ALTERNATIVE SEWER STUDIES
-------
ALTERNATIVE SEWER STUDIES
by
Urban Systems Research & Engineering, Inc
Cambridge, Massachusetts 02140
Contract No. 68-03-3057
Project Officers
James F. Kreissl
Robert P.Q. Bowker
Wastewater Research Division
Water Engineering Research Laboratory
Cincinnati, Ohio 45268
WATER ENGINEERING RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
The Information in this document has been funded wholly by the United
States Environmental Protection Agency under Contract No. 68-03-3057 to
Urban Systems Research and Engineering, Inc. It has been subject to the
Agency's peer and administrative review, and it has been approved for
publication as an EPA document.
ii
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with
protecting the Nation's land, air, and water systsnis. Under a mandate of
national environmental laws, the agency strives to formulate and imple-
ment actions leading to a compatible balance between human activities and
the ability of natural systems to support and nurture life. The Clean
Water Act, the Safe Drinking Water Act, and the Toxics Substances Control
Act are three of the major congressional laws that provide the framework
for restoring and maintaining the integrity of our Nation's water, for
preserving and enhancing the water we drink, and for protecting the
environment from toxic substances. These laws direct the EPA to perform
research to define our environmental problems, measure the impacts, and
search for solutions.
The Water Engineering Research Laboratory is that component of EPA's
Research and Development program concerned with preventing, treating, and
managing municipal and industrial wastewater discharges; establishing
practices to control and remove contaminants from drinking water and to
prevent its deterioration during storage and distritJtion; and assessing
the nature and controllability of releases of toxic substances to the
air, water, and land from manufacturing processes ar J subsequent product
uses. This publication is one of the products of th-.t research and
provides a vital communication link between the res. ircher and the user
community.
This report provides additional design and operational information
on two of the most effective and widely applied alternative sewer systems,
i.e., smal1-diameter gravity and pressure sewers, over and above previously
published reports. The information provided herein will assist system
designers and operators to avoid and/or rectify problems resulting from
sulfides and downhill hydraulics which could otherwise represent major
impairments to the successful application of the^e technologies.
Francis T. Mayo, D-"rector
Water Engineering Research Laboratory
iii
-------
ABSTRACT
This study deals with design and operational considerations not
practically addressed in previous reports on smal1-diameter gravity (SDG)
and pressure sewer systems. These issues are sulfide generation and
hydraulic design of downhill runs.
Sulfide generation is demonstrated to be a problem in both types of
pressure sewers, i.e., grinder-pump (GP) and septic tank effluent pump
(STEP), and SDG sewers. The highest values were found in pressure sewers,
but sulfides were also found in significant concentrations in the SDG
sewer. Other than the verification of substantial sulfide concentrations
in these types of sewers, no quantitative modeling of sulfide profiles
in sewers was attempted.
Early pressure sewer designs were based on water supply system
design technology. Those systems which were not generally flat or rising
in profile have experienced numerous problems due to air introduction and
accumulation in the mains. This study details the development and full-
scale results of an improved design approach for downhill portions of
pressure and SDG sewers to minimize air accumulation and its attendant
operational problems.
This report was submitted in partial fulfillment of Contract No.
68-03-3057 by Urban Systems Research and Engineering, Inc. under the
sponsorship of the U.S. Environmental Protection Agency. This report
covers a period from March, 1981 to July, 1983, and work was completed
as of September, 1983.
iv
-------
CONTENTS
Foreword Ill
Abstract iv
Figures vi
Acknowledgement vii
1. Introduction 1
2. Conclusions 3
3. Recommendations 5
4. Sulfide Generation in Alternative Sewers 6
5. Two-Phase Flow in Pressure and Small-Diameter Gravity Sewers 14
References 38
-------
FIGURES
Number Page
1 Cross-section at pump Installation where pump Is above 16
HGL of main '
2 Profiles of uphill and downhill flow In sewers 17
3 Air-bound hydraulic gradient 18
4 Pipe check and vent 21
5 Air pocket and bubbles encountered in downhill flow 22
6 Test apparatus using transparent piping 24
7 Profile of two mile test section of glide pressure sewer 27
system, ignoring air consideration
8 Measured HGL at 600 gpm 28
9 Pressurized static HGL 29
10 Pipeline .profiles at stations 69 and 81 30
11 Measured HGL @ 600 gpm after alterations 32
12 Pipeline profile with numerous downhill flow sections 33
13 Standpipe 34
14 Pipeline profile using standpipes, showing static and 35
dynamic HGL's
vi
-------
ACKNOWLEDGEMENTS
A number of people contributed to the preparation of this report.
Special thanks go to the Project Officers and staff of the Water Engi-
neering Research Laboratory, James F. Kreissl, Robert P.G. Bowker, and
James W. Cox. Their guidance and direction have been essential to en-
suring the quality of the project and report. Particular note should
be made of the major contributions of Mr. W.C. Bowne in developing much
of reported material and to the staff of Rural Systems Engineering, Inc.
for their substantive and review contributions. Finally the physical
development of this report was the responsibility of Patricia L. Deese,
Susan Farrell, and David Bur-master of USR&E and Kathryn G. Wiker of the
USEPA.
vii
-------
SECTION 1
INTRODUCTION
Small communities in need of new or expanded sanitary sewers are faced
with a severe financial burden. Low densities and unfavorable geological
situations increase per capita costs of conventional sewers, which often
account for up to 80% of the total capital costs of a new wastewater man-
agement system (9).
Conventional sewers are expensive. In order to insure that raw sewage
flows freely conventional sewer systems use large diameter pipes set in
the ground at minimum slopes. Often pumping stations are required as well.
Extensive excavation is usually necessary to achieve the desired slopes.
Flat terrain, high groundwater, and waterfront areas all add to construc-
tion costs and difficulties. Finally infiltration and inflow (141) can-
not be eliminated in large pipes. The added wastewater volume and solids
mean that the treatment plant must have a greater capacity than would be
required to only treat the sewage.
Alternative approaches to sewering that address some of the problems
encountered with conventional systems can reduce collection and treatment
costs. Three types of alternative sewers are discussed below:
o small diameter gravity sewers (SDGS),
o pressure sewers and
o vacuum sewers.
Small diameter pipes reduce excavation and construction costs. Vacuum
or pressure sewers offer the dual advantages of small diameter pipes and
the ability to follow the natural topography without risk of clogging which
further lowers excavation costs. Furthermore all three of these sewers
provide a system virtually sealed against infiltration inflow.
The two major types of pressure sewers are grinder pumps (GP) and
septic tank effluent pumps (STEP). These two systems differ in the onsite
equipment, layout, and the quality of the wastewater conveyed to the pressure
sewer. In GPs solids are ground to a slurry and discharged through pressure
lines. In STEPs wastewater from a home first flows into a septic tank from
which treated effluent is pumped to pressurized lines.
-------
Vacuum sewers use a central vacuum source to constantly maintain a
vacuum on small-diameter collection mains. Periodically the pressure
differential created by the vacuum source draws a slug of sewage from a
holding tank at each home into the line. When sufficient volume of sewage
is collected at a central vacuum station, it is pumped to the treatment
plant or main interceptor.
Like the STEP system, a small-diameter gravity (SOG) sewer is used
with individual septic tanks. Because solids are removed by the septic
tank, pipes as small as 4 inches in diameter can be used at very shallow
slopes without risk of clogging. The effluent requires little or no
pumping and flows by gravity to the treatment facility.
Despite their many advantages, several concerns have been raised
about alternative sewers. The most important of these potential problems
are:
o excess sulfides generation;
o two-phase flow in pressure and SOG sewers.
Sulfides generation effects all types of sewers. The problem mani-
fests itself in unpleasant odors and corrosion produced by hydrogen sul-
fide. Years of experience have gone into the design of conventional
sewer systems to minimize sulfide generation. In 1974 EPA published a
design manual setting forth the best engineering practice (7).
Based on experiences with sulfides in conventional sewer systems,
two main concerns have been raised with regard to alternative systems.
Is the septic effluent in SDG's and STEP systems more prone to sulfides
generation, and does the anaerobic nature of pressure and SDG sewers
contribute to sulflde generation?
Two-phase flow refers to a hydraulic problem of particular concern
in pressure systems. In downhill sloping sections of pressure sewers,
gas bubbles present in the pipeline can adversely affect flow. The
typical solution is to install air release valves at summits within the
pipeline. However, In many cases, this action does not work and addi-
tional steps must be taken to solve the problem.
-------
SECTION 2
CONCLUSIONS
Although a significant number of pressure and small-diameter gravity
(SDG) sewers have been designed and constructed there remain some signifi-
cant gaps in understanding the technology. These studies provided some
insights as to two of the major technology gaps. The major conclusions
are:
1. Grinder-pump (GP) systems can produce" sul'fides~at__a_rate of 3
to 4 times that of septic tank effluent pumping (STEP) systems
and about twice that of conventional sewer force mains due to
the high organic strength of the wastewater.
2. There appear to be unexplained losses of sulfide and gains in
dissolved oxygen in STEP systems based on analyses performed
in this study and previous data for septic tank effluents.
3. Both pressure systems, i.e., GP and STEP, can be expected to
have some sulfide concentration in their wastewaters, with value
varying from 1 to 14 mg/L, based on this study.
4. GP sulfide concentrations will generally increase in the direc-
tion of mainline flow, but random locations of service lines and
branches may mask this trend.
5. Concentrations of sulfides in pressure sewers cannot as yet be
quantifiably predicted, due to the empirical nature of the
available equations and their derviation for weaker conventional
wastewaters.
6. SDG sewers should not be designed to minimize pipe sizes, to
flow full for substantial periods, or to proliferate substantial
periods, or to proliferate substantial inundated sections of
mainline if sulfide minimization is desired.
7. For conventional gravity sewers, equilibrium sulfide concentra-
tions result, from long pipe segments with relatively uniform
conditions. Applying the equilibrium equation for conventional
gravity sewers to SDG sewers results in concentrations comparable
to those observed in SDG sewers. These concentrations are much
lower than the levels reported to occur in septic tanks.
8. Due to the phenomenon of #7 above, SDG sewers appear capable of
producing terminal wastewater sulfide concentrations lower than
-------
those of pressure sewers.
9. Conventional placement of air-release valves at high points of
a pressure sewer system does not preclude the entrainment of air
and results in headlosses greatly exceeding design calculations.
10. In downhill runs where the pressure main intersects the dynamic
hydraulic grade line (HGL) a hydraulic jump is formed which
generates gas bubbles to downstream segments of the main.
11. Placement of sewage-type automatic air release valves at points
at least 14 pipe diameters below hydraulic jump locations was
effective in removing entrapped air and reducing headlosses to
near theoretical levels. =— —__—•— -.
12. Backpressure sustaining valves were found to be inadequate for
control of downhill hydraulics in the pressure sewer owing to
high capital cost, intensive maintenance requirements and un-
reliable operation.
13. On downhill runs with irregular terrain which provide numerous
opportunities for the formation of smaller hydraulic jumps
standpipes were found to be inexpensive and reliable. The stand-
pipes employed large diameter downlegs to prevent the escape and
conveyance of air bubbles into the downstream segment of the
mains and automatic air-release valves at their summits to expel
the trapped gases.
14. Soil absorption beds were successfully used for the vented
gases from the air-release valves to prevent odors from hydrogen
sulfide.
-------
SECTION 3
RECOMMENDATIONS'
There is a need to improve the designer's capability to predict
sulfide concentrations in pressure and SDG sewers. To accomplish this
end comprehensive studies of sulfides should be made from the septic tank
to the terminus of a number of these systems to identify the gains and
losses of sulfide concentration, to quantify the mechanisms responsible,
and to develop predictive equations. Once this is accomplished, a study
to develop a corrosion-based methodology for evaluating the alternatives
of designing a transition station or modifying a receiving conventional
sewer should be undertaken. Such a methodology would provide a quanti-
tative solution to one of the major design obstacles in the design of
pressure and SDG sewers which terminate at larger conventional sewers.
There is also a need to more quantitatively assess the need for
location of, and design and 0/M requirements for air-release valves in
pressure and SDG sewer systems, along with the use of soil absorption and
other low-maintenance odor control methods appropriate for these alterna-
tive sewers.
-------
SECTION 4
SULFIDE GENERATION IN ALTERNATIVE SEWERS
INTRODUCTION
Hydrogen Sulfide (^S) Is a gas that can be found In any type of
sewer. Hydrogen sul fides and other sulfur-containing substances such
as thiosul fates can cause two main problems In sewers. They can be a
source of noxious odor, and they can cause corrosion. H2S is also very
toxic. Workers have been known to succumb to H2$ poisoning when concen-
trations were as low as 300 ppm in the sewer air.
is formed by the action of anaerobic bacteria on sulfates dis-
solved in the wastewater. These dissolved sulfates may be indigenous
to the water supply, or be themselves the product of microbial breakdown
of waste proteins. Since most sewers contain some oxygen, sul fide genera
tion can only occur in a relatively small part of the sewer that is
anaerobic: the layer of slime growing on the walls. Sulfides are formed
in a narrow band within the slime — deep enough so that oxygen does not
penetrate, yet shallow enough so that the bacteria can still obtain the
organic nutrients they need to survive.
The amount of sul fides generated depends on the amount of sul fate
in the water, pH, the available organic carbon, and the thickness of the
slime layer. The thickness of the slime layer depends in turn on the
composition and abrasiveness of the sewage, and the velocity of the flow.
concentrations will only increase if the gas is created faster
than it is destroyed. Under most circumstances, the sul fides formed
within the slime are reconverted back to sulfates by aerobic bacteria
living on the slime surface. Because of this, H2S levels do not usually
build up in sewers. Gas concentrations will only rise when the DO levels
in the sewer are less than .1 mg/L.
Designers of sewer systems have had to cope with sul fides generation
in conventional designs for many years. Some researchers have theorized
that small diameter pipes, pressurization, septic tank pre-treatment,
grinder pumps and frequent intersection with large diameter gravity col-
lectors amplify the problems of sulfide generation. This section presents
the reasons why differences in sulfide generation do or do not exist, and
identifies gaps in our current knowledge.
Designers have developed many ways to control sul fides. The tech-
niques that have proven effective involve either preventing the anaerobic
-------
conditions conducive to sulfide formation or limiting exposure of the
waste stream to the air where sulfides can be released as a gas.
This portion of the study involved onsite sulfide monitoring in three
different alternative sewer systems, review of the monitoring results by
.Richard D. Pomeroy, an internationally known expert in the field, and
finally, a simple analysis of how the physical factors that affect sulfide
generation in conventional sewers compare to alternative sewers.
SULFIDES GENERATION IN THREE DIFFERENT ALTERNATIVE SEWER SYSTEMS
It is difficult to generalize across the range of alternative systems.
This report does not attempt a comprehensive discussion of all types, but
rather-reviews-sul_fideHConditions found during this study in three separate
systems. " .
Quaker Lake Pa. Grinder Pump System
The Quaker Lake sewer system consists of two separate lines serving
39 and 71 residences, respectively, carrying a total of 18,000 gallons/day.
Each house has a wet well in which a grinder pump is located. As sewage
fills the wet well, it activates the grinder pump which forces the waste
into the pressure mains. The lines are designed to flow under intermittent
pressure. The two lines flow to two 750,000 gallon aerated lagoons. The
system was installed in 1977 and there have been no reported corrosion or
odor problems since that date.
Over a two day period in August 1981, four rounds of samples were
taken at six stations along the north branch of the system which serves
39 homes. The samples were analyzed for total sulfides, dissolved sulfides,
sulfates, BOD, temperature, and pH. The dissolved sulfide tests were made
on samples preserved by the use of zinc salt, which converts dissolved
sulfide into insoluble zinc sulfide. This error does not impair the value
of this report for its intended purpose because insoluble sulfide is
usually only a few tenths of a mg/L in domestic sewage.
The average sulfide concentrations (mg/L) at six stations along the
collector, starting at the upstream station were: 9.4, 10.1, 6.7, 8.9,
5.4, 9.0. (Sulfide concentrations in sewers are typically erratic). The
sewage temperatures averaged 18.4°C. The pH averaged 7.06.
For sulfates it was found, in line with experience elsewhere, that
the results obtained by the standard method normally used for water anal-
yses gave unreliable results when applied to wastewaters. The reasons
are not clear, but until dependable methods are found for determining
sulfate in sewage, the test should be omitted. The sulfate concentration
has little to do with the rate of sulfide generation, but sometimes it is
a limiting factor on the amount of sulfide that can be produced.
-------
BOD determi nations showed erratic results. The average was 390
mg/L, which is not exceptional for ground domestic sewage near the source.
Glide. Oregon; Septic Tank Effluent Pump System
The sewage first passes through septic tanks and then is pumped into
the pressure main and thence to the large oxidation ponds of the City of
Glide. There are numerous pressure relief valves. The profile is gener-
ally downhill, but there are grade reversals. Falls occur within the
mainline at five locations, where oxygen may be absorbed. Despite the
fact that the system Is not airtight, there have been no odor or corrosion
problems since this system was installed in 1975.
____ Jests for sulfide, BOD, temperature and pH were made at four sampling
statioris-along the main-trunk. On September 23, two rounds of tests were
started at 4:45 and 7:20 a.m. and on September 24, rounds were started at
6:05 and 7:10 a.m., each round requiring about two hours.
There was not a significant difference between the earlier and later
tests. The average measured sulfide concentrations for the four stations,
starting at the uppermost, were 7.1, 7.2, 9.2, and 8.5 mg/L. Average
temperature of the sewage was 18°C. The pH averaged 6.95. BOD values
appeared to reduce with collection system length, while both total and
dissolved sul fides increased.
Westboro, Wisconsin; SPG System
This is a complex system, including small diameter gravity (SDG)
sewers, a pump station and pressure main, and septic tanks pumping into
small diameter pressure mains. Single sulfide tests are reported for
samples taken in August from nine manholes. Sulfide concentrations
ranged from 1.25 to 3.9 mg/L averaging 2.7 mg/L.
It would be difficult to draw any generalization from a small number
of tests, or even from a considerably greater number, in such a complex
system.
FURTHER DATA COLLECTION
Sewage is a highly variable material, both in time and place. It is
difficult to quantify all of the factors that influence the rate at which
a wastewater stream will generate sulfide. The -"practical" approach to
finding out how various system designs influence sulfide conditions by the
procedure of measuring sulfide concentrations in existing systems is not
really a very practical approach unless a very large number of systems are
studied, with observations being made on many variables other than those
reported.
Two essentially different sources of sulfide may exist in these
systems: septic tanks and sewers. There can be wide differences in the
8
-------
condition of septic tank effluents. In one pressure system (in California)
the effluents of septic tanks carry an average of about 50 mg/L of sulfide.
By contrast there is a large Third World city in which all of the build-
ings have septic tanks. Only a small part of the city is sewered. Else-
where in the city, the effluents discharge to channels and storm drains.
No more than a few tenths of a mg/L of sulfide was found in any of those
flows. This was very puzzling until it was realized that the regular
pumping of the tanks was only theoretical; in reality they are very rarely
pumped. When a septic tank is full of sludge, the water maintains a
channel throught the .tank. Because of the slowness of molecular diffusion,
very little sulfide can diffuse out into the wastewater stream.
In an alternative system that employs septic tanks, a comprehensive
view of sulfide conditions would include information on the sulfide con-
centrations in the septic tank effluents. Also, transit times of the
sewage in the pipelines should be observed by use of dyes. Sulfide
build-up rates could be compared using equations such as those presented
i n the EPA Sulfi de Control Manual.
Altogether, it appears that it would be a formidable task to extend
this project to a degree that would support any quantitative generali-
zations. It is unlikely, as these three cases point out, that any lesser
effort would produce statistically significant data.
THEORETICAL COMPARISON OF SULFIDE GENERATION IN ALTERNATIVE AMD
CONVENTIONAL SEWERS
Nearly all of the sulfide produced in small pipes comes from the
slime layer on the wall of the pipe. The amount of sulfide produced
upstream from a given point is proportional to the upstream area of pipe
wall. If the houses contributing flow to the pipeline are more or less
uniformly distributed, the flow in the pipeline is expected to be roughly
proportional to the distance. If the pipe were uniform in diameter, the
wall area, and hence the amount of sulfide, would increase in porportion
to the distance, and the sulfide concentration would remain about the
same. Actually, the diameter of the pipeline in the Quaker Lake system
is a little greater at the downstream end, so the sulfide concentrations
might increase slightly toward the end. The randomness of the flow pre-
vents an accurate comparison, but it can be said that there is no clearly
demonstrated trend.
From various research studies on this process in conventional sewer
pressure mains, an empirical equation was derived to show the expected
rate of sulfide build-up in the slime layer on the pipe walls and in the
waste stream:
-------
d[S] = 3.28 M [EBOD] (1 + 0.48 r)
~3F~ r
in which:
dCS] = rate of change of total sulfide concentration, mg/L - hr
dt
M = specific sulfide flux coefficient, m/hr
EBOD = effective BOD5 = BOD5 (1.07) (T-20) (mg/L)
T = temperature, 8C
r = hydraulic radius * d/4 (m)
3.28 = conversion factor (ft/m)
(l+0.48r) = empirical factor for sulfide production in the stream
Where the dissolved oxygen concentration is not more than a few
tenths of a mg/L, M is normally about 0.001. Sulfide buildup rates can-
not be predicted precisely, because the quantities measured by the BOD
test are not precisely proportional to the sulfide generating nutrients.
The calculation, using the- coefficients suggested, yields results that
will exceed the actual buildup most of the time. The currently available
data are not adequate to support a calculation of the expected sulfide
concentrations in the alt;rnative systems that were studied.
The factors which contribute to (and can limit) sulfides production
include:
o availability ot nutrients (EBOD),
o sulfates, presumed to be available in sufficient concentration,
o favorable temperature conditions,
o no inhibiting biological conditions, and
o anaerobic conditions.
Assuming the last three conditions are met, the maximum amount of
sulfide which can be generated, or the Sulfide Potential, is dependent
on the quality of the influent wastewater, the hydraulic conditions of
the sewer and temperature. Special attention should be given to sulfide
generation in areas with high sulfate concentrations in the water supply.
The rate at which,sulfate is converted to sulfide is dependent on
10
-------
the conditions within the sewers. The rate of change of sulfide concen-
tration in pressure sewers is expressed by the above empirical formula.
It is important to note that this formula is based on observation of
sulfide generation in conventional system force mains. Since the formula
is empirical, it is only applicable for conventional force mains. However
it is interesting to review this formula and determine if what is known
about the parameters affecting sulfides generation in conventional forc«
mains enables us to forecast significantly different patterns in pressure
sewers.
In comparing the basic pressure sewer concept with the equation and
other information about the technology, the following appear to be
reasonable assertions:
1. Based on EBOO, the rate of. sulfide production in GP systems is
about'3 to 4 times that in'STEP systems and nearly twice that
in conventional force mains, all other conditions being equal.
2. If septic tank effluent sulfide concentrations are as high as
reported elsewhere (12), I.e., 20-100 mg/L, the fact that in-
system and system effluent concentrations are reported herein
and elsewhere (1) (2) to be in all cases less than 15 mg/L and
normally less than 10 mg/L, there is a consistent loss of sul-
fide in STEP systems which is not yet obvious. These systems
also consistently display a low-level dissolved oxygen (<0.5 mg/L),
the source of which is not clear. More thorough sampling and
study are required to explain these phenomena.
3. Grinder-pump systems generally receive a wastewater which is
devoid of significant sulfide concentration, but consistently
displayed in the systems evaluated herein a minimum mainline
sulfide concentration of 1.0 mg/L, progressing higher in the
direction of mainline flow to values as high as 14 mg/L. These
results are consistent with the predictive equation, although
not predictable due to varying flows and input locations.
The basic equation for the change of sulfide concentration with
time for gravity sewers employs the same parameters as the above equation,
without the empirical factor for sulfide production in the stream, in
the first term and a sulfide reduction or loss term. When equilibrium
conditions occur in the gravity sewer, e.g., a long segment with continuity
of conditions affecting sulfide concentration, the gain and loss terms
become equal and the equation is expressed as follows (13):
Se = CiCEBOD] P
(SV)0.375 b
in which
Se = equilibrium sulfide concentration, mg/L
11
-------
GI = coefficient, in English units = 0.00052
S = slope
V = velocity, ft/sec
P = wetted perimeter, ft
b = liquid surface width, ft
Under a given set of 'hydraulic conditions, sulfides will approach (Se)
as an asymptote, either in an increasing or a decreasing mode (13).
Based on the equation for conventional gravity sewers and other known
information on SDG sewers, the following statements appear reasonable:
1. SDG sewers should not be designed to minimize pipe sizes or to
flow full for substantial periods.
2. The mainline sulfide concentrations found at Westboro which are
substantially lower than reported septic tank effluent levels
(12), are consistent with the above equation and sewer reaeration.
Whether these concentrations are representative of other SOG
sewers due to its unique design with multiple manholes remains
to be determined, but the potential for lower sulfide concentra-
tions in SDG sewers in comparison to pressure sewers exists.
The parameters of the equation are EBOD (which is directly propor-
tional to BODs and Temperature), M, and empirical constant, and r, the
hydraulic radius.
BODs Grinder-pump pressure systems have BODs concen-
trations of about twice that in conventional sys-
tems. STEP systems will have about half the BOD
of conventional systems.
Temperature Should be virtually the same for all sewer types,
although STEP wastewaters might be slightly cooler
due to lengthy residence time in interceptor or
septic tanks.
Hydraulic Radius Smaller pipes present a higher slime area to waste-
water ratio. This increases the rate of sulfide
production.
It then follows that designers of alternative systems should to the
extent possible follow good design practice to minimize sulfides genera-
tion where possible and, where not possible, take appropriate steps to
minimize the adverse impacts of odors and corrosion.
12
-------
It is interesting to review these recommendations in light of current
design practice for SDG systems and to highlight cases where different
design objectives might result in a system more conducive to sulfides
generation.
By transporting septic tank effluent rather than raw waste SDG sewers
can hydraulically have smaller diameter pipes placed at less steep slopes.
The net result is that the wastewater travels more slowly. Also in some
design approaches portions of the line may be continually full. In addi-
tion, the lack of abrasive solids limits the scouring action on the slime
layer of the pipe wall. These are conditions which clearly favor sulfides
generation. On the other hand, a significant portion of the sulfate may
have been converted to sulfide in the septic tank and may be vented to the
atmosphere in the sewer, since conditions in the latter would favor a
lower (Se). This could dramatically reduce the Sulfide Potential within
the SOG system.
It is impossible to quantitatively predict the relative importance
of these factors. However it is important that a designer consider the
sulfate levels in the water supply to assess the overall risk of sulfide
generation within an SDG sewer and then.design the system accordingly.
It is evident that alternative systems of the types that have been
considered will generally deliver wastewater carrying relatively high
sulfide concentrations. This does not invalidate the choice of such a
system, but some design modifications at the plant and of the collection
system may be necessary.
Unless a sewage stream will be sulfide free, some ways to cope with
the H2$ related odor and corrosion problems at the plant are necessary.
Both Quaker Lake and Glide systems show that the odor problem can be
solved by submerged discharging to an amply sized pond. With some other
systems it may be necessary to consider some form of chemical or biolo-
gical treatment for sulfide control before the wastewater enters the
treatment processes. It is not within the scope of this report to trans-
mit a comprehensive discussion of the complex subject of sul fide control
technology.
13
-------
SECTION 5
TWO-PHASE FLOW IN PRESSURE AND
SMALL-DIAMETER GRAVITY SEWERS1
INTRODUCTION
In pressure sewer systems, a small diameter sewer main similar to a
water main installation is buried shallowly and follows the ground contours,
Each home or group of homes has a small pump. The pump may be a grinder.
pump (GP) which grinds the solids to a slurry, or an effluent pump septic
tank combination (STEP).
Small diameter gravity (SDG) sewers are similar to STEP systems in
that a septic tank is used at the home to capture the troublesome matter
in sewage: grit, grease, bulky and stringy material. As the name implies,
flow is by gravity, negating need for pumps. SDG systems take a variety
of forms. Mains can be placed at a constant downgrade slope between
cleanouts. This is similar to conventional sewer design practice, although
the slope may be less steep due to the reduced scouring velocity required
since septic tank effluent is conveyed. Another form of SDG sewer is
like a pressure sewer main in that is can (within limits) follow the
contours of the terrain. Sewage may flow in an upslope direction so long
as the energy grade line slopes properly. This type of SDG sewer is
sometimes called a variable grade sewer (VGS).
Technologies of STEP pressure sewers (PS) and SDG systems may be
combined into one collection system.
Like gravity sewer technology, the hydraulic analyses of common
domestic water supply systems apply somewhat to PS and SDG systems.
However, there are important differences. One important difference,
the matter of two-phase flow (air and water), is discussed herein.
ENTRANCE OF AIR
When air or gas bubbles are present in a hydraulic pipeline, flow
problems can occur. To expel the air, air release valves are typically
placed at summits within the pipeline. However, as explained later,
this practice is insufficient when used with some configurations of
PS or SDG systems.
1. Written originally by William C. Bowne, Roseburg, OR, for USR&E.
14
-------
Air may enter a pipeline in a variety of ways. A one mile water
conveying pipeline contains enough dissolved air to completely fill over
100 feet of the pipe, assuming that water contains about 2 percent dis-
solved air by volume. The air (or gas) volume present in sewage would
differ by some unknown amount from that in water, but the concept is the
same. Because the solubility of gases depends on pressure, as the waste-
water is conveyed through a pipeline over varying elevations, gas comes
out of solution at higher points. Once out of solution, the gas will not
readily redissolve.
If pump intake submergence is shallow in a pressure sewer, air may
also enter the main by vortexing at the pump. For this reason, it is
preferable to submerge the pump intake sufficiently. The exact depth
depends on discharge rate.
When the pipeline is such that the static hydraulic gradient eleva-
tion of the main is lower than the elevation of an adjacent pump intake,
there is another source of air entrance. When flow in the system is zero
or minimal, the hydraulic gradient of the main is at or near static eleva-
tion. A siphoning condition develops, causing flow through the pump even
though the pump is not running. This flow continues until the liquid
level in the pump vault is lowered to the elevation of the pump intake,
which breaks the siphon. Air then enters the service line connecting the
main and the pump, and is forced into the main when the pump again operates.
This is shown diagrammatically in Figure 1. An enormous amount of air
may be introduced to a system at start-up, or following the repair of a
ruptured main. System design should provide means of purging such large
volumes of air and the methods should be described in maintenance training.
There is another source of air in SDG and PS systems. PS and SDG
mains may flow "uphill" or "downhill", as shown in Figure 2. When some
portion of the collection system is at an elevation higher than the point
of discharge to atmosphere, flow is "downhill" within some reaches of the
piping. During static conditions with downhill flow, portions of the
system will drain and fill with air. The air is drawn into the main as
previously described, through the main terminus, or through leaky pipe
joints at zones of negative pressure.
While it is important to totally prevent the accumulation of air (or
gases) within the piping system, the volume can be minimized by design,
and negative effects can be attenuated.
EFFECTS OF AIR IN PIPELINES
A classic air bound pipeline is shown in Figure 3. Suppose a pump
is used to move water from one reservoir to another. Water in the first
reservoir is at elevation zero, and the pump has a shutoff height of 105
feet. Assume also that water in the receiving reservoir is at elevation
80, no portion of the interconnecting piping is at an elevation higher
than 80 feet, the pump is running, and there is zero discharge.
15
-------
A1r fills this portion
of the service line.
Static
hydraulic grade
Main
Pump
Pump located
higher than
static HGL
Figure 1. Cross-section at pump Installation where pump is above
hydraulic grade line of main.
16
-------
Direction of flow
Mrcharge elevation-
a) "Uphill" flow. All parts of collection system are lower
than the point of discharge ta atmosphere.
Downhill
flow
Downhill
flow
Discharge elevation
a static HGL
Direction of flow
b) "Downhill" flow in portions shown.
Figure 2. Profiles of uphill and downhill flow in sewers.
17
-------
Air bound
hydraulic gradient
00
100
80
60
40
40 feet
40 fett
25 feet
106 feet
Reservoir
Figure 3. Air-bound hydraulic gradient.
-------
In this example, air is trapped at two summits, A and B. The air is
displaced some in the direction of intended flow, as shown. At summit A,
the pressure at a1 and a" is the same (65 feet of head) even though the
liquid level of a" is ten feet lower in elevation than at a1 due to dis-
placement of the water. A similar condition exists at summit B. Summing,
a, b, and c results in 105 feet, the shutoff head of the pump. No flow
is occurring, and the static hydraulic grade line is not at one elevation,
but is instead stepped as shown. One method of determining the presence
and location of pipeline air is to create a pressure on a static main and
then determine the hydraulic gradient at particular stations. This example
of air binding shows how many pumps routinely operate at higher heads than
necessary and points out a useful tool in hydraulic analysis.
There are other concerns in addition to air bound situations. When
air is present in a hydraulic pipeline:
o Flows are erratic, unpredictable, and have high head losses.
o Check valves are hammered and sometimes destroyed.
o Pumps fluctuate over a wide range of their H-Q curves.
o Acid may form on the pipe wall, creating a corrosive situation
if the air pocket is stationary.
o Gravity flow or intermittent gravity flow may allow grease and
other solids to air plate on the pipe walls.
Because the grease and other solids plating problem, GP systems are not
recommended for use in downhill flow situations.
When flow is downhill and siphoning through the pump occurs, the
pump itself will often become airbound. When the liquid level in the
pump vault again rises to "pump on" level, air is trapped in the volute.
The pump will run, but will not pump water. This condition may continue
until the motor burns out, necessitating a maintenance call.
Submergence is required for two reasons. Pump and motor submergence
cools equipment, prolonging life. Nonsubmergence of the motor violates
explosion proof requirements (NEC 501-8(a)) (10).
PREVIOUS RESEARCH
Problems of pipeline air have been the subject of intensive design
efforts on some larger systems. Typically, these pipelines convey water
from mountain regions to distant population centers, and are on overall
descending grades. Although the large size of the systems make them
dissimilar to PS or SDG projects, the problems are similar, though of
smaller magnitude.
19
-------
Upstream control on a closed conduit is illustrated in Figure 4.
Pipe stand and pipe check and vent structures is used to maintain a full
pipeline upstream from the control structure. This practice reduces or
eliminates surge and air entrainment problems.
In his doctoral thesis, Kent investigated the entrainment of air in
downhill flow (11). A four-inch diameter clear plastic pipe was used,
which allowed observation and scientific evaluation.
When air pockets form in the pipeline, as shown in Figure 5 and the
initial depth of flow is not below critical, a hydraulic jump cannot be
formed. Nevertheless, the action below an air pocket closely resembles a
jump. The jump is violent, with associated high head losses, and bubbles
are ripped from the pocket. The bubbles move downstream, with smaller
bubbles traveling faster than large ones. The bubbles tend to join, and
form another pocket from which bubbles are again torn, and the process
continues. The pocket may be stationary, move upslope, or move downstream.
Depending on the diameter and slope of the pipe, Kent developed equations
below to describe the velocity required for the pocket to remain stationary
(the point where drag and bouyant forces are equal). The velocity required
to attain this condition is a direct function of pipe diameter and pipe
slope from the horizontal plane
vmin =c
where Vmi-n = the minimum velocity to keep an air pocket stationary,
ft/sec.
Co1'2 = constant when the air pocket length is 1.5 pipe diameters,
dimensionless.
g = acceleration due to gravity, 32.2 ft/sec^
D = pipe diameter, ft.
8 = angle of pipe from horizontal
vmin in PS or SDG applications will typically range from 1/2 to 2
feet per second, the higher values related to larger diameter pipes and
steeper slopes. In most PS and SDG systems, movement to the terminus of
the pipeline will not purge the air because flows of sufficient velocity
are not of long enough duration. The air will'move only a limited dis-
tance downstream and return upslope when the peak flow condition has
passed.
The equation for head loss due to the presence of air in the pipe
developed by Kent indicates that head loss is directly proportional to
the percentage of air by volume and slope. Therefore, head loss increases
with quantity of air and as pipe slope increases. A few calculations show
that head losses due to air can be large and of substantial importance to
20
-------
Upstream n vent
TSii vS\
Downstream
pool
level
Pipe check
Vent
Direction of flow
Figure 4. Pipe check and vent.
21
-------
Air
pocket
Turbulant action of
hydraulic jump
Bubbles torn
from pocket
Figure 5. Air pocket and bubbles encountered In downhill flow.
22
-------
PS or SDG design. For example, a one percent volume of air, with a 10
percent (5.7 degree) pipe slope would correspond to a head loss of 0.001
ft./ft. A head loss of 0.01 ft./ft. would occur at 10 percent air. Recog-
nizing that these head losses would be additive to losses determined by
hydraulic equations, we see that they can easily become excessive.
LABORATORY AND PILOT EXPERIMENTS
To better understand the phenomena of two-phase flow, several experi-
ments were performed.
A 1-inch diameter transparent plastic tube, about 30 feet long, was
fastened to wal1-mounted pegboard with hooks. The hooks could be easily
adjusted to alter the pipeline profile. Tees fitted with vertically
rising 1/2-inch clear tubing were placed periodically in the tubing. The
1/2-inch tubing served as piezometers or as air release points, depending
on location. A pressurized water supply was connected and a venturi aspi-
rator used to inject dye to make the flowing water (and air pockets and
bubbles) more visible, and to introduce air to the pipeline.
The experiment was operated for several hours daily over a period of
about two weeks. When flow was uphill, the system ran quietly and pre-
dictably. When air was introduced to the system, it was effectively
expelled by the air release vents. At high flow velocities, some air
bubbles would pass the vents, but the volume was small, and no particular
difficulties were noted.
It was a different matter when flow was downhill. The larger volume
of air created by the draining of portions of the tubing changed the
hydraulics completely. Pockets and bubbles similar to those described by
Kent formed. The hydraulic jumps were violent, while flow was highly
erratic, unpredictable, and accompanied by high head losses. Water would
suddenly spurt out of the piezometers and air release vent tubing. The
hydraulic grade line was much steeper than when flow was uphill. Under
uphill flow conditions, water would never spurt from the vertical tubing,
nor would the hydraulic gradient rise to near the top of the tubing, even
at much higher flow rates.
This was a worthwhile experiment, easy to construct and operate. It
is a recommended experience for engineers contemplating two-phase flow
design.
Because of the small diameter of the experimental tubing, the experi-
ment was not entirely representative of PS or SDG sewer application, so
another test was conducted. In this experiment we used 100 feet of 6-inch
diameter transparent plastic piping, sloped at 2.5 percent, and flowing at
185 gpm (about 2 feet per second velocity). Discharge was to a receiving
vessel with a liquid level at the elevation where the static hydraulic
grade line intersected the sloping pipes (See Figure 6). During static
23
-------
Direction of flow
static
pool
hydraulic jump
during flow
Receiving
vessel
6'xlOO1 piping.
S - 2.51
no scale
Figure 6. Test apparatus using transparent piping,
24
-------
conditions, the length of the ellipse formed by the water surface in the
pipe at the static HGL was about 20 feet long. When flow began, the jump
would develop about 3 to 4 feet upstream from the toe, or most downstream
point of the ellipse.
Except for a few parameters, results were fairly similar to those
obtained in the previous test and as described by Kent. Bubbles traveling
downstream would often pass pockets rather than joining them. Very small
bubbles (approximately 2 millimeters diameter) seemed to move downstream
at a slower effective- rate than larger bubbles, because the small bubbles
took a more erratic path, darting about as they traveled.
Air bubbles and pockets were effectively removed by a 1/4-inch diam-
eter hole drilled in the crown of the pipe. When the hole was placed U
diameters below the jump, the vent was an estimated 95 percent effective
in removing air. Effectiveness did not increase at a greater distance
downstream.
FIELD STUDIES
This work was conducted to advance the knowledge of PS and SDG hydrau-
lics generally and to assess the design of the Glide, Oregon pressure sewer
system. Glide is the world's largest STEP system, and combines a few SDG
connections (4). Topographic conditions made downhill pumping manadatory
in some instances, and it was necessary to provide the most reliable hydrau-
lics possible. The system is sized to serve 7,300 people, and initially
serve 2,000. Predicted peak flows were 350 gpm initially and 1,050 when
the system was fully operational.
Much of the collection system had been installed at the time the labora-
tory experiments were conducted, yet design refinements were continuing. It
was decided to investigate the hydraulics of two-phase flow in a two-mile
portion of the main.
,1 0T5?«t!st Sect1on Involved 2,350 lineal feet of 12-inch diameter main,
and 8,270 feet of 10-inch main. The maximum elevation difference of the
pipeline profile was 58 feet. Two automatic air release valves (AARV) were
a part of the original installation, but by the time testing was completed,
one more had been added, and one relocated, for a total of 3. The ARRV's
were sewage-type air release valves, not vacuum or combination, with 5/16
inch orifice.
Flows of from 200 gpm to 700 gpm were produced by pumping river water
with a diesel fuel powered pump. Flows were measured using an impact tube
meter (Metraflex™ model 23) and checked with reasonable agreement using
the XY coordinate method calculation of water drop and the pump curve.
Pressure recording stations were located at four points along the main.
Laboratory test gauges were used, the accuracy checked daily with a mercury
manometer.
25
-------
The main was slowly filled with water over a period of two days, and
air purged. A plot of the profile of the main is shown in Figure 7. Drawn
on that plot are the static hydraulic grade lines, as well as the theoret-
ical dynamic hydraulic gradeline based on the Hazen Williams equation using
an assumed C value of 140 and a flow rate of 600 gpm. The assumption was
temporarily made that flow would break to gravity flow in the steeply
sloped pipe reach between stations 70 and 80. In developing the dynamic
gradient, no allowance was made for headlosses due to air. The dynamic
HGL was drawn to allow a quick visual comparison between actual conditions
and theoretical conditions without allowances for air.
The system was then slowly started up, and the air release valves
constantly checked. The system was operated for a full day at flows
ranging from 200 to 600 gpm. As expected, headlosses were high. The
elevation of the hydraulic gradeline was determined from pressure readings
at the observation station. A typical HGL for the day, at a flow rate of
600 gpm is shown in Figure 8. The HGL has been drawn as a straight line
between pressure stations for ease in interpretation, but in actuality
the headlosses would vary within each reach and the true HGL would take
on other shapes. At 600 GPM, notice how much higher the headlosses
actually are than hydraulic calculations would predict without accounting
for air. Realize that this two-mile test section was typical of the 25
miles of main comprising the system. Such a system that ignored air
considerations in the design would not operate suitably.
The shutoff valve at station 0 was then slowly closed while the pump
was still running, and pressure readings taken at the stations at one
minute Intervals. The close time intervals were necessary since air made
the static pressures constantly change. A typical plot is shown in Figure
9. Note that the pressurized static HGL does not attain a single elevation.
This reveals trapped air, as seen in the example in Figure 3. Clearly, air
was trapped In the reach between approximate stations 70 and 80. No doubt
much air was also contained between pressure station 1 and 2, but the fairly
flat shape of the pipeline profile between these points prevented the pres-
sure readings from revealing its presence.
Next, the piping system was modified. The air release valve near
station 81 which had been placed on the exact summit of the pipeline was
relocated to a point several hundred feet downstream. An automatic air
release valve was also added at station 69.
These changes had been anticipated from experience in operating model
studies. The reason for the change at station 81 is that under static
conditions, water would stand to the right of the Invert summit as shown
in Figure 10. As flow began, the profile of water surface would soon seal
off the entrance to the air release at the summit. Yet, as flows increased
and the hydraulic grade line rose in the reach between stations 70 and 80,
over 4,000 gallons of water (septic tank effluent) filled the reach. The
displaced air had to be expelled.
26
-------
to
ro
-j
I/I O
€-»• -«l
(D -*
- CD
-•• o
10 -h
Is?
_•. o
5 4
o> — «
_•. fD
O (D
O (A
3 C*
VI
-»• «/l
O. 0>
tt> O
-I rfr
Q| -i.
rt O
-«• a
o
O
o.
(D
•o
10
I/I
I/I
c
(D
I/I
0>
(D
-I
170
160
150
140
130
120
110
100
90
Calc. HGL 9 600
O* Automatic air release.
a- Pressure recording sta.
JL
JL
J_
Station 20
30
40
SO
60
70
80
90
100
-------
ro
00
ua
(D
00
if
Oi
I/I
c
3
o.
o.
0>
_k
o
(O
01
a.
a>
n>
o>
r*
a\
o
o
1
Neas. HGL 9 600 gpm
^Static HGL
Calc. HGL 9 600
O- Automatic air release.
D- Pressure recording sta.
Station
-------
ro
to
170
160
150
140
1JO
120
110
100
90
Pressurized static HGL
^•Static HGL
m _^ — — -*- Calc. HGL 9 600 gpa
'- Static HGL
12" ft 10" 0
« T >
o- Automatic air release.
Q* Pressure recording sta.
Station 20
30
40
SO
60
70
80
90
100
-------
Dynamic water
surface
Static water
surface
Location of air
release valve
a)Enlargement of profile at
Station 81 showing water surface
profile during flow. Water seals of
air release at summit.
Direction
of flow
b)Enlargement of profile near
Station 69, showing hydraulic
jump, air bubbles and pockets
that require ventilation.
Figure 10. Pipeline profiles at stations 69 and 81,
30
-------
The situation at station 69 is shown in Figure 10. Under flow, a
hydraulic jump is formed where the dynamic hydraulic grade line intersects
the main. The jump is violent and pumps air bubbles into downstream reaches.
To capture and ventilate that air, the connection of the air release assembly
to the main must be made below the jump.
Once installation of the air release valves had been made in the revised
locations, flow at various rates was again introduced to the system. Time
was required to work the air out of the pipeline. During this period, gauge
readings were taken at the pressure stations. Typical results are shown in
Figure 11. Note that the HGL is much lower than in Figure 8, and as time
progressed with flow occurring, the HGL continued to drop and become of
flatter slope, approaching the theoretical.
DESIGN OF THE GLIDE, OREGON SYSTEM
A portion of the Glide piping system was designed and now operates as
previously described, and as shown in Figure 11. Another portion of the
collection system had a different profile, characterized by downhill runs
and multiple locations of hydraulic jumps that were less steep and dis-
tinct. Treatment of the situation by installation of air release valves
as used at stations 69 and 81 .was not practical.
The type of profile referred to is shown in Figure 12. To obviate
the problems of downhill flow, standpipes were used to keep upstream
reaches full. The standpipe design is shown in Figure 13, and its appli-
cation shown in Figure 14. The PVC standpipe was located in an area of
relatively steep topography so that the it could be buried at the usual
depth and run a short distance to attain the desired elevation at the
crest to keep the upstream reaches submerged. A hydraulic jump occurs
in the downleg of the standpipe. This down!eg is steeply sloped, and of
larger diameter to reduce flow velocity and prevent the conveyance of air
bubbles into the downstream reaches. Air release valves were used to the
exact summits of the pipeline upstream from the standpipes.
An investment of several months engineering time was made to inves-
tigate the use of special backpressure sustaining valves in lieu of stand-
pipes. Despite special attention given to the design and selection of the
valves by valve company engineers, none worked properly. A variety of
problems developed. In some cases, the valve opening and closing pressures
differed too much for them to be used. Some valves would not close com-
pletely enough, especially if small amounts of solids were caught on the
valve seat. Other valves would constantly hunt for position, while still
others were so complicated and expensive that they were impractical.
Valve failure modes were undesirable. Neither failing open nor
failing closed were suitable options. The standpipes were passive and
accomplished the goal of maintaining full upstream reaches simply, reliably,
and economically.
31
-------
I
to
ro
to
o.
CD
Ol
o\
o
o
ua
•a
rf
a>
c*
ro
to
r*
170
160
150
140
UO
120
110
100
90
Measured H6L 9 600 gpm
o- Automatic air release.
a- Pressure recording sta.
Station
20
30
40
SO
60
70
80
90
100
-------
Flow
Station
Figure 12. Pipeline profile with numerous downhill flow sections.
33
-------
AARV
Larger diameter down-
leg reduces velocity
to keep air bubbles
froo traveling Into
downstream piping
Downstream static
liquid level
AARV*
X
Upstream
liquid
level
Isolation valve,
normally closed
AARV* • Automatic air release valve vented to soil bed for
odor absorption
Figure 13. Standpipe.
34
-------
Dynamic
HGL
\J
Static HGL
Station
Figure 14. Pipeline profile using standpipes, showing static and
dynamic HGL's.
35
-------
As air release valves are normally used on a force main, only small
volumes of air are expelled. Large volumes are expelled at installations
such as at station 69, 81 and the standpipes. Since odor from hydrogen
sulfide gas was expected, the air release valves were vented to soil beds
for odor absorption (4).
When a septic tank was located at an elevation higher than the dynamic
HGL, a gravity installation was used. The tank was fitted with a 1-1/4 inch
diameter outlet tee, located at an elevation lower than would normally be
found in septic tank practice. The purpose of the lowered level was to
provide reserve volume within the tank. In the event valves on the pressure
sewer main were closed to allow for repairs on the main, flow from the home
could continue for at least one day before the tank overflowed.
The gravity service line was 1-1/4 inch PVC, placed without regard to
grade, the same as used on the STEP Installations. A check valve was
used on the service line at the main in those cases where a STEP pump
could pump a gravity tank if an isolation valve on the main were closed.
OPERATING EXPERIENCE WITH THE GLIDE SYSTEM
Over three years have passed since the Glide, Oregon system became
fully operational. During this time, performance has been carefully
monitored. Air release stations are inspected frequently. Pressure
readings are taken at designated stations throughout the system, and
compared with previous charts.
The automatic air release valves have proven reliable. Rarely does
one malfunction, and never to cause any difficulty, contrary to the general
reputation of sewage air release valves. This performance is attributed
to the low grease and solids content characteristic of septic tank effluent.
No odor has ever been detected at air release stations vented to soil
beds. It is not known if odor problems would exist if the valves were
vented in the usual way, but it is logical to expect they would.
Peak flows are on the order of 350 gpm. Since the system is sized
for 1,050 gpm, it operates far below its ultimate capacity. However, on
two occasions the service area was flooded, the standing water on the
ground surface entered the pressure sewer pump vaults, and peak flows to
an estimated 1,200 gpm occurred at those times. Some of the pressure
sewer pumps were dominated by others to shut off head during these periods
as would be expected since flow exceeded design. The fact that design
flows could be exceeded testifies that the system was performing properly.
There has been no evidence of air plating problems, or any other
problem with the collection system piping. Some treatment occurs in the
pipeline, but it is unknown how much can be attributed to air entrain-
men t at the hydraulic jumps.
36
-------
^There are a few places where pumps are above the static hydraulic
gradient. These pumps occasionally became air bound. In one case, a
pump was located at an elevation substantially higher than the static HGL,
and higher than the dynamic gradient. About weekly, the service line at
this installation became air bound as shown in Figure 3, and the pump
would be driven to shutoff head. No problems have occurred since the
installation was switched to a gravity connection. The trickle gravity
flow from the septic tank to the service line allows air to escape from
the service line back through the tank outlet tee rather than becoming
trapped by the sudden and larger flow produced by a pump.
The gravity installations have had no problems. Concerns that
zooglea or other matter might plug the service lines have proven unfounded
so far. Evidence indicates that the effluent level in the interceptor
tanks has risen perhaps a foot at times. Overflowing of the tanks has
not occurred, so apparently this small rise in liquid level has developed
enough head to overcome whatever restriction developed in the service
line.
When flow is "downhill" in a closed conduit, such as a pressure
sewer or a small-diameter gravity sewer, a hydraulic jump is formed at
the intersection of the pipe and the dynamic hydraulic gradient. Bubbles
are ripped from air pockets and conveyed further into the piping system.
The presence of air (or gas) causes flow to occur in ways that would not
be predicted by standard hydraulic equations such as Manning or Hazen-
Williams. Head losses are high, and flows and pressures are erratic.
By proper design, systems can be engineered to flow downhill and
avoid air problems. The techniques are not in common knowledge. An air
release station is placed a short distance downstream from the jump, and
at the upper reach to vent the air displaced. Standpipes may be used to
locate the jump in a steeply sloped pipe section that may be of larger
diameter to produce a low flow velocity.
Pressure and gravity sewers conveying septic tank effluent have been
effectively combined on the Glide, Oregon project using hydraulic techni-
ques herein described.
37
-------
REFERENCES
1. Eblen, J.E. and Clark, L.K. Pressure and Vacuum Sewer Demonstration
Project. Bend, Oregon. U.S. EPA, Publication No. 600/2-/y-166, IRITIS No,
2. Winzler and Kelly. Engineers. Manila Community Services District Septic
Tank Effluent Pumping and Sewer Demonstration Project. Final Report.—
California state Water Resources Control Board, 1982.
3. University of Wisconsin. Management of Small Waste Flows, U.S. EPA
Publication No. 600/2-78-173, MTIS Mo.
4. Bowne, W.C. and Ball, H.L. "Pressure Sewer System Proves Effective,
Economical", Public Works, March 1981.
5. WPCF-ASCE, Municipal Wastewater Treatment Plant Design. WPCF MOP #8,
iy / /.
6. WPCF-ASCE, Design and Construction of Sanitary and Storm Sewers,
WPCF MOP #9, 1369. —
7. U.S. EPA Process Design Manual for Sulfide Control in Sanitary
Sewerage systems. U.S. EPA. J974.~"
8. Cooper, I.A., and Ryzek, J.W., Treatment of Pressure Sewage" in
NSF 5th Conference on Individual Onsite Wastewater Systems. 1979.
9. Kreissl, J.F., Smith, R., and Heidman, J.A., "The Cost of Small
Community Wastewater Alternatives", in Wastewater Alternatives for
Small Communities. U.S. EPA publication NO. 60U/9-80-062, NTIS
Pub. No. PB 81-131658, 1980.
10. Anon., National Electric Code. 1981.
11. Kent, J.C., The Entrapment of Air by Water Flowing in Circular
Conduits wi tTTDowngrade Slopes, Doctoral Thesis. U. of California.
1952.
12. Weibel, S.R., Straub, C.P., and Thoman, J.R., Studies on Household
Sewage Disposal Systems. USPHS - Federal Security Agency, 1949.
13. Kienow, K.E., Pomeroy, R.D., and Kienow K.K., "Prediction of Sulfide
Buildup in Sanitary Sewers", in ASCE-EED, 108, 941, 1982.
38
-------
TECHNICAL REPORT DATA
(Plcese read [inductions on the reverti! bef
I. REPORT NO.
3. RECIPIENT'S ACCHSSIC.VNO.
-. TITLE ANO SUBTITLE
Alternative Sewer Studies
5. REPORT OATS
6. PERFORMING ORGANIZATION COOS
'. AUTHORIS)
3. PERFORMING ORGANIZATION REPORT NO.
Urban Systems Research and Engineering,.Inc.
9. PERFORMING ORGANIZATION NAME ANO ADDRESS
Urban Systems Research and Engineering, Inc.
Cambridge, MA 02138
10. PROGRAM cLs.VlE.MT NC.
CAZ31B
11. CCNTRACT»GRANT NO.
68-03-3057
12. SPONSORING AGENCY NAME ANO AOOHS3S
Water Engineering Research Laboratory—Cincinnati, OH
'Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
13. TYPS OF REPORT ANO PERIOD CSVSnE3
Final (12/80 - 6/83)
14. SPONSORING AGENCY CCO£
EPA/600/14
1S. SUPPLEMENTARY NOTES
Contact: James F. Kreissl C"513) 684-7611 - FTS; 569-7611 - Comm.
16. ABSTRACT
An evaluation of design and operational aspects for two popular
alternative sewering techniques, i.e., pressure and small-diameter
gravity sewers, was performed. The results provided insights useful
to designers of these systems with regard to'sulfide concentrations
in these sewers and hydraulic designs for downhill runs.
17. KEY WCRCS ANO DOCUMENT ANALYSIS
a. DESCRIPTORS
13. DISTRIBUTION STATEMENT
Release to Public
b.lOeNTIFIERS/CPSN ENDED TERMS
19. SECURITY CLASS (Tnti Rtporrj
Unclassified
20. SECURITY CLASS iTiia pagtj
Unclassified
c. C=SATI r:«!d/Grcu3
21. NO. 0? PAGES
22.* PRICE
EPA Farm 2220-1 (9-73)
------- |