54-inch pipe in the
middle and near the downstream end of the reach, respectively. Wastewater sulfide
content was 1.1 ppm at the Contour Drive site; wastewater pH averaged 6 at the two
sites. Pipe surface pH measurements were between 2.5 and 4 at the Contour site, and
2.5 and 5 at Staring Lane. Corrosion at these sites was limited to about 0.50 to 1 inch
of concrete loss, exposing only the first layer of aggregate.
Hydrogen sulfide gas levels of 3 to 4 ppm were measured at the Contour Drive
site. The Staring Lane site is just downstream from a 36-inch-diameter force main
terminus. In addition, the downstream pipe at the Staring Lane site had a broken invert
near the manhole, which has created a backwater condition and turbulence at the
manhole.
2.2_3 Boise, Idaho
The City of Boise maintains approximately 325 miles of sewer which transport an
average of 24 mgd of wastewater to the city's three treatment facilities. Boise provides
sewer service to three sewer districts and to Garden City. Boise has recognized a
hydrogen sulfide corrosion problem in its system since 1983. Concrete sewers and
manholes in at least four areas have experienced severe corrosion. Some of their most
seriously damaged manholes have been coated recently with materials to resist further
sulfide attack.
Hydrogen sulfide corrosion in Boise is system-wide. Boise officials feel their cor-
rosion problem can be correlated to low flows in hydraulically oversized sewers and to
turbulent flows created by force main discharges and drops in manholes. There is very
little industry in the area, and Boise operates a completely separate sewer system.
Boise, in consultation with CSDLAC, has tried Polymorphic resin and Zebron
2-10
-------
coatings and Chiystallok and fiberglass liners in several of their manholes. Insituform
and sliplining have been and are presently being used in Boise to rehabilitate corroded
sewers.
After Boise discovered its problem in~1983, it realized that the 1977 television
monitoring tapes indicated previously overlooked signs of corrosion such as concrete
swelling and spalling. During the visit, Boise displayed over 12 samples of 4-inch-
diameter cores, recovered from a 1984 coring program, which showed the extent of
corrosion in different pipe sizes, ages, and areas of the system.
Based on measurements of core thickness and the known age of the pipe, Boise
has-calculated that lifetime corrosion rates are as high as 0.12 inches per year in the
sewer pipe at Glenwood and Chinden streets, and 0.15 inches per year in the sewer pipe
at Canal and Columbus streets. Corrosion rates calculated similarly for pipe in the
warm springs area was 0.03 inches per year over a 37-year period, and 0.06 inches per
year at Protest and Federal streets.
About 30 homes in the Warm Springs area use a geothermal water source for
home heating. The water is extracted from the ground at about 175°F and discharged
from homes to the sewer at about 130°F. The sewage in this area of town averages
between 90 and 100°F. The sulfate concentration of this water source is about 23 mg/1.
Corroded manholes were observed in this area.
The maintenance supervisor from the neighboring West Boise Sewer District
(West Boise) described a serious problem in his system. West Boise replaced six
manholes after a 5-year-old sewer collapsed due to hydrogen sulfide corrosion in 1983.
There were 10-foot-drop laterals at some of these spun concrete, Type 2 concrete
manholes. West Boise feels that hydrogen sulfide conditions are worse at turbulent flow
areas (e.g., drop manholes). In addition, the supervisor cited uneven slope during
installation of the system as contributing towards solids deposition in the lines.
West Boise previously used chlorine and hydrogen peroxide dosing and
experimented unsuccessfully with bacterial seeding to control sulfide generation. The
chemical treatment program was successful once the proper dosing was defined, but very
expensive. The West Boise maintenance supervisor also feels that hydrogen sulfide
conditions are worse at turbulent flow areas resulting from drop manholes. He noted
that Garden City, a nearby area with high infiltration and inflow, has little conosion.
The West Boise Sewage Treatment Plant, owned and operated by Boise City, has
had air scrubbing equipment installed to reduce odor emissions.
A tour of Boise's treatment plant revealed some concrete corrosion. Influent
channels covered for three years at the plant's headworks have experienced corrosion,
particularly the channel that formerly carried sludge. The covered wet wells had no
2-11
-------
corrosion, but the air has been scrubbed since the 1970s to reduce odor complaints.
The remainder of the process tanks at the plant axe not covered (except for the
anaerobic digesters), and are not experiencing any concrete conosion.
The field team members made observations at 12 sites in Boise. Of these, five
sites (i.e., one at Protest Avenue and Federal Way, and four along a segment of another
sewer between North Gary Street at West Baxon Street and Glen wood Street at
Chinden Boulevard) showed a high rate of corrosion. The remaining sites, although
they often have acidic pH levels on walls, do not yet show evidence of corrosion.
The Protest Avenue site is located only 2 miles from the upstream end of the
collection area and has a 10-foot-drop inlet The long drop creates turbulence that is
believed to accelerate release of hydrogen sulfide and corrosion. A screwdriver could
be pushed up to 2 inches into the remaining concrete of the manhole wall.
Measurements of pH were 2 on the manhole wall. This site is 14 years old.
Four sites along a single 12-year-old line between North Gary Street at West
Baron Street and Glenwood Street at Chinden Boulevard also have high-rate corrosion.
Pipe at the downstream end (Glenwood Street at Chinden Boulevard) showed deep
corrugations at reinforcing steel, indicating that corrosion had penetrated deeper than
the reinforcement. Surface pH levels were 6 in the pipe and 1 in the manhole at North
Gray Street, 3 in the pipe and 1 in the manhole at Bluebird, and 3 in the manhole at
State Street The wastewater sulfide concentration was 2.25 mg/1 near the upstream
end.
A brick manhole and a previously corroded concrete manhole coated with Poly-
morphic resin were inspected in the Warm Springs area of Boise. The surface of the
brick manhole had a pH of 5, and the coated concrete manhole had a pH of 2.0 - 3.0.
Both the brick manhole and the resin coating appeared to be in good condition.
Additional observations at one unlined and two lined manholes did not reveal
corrosion. Shallow corrosion, zero to 0.50 inches deep, was observed at a pump station
wet well.
2.2.4 Casper, Wyoming
Casper officials feel that a severe hydrogen sulfide corrosion problem exists in
that city. The problem first came to light in 1975 during reconstruction of the
wastewater treatment facility when a severely corroded influent line to the primary
clarifier needed replacement Since that time, the city has begun looking for corrosion
in manholes as part of its manhole inspection program. In addition, the city tried
sodium hydroxide dosing once in 1986 and once in 1987 to control the slime layer inside
sewer pipes and has added clean water to upstream portions of the system to increase
flow rates and decrease detention times. The sodium hydroxide treatments were
2-12
-------
effective for approximately three-week periods. Generation of hydrogen sulfide is only
a problem during summer months. Casper also has a problem with hydrogen sulfide
corrosion of the engines being fueled with digester gas in its cogeneration plant A
$16,000 rebuild was recently completed. City staff reported that this digester gas
cogeneration problem is shared with Billings, Montana, and Boulder, Colorado.
Casper officials identified seven manholes for the visit The first observation was
in a manhole on a 29-year-old 36-inch sewer line about 0.75 miles from the wastewater
treatment facility. The remaining observations were along a 10-mile segment of a 6- to
7-year-old sewer that transports wastewater from the western side of Casper to the
wastewater treatment facility.
Corrosion is clearly evident in the 29-year-old manhole. Aggregate is exposed
and loose in some instances. Up to 1.5 inches of pipe wall may have washed away.
Corrosion product was not observed at this location; however, a pH of 3 was measured
on the manhole wall and a pH of 4 to 5 was measured in the crown of the downstream
pipe. Corrosion is evident at all the manholes observed on the 6- to 7-year-old sewer.
Furthermore, corrosive conditions appear to worsen the farther downstream that
observations were made. The farthest upstream observation was at a manhole located
about 200 feet below a force main river crossing. The sewer pipe appeared in very good
condition, except for 0.125 inch of erosion evident along the side of the outlet Pipe
and manhole surface pH was 6 at this location; there was no corrosion product
As the observers progressed to downstream locations, the presence of corrosion
product increased and pH levels on pipe and manhole surfaces decreased.
Measurements at three downstream locations showed pH levels of 2 or less. At the
farthest downstream location, Center and G St, approximately 1.5 inches of soft mushy
corrosion product was evident on the walls of the manhole. Because of the short length
of time that this sewer segment bad been installed, it was difficult to estimate the
amount of concrete that had corroded. However, corrosion was clearly occurring.
The effluent channel of the primary clarifiers at the wastewater treatment facility
at Casper had severe conosion. Up to 2 inches of concrete may be missing from parts
of the channel. The facility superintendent believes that a major contributing factor to
sulfide generation in that city is excessive sewage detention time. This results from
hydraulically oversized sewers constructed in anticipation of growth that did not occur
because of a regional economic downturn. In addition, high sulfate concentrations in
the local drinking water, 180 to 200 mg/1, may aggravate the problem.
2.2.5 Fort Worth, Texas
The City of Fort Worth maintains approximately 2,000 miles of sewer which
transport wastewater to a single treatment facility located adjacent to Village Creek, a
tributary of the Trinity River. A second facility, the Riverside facility, used to treat
2-13
-------
wastewater for the city, however, flow to that facility was diverted to the Village Creek
facility several years ago.
Fort Worth experiences hydrogen sulfide odor problems during warm weather
and has had a pipe collapse that is attributed to hydrogen sulfide corrosion. In particu-
lar, one of the Village Creek collectors collapsed. The city now injects chlorine into the
two main interceptors (90 and 96 inches) to control sulfide and odor. The closing of
the Riverside Treatment Facility and concomitant shifting of flow to Village Creek have
decreased detention time and hydrogen sulfide levels in these two interceptors.
The industrial contribution of wastewater is a fairly uniform 10 to 20 percent
throughout the collection system. The major sources are from electroplating, brewing,
food processing, and aircraft manufacturing.
The levels of metals in the wastewater have declined dramatically during the past
five years. However, levels of aluminum and iron are high because of the discharge of
drinking water treatment sludge to the wastewater collection system at several locations.
A pipe collapse was reported to have occurred at the end of a force main in the
neighboring City of Grand Prairie. The City of Pantego, also a neighbor, was said to
have a major problem.
Field team members entered five manholes in Fort Worth to assess the presence
and effects of corrosion in the city sewer system. The manholes are spread out across
the city and represent several sewer main subsystems. Two manholes manifested severe
corrosion. At Rosedale Street, a section from the crown of a 36-inch pipe is clearly
visible lying on the pipe floor. The pipe walls have corrugations 1.5 to 2 inches deep;
an estimated 2 to 3 inches of pipe is missing. The city is aware of problems in this 30-
plus-year-old line and has rerouted wastewater to allow replacement of this sewer. This
sewer has a steep, easily observable slope that increases sewage velocity and could ac-
celerate the release of hydrogen sulfide gas. The pH of the pipe surface at this
manhole was approximately 6, indicating that conditions were not as corrosive at the
time of the visit as in the past, probably because of the rerouting of the wastewater.
The second location with severe corrosion was a 65-year-old, 54-inch pipe on
Bomar Street At this location, the pipe upstream and downstream of the manhole had
corrugations 1 to 2 inches deep. In addition, a section of pipe wall approximately 1 foot
high by 6 feet long is missing from the right side of the pipe approximately 15 feet
downstream. An estimated 2 inches of concrete has eroded from the lower portion of
the manhole, and the joint between the manhole and outlet pipe has deteriorated. Two
15-inch laterals enter this manhole, but do not appear to be very active. The upstream
manhole has an active drop lateral, and flow in the downstream manhole is very
turbulent In both instances, these factors could have contributed to release of
hydrogen sulfide gas and an increased corrosion rate. The pH of the pipe surface at
2-14
-------
this manhole was approximately 6. There is no apparent indication of current or
ongoing coiTosion.
The other three manholes observed in Fort Worth are approximately 30 years
old. Even though these locations had lower pH levels of 4 to 5, corrosion is not as
severe as at the other two locations. Drop laterals were not observed at or near these
three locations.
2.2.6 Seattle, Washington
The Municipality of Metropolitan Seattle (Metro) maintains approximately 247
miles of sewer which transport 186 mgd of wastewater to Metro's treatment facilities.
Metro has had a hydrogen sulfide odor problem for many years. A large number of its
concrete sewers and sewage structures have experienced extensive corrosion damage.
The most serious identified cases of hydrogen sulfide coiTosion have been replaced or
repaired by coatings or liners.
Construction of Metro's interceptor facilities began in 1963; corrosion is
widespread in this relatively new system. Local municipalities provide smaller-diameter
sewage collection systems which were not investigated during this study. The Seattle
area is heavily industrialized, and industrial flow represents about 25 percent of the total
flow; however, industrial discharges have not been correlated with sulfide generation or
concrete coiTosion. Areas served by Metro to the east and north of Lake Washington
have separate sewer systems for stormwater transport Areas to the west of Lake
Washington are served predominantly by a combined sanitary-stormwater sewer system.
Metro has an extensive sulfide monitoring program, and has had full-time staff
working on the problem since early 1987. Metro personnel look for hydrogen sulfide
damage as part of sewer inspections during which headspace hydrogen sulfide
concentrations and pipe surface pH levels are also measured. Hydrogen sulfide
concentrations from 0.1 to over 50 pptn have been found along with pH readings as low
as 2. Metro's records indicate lower pH readings occur at sites with higher hydrogen
sulfide gas concentrations.
Metro has tried various concrete liners and coatings in pipes and on structures to
control con-osion as well as chemical addition to control sulfide. Sliplining, epoxy,
polyethylene (PE), PVC, UPC (a polyurethane polyethylene copolymer), Ameron lining,
polyurethane (Sancon), C.T.E. coating, and Aquatapoxy all are being or have been
tested by Metro since 1974. Both satisfactory and unsatisfactory performances have
been observed. For example, the PE liner on the East Bay Interceptor - Section 8 is in
good shape and is protecting the concrete behind it, but the UPC coating on the Lake
Sammanish Interceptor failed and is peeling off. Hydrogen peroxide addition to control
sulfide was tried but abandoned for monetary reasons. However, Metro did find that
once a large shock dose of peroxide was added, subsequent dosages could be reduced to
2-15
-------
control sulfide.
Metro has been involved in other activities related to hydrogen sulfide corrosion
control. Power cleaning of sewers, use of sacrificial concrete in its sewers, and sonar,
radar and ultrasonic measurement of pipe wall thickness have been tried. Metro has
also tried to monitor corrosion rate with concrete coupons and copper shavings hanging
in pipes; but found reactions too slow to provide useful data. In addition, the Renton
Treatment Plant has a $5,000,000 odor-control system employing scrubbers, activated
carbon, impregnated carbon, and chlorine addition. A facilities plan study by a
consultant included sulfide-control recommendations. Concrete corrosion at Metro's
treatment plants is not a problem.
Seattle Metro personnel recommended five sites for observation. The sites are
widely distributed throughout the system and in parts of different subsystems. Three
sites are directly downstream of force main discharges: a manhole at East Marginal
Way and South 112th Street, downstream of the Renton sludge force main; a manhole
near 15th Avenue W and W Raye streets, downstream of the Interbay Pump Station
force main; and the Hollywood Pump Station discharge structure. One of the remaining
sites, a manhole at 15th Avenue NW and 188th Street NW, is a few blocks downstream
from a force main. The fifth site, a manhole on the Lake Sammanish Interceptor at NE
Union Hill and Avondale roads, is not downstream of a force main.
Concrete pipe downstream of both the Renton and the Interbay force main dis-
charges has experienced severe corrosion. Corrosion appears to have penetrated the
second layer of aggregate (1-inch loss) leaving only a short distance to reinforcing steel
in the pipe downstream of the Renton force main. The surface pH averaged 1.8. The
sewer downstream of the Interbay force main carries combined flow and occasionally
surcharges. The 21-year-old sewer pipe was PVC-lined in 1978 for about 200 feet
downstream of the Interbay force main; however, severe corrosion begins where the
liner ends. Assuming that the corrosion all occun-ed in the seven years following the
lining, the corrosion rate at the Interbay site is over 0.2 inches per year. Rust spots are
visible on the unlined concrete pipe wall, indicating that reinforcing steel will likely be
exposed soon. One and one-half inch is estimated to be missing. Measurements of
surface pH average 1.3 at Interbay.
Exposed aggregate and corrosion were observed around the flap gates at the
Hollywood Pump Station discharge and on concrete not protected by a PVC lining.
However, most of this structure is PVC-lined. The exposed portions are probably
exposed to erosional forces when the pumps discharge.
The manhole at Union Hill and Avondale roads showed shallow corrosion, zero
to 0.50 inches deep, and had a surface pH of 3. The inlet and outlet pipes to this
manhole were in good condition, even though the UPC lining was in poor condition.
The site at 15th Avenue and 188th Street NW also showed only shallow corrosion,
2-16
-------
which was limited to the outlet pipe. This sewer carries combined sanitaTy-stonnwater
flows. The surface pH level averaged 1.2 at this site. A wastewater total sulfide
concentration of 0.6 mg/1 was measured at the Union Hill Road site.
The frequent presence of force mains, required to overcome topographic barriers,
appears to increase the hydrogen sulfide corrosion problem in Seattle. Seattle feels
industrial metal bearing discharges have no correlation with corrosion, since that
industry has always had pretreatment standards.
2.2.7 Charlotte, North Carolina
In the Charlotte-Mecklenburg Utility District (CMUD) system, EPA compared
corrosion conditions in purely domestic sewers with conditions in sewers that cany
industrial flow. Approximately 12 metal finishers and a large foundry are permitted for
discharge into the CMUD sewer system. The field team entered six sewers with a large
flow contribution from industry and four sewers with only domestic flow.
CMUD personnel were not aware of system-wide hydrogen sulfide corrosion
problems, although a failure occurred in the Briar Creek sewer sometime prior to 1973.
Since that time, CMUD has been specifying tricalcium phosphorus as an additive to its
concrete pipe. CMUD also currently specifies a 1-inch sacrificial layer of concrete in its
pipe. In the late 1960s, CMUD had an odor study done on the Briar Creek Sewer, it
implemented a program of hydrogen peroxide addition for odor control in 1974. The
hydrogen peroxide was added to a point about 3 miles upstream of the Sugar Creek
Treatment Plant to which the Briar Creek sewer is tributary. This action was unrelated
to the prior Briar Creek failure. Strong odors at the Sugar Creek treatment facility
prompted another odor study in the late 1970s. The second study lead to the injection
of hydrogen peroxide at a location 0.50 miles upstream in both 54-inch influent lines to
the plant
The Charlotte water supply is categorized as "soft" by CMUD and has a 8.0- to
9.0-ppm total sulfate concentration.
Two of the domestic sites (Davidson Street at East 22nd Street, and Myers Street
at East 12th Street) are in the Sugar Creek drainage area and two (Arborway near
Sedley Road, and Old Providence Road near Sharonview Road) are in the McAlpine
Creek drainage area. The Sugar Creek sites are 7 and 6 miles from the treatment plant,
and 5 and 6 miles from the upstream end of the same drainage area, respectively. The
McAlpine Creek sites are 7 and 8 miles from the treatment plant, and 3 and 10 miles
from the upstream end of their respective drainage areas. All pipe observed in the
CMUD system is 20 to 25 years old. Wastewater sulfide concentrations at the four sites
ranged from 0.2 to 0.6 mg/1. Wastewater pH measurements were 6 at three sites, and
5.5 at the Old Providence Road site. It was the only site in Charlotte with pipe and
manhole surface pH measurements below 6. Pipe surface pH measurements were 4.5 to
2-17
-------
5. Corrosion product extended about 0.25 inches deep, exposing "peastone" aggregate at
this site.
Two of the four domestic sites experience turbulent flows due to a bend and an
obstruction. Although the wastewater contained measurable concentrations of sulfide at
each site (0.6 mg/1 and 0.4 mg/1), there was no measurable headspace hydrogen sulfide.
The Old Providence Road site has a 42-inch pipe and was flowing half full at about 2
fps when observed. The three clean pipes ranged in size from 24 to 54 inches in
diameter.
Three of the industrial sewers (Ganton Road at the Irwin Creek Bridge,
Remount Road at the municipal park, and Freedom Drive at Thrift Road) are in the
Irwin Creek drainage area, 1 to 6 miles from the treatment plant, and 5 to 10 miles
from the upstream end of the same drainage area. Two of the industrial sewers (Old
Nations Ford Road near Ervin Lane, and Granite Street near Continental Boulevard)
are located in the McAJpine Creek drainage area. The remaining industrial site is
located next to Park Road near Moncure Drive in the Sugar Creek drainage area. The
McAlpine Creek sites are located approximately 10 and 7 miles, respectively, from the
farthest upstream points in their drainage areas. The Granite Street site is about 1 mile
downstream of a 12,000-foot, 24-inch-diameter force main; the wastewater pH was 5.5 at
this site. The Park Road site is located about 7 miles from the farthest upstream point
in its drainage area.
Two of the six industrial sites showed signs of very shallow hydrogen sulfide
corrosion. The Remount Road site had lost just enough concrete to expose aggregate
and also had turbulent flow. The Granite site had turbulent flow and an observed
velocity of approximately 10 fps. This site also had four consecutive drop manholes
upstream. Pipe wall and manhole surface pH measurements were pH 6, and some
corrosion product was observed. Wastewater pH measurements were 6 at four of the
industrial sites, 5.5 at one site, and 10 at the remaining site. Wastewater sulfide ranged
from 0.0 to 0.3 mg/1. The wastewater sulfide level was 0.05 mg/1 at the site where
wastewater pH was 5.5, and 0.0 mg/1 at the site where wastewater pH was 10. There
was no measurable headspace hydrogen sulfide gas at any of the six industrial sites.
Pipe diameter at the industrial sites ranges from 21 to 54 inches, and all pipes are
approximately 20 years old. The observed flows range from one-third to two-thirds full,
from smooth to extremely turbulent, with velocities typically 2 to 4 fps.
2.2.8 Milwaukee, Wisconsin
The Milwaukee Metropolitan Sewerage District (MMSD) maintains
approximately 305 miles of sewer and two treatment facilities which serve approximately
one million people in the Milwaukee area. The average daily wastewater flow is 190
mgd, of which industrial flows represent over 25 percent MMSD estimates that 15
2-18
-------
percent of the area served by its system contributes storm flow. Wet weather flows at
both treatment plants double diy weather flows. The average biochemical oxygen
demand (BOD) is 200 mg/l, and total suspended solids (TSS) is 250 mg/1 at the two
plants.
In the MMSD system, EPA compared corrosion conditions in purely domestic
sewers to conditions in sewers that cany industrial flow. Approximately 90
electroplaters and metal finishers and about 15 tanneries aTe permitted for discharge
into the MMSD sewer under its pretreatment program. Some of the permitted
tanneries have waivers to discharge wastewater without pretreatment for sulfide, making
data obtained from the MMSD system particularly pertinent to this study. Observations
covered five sewers with only residential flow and five sewers with a heavy industrial
contribution to the flow.
MMSD personnel were not aware of any hydrogen sulfide corrosion problems.
The District recently inspected (by television) 20 percent of its large-diameter pipe.
Annually, it inspects an additional 40,000 feet MMSD also manually inspects manholes
and sewers during a standard manhole step replacement program and a seasonal
manhole cleaning program. MMSD has some odor problems; however, these are
located in parts of the system where the odors do not generate public complaints.
Three of the residential sites are in the northern part of the service area, 4 to 6
miles from the Jones Island Treatment Plant, and 3 to 5 miles from the upstream end of
the system. Pipe ages at these sites range from 50 to 70 years old. None of the three
sites revealed any wastewater sulfide. Wastewater pHs were all 6.5, and pipe and
manhole surface pHs were all 6.5. (According to carbonate chemistry, one would expect
weathered concrete to be about pH 6.3.) No corrosion or signs of corrosion of pipe or
manhole concrete were observed at these sites, even though one site is a junction
structure and another site is located just downstream of a pump station. In both cases,
these locations often experience turbulent flow and potential release of hydrogen sulfide
gas.
The other two residential sites, located in the South Shore Treatment Plant
basins, are 8 to 10 miles from the treatment plant, and 3 to 5 miles from the upstream
end of the basin. The first site is less than 20 years old, and the second site,
Kinnickinnic, is 50 years old. Observations at the 20-year-old site were similar to those
at the first three residential sites. However, a wastewater sulfide content of 0.5 mg/1 was
found at the Kinnickinnic site and pH of 3.5 was measured on the crown of the
downstream pipe. Kinnickinnic had severe corrosion from the water line up the pipe
about 1 foot Up to 1 inch of concrete appeared lost as estimated by aggregate
exposure in this 36-inch-diameter pipe. A black slime growth was observed from 1 inch
above the normal water line to 2 inches below.
Five sites had large amounts of industrial flow. Three sites are about 2 to 3
2-19
-------
miles north of the Jones Island Treatment Plant and two are immediately upstream of
the plant All five sites are at least six miles from the upstream end of the system and
are at least 40 years old. Corrosion was not observed at any of these sites.
Two of the industrial sites had measurable sulfide in the wastewater: 0.18 and
0.40 mg/1. Wastewater pH ranged between 6.5 and 7.5. Pipe and manhole surface pH
measurements ranged between 6.0 and 7.0. One of the industrial sites was located less
than 0.5 miles downstream from a tannery. Two sites had initial hydrogen sulfide gas
concentrations of between 0.5 and 0.6 ppm in the pipe headspace. One site located in
the downtown industrial area could not be entered because of a photoionization meter
reading of greater than 1,000 ppm. Two sites had abrupt changes in direction 20 to 30
feet upstream from the manhole and 6 to 8 inches of bottom debris. Typical at these
sites was a grease buildup on pipe and manhole walls, calcium buildup, and slime, but
solid concrete pipe underneath.
23 Site Visits to Assess Hydrogen Sulfide Corrosion at Wastewater Treatment
Plants and Pump Stations
Site investigations were conducted at five wastewater treatment plants in three
cities. The purpose of these investigations was to document the location, nature and
severity of hydrogen sulfide corrosion problems at these facilities. The wastewater
treatment plants included the Hookers Point facility in Tampa, FL, the East Bank and
West Bank plants in New Orleans, LA, and the Hyperion and Terminal Island plants in
Los Angeles, CA. Pump station corrosion was also investigated as part of these site
visits.
The type and extent of information available from the various cities varied
widely. Some cities closely monitored hydrogen sulfide levels in the wastewater and in
the atmosphere, and maintained detailed records of corrosion repair and rehabilitation
efforts. Others had done little to monitor or control corrosion.
The following provides a summary of the information collected from the site
visits to cities where corrosion was believed to be a problem in the wastewater treatment
plant and pump stations.
2J.1 Tampa, Florida
2J.1.1 Wastewater Treatment Plant
The Hooker's Point Advanced Wastewater Treatment Plant was expanded in
1978 to handle a design flow of 60 mgd. The plant is averaging approximately 57 mgd,
and employs advanced waste treatment (AWT) for biological nitrogen removal. Unit
processes at the plant include influent screens and grit chambers, primary clarification,
two stage activated sludge treatment, secondary clarification, denitrifying filtration,
2-20
-------
chJorination and dechlorination. The plant achieves nitrificationydenitrification before it
discharges to Tampa Bay. Sludge handling processes are varied, and consist of gravity,
dissolved air flotation, or belt filter thickening of waste activated sludge, anaerobic or
aerobic digestion, and belt press or drying bed dewatering. A new sludge dryer and
pelletizer will come on-line in the fall of 1990.
Hydrogen sulfide corrosion at the wastewater treatment plant is very severe. The
walls of the influent junction box were constructed with a corrosion-resistant plastic
liner. H2S corrosion is also severe in the screen and grit building and in the effluent
chamber in the grit building. Dissolved sulfide is approximately 10 mg/1 in the influent
wastewater. Concrete on the roof of the junction box had also corToded to an extent
that-the aggregate was exposed. All mechanical equipment showed mild to severe
corrosion. Hand rails, platform, and other structures at the primary clarifiers were
corroded.
The plant expends significant resources to combat hydrogen sulfide corrosion.
All carbon steel parts have been replaced by stainless steel parts wherever possible.
Electrical components have been covered and electrical sockets replaced using plastic
materials. A very rigorous painting schedule is maintained on all equipment and parts
at the junction chamber. H2S levels in the atmosphere of the screen and grit building
are as high as 20 ppm. A fine-mist scrubber was installed to treat the H2S-laden air
emissions from the junction box. Although designed to handle SO ppm of H2S, levels
entering the scrubber range from 400 to over 1000 ppm. The capital cost of the
scrubber system was approximately $1,000,000. Annual operating cost is estimated to be
S400,000/yr.
The primary clarifiers at the wastewater treatment plant are also at an advanced
stage of corrosion. Some clarifiers are 40 years old and the others were built during the
expansion. There is little corrosion at the influent end of the clarifiers but severe
corrosion at the effluent end. The wastewater has a fall of four feet in the effluent
channel thereby creating turbulence and releasing H2S to the atmosphere with the result
that the concrete structure at the effluent channel is severely corroded.
Most of the moving parts on the clarifiers have been replaced by plastic,
including the scraper mechanism. Gear motors and electrical/mechanical components
are covered with corrosive-resistant materials. Approximately 2 to 4 inches of the side
walls at the effluent channel in the primary clarifiers have been lost due to corrosion.
At some locations, reinforcing steel was visible. The rehabilitation of the clarifiers is
now under contract and includes the installation of a plastic liner on the walls.
Hydrogen sulfide corrosion downstream of the clarifiers is very limited. There is very
little hydrogen sulfide corrosion found at other treatment processes and sludge handling
facilities.
Hydrogen sulfide corrosion of instrumentation and controls at the wastewater
2-21
-------
treatment plant was severe at the transformer cabinets. All copper tubing and wiring
corrodes rapidly. Corrosion of electrical contacts was widely observed. Switchgear at
the influent junction chamber also corrodes rapidly. Corrosion prevention measures for
instrumentation and control equipment includes covering the instruments, purging with
clean air, and air conditioning control rooms. All electrical equipment at the plant is on
a preventative maintenance and painting schedule. Contacts and relays are cleaned
regularly. Transformer housings must be replaced periodically.
Although corrosion of sludge handling components and structures has been a
problem in the past, such problems have largely been eliminated through gradual
replacement with corrosion resistant materials such as galvanized and stainless steel.
Spare parts are stored in an air-conditioned warehouse to prevent corrosion. Minor
corrosion problems are still evident where components such as conduit fittings are not
available in corrosion resistant materials.
2-3.1.2 Lift Stations
There are 160 lift stations in the sewer system that collect and transport
wastewater to the treatment plant The more recent pump lift stations are built of
concrete.
Medium to very high rate corrosion was found at many of the lift stations. Most
of the manholes, wet wells and interior control room walls in lift stations have sulfur
(yellow) deposits. There was severe corrosion near turbulent areas of the lift stations.
The concrete was corroded and reinforcing steel was visible. Most of the lift stations
have mild to severe corrosion present Steel sound enclosures over wet wells had to be
replaced by fiberglass buildings. Most of the larger pump stations have fine-mist
scrubber systems. The City tried a hydrogen peroxide dosing system, but it was judged
to be too expensive to operate. A few lift stations have used a ferrous sulfate dosing
system to control H2S. The City also tried packed tower air scrubbers. They were very
high in maintenance. Carbon adsorption systems were also installed on some lift
stations.
Corrosion of instrumentation and control systems at the lift stations was not
quite as severe as at the plant This was primarily due to the active preventative
maintenance program imposed by the City. Copper tubing and exposed wiring were
seen to be corroded. All motor control centers and electrical equipment were covered.
2_3.2 New Orleans, Louisiana
2.3.2.1 Wastewater Treatment Plants
The East Bank and West Bank wastewater treatment plants of the City of New
Orleans were visited to document the extent of hydrogen sulfide corrosion at the
2-22
-------
facilities. The East Bank plant treats the sanitary flows from downtown and the
northeast part of the City. The plant was originally built in 1963 for primary treatment
and was later expanded for secondary treatment in 1980. The original design flow at
the plant was 30 mgd but the facility has been expanded to handle 122 mgd. A total of
1500 miles of collection system comprised of gravity and force mains collect and convey
sewage to the plant The treatment plant consist of screens and grit removal, pure
oxygen activated sludge system and secondary settling. Effluent is discharged to the
Mississippi River. Secondary sludge is dewatered and then incinerated. The ash, along
with screenings and grit, are disposed of in a sanitary landfill.
v Plant headworks at the East Bank plant had severe corrosion in the screen and
grit basins. Some parts of the grit basins were built in 1963 and were then expanded to
meet the new design flows. Three force mains feed wastewater to these grit basins.
One force main conveying flows from the City has long detention times, and hence the
wastewater is very septic when it reaches the plant The color of the wastewater was
very dark (black) and was deficient in D.O.
The side walls of the grit chamber were severely corroded. Approximately 1 to
V/z inches of concrete was corroded away at some locations. Severe corrosion was also
observed at the effluent end of the grit box where the wastewater spills into a channel
which led it to the pure oxygen activated sludge tanks. The grit chambers were installed
with screens on each pass. These screens were in a deteriorated condition. Many of
the-components of the screens had rusted and the metal frames on which they were
attached were corroded along with the concrete below the frames.
Corrosion of instrumentation and controls was found to be severe at the East
Bank plant Contacts on electrical equipment were oxidized. The plant personnel
replace small items and clean contacts and equipment on an annual basis. They
sometimes must take equipment off-line for service and maintenance. As preventative
maintenance, they use a light coating of oil, and cabinets purged with cleaned air. The
plant has entered into an annual preventative maintenance contract They allocate two
men 1 to 1-1/2 days/wk for electrical equipment maintenance. The electrical contacts on
indicator lights, pump relays, and contacts operate intermittently because of oxidation
problems at the contacts. The instrument control room is fully air conditioned. Air
cleaning is done through permanganate beads which are replaced every month. The
plant expends significant effort for replacement and maintenance of the electrical and
instrumentation components.
The plant does not have any control measures to prevent future corrosion. No
efforts have been made to rehabilitate the corToded structures. The plant has a limited
budget and does not plan to employ rehabilitation of structures as a corrective action
until there is a failure.
The West Bank plant serves the population of the western side of the City of
2-23
-------
New Orleans. The plant was originally built in 1971 for a design maximum flow of 15
mgd. The average dry weather flow (ADF) to the plant is approximately 7 to 8 mgd.
The plant is dow under design for expansion to 40 mgd. The treatment plant consists of
influent bar screens, grit removal, primary sedimentation, high rate trickling filters,
secondary sedimentation, chlorine contact and final discharge to the river. The sludge
from the clarifiers goes to a thickener and a vacuum filter and is then incinerated. The
ash from the incinerator is disposed of in a local landfill.
The West Bank plant also has severe corrosion at the influent head box where
the screens and grit chamber are located. Corrosion has degraded the sidewalls on the
grit chamber to a depth of 1 to 1 Yi inches. Again, corrosion was found to be severe at
areas of high turbulence i.e. at the influent and effluent end of the grit basins. The
metal grating and handrails on the grit basins were also corroded. The wastewater
entering the plant was septic and the dissolved oxygen was always found to be 0 mg/1
except during heavy rainfalls when the D.O. would increase to 0.2 mg/1. As the plant is
located adjacent to a golf course, there are plans to cover the plant headworks, the
sludge thickener and some other tanks to control odor emissions.
There are no efforts being taken to rehabilitate the degraded structures. No
rehabilitative techniques have been employed to correct the odor and corrosion
problems.
The vacuum filters at the West Bank Plant are located in a building that is
equipped with a passive air ventilation system. The mechanical and support parts of the
vacuum filters are in a severely corroded state. The plant had to replace grating over
the filter supports. When the filters are operating, high H2S levels are reported in the
building. There is no corrosion reported at other parts of the plant Corrosion at
instrumentation and controls is minimal. Corrective action at this plant is based
primarily on minimizing odors which are affecting the neighboring golf course.
2.3.2.2 Lift Stations and Collection Systems
There are a total of 87 lift stations and 1500 miles of sewers that serve both the
East and West Bank Treatment Plants in the City of New Orleans. The lift station wet
wells are made of brick and concrete. Force mains range in size from 42 to 52 inches
and are constructed of cast iron, steel or concrete. Ninety to 95 percent of the
collection system is 8 to 10 inch diameter pipes. Concrete pipes were laid in late 1930's.
There are a few older pipes made of clay. Since the 1970's, plastic pipe has been used
where possible.
All of the 87 lift stations employed in the collection systems for the East Bank
and West Bank plants are in some stage of corrosion. The older lift station wet wells
were built of brick and are severely deteriorated. The pump base and supports have
corroded and at some places are on the verge of falling down into the wet well.
2-24
-------
Rehabilitation of brick wet wells consists of coating by gunite. The New Orleans
Sewage Board experimented with pump cycle times to minimize detention times and
decrease H2S levels. Continuous ventilation is provided in the lift stations at six air
changes per hour. At some places the Board has tried adding ferric chloride but found
that it forms clinkers in the incinerator at the plant H2S levels in the atmosphere of
the wet wells average approximately 100 ppm. The Board spends around $5.2 million
per year for lift station maintenance. About 30 percent of total man hours is utilized
for lift station maintenance. Electrical and instrumentation equipment have minor
corrosion problems. New electrical equipment has been installed with clean air supplied
by treatment through potassium permanganate. There is reported to be more corrosion
in lift station wet wells at the east side of town.
2J.3 City of Los Angeles
2J .3.1 Hyperion Wastewater Treatment Plant
The plant is designed for 400 mgd through primary treatment and 150 mgd
through secondary treatment Present day flows are 370 mgd and 200 mgd, respectively.
The ability of the secondary process to handle the additional flow is attributed to the
addition of fine bubble diffusers. The headworks, primaries, secondaries and anaerobic
digesters are approximately 40 years old. Regulations eliminating ocean sludge disposal
and requiring full secondary treatment, along with population growth, have resulted in
10 years of construction at the plant The City foresees at least another 5 to 10 yean at
the same pace. The latter includes replacement of the existing secondary process with a
pure oxygen process.
With the exception of the gravity degritter in the east headworks, all trash and
grit removal tankage are under cover, making direct observation of corrosion on these
processes difficult without considerable expenditure of staff manpower. The covers on
the west aerated grit chamber effluent channel were small enough to be managed by
one person and were lifted for observation. Corrosion of the concrete sewer at those
points was observed to be severe, with penetration to at least 12 inches at the water line
diminishing to 1 to 2 inches in the closed channel and 0 to 1 inch at ground level of the
open tank. The plant carpentry superintendent in charge of all in-house concrete repair
indicated the observed areas were typical of all headworks tankage of the same age.
The cost of these repaid are not segregated from general plant maintenance costs.
The extent of corrosion below the water line in both tanks and channels was
described as minor (less than 1 inch) even in the oldest tankage. All covers (tank and
channel) and deck plates are made of aluminum, as were handrails, conduit and other
hardware (some stainless steel). No corrosion of these materials was apparent
The headworks processes are all contained in buildings. The ambient
atmosphere of the buildings is swept by fans and discharged to a collection point at the
2-25
-------
suction of the secondary process blowers. Thus a slight negative pressure is maintained
in each building. This prevents noxious odors from escaping the plant and with normal
infiltration plus some outside air intakes, avoids the build up of coirosive gases in the
atmosphere of the process buildings. All windows in the aerated grit chamber building
were sealed in order to reduce escape of hydrogen sulfide, even though the tanks are
covered. The few pieces of carbon and galvanized steel found in the buildings were
severely coaoded. This was especially true of steel doors. No maintenance program is
in force for the doors other than repainting when scratched or chipped. The ambient
air removal system piping is fiberglass and most other piping is PVC. Conduit is
aluminum or PVC.
A short section of the force main entering the plant collapsed and was replaced
in 1987. The collapse was attributed to corrosion-weakened concrete pipe combined
with the ground vibration caused by heavy construction equipment
The decision to rehabilitate or replace all or part of the headworks has yet to be
made. There is obvious structural damage in some places and some doubt in the mind
of staff as to the structural integrity of a rehabilitation effort given the frequency of
earthquakes in the area. In either case, PVC liners with concrete slabs will be used in
all channels and the inside of all tankage will be at least coated with coal tar or an
alternative coating material.
The primary clarifiers are covered with concrete slabs so casual inspection was
not possible. The influent and effluent channels are covered by aluminum plates which
can be easily removed for inspection. Like the headworks, concrete exhibited deep
corrosion penetration from the water line to the surface, with some of the deepest
penetration (6 to 8 inches) at the surface adjacent to the channel covers. Most of the
corrosion at the top has been repaired by cutting back to good concrete, reforming to
the original geometry and grouting. These repairs are recent, and are not covered by
any protective coating. An epoxy-type coating had been applied to the early patches
and began peeling almost immediately, so coating was discontinued. The channels will
be covered with PVC liners. The type of coating for the inside of tank walls and covers
is as yet undetermined. They are in the process of converting from steel to plastic chain
and from wood to fiber glass boards for the sludge rakes. The existing primaries will be
rehabilitated once new primary construction is complete.
With the exception of anaerobic digestion, the sludge handling processes came
on-line in 1985-1986. Ocean disposal of sludge ceased in 1987, and digested sludge is
now either dried and applied to power generation (Carver-Greenfield process) or
dewatered by centrifuges and transported to a Yuma, AZ land application site. Due to
safety regulations for construction at the site, the sludge handling facility was off-limits
to visitors. The addition of fenous chloride (280 mg/1) for hydrogen sulfide reduction
after sludge digestion is to control sulfur emissions as opposed to corrosion control.
2-26
-------
All instrumentation and control electronic equipment is conformably coated (a
thin lacquer-like coating applied to circuit boards and components to seal them from
the atmosphere) in the manufacturing process. This is standard practice in the industry
for wastewater treatment equipment suppliers. In addition, all field mounted
instrumentation (sensors, transmitters, etc.) are nitrogen purged. The case of each
instrument is connected to a low pressure nitrogen supply which maintains a slight
positive pressure in the instrument housing to prevent exposure of the components to
ambient air. Inspection of the equipment disclosed no sign of corrosion. All circuit
boards, contacts, wire terminations and other exposed metal was bright and shiny. The
annual cost of nitrogen is estimated at less than S3,000. The only sensing elements
immersed in liquid process streams are DO probes. These are newly installed and as
yet have no track record. The control room is isolated from ambient atmosphere by
scrubbing, filtering, and air conditioning. No problems were reported or apparent with
these systems.
Although not as severe, there is ample evidence of concrete corrosion in
secondary treatment The worst is at the aeration basin influent mixing channel where
corrosion has penetrated to the reinforcing steel (2 inches). Other areas of the reactors
have exposed aggregate. Steel hand rails and steel plate on the side of the reactors are
pitted and rusted where chipped or peeled paint allowed exposure to atmosphere.
Since a new oxygen activated sludge system is planned, only those repairs
necessary for the existing system to remain operational will be made.
The scavenged air recovered from buildings and below tank covers is ducted to
the aeration basin blowers for scrubbing in the activated sludge mixed liquor. This air is
not cleaned by other than conventional blower inlet air filters, nor are the blowers
constructed of special corrosion resistant materials. This has not caused any additional
blower maintenance or reduced the useful life of the blowers. The only impact is on the
carbon steel linkage that moves the internal guide vanes and this impact is considered
minor by the maintenance staff.
2JJ.2 Terminal Island Wastewater Treatment Plant
The original plant was constructed in 1935 and completely rehabilitated in 1977.
The plant is designed for full secondary treatment of 30 mgd. Present day dry weather
diurnal flow ranges were modified from 5 to 35 mgd, a 7:1 ratio, to 10 to 30 mgd, a 3:1
ratio, by requiring (as part of pretreatment enforcement) local industries to shift
discharges to off-peak hours. Over 50 percent of the flow and 70 percent of the load is
of industrial origin.
With the exception of the bar screen, all trash and grit removal tankage is under
cover, making direct observation of corrosion on these processes difficult without
considerable expenditure of staff manpower. Corrosion of the concrete at the bar
2-27
-------
screens was negligible at those points observed, with penetration barely to the aggregate
at the water line. The extent of corrosion below the water line in both tanks and
channels was described as minor (less than 1 inch) in the oldest tankage. All covers
(tank and channel) and deck plates are made of aluminum, as were handrails, conduit
and other hardware (some stainless steel). No corrosion of this material was apparent
The bar screen frame and sheet metal is of coated (coal tar) carbon steel, which was
severely corroded. Most of the sheet metal has been replaced with sheet PVC. The
frame (Vi inch angle iron) will probably be replaced with stainless steel.
The headworks processes are all contained in buildings. The ambient
atmosphere of the buildings is collected by the suction of the secondary process blowers
(no fans). Thus a slight negative pressure is maintained in each building. This prevents
noxious odors from escaping the plant and with normal infiltration plus some outside air
intakes, avoids the build up of corrosive gasses in the atmosphere of the process
buildings. The few pieces of carbon steel found in the buildings were severely conoded
including galvanized steel hardware. This was especially true of steel doors. No
maintenance program is in force for the doors other than repainting when scratched or
chipped.
The ambient air removal system piping is fiberglass and most other piping is
PVC. Conduit is aluminum or PVC.
The primary clarifiers are fitted with aluminum covers. The influent and effluent
channels are also covered by aluminum plates which can be easily removed for
inspection. Plant staff had previously converted from steel to plastic chain and from
wood to fiberglass boards for the sludge rakes. Because of problems with the plastic
chain jumping the sprockets, they are converting back to steel chain.
The egg shaped anaerobic digesters appear to be in good condition externally.
An external pipe that collects gas for mixing has been replaced with a welded stainless
steel pipe. The sacrificial anodes are replaced routinely as part of the maintenance
program. The motorized valves located on top of the digesters are also being replaced,
but this is because they do not have weather proof housings, although the problem may
have been exacerbated by hydrogen sulfide. The earth ground bonding wire (bare
copper) in this location has almost turned to dust and is being replaced with an
insulated wire. This location is also exposed to winds from the sea, and the corrosion
observed may be the result of salt air. The elevator at this location is a high
maintenance item, since it is exposed to both sea air and ambient hydrogen sulfide.
The addition of ferrous chloride (450 mg/1) for hydrogen sulfide reduction (10
fold) after sludge digestion is to control sulfur emissions as opposed to corrosion
control.
All instrumentation and control electronic equipment is conformably coated. In
2-28
-------
addition, all field mounted instrumentation (sensors, transmitters, etc.) are nitrogen
purged. The case of each instrument is connected to a low pressure in the instrument
housing to prevent exposure of the components to ambient air. Inspection of the
equipment disclosed no sign of corrosion. All circuit boards, contacts, wire terminations
and other exposed metal was bright and shiny. The annual cost of nitrogen is estimated
at less than $2,000. The only sensing elements immersed in liquid process streams are
DO probes. These are relatively new yet have performed well to date. The control
room is isolated from ambient atmosphere by scrubbing, filtering, and air conditioning.
No problems were reported or apparent with this system.
The scavenged air recovered from buildings and below tank covers is ducted to
the. aeration basin blowers for scrubbing in the activated sludge mixed liquor. This air is
not cleaned by other than conventional blower inlet air filters, nor are the blowers
constructed of special corrosion resistant materials. This has not caused any additional
blower maintenance or reduced the useful life of the blowers. The only impact is on the
carbon steel linkage that moves the internal vanes and this impact is considered minor
by the maintenance staff.
2.4 Site Visits to Investigate Corrosion Mechanism
Although many authors have discussed the role of the sulfate-reducing bacteria
and sulfur-oxidizing bacteria in hydrogen sulfide corrosion, little is known about these
organisms or the factors that may promote or inhibit their growth. To confirm existing
theories concerning hydrogen sulfide corrosion, gain additional insight into the
mechanisms of corrosion, and lay a foundation for future work, a multidisciplinary field
team was assembled to make observations and collect samples for microbiological,
physical, and chemical analysis from locations in the CSDLAC system and the
Metropolitan Seattle system. The field team included microbiologists who had studied
sulfate-reducing bacteria, sulfur-oxidizing bacteria, and microbiologically-induced
corrosion, and a structural consultant who had studied corrosion of concrete.
The field program for this work segment involved collecting samples from three
locations in each of the two sewer systems. Two of the sites in each system were
locations where corrosion was well established and ongoing. A third site in each system
where corrosion had not been observed was selected as a control site. Samples of
surface deposits were collected from the crown, sidewall, and waterline areas of the pipe
at each location, as well as samples of the wastewater itself for microbial and chemical
analysis. Samples of concrete were chipped from the crown and sidewall areas for
chemical and physical analysis.
2.4.1 Results of Microbial Analysis
The microbial analyses showed that a large and complex microbiological
community is present in the wastewater and on the structure walls and crown at the
2-29
-------
locations sampled. Veiy high levels of aerobes, anaerobes, acid-producing bacteria, and
sulfate-reducing bacteria were found in most wastewater, sidewall, and submerged
sidewall samples. Relatively large populations of sulfate-reducing bacteria were found
in the bulk wastewater samples. However, CSDLAC slug dosing of caustic has been
shown to suppress sulfide generation for 7 to 10 days after treatment, indicating that the
slime layer is the predominant site for reduction of sulfate to sulfide (1).
The microbial community of the pipe crown region appears, particularly at
conoded sites, to be significantly different. A much lower level of viable organisms,
including aerobes, organic acid-producing bacteria, and sulfate-reducing bacteria was
found in this region. It is possible that the low pH of corroding crown regions reduces
the total level of viable organisms and selects for the organisms that do live there.
The microbial analyses showed, nonetheless, that samples from areas of crown
corrosion in both the CSDLAC and Seattle systems contain large numbers of
acidophilic, sulfur-oxidizing bacteria of the genus Thiobacillus. and probably the species
thiooxidans. Samples from crown areas in the CSDLAC system where reinforcing steel
was exposed also showed the presence of T. ferrooxidans. an iron- and sulfur-oxidizing
bacterium. Samples from the location in the CSDLAC system where wastewater was
being treated with ferrous chloride to control hydrogen sulfide showed fewer T.
ferrooxidans and acidophilic sulfur oxidizers in the crown area than in the corroded
location being slug dosed with NaOH.
Large numbers of sulfur-oxidizing organisms that grow at neutral pH were also
found in all samples, including the wastewater The wastewater could continuously
inoculate these organisms on the concrete surface. These organisms can oxidize sulfur
and produce sulfuric acid, which would lower the pH of the concrete surface. Lowering
the pH would allow more acidophilic organisms, (e.g., T. thiooxidans. which has a pH
preference of 4.5 to 1.0) to grow. Additionally, large numbers of acid-tolerant fungi and
yeasts were present in most samples. The role of these organisms in contributing to
corrosion is not understood.
All the sites in Seattle appeared to have similar microbial communities which
included high levels of sulfate-reducing bacteria in the wastewater and in the surface-
associated populations. More variety of species were observed in the Seattle sites,
however. The sites which had high corrosion rates had very turbulent wastewater flows
which could increase the out-gassing of hydrogen sulfide to the sewer headspace. This
is supported by the fact that these sites had high levels of hydrogen sulfide in the sewer
headspace.
At CSDLAC, there were major differences between the control site with low
corrosion rate and those sites with high corrosion rates.
• The amount of moisture on the crown at the control site was low
2-30
-------
compared to the other sites.
• The levels of metals in the springline sample were higher at the control
site than at the other sites.
• The average sewage age (the length of time the sewage is in the sewer
system) was less at the control site than at the other sites.
• The levels of sulfate-reducing bacteria at the control site were 10,000 times
lower than at the other sites.
The low level of moisture on the crown at the control site certainly would inhibit
both growth of bacteria on the crown and the conosion reaction. However, even if
moisture was present, the corrosion rate would be low due to the low levels of sulfate-
reducing bacteria and hydrogen sulfide. The temperature differences are not considered
sufficient to affect the growth of sulfate-reducing bacteria and Thiobacilli. This leads to
the conclusion that the low levels of sulfate-reducing bacteria and hydrogen sulfide
production are responsible for the low rate of conosion at the control site, and by
implication that the high levels of sulfate-reducing bacteria at the other sites are
responsible for the high rates of conosion.
The results of microbial analysis support the accepted theory of hydrogen sulfide
corrosion: production of hydrogen sulfide by bacteria in the wastewater and on the
sewer wall, followed by oxidation of the hydrogen sulfide to sulfuric acid by bacteria on
the sewer crown. However, unanswered questions remain concerning:
• The role of acidophilic heterotrophic organisms in the conosion process.
• The minimum sulfide levels needed to sustain the growth of sulfur-
oxidizing bacteria.
• Whether techniques could be developed to interrupt the microbial
pathways of sulfide generation and sulfuric acid production.
2.4.2 Results of Physical Analyses
The samples of concrete collected in the CSDLAC and Seattle Metro sewer
systems were evaluated using standard petrographic examination techniques and X-ray
diffraction techniques. In addition, chloride ion content was determined on selected
samples.
Petrographic examination revealed the presence of conosion at all six of the
sampled sites, including the "control" sites. Corrosion at the control site in the
CSDLAC system was shallow (0.04 to 0.08 inches) and the high surface pH values
2-31
-------
measured at this site, as well as observations, indicate that this corrosion may have
occurred in the past At the remaining five sites, petrographic examination and X-ray
diffraction analysis of samples collected above the waterline showed gypsum deposits
covering a zone of deterioration extending up to 0.5 inches beneath the concrete
surface. Within this zone, the normally crystalline concrete paste showed reduced levels
of crystallinity and an increasingly amorphous nature depending on the extent of
deterioration. This concrete was soft in comparison to non-deteriorated concrete.
Microcracks in the concrete (typical of most concrete) were usually lined with gypsum
and, in some instances, ettringite also. Below the waterline, the concrete samples were
sound. These conditions are considered typical of classical hydrogen sulfide corrosion
of concrete.
Determination of chloride ion levels in the concrete samples indicated that, with
the exception of one location in Seattle, chloride levels in the concrete of the original
pipe were below the threshold needed to initiate chloride-induced electrochemical
corrosion of reinforcing steel.
The observations at the two control sites point out the difficulty involved in field
assessment of the preseace and extent of corrosion. The concrete surface at both
control sites looked smooth and in good condition, and the pipe wall at both sites
produced a sharp ringing sound indicative of sound concrete when struck with a rock
hammer. However, the surface pH at the CSDLAC site was 4.0, and laboratory
examination of samples revealed that shallow con-osion bad occun-ed, perhaps in the
past At the Seattle Metro control site, surface pH levels were between 1.0 and 2.0, and
laboratory examination indicated that corrosion had penetrated up to 0.25 inches into
the concrete.
2.5 Other Cities Reporting Hydrogen Sulfide Corrosion
Information was analyzed from surveys conducted by the County Sanitation
Districts of Los Angeles County (CSDLAC), Association of Metropolitan Sewerage
Agencies (AMSA), and the Water Pollution Control Federation (WPCF). Results of
the CSDLAC survey are summarized in Table 2-2. Of the 89 cities responding to this
survey, 32 cities (36%) reported sewer collapses. Twenty-six cities experienced collapses
that were believed to be due to hydrogen sulfide conosion. Thirty cities (34%) reported
taking measures to control sulfide generation in sewers. Fifty-six cities (63%) had taken
steps to protect pipe from corrosion, or had rehabilitated pipe damaged by corrosion.
The AMSA survey asked if the municipalities experienced hydrogen sulfide
corrosion at the treatment plant Almost 70 percent of the 61 respondents responded
positively. In addition, 34 percent of the respondents indicated that they were currently
employing techniques to control sulfide generation. Results of the AMSA survey are
summarized in Table 2-3.
2-32
-------
TABLE 2-2
SUMMARY OF CSDLAC SURVEY DATA
__ CorTosion
Sewer Sulfide Protection/
City
State
Collapse
Control
Rehab.
Birmingham
AL
X
X
Phoenix
AZ
X
X
Tucson
AZ
X
Little Rock
AR
X
Pine Bluff
AR
X
X
Carlsbad
CA
X
X
Cucamonga
CA
X
Orange County
CA
X
X
X
Whittier
CA
X
X
X
Colorado Springs
CO
Denver
CO
X
Hartford
CT
Washington
DC
Fort Lauderdale
FL
X
Jacksonville
FL
X
X
Miami
FL
X
Orlando
FL
X
X
X
Tampa
FL
X
X
Atlanta
GA
X
X
Honolulu
HI
X
X
X
Boise
ID
X
Chicago
IL
Downers Grove
IL
Elgin
IL
Kankakee
IL
2-33
-------
TABLE 2-2 (cont)
SUMMARY OF CSDLAC SURVEY DATA
Corrosion
Sewer Sulfide Protection/
Citv State Collapse Control Rehab.
Rockford IL X
Springfield IL X
Urbana IL X
Wichita KS X
Louisville KY XX
Jefferson Parish LA X X
New Orleans LA X X
Baltimore MD X
Glen Buroie MD
Hyattsville MD X X
Boston MA X X
Salem MA X
Detroit MI X
Kalamazoo MI
Duluth MN X
St Paul MN X X
Kansas City MO X
St. Louis MO X X X
Omaha NE X X X
Bayville NJ XX
Elizabeth NJ
Little Feny NJ XX
Newark NJ X
Sayreville NJ X
2-34
-------
TABLE 2-2 (conL)
SUMMARY OF CSDLAC SURVEY DATA
Corrosion
Sewer Sulfide Protection/
City State Collapse Control Rehab.
Albuquerque NM XX X
LasKZruces NM X
Albany NY
Buffalo NY
New York City NY
Mineola NY X X X
North Syracuse NY X
Rochester NY X
Greensboro NC X X
Akron OH X
Cincinnati OH X
Cleveland OH
Columbus OH X
Dayton OH
Toledo OH
Tulsa OK
Hillsboro OR X
Portland OR
Oregon City OR
Philadelphia PA X X
Pittsburgh PA X
Providence RI
Chattanooga TN X X
Knoxville TN X X
Memphis TN X
Nashville TN X X X
2-35
-------
TABLE 2-2 (conL)
SUMMARY OF CSDLAC SURVEY DATA
Corrosion
Sewer Sulfide Protection/
City
State
Collapse
Control
Rehab.
Arlington
TX
X
Dallas
TX
X
X
El Paso
TX
X
X
X
Fort Worth
TX
X
X
X
Houston
TX
X
X
San Antonio
TX
X
X
Salt Lake City
UT
Fairfax
VA
X
X
Virginia Beach
VA
X
X
Seattle
WA
X
X
X
Tacoma
WA
X
X
Green Bay
WI
X
Madison
WI
X
X
X
Milwaukee
WI
2-36
-------
TABLE 2-3
AMSA SURVEY SUMMARY
Question
Percent
Responding Yes
Percent Number of
Responding No Responses
1. Are any or your sewers unlined reinforced
concrete pipe (RCP)?.
2. Has the rate of corrosion of the RCP increased
since implementation of national pretreatment
standards?
89
11
95
62
37
N>
I
u>
-J
3. Have core borings of the RCP been taken to
determine how much the pipes have corroded?
4. Does the industrial wastewater conveyed by
the RCP contain heavy metals?
5. Have you taken measures (e.g., addition of metal
salts to the sewers) to reduce the sulfide in
your sewers?
6. Has your method of sulfide reduction reduced
the rate of corrosion of the RCP's?
19
82
34
78
81
18
66
22
59
61
61
18
Have you experienced sulfide corrosion of
structures at your treatment plant(s)?
69
31
61
-------
Id 1989, the Water Pollution Control Federation received survey responses from
1003 wastewater treatment plants in the United States. The survey of "Problem
Technologies and Design Deficiencies at POTWV was intended to identify information
needed to improve wastewater treatment plant design and operation. The WPCF Plant
Survey database was utilized by this study as a means to determine the source of
corrosion problems at wastewater treatment plants. The survey asked the 1003
participants to rate the level of problems in various areas of the plant and also asked
questions about plant age, daily flow, and methods of operation.
Several questions dealt with the level of corrosion experienced in major portions
of the plant, as well as the incidence of influent H2S and odors. A summary of the
responses related to corrosion problems is shown in Table 2-4. The approach taken to
determine the source of corrosion was to develop a matrix of responses in an attempt to
correlate the incidence of corrosion to such factors as influent H2S, odors, age and
recycle streams.
Overall, it was found that 68 percent of the plants surveyed experienced some
level of corrosion within the plant The same level of corrosion is generally experienced
in all portions of the plant. For example, if the respondent classified coirosion
problems at preliminary treatment as major but periodic, then the corrosion problems
experienced at the secondary clarifier tended to be rated the same.
The age of the plant does not explain the consistent level of corrosion problems
in all sections of the plant When compared to the level of corrosion at preliminary
treatment, the same age distribution was seen for all levels of corrosion. The plants
reporting no corrosion at preliminary treatment are generally over 20 years old and have
had major liquid and sludge train expansions in the last 5 years. The plants showing
some degree of corrosion are between 1-15 years old but have also bad train expansions
in the last 5 years.
As for plant size, the majority of wastewater treatment plants without corrosion
problems in preliminary treatment handle 0.1 to 1.0 mgd (42%) or 1 to 5 mgd (31%)
average flow. The plants exhibiting corrosion are rather large - 34 percent treat
between 1 and 5 mgd. One interesting finding is that of the largest wastewater
treatment plants (>10 mgd), those reporting corrosion outnumber those without
corrosion by almost 7 to 1. The distribution of corrosion severity for each size category
is generally the same, but with a tendency for larger plants to show slightly more severe
corrosion problems.
The recycle of filtrate or supernatant streams from digesters, thickeners and
dewatering equipment does not appear to increase the incidence of corrosion at
preliminary treatment The distribution of responses was similar for those plants which
recycled and those which did not There was also no tendency for an increase in
corrosion with a particular recycle stream. For all recycle streams, the level of corrosion
2-38
-------
TABLE 2-4
SUMMARY OF RESPONSES TO WPCF SURVEY - CORROSION
OF WASTEWATER TREATMENT SYSTEMS
Corrosion Problem Severity and Frequency,
Percent of Responses
Minor Problem Major Problem Total
Periodic
Continuous
Periodic
Continuous
Preliminary Treatment
30
23
6
9
68
Aeration Basins
27
15
3
4
49
Fixed Film Systems
23
13
3
5
44
Susp. Growth Clarifiers
27
13
2
3
45
Fixed Film Clarifiers
23
7
1
3
34
Disinfection
23
12
2
2
39
Sludge Thickening
30
15
5
5
55
Sludge Digestion
26
15
6
4
51
Sludge Treatment
24
11
4
2
41
-------
is rated as minor periodic.
The primary relationship of concern to this study is that of influent HZS to
corrosion. For preliminary treatment the survey participants were asked to rate the
severity of problems with corrosion and hydrogen sulfide. The matrix of responses
indicates that for any level of corrosion problem most wastewater treatment plants
generally have minor periodic or no troubles with sulfide. It is also noted, however, that
the majority of plants reporting major continuous problems with H2S also report the
same level of severity for corrosion.
The severity of odor releases was also compared to the level of corrosion
problems in preliminary treatment. Of those plants reporting major continuous
corrosion problems, the majority of responses were evenly divided among minor and
major odor problems. Only a few of these plants reported no odor problems.
The matrix of responses for corrosion and odor was similar to that for corrosion
and sulfide. It was found that wastewater txeatment plants seem to recognize a
relationship between odor releases and the presence of hydrogen sulfide. In general,
the same response to the severity of H2S was also given to the problems with odor
releases. Approximately 30% of the wastewater treatment plants surveyed reported no
problems with either.
As part of the selection process for site visits, information was analyzed from 34
cities reported to have hydrogen sulfide corrosion problems. This information is
summarized in Table 2-5.
Other organizations, manufacturers, and contractors were contacted to gain
additional perspective on the national extent of hydrogen sulfide corrosion. These
entities included the Clay Pipe Institute, National Association of Sewer Service
Contractors, Insituform of North America, Spirolite Corporation, Ameron Corporation,
Sauereisen Cement Co., LaFarge Cement Co., and Specialty Sewer Services, Inc., and
others. Highlights of information collected from these sources are briefly summarized
in Table 2-6.
Figure 2-1 is a map pinpointing locations where severe corrosion problems are
judged to exist in the sewer system or treatment plant This is based on EPA site
investigations, surveys conducted by other organizations and the experiences of
professionals active in the field of hydrogen sulfide coirosion control. This does not
represent all the cities experiencing severe corrosion problems.
Figure 2-2 is a map which shows the frequency of use of a proprietary, corrosion-
resistant liner for concrete pipe. This type of liner is specified during design for
concrete pipes which may be subjected to hydrogen sulfide corrosion. The map does
not represent actual corrosion problems.
2-40
-------
TABI-E 2-5
SELEC1EU INFORMATION IRON TNIRTTFOIIH CITIES
N>
I
U
TOIAL
RCP
SEVER
length
NO.
AVC.
IfRCENT
PERCENT
population
LENGTH
>24"
•>F
HjOV
COMBINED
INDUSTRIAL
CITY
(MIL)
-------
TABU 2 5 (coni.)
N)
I
4-
N)
PIPE ACE
RATE
ANNUAL
ANNUAL
AVG
WIIEH AG/S1L
men
INCREASED
Ihi'l EMENIEU3
CORROSION
RAIN
SNOW
6UIIIER
FIRST
CORROSION*
RA1E
SINCE
CORROSION
CONIROI.S
FRECIP
FRECIP
TUIP
CITY
EXPOSED (IK)
LOCATION
CORHOSION
19)0
COHIROLS
EFFECTIVE
(IN)
(IN)
(FAR)
Jeiietaon City, NO
36.0
II.0
IB.6
Milwaukee, VI
C
C
29.0
45.0
69.9
Topeka, KA
20
C
KM
37.0
20.0
78.B
Duluth, HN
n
10 0
71 0
6S.6
Indianapolis, IN
CI
19.0
21.0
75 0
Charlotte, NC
B
N
N
B
Y
410
6 0
71. S
Orlando, FL
BCE
N
B
SI.O
0.0
BIB
Uibana, 1L
15.0
22.0
76.1
El Paao, TX
Ctrl
Y
10
SO
62.3
Sacraaeoto, CA
N
M
ABU
Y
1)0
0.1
IS .2
Albuquerque, Hrt
15
J
Y
M
ABC
V
>7
110
76.7
Baltiaore, Ml)
16
C
N
N
40.0
22.0
76.6
Baton Rouge, LA
IS
ACUEGUI
1
¥
A
57 0
0 2
61.9
Battle Creek, HI
10
BG
I
.
lit
Y
32.0
39 0
72.3
Boiae, ID
10
ACOE
1
N
12.0
210
74.S
Caaper, UK
5
tDEGtl 1
V
N
F
N
11.0
77 0
71.0
Dallaa, TX
10
BlUECHI
N
N
Ut
Y
32.0
3.0
64.9
Denver, CO
10
ACDEFGIII
I
B
Y
16.0
60 0
73.0
Forth Worth, TX
10
El
I
N
B
Y
32.0
3 0
64.9
Honolulu, HI
10
CDECI
Y
N
H
T
23.0
0.0
BO.7
Houatoo, TX
20
CUE Gil
Y
41.0
0 0
B1.4
Jackaonvllle, FL
10
ACUE
I
N
54.0
0.0
SI.O
Naalivillc, TN
BLUE Gill
Y
H
Y
46.0
10 7
79.6
Oaaba, HE
Bf.DECI
1
N
1
Y
30.0
32 0
77.2
fboenii, AK
10
0
N
N
ABUtl
Y
7.0
0 0
91.2
Fine Blull AK
10
CU
V
V
J
Y
49.0
SO
• 1.4
Raleigh, HC
IS
G
N
1
AM
Y
43.0
7 0
77.5
St. Lou la, HO
10
BDGIIt
H
N
D
Y
36.0
IB 0
76.6
San Aotonlo, TX
IS
ACUEH
N
21.0
0 5
64.7
San Diego, CA
e
ACUECIII
V
N
H
9.0
0 0
71.4
Seattle, UA
IS
ACDEFGHI
N
H
B
39.0
15.0
64.5
Taapa, FL
20
ACDEGI
N
N
J
N
49 0
0.0
62 2
Tuicon, AR
20
BCUEHI
N
N
N
II.0
2.0
66.3
Virjlnla Beacb, VA
4
ACIIECHI
V
N
AHJ1
45.0
7 0
76.3
0
HI
uir
FAN)
AVG
WW
TEMP
SS
70
70
67
60
73
55
1
I N-no, N=alnor, f=|>nt, Y=current aajor pcoblea
* A-tbiou|l>out ayatea, B-ouly apeciiic locatlona, C-allcr loice ailni, U=»l aanliolea, i-it (low tianaltion aliucluiea, F-aaaociated witb induatrlal
diacbar|ea, G-1 n large diaaeter pl|>e, "''n aaall diaaeter pipe, l-long detention tiaca, J=otbcr.
1 A=CI injectioo, B=ll 0^ Injection, C=0 injectiou, D=alr injection, E-KIINO injection, F=NaUII iojcclion, G-NaOcI injection, lUFeSO injection,
1-leCi. injection, J-unapecifled cbcalcal injection, K-CAC illter, L=aaiiitlin aiuiatia pipe alope, H=u«e corcoaion icaiatant aaterial (viltilied
clay, 6t PVC)
-------
TABLE 2-6
SUMMARY OF INFORMATION FROM SELECTED ASSOCIATIONS,
MANUFACTURERS,. AND CONTRACTORS
Source
Clay Pipe Institute
Spirolite Corporation
Insituform of North America
Comments
From 1950 to 1969, over 154 miles of
severely corroded concrete pipe in 57
cities was replaced with clay pipe, with
over 5% of production tonnage used
for this purpose
From 1986 to 1989, approx. 20 miles
of sewer was sliplined using Spirolite
polyethylene pipe
Over 100 miles of large diameter
sewer was lined using cured-in-place
inversion lining in more than 36 U.S.
cities between 1977 and 1988
Ameron Corporation From 1947 to 1988, over 900 miles of
sewer pipe in 500 U.S. projects was
specified with T-lock liners to prevent
crown corrosion
2-43
-------
FIGURE 2-1
STATES HAVING SEVERE CORROSION PROBLEMS IN
WASTEWATER SYSTEMS OF FOUR OR MORE MUNICIPALITIES
N>
I
4±
-------
FIGURE 2-2
USE OF PROPRIETARY PVC LINING TO PREVENT
CORROSION OF CONCRETE PIPE
.49
2ZZ
H OF CITIES
11-20
> 20
-------
2.6 Case Studies
To gain further insight into the causes and prediction of hydrogen sulfide
generation and subsequent hydrogen sulfide corrosion, case histories were prepared
from previous evaluations of five wastewater collection systems located in three cities.
The following systems were investigated:
• Sacramento County, California, Central Trunk Sewer
• Sacramento, California, Regional Interceptor System
• City of Lakeland, Florida, Western Trunk Sewer
• City of Omaha, Nebraska, Papillion Creek Wastewater System
• City of Omaha, Nebraska, South Interceptor Sewer
A brief summary of each of the five case studies is provided below (2)(3):
2.6.1 Sacramento County, California, Central Truck Sewer
2.6.1.1 Description and History
The Central Trunk Sewer conveys both domestic and industrial wastewater and,
for several years, conveyed sludge from two upstream wastewater treatment plants. The
trunk is approximately 16 miles long. About 2 miles of the upper reach is vitrified clay
pipe ranging from 18 to 24 inches in diameter. The remaining 14 miles is granitic-
aggregate reinforced concrete pipe 27 to 60 inches in diameter. The slope varies from
0.18 percent in the upper reaches to 0.05 percent in the lower end of the trunk.
The Central Trunk Sewer was constructed in 1962. The following is a brief
history of the system:
• 1964—The entire line was visually inspected by County personnel by
floating on rafts through each reach. A powder deposit on the inside of
the pipe was the only evidence of corrosion.
• 1968-AnaerobicalIy digested sludge discharge to the Central Truck Sewer
was initiated.
• 1968—The entire pipeline was again visually inspected and samples of
concrete were taken. The pH of the concrete walls was greater than 3.0.
Maximum corrosion at that time was estimated to be 1/4 inch, and no
coarse aggregate was exposed.
• 1969--Dr. Richard Pomeroy completed an analysis of portions of the
trunk. Existing corrosion was estimated to be 1/8 or 1/4 inch. The useful
life of the trunk was estimated to be 100 years.
2-46
-------
• 1972-The County studied the effects of chlorine on sulfide generation in
the Central Trunk system.
• 1972-1973—Vitrified clay plugs-were installed at the manholes in the trunk.
Three sulfide control facilities were designed and one was bid. Due to the
high price of the bid, the sulfide control facilities were not constructed.
• 1976—The clay plug locations were inspected. Significant corrosion was
evident, and a hydrogen sulfide coiTosion study was initiated in September
1976.
• 1979—Three chlorine injection stations were placed in service along the
Centra] Trunk to reduce sulfide levels and to minimize further coiTosion.
• 1982-83-Sludge discharge to the Central Trunk was stopped.
• 1984—The County completed another field investigation which showed
continued corrosion, but at a reduced rate.
• 1984—A 1,800-foot reach of 60-inch-diameter pipe from the Central Trunk
Sewer was abandoned primarily due to a new alignment, and partially due
to concern about the structural integrity of the pipe.
• 1987—County inspection of the Central Trunk showed reduced sulfide
concentrations and indications of reduced corrosion rates.
2.6.1.2 Summary of Results
Average annual total sulfide concentrations ranged from approximately 0.5 to 1.5
mg/1 from 1965 to 1976, based on weekly or monthly sampling at 12 noon at one
location. Average 24-hour sulfide concentrations were approximately 40 percent higher.
Corrosion penetration was as much as 1.5 inches at some locations in the sewer.
Cores taken from the crowns of the pipes showed that the worst corrosion conditions
usually existed in the first 30 ft downstream of the manholes.
Predicted sulfide concentrations from the Pomeroy-Parkhurst equation were
compared with measured values. When the effect of sewer junctions was considered,
the predictive equation estimated sulfide concentrations with reasonable accuracy.
Predicted corrosion penetration was also compared with measured values, and showed
an excellent correlation. It was also found that peaking factors of 2 were justified to
account for minor turbulence at manholes. Higher peaking factors would be required
where high turbulence levels were encountered to account for the increase in off-gassing
of H2S.
2-47
-------
Dead load testing was conducted on a 12 ft section of 60 inch diameter pipe in
1984. The pipe was highly corroded, but reinforcing steel was not exposed. Tests
showed that, although concrete loss had occurred, the pipe strength was still significantly
greater than the original specification for dead load.
Based on inspections in 1987, the rate of corrosion was reduced from that
observed in the 1970's due to elimination of sludge discharges and installation of
chlorination stations.
2.6.1.3 Findings and Conclusions
1. The observed corrosion information from 1976 is in general agreement with
Pomeroy's corrosion predictive equation, if conservative assumptions are made
and input data are based on field measurements and monitoring data.
2. Valid, positive measurements of the depth of actual corrosioa of in-place pipe for
the Central Trunk were difficult Accuracy of all methods used was less than
desirable.
3. The discharge of anaerobically digested sludge to the Central Trunk for 14 years
was partly responsible for the higher corrosion rates over those years. Sludge
discharges increased the BOD and the temperature of the wastewater in the
Central Trunk.
4. With removal of the sludge discharges and installation of chlorination stations
along the Central Trunk Sewer, the sulfide concentration has been reduced in the
1980's, and the rate of conosion appears to have dropped significantly.
5. The Central Trunk is believed to be made from spun RCP. It seemed to take
several years before corrosion penetrated the surface layer of this gTanitic
aggregate pipe. Once the high-alkalinity surface layer had been con-oded, the
corrosion rate increased, since the alkalinity of the rest of the pipe concrete was
only 16 percent
6. The corrosion information on manholes and structures with turbulent flow
characteristics is particularly interesting. This information points out the need to
use conservative corrosion rate peaking factors in predicting hydrogen sulfide
corrosion rates within close proximity to these locations (within several pipe
diameters).
2-48
-------
2.6.2 Sacramento, California Regional Interceptor System
2.6.2.1 Description and History
The Sacramento terrain is quite flat, and the climate features hot summers and
mild winters. The collection system extends 20 miles east of the Sacramento River, and
for more than 20 miles north to south. The longest interceptor, nearly 30 miles long,
drops less than 200 feet from end to end. Several pumping stations are included in the
system.
*, The Regional Interceptor System is extensive in scope, with a total capital cost of
about $143 million. It was constructed during the period 1975 to 1982, and
encompasses about 62 miles of gravity sewer, over 25 miles being pipe in the 60- to 120-
inch diameter range.
In designing the regional interceptors, a number of passive measures were
implemented:
1. County Source Control Ordinance. Sacramento County passed an ordinance in
1977 controlling the quality of industrial waste discharges into the Regional
System.
2. Calcareous Ag^re^ate. Use of sacrificial calcareous (rather than granitic)
aggregate was specified for all concrete pipe construction.
3. Turbulence Control. Junction structures were designed for smooth transitions to
minimize wastewater turbulence.
4. Lining of Hydraulic Structures. Junction structures, and other hydraulic elements
especially vulnerable to corrosion, were lined with plastic (locking PVC liners).
5. Slope/Velocity. Pipe slope and velocity were carefully evaluated to limit solids
accumulations within the system.
An extensive study to determine needed sulfide controls commenced in 1974 and
was completed in 1976. Several sulfide control measures appeared feasible from this
study for use in the Regional Interceptor System. Except for chlorination, no reliable
performance data existed prior to the mid-1970's for the control measures being
considered. To develop this information, a field testing program of several sulfide
control measures was undertaken in the summer of 1974. Information was obtained for
the following:
1. Air injection in a force main (injected at pumping station)
2. ' Hydrogen peroxide injection in a force main
2-49
-------
3. Oxygen injection in a force main
4. Chlorination in a force main and gravity sewer
5. U-tube aeration (air) at the end of a force main
6. U-tube aeration (oxygen) at the end of a force main
For upstream pumping stations, chlorine was selected to provide control of
hydrogen sulfide generation. However, for a 60-inch diameter and a 72-inch diameter
force main, injection of high purity oxygen into a fall structure was selected as the most
cost-effective alternative. The injection of high purity oxygen enriches the atmosphere
of the fall structure with oxygen to allow entrainment of oxygen through turbulence
induced by the fall, and elevation of dissolved oxygen levels in the wastewater, thus
allowing oxidation of existing sulfide and prevention of further sulfide generation.
Attention was also focused on the potential problems of solids deposits,
especially during the early years of system operation. Since solids deposits can generate
sulfide, the objective of the design was to eliminate the possibility of solids
accumulation, or otherwise plan for removal of any such deposition.
The predicted lack of adequate velocities during early years of operation in
several interceptors led the designers to a decision to construct flushing facilities at two
locations. In this manner, flows could be greatly increased for a few hours at a time
during the long periods of low dry weather flows, flushing solids down the system of
interceptors.
The Pomeroy-Parkhurst sulfide prediction equations and the Pomeroy conosion
prediction equation were used extensively during design of all regional interceptors.
Many different assumptions were used to determine the best mix of passive and active
sulfide controls to provide assurance of long pipe life (100-year minimum) and lack of
odor problems.
2.6.2.2 Summary of Results
Chlorine gas is used at upstream pumping stations and at three locations along
the Central Trunk. The chlorination stations have been successful in maintaining low
dissolved sulfide levels (generally less than 0.3 mg/1) at the force main discharge points.
Solids deposition in the interceptor system has not been a problem. Peak wet
weather flows have provided adequate flushing of the lines, and use of the flushing
stations has not been required in the northern portion of the system.
Two oxygen injection stations were constructed and operated at fall structures
(4). In one system, DO concentrations of 13 mg/1 have been achieved in the wastewater
at a point nearly 4000 ft downstream of the fall structure. Similar results have been
reported for the other oxygen injection system. Dissolved sulfide concentrations
2-50
-------
downstream of the fall structures are consistently less than 0.5 mg/1.
The primary problem with these stations has been that hydrocarbon compounds
are stripped from the wastewater under the turbulent conditions found in these
structures. Explosion meters that monitor the Lower Explosive Limit (LEL) are used in
each station. When 25 percent LEL is reached, pure oxygen addition is shut down.
This shutdown event occurs often. In the summer months, 25 percent LEL is reached
on almost a continuous basis. Investigation by the District into the source of the
hydrocarbons has shown that it may be methane gas generated under the anaerobic
conditions in the upstream force mains and gravity sewers, and within the combined
sewer system of the City of Sacramento.
Vitrified clay plugs were installed at many locations in the pipe of the Regional
Interceptor System to allow future measurement of the corTosion of the pipe
surrounding the plug. The District has had difficulty obtaining measurements from
these plugs due to a variety of problems including: problems in finding the plugs in the
dark, uncomfortable environment of the sewers, and, if found, difficulty achieving highly
accurate measurements due to the conditions. Collected data have been analyzed and,
in general, indicate that corrosion to date is very low.
2.6.13 Findings and Conclusions
1. Although only a few years of monitoring data are available, the information
suggests that the Regional Interceptor System design, and its sulfide control
systems, are providing the level of protection anticipated.
2. District monitoring work has confirmed the need for PVC, or other type of non-
conrodible lining material, for all junctions and structures where even limited
turbulence occurs and where wastewater contains minor amounts of dissolved
sulfide.
3. The lack of accurate measurements on the amount of corrosion in the early years
of operation of the Regional Interceptor System has been frustrating. The level
of accuracy needed is in the range of hundredths of an inch. The level of
accuracy for corrosion measurements needs to be improved.
4. The stripping of hydrocarbon compounds in the fall structures has caused
additional safety considerations that were not initially anticipated during design.
In the summer, one of the fall structures is now bypassed, and oxygen is injected
in the bypass. This results in satisfactory oxidation of sulfide at the ends of the
two long, large-diameter force mains.
5. Solids deposits have not been a problem, and do not appear to produce any
significant sulfide.
2-51
-------
6. Chlorination in upstream force mains has performed satisfactorily to minimize
sulfide at the discharge points of these force mains.
7. Data on wastewater sulfide concentrations and sewer atmospheric H2S levels have
been difficult to correlate at specific locations in the Regional Interceptor —
System. This is probably due to analytical inaccuracy at low dissolved sulfide
concentrations (typically less than 0.3 mg/1), slight wastewater pH variations, and
the degree of wastewater turbulence, all of which are critical to H2S off-gassing
rates.
8. The operating data from the Regional Interceptor System show that careful and
conservative use of the sulfide and corrosion predictive equations can be of major
assistance in designing long interceptor systems in warm climates to meet
stringent corrosion standards.
2.6.3 City of Lakeland, Florida, Western Trunk Sewer
2.6.3.1 Description and History
The City of Lakeland is a growing community of approximately 66,000
population. Lakeland has experienced corrosion problems in portions of its sewer
system due to sulfide generation. An engineering study was undertaken in early 1988 to
assess the existing conditions and to develop a plan to renovate portions of the Western
Trunk Sewer.
The Western Trunk Sewer receives wastewater discharges from food processing,
other industrial, commercial, and residential areas. The collection system consists of
both force mains and gravity sewers.
The gravity portion of the Western Trunk consists of about 27,300 lineal feet
(LF) of primarily reinforced concrete pipe (RCP) and vitrified clay pipe (VCP) ranging
in size from 24- to 48-inch diameter. There are variable slopes on most reaches which
cause changes in velocity. Most dry weather velocities are greater than 2 fps, and some
reaches have velocities of 7 fps.
The Western Trunk Sewer was constructed in 1960 and 1961. Lakeland has
rehabilitated or replaced portions of this sewer in recent years because of pipe collapses.
An odor control study conducted by the City of Lakeland in 1987 confirmed high levels
of H2S gas and high wastewater sulfide concentrations. The City undertook a sewer
system evaluation study to determine the extent of hydrogen sulfide corrosion and
appropriate solutions to correct the deficiencies.
2-52
-------
2.6J.2 Sewer System Corrosion Evaluation
The corrosion evaluation consisted of detailed inspection and analysis of the
conditions of both manholes and truck sewer. A total of 79 manholes were inspected.
Internal television inspections were conducted on nearly all the gravity portions of the
trunk sewer. Core borings of concrete pipe were taken at eight representative locations
along the trunk.
In general, the manholes were found to be coiToded and in need of
rehabilitation. Seventy-seven manholes had coiToded barrels, and 66 had deteriorated
frames.
In the trunk sewer system, all reaches of reinforced concrete pipe (RCP) had lost
from one to four inches of concrete due to corrosion. Reinforcing steel was found
exposed or missing in numerous locations, and aggregate was exposed in all reaches of
RCP. A section of 30-inch diameter ductile iron pipe, installed in 1966, was severely
blistered and brittle enough to break by hand. While the vitrified clay pipes did not
suffer from corrosion damage, there were numerous cracks and leaking joints in the
lower portion of that segment; aad manhole corrosion was worse in the VCP reach.
Predictive models were utilized to estimate sulfide generation, corrosion rates,
and remaining useful lifetimes of pipes. The model predicted sulfide build-up of 1.5 to
2.0 mg/I. Recent field data show levels of 1.0 to 1.5 mg/1. Approximate correlation was
shown by the model. Similarly, predicted corrosion rates approximated estimated
corrosion rates based on field measurements. Predicted average corrosion rates were
0.03.to 0.15 inches/year, with peak rates approximately double these values. It was
estimated that, at turbulent structures, corrosion rates could be five times predicted
average rates. This underscores the importance of turbulence on hydrogen sulfide
corrosion, and the difficulty in estimating corrosion rate based on sulfide levels.
Based on existing depth of corrosion and estimated corrosion rates, forty-six
percent of the RCP was determined to have no remaining useful life, and 54 percent
was determined to have a useful life of 1 to 8 years.
Alternatives were analyzed for rehabilitation of the corroded pipe. The following
five methods of rehabilitation were considered feasible for the Western Trunk sewer:
1. Slip lining with high density polyethylene pipe (with fusion joints, or bell and
spigot joints) or with fiberglass pipe (filament wound, or centrifugally cast).
2. Inversion lining with polyester, resin-impregnated fabric,
3. Removal of existing pipe and replacement with one of the following:
1) reinforced concrete pipe with PVC liner, 2) fiberglass pipe (centrifugally cast)
2-53
-------
designed for direct burial.
4. Parallel replacement of pipe using the same two options listed above.
5. Chemical grouting for specific locations (limited on this project to grouting
structurally sound vitrified clay pipe for infiltration control).
2.633 Findings and Conclusions
1. Conditions which contribute to the high corrosion rates in the Western Trunk
Sewer include at least the following: high wastewater BOD, high soluble BOD
fractions, high wastewater temperature, low wastewater pH, variable slopes,
formation of deposits, use of drop structures, and existence of upstream force
mains.
2. The Pomeroy-Parkhurst sulfide predictive equation and the Pomeroy corrosion
prediction equations were used with a series of conservative assumptions for the
Western Trunk evaluation. In doing so, and by assuming that historical system
flows, characteristics and operation were similar to current situations, the
"modeled" corrosion approximated the actual corrosion. This tends to indicate
the quantity of sulfide and corrosion expected in similar situations. However, it
also points out that the equations should be used cautiously and with substantial
conservative assumptions and safety factors.
3. The velocities in the Western Trunk system are insufficient, in some reaches, to
transport the grit and heavier solids. Additional sulfide production can be
expected in these reaches due to the ability of high soluble organic and sulfate
concentrations to penetrate into these deposits.
4. The high likelihood of substantial sulfide generation in the Western Trunk in the
future was judged to preclude the use of corrodible materials, or at least
minimize use of these materials, in the rehabilitation/replacement of the Western
Trunk Sewer.
2.6.4 Omaha Nebraska, Papillion Creek Wastewater System
2.6.4.1 Description and History
Corrosion and odor problems have occurred in Omaha's Papillion Creek
Wastewater System over the past decade. The start-up of the expanded Papillion Creek
Interceptor System in the mid-1970's brought new dischargers into the system, and
substantially increased the transit time of wastewater to reach treatment facilities. The
new Papillion Creek Wastewater Treatment Plant has experienced corrosion and odor
problems which are partly related to interceptor sulfide problems. Safety has been an
2-54
-------
additional problem due to high concentrations of hydrogen sulfide in the confined
spaces of interceptors and treatment facilities.
Omaha's Papillion Creek Wastewater System has evolved over several decades of
growth and urban expansion. It now includes a service area bringing wastewater flows
over 25 miles to the Papillion Creek Wastewater Treatment Plant (Papio Plant).
Only the largest Papillion Creek interceptors were evaluated in the 1984/85
corrosion and sulfide study because these were the interceptors suspected of corrosion
damage. The interceptor downstream from the old Papio Plant was a primary target of
the study. These interceptors were put into service in the mid-1970's to transport raw
wastewater to the new Papio Plant This system of interceptors allowed several
treahnent plants in the Papillion Creek drainage area to be abandoned.
Ill 1971, a limited study was completed on potential sulfide and corrosion
problems in this interceptor system, which was then being designed. This report made
predictions of sulfide levels in the interceptors and confirmed the need for lining the
sewer with either sacrificial concrete or plastic.
Between 1973 and the start-up of the new Papio Plant in August, 1977,
communities and industries were allowed to discharge treated effluent and, in some
cases, raw wastewater to the new Papio Interceptor System. Many odor complaints
from residents living near manholes and structures were received during this period.
The worst odor conditions occurred in the last five miles of the system, where high
BOD wastes and flows of only a few mgd probably caused very high sulfide production
in at least 1974 and 1975.
Significant corrosion of the unlined outfall portion (last 3,200 feet to the
Missouri River) was noted in 1977. This corrosion was evidently a result of these raw
and partially treated flows. The unlined outfall portion of the interceptor was estimated
to have 1/2-inch of corrosion in August 1977. Extensive corrosion of unlined manhole
risers in the lower interceptor reaches also occurred prior to 1977, to the extent that
some risers were replaced during the mid-1970's.
Odor problems became a major issue in Sarpy County, and the City of Omaha
decided to seal all manhole covers from the old Papio Plant south to the new Papio
Plant in 1975. Little ventilation of the system can occur, and only one siphon has
ventilation stacks. The West Branch has little ventilation since manhole covers are
solid.
Data collected at the Papio Plant show influent total sulfide levels of 4 to 5 mg/1
from July through February, dropping to between 1 and 3 mg/1 during spring months.
2-55
-------
2.6.4.2 Summary of Results
Inspection of the interceptor was conducted in 1984. Manhole risers were badly
corroded, in some cases, to over 1 inch. The risers were constructed of concrete with
non-calcareous aggregate. Measured pipe cotTosion in calcareous aggregate reaches
ranged from zero to 0.5 inches. At points of high turbulence such as discharge
structures, up to 0.75 inches of calcareous/granitic aggregate concrete had been lost over
a 10 year period.
Use of predictive models was unsuccessful in that actual sulfide concentrations at
the Papio plant were double those predicted by the model. The reasons are judged to
be 1) high fraction of soluble BOD, 2) low oxygen content in the sewer atmosphere due
to sealing of manhole lids, 3) need for more conservative coefficients in the predictive
equation.
The Papio plant, commissioned in 1977, was also inspected for corrosion in 1984.
CotTosion at the headworks was significant, with up to 0.75 inches of concrete lost over
a 6-1/2 year period. Several locations within the plant had little or no concrete cover
remaining. Reinforcing steel was exposed in the sludge storage tank, decant tank, drain
manholes, and trickling filter walls. Hydrogen sulfide corrosion problems were caused by
three basic factors:
1. High sulfide concentrations and low pH of influent wastewater
2. Long storage times for sludge and recycle streams
3. Recycle of streams from anaerobic digestion process
2.6.4.3 Findings and Conclusions
1. The large Papillion Creek interceptors were designed with attention to sulfide
and corrosion issues, and based on 1984 inspection work, most of the interceptor
system was in good condition. Use of sacrificial concrete, coatings, PVC linings,
and calcareous aggregate prevented serious corrosion problems in the interceptor
system.
2. There have been a few specific corrosion problems in the interceptor system.
These include metal gates and aluminum grating which was subject to destructive
acid attack. H2S off-gassing from highly turbulent wastewater in certain
structures has caused corrosion rates that are largely unpredictable. Unprotected
manholes have been subject to extensive corrosion.
3. If structural rehabilitation work is required on any of the primary interceptor
structures, costs could be extensive for diversion and rerouting of raw wastewater
2-56
-------
flows.
4. Assessment of concrete corrosion rates over time is difficult without historical
sulfide data and/or historical inspection information identifying the depth of
corrosion from the original surface.
5. Actual sulfide buildup in the interceptor system seems to be at least twice that
predicted by the Pomeroy-Parkhurst equation.
6. s.- Actual concrete corrosion rates are greater than predicted in the design for
portions of the interceptor system due to: 1) greater sulfide production than
planned, and 2) reduction in pH of wastewater as it travels down the long
" anaerobic interceptor system.
7. Concrete, metal, and instrumentation corrosion in the Papio Plant occurred at
many locations over the period of 1977 to the mid-1980's. The corrosion was
caused by high wastewater sulfide levels, and high H2S off-gassing rates. High
dissolved sulfide levels in the wastewater were partially caused by high influent
sulfide concentrations, and partially by sulfide produced within the plant
8. Solutions to plant hydrogen sulfide corrosion problems are now mostly in-place.
These solutions encompassed rehabilitation, process changes, chemical addition,
and ventilation improvements. The solutions implemented thus far appear to be
working satisfactorily.
9. There remains an odor problem at the Papio Plant caused in large part by the
high influent sulfide levels. High sulfide production in the interceptor system
and high plant influent sulfide concentrations constitute the most significant
unresolved sulfide issues in the Papillion Creek Wastewater System.
2.6.5 Omaha, Nebraska, South Interceptor Sewer '
2.6.5.1 Description and History
In the late 1950's, planning was initiated for collection, diversion, and treatment
of raw waste discharges to the Missouri River. By 1965, the system of diversion
structures, interceptors, pumping stations, and primary treatment facilities had been
constructed and placed in operation. The system involves a series of structures along
the west bank of the Missouri River to intercept flows and pump them to the Missouri
River Wastewater Treatment Plant (MRWTP). The area served by the MRWTP is the
older and more highly developed portion of Omaha. It contains Omaha's central
business district and industrial centers which are located adjacent to, or near, the
Missouri Rjver.
2-57
-------
The South Interceptor Sewer (SIS) is a 4-1/2 mile long force main that brings the
majority of the plant flows to the MRWTP. Flow from the SIS is discharged at the
North Inlet Dry weather flow in the winter is about 20 million gallons per day (mgd)
from the SIS. This flow increases in the warmer months due to infiltration, runoff from
lawn watering, and discharges to the sewer system from drawdown of Carter Lake.
Velocities in the 66-inch diameter South Interceptor Sewer are about 0.7 foot per
second (fps) or less at night and typically average 1.2 to 1.6 fps in dry weather.
Deposition of solids is no doubt occurring under these conditions. Dissolved sulfide
levels in the plant influent are typically 3 to 6 mg/1 during warm months.
Additions to the MRWTP completed in 1980 included covering open tankage at
the North Inlet, as well as other locations throughout the plant Fans were installed to
exhaust foul air from under most of these covered areas and scrub it in chemical mist
units prior to discharge to the atmosphere. The City has not found the chemical
scrubber at the North Inlet to be effective, and does not use it Hypochlorites and
permanganate solutions were attempted in the scrubber with little success. The covers
have remained in place, but the ventilation system is not used, since very high HZS
concentrations would be discharged to the atmosphere. Corrosion of the concrete is
occurring faster since installation of the covers, due to high levels of hydrogen sulfide in
the atmosphere of the tanks and channels of the North Inlet Gaseous hydrogen sulfide
levels under the North Inlet covers have reached the 50 to 300 ppm range during
summer and fall months.
In 1984, a study was initiated to evaluate odor and corrosion problems in the
South Interceptor Sewer and at the Missouri River Wastewater Treatment Plant The
results are summarized below.
2.6.5.2 Summary of Results
Hydrogen sulfide corrosion has occurred at a number of locations in the
MRWTP, but the North Inlet is the location where the problem is most apparent at this
time. Besides the North Inlet, concrete corrosion has occurred at the South Inlet and in
various wastewater channels and boxes prior to, and following, primary clarification.
Corrosion in foul air ducts from the biological filters has also occurred, as well as
corrosion in the anaerobic digestion portion of the plant
Inspection work in 1984 showed that concrete spalling was occurring above the
water line in the Parshall flume area and in other channels. The worst corrosion was at
the drop structure where at least 1/2- to 3/4-inch of concrete was estimated to have
become conoded from walls and structural support members. Most of this corrosion
probably occurred in the 1980 to 1984 period after the tanks were covered. No
reinforcing bars were exposed. Metal corrosion at gates was also significant In the bar
screen building, continued corrosive environmental conditions had deteriorated many
2-58
-------
metal components, and the exposed electrical and instrumentation equipment had
suffered irreparable damage.
The recommended approach to sulfide control at North Inlet was as follows:
1. Reduce the high sulfide and H2S concentrations through sulfide control methods
in the South Interceptor Sewer.
2. Leave the covers in place at the North Inlet and implement improved ventilation
^ and odor scrubbing to reduce corrosive atmospheres and treat residual H2S and
other odorants more reliably.
3. Once improved ventilation and H2S control is in place, conduct structural
rehabilitation and replace equipment as necessary.
Alternatives evaluated for the SIS included chlorination, iron chloride addition
upstream from the plant, and caustic slugging. Costs favored caustic slugging, although
the performance was not expected to be as consistent as other alternatives. Full-scale
testing of caustic slugging was undertaken in late 1984 to confirm its effectiveness and
define costs more accurately.
The objectives of a caustic injection program is to inactivate sulfide-producing
bacteria which grow in the slime layer on the walls of the pipe. High pH (12 to 12.5
and above) is toxic to these bacteria, and interim application of a strong alkaline
solution has proven effective in depressing sulfide production within sewers. The test
program was successful, and the technique has been used on the SIS each summer since
that time. Caustic slugging is effective for reducing dissolved sulfide levels in the plant
influent from over 3 mg/1 to an average of 0.4 mg/1. In 1985 the City spent $44,000 on
17 truckloads of caustic, estimated to eliminate 47,000 pounds of sulfide. In the
summer of 1988, the effectiveness of caustic slugging was reduced. This was believed to
be due to the dry weather conditions which led to higher wastewater strengths, and
lower velocities and greater solids deposition in the SIS. Evidently, the caustic slug is
unable to penetrate sludge deposits and inactivate the bacteria.
The City is currently evaluating alternatives to caustic slugging for control of
sulfide generation in the South Interceptor Sewer.
2.6.53 Findings and Conclusions
1. Significant corrosion and odor problems have occurred at the North Inlet, due to
sulfide production within the SIS. Attempts to control the odor problem in 1980
by covering tankage and scrubbing the foul air resulted in worsening of the
corrosion problem due to failure of the odor scrubbing system, and subsequent
shutdown of the ventilation system.
2-59
-------
2. Rapid rates of corrosion can be expected for concrete and metal exposed to the
high H2S concentrations (50-300 ppm) and moist environment which exist
beneath the covers at the North Inlet The life of some of these facilities appears
to be 10 years or less under the severe conditions which have existed in the
1980's. The high H2S concentrations constitute a safety hazard.
3. Caustic slugging has provided a cost-effective reduction iQ high sulfide
concentrations from the SIS since late 1984 when it was initiated. Its
performance is variable, primarily because of the solids deposition problem in the
SIS, caused by low wastewater velocities. Atmospheric H:S levels beneath the
primary clarifier domes have dropped from over 50 ppm to below 10 ppm.
4. Other sulfide control methods are likely to be more reliable and have better
overall performance than caustic slugging; however, the costs for other control
methods appear to be substantially higher than caustic slugging.
5. The low velocity and resulting deposition problem in the SIS is likely to continue
to plague attempts at reliable sulfide control in this pipeline. Velocities of about
4 to 5 fps are needed on a regular basis to scour grit deposits from a force main
of this size. The periodic scouring of the SIS causes considerable solids loading
fluctuations to the MRWTP which may affect treatment performance.
6. The City, with the use of caustic slugging, has implemented the first phase of a
program to solve the sulfide problems at the North Inlet Substantially more
work is needed to restore the North Inlet to a sound, safe, and acceptable
condition.
2.7 Hydrogen Sulfide Corrosion In Other Countries
Hydrogen sulfide corrosion has been reported in the literature of many countries
including France, Germany, Italy, United Kingdom, the Netherlands, Denmark,
Czechoslovakia, Iraq, India, China, the Soviet Union, Japan, Saudi Arabia, Kuwait
Egypt South Africa, Venezuela, Brazil and Australia. TTiis observation is based on a
series of literature searches conducted for EPA in 19S2 and 1988 on the subjects of
odor and corrosion in wastewater systems.
Several severe cases of hydrogen sulfide corrosion are brieflv summarized below
(5)(6):
1. Venezuela: Reinforced concrete pipe was corroded to a depth of 2.8 inches
within eighteen months of construction in an area downstream from the
discharge of a force main. Vitrified clay pipes used in the same system were not
affected.
2-60
-------
2. Cairo, Egypt: As early as 1920, hydrogen sulfide corrosion of sewer pipes was
recognized as a problem in the Cairo system. By 1922, in spite of aeration and
regular flushing, the crown of the original main outfall sewer, 60 inches in
diameter and made of local cement concrete, was corroded to a depth of 3.9
inches over a length of 8 miles. By 1930, the depth of corrosion had increased to
5.9 inches, nearly half the thickness of the pipe.
3. Baghdad, Iraq: A system of reinforced concrete interceptors was put in
operation in 1963, and by 1977 the maximum depth of corrosion in extended
sections of the interceptors varied from three to four inches at the crown where
reinforcing steel, which was designed to be protected by a two-inch concrete
cover, was exposed and had corroded away in places. The walls were also
damaged to a depth of two inches and the access manholes showed severe
corrosion. The rate of internal corrosion was estimated at 0.3 inches per year,
which would have led to a critical situation in approximately 15 years.
Rehabilitation had begun in certain sections of the system, and new extensions
have been made with PVC or fiberglass-reinforced plastic lining.
4. Singapore: The sewer of concern was 71 inch diameter concrete pipe internally
lined with 0.5 inches of high alumina cement mortar. The line was over 1600 ft
long with a design capacity of 81 mgd. The sewer was commissioned in 1961.
This portion of the line received almost entirely pumped sewage, much of which
was pumped more than once.
Inspections since 1970 revealed extensive and continuing hydrogen sulfide
corrosion. The high alumina lining was completely corroded away or reduced to
a soft past above the waterline. Corrosion in some areas had proceeded to a
depth of over 1.5 inches, and had gone beyond the reinforcing steel, in less than
15 years.
Several rehabilitation options were under consideration including sliplining with
glass-reinforced polyester or high-density polyethylene, installation of coiTOsion-
resistant panel liners, and cured-in-place inversion liners.
2.8 Conclusions
Attempts to gain a thorough understanding of the severity and extent of
hydrogen sulfide corrosion problems in U.S. were thwarted by the lack of historical data
on sewer corrosion, the lack of a standardized technique to measure corrosion, and the
poor documentation by municipalities of sewer corrosion and the expenditures for sewer
rehabilitation or replacement Upon review of information gathered, the following
findings and conclusions are presented:
• Severe hydrogen sulfide coirosion problems are not limited to CSDLAC.
2-61
-------
Extensive corrosion damage requiring immediate repair or rehabilitation
has been observed in sewers in other cities. In some cases, corrosion
damage is so extensive as to compromise the structural integrity of the
pipe, which could lead to collapse.
• Hydrogen sulfide corrosion problems in operating systems are often not
recognized early enough to take corrective action before considerable
damage has occurred.
• In a 1984 survey, approximately 30 percent of the 89 cities reported sewer
collapses that were judged to be due to hydrogen sulfide corrosion.
• In two independent surveys, 60 to 70 percent of the municipalities
reported hydrogeo sulfide corrosion at their wastewater treatment plants.
Of those plants experiencing corrosion, about 20 percent are considered to
have severe problems.
• Hydrogen sulfide corrosion problems have been documented in the
literature of at least 20 foreign countries.
• Due to lack of historical data, corrosion rate is estimated based on depth
of corrosion and age of pipe. This may not reflect the true corrosion rate,
which may be substantially higher at a given time and condition.
• No entities other than CSDLAC had sufficient data on corrosion rate to
establish whether the rate of corrosion had changed over time.
• Due to changing alkalinity in spun vs. cast concrete pipes, corrosion rate
can change over time.
Evidence of severe corrosion may be found in cities throughout the United States
and other countries. Cases of "high rate" corrosion are also common. However, at this
time EPA has been unable to document other cases of "accelerated" corrosion of the
type that has been experienced in the sewers of CSDLAC.
2-62
-------
REFERENCES
1. Jin, Calvin, "Sulfide Control with Sodium Hydroxide in Large Diameter Sewers,"
internal report, CSDLAC, March 1987.
2. Witzgall, R.A., Horner, I.S., and P.L. Schafer, "Sulfide Coirosion in Collection
and Treatment Systems - Case Histories and Evaluation of Predictive Equations,"
prepared for the 62nd Annual WPCF Conference, San Francisco, CA, October,
19, 1989.
3. Witzgall, R.A., Homer, I.S., and P.L Schafer, "Case Histories of Sulfide
Corrosion: Problems and Treatment," Water Environment and Technology,
Water Pollution Control Federation, 42-47, July, 1990.
4. EPA, "Odor and Corrosion Control in Sanitary Sewerage Systems and Treatment
Plants," EPA/625/1-85/018, Cincinnati, OH, October, 1985.
5. Prevost, R.C., "Conrosion Protection of Pipelines Conveying Water and
Wastewater - Guidelines," World Bank Technical Paper No. 69, Water Supply
Operations Management Series, The World Bank, Washington, D.C., 1987.
6. Nadarajah, A, and J. Richardson, "Prevention and Protection of Sewerage
Systems Against Sulphide Attack with Reference to Experience in Singapore,"
Prog. WaL Tech. Vol 9: 585-598, Pergamon Press, Great Britain, 1977.
2-63
-------
3.0 EFFECTS OF INDUSTRIAL PRETREATMENT
3.1 Overview
Section 522 of the Water Quality Act of 1987 requires EPA to study the extent
to which implementation of the categorical pretreatment standards will affect hydrogen
sulfide corrosion in wastewater collection and treatment systems in the United States
(U.S.). EPA's industrial pretreatment program regulates the discharge of certain
constituents such as metals and toxic materials into municipal sewer systems. The
County Sanitation Districts of Los Angeles County (CSDLAC) implemented industrial
pretreatment standards in 1975-1977 to meet ocean discharge requirements, and
additional controls starting in 1983 to comply with the EPA-mandated industrial
pretreatment program.
Metals and other constituent levels in CSDLAC wastewater dropped substantially
between the early 1970's and the mid 1980's as a result of implementation of these
industrial pretreatment programs. A concomitant rise in both total and dissolved sulfide
levels in the wastewater occurred over this same time period. Further, CSDLAC
observed an increase in the rate of corrosion in their concrete sewers.
Two hypotheses have been set forth to explain increased corrosion rates in
CSDLAC sewers due to the reduction in levels of metals. The first is that at the higher
levels of metals, a significant amount of sulfide was rendered insoluble in metal-sulfide
compounds, reducing the amount of dissolved H:S available for release to the sewer
atmosphere. The second is that the higher levels of various metals and other
compounds in the wastewater had a toxic effect on the sulfate-reducing bacteria
responsible for the generation of sulfide. When the metal concentrations were
significantly reduced, the sulfate-reducing bacteria flourished, increasing sulfide levels in
the wastewater, which generated more dissolved H2S. Both phenomenon would increase
the amount of H:S available for release to the sewer atmosphere, and subsequent
corrosion of the sewer crown due to increased sulfuric acid production.
A thorough search of the literature and contacts with municipalities throughout
the U.S. revealed that no data existed from other cities to show a correlation between
implementation of industrial pretreatment standards and increased sulfide generation
and corrosion. Municipalities simply do not have historical data on corrosion rates or
sulfide levels that would allow establishing a correlation such as was found in CSDLAC.
Given the unavailability of full-scale data to support the theory proposed by
CSDLAC, the study objectives were determined to be as follows:
1. Investigate the theoretical impacts of metals on sulfide levels.
2. Review and analyze research conducted or supported by CSDLAC.
3-1
-------
3. Compare metals levels of CSDLAC with other cities to assess whether
other municipal sewerage systems could potentially experience a similar
phenomenon (decrease in industrial wastewater constituents and increase
in corrosion).
4. Review data from site visits to industrial cities to determine if corrosion
rates differed in sewers with high industrial contributions vs. those with
predominately residential contributions.
5. Review other potential impacts of implementing pretreatment standards.
3.2 Theoretical Impacts of Sulfide Reaction with Metals
It is well known that many metals "bind" with sulfide to produce a precipitate
which is insoluble, effectively preventing release of hydrogen sulfide gas to the sewer
atmosphere and preventing formation of corrosive sulfuric acid. As discussed in Section
4, salts of metals such as iron and zinc are routinely added to wastewater to prevent
odors and corrosion associated with hydrogen sulfide.
The weight of metal required to precipitate a given weight of sulfide can be
predicted theoretically using chemical reaction equations. Table 3-1 shows the probable
precipitation reactions of metals with sulfide in wastewater devoid of oxygen. Based on
the stoichiometry of the reactions, the necessary concentration of each metal required to
precipitate 1 mg/1 of sulfide has been calculated. The last column shows the inverse of
this value, or the theoretical concentration of sulfide that would be precipitated by 1
mg/1 of metal. However, in wastewater containing a complex mix of organic and
inorganic compounds which interfere with such reactions, the amount of metal required
to precipitate a given weight of sulfide may be much greater than what would be
predicted from the equations.
Table 3-2 shows the theoretical stoichiometric increase in dissolved sulfide
concentration based on the reduction in metals concentration experienced in CSDLAC
between the periods 1971-1974 and 1983-1986. The total theoretical increase in
dissolved sulfide due to reduced availability of metals to precipitate the sulfide is
approximately 4 mg/1. The measured increase in dissolved sulfide during that same
period was approximately 1 mg/1.
Reduction in iron alone accounts for 69 percent of the theoretical increase in
sulfide. Zinc accounts for 16 percent, and chromium 10 percent. The reduction in
these three metals accounts for 95% percent of the theoretical increase in dissolved
sulfide levels.
Because of its toxicity, chromium is not used for sulfide control. However, data
are available on dosage requirements for iron and zinc to precipitate sulfide that are
3-2
-------
TABLE 3-1
PROBABLE METAL - SULFIDE PRECIPITATION REACTIONS
IN WASTEWATER DEVOID OF OXYGEN
Theoretical mg/1
of Metal to Theoretical mg/1
Precipitate or Sulfide Precipitated
Reactions 1 my/1 of Sulfide bv 1 mg/1 of Metal
Fe*2 + S"2 —> FeS 1.74 0.57
Zn*2 + S'2 —> ZnS 2.04 0.49
Ni*2 + S'2 —> NiS 1.83 0.55
Cd"*2 + S"2 —> CdS 3.51 0.28
Pb*2 + S'2 —> PbS 6.48 0.15
Or1 + S"2 —> Cu^S 3.97 0.25
Cr+2 + S"2 —> CrS 1.63 0.61
3-3
-------
TABLE 3-2
THEORETICAL INCREASE IN DISSOLVED SULFIDE
BASED ON METAL PRECIPITATION; LA COUNTY
Metal
Reduction
in Metals'
Theoretical
Increase in Dissolved
Sulfide Concentration2
Expected
Increase Based
on Field Studies3
mg/1
mg/1
mg/1
Chromium
0.68
0.42
~
Copper
0.38
0.10
—
Lead
0.17
0.03
—
Zinc
1.34
0.66
0.06 - 0.1
Nickel
0.14
0.08
—
Iron
4.92
2.83
0.1 - 0.7
Cadmium
0.01
0.00
--
TOTAL
4.12
1 Difference in average values for the periods 1971 • 1974 and 1983 - 1986.
2 Based on stoichiometry of chemical precipation reactions
3 Based on field dosages required to precipitate dissolved sulfide;
LA County research data.
3-4
-------
useful to estimate the actual increase in sulfide levels that might be expected by
reduction in metals.
Studies by CSDLAC in 1985-1988 on the use of iron addition to control sulfide
showed that when the dissolved sulfide levels were between 1 and 4 mg/1, a dosage ratio
of six to seven parts iron to 1 part dissolved sulfide was required to achieve 90 percent
removal. When the dissolved sulfide was less than 1 mg/1, a dosage ratio of 44 to 1 was
required (1). The theoretical dosage ratio is 1.7 to 1. Thus, four to 25 times the
theoretical dosage was required for iron. Other studies conducted in 1971-1972 in
CSDLAC have showed that five to seven times the theoretical dosage is required to
remove sulfide using zinc.
Based on this analysis, it is possible that the reduction in metals experienced by
CSDLAC could account for some portion of the observed increase in dissolved sulfide,
considering that iron and zinc can account for between 0.2 to 0.8 mg/1 of the increase in
dissolved sulfide. However, it is unlikely that precipitation could account for all of the
measured increase in dissolved sulfide (over 1 mg/1) in CSDLAC wastewater.
During the period 1971 to 1986, total sulfide increased from 0.4 mg/1 to 3.0 mg/1,
and dissolved sulfide from 0.1 mg/1 to 1.4 mg/1. If sulfide precipitation with metals was
the -only mechanism, the fraction of dissolved sulfide would increase, but the total
sulfide level would remain essentially constant as metals were reduced. This is because
the insoluble metal-sulfide precipitates are still detected in the total sulfide test
However, research by Pomeroy found that when iron was added to wastewater
containing sulfide, a reduction in total sulfide was observed. Two explanations were
suggested. The first is that one of the products of the reaction is iron disulfide, which
when treated with acid in the sulfide test, forms H2S and elemental sulfur. Elemental
sulfur is not measured in the test The second is that the iron may act as a catalyst to
oxidize sulfide to a product which is not detected in the sulfide test (3). Thus, it is
unlikely that chemical precipitation or the presence or absence of iron could account for
the increase in total sulfide between 1971 and 1986.
33 Biological Inhibition by Metals and Toxic Compounds
CSDLAC has conducted in-house experiments to investigate inhibition of
microbial sulfide generation by constituents present in wastewater. In addition,
CSDLAC is partially funding research at the University of Arizona to investigate the
toxic effects of metals on the sulfide-oxidizing bacteria Thiobacilli. which are responsible
for the production of sulfuric acid on the sewer crown. Other research at the University
of California at Los Angeles funded by CSDLAC has considered the effects of metals
concentrations on both sulfate-reducing bacteria and sulfide-oxidizing bacteria.
The first in-house experiments conducted by CSDLAC were bench-scale
laboratory studies designed to determine the acute toxicity levels of selected metals and
3-5
-------
cyanide on sulfate-reducing bacteria. Growth medium for sulfate reducers was dosed
with varying concentrations of metals and cyanide. Tubes were inoculated with primary
effluent containing the sulfate-reducing bacteria and incubated for three weeks. The
tests involved dosing with individual metals and cyanide as well as with a stock solution
containing nickel, chromium, zinc, copper, lead, cadmium, and cyanide in ratios
approximating that in sewage to determine if a synergistic effect existed.
Table 3-3 shows the results of the experiment with the individual metals and
cyanide. Copper was the most toxic at 6 mg/1, while cyanide was the least toxic at 50-55
mg/l(4).
The stock solutions containing the mixtures of metals and cyanide were added to
the tubes at various dilutions to simulate total constituent concentrations of 1, 5, 10, 15,
20, and 25 mg/1. The distribution of constituents was as follows: nickel - 6.2%,
chromium - 20.4%, zinc - 44.3%, copper - 13.3%, lead - 6.6%, cadmium - 0.7%, and
cyanide - 8.5%. The results of this experiment showed that growth was completely
inhibited at a total constituent concentration of 10 mg/1, but was not visibly affected at a
concentration of I mg/1. At 5 mg/1, growth was notably retarded. A synergistic effect
apparently existed wben the combination of metals and cyanide were added to the
growth medium.
CSDLAC performed a second experiment using column tests to determine on a
larger scale how changes in wastewater metal concentrations affect the generation of
sulfide. The testing was carried out at the Joint Water Pollution Control Plant
(JWPCP) in Carson, California.
The test apparatus consisted of three 8-inch diameter, 6-feet high, polyvinyl
chloride (PVC) pipes that functioned as test columns. The characteristics of the
columns are shown in Figure 3-1. Each column was filled with 4 feet of hand-cut
polystyrene cubes measuring 1.5 inches on each side. This provided a substrate for the
attached growth of sulfide-generating bacteria. Primary effluent from the JWPCP was
pumped through each column at a rate of 0.75 liters per minute at a temperature of
80°F. One column was used as a control while solutions with known metal
concentrations were added to the influent of the other two columns. Dissolved sulfide
and total sulfide levels were measured at the influent and effluent ends of each column.
After each solution was tested, all the columns were cleaned, the polystyrene was
replaced, and the system was flushed. As of April 1, 1989, seven different metal
solutions had been tested.
The first of the seven trials was performed with a cocktail of chromium, copper,
zinc, and cyanide. The feed rate of the cocktail was controlled so that the
concentrations of metals in the wastewater approximated those of the early 1970's. The
second trial was performed with five times the concentration of metals and cyanide in
the wastewater. In the remaining trials, chromium, copper, nickel, and cyanide were
3-6
-------
TABLE 3-3
TOXICITY OF WASTEWATER CONSTITUENTS ON SULFATE-REDUCING BACTERIA1
Compound
Copper
Nickel
Chromium
Lead
Zinc
Cyanide
In-bouse experiment conducted by CSDLAC
Toxic Cone, mg/1
6
13
23
25
25
50-55
3-7
-------
Figure 3-1
Sulfide Generation Pilot Plant
Upflow Packed Columns
COLUMN NO. 3
WASTEWATER EFFLUENT
FROM SEDIMENTATION TANKS
COLUMN NO. 2
CHEMICAL
ADDITION
TANK No. 3
COLUMN NO 1
-^CHEMICAL
ADDITION
TANK No. 2
EFFLUENT TO
SEDIMENTATION TANKS
-------
each tested separately at wastewater concentrations approximating those of the
early 1970's. The target concentrations for the trials are listed in Table 3-4.
Results of the column experiments are summarized in Table 3-5 and 3-6.
Examination of the data yields several pertinent conclusions. Clearly, when the
"cocktail" of metals and cyanide were added at five times the 1971-1974 concentration,
sulfide generation was significantly inhibited. These results are supported by bench
scale studies at the CSDLAC laboratory which showed that inhibition of sulfide-
reducing bacteria requires metals concentrations much higher than that observed in
wastewater. However, the concentration of metals causing sulfide generation inhibition
in these pilot scale studies cannot be used to predict the increased levels of total and
dissolved sulfide observed by CSDLAC in their sewers.
Of greater significance is the comparison of performance when the "cocktail"
containing metals and cyanide was fed to the columns at the 1971-1974 levels. Total
sulfide concentrations in the effluent from the control column and test column were
16.6 mg/1 and 12.4 mg/1, respectively. Dissolved sulfide levels were 13.5 mg/1 and 9.7
mg/1 for the control and test columns, respectively. A statistical analysis was conducted
of the data from the experiment in which metals and cyanide were added at the 1971-
1974 levels, assuming identical influent metals concentrations in both control and test
columns. The analysis indicated that at a 0.01 level of significance (highly significant),
the total and dissolved sulfide concentrations in the effluent from the control columns
were higher than the levels in the effluent from the test column. Further, it was
determined that, at the 90% confidence level, the actual difference in effluent total
sulfide concentrations was between 2.8 and 5.6 mg/1. For effluent dissolved sulfide, the
actual difference was between 2.7 and 5.1 mg/1.
Based on the difference in sulfide levels in the effluent and influent for both the
control column and the test column, CSDLAC staff prepared Figure 3-2 showing the
percent change in sulfide generation upon addition of metals and cyanide. At the lx
levels (corresponding to the early 1970's), generation of total and dissolved sulfide was
reduced by 34 percent At the 5x levels, generation of total and dissolved sulfide was
reduced by 114 and 106 percent, respectively. Interestingly, addition of zinc alone at the
1970's levels resulted in reduction of sulfide generation by approximately one third.
Another pilot study conducted by CSDLAC involved the construction of two 3/4-
inch diameter, PVC "force mains," each 150 feet in length. One served as a control
pipeline, receiving primary effluent, and the other served as a test pipeline receiving a
"cocktail" of metals and cyanide similar to that used in the column experiment One
difference in the feed material to the pipeline pilot system was that iron was also
included in the metals cocktail at a level approximately that of the early 1970's (10.7
mg/1). Only the lx levels of metals and cyanide were used during the pipeline
experiments. Figure 3-3 is a diagram of the pipeline pilot plant system.
3-9
-------
TABLE 3-4
CONCENTRATION OF AGENTS ADDED TO UPFLOW PACKED COLUMNS (5)
1971-74 1987 Ix 5x
Agent Concentration Concentration Concentration Concentration
mg/1 mg/1 mg/1 mg/1
Chromium 0.92 0.178 1.00 5.00
Copper 0.60 0.18 0.50 2.50
Cyanide 0.32 0.02 0.40 2.00
Nickel 0.285 0.087 0.25 2.25
Zinc 2.17 0.60 2.00 10.00
3-10
-------
TABLE 3-5
AVERAGE INFLUENT SULFIDE, TOTAL COD, SUSPENDED SOLIDS AND
EFFLUENT SULFIDE; UPFLOW PACKED COLUMNS (5)
CONTROL COLUMN TEST COLUMNS
Influent -Effluent- Influent --Effluent--
Additive
DS
TS
COD
SS
DS
TS
DS
TS
COD
SS
DS
TS
mg/1
mg/1
mg/1
nig/l
mg/1
mg/1
mg/1
nig/l
mg/1
mg/1
mg/1
mg/1
lx Mixture
7.3
9.5
413
84
13.5
16.6
5.6
7.6
399
112
9.7
12.3
5x Mixture
6.2
8.2
420
71
11.3
14.6
0.8
3.0
426
70
0.5
2.1
Cyanide
8.9
11.0
381
80
15.0
17.8
5.7
7.6
360
61
13.9
17.4
Chromium (VI)
7.5
10.0
289
58
12.7
15.8
4.9
6.9
407
88
9.9
12.5
Nickel
6.5
8.4
509
77
15.1
17.4
6.1
8.8
608
211
13.8
16.1
Copper
6.5
8.4
509
77
15.1
17.4
4.6
6.9
513
149
13.4
16.3
Zinc
4.2
7.0
440
86
13.1
15.9
2.8
5.8
407
135
9.0
11.7
Chromium (III)
6.2
6.8
681
342
12.0
14.1
3.8
5.0
537
238
11.6
13.5
3-11
-------
TABLE 3-6
COMPARISON OF CONTROL AND TEST COLUMNS' SULFIDE GENERATION;
UPFLOW PACKED COLUMNS (5)
EFFLUENT MINUS INFLUENT SULFIDE PERCENT CHANGE
Control Test Control vs. Test
Additive PS TS PS TS DS TS
mg/1
mg/1
mg/1
mg/1
%
%
lx Mixture
6.2
7.1
4.1
4.7
-34.0
-34.0
5x Mixture
5.1
6.4
-0.3
-0.9
-106.0
-114.0
Cyanide
6.1
6.8
8.2
9.8
34.0
44.0
Chromium (VI)
5.2
5.8
5.0
5.6
-4.0
-3.0
Nickel
8.6
9.0
7.7
7.3
-10.0
-19.0
Copper
' 9.6
9.0
8.8
9.4
2.0
4.0
Zinc
8.9
8.8
6.1
5.9
-32.0
-33.0
Chromium (III)
5.8
7.3
7.9
8.5
36.0
16.0
3-12
-------
100
Figure 3-2
Percent Change in Sulfide Generated Due to Metals and Cyanide
Upflow Packed Column Pilot Plant
I
U>
50
-------
Figure 3-3
Sulfide Generation Pilot Plant
Pipeline System
EFFLUENT FROM PRIMARY
SEDIMENTATION TANKS
CHEMICAL ADDITION
TANK
LINE NO. 2
LINE NO. 1
EFFLUENT TO
SEDIMENTATION TANKS
EFFLUENT TO
SEDIMENTATION TANKS
-------
Tables 3-7 and 3-8 provide the results of the experiment At the lx metals and
cyanide dosage without supplemental iron, the generation of total and dissolved sulfide
was reduced by 36 to 25 percent, respectively. When iron was added to simulate
concentrations in the early 1970's, total dissolved sulfide generation was reduced by 51
and 77 percent, respectively. These results are depicted graphically in Figure 3-4.
The results of these experiments strongly suggest that the generation of hydrogen
sulfide in the wastewater of CSDLAC was suppressed due to the presence of
constituents associated with industrial discharges of the early 1970's. Higher levels of
sulfide in the wastewater would be expected to result in higher concentrations of
hydrogen sulfide gas in the sewer atmosphere and higher sewer corrosion rates.
However, the relationship between wastewater sulfide levels and conosion rate is not
well established.
3.4 Comparison of Metals at CSDLAC with Other Cities Before Pretreatment
Using available data, pre-1975 levels of metals and cyanide entering the main
CSDLAC wastewater treatment plant were compared with levels in the wastewater of
other municipalities across the U.S.. Data were analyzed for 50 cities from the EPA
report, "Fate of Priority Pollutants in Publicly Owned Treatment Works"(6)(7). These
data were collected in 1978-1979 prior to any significant implementation of industrial
pretreatment standards. The fifty cities typically had estimated industrial flow
contributions ranging from ten to fifty percent of the total flow. Analysis of these data
allowed determination of the number of cities with metals and cyanide levels similar to
those of CSDLAC prior to pretreatment, and assessment of whether other cities may
have bad the potential to experience suppression of sulfide generation and conosion
due to the presence of these constituents.
Table 3-9 shows a ranking of the 50 cities plus CSDLAC based on the
concentrations of selected metals and cyanide in the wastewater. This was developed
from the sum of the equivalent weight concentrations of each of the constituents, and
does not account for the relative toxicity of the constituents on sulfide-producing
bacteria. Of the 50 other municipalities, only three (six percent) are ranked higher than
CSDLAC, while 47 (94 percent) are ranked lower. The total concentration of metals
and cyanide in CSDLAC wastewater was approximately three times the median
concentration for the 51 cities. Table 3-10 shows the actual constituent concentrations,
in ug/1, for the 51 cities.
The total metals levels in CSDLAC wastewater in 1986 are also shown in Table
3-9. On an equivalent weight basis, 1986 levels were 42 percent of 1971 - 1974 levels.
Comparing 1986 CSDLAC levels with 1978 - 1979 levels of 50 other cities, 16 cities (32
percent) were higher than CSDLAC, and 34 cities (68 percent) were lower.
Clearly, sulfide generation and con-osion in CSDLAC sewers increased
3-15
-------
TABLE 3-7
AVERAGE INFLUENT AND EFFLUENT SULFIDE;
PIPELINE PILOT PLANT (5)
CONTROL PIPELINE TEST PIPELINE
—Influent— -Effluent-- —Influent— -Effluent-
Additive
DS
TS
DS
TS
DS
TS
DS
TS
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
lx Mixture
2.5
3.8
7.9
9.7
2.7
3.7
6.4
7.5
lx Mixture + Fe
2.5
3.9
7.8
9.8
2.5
3.6
3.9
6.4
TABLE 3-8
COMPARISON OF CONTROL AND TEST PIPELINE SULFIDE GENERATION:
PIPELINE PILOT PLANT (5)
EFFLUENT MINUS INFLUENT SULFIDE PERCENT CHANGE
Control Test Control vs. Test
Additive
DS
TS
DS
TS
DS
TS
mg/1
mg/1
mg/1
mg/1
%
%
lx Mixture
5.2
5.9
3.8
3.8
-25
-36
lx Mixture + Fe
5.3
5.9
1.4'
2.8
-77
-51
3-16
-------
50
Figure 3-4
Percent Change in Sulfide Generated
Due to Metals and Cyanide
Pipeline Pilot Plant
u>
I
t—'
-J
UJ
O
z
<
X
o
LLI
o
cc
UJ
0-
0
-50
-25%
-36%
-100
-51%
TOTAL SULFIDE GENERATED
*
DISSOLVED SULFIDE GENERATED
::: i: r:;:::: ^:
liiiiiiiiliiiiii
I'iHilHi
-77%
-150
_L
_L
1X ADDITION
1X ADDITION + IRON
-------
TABLE 3-9
COMPARISON OF CSDLAC METALS LEVELS BEFORE AND AFTER
PRETREATMENT WITH METALS LEVELS OF 50 CITIES IN 1978-1979
PLANT
1
2
3
LA Cointy
-------
TABLE 3-10
METALS AND CYANIDE CONCENTRATIONS IN WASTEWATER
FROM 51 CITIES1
Concentration
CADMIUM
CHROMIUM
COPPER
CYANIDE
LEAD
MERCURY
NICKEL
ZINC
IRON
PLANT
ug/1
ug/l
ug/l
ug/l
ug/l
ng/l
ug/l
ug/l
ug/l
1
17
572
267
61
583
133
189
6717
151917
2
10
226
123
4747
136
333
98
486
4267
LACotnryJ
32
915
596
322
312
1400
286
2164
10706
4
966
1422
803
123
164
5000
440
3897
8740
5
107&
1390
760
99
199
3233
701
4935
7363
6
100
1252
138
84
217
300
218
928
11827
7
4
28
269
100
98
1000
61
317
13074
S
51
170
1704
414
1223
617
91
7170
3072
9
2
55
117
2122
58
67
10
233
2212
10
1
159
330
5
194
1667
46
806
10172
11
4
427
152
1323
132
983
1097
283
3750
12
2
410
20
1560
50
295
50
2769
1264
13
1
163
82
1753
45
1050
10
261
2373
14
174
414
922
891
58
517
164
1615
2888
15
2
1
39
2003
50
200
10
103
703
16
2
80
47
37
127
1250
20
494
7358
17
1
256
337
11
329
350
427
1722
5460
IB
2
155
291
337
7
333
394
799
5043
19
2
4175
104
54
260
2000
140
1520
2546
20
4
80
107
1215
55
600
6
179
1768
21
4
460
378
141
281
117
300
927
4755
22
1
239
59
83
50
900
25
189
5568
23
12
419
183
17
48
305
93
260
5330
26
10
107
98
714
47
600
54
224
2035
25
2
51
117
169
136
517
24
330
4103
26
5
289
223
42
72
1000
345
619
3463
27
9
152
245
713
158
667
64
196
961
2B
3
49
102
71
35
283
21
381
3790
29
2
39
185
243
26
767
63
370
2925
30
2
16
72
822
34
633
4
208
712
31
5
109
105
452
51
300
86
232
1997
32
15
172
165
255
105
933
69
591
2222
33
4
15
21
2
123
833
11
114
3700
34
2
101
221
500
16
833
5
120
1266
35
2
t
20
175
50
67
10
93
2890
36
100
458
170
38
200
400
170
793
2063
37
2
125
241
28
50
617
55
395
2665
3a
2
33
110
370
50
533
13
117
1502
39
67
132
358
267
91
550
108
352
1289
40
8
96
252
164
200
333
265
181
1829
41
2
44
336
214
38
683
62
226
1693
42
9
100
K2
289
135
817
80
242
1370
43
25
82
62
204
9
50
22
274
1678
44
27
108
185
277
81
350
34
272
1103
45
3
12
119
102
29
833
117
294
1462
44
4
71
54
77
16
2K
30
278
1640
47
3
55
70
82
91
999
38
160
1505
48
6
102
70
12
67
12
248
1192
49
2
1
23
20
40
1167
8
89
1400
50
1
9
58
125
9
483
96
145
732
51
1
13
198
121
24
500
4
120
679
CAONIlM
CHROMIUM
COPPER
CYANIDE
LEAD
MERCURY
NICKEL
ZINC
IRON
AVG:
55
314
232
472
131
757
135
911
6393
STD:
198
638
284
799
187
812
19a
1539
20791
Without PLANT #1
AVG:
56
299
226
477
120
770
133
771
3416
STD;
200
644
287
805
177
816
199
1315
2948
' Compiled froM data contained in Raft. 5,6,7
J LA County - average levels during 1971-74. All other data for 1973-79.
3-19
-------
dramatically between 1971 and 1986, and metals and cyanide levels dropped
significantly. Unfortunately, it is not known what levels of specific metals, cyanide, and
combinations cause suppression of hydrogen sulfide corrosion. Although CSDLAC
appears to have passed a threshold level of metals and cyanide which resulted in
increased sulfide levels and corrosion, it is difficult to predict whether other cities could
experience a similar increase in sulfide generation and corrosion upon reduction in
metals levels resulting from industrial pretreatment.
3.5 Site Visits to Industrialized Cities
EPA conducted site visits to three cities having portions of the sewer system with
high industrial contributions. Initially, it was believed that comparison of corrosion in
residential vs. industrial sewers might show differences attributable to the metals and
other constituents present in the wastewater. The cities were Charlotte, NC;
Milwaukee, WI; and Tempe, AZ.
3*5.1 Charlotte, North Carolina (see also Section 2.2.7)
In the Charlotte sewer system, four sewers conveying primarily residential
wastewater and six sewers with a large industrial flow contribution were inspected. All
of the sewers were 20 to 25 years old.
Total sulfide levels in the residential sewers ranged from 0.2 to 0.6 mg/1. Pipe
surface pH measurements were generally 6.0, with the exception of one site where
surface pH levels ranged from 4.5 to 6.0. At that site, corrosion penetration was
approximately 0.25 inches, exposing aggregate. No measurable H2S was detected in the
sewer head space at any of the four sites.
Two of the six industrial sites showed signs of shallow hydrogen sulfide corrosion,
with penetration estimated to be up to 0.12 inches. Wastewater sulfide levels at the six
sites ranged from 0.0 to 0.3 mg/1. At sites where corrosion was observed, high
turbulence levels were noted. Wastewater pH measurements were 6.0 at four of the
industrial sites, 5.5 at one site, and 10.0 at the remaining site. No H2S was detected in
the headspace at any of the six industrial sites.
3.5.2 Milwaukee, Wisconsin (see also Section 2.2.8)
In Milwaukee, observations included five sewers conveying primarily residential
wastewater and five sewers with a heavy industrial flow contribution. Three of the
residential sites were located in the Jones Island WWTP service area, and were concrete
pipes ranging in age from 50 to 70 years. No corrosion was observed, and the surface
pH of the concrete was measured to be 6.5. Two other residential sites were sewers in
the South Shore WWTP service area. One site was a sewer less than 20 years old; the
other was 50 years old. The 20 year old sewer was similar to the first three - no
3-20
-------
corrosion and high surface pH. However, at the other site wastewater sulfide content
was 0.5 mg/1 and the crown pH was 3.5. Severe corrosion was observed, with up to one
inch of concrete lost
All five industrial sites are at least 6 miles downstream from the beginning of the
collection system and at least 40 years old. No corrosion was observed at any of these
sites. However, two sites had measurable wastewater sulfide levels of 0.18 and 0.40
mg/1. Crown pH levels were between 6.0 and 7.0 at all of the industrial sites. One site
was .less than 0.5 miles downstream of a tannery.
3-5-3 Tempe, Arizona
A site visit was made to Tempe because one area of the City generated
wastewater primarily of industrial origin. These industries included four circuit board
manufacturers, two electroplaters, two metal finishers, one coating operation, and two
dry cleaners. It was believed that results of detailed monitoring data from industrial vs.
residential areas could provide some insight as to the impact of industrial discharges on
the extent of hydrogen sulfide corrosion.
Data from previous monitoring by City of Tempe staff were reviewed, and
inspections were made of industrial and residential sewers. The City had measured
sulfide, pH, temperature, ORP, and D.O. of the wastewater, and had contracted with a
° local laboratory for metals analysis. At the industrial site, dissolved sulfide levels ranged
from 0.05 to 1.9 mg/1 on five separate days (grab samples). Dissolved sulfide levels
averaged 0.6 mg/1. Atmospheric H,S levels were 1 to 2 ppm. Wastewater pH ranged
from 8.0 to 8.9. Inspection of a manhole in the 27 inch sewer conveying the majority of
wastewater from the industrialized area showed no evidence of corrosion. The sewer
system in this area was approximately 20 years old, and appeared to be in excellent
condition with no sign of deterioration. On the day of inspection, no atmospheric H2S
was detected in the manhole near the liquid surface.
A large trunk sewer conveying residential wastewater from the adjoining City of
Mesa was inspected at several locations. Previous monitoring by the City of Tempe had
indicated dissolved sulfide levels of up to 9 mg/1, and atmospheric H2S levels of up to 68
ppm. The concrete manholes which were unprotected exhibited severe corrosion above
the waterline, with abundant quantities of corrosion product Several manhole
chambers had been lined with plastic, and those manholes that were sources of odor
complaints were equipped with carbon canisters to control odor emissions.
No conclusions could be drawn from existing data regarding the impact of
industrial discharges on hydrogen sulfide corrosion. A further, detailed monitoring
program was abandoned because of the high pH (8.0 to 8.9) of the wastewater
emanating from the industrial zone. This high pH prevents significant amounts of HZS
from being released from solution, since over 90 percent of the dissolved sulfide is
3-21
-------
present as the hydrosulfide ion, not as the dissolved gas. In general, it is believed that
attempts to assess the impacts of metals and other industrial constituents on hydrogen
sulfide corrosion by monitoring industrial vs. residential sewers are futile due to the
many factors which affect sulfide generation and coiTosion.
3.6 Beneficial Effects of Local Industrial Pre treatment Programs
It is important to recognize that several aspects of the industrial pretreatment
standards may actually lower the potential for sulfide generation and corrosion in sewer
systems. Among the more important of these are 1) reduction of sulfide-bearing wastes,
2) reduction of high strength organic waste discharges, 3) reduction of high temperature
discharges, 4) reduction in fats, oils, and grease, and 5) reduction in acidic wastes.
Because of the complex interaction of all the factors that affect sulfide generation, it is
very difficult to quantify these effects for a broad base of sewer systems. Beneficial
impacts of local regulation of industrial waste discharges on sulfide generation in
municipal sewers are summarized in Table 3-11. In this table, sulfide is the only
parameter specifically regulated by the EPA Categorical Pretreatment Standards.
3.7 Conclusions
The national effects of industrial pretreatment on hydrogen sulfide corrosion are
impossible to ascertain since no municipalities other than CSDLAC were found to have
sufficient data to establish ^correlation. Based on theoretical analysis, review of full
scale and pilot scale research data from CSDLAC, and a series of site investigations, the
following conclusions are presented.
• The reduction in metals and other industrial constituents in CSDLAC
wastewater apparently caused an acceleration in corrosion rate, possibly
due to biological inhibition and/or chemical precipitation.
• Two pilot studies conducted by CSDLAC demonstrated that sulfide
generation was reduced when metals were added to the wastewater at
levels approximating those in the early 1970's.
• When comparing 1970's data from 50 other cities having 10 to 50 percent
industrial flow input, total metals and cyanide levels in CSDLAC
wastewater were higher than levels in 94 percent of 50 U.S. cities.
• If current (1986) CSDLAC data are compared with 1970's data from 50
cities, CSDLAC levels would be lower than 32 percent of the cities.
• It is difficult to project how many cities could potentially be adversely
affected by industrial pretreatment since it is not known at what levels
industrial constituents begin to suppress sulfide generation.
3-22
-------
TABLE 3-11
BENEFICIAL IMPACTS OF CONTROLLING
INDUSTRIAL DISCHARGES ON SULFIDE CORROSION
Tvpe of Discharge Controlled
Sulfide-bearing wastes
Benefit
Lowers sulfide levels,
corrosion potential
High organic strength
(BOD) wastes
Sulfide generation rate
proportional to BOD; reduction in
organic strength reduces oxygen
uptake and depression of dissolved
oxygen
High temperature wastes
Lower temperature reduces sulfide
generation rate; increases
solubility of HZS, reducing release
of H2S; increases solubility of
oxygen
Wastes containing fats,
oils, and grease
Reduces potential for sewer
clogging, reduced velocities, solids
deposition, and sulfide generation
Low pH wastes
Maintaining pH at or above
neutral reduces release of H2S to
the sewer atmosphere
3-23
-------
Site visits to inspect corrosion in residential vs. industrial sewers were
inconclusive regarding the impacts of metals and other industrial
constituents on hydrogen sulfide corrosion.
Local regulation of certain non-tone constituents in industrial waste
discharges has likely had a beneficial impact in reducing the potential for
sulfide generation and corrosion.
Additional research is necessary to establish the constituents and their
associated levels at which sulfide generation is suppressed or accelerated.
3-24
-------
REFERENCES
1. Won, D.L., "Sulfide Control with FerTous Chloride in Large Diameter Sewers,"
internal report, County Sanitation Districts of Los Angeles County, November,
1988.
2. Pomeroy, R.D., Parkhurst, J.D., Livingston, J., and H.H. Bailey, "Sulfide
Occurrence and Control in Sewage Collection Systems," U.S. EPA, EPA 600/X-
85-052, Cincinnati, OH, 1985.
3. Pomeroy, R.D., and F.D. Bowlus, "Progress Report on Sulfide Control Research,"
Sewage Works Journal, Vol. 18, No. 4, pp 597-640, July, 1946.
4. Internal Monthly, Reports, County Sanitation Districts of Los Angeles County,
1988.
5. Morton, R., Caballero, R., Chen, C.L., and J. Redner, "Study of Sulfide
Generation and Concrete Corrosion of Sanitary Sewers," prepared for the 62nd
Annual Conference of the Water Pollution Control Federation, San Francisco,
October, 1989.
6. "Fate of Priority Pollutants in Publicly Owned Treatment Works - Volume I,"
EPA 440/1-82/303, USEPA, Washington, D.C., Sept, 1982.
7. "Fate of Priority Pollutants in Publicly Owned Treatment Works - Volume II,"
EPA 440/1-82/303, USEPA, Washington, D.C., Sept, 1982.
3-25
-------
4.0 DETECTION, PREVENTION AND REPAIR OF HYDROGEN SULFIDE
CORROSION DAMAGE
Alternatives for the detection and prevention of hydrogen sulfide corrosion in
both existing and new wastewater systems, and techniques for repairing hydrogen sulfide
corrosion damage, are summarized in this section. Additional detailed information may
be found in publications prepared by the Environmental Protection Agency, The
American Society of Civil Engineers, the American Concrete Pipe Association, and the
U.S. Department of Housing and Urban Development (1)(2)(10)(7).
4.1 Detection and Monitoring of Hydrogen Sulfide Corrosion
One of the most useful "early warning" indicators of potential hydrogen sulfide
corrosion problems is pH of the pipe crown or structure wall. This is a simple test using
color-sensitive pH paper which is applied to the moist crown of the pipe. New pipe has
a high pH of 10 to 11. After aging the pH of the crown under non-corrosive conditions
may drop to near neutral. Pipe experiencing severe hydrogen sulfide corrosion may
have a pH of 2 or lower.
Dissolved sulfide levels in the wastewater and hydrogen sulfide levels in sewer
headspaces can be checked to determine if sulfide is being generated in the sewers and
where and to what extent it is being released from solution. Routine monitoring may be
justified at the lift stations, junction structures, discharges of force mains, treatment
plant headworks, or other locations in the collection and treatment system. Such tests
indicate whether conditions are present for hydrogen sulfide corrosion to occur.
Routine visual inspections are essential. Where accessible by a worker, this can
be done by entering manholes or sewers and checking the soundness of the pipe
material. A screwdriver or other sharp tool can be used to determine the depth of
penetration into soft corrosion product Since corrosion products occupy greater
volume than the original concrete, depth of penetration is not an accurate measurement
of concrete lost to corrosion (section 1.4). Concrete loss can be approximated by
measuring the depth of aggregate protrusion from the surface. Sewer pipe may also be
inspected remotely through the use of television cameras. With improvements in the
resolution of camera equipment, TV inspections can often identify coirosion problems,
although considerable damage may already have been done.
A relatively recent development in remote sewer inspections is the use of "sonic
caliper" technology to measure the inside dimensions of the pipe. Sonic signals are
transmitted from a floating raft to the pipe walls, and the signal is detected after
reflecting off the wall. Software developed by a proprietor is used to process the signals
and determine the variation in pipe diameter along its length. Areas where loss of pipe
material has occurred can thus be detected. The technique was successfully used to
inspect over 40,000 feet of 36 in. to 54 in. diameter pipe in Tampa, Florida (9).
4-1
-------
Core borings of the pipe crown and submerged pipe may be taken to calculate
the extent of the corrosion loss. Some municipalities only take cores of corroded
portions of pipes to determine how much pipe remains. Expandable rods have also
been used to measure the inside pipe diameter, rather than taking core borings, as a
means of estimating the extent of corrosion, although errors are introduced due to
variation in wall thickness and pipe "roundness/
To determine the rate of corrosion, the thickness of corroded pipe must be
compared at two different points in time, since no instantaneous technique has been
proposed for monitoring corrosion rates. CSDLAC is the only entity known that has
determined corrosion rate on a regular basis using core borings. Some sewers may be
installed with vitrified clay "plugs" or stainiess steel rods in the crown at accessible
locations, providing a direct visual comparison of adjacent corroded and non-corroded
material
4.2 Prevention of Hydrogen Sulfide Corrosion in Existing Systems
A number of techniques have been used to control corrosion and odors
associated with hydrogen sulfide generation in existing systems. The most common
techniques can be divided into the general categories of oxidants, precipitants, or pH
elevators. Oxidants control sulfide by chemically or biologically causing the oxidation of
sulfide to thiosulfate or sulfate. Such techniques include air or oxygen injection, or
addition of chemicals such as hydrogen peroxide, chlorine, or potassium permanganate.
Precipitants control sulfide by precipitation with a metal salt such as ferrous chloride,
ferrous sulfate, or zinc salts. The dissolved sulfide is converted to an insoluble
precipitate, preventing release of gaseous H2S. Elevation of the pH through shock
dosing of caustic controls sulfide generation by inactivation of sulfide-producing slimes
present on the wall of the sewer pipe. A summary of sulfide control techniques is
provided in Table 4-1.
All of the above control techniques are oriented towards reducing the levels of
dissolved sulfide in solution such that less sulfide is released to the sewer atmosphere.
Work conducted by CSDLAC indicated that, although significant reductions (75 to
95%) in dissolved sulfide could be obtained with chemical addition, only modest
reductions (50 to 60%) in H2S levels in the sewer atmosphere were realized (4). Thus,
a 90% reduction in dissolved sulfide does not necessarily indicate that the rate of
corrosion will be reduced by 90%. Although empirical predictive corrosion equations
assume that corrosion rate is directly proportional to the rate of H,S flux from the
wastewater to the sewer walls, this relationship is very difficult to verify with field data.
No one sulfide control technique can be generalized as being the most cost-
effective. Dosages of chemicals to control sulfide vary widely from one wastewater
system to another, and are dependent on wastewater characteristics and other site-
specific factors. Sulfide control options must be considered on a case-by-case basis.
4-2
-------
TABLE 4-1
SUMMARY OF SULFIDE CONTROL TECHNIQUES (4)
Technique
I. OXIDATION
Air Injection
Direct oxygen
injection
Sidestream 02
injection
Hydrogen peroxide
Chlorine
Potassium
permanganate
Frequency
Of Use
Low; limited
by application
Low in U.S.;
high in U.K.,
Australia
Very low
Relative
Cost
Low
Low
High
Med.
Med.
High
Low
Med.-
High
High
Advantages
Low cost, adds
DO to wastewater
to prevent further
sulfide generation
5 x solubility of
air; very economical
for force mains; adds
DO
Applicable to
oxygenating gravity
sewers and wet wells
Effective for odor/
corrosion control in
grav. sewers or force
mains; simple
installation
Applicable to grav.
sewers or force mains
Effective, powerful
oxidant
Disadvantages
Applicable only
to force mains;
potential for air
binding
Achieving good
02 transfer may
be difficult
Potential for
degassing of 02
from solution
Costs can be high
if dosages much
greater than
stoichiometric
Safety considerations
High cost,
difficult to
handle
-------
TABLE 4-1 (conl.)
SUMMARY OF SULFIDE CONTROL TECHNIQUES
Technique
Frequency
Of Use
Relative
Cost
Advantages
Disadvantages
II. PRECIPITATION
Iron salts High
Low -
Med.
Zinc salts
Very low
Med.-
High
Can be used for
sulfide control in
gravity sewers or force
mains
Lower solubility than
iron; may inhibit
sulfate-reducing bacteria
due to intrinsic toxicity
Does not control
non-H2S odors;
sulfide control
to low levels may be
difficult
Not economical
compared to iron salts;
discharge is regulated
III. pH ELEVATION
Sodium hydroxide Med
(shock dosing)
IV. OTHERS
Sodium nitrate Very low
Low
Med.-
High
Economical intermittent
application
Prevents reduction of
sulfate to sulfide
Special handling
of high pH slug may
be required at treatment
plant
Applicable only for the
prevention of sulfide
generation
-------
Brief descriptions of sulfide control methods applicable to existing wastewater systems
are provided below. More detailed information may be found in references (1) and (2).
4.2.1 Air Injection
Injection of compressed air into the wastewater is most applicable to force mains,
siphons, and pipes flowing under pressure. Often, air is injected on the discbarge side
of sewage pumps to provide dissolved oxygen which promotes oxidation of existing
sulfide and prevention of further sulfide build-up. Tlie pressure in the pipe, being
greater than atmospheric, allows dissolution of greater quantities of oxygen. Air
injection is an economical alternative for sulfide control in pressurized lines. Because of
the large quantities of air injected, potential exists for gas accumulation and increased
head losses. Although research has been conducted on pressure tank dissolvers and U-
tubes for use in aerating gravity sewers, such devices appear to be marginal for this
purpose.
4.2.2 Oxygen Injection
Oxygen is five times more soluble in water than air, and thus it is possible to
achieve higher DO levels in the wastewater. As with compressed air injection, oxygen is
most applicable to sulfide control in force mains and pipes under pressure. However, it
is currently being used in Sacramento to oxygenate wastewater in a fall structure, and
CSDLAC has conducted demonstrations of pressurized sidestream dissolution for
oxygenation of gravity sewers. In the sidestream dissolution system, a portion of the
flow is directed through a pressurized pipe into which oxygen is injected. The oxygen-
saturated sidestream is then introduced back into the gravity main.
Oxygen is generally an economical technique for sulfide control. However, the
annual costs for purchased oxygen are highly dependent on how efficiently the oxygen is
transferred into solution.
4.2J Hydrogen Peroxide
Hydrogen peroxide is widely used for sulfide control in force mains and gravity
sewers. At neutral and acidic pH, H202 oxidizes H,S to elemental sulfur. Dosage
weight ratios of H202 to H,S vary from near stoichiometric (1:1) to over 5:1, depending
on degree of control desired, wastewater characteristics, initial sulfide level, and
wastewater travel time between injection station and control point Costs for sulfide
control using hydrogen peroxide are competitive with other sulfide control chemicals.
4.2.4 Chlorine
Chlorine oxidizes sulfide to sulfate or elemental sulfur. Chlorine can be
purchased as a gas or as hypochlorite solution. In practice, C12:H2S dosage weight ratios
4-5
-------
are typically in the range of 10:1 to 15:1. Although commonly used for sulfide control,
dosage requirements and unit chemical costs often make chlorine uneconomical
compared to other chemicals used for sulfide control. The hazardous nature of chlorine
gas make it less attractive for use near residential areas.
4.2-5 Potassium Permanganate
Potassium permanganate is a powerful oxidant that is effective for sulfide control.
In general, dosage weight of KMn04 to H2S are approximately 6:1 to 7:1. Potassium
permanganate is purchased as dry crystals. Chemical costs are high, and use of KMn04
for wastewater applications is generally not cost-effective.
4.2.6 Metal Salts
Iron salts such as ferrous sulfate and ferrous chloride are widely used,
economical chemicals for sulfide control. Iron reacts with HjS to form an insoluble
precipitate, preventing release of H2S from solution. In practice, dosage weight ratios of
FeS04 to HjS are approximately 5:1, although higher dosage ratios may be required
depending on wastewater characteristics, initial sulfide levels, and degree of sulfide
control required.
4.2.7 Sodium Hydroxide
Sodium hydroxide is added in "shock doses'1 to sewers for sulfide control. Caustic
soda (NaOH) is added over a period of 20 to 30 minutes at sufficient dosages to elevate
the pH to between 12.5 and 13.0. The high pH slug temporarily inactivates sulfate
reducing bacteria and greatly reduces hydrogen sulfide generation. Within a period of
several days to two weeks, the sulfate reducing bacteria become re-established, and
caustic dosing must be repeated. If this approach is employed in the collection system
near the treatment plant such that dilution of the high pH slug does not occur,
provisions must be made to store the high pH wastewater and gradually release it to the
plant to avoid biological upset
4.2.8 Other Chemicals
1 Other chemicals have been used for sulfide control with varying degrees of
success. Sodium nitrate has been used for H2S control in lagoons, trickling filters, and
carbon columns, but has not been widely used for sulfide control in sewers. Nitrate
prevents sulfide generation by acting as a hydrogen acceptor which is used preferentially
by bacteria over sulfate.
Several proprietary bacterial cultures and enzyme preparations are claimed to be
effective for sulfide and odor control although their effectiveness has yet to be
demonstrated.
4-6
-------
43 Prevention of Hydrogen Sulfide Corrosion in the Design of New Systems
Consideration of sulfide generation and corrosion is critical in the design of
wastewater collection systems. While it is possible, and sometimes necessary, to
incorporate chemical addition stations for sulfide control as part of the overall system
design, the most cost-effective and rational engineering approach is to develop a
hydraulic design that minimizes sulfide generation. In general, such an approach strives
to maintain aerobic conditions in the wastewater by providing adequate wastewater by
providing adequate wastewater velocities, and by minimizing the use of force mains,
inverted siphons, and surcharged sewers in which anaerobic conditions can develop,
resulting in sulfide generation.
Under certain conditions, sulfide generation may be unavoidable. Empirical
equations have been developed to allow prediction of sulfide build-up and rates of
corrosion. Where sulfide generation is anticipated, conosion resistant materials can be
selected, or the alkalinity and thickness of concrete pipe can be specified to help reduce
the effects of hydrogen sulfide corrosion.
Table 4-2 summarizes various approaches used to minimize sulfide generation
and corrosion during the design of wastewater collection and treatment facilities.
Several key design elements are summarized below. More detailed discussions of
conosion prevention during design may be found in references (1),(2), and (10).
4.3.1 Wastewater Velocity
Wastewater velocity is critical in designing sewer systems to prevent or minimize
sulfide generation. Adequate velocity 1) prevents deposition of solids which can cause
flow obstructions and increase sulfide generation and 2) provides surface reaeration
which helps to maintain aerobic conditions and prevent sulfide generation. Although a
minimum scouring velocity of 2 ft/sec has been historically used by engineers designing
gravity sewers, large diameter sewers require much higher scouring velocities, on the
order of 3 to 4.5 fVsec. Minimum scouring velocities for force mains are typically 3 to 5
ft/sec depending on pipe size.
The impact of velocity on reaeration rate is significant For a 24 inch diameter
flowing half full, increasing the velocity from 2 to 3 ft/sec increases the reaeration rate
by a factor of 2. The effect is somewhat less dramatic for sewers larger than 36 inches.
Velocity in sewers is controlled by flowrate, slope, and pipe diameter. Figure 4-1
is a generalized guide showing the potential for sulfide generation as a function of flow
and slope.
4-7
-------
TABLE 4-2
APPROACHES TO PREVENT HYDROGEN SULFIDE CORROSION DURING DESIGN
Techniques to Minimize Sulfide
Generation and Corrosion
Choose pipe sizes and slope to provide sufficient
velocities to maintain aerobic conditions
and prevent solids deposition.
Limit use of force mains, siphons, and surcharged
sewers which promote anaerobic conditions.
00
Impose local control of industrial discharges to reduce
wastes with sulfide, high BOD, high temperature, low
pH, and high grease content.
Avoid excessive detention times in wet wells, holding
tanks, etc.
Techniques to Minimize Corrosion
when Sulfide Generation is Anticipated
Utilize corrosion resistant pipe materials such as PVC,
PE and vitrified clay.
Specify calcareous aggregate (high alkalinity) concrete
with additional sacrificial cover over reinforcing steel.
Specify corrosion-resistant PVC or PE liners for
concrete pipe, junction structures, etc.
Design junction structures, manholes, etc. to minimize
turbulence and release of H2S.
Consider air/oxygen injection or chemical addition
stations where appropriate.
-------
0.6
Effective BOO
500 mg/1
0.5
0.4
Little sulfide
generation potential
0.3
Moderate sulfide
generation potential
0.2
Severe sulfide
generation potential
0 0.1 0.5 1.0
5.0 10 15 20 30 40
Row. cu ft/s
FIGURE 4-1 GUIDE FOR ESTIMATING SULFIDE GENERATION POTENTIAL
4-9
-------
4.3.2 Design of Junction and Drop Structures
Turbulence created by junctions and drop structures can have opposite effects on
sulfide build-up and corrosion, depending on the characteristics of the wastewater when
it reaches the structure. If no sulfide is present, turbulence will increase reaeration
rates, thereby adding dissolved oxygen and maintaining aerobic conditions. On the
other hand, if dissolved sulfide is present, turbulence will increase the release of
hydrogen sulfide to the atmosphere and increase the rate of corrosion. Therefore, if
potential for sulfide generation exists, designs of such structures should be such that
turbulence is minimized.
433 Force Mains, Siphons, and Surcharged Sewers
Force mains, siphons, and surcharged sewers have one thing in common: the
sewer flows full with no opportunity for reaeration. This results in anaerobic conditions,
generation of H2S, and often severe corrosion at the outlet In general, use of force
mains, siphons, and surcharged sewers should be avoided whenever possible.
Where required, design velocities should be selected to avoid solids deposition,
and detention times should be minimized. Where long force mains or siphons are
necessary, consideration should be given to positive sulfide control systems (e.g.,
chemical addition) and/or use of corrosion resistant materials such as PVC liners at
discharge points.
4.3.4 Sewer Ventilation
Sewers are naturally ventilated through building vents and manholes, occurring
from factors such as changes in barometric pressure, wind, air density differences, and
flow conditions. Wet wells at pumping stations have 12-30 changes/hour, and dry wells
have 6-30 changes/hr. Ventilation is often practiced at wastewater treatment plants,
where air is withdrawn at the headworks and either treated separately or piped to
existing biological processes (1).
4J.5 Local Control of Industrial Discharges
Local control of industrial discharges as a means of minimizing sulfide generation
is applicable to both existing and new wastewater systems. Beneficial impacts of
controlling industrial discharges on sulfide generation were summarized in Table 3-11.
Adequate industrial pretreatment to make such wastes compatible with municipal
wastewater can eliminate their contribution to sulfide generation potential.
4-10
-------
4J.6 Design Considerations When Sulfide Generation is Anticipated
4.3.6.1 Prediction of Sulfide Generation and Corrosion
In some cases, it is difficult or not cost-effective to design a wastewater collection
system that will be free of sulfide problems. It is then useful to know what levels of
sulfide can be expected.
Empirical equations have been developed to allow prediction of sulfide levels. In
addition, a model has been developed to allow prediction of corrosion rates where H:S
is present
The Pomeroy-Parkhurst equations that predict sulfide build-up are given below:
where:
'Urn "
S,i
.-s,
log1
1 m(su)3flt
2.31 dm
S2
=
predicted sulfide concentration at time tj
s,
=
sulfide concentration at time tt
=
theoretical upper limit of sulfide concentration
s
=
slope of the pipe
u
stream velocity
t
=
(tj-t,) flow time
m
=
empirical coefficient for sulfide loss
=
mean hydraulic depth
Pipes Flowing Full
S2 = St + (M)(t)[EBOD (4/d) + 1.57)1
where:
M = experimentally determined empirical constant
representing the sulfide flux
EBOD = BODfl.O?™] (T = temperature, °C)
4-11
-------
The rate of corrosion of concrete pipe can be predicted using the following equation:
C„f = 11.? k0„
A
where,
CW| = average rate of penetration, mm/yr
k = Coefficient of efficiency for acid reaction considering the
estimated fraction of acid remaining on the wall. May be as low
as 0.3 and will approach 1.0 for a complete acid reaction.
0W = flux of H2S to the pipe wall, gm/m2-hr
A = Alkalinity of the cement bonded material, expressed as CaC03
equivalent Approximately 0.18 to 0.23 for granitic aggregate
concrete, 0.9 for calcareous aggregate, 0.4 for mortar linings, and
0.5 for asbestos cement
11.5 = constant
0„ - 0.69(su)wj[DS](b/P')
where,
s = energy gradient of wastewater stream, m/m
u = stream velocity, m/s
j = fraction of dissolved sulfide present as H2S as a function of pH
[DS] = average annual concentration of dissolved sulfide in the
wastewater, mg/1
b/P" = ratio of width of wastewater stream at surface to exposed
perimeter of the pipe wall above the water surface.
Peak corrosion rates occur at the sewer crown, and may be higher than the
average corrosion rate by a factor of 1.5 to 2.0. This is the crown corrosion factor
(CCF). Another factor, the turbulence corrosion factor (TCF) is used to account for
greater flux of H2S to the pipe wall, and may vary from 1 to 2.5 for well-designed
junction structures or other areas with nonuniform flow conditions. At drops or
turbulent junctions, the turbulence corrosion factor may be 5 to 10.
Thus, the peak crown corrosion rate is given as follows:
= Cwg x CCF x TCF
4-12
-------
4J.6.2 Selection of Materials
When it is anticipated that sulfide will be present, consideration must be given to
use of corrosion-resistant materials. Pipe materials such as PVC, PE, and vitrified clay
are virtually unaffected by sulfuric acid produced by the biological oxidation of hydrogen
sulfide, and should be specified if hydrogen sulfide corrosion is anticipated. For larger
diameter sewers, PVC or vitrified clay pipes are not available, and concrete pipe is often
used. However, PE pipe is available in diameters up to 120 inches. Two alternatives
are frequently used to protect the concrete pipe from corrosion or to extend its useful
design life. These are 1) use of PVC liners which are imbedded into the concrete pipe
during manufacture, and 2) use of high alkalinity calcareous aggregate and/or additional
sacrificial concrete cover over the reinforcing bars.
The proprietary "T-lock" PVC liner has been successfully used for concrete pipe
protection for many years. Although the cost of the pipe is increased, follow-up tests
have shown that the PVC lioer provides excellent protection of the underlying concrete.
The PVC is immune to sulfuric acid attack, and proper installation of the liner prevents
migration of acid to the concrete. Surface-applied synthetic coatings have yet to
demonstrate the longevity and acid resistance of PVC liners.
Use of calcareous aggregate in concrete pipe increases the alkalinity of the
concrete and thus increases the resistance to sulfuric acid attack. An additional
sacrificial layer of concrete over the reinforcing steel increases the useful design life of
the pipe. An equation has been developed to assist in proper selection of pipe materials
(alkalinity) and thickness. This is the "Az" or "life factor" equation, shown below.
Az = 0.45k L(CCF)(TCF)
where,
Az = life factor, equal to the product of alkalinity and thickness of
allowable concrete loss
L = desired design lifetime, years
Thus, assuming a design lifetime of the pipe, it is possible for the engineer to
specify the desired combination of pipe wall thickness and concrete alkalinity to achieve
the target lifetime.
4.4 Repair of Damage Caused by Hydrogen Sulfide Corrosion
Once corrosion damage has occurred, it may be necessary to repair a structure to
reduce the potential for failure or collapse. In the past, excavation and replacement was
a common repair solution to corroded pipes and structures. However, due to the
4-13
-------
expense, the disruption to traffic, the potential for damage to other underground
utilities, and the interruption to the service itself, in-line rehabilitation techniques have
become more attractive. Rehabilitation techniques are those methods and repairs
applied to an existing structure to prolong its useful life. With such techniques,
municipalities can repair existing structures at a lower cost than replacement, and
without public inconveniences due to traffic disruptions and service interruptions. In
many situations, pipelines can be rehabilitated at somewhat less than the cost of
replacement Rehabilitation techniques are not acceptable under the following
conditions:
• where significant additional capacity is needed
• where rehabilitation methods that are adequate to restore pipeline
structural integrity would produce an unacceptable reduction in service
capacity
• for point repair where short lengths of pipeline are too seriously damaged
to be effectively rehabilitated by any means
• where entire reaches of pipeline are too seriously damaged to be
rehabilitated
• where removal and replacement is less costly in dollars and urban
disruption than other rehabilitation methods
Numerous rehabilitation techniques exist, but not all are applicable to corrosion
repair. The selection of a particular method of rehabilitation depends on many factors
such as economics, extent of damage and structural integrity, disruption of traffic and
excavation requirements. High concentrations of sulfuric acid may be detected on the
walls of sewers where H2S is being generated. Sulfuric acid can quickly deteriorate
crown and sidewall concrete, thereby exposing aggregate and reinforcing steel, and
potentially weakening structural integrity. Therefore, corrosion rehabilitation techniques
must focus on internal repairs to ensure structural integrity and provide a protective
barrier against subsequent acid attack, rather than on external repairs (e.g., soil
stabilization). After surveying consulting engineers, municipal engineers, and
manufacturers, the following seven generic rehabilitation measures were identified as
appropriate for acid corrosion repair:
• insertion renewal (sliplining)
• deformed pipe insertion
• cured-in-place pipe
• specialty concrete
• coatings
• liners
• spot replacement
Table 4-3 describes various methods of pipeline rehabilitation, indicating their
4-14
-------
TABLE 4-3
PRINCIPAL METHODS FOR PIPELINE REHABILITATION
Method
Insertion Renewal
(Sliplining)
Deformed Pipe Insertion
¦P-
~—*
Description
A liner pipe of slightly smaller
outside diameter is inserted
into existing pipe, tben
connected to service laterals.
Materials include polyethylene,
polybuiylene, reinforced
thermosetting plastic, reinforced
plastic mortar.
A thermoplastic pipe is
deformed by folding or
compression and inserted into
existing pipe and expanded
naturally, hydraulically, or
mechanically
Application
Leading method for gas pipe
rehabilitation. Also used for
cracked or deteriorated sewer
pipes and, to lesser extent,
water distribution pipes.
Similar applications as for
sliplining but for relatively
small (<24 inch) circular pipe.
Advantages
Less time, lower cost than
excavation and replacement,
minimal disruption. May
improve hydraulics in some
cases. Provides some structural
reinforcement when properly
grouted. Bypassing not req'd.
Close fit of liner to pipe; may
not require grouting. Requires
no mixing of resins, curing.
May not require excavation.
May improve hydraulics.
Disadvantages
If original pipe is deformed,
liner pipe may have to be m
smaller diameter.Excavation
required for access pits, serv
laterals. Only large radius
bends are easily accotnmoda
May decrease capacity.
Limited track record. Curre
applicable only to small
diameter, circular pipe.
Bypassing required.
Cuted-in-Place Inversion Lining
Flexible liner installed through
inversion process, thermally or
steam hardened. Laterals cut
by remote control.
Sewer pipe of any geometry,
largest current application is for
96 inch diameter pipe.
For repairs under busy streets,
buildings as well as normal
locations. Return to service in
12 to 48 hrs. Excavation
normally not req'd. May
improve hydraulics.
Only used for mainline rep
Patented system handled by
relatively few contractors. S
set-up costs high for small j<
Bypassing req'd.
Specialty Concrete
(Spot repair)
After placement of steel
reinforcement, a mixture of fine
aggregate cement and water is
applied by air pressure.
Large sewers needing structural
repairs.
Higher strength than cement
mortar linings. Requires no
excavation. Variations in cross-
section readily accommodated.
Only suitable for large pipe:
Difficult to supervise and
depends on operator skills.
Control of infiltration requii
Susceptible to acid corrosioi
but at somewhat reduced ra
Bypassing req'd.
-------
TABLE -1-3 (conL)
PRINCIPAL METHODS FOR PIPELINE REHABILITATION
Method
Coatings
Description
Different materials that can be
applied by spray or brush to
sewer lines.
Liners
i
i—»
o\
Prefabricated panels or sheets
of PE or PVC that arc installed
manually, or a continuous,
interlocking strip that is
installed using a special
machine or by hand.
Application
Rapidly growing method for
pipes with new application
methods being marketed
coounually.
Not designed to support earth
loads and should be used only
in structurally sound sewers.
Can easily fit variations in
grades, slopes, cross section for
manually applied strip
applications
Advantages
A
Material cost is low, leu
disrupuon to traffic as
installation in-line; smooth
surface provides good
hydraulics.
Less disruption to traffic and
urban activity, less costly than
replacement
Disadvantages
Many applicable only to man
entry size sewers; prolonged by-
pass; greater surface
preparation; most coatings do
not provide long-term
resistance to acid attack.
Imperfect installations fail
completely, bypassing req'd.
Possible small reduction in pipe
capacity, requires obstruction
removal; susceptible to leakage
due to numerous joints;
bypassing req'd.
Spot Replacement
Exterior Wrap and Cap
Replacement in original trench.
Panels of ribbed PVC are
placed on the exterior of
corroded pipes and capped with
reinforced concrete.
Any pipe with major or
structural defects.
Provides back-up corrosion
protection and structurally
reinforces existing pipeline.
Similar method applicable to
monolithic structures.
Only method that can
significantly increase flow
capacity by replacement with
larger diameter pipe; allows
substitution with corrosion
resistant pipe materials having
long service life (e.g. PVC, PE,
VCP). Handles tight curves.
Service is not interrupted.
Bypassing is not required. Less
costly than complete
replacement
Most costly and disruptive
method; bypassing required.
Pipeline or structure must be in
a location where open
trenching or excavation u
practical.
-------
applications, advantages, and disadvantages. More detailed information may be found
in references (5),(6),(7), and (8).
Sliplining or insertion renewal is the most widely used rehabilitation technique.
Sliplining involves inserting a new, continuous length of pipe or segments of pipe inside
an existing pipe. Pipes are made by joining individual lengths with heat fusion or
various types of bell and spigot joints. Materials used for sliplining pipes include
polyethylene (PE), polybutylene, fiberglass-reinforced polyesters, reinforced
thermosetting resins, and PVC. Material selection is based on application, design needs,
economics, and to some extent, space availability at the installation working area.
Sliplining is capable of dealing with a variety of serious structural problems if the
annular space between the existing pipe and the sliplining pipe has been properly
grouted. Exceptions are severely crushed or collapsed pipes. Installation is done by
excavating an access pit and pushing or pulling the slipline pipe into the existing pipe.
It is widely used for cracked or deteriorated sewer pipes and to a lesser extent, water
pipes. It takes less time for installation, has lower cost than excavation and
replacement, and requires minimal excavation and disruption. It has the ability to
accommodate large radius bends and may improve hydraulics in spite of the reduction
in the overall pipe diameter. Most of the sliplining materials available are acid-resistant
and provide good hydraulics.
Deformed pipe insertion lining is a new process with a limited experience base.
Several variations exist In general, folded or compressed plastic pipe is inserted into
the existing pipe. Pipe material may be polyethylene or plasticized PVC. The pipe may
be heated to increase flexibility prior to installation. In one process, the heated pipe is
pulled through a die to reduce the diameter prior to insertion. In processes using
folded pipe, steam is added after insertion and a ball is propelled through the folded
pipe to expand it to conform with the existing pipe, or the pipe is expanded
hydraulically using steam pressure. In the process using the mechanical die to reduce
the diameter, the pipe naturally reverts to its original diameter within several hours.
After insertion and expansion, the ends of the liner are cut off, trimmed, and likely
sealed. Proprietors claim that the process results in a tight fit of the liner to the pipe,
eliminating the need for grouting of the annular space. Mixing of resins and curing, as
required with cured-in-place systems, are unnecessary. The folded liner systems are
supplied in rolls, and insertion pit excavation is not required. For the system which
reduces pipe diameter using a die, an insertion pit is apparently required as with
conventional sliplining. Because of the limited experience with deformed pipe insertion
methods for rehabilitating sewers, little information is available on the applicable size
ranges, ease of installation, and cost-effectiveness of these systems.
Cured-in-place pipe is formed from a resin-impregnated felt tube which is
inverted into an existing pipe and allowed to cure. After curing, the felt tube becomes a
smooth, hard pipe of slightly less diameter and of the same shape as the original pipe.
Cured-in-place pipe can be installed in pipes of all shapes up to diameters of 96 inches.
4-17
-------
It can also adhere to beads present in the original pipe. Inversion lining is successful in
dealing with a number of structural problems. Caution must be used in the application
of this method to any structural problems involving major loss of pipe wall, reinforcing
steel, or exterior pipe and bedding support Within limits, the liners can be designed to
deal with these more serious structural problems. Precautions should be taken in
determining the selection of an inversion lining method for sewer rehabilitation.
Various resins can be used to provide different degrees of acid resistance. Standard
polyester resins are suitable for most sewer applications. Cured-in-place inversion
linings have found more acceptance where minimal excavation and traffic disruption is
required. This technique is one of the most widely used rehabilitation methods. The
long-term acid resistance of the liner is unknown. Observations made within the past 15
years have not indicated corrosion problems.
Specialty concretes are sulfate-resistant cements applied to corroded surfaces
primarily for structural reasons and to resist corrosion. Sulfate resistant cements
include potassium silicate. Typically, reinforcing steel is added for additional support
Specialty concretes include cement mortar, shotcrete, and cast concrete. Cement mortar
is applied with a hand trowel for spot repairs in man-entry size (i.e., greater than 32-
inch-diameter) sewers or with a centrifugal lining machine for complete coverage within
a stretch of pipe. Shotcrete is applied with a special nozzle using compressed air. Cast
concrete is installed using prefabricated or hand-built interior pipe forms. The
development of mechanical, in-line application methods (e.g., centrifugal and mandrel)
has established mortar lining as a viable rehabilitation technique. Both shotcrete and
cast concrete are used in large-diameter sewers where adequate space is available to
handle materials and equipment The specialty concrete applications depend on the
degree of corrosion present and the structural integrity of the unit in question.
Generally, thin film concrete will perform best on relatively non-corroded concrete
whereas an elastic membrane concrete system will work for all cases. Testing performed
by CSDLAC revealed that seven out of eight supposedly acid-resistant specialty
concretes failed after submersion from 0.1 to 488 days in a 10-percent sulfuric acid
solution. Only one concrete has maintained its acid resistance even after 605 days of
testing. Specialty concretes are used mostly for spot repairs such as manhole barrels,
wet wells, junction chambers, and sections of pipe.
•» Coatings include a myriad of proprietary epoxies, resins, sealers, silicones,
urethanes, and coal tars applied by spray or brush. Coatings are experiencing rapid
growth with new products being marketed continually. Unfortunately, field testing has
not kept up with the rapid growth. CSDLAC have been testing new products, and cities
such as Seattle have been using those products that exhibited good results. Only about
25 percent of coatings tested by CSDLAC exhibited good acid resistance. The majority
of failures can be attributed to application difficulties (e.g., pinholes and blowholes).
All coating systems require some form of surface preparation. The surface preparation
and conditions under which coatings material is applied are extremely critical. The
specification of any acid resistant epoxy should require the minimum application of 1.5
4-18
-------
mm of material for rehabilitation. Consideration should be given to the use of coatings
that will cure underwater for projects that require either short down times, or where it
is impractical to completely remove the structure from service, thereby requiring coating
application to an intermittently wet area.
Liners used for rehabilitation may be prefabricated panels or flexible sheets that
are installed manually with anchor bolts or with concrete-penetrating nails; or
continuous, interlocking strips that are installed in a spiral fashion using a special
machine or manually. Manual methods are applicable only to man-entry sewers.
Common liner materials are fiberglass-reinforced cement, fiberglass-reinforced plastic,
PVC, and PE. The liner materials themselves are acid-resistant but problems have
occurred due to poor jointing. Liners may be susceptible to acid leakage due to
numerous joints. Some panel systems are time consuming to install and thus prolonged
bypass is required. Hydrogen sulfide gases have been documented to pass through poor
joints and cause failure by attacking the concrete substrate behind the liner. A recent
design introduced to the U.S. involves a continuous, helical, interlocking strip with
improved joints which may overcome such gas-penetration problems. Also, a new acid-
resistant urethylene mastic has shown excellent results in bonding PE and PVC sheets
to concrete surfaces, and may eliminate problems with mechanical anchoring and poor
jointing. Cracking of polyethylene liners has been observed in areas of high turbulence.
4.5 Conclusions
No material or technique is effective for controlling sulfide generation or sulfide-
induced corrosion in every situation. The environmental variables that determine the
success of a prevention or repair method (e.g., the characteristics of the wastewater and
the collection system and the severity of corrosion) vary with each system and within
each reach of sewer. It is only after a system has been analyzed that these variables can
be taken into account and an effective measure selected. Often, several different
methods must be used in combination or at different points in the same system to
combat corrosion under various conditions. The effectiveness of a method is deter-
mined by more than just its physical properties and/or theory of operation. Proper
design, installation, operation, and maintenance are required to ensure that the material
or technique is effective. However, even if studies indicate that a method has the best
long-term cost- effectiveness, initial costs may exceed the budgetary constraints of a
municipality and force it to use a less expensive and less effective method. Because of
the variations in effectiveness, affordability, availability, convenience, and applicability,
sulfide control techniques and corrosion rehabilitation methods must be evaluated on a
case-by-case basis.
4-19
-------
REFERENCES
1. "Design Manual - Odor and Corrosion Control in Sanitary Sewers and Treatment
Plants," EPA/625/1 -85/018, EPA, Cincinnati, OH, 1985.
2. "Sulfide in Wastewater Collection and Treatment Systems," ASCE Manuals and
Reports on Engineering Practice - No. 69, ASCE, New York, 1989.
3. "Sulfide and Corcosion Prediction and Control." American Concrete Pipe
Association, Vienna, VA, 1984.
4. Stahl, J.S., Redner, J., and R. Caballero, "Sulfide Corrosion in the Sewer System
of Los Angeles County," presented at 11th U.S./Japan Conference on Sewage
Treatment Technology, the Public Works Research Institute, Tsukuba, Japan,
October, 1987; and ASCE Conference on Sulfide Control in Wastewater
Collection and Treatment Systems, Tucson, AZ, February, 1989.
5. "No-Dig Technology Outline," prepared by National Association of Sewer Service
Companies, Altamonte, FL, 1989.
6. Schrock, B.J., "Pipeline Rehabilitation Seminar," Portland, Maine, August 8,
1988.
7. U.S. Department of Housing and Urban Development, Utility Infrastructure
Rehabilitation: Office of Policy Development and Research - Building
Technology, November, 1984.
8. Redner, John, A, "Evaluation of Protective Coatings for Concrete," presented at
the 59th Annual Water Pollution Control Federation Conference, Los Angeles,
CA, October 7, 1986.
9. Cronberg, Andrew T., Jack P. Morris, Ted Price, "Determination of Pipe Loss
Due to Hydrogen Sulfide Attack on Concrete Pipe," prepared for the 62nd
Annual Water Pollution Control Federation Conference, San Francisco, October,
1989.
10. "Concrete Pipe Handbook," American Concrete Pipe Association, Vienna,
Virginia. 1988.
4-20
-------
APPENDIX A
ANNUAL AVERAGE WASTEWATER
CHARACTERISTICS FOR LA COUNTY
1971-1986
-------
x ear
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
July 1987
JOINT WATER POLLUTION CONTROL PLANT
RAW SEWAGE PARAMETERS
1971 - 1986 YEARLY AVERAGES
(Based on Water Quality Characteristics Monitoring Program)
Alkalinity
Total (mq/l)
Arsenic
(mq/1)
Barium
(mq/l)
BOD
Total
(mq/l)
Boron
(mq/l)
307
0
-
384
1.03
302
0
-
319
1.11
316
0 .0250
0 .53
357
1. 14
298
0.0354
0.55
314
1.35
289
0 .0155
0.75
' 302
1.49
307
0.0073
1.07
306
1.54
330
0 .0114
0.91
334
1.51
322
0.0135
0.78
324
1.64
316
0.0188
0.67
322
1.50
314
0.0064
1.02
335
1.52
317
0 .0067
0.80
322
1.41
323
0.0079
0.82
313
1. 66
340
0.0087
0.91
291
1.68
338
0.0257
0.83
317
1.76
329
0 . 0180
1.03
329
1.72
339
0.0101
0.97
328
1.58
-------
JWPCP
1971 -
- Raw Sewage
1986 Yearly
Parameters
Averages
Page 2
Year
Cadmium
(mg/1)
Chloride
(mq/1)
Chromium
Hexavalent
(mg/i)
Chromium
Total
(mg/i)
COD
Soluble
-------
JWPCP - Raw Sewage Parameters
1971 - 1986 Yearly Averages
Page 3
xear
COD
Total
(mg/1)
Conductance
MMHO
cm
CODDer
(mg/1)
Cyanide
(mg/"i)
DDT
Total
(mg/1)
1971
680
2992
0.450
0 . 200
0.01527
1972
721
2813
0.736
0.293
0.02132
1973
925
2785
0.563
0. 363
0.01802
1974
844
2369
0. 635
0. 430
0.00278
1975
818
2145
0.580
0.280
0 .00172
1976
1022
2162
0. 430
0. 290
0.00325
1977
863
2353
0.430
0.240
0.00211
1978
923
2185
0.360
0.180
0. 00273
1979
910
2312
0.337
0. 178
0.00234
1980
813
2273
0.334
0. 118
0.00177
1981
759
2355
0. 268
0.080
0.00172
1982
916
2524
0. 230
0 .063
0.00076
1983
904
2582
0.245
0 .042
0.00049
1984
900
2679
0. 240
0. 040
0.00096
1985
881
2632
0.197
0.020
0 .00037
1986
879
2620
0. 179
0.022
0.00019
-------
JWPCP - Raw Sewage Parameters
1971 - 1906 Yearly Averages
Page 4
Year
Detergents
(MBAS)
(nig/ 1)
Flow
(MGD)
Fluoride
(mq/1)
Hardness
Calcium
(mq/1)
Hardness
Magnesium
(mq/1)
1971
7.23
372
1.05
273
165
1972
7. 40
351
1.24
259
162
1973
7. 49
359
1.51
215
143
1974
6.60
347
1.40
215
109
1975
6.50
342
1.37
176
81
1976
6.74
353
1.48
165
79
1977
6.31
330
1.45
197
97
1978
8. 17
345
1. 56
197
105
1979
7.62
367
1.66
191
94
1980
6.97
374
1.56
191
32
1981
5. 97
364
1.51
204
97
1982
6.20
360
1. 39
194
94
1983
5.67
353
1.99
210
101
1984
6.24
352
2.61
211
104
1985
6.44
361
2. 17
192
99
1986
6. 10
364
2.01
183
98
-------
JWPCP
1971 -
- Raw Sewage
1986 Yearly
Parameters
Averages
Page 5
Aear
Hardness
Total
(mg/1)
Iron
(mg/1)
Lead
(mq/1)
Lithium
(mq/1)
Manganese
(mq/1)
1971
438
13.130
0. 280
0.070
0. 11
1972
421
11.340
0.306
0.070
0 .12
1973
371
8.875
0. 292
0.060
0. 11
1974
350
9 . 480
0.371
0.050
0.12
1975
282
13.950
0.370
0.042
0.15
1976
259
6 .840
0 . 272
0.058
0.10
1977
294
8. 290
0. 340
0. 100
0. 12
1978
265
7.562
0.270
0 .059
0.13
1979
298
7. 367
0.225
0.061
0.11
1980
294
6. 4 29
0 .212
0.053
0 . 10
1981
29 6
6. 106
0. 164
0.063
0.09
1982
289
5.430
0.160
0.066
0 . 09
1983
307
7.840
0.150
0.063
0. 12
1984
310
5. 290
0.140
0.071
0 .11
1985
292
4.870
0.129
0.071
0. 11
1986
287
5. 138
0.155
0.065
0 .10
-------
JWPCP -
1971 -
Raw Sewage
1986 Yearly
Parameters
Averages
Page 6
Year
Mercury
(mq/1)
Nickel
(mq/i;
Nitrogen
Ammonia
(mq/1)
Nitrogen
Organic
(mq/1)
Oil
and Grease
(mq/1)
1971
0.0022
0. 230
80.6
26.7
-
1972
0.0010
0. 310
41.3
17.4
-
1973
0.0012
0.324
58.5
20.8
-
1974
0.0012
0. 280
34.4
16.6
-
1975
0.0014
0. 280
33.6
18.8
91
1976
0.0015
0.340
32.9
18 .9
219
1977
0.0014
0. 310
33.6
17.9
124
1978
0.0014
0.342
34. 4
21.0
91
1979
0.0014
0.245
34.2
19.6
90
1980
0.0011
0.245
34. 2
20. 2
76
1981
0.0009
0. 210
34.0
20.3
62
1982
0.0011
0. 200
34. 6
19 .6
73
1983
0.0013
0. 220
35.0
23.1
69
1984
0.0011
0. 150
31. 6
21. 1
79
1985
0.0010
0 .129
31.9
19.5
72
1986
0.0011
0. 099
34.1
21.1
64
-------
JWPCP
1971 -
- Raw Sewage
1986 Yearly
Parameters
Averages
Page
Year
PCB Total
(mq/1)
PH
Phenols
(mq /1)
Phosphate
(mg/1)
Po tas
(mq.
1971
0.02126
7.92
3.83
45.8
20
1972
0.01077
7.61
2.51
59.5
20
1973
0.01233
7.65
4.64
53.9
19
1974
0.01685
7.55
3.27
52.2
19
1975
0.00531
7.51
3.74
47.5
19
1976
0.00061
-
3.17
38.6
20
1977
0.00242
-
3.02
24.5
19
1978
0 .00180
-
2.69
13.3
19
1979
0.00098
-
2.40
11. 5
19
1980
0.00092
-
2. 28
11.5
16
1981
0.00070
-
2.51
10.7
18
1982
0.00081
-
2.07
10.6
18
1983
0.00061
-
2.37
11. 6
18
1984
0.00064
-
1.93
12.6
13
1985
0.00031
-
1.70
12.4
18
1986
0
2.30
13.2
17
-------
JWPCP -
1971 -
Raw Sewage
1986 Yearly
Parameters
Averages
Page
8
Year
Selenium
(mq/1)
S ilver
(mq/1)
Calculated4
Sludge
Concentration
(mq/1)
Calculated4
Sludge
Mass
(lbs/d)
Sodium
(mq/1)
1971
-
0.0140
-
-
369
1972
-
0.0170
-
-
381
1973
-
0.0126
-
-
357
1974
-
0.0109
58.57
169 , 656
336
1975
-
0.0140
57. 57
164,320
308
1976
0. 015
0.0140
74.38
219,122
306
1977
0.016
0.0106
76.47
210,597
336
1978
0. 009
0.0191
76.61
220,572
357
1979
0.012
0.0197
72.00
220 , 524
331
1980
0.010
0.0153
78.52
245,075
333
1981
0.009
0.0150
78. 39
238,127
338
1982
0. 014
0.0175
92. 27
277,223
368
1983
0.016
0.0182
115.44
340,082
369
1994
0. 015
0.0170
120.99
355,420
407
1985
0.013
0.0200
108.33
326,353
395
1986
0. 017
0.0198
107.54
326,691
378
-------
TWPCP - Raw Sewage Parameters
l971 - 1986 Yearly Averages
Page 9
ear
Solids
Total
(mg/1)
Sulfate
(mq/1)
Sulfide
Total
(mq/1)
Sulfide
Dissolved
(mq/1)
Suspended
Solids
(mq/1)
.971
2112
347
0.4
0 .1
397
.972
2040
330
0.5
0.2
416
.973
1975
349
0.5
0.2
518
.974
1828
320
0.5
0.2
459
.975
1681
258
0.5
0.2
484
.976
-
224
0.6
0. 3
424
.977
-
260
1.0
0.5
463
.978
-
270
1. 2
0.6
448
.979
-
275
1.3
0.6
435
.980
-
275
1.3
0.7
442
.981
-
286
1.6
0.7
442
982
-
240
1.6
0. 6
442
.983
--
282
1.6
0 . 6
463
.984
-
296
2.0
0.8
455
.985
-
268
2.6
1.2
445
986
—
285
3.0/2.0
1.4/0.6
454
-------
JWPCP - Raw Sewage Parameters
1971 - 1986 Yearly Averages
Page 10
Year
TDS
(Jtiq/1)
Tftia££.anate
(ma/1)
TICH
(mcf /1)
Zinc
1971
1716
-
0.03680
1.930
1972
1624
-
0.019 8.6
2.269
1973
1458
2.76
0.03032
2.470
1974
1368
2.00
0.02157
1.990
1975
1199
1.80
0.00724
1.640
1976
1191
2.00
0.00456
1. 420
1977
1300
1.89
0.00459
1.460
1978
1404
2.34
0.00470
1. 278
1979
1262
2.11
0.00316
1.000
1980
1259
1.18
0.00302
0.952
1981
1333
1. 27
0 .00281
1.012
1982
1318
1.53
0.00227
0. 960
1983
1474
1.50
0.00166
1.011
1984
1523
1.51
0.00318
0.830
1985
1522
1. 09
0.00088
0.720
1986
1426
1.42
3.000 40
0.751
-------
JVJPCP - Raw Sewage Parameters
l971 - 1986 Yearly Averages
Page 11
'otes:
1. The values in the tables are based on the Water Quality
Characteristics Monitoring Program and were used for
the correlation with dissolved sulfide. The values,
in some cases, do not exactly match the graphs and
slides previously prepared. Some of the data for the
graphs and slides were from sampling conducted for the
Industrial Waste Section. However, the industrial waste
data is incomplete, so to be consistent, only data from
the Water Quality Characteristics Monitoring Program
were used for the correlation.
2. Total solids and pH were not correlated with dissolved
sulfide because of insufficient data.
3. Sludge mass was not correlated with dissolved sulfide
because it doesn't account for any change in flow.
Sludge concentration was used instead.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
February 1988
CORRELATION OF JWPCP RAW SEWAGE PARAMETERS
WITH DISSOLVED SULFIDE FROM 1971-1987
Correlation
Parameters
Coef f ic:
Nickel
-0 .83
Chromium-Total
-0.81
Sludge-Concentration (theoretical)
0.78
Zinc
-0.78
Copper
-0.77
Cyanide
-0.77
Lead
-0 .76
Iron
-0.70
Potassium
-0.69
Silver
0.69
Fluoride
0.68
Phenols
-0.67
Alkalinity-Total
0.67
Phosphate
-0.66
Sodium
0.63
TICH
-0.61
PCB-Total
-0.60
Barium
0.57
DDT-Total
-0.53
Boron
0.51
Oil and Grease
-0. 49
Mercury
-0.48
Nitrogen-Ammonia
-0.43
Hardness-Calcium
-0.43
COD-Soluble
-0.4 2'
Cadmium
-0.40
Detergent (MBAS)
-0. 39
Hardness-Magnesium
-0 . 38
Selenium
0 . 37
Sulfate
-0 .33
Flow
0.33
Maganese
-0.32
Thiocyanate
-0. 28
COD-Total
0.28
Hardness-Total
-0 . 21
Chloride
0 . 20
Chromium-Hexavalent
-0. 13
Lithium
0 . 12
Nitrogen-Organic
0. 11
Suspended Solids
0.07
Conductance
0.05
Temperature
-0.05
Pt. "-Total
0.04
TDS
0.02
Arsenic
-0 .02
All parameters correlated with dissolved sulfide without sulfide
control except for Iron which was correlated with dissolved
sulfide with sulfide control in 1986 and 1987.
-------
REPORT DOCUMENTATION PAGE
Form Approved
OMB No. 0704 0188
Public repeung fcjrden 8
298-102
------- |