625477011B
Alternatives for
Small Wastewater
Treatment Systems
Pressure Sewers/Vacuum Sewers
EPA Technology Transfer Seminar Publication
This document has not been
submitted to NTIS, therefore it
should be retained.
-------
EPA-625/4-77-011
ALTERNATIVES FOR SMALL
WASTEWATER TREATMENT SYSTEMS
Pressure Sewers/Vacuum Sewers
ENVIRONMENTAL PROTECTION AGENCY* Technology Transfer
October 1977
-------
ACKNOWLEDGMENTS
This seminar publication contains materials prepared for the
Environmental Protection Agency Technology Transfer Program and
presented at Design Seminars for Small Wastewater Treatment
Systems.
Part I was prepared by James F. Kriessl, Sanitary Engineer,
Municipal Environmental Research Laboratory, Environmental Pro-
tection Agency, Cincinnati, Ohio.
Part II was prepared by I. A. Cooper, Senior Project Manager,
and J. W. Rezek, President, Rezek, Henry, Meisenheimer & Gende,
Inc., Consulting Engineers, Libertyville, 111.
NOTICE
The mention of trade names or commercial products in this publication is
for illustration purposes, and does not constitute endorsement or recommenda-
tion for use by the U.S. Environmental Protection Agency.
-------
CONTENTS
Part I. Pressure Sewers 1
Background 1
Introduction 2
Description 5
Previous Experience 6
System Description 10
Design Alternatives 12
Construction Considerations 42
Operation and Maintenance 44
Variations 47
Codes 48
Wastewater Characterization and Treatment 48
Costs 50
Conclusions 56
References 57
Part II. Vacuum Sewers 61
Background 61
System Parameters 66
Vacuum Sewage Characteristics 69
Cost Comparisons 70
Critique of Existing Systems 72
Operation and Maintenance 73
Conclusions 76
Recommendations 77
References 77
Appendix A. Vacuum Sewage Flow Characteristics 79
Flow Regimes 79
Slip 79
Friction Headloss 82
Nomenclature 85
References 86
Appendix B. Design Example 88
Base Conditions 88
Piping Profile 88
Flow Conditions 90
Dynamic Headloss 95
Discharge Pump 96
References 97
in
-------
Port I
PRESSURE SEWERS
BACKGROUND
The disposal of human wastes in an environmentally acceptable manner has been a problem
throughout history. Historical evidence of early man's struggle to overcome this problem has been
recorded and related elsewhere.1"3 Little progress has been recorded between the pre-Christian solu-
tions described in the literature and the mid-19th century rebirth of municipal sewage in
Europe and the United States, a development that has continued to the present. Between 1857,
when the first large American system was installed in Brooklyn, N.Y., and 1905, about 28 million
Americans were "sewered."4 By 1940 more than half of the total U.S. population was served by
sewers, and the figure now stands at about 71 percent.
A recent review of nearly 300 facilities' plans for rural communities in the United States pro-
duced the relationship shown in figure I-l.5 Monthly charges much above $20 are considered ex-
cessive in rural areas where median incomes are generally significantly lower than in urban areas.
Because most on-site wastewater disposal systems would cost significantly less than $20 monthly,
the on-site approach has been generally used in these areas. Difficulties have arisen in areas where
conventional on-site systems have failed because of unfavorable soil conditions. Typically, the result
from this condition has almost invariably been a recommendation to install sewers in the commun-
ity. Implementation of this recommendation depended on the financial status of the community,
availability of Federal grants, and public attitude. Without getting into a lengthy discussion on the
merits and demerits of the grant programs and centralized collection and treatment systems, it suf-
fices to note that the cost of conventional sewers is extremely high for most small communities. In
fact, it is not uncommon to see engineering estimates in excess of $10,000 per home. Also, the cost
of the conventional collection system generally represents more than 80 percent of the total system
capital cost in rural areas. Figure I-l clearly illustrates the relationship between cost and population
density, which is primarily explained by the greater length of sewer per contributor, greater prob-
lems with grade resulting in more lift stations or excessively deep sewers (see fig. 1-2), and regula-
tions that limit the smallest sewer pipe diameter.
Essentially because of the foregoing economics, the primary form of wastewater treatment and
disposal in rural areas has been the septic tank-soil absorption system (ST-SAS). Before the passage
of the Norris-Rayburn Act during the depression of the 1930's, few rural areas had the electricity
necessary to provide for water carriage of human wastes. However, as the rural electrification pro-
gram took effect the following decade, two major events occurred. First, the rural population ob-
tained electricity that upgraded the standard of living, including pressurized water supplies and
water carriage of wastes. Second, urban dwellers emigrated to previously rural areas, where they
could enjoy the best of both societies. The disposal of wastewater generated in these unsewered
areas was best accomplished by ST-SAS's, as shown in figure 1-3.6 Developers of these areas also
found advantages in these systems because costs were directly related to the dwelling and offered a
minimum of postconstruction responsibility. Unfortunately, many of these systems have failed be-
cause of faulty design and construction, unsuitable soil conditions, and owner negligence. Present
estimates indicate, however, that 15 to 20 million ST-SAS's still exist in the United States, serving
more than one-fourth the population.
Unfortunately, many situations have come about in recent years that cannot be solved by
either of the traditional alternatives, and the results of attempting to apply either technology in
-------
40
30
"o
-a
en
O
I
IT
ID
O.
8
20
10
Cost (dol/mo) = 43e~°-1 (persons per acre)
I
I
4 6 8 10
POPULATION DENSITY, persons per acre
Figure 1-1. Monthly cost of gravity sewers.
12
14
these situations have been unsatisfactory to all involved. For example, building moratoriums have
been imposed that prevent the development of highly desirable land parcels; conventional sewers
have been installed at tremendous cost to the homeowners serviced; and ST-SAS's have been con-
structed that cannot function properly, therefore contaminating the very environment that made
the site so attractive originally. The problem has become so acute that the 92d Congress specifi-
cally directed the Environmental Protection Agency (EPA) in Section 104(q)(l) of Public Law
92-500 to
conduct a comprehensive program of research and investigation. . . eliminating pollution from sewage in rural
and other areas where collection of sewage in conventional, community-wide sewage collection systems is im-
practical, uneconomical or otherwise infeasible, or where soil conditions or other factors preclude the use of
septic tank and drainage field systems.
INTRODUCTION
Although sewage pumping has been practiced for many years in municipal systems in the form
of lift stations and force mains to avoid excessive depths of cut, and in many individual homes in
the form of ejector or sump pumps, the wholesale use of small-diameter pressure collection systems
-------
28-32
24-28
20-24
£
t-~
o
LL
O
I
a.
LU
Q
16-20
14-16
12-14
10-12
8-10
<8
10
15
20
25
30
35
40
45
50
55
60
CONSTRUCTION COST, dol/l ft
Figure I-2. Cost of sewer construction.
did not emerge until the latter part of the 1960's. The chief impetus at that time was provided by
the late Gordon Maskew Fair who proposed that small-diameter pressurized sewers be placed inside
larger combined sewers to carry sanitary sewage. Although the "sewer-within-a-sewer" concept was
not found to be entirely feasible in a resulting study, the use of pressurized sewers carrying ground
sewage was feasible.7 This study was performed by the American Society of Civil Engineers (ASCE)
for the Federal Water Pollution Control Administration and included numerous research studies on
such topics as household wastewater generation patterns, critical velocities of flow, alternative sys-
tem layouts, and prototype grinder pump (GP) performance.
Although experience with pressure sewer systems is limited in both number of installations and
duration of service, some information is available on their economics. The EPA has sponsored full-
scale evaluations of GP-pressure sewer systems at Albany, N.Y.; Phoenixville, Pa.; and Grandview
Lake, Ind. Other communities have also used this technology. Significant data are available from
these sources, and the purpose of this report is to present as much of this information as possible to
assist the engineering profession in determining the applicability of and design criteria for pressure
sewer systems.
A number of advantages of pressure sewers have been presented in the literature.1 "3> 7'12
These benefits are primarily related to installation costs and inherent system characteristics. Because
these systems all use small-diameter plastic pipes buried just below the frost penetration depth, their
installation costs can be quite low compared to conventional gravity systems in low-density areas.
Other site conditions that enhance this cost differential include hilly terrain, rock outcropping, and
-------
Plumbing fixtures to be
properly trapped and
vented
House
drain
House
sewers, laid
on well-compacted
earth
Absorption
field
Nonperforated
tile
Tile
drainage
lines
Figure 1-3. Typical on-site system.
high water tables. Because pressure sewers are sealed conduits, there should be no opportunity for
infiltration, and treatment plants can be designed to handle only the domestic sewage generated in
the homes serviced, excluding the infiltration that occurs in most gravity systems.
As with any technology, certain disadvantages also exist. The disadvantages of pressure sewers
include high operation and maintenance costs related to the use of mechanical equipment at each
-------
point of entry to the system. Also, depending on the type of system used, the wastewater conveyed
to the treatment facility may be more concentrated than normal wastewater. It may require, there-
fore, a higher level of treatment to satisfy effluent requirements. A wastewater will also be devoid
of oxygen.
DESCRIPTION
Essentially, a pressure sewer system is the reverse of a water distribution system. The latter
employs a single inlet pressurization point and a number of user outlets, while the pressure sewer
embodies a number of pressurizing inlet points and a single outlet, as shown in figure 1-4. The user
input to the pressure main follows a generally direct route to a treatment facility or to a gravity
sewer, depending on the application. The primary purpose of this type of design is to minimize sew-
age retention time in the sewer.
The two major types of pressure sewer systems are the GP system and the septic tank effluent
pump (STEP) system. These are depicted in figures 1-5 and 1-6. From these figures it is obvious that
the major differences between the alternative systems are in the on-site equipment and layout. But
some subtle differences also exist in the pressure main design methods and in the treatment systems
required to reduce the pollutants in the collected wastewater to an environmentally acceptable
level. Neither pressure sewer system alternative requires any modification of household plumbing,
although neither precludes it if such modifications are deemed desirable.
T
Pressure sewer
Water main
Figure 1-4. Pressure sewer vs. water main.
-------
Pressure
sewer
PVC piping
Existing
gravity
sewage
piping
Storage
tank
•^Drainage field
Existing
septic tank
Overflow level sensor
On-off level sensor
Figure 1-5. Typical grinder pump installation.
Junction box and
high level
alarm-
2-inch plastic pipe for electricity
PVC plastic main
1 1/4-inch plastic
service
Existing or new
septic tank
Check'
valve
Ball or gate valve
24-inch concrete pipe with floor and lid
1/3-hp sump pump
Figure I-6. Typical STEP system.
PREVIOUS EXPERIENCE
As noted earlier, the isolated pumping of sewage and septic tank effluent has been practiced
for many years. The first attempt to use a pressure sewer system was reported by Clift9 in 1968,
but the system described did not employ the techniques and materials that are now considered
standard design practice. To serve 42 customers in low-lying areas of Radcliff, Ky., it would have
-------
been necessary to finance $3,170 per connection, while the prototype pressure sewer system cost
only $1,346 per connection. This prototype design used pneumatic ejector units at each connec-
tion, which discharged into a 3-inch (7.6-cm) cast iron lateral and a 4-inch (10.2-cm) cast iron main
that emptied into a gravity sewer. Even though mechanical and electrical problems were encoun-
tered that eventually caused abandonment of the system, Clift reported that during the first 3 years
of operation, no odors or blockage of pressure lines occurred.9 Severe corrosion was encountered,
however, and found to be the primary cause of the electrical and mechanical component short-
comings.
Clift also performed preliminary estimates on similar prototype pressure sewers for two other
locations. In one case, 120 out of 280 lots around a lake were considered well suited to using a sys-
tem with conventional gravity sewers, while the more inaccessible lots were thought to be better
served by pressure sewers. In the second case a similar hybrid design approach was estimated to save
5.5 percent in capital costs.9
The most highly instrumented study of pressure sewers was performed on a group of 12 homes
in Albany.2'8'10 Each dwelling was equipped with a commercial GP and connected by laterals to a
pressure main that emptied into a gravity sewer, as shown in figure 1-7. The system operated well
after the original prototype GP units were replaced with improved models. The pressure main had
been oversized to allow all units to operate simultaneously. Subsequent accumulations of grease and
fibrous materials reduced some pipe cross-sectional areas by as much as 40 percent. Valuable infor-
mation was reported on design and construction methods and on the operational characteristics and
maintenance requirements of the GP units. Monthly power costs of 10 to 27 cents per home were
incurred, based on a rate of 2.3 cents per kilowatt hour. Wastewater from the pressure sewer was
characterized and found to be more concentrated than normal municipal wastes, ostensibly because
of the lack of sewer infiltration.
Another relatively short-term (6-month) study of a pressure sewer system with GP's was made
at Phoenixville.1 1 This system, as shown in figure 1-8, was 2,800 feet (854 metres) long and dis-
charged into a gravity sewer more than 60 feet (18.3 metres) above the farthest GP location. An-
other unique feature of this system was the inclusion of multiple-family dwellings serviced by a
single GP. Data reported on construction costs were excellent. Some indirect evidence of pipe cross-
sectional area reductions, similar to the Albany study, was also noted. The GP units used were simi-
lar to the Albany units, and their operation resulted in a monthly power cost of 11 to 25 cents per
capita.
The Grandview Lake pressure sewer system12'14> a> b was much larger in size (it served 93
homes). The need for this system was related to an engineering estimate for conventional sewerage
Ell I 1 1 1 1 1 1 II
12
11
10
9
8
7
6
5
4
3
2
1
v . Sampling and
\l control box
t •*• Gravity sewer
Manhole
Figure 1-7. Twelve homes in Albany system.
aS. M. Rees, SIECO, Inc., personal communication.
bR. P. Farrell, Environment/One Corporation, personal communication.
-------
Existing
gravity
sewer
Location of pump storage grinders
2 apartments
7 apartments
Single home
500
1,000
1,500
2,000
2,500
3,000
Figure I-8. Phoenixville pressure sewer system.
of $3,000 per lot or $10,000 per existing home because of unfavorable terrain and the resulting
need for nine lift stations. This pressure sewer system included different types of GP units and a few
STEP units. Approximately 29,000 feet (8,840 metres) of pressure main, six automatic and seven
manual air release valves, and treatment by stabilization pond with effluent land spreading were
used in the Grandview system, as illustrated in figure 1-9. Grease problems plagued the system by
causing faulty operation of automatic air-release valves and by promoting deposits on flow measur-
ing devices at the plant. The installed cost of pressurization equipment and ancillary on-site com-
ponents varied from $1,000 to $1,500 per home. A contingency provision for potential on-lot over-
flows during equipment or electrical outages was included in the system design. Existing SAS's were
used whenever possible. Where these were not available, a small (2-day capacity) gravel-filled absorp-
tion bed was provided. Generally, 1-inch (2.5-cm) service connections were used to feed 3- and 3.5-
inch (7.6- and 8.9-cm) pressure mains.
Other installations of GP-pressure sewer systems being designed, installed, or operated have
been noted.10-15-17, b, c, d Bowles1 and Cochran15 describe an installation at Horseshoe Bay on
Lake Lyndon Baines Johnson, Tex. As many as 4,000 connections are planned for this develop-
ment, with about 106,000 feet (32,300 metres) of a 2- to 12-inch (5.1- to 30.5-cm) pressure main.
About 200 GP units were in operation.15 Equipment problems relating to installation and design
have been experienced, but corrections have been made and the system is now functioning satisfac-
torily. The sewage is treated by an activated sludge system with tertiary chemical clarification and
filtration. Another Texas system, which has been in partial operation since 1972, is located at Point
bR. P. Farrell, Environment/One Corporation, personal communication.
CR. E. Lawford, Peabody Barnes Company, personal communication.
dj. Schultz, Becher-Hoppe Engineers, Inc., personal communication.
-------
—— PVC pipe (size as shown)
H Lagoon cells
*•-
3/4 acre total
Figure 1-9. Grandview Lake sewage research and demonstration project.
-Control plot
-Irrigated plot
i
Venture on Lake Travis. This GP installation also suffered from initial problems resulting from con-
struction activity but was functioning in an acceptable manner in 1974. Two more large GP systems
and two smaller STEP systems have been approved in Texas.15
Gray16 has reported on the circumstances that led to the design and construction of a GP-
pressure sewer system at Weatherby Lake, Mo. The system now serves 330 homes and is expected
ultimately to serve 900. The total bid cost of the pressure system was $1,030,108, compared to a
conventional system (including eight pumping stations) estimate of $2,250,000. This system con-
sists of 309 GP units, 35,000 feet (10.7 km) of pressure main, 37,100 feet (11.3 km) of service lines
(polyvinyl chloride (PVC), SDR-26 with gasketed joints), 42 air-release valves, and 24 flushing and
cleanout connections. Included in the above bid cost is 5,300 feet (1.6 km) of gravity interceptor to
deliver the pressure sewer effluent to the Kansas City municipal system for treatment and disposal.
Other GP projects have been proposed, designed, or constructed in Saratoga, N.Y.; Clifton
Park, N.Y.; and Kinnelon, N.J.10- 18 A GP pressure sewer project on Madeline Island, Wis., is being
built to serve a recreational development.'1
dj. Schultz, Becher-Hoppe Engineers, Inc., personal communication.
-------
The most noteworthy STEP-pressure sewer installations are located in Florida and Idaho. Gen-
eral Development Utilities, Inc., of Miami has installed two large systems—one in Port Charlotte,
Fla., and the other in Port St. Lucie, Fla. The Port St. Lucie pressure sewer (125 homes), buried at a
depth of 2 feet (0.61 metre), discharges into a gravity sewer; while the smaller (26 homes) system
at Port Charlotte discharges into an extended aeration treatment plant. The Port Charlotte system is
the oldest, having been in operation since August 1970. The pumping units are small centrifugal
sump pumps, and the pump pits are vented via the building sewers in the same manner as the 900-
gallon (3.4-m3) septic tanks that pretreat the wastewater.6
Two separate pressure sewer installations located at Coolin and Kalispell Bay at Priest Lake,
Idaho, serve 348 and 200 homes, respectively. One-third- and one-half-hp (0.25- and 0.37-kW) sump
pumps, equipped with bronze impellers, are used to pump the septic tank effluent through 1.5-inch
(3.8-cm) PVC, SDR-26 service lines and 3- to 6-inch (7.6- to 15.2-cm) PVC mains to lagoons for
treatment.f Although some initial problems resulted from improper impellers that were supplied
with the pumps, the operation of these systems and treatment facilities has been capably handled
by one individual.
Sanson has described design methods used in planning STEP systems for two Indiana commun-
ities.19 Additional pressure systems using STEP concepts have been planned or approved in Florida,
Oregon, Idaho, Washington, Ohio, Wisconsin, and Arkansas.
Two privately owned and operated pressure sewer systems located in Oregon have been in
operation for a significant period of time using centrifugal pumps to pressurize raw sewage directly
from the source. One system services houseboats (approximately 500), while the other services a
private housing development (approximately 150 homes). These systems are operating without ex-
cessive operation and maintenance requirements, despite the higher potential operation and main-
tenance costs with this design.20' 8
Two major manufacturers of pressurization equipment have supplied information on present
and future installations of pressure sewer systems.1*' c The States listed below have approved at least
one project that is either being designed, constructed, or operated:
Arkansas
California
Delaware
Florida
Idaho
Illinois
Indiana
Kentucky
Michigan
Mississippi
Missouri
Nebraska
New Jersey
New York
North Carolina
Ohio
Oregon
Pennsylvania
South Carolina
South Dakota
Texas
Vermont
Virginia
Washington
West Virginia
Wisconsin
SYSTEM DESCRIPTION
A pressure sewer system consists of two major elements: the on-site or pressurization facility
and the primary conduit or pressurized sewer main. Probably the widest divergence of opinion ex-
ists on the proper design and equipment selection for the pressurization facility. Opinion varies be-
bR. P. Kartell, Environment/One Corporation, personal communication.
CR. E. Lawford, Peabody Barnes Company, personal communication.
"H. Schmidt, General Development Utilities Company, personal communication.
*C. W. Rose, Farmer's Home Administration, personal communication.
gj. Ward, personal communication.
10
-------
cause of the competition between proprietary mechanical devices of different designs and because
of some basic attitudes on the relative merits of the available alternatives.
In all designs household wastes are collected in the building drain and conveyed therein to the
pretreatment or pressurization facility. In most cases the piping arrangement includes at least one
check valve and one gate valve to permit isolation of each pressurization system from the main sew-
er. The two major alternative systems, which are illustrated in figures 1-5 and 1-6, use a pressuriza-
tion device that is located below ground in a manhole or access hole to collect the household wastes
by gravity discharge. GP's also can be installed in the basement of a home to provide easier access
for maintenance and greater protection from vandalism.21
The pressure main can take many forms, but it generally consists of a single, small-diameter
conduit that has numerous feeder lines from each pressurization inlet. This type of arrangement has
been deemed desirable to minimize sewer retention time.7 A typical example of a pressure sewer
flow diagram is illustrated in figure I-10.22
O Pressurization device
|» Combination cleanout,
manual air release, and
flushing station
D Branch number
O Manhole
Pressure sewer
Gravity sewer
-*'-- Contour line
> Flow direction
l * ^
~~~~HzSi
s IW_J I ^-
Street
^,,--790
Figure 1-10. Typical pressure sewer layout.
11
-------
DESIGN ALTERNATIVES
Pressurization Facilities
STEP. As noted earlier, household wastewater is collected by the building drain and trans-
ported from the home through the building sewer to a septic tank. A few investigators have charac-
terized household wastewaters and per capita flows.23"25 Mean flows have been found to vary from
43 to 50 gallons per capita per day (0.16 to 0.19 m3 per capita per day). Table 1-1 represents the
results of the most extensive of these studies24 for homes with and without garbage grinding against
typical municipal wastewater analyses for the same parameters.26 Generally, the wastewater gener-
ated at the home is more concentrated in most pollutant categories than a normal municipal waste
(primarily of domestic origin), which has been diluted by infiltration and other extraneous water
sources in municipal gravity sewer systems.
Significant treatment occurs in a septic tank. Primarily, the septic tank serves as a device for
removing settleable solids and grease. Heavy solids settle during the multiday nominal detention
period, while grease and other floatables collect in the scum layer. A cutaway view of a septic tank
is shown in figure 1-11. Anaerobic biological activity occurs sporadically, which causes some lique-
faction of accumulated solids. This digestive action produces gas that rises as bubbles in the system,
and the inlet flow patterns are quite variable. Both of these occurrences reduce the effectiveness of
the septic tank in retaining captured solids. A well-designed septic tank generally removes from 80
to 90 percent of the hexane extractables (grease), 70 to 90 percent of the suspended solids (SS)
(including all the grit), and 50 to 80 percent of the biochemical oxygen demand (BOD).27'30 In the
case of grease removal, the septic tank is an excellent grease trap, not only because of its inlet and
outlet configurations but also because its size allows the grease to cool and congeal for easier separa-
tion. SS removal may be temporarily reduced or negated during extended hot summer periods be-
cause of increased anaerobic digestion and resulting gas production and mixing from rising bubbles.
BOD removal is higher than that normally credited to primary sedimentation. Typical septic tank
effluent may have the following analysis:
• BOD5, 100-180 mg/1
• SS, 50-75 mg/1
• Grease, 10-20 mg/1
Table 1-1 .-Household wastewater characterization
[mg/l]
Parameter
BOD5
TSS
vss
TKN
NH3-N
TP '
Grease
Household
wastes2 4
Without
grinder
415
296
222
51
11
33
123
With
grinder
465
394
309
52
10
32
129
Typical
munici-
pal26
medium
strength
200
200
150
40
25
10
100
12
-------
Inspection
ports
Figure 1-11. Cutaway of septic tank.
The septic tank effluent then flows to a receiving tank, as depicted in figure 1-6, which houses
the pressurization device, control sensors, and valves required for a STEP system. The heart of the
STEP system is, of course, the pressurization device. Normally, small centrifugal pumps have been
designed or employed for the STEP systems. The oldest STEP systems are in the Miami area. The
pumps used at Port Charlotte and Port St. Lucie are all manufactured by the Hydromatic Pump
Company. Almost all of these units are Models SP 33A, with the head-discharge curve shown in
figure 1-12. The same pumps also are used in two STEP systems at Priest Lake, although some 0.5-
and 1-hp (0.37- and 0.75-kW) pumps are also included in these systems for locations where higher
heads were required. The EPA demonstration project at Bend, Ore., uses similar pumps manufac-
tured by Peabody-Barnes, which are driven by 0.5-hp (0.37-kW) motors.
All of the above sump pumps are submersible and generally retail for around $200. Several
units are equipped with bronze impellers said to reduce potential corrosion problems. Pumps with
impellers made of plastic materials are now available from several manufacturers. This development
should significantly extend the present 10-year life estimate for the pumps now in use. The cost of
the new pumps is not significantly higher than the standard units.
The design of these systems requires a proper septic tank installation and an effluent holding
tank containing the pump, level controls, valves, and piping. The discharge piping is essentially the
same for any design alternative and will be discussed along with the pressure main design alterna-
tives. Design decisions for effluent holding tank installations include material of construction, diam-
eter, and working levels for the tank, pump choice, and ancillary needs.
The effluent holding tank can be made of any material suited for septic tank use, including
properly cured precast or cast-in-place reinforced concrete and steel tanks meeting Commercial
Standard 177-62 of the U.S. Department of Commerce, with proper anticorrosion coatings.6 The
Albany project report8 indicated the epoxy-coated steel tanks underwent severe corrosion during
13
-------
30
20
a
00
o
o
LU
I
10
20
40
60
80
DISCHARGE, gal/min (1 gal/min = 15.85 l/s)
Figure 1-12. Head-discharge curve for SP33A pump.
the study. Molded fiberglass, reinforced polyester (FRP) resin tanks were found to be quite accept-
able on that project. The Phoenixville and Bend studies used concrete tanks with no apparent diffi-
culty11'11 while the Grandview Lake project12 used FRP, precast concrete and steel tanks, with no
mention of corrosion problems. The Miami systems use fiberglass tanks and the Priest Lake systems
use steel tanks with a litumastic coating.6 It should be noted, however, that these experiences have
been of a relatively short duration and do not provide long-term information.
The size of the effluent holding tank is a function of a number of variables. A typical unit
design is shown in figure I-13.h For single-family dwellings, the ASCE report7 determined that a
minimum of 30 gallons (114 litres) of net storage capacity was required for a 12 gal/min (0.76
litre/s) discharge rate. The concern for storage capacity relates to the submersible pump's ability to
handle the maximum short-term flow from a home. Simultaneous discharge of a bathtub and an
automatic washing machine was cited as the most likely maximum condition, producing a flow of
46 gallons (174 litres) in 2 minutes. The use of 30 gallons of net storage capacity (volume between
cycle initiation level and overflow pipe) has proved adequate at Albany and Grandview Lake. A
70-gallon (276-litre) storage capacity was used at Phoenixville. Other considerations for required
tank diameter include providing repair personnel with access to the defective pump or other mal-
eH. Schmidt, General Development Utilities Company, personal communication.
hL. R. Clark and J. E. Eblen, C & G Engineering, personal communication.
14
-------
M
Control
box
Locking
device
extension
Gate
valve
extension
Figure 1-13. Typical effluent pump chamber.
15
-------
functioning item for repairs. If the unit is located within 3 feet of the surface, required diameter
may be reduced to 24 to 30 inches (0.61 to 0.76 metre), but small diameter units have been trou-
blesome at one GP location where grease accumulations interfered with float switches.14 In loca-
tions where soil conditions do not provide adequate bearing strength or high ground water levels
occur, a concrete pad or collar may be required, as shown in figure 1-13.
The working levels in a tank are the levels at which the pump originates and terminates opera-
tion (fig. 1-13). The volume of the tank between these levels is considered the working volume,
which must be discharged during each cycle of operation. In actuality, the volume discharged during
a cycle is always greater because the influent flow continues during the pumping cycle. This increase
above the working volume is minimized by higher discharge (pumping) rates. At Albany the average
operating cycle varied from 39 to 112 seconds for GP units having working volumes between 10 and
30 gallons (38 and 114 litres), while at Phoenixville monthly working volume averages varied from
33 to 137 seconds per 20 gallons (76 litres).8-11 These times are a function of the pump character-
istics, the total dynamic system head at the time of each cycle, the placement of on-off control
sensors, and the wastewater flow and duration at the time of cycle inception. For a single home
installation the Environment-One (E/One) design (Model GP 210) indicates a 10- to 14-inch (25- to
35-cm) differential that corresponds to about 20 to 28 gallons (77 to 106 litres) per cycle for their
standard tank.22 The Hydromatic GP design (Model CSPG-150A) indicates a 6-inch (15-cm) differ-
ential that would correspond to approximately 12 gallons (45 litres) in a similar tank.31 The Miami
STEP systems use 11- and 24-inch (28- and 61-cm) differentials, which correspond to 22 and 47
gallons (83 and 178 litres), respectively.6 The differential determines the number of actuations or
cycles per day and their duration if all other factors are equal. The relative value of fewer, longer
cycles versus more, shorter cycles is not now quantified, but implications as to the relative merits of
each can be derived in other sections of this seminar publication. Now it appears that manufac-
turers' standard designs control the issue.
A similar statement can be made about the type of pump control switches used in STEP sys-
tems. Many pump manufacturers offer "packages" that may include level control switches, control
panels, wiring, and simplified maintenance systems. Although several types of control switches ex-
ist, only two types have been used in the manner required by pressure sewer designs. The first type
is the pressure sensing tube, which is standard equipment on one GP unit and has been used in a
privately owned system near Bend. In the Albany project 1-inch (2.5-cm) tube openings were rejec-
ted for GP units because of grease buildup that caused them to malfunction. After replacement with
pressure-sensing tubes having 3-inch (7.6-cm) openings, no further problems occurred.8 No pressure-
sensing tube malfunctions have been noted in either the Phoenixville or Grandview Lake proj-
ects.11-14 The Miami STEP systems use diaphragm-type pressure switches, and it is reported that
these devices become the major source of maintenance problems after 2 years of service because the
diaphragms lose their elasticity.6
The other major type of level control is the mercury float switch. This type of control device
consists of a mercury switch sealed within a float made of a noncorrosive material. As the water
level rises, it causes the float to either rotate or keel over to a position where the mercury switch
actuates the pump or, conversely, terminates its operation. Several forms of mercury switches have
been used in pressure systems, but usually the switch is either attached directly to the pump hous-
ing or is suspended from a stationary point above the liquid. This type of control is also standard
equipment for several pump and GP malfunctions. Some difficulties were experienced with this
type of control at Grandview Lake, but most problems related to faulty manufacture and shipping
and installation problems rather than conceptual shortcomings. The Priest Lake and Bend systems
use mercury float switches, with few reported problems.
Some of the ancillary factors that must be considered are the tank location, depth, covering,
electrical connections, warning signals, and contingency items. Because the effluent holding tank
eH. Schmidt, General Development Utilities Company, personal communication.
16
-------
follows a septic tank, gravity flow would dictate that the tank inlet be below the liquid level in the
septic tank. In most cases it is good practice to locate the pumping chamber as close as possible to
or as an integral part of the septic tank. Certain circumstances, however, may require greater physi-
cal separation of these tanks. Because septic tanks are usually located in the rear of a house and
sewers are generally in the front, it may be more economical to locate the effluent holding tank in
the front in cases where a natural slope exists toward the pressure main. Tanks or manways above
tanks may be covered with any load-bearing material, such as prestressed concrete, protective-
coated steel, or plastic. The covers should be attached in a watertight manner by gaskets or grooves
and should be sufficiently above the ground to prevent entrance of normal surface runoff. They
should be made as decorative as possible without impairing their accessibility. Trade-offs must be
made between ease of access and protection against vandalism. The final design will have to take
into account both factors as they relate to the locality under consideration. One suggestion is the
use of covers that incorporate locks requiring the use of a special tool to open them.
Electrical connections to the main panel in the house must be made according to local codes.
Approved underground wiring is recommended for both the pump and control circuits, which
should be wired separately—that is, have separate fuses or circuit breakers—from other household
lines and incorporate a fused electrical disconnect adjacent to the controls for use by the service-
man. The controls should be located in the garage or the basement, if an outside entrance is avail-
able. Outdoor locations must be designed to thwart would-be vandals. The choice of pump must be
compatible with the available electrical service, for example, 110 or 220 volts, single or three phase.
A high-level alarm light or audible device (bell or buzzer) should be located in the house where any
malfunction can be quickly noted by the occupants. If an audible alarm is used, a reset button
should be conveniently located so that relief can be easily and quickly obtained. The pump should
be wired for automatic level control with a manual override located at the control panel. Electrical
connections to the pump should be easy to disconnect if the pump must be removed for servicing,
but they must also be completely watertight.
The primary contingency concerns of the designer are the possibility of a power failure and the
ability of the system to handle a pump malfunction. It has been noted22 that, during the period
from 1968 to 1972, the 187 power outages recorded in the United States lasted for the times shown
in table 1-2. Outages of more than 9 hours' duration were caused by major natural disasters, such as
floods, hurricanes, and earthquakes. Under such conditions, it is unlikely that some septic tank
effluent overflow will significantly add to the total effect of the tragedy. Because 9 hours appear to
be a reasonable maximum outage, the system should be able to absorb the flow from the house for
that period. One can very conservatively assume that no more than 50 percent of a daily household
wastewater flow of about 200 gallons (0.76 m3) would occur in that period considering the prob-
ability of reduced water use during the power outage. Because septic tanks usually have 6 to 12
inches (0.16 to 0.33 metre) of freeboard, a rectangular 1,000-gallon (3.8-m3) tank could hold any-
where from 100 to 200 gallons (0.38 to 0.76 m3), excluding the capacities of the effluent holding
tank and house sewer. Also, the loss of power in rural areas that are served by individual wells and
cisterns essentially eliminates any possibility for wastewater generation because water supplies be-
Table (-2.—Power outages recorded in the United States, 1968-72
Percent of total outages
53
81
89
95
97
Duration,
hours
<1
<2
<3
<5
<9
17
-------
come inaccessible. Because the system seems to handle power outages, the primary potential diffi-
culty appears to be malfunctioning mechanical components.
The time involved between determination of a malfunction by the alarm light and the arrival of
a repair crew is a function of the institutional approach of the sewer district. The approach, in turn,
would be influenced by such factors as the prior existence of soil absorption fields, the size of exist-
ing septic tanks, and the number of system contributors. For example, if solid absorption beds were
previously in use, an overflow from the effluent holding tank to the bed could be sufficient to per-
mit a normal 5-day workweek for repair personnel. Also, a larger septic tank, with its increased stor-
age capacity above its normal water level, would allow a somewhat more generous response time
than would a smaller one. A larger number of contributors would justify having a larger repair staff
employed by the authority; if the number is smaller, a contract servicing arrangement with a private
firm might be more advantageous.
Rose3 has posed the question of who should purchase, install, and maintain pressurization
facilities. The unanimous opinion of several authors1-15'32"34has been that the maintenance of the
pressurization unit, house service connection, and pressure mains, and the installation of the mains
should be the responsibility of the authority (district, county, and so forth). Bowles1 and Coch-
rane15 have recommended that a sewer district own the pump, install it and the service line, and tap
the pressure main for a fee. The homeowner would then be responsible for installing all lines, tanks,
and electrical connections to the pump and paying for the electricity to operate the pump. The
district would perform maintenance and make repairs, on request, for a service charge. Voell indi-
cates that homeowners prefer to have a utility arrangement whereby regular maintenance and repair
would be performed by utility employees.33 Leckman has discussed the option of public versus
private ownership of the pressurization facility, with the authority providing all maintenance and
repair services in preference to a private contractor.34 All sources recommend a well-stocked repair
shop with sufficient replacement units for quick and easy exchange with malfunctioning units to
allow for repairs to be made at the shop.
The number of replacement pumps that must be available is primarily a function of their reli-
ability, the number of units in the system, and the rate of repair. For example, if a system of 100
pumps with a reliability factor of 0.99 (work 99 percent of the time) were involved, the necessary
number of standby units could be computed by use of the binomial distribution. Accordingly, the
following probabilities exist for the respective situations:
• One pump fails, p = 0.370
• Two pumps fail, p = 0.185
• Three pumps fail, p = 0.061
• Four pumps fail, p = 0.015
On the basis of this type of analysis, the authority would then have to make a decision on the num-
ber of spare pumps. Other factors involved are the average time required to repair a pump in the
shop, the size of the maintenance staff, and delivery times for spare pumps and parts. Under present
conditions, the last item may be a major concern.
Assuming that a reasonable response time by the authority's maintenance and repair crews
would be less than the 9-hour maximum power outage figure, the STEP system appears to require
no auxiliary holding capacity. If, however, an existing SAS is available, the minor cost of installing an
emergency overflow drain from the effluent pumping tank to that system could be a prudent invest-
ment. The major concern in using an existing soil field is infiltration during wet periods that could
result in a reverse flow from the field to the effluent pumping tank through the overflow drain. 14>a
aS. M. Rees, SIECO, Inc., personal communication.
18
-------
The pump chosen for pumping septic tank effluent should be selected on the basis of reason-
able cost, reliability, proper head versus discharge characteristics, and compatibility with the appli-
cation. To be compatible the pump must be able to handle organic and some light inorganic solids
(negligible grit) and be reasonably resistant to sulfide and other septic effluent corrosiveness. The
reliability criterion should be satisfied by rugged construction and resistance to moisture penetra-
tion, for example, submersible or watertight properties. If the above criteria are satisfied, the pump
choice becomes one of economics and proper performance characteristics. In this type of applica-
tion a centrifugal pump is probably the most economic selection.
Although proper pump selection is well discussed in the literature,7'35>c a few other items
should be taken into consideration. To avoid major problems arising from dynamic dissimilarity, it
would be prudent to install only one kind of pumping unit for all installations. In this way the units
would be both geometrically and dynamically identical. Because little design information is avail-
able on this specific application of multiple pumping units, the methods of analysis discussed by
Metcalf and Eddyc and Flanigan and Cudnik35 should be helpful to the designer. The original
ASCE report7 indicated that the maximum economical curb pressure head should be equivalent to
69 feet (21 metres) of water and that the minimum pressurization unit discharge pressure head be
equivalent to 0 to 11.5 feet (0 to 3.5 metres) of water. Therefore, a maximum discharge pressure
head of about 81 feet of water (24.7 metres) was required. These numbers are only a function of
the assumptions made in this study, but they represent reasonable target pressures. Higher working
pressures may require stronger and more expensive piping materials, while lower pressures may re-
strict the capabilities of the system. There is no reason, however, not to use less restrictive pressuri-
zation criteria when the conditions of a particular site do not demand them.
As noted earlier, the most popular pump for STEP systems has been the low-head, submersible
sump pump. The primary reasons for this pump's popularity have been generally flat terrain ap-
plications, low cost, and availability of parts. Unfortunately, until very recently no commercial
pumps were available that were specifically designed for this type of application. Alternative units
are needed to meet the requirements of these applications. One alternative approach for the STEP-
type pressure sewers is to use a pneumatic ejector. The unit shown in figure 1-14 is one of two types
of pneumatic ejectors that have been developed expressly for pumping septic tank effluents. These
units are manufactured by Clow Corporation and Franklin Research Company. Both units require
an air compressor to impart discharge pressures.
Design methods for selection of STEP'S are not easily found in the literature. A number of
these types of installations have been designed and built, but very little information on them has
been published. The hydraulic conditions that must be satisfied are primarily related to system size,
pipe sizes and lengths, probability of simultaneous pumping, and growth characteristics of develop-
ment. Most of these conditions will be discussed with the pressure main and service line design alter-
natives. The methods used to determine the applicability of a given pump have been outlined by
Bowne,32 Metcalf and Eddy,c and Flanigan and Cudnik.35 Starting from the head-discharge curve
shown in figure 1-15, the operating point of each pump can be found by the methods that follow
(using English units).
The static head on the pump is determined by
H8 = ht- hp
where
Hs = static head, feet
ht = elevation of discharge point of pressure main, feet
hp = elevation of pump, feet
CR. E. Lawford, Peabody Barnes Company, personal communication.
19
-------
Check valves
Figure 1-14. Pneumatic ejector (courtesy of Clow Corporation).
The approximate dynamic head caused by pipe friction and other pipe constrictions, such as
valves, bends, elbows, and other fittings, is determined. The pipe friction losses are usually com-
puted by use of the Hazen-Williams formula,
where
H
F1 = pipe friction head, feet
d = pipe diameter, feet
V = velocity of flow, ft/s
C = Hazen-Williams coefficient
20
-------
3:
Q"
LU
I
O
H
"S
O
= F(Q)
CAPACITY, Q
Figure 1-15. Operating point of pump with characteristic H = F(Q) and a pipeline with characteristics H$+HL
while losses in valves and fittings are computed by use of the formula,
y2
"UFO — •** ~rr~
where
g
= fitting friction head, feet
= gravitational constant (32 ft/s2)
= fitting coefficient
Values of K for each type of fitting can be found in various hydraulics textbooks. The value of
the Hazen-Williams coefficient C has been subject to interpretation. The Plastics Pipe Institute indi-
cates that tests in several laboratories of new and used thermoplastic pipe resulted in C values rang-
ing from 155 to 165, and recommends the use of a conservative value of 150.36 The Albany and
Phoenixville designs also used this value. Flanigan and Cudnik indicate that a C value of 150 is
21
-------
proper for clean water applications, but because of grease and other interfering matter present in
wastewater, they recommend a C value of 140. They note that this conservative value should permit
easier operation of the system during periods of stress and, if found through experience to be overly
conservative, it can be revised upward. The Grandview Lake system was designed using a C value of
130.a
Applying the foregoing information at various values of Q (discharge, gal/min) will yield the
total dynamic head (TDK) by:
TDH=tfs +HFl=HF2
Because Hp^ and Hp2 are functions of discharge, the TDH is represented by a nonlinear increasing
curve in figure 1-15, where H± = HFl + HF2. The intersection of these curves is called the "operat-
ing point." In normal pump selection design practice, the pump that has its optimum efficiency at
this point would be chosen. However, because these pumps operate under varying conditions of
TDH in a pressure sewer system and the cost of inefficient operation is negligible in this type of
operation, this requirement is not very important.
This analysis is used to "plug in" the information of a tentative system design to the extreme
cases, for example, the pumps requiring the most and least heads for operation. In the former case,
a test of the highest TDH versus Q can be related to the adequacy of a particular pump design. In
the latter case, a test can be made to determine any possible difficulties related to pump overloads
and cavitation. Cavitation and overloading are unlikely with low-specific-speed pumps of the type
used for this application, especially with discharge line losses. The maximum and minimum TDH
analyses determine the variation in flows that can be expected from single-pump operation.
The problem of multiple-pump operation is far more complex. The pump head-capacity curve,
shown in figure 1-16, is assumed for all pumps in the system. Line and fitting losses, related to the
service lateral that feeds the pressure main, are combined with the original head-capacity (H-Q)
curve to produce a modified H-Q curve, as shown in figure I-16(c). Each pump then must be refer-
enced to a single location on the main line, usually the point where the pump closest to the dis-
charge end of the main enters the main line. This point or station is shown in figure 1-17.
This referencing can be accomplished by a repeated series of combinations, that is, pumps 1
and 2 referenced to station 2 to get a combined H-Q curve at that point. This combination is then
referenced to station 3 and combined with pump 3 and so on, until all the operating pumps are
combined at the final point (in the case of fig. 1-17, it is station 4 on the main line). The referencing
process involves the conversion of the modified H-Q curve of the farthest pump (1 in this case) to
the conditions at the main line connection of the closest pump (station 2 in this case) by converting
the static and main line friction heads between the two stations. The major steps in the referencing
process are shown graphically in figure 1-16. This simplified example assumes that the pump and
main line station are at the same elevation. In figure 16(a), the service discharge line losses are sub-
tracted from the original H-Q curve to get the modified H-Q curve, and these losses are assumed to
be the same for both identical pumps. In figure 16(b), the elevation difference and piping losses be-
tween stations 1 and 2 are shown and are applied to the modified curve from figure 16(a), to get the
new H-Q curve for pump 1 at station 2. In figure 16(c), the pump curves at station 2 are shown
separately and combined, along with the system curve, from station 2 to station 3. Theoretically,
this type of analysis must be repeated until all the pumps in the system are related to station 4 in
this example, or whatever the final station might be. Actually, a limited number of combinations
will suffice, as discussed later in the section on pressure main design.
Bowne32 has proposed using a much-simplified pump selection method, which takes advantage
of a centrifugal pump's flexibility. Essentially, he establishes the hydraulic gradeline for the system
aS. M. Rees, SIECO, Inc., personal communication.
22
-------
H
Original pump H-Q curve
Modified H-Q curve
Discharge piping losses
H-Q curve
for pump 1
at station 2
Combined H-Q curve
Operating point
Modified (as in a)
H-Q Q curve for pump 2
Figure 1-16. Multiple pump operating analysis.
and determines the difference in elevation between the pump level and the hydraulic gradeline.
Then, knowing the length and size of the service line, the modified pump H-Q curve will yield the
adequacy of the pump at that location. By allowing a variance in the size of the pump discharge
line, further design flexibility can be obtained. The number of simultaneous operations that may
23
-------
Reference point
Flow
Figure 1-17. Reference point for multiple pump operating analysis.
reasonably be expected will be discussed later, as will pipe sizes and other main line design factors
that may impact pump selection. One of the advantages of centrifugal pumps is that they can oper-
ate for reasonable periods in a no discharge condition at heads greater than the maximum or shut-
off head without significant damage to the unit. Flanigan and Cudnik relate an experiment wherein
a 1.5-hp (1.2-kW) pump immersed in 70 gallons (265 litres) of water for 4 hours raised the tempera-
ture of the water to 122° F (50° C). This characteristic provides a cushion if a temporary high-head
condition occurs in the pressure sewer system because of a malfunction or extraordinary flow. But
pump discharge connections must be well designed and properly installed to resist the high pressure/
high temperature conditions of significant duration without serious damage.
GP's. Much of the preceding analysis provided for STEP stations is applicable to GP installations
because the pressurization operation is essentially the same. Obviously, the septic tank is eliminated
from the system along with the necessary inspections and cleanings that are associated with it. This
exclusion may well be a determining economic factor when choosing the appropriate system for a
home where no septic tank exists or for a multifamily dwelling that may require an exceedingly
large septic tank to conform to local codes. It is also possible that a significant reduction in on-lot
piping can be accomplished by not having to be tied to the existing septic tank location. Another
degree of freedom is available with GP designs: the ability to install the unit in the basement of a
home, as shown in figure 1-10, for ease of maintenance and less-severe operating conditions. One
other significant difference is the fact that more data are now available for GP systems than for
STEP systems.
Design techniques will vary somewhat for GP installations compared to STEP installations,
with respect to such things as emergency storage provisions and commercial package availability.
But the basic design considerations are quite similar. The commercial aspect of the GP designs is
worthy of discussion. As noted earlier, the concept of a GP for sanitary sewage transmission was
integral to the original ASCE study.7 As part of that study, General Electric Company developed a
commercial GP unit in concert with waste generation, hydraulics, and other engineering factors.
Certain members of the General Electric staff who were involved in this study later left and formed
the E/One Corporation, which became a pioneer in GP-pressure sewer development. Since that time,
E/One and the Hydromatic Pump Division of Weil-McLain Company have become the leading sup-
pliers of GP units. There are other firms in competition with them, such as Robbins & Meyers,
Toran, Peabody-Barnes, and Empo-Cornell (no attempt has been made to compile a complete list
because any such list would be accurate only at the time it was compiled). E/One and Hydromatic
units represent the two major GP design choices, that is, a progressing-cavity (semipositive-
displacement) pumping element and a centrifugal pumping element, respectively.
The H-Q curves for the basic single-home models of each manufacturer presented in figure 1-18
differ markedly, as would be expected. The E/One pump has been shown to be capable of operation
24
-------
•E/One (model Farrell 210)
90
LJJ
I
O
60
o
LU
X
o
30
Hydromatic (model SPG 150)
10
20
30
40
CAPACITY, Q, gal/min
Figure 1-18. Grinder pump characteristics.
25
-------
above the 81-foot (24.7-metre) design limit for a considerable number of cycles by the National
Sanitation Foundation (NSF).37 The extreme condition of operation would occur immediately
following restoration of power after a prolonged outage. Although it is unlikely that normal waste-
water generation patterns would exist during such an outage, the assumption is that all or a signifi-
cantly large number of the total units could be at or above their discharge or actuation levels. There-
fore, at the instant power is restored, it could be assumed that all units would commence to dis-
charge. However, because the resultant TDH would be greater than both maximum heads, a sequen-
tial pumpout would likely occur. The sequence would initially permit discharge from those units
that pump against the least TDH, for example, in the case of a "flat" system, the units closest to the
discharge point. The other units in the system that cannot discharge because of excessive TDH must
wait their turn (thus, the sequential pumpout). During the period of excess TDH, centrifugal units
will rotate without discharging, with the input energy being dissipated as heat. The previous section
indicated that 4 hours of such operation, with a 1.5-hp (1.1-kW) GP, raised the temperature of 70
gallons (263 litres) of water to 122° F (50° C).35 Because all the units in a small community system
would likely be emptied in less time than this, no difficulties should be experienced. The pro-
gressing-cavity design, when pumping against excessive TDH, uses a thermal overload protector with
automatic reset capability. Because this type of unit can pump at destructive pressures when un-
checked, the thermal overload feature is intended to protect the motor and discharge piping from
potentially damaging excessive pressure development. The automatic reset then allows the unit to
cycle as many times as necessary until the tank contents can be discharged.
The minimum TDH operating condition is the single unit discharge. From figure 1-18, it is
apparent that the centrifugal unit will pump a given quantity of sewage in a significantly shorter
time at any TDH below 88 feet (26.8 metres) of water. This provides a higher velocity in the dis-
charge system and reduces the probability that simultaneous pump operations will occur.
Each manufacturer offers a package that includes the pump and its mounting, holding tank,
controls, and other accessories. The E/One unit includes two flapper check valves, a pressure-sensing
switch, and an antisiphon valve. The Hydromatic unit includes a ball check valve, a mercury level
switch, a gate valve, and a control box. Each includes various ancillary items in their commercial
packages.
The problem of emergency overflow storage was addressed in the previous section where septic
tanks were found to offer sufficient storage capacity for extended power outages or system mal-
functions. GP holding tanks are not normally large enough to provide proper storage of raw sewage.
A solution proposed to handle this contingency includes the one shown in figure 1-5. These absorp-
tion pits and beds were installed when an existing ST-SAS was not available. The specifications re-
quired that the bed have a minimum volume of 500 ft3 (14.2 m3) and be filled with pea gravel. The
operating principle of this type of system is that the overflow will percolate into the ground at some
finite rate. As pointed out previously, one caution with this type of system is that infiltration may
occur from the bed or pit back to the GP tank during wet periods.3
Another design proposed by Schultzd is shown in figure 1-19. This 200-gallon (0.76-m3) hold-
ing tank lies between the house and the GP unit. Under normal conditions, the raw sewage flows
along the bottom of the tank and does not accumulate. During an emergency overflow condition,
the sewage backs up into the tank and not into the house because the top of the tank is at least 2
inches (5.1 cm) below the level of the lowest house drain. When power or system operation is re-
stored, the emergency tank drains by gravity into the GP tank. The Phoenixville and Bend systems
also use existing septic tank systems for emergency overflows.11»h
Leckman has discussed various other alternatives, such as standby power, water service termi-
nation, other holding tank designs, and interconnection with an adjacent GP unit.34 He estimated
aS. M. Rees, SIECO, Inc., personal communication.
dj. Schultz, Becher-Hoppe Engineers, Inc., personal communication.
hL. R. Clark and J. E. Eblen, C & G Engineering, personal communication.
26
-------
Existing ground
Vent cap
4-inch valve with box
4-inch cleanout
NOTE- All holes in vault walls
shall be sealed and
watertight solvent
welds may be utilized.
NOTE- The top of the holding
tank shall be at least two
inches below the lowest
floor level of the
dwelling.
To GP unit -« V [X] > ~|T
fs-g".-. -
•t-'
1 M—
1 CO
1
1
200-gallon ;
emergency holding |
tank
4 feet
Slope 1/8 inch per foot
4 feet
Building
sewer
4 inches
Figure 1-19. Emergency tank design.
the cost of a standby generator at $750 (1972 prices) with substantial maintenance costs. Water
main service could be terminated at the curb during power failures or planned maintenance activi-
ties to prevent overflow conditions. Holding tanks could be conveniently located and pumped out
by septic tank pumpers, that is, interim holding tank systems. If GP units were close enough, inter-
connecting overflows would be possible during periods when malfunctioning of one unit occurs.
It is essential that GP systems have some type of contingency arrangement, based on the
assumption that public health considerations preclude the possibility of raw sewage contamination
of lawns, basements, and so forth. Confronted with this problem, the designer must choose one of
the foregoing alternatives or develop a new means of storage for such emergencies. The choice
should be influenced by the servicing arrangement, fluctuation of groundwater levels in the area,
and available existing facilities. Also, the contingency approach should minimize the cost to home-
owners.
System
Service Connections. Service connections between each pressurization facility and the pressure
main would be similar in almost any design. Key elements in this design are pipe material, pipe size,
valves, and the connection to the pressure main. The use of 1.25-inch (3.18-cm) inside diameter
plastic pipe has been documented as the best compromise between minimum required scouring
velocities for GP systems and minimum headloss considerations.6 This analysis was based on the
H-Q curve of the E/One GP unit, however, and is not necessarily valid for other pressurizatior,' facili-
27
-------
ties. Under these conditions the minimum discharge pipe velocity was greater than 2 ft/s (0.61 m/s)
and headless was less than 4 ft/100 ft (4 cm/m) of pipe length. The Albany project used 1.25-inch
(3.18-cm) PVC, Type I, schedule 40 service lines with PVC, drain, waste and vent (DWV) fittings.8
The DWV fittings were used because of their smoother transition properties compared to schedule
40 or 80 fittings. At the end of this project grease accumulations were found that resulted in reduc-
tions of as high as 40 percent of the lateral (service line) cross-sectional area. The Phoenixville sys-
tem used 1.5-inch (3.81-cm) PVC, SDR-26 service lines with schedule 40 fittings.11 The Grandview
Lake project used two types of PVC (SDR-21 and SDR-13.5), polypropylene, and polybutylene
service connections. The latter two types, with brass fittings, were found to be more costly and
more difficult to install but performed in an excellent manner when properly installed. The PVC,
SDR-13.5 was considered the best of all service lines, and the PVC, SDR-21 was considered the
worst on the basis of installation and operation performance.14 Most of the Grandview Lake service
lines used 1.0-inch (2.54-cm) piping. Bowles1 noted that the Lake Lyndon Baines Johnson Munici-
pal Utility District project uses PVC service lines from 1.25 to 2.0 inches (3.18 to 5.08 cm) in nomi-
nal diameter. The STEP system at Priest Lake incorporates 1.5-inch (3.81-cm) PVC, SDR-26 service
laterals.f The Miami systems use either 1.25- or 1.5-inch (2.54- or 3.81-cm) PVC, SDR-26 service
lines, while the Bend system uses only 1.25-inch (2.54-cm) PVC, SDR-26 laterals.6-11
The original ASCE study report7 indicated that polyethylene pipe had the advantage of being
plowed in for quick sewer installation. Essentially, "plowing in" refers to a system in which trench-
ing, feeding of coiled tubing or pipe, and backfilling are accomplished in a single operation. Despite
this advantage and the lower cost, polyethylene has not been used in any pressure sewer systems.
The reason appears to be that pressure-resistant fittings are not available for polyethylene. There-
fore, PVC has been almost exclusively used for pressure sewer mains and service connections, de-
spite its greater cost and its unadaptability to plowing methods. It is not inconceivable, however,
that future development will allow the use of polyethylene.
The need for check and gate valves in service connections is obvious. Because the main is under
pressure at all times, especially when one or more pressurization facilities are operating, check valves
or backflow preventers are required on all pressurization units. In addition, some type of gate valve
or equivalent is necessary to allow the isolation of each unit for repairs. As a result of the Albany
study, Carcich et al.2 indicated that shut-off valves are necessary both on the discharge side of the
pressurization facility (GP or effluent pump) and on the service connection just before the pressure
main (curb stop) to allow isolation of the pressurization facility or the pressure main. They also
noted that the flapper-type pump discharge check valves required a horizontal run of pipe on the
pump's downstream side to prevent the accumulation of solids that impeded normal operation of
the flapper. Their project report8 recommended that a 1-foot (0.30-metre) run of horizontal pipe
be used, while Leckman14 recommended 2 feet (0.61 metre). The Phoenixville system used a
bronze swing-type check valve in a horizontal section of pipe.11 The Grandview Lake installation
used two types of bronze check valves, one with a vertical hanging gate and one with a 45° seating
from the vertical. This latter design was reported to be superior.14 The need for a horizontal run
following GP discharge check valves was also documented, along with a preference for swing check
valves over ball check valves.14
The Miami systems use a single check valve and gate valve on the discharge line of the pressuri-
zation unit. The check valves successfully used include plastic ball check valves, brass flapper valves,
and plastic flapper valves. The gate valves used for these systems are made of brass or plastic. All of
the above types of valves have been trouble free for as long as 5 years. Serious problems have been
noted in these systems with valves made of steel or iron because of corrosion (iron sulfide), which
caused failure within 2 years.6
eH. Schmidt, General Development Utilities Company, personal communication.
*C. W. Rose, Farmer's Home Administration, personal communication.
hL. R. Clark and J. E. Eblen, C & G Engineering, personal communication.
28
-------
At least two manufacturers now offer flexible rubberlike check valves for pressure sewer sys-
tems. Although conceptually attractive, these devices will require field experience to determine if
they can retain their elasticity for an acceptable number of years under these severe conditions of
service.
On-off valve selection and location should also be described. A study of the relative merits of
gate and ball valves indicates that ball valves are more compact and lighter than gate valves, and they
offer slightly less fluid friction (pressure drop) when in the fully open position. These differences
are not normally so significant, however, that relative economics will not be the primary deter-
minant of final choice.
Although some designers have indicated a preference for locating the on-off valve outside of
the pressurization holding tank in order to permit valve operation without removing the tank cover,
such a design decision would necessarily be a function of the ease with which the cover can be re-
moved, redundancy of on-off valves on the service line, and economic trade-offs for the initial cost
of the on-lot facilities.6'32 It is likely that the valve would be located within a holding tank of suffi-
cient size where a curb valve is also used for the service line.
Some designers prefer backup check and on-off (gate or ball) valves; others feel that such back-
up is superfluous.6-32 The use of two on-off valves has been more common than dual check valves.
Usually, a corporation stop or curb valve is located near the service line-pressure main connection to
permit isolation of the entire service line and pressurization facility from the main, in addition to
the on-off valve, which is generally located at the pump holding tank. This arrangement permits
isolation of the pressurization unit from the service line in addition to providing a redundancy in
isolation from the main. The use of two check valves is less clear. If such a design is chosen, how-
ever, it could be accomplished by two separate check valves or a dual check valve in a single hous-
ing.
In order to connect each service lateral to the pressure main at Albany and Phoenixville, sani-
tary tees with 45° elbows for 1.25-inch (3.18-cm) connections were used.8'11 The Grandview Lake
system used curb, cock-tapping saddle connections to the main, as shown in figure I-20.14 The
choice of service line connection may be based on relative economics or headloss considerations.
One alternative would be to use standard 20-foot (6.1-metre) lengths of PVC pipe in laying the pres-
sure main with later return to tap the main for each service line connection at its most advantageous
location based on the user's lot geometry. Saddles and tapping tools for PVC are commonly avail-
able in most locations. Another alternative would be to provide a tee or valve connection between
standard lengths of PVC main during construction so that each potential user would have a connec-
tion for his lateral. Variations on these two alternatives are manifold, and the economic trade-offs
must be weighed against the objectives of the sewer authority. A list of suppliers and manufacturers
of the necessary service line pipe and fittings, as well as pressure main items, can be obtained from
the Plastics Pipe Institute, 250 Park Ave., New York, N.Y., 10017.
As noted previously, the size of service connection lines is normally between 1 and 2 inches
(2.5 and 5.1 cm) in diameter. However, the designer should determine the proper size based primar-
ily on the characteristics of the pressurization pumps used in his particular system; that is, a pipe
size that is proper for a 15-gal/min (0.95-1/s) pump may not be proper for a 30-gal/min (1.89-1/s)
pump. The choice, which is based on the trade-off between headloss (which includes service line
length) and minimum scouring velocity, becomes more difficult with increasing system size. For
smaller installations, where the static head pumped against is minimal, the use of smaller service lines
may be prudent. In larger systems, having significant static and dynamic head, the use of larger lat-
erals may be necessary. Greater flexibility in pipe sizing is available with STEP systems because of
less-stringent minimum velocity concerns. When the complete system hydraulics is analyzed, the
eH. Schmidt, General Development Utilities Company, personal communication.
29
-------
'9d
-.leanout
t
'
3 P
'
_^^- Solvent weld
S'/z-inch main line
3'/2-inch PVC, SDR-26, OD 4.0 inches
Tapping saddle
with curb valve for
pressure tap
1-inch Service line
JTo home unit
. > •
Bricks, stones, concrete pad
Figure I-20. Curb valve and riser installation.
decision can be made in context with all other hydraulic considerations, including friction losses
caused by valves and fittings. In the Albany study approximately 32 percent of the total friction
loss was computed to be caused by these items.8
Pressure Mains. Design considerations for pressure mains include the general layout, hydraulic
load factors, appurtenances, maintenance provisions, and the type and size of piping. The general
layout of a pressure sewer system was given careful scrutiny during the ASCE study.7 One of the
original objectives was to minimize the residence time of sewage in the system. Eliminating loops in
the system results in reduced storage time, less opportunity for solids settling and odor production,
fewer abrupt direction changes, and less headloss because of fittings. One potential difficulty with
eliminating loops is problems with pressure main repair, which will be discussed later. By employing
a single main with unidirectional flow, the designer can route that main via the most direct path to
its effluent point, taking into account grade and existing natural and manmade barriers. The E/One
design handbook also recommends the use of separate sections or branches, where possible, to mini-
mize difficulties that result from occasional main repairs.22 An example of this approach is shown
in figure 1-11, where one alternative could have been to extend the main (branch 7) vertically to the
top of the figure and feed it directly from each pressurization unit.
Because the pressure sewer systems at Albany and Phoenixville pumped against a static head,
that is, the terminal level of the pressure main was at a higher elevation than the remainder of the
system, maintaining a positive pressure in the main was no problem. Difficulties can occur if a posi-
tive pressure cannot be maintained because of high points in the system. Several investigators7'32'35
have recommended positive pressure maintenance at all points within a system, but such a conclu-
sion must be weighed against a possible combined pressure gravity system where the latter is feasi-
ble. If a system were located on flat terrain, the use of a standpipe could accomplish the need for a
positive pressure gradient. The investigators7'32'35 all describe the use of pressure-sustaining and
pressure-control valves to provide sufficient back pressure or artificial head to prevent draining of
30
-------
portions of the line and the associated problems of siphoning and flow impairment. If undulating
terrain or long downhill runs are involved, these devices may be necessary if a combined system is
not feasible. There is no reported experience with these devices, but the ASCE report7 and
Bowne26 describe their function in greater detail.
The Grand view Lake, Phoenixville, Radcliff, and Priest Lake systems used air-release valves;
the E/One design handbook,22 the ASCE report,7 and Flanigan and Cudnik35 recommend them.
However, exact criteria for their need and placement are lacking. Air or any gas accumulation in
pressure mains will increase flow resistance and the headloss against which the sewage must be
pumped. The potential sources of air in the sewer include insufficient purging after main filling and
testing, malfunctioning pumps, or the release of air that was in solution at the time of pumping.
Other gases could also be released from the wastewater in the pressure main by pressure reduction,
biological activity, or chemical reactions. Gas bubbles accumulate along the crown of the sewer and
ultimately move toward the high points in the pressure main. Once located there, the multibubble
or "foamlike" structure tends to disappear in favor of a single large bubble configuration. This large
bubble continues to grow until the drag force of the flowing liquid exceeds the pipe centerline com-
ponent of the buoyant forces long enough to carry the bubble in the direction of flow (with mini-
mum slippage) beyond the next low point in the system. If the duration of necessary flow is insuffi-
cient, the bubble will return to its original location. Kent has determined the relationship between
slope, velocity, and bubble size.38 According to his studies, at a 10° downslope, a bubble 3 inches
(7.6 cm) long requires a velocity of almost 1.2 ft/s (0.36 m/s) to move through a 2-inch (5.1-cm)
pipe. The additional headloss caused by the bubble is a function of relative volume of air to water.
For the example above, the additional headloss because of air could be as high as 50 ft/1,000 ft (1.5
to 15 mm/m) of pipe length. Flanigan and Cudnik35 and Farrell39 have also examined this required
bubble transport velocity but have not defined the requirement for sufficient duration of velocities
in excess of this minimum. The hydraulic calculations are best described by Kent, but the require-
ment for sufficient duration is quite difficult to achieve because pressurization units rarely operate
for much more than 1 minute or 2. If sufficient velocity were present in the main to move the bub-
ble at a velocity of 1 ft/s (0.30 m/s), a 1-minute operation would only move the bubble 60 feet
(18.3 metres). Because each pumping generally contains about 15 gallons (56.9 litres) of sewage,
the internal volumes of three common PVC pipes would correlate to the displacement distances
(1 foot = 0.305 metre) shown in table 1-3. Minimum slippage is likely to occur between the liquid and
the bubble so that the bubble will be displaced at a rate essentially equal to the average velocity of the
fluid times the duration of flow.38 By using some engineering judgment, combined with calculated
velocities, displacement volumes, and system geometry, the designer can make reasonable decisions
about the need and placement of air-release valves. (Bowne presents an interesting discussion on this
subject.32)
When it is determined that an air-release valve is required, the type (automatic or manual) must
be chosen. Automatic air-release valves are available for sewer application that permit accumulated
gas to escape at the valve without loss of liquid. Such valves require regular maintenance in the form
of inspection and flushing to minimize clogging by sewage solids and grease. Manual valves are essen-
tially vertical risers attached by a corporation cock to the crown of the pressure main at a high
point. Although maintenance is minimal, a regular schedule of manual operation must be used to
release trapped air. Both types require an access way for required operation and maintenance.
Table \-3.-lnternal volumes of PVC pipes and displacement distances
Nominal size, in
1.25
1.5
2.0
3.0
Schedule 40, ft
193
142
86
39
SDR-21,ft
163
124
77
37
SDR-26, ft
1R7
120
7fi
qc
31
-------
The pressure sewer main is subject to malfunctions that directly or indirectly cause shutdowns.
When these occur, all or some of the homes will be deprived of service for a period of time. Because
the branched or dendriform design of a pressure sewer system has already been shown to be desir-
able, all homes upstream of the shutoff point will be without service. Two questions that then must
be answered are:
• How quickly can repairs be made?
• Can the shutdown area be bypassed?
The design aspects of these questions are:
• What main line ancillary facilities are required?
• What would be an optimum spacing?
The question of main line valve requirements and spacing has been dealt with in numerous
cases, even though they were not present in the Albany system and were not adequately used either
at Phoenixville or Grandview Lake. In the Grandview Lake project, however, the desirability of
using main line gate valves to isolate sections for repair was noted by the engineer.14 The NSF
sewer layout described in the ASCE report7 suggested 600-foot (183-metre) intervals between valve
and cleanout facilities, while Leckman34 suggested a maximum distance of 400 feet (122 metres).
The typical in-line cleanout facility is shown in figure 1-21. Two facts should be noted regarding this
figure. First, this design does not provide for a temporary bypass to minimize the number of units
out of operation; second, the space and depth requirements of the valve box are obviously reduced
compared to a conventional manhole. The second factor constitutes additional capital cost savings.
The distance between cleanouts finally chosen by the designer will be a function of the size of the
system, topography, and layout. On long runs in larger systems the 400- to 600-foot (122- to 183-
Meter box
Figure 1-21. In-line cleanout.
32
-------
metre) maximum separation would relate to the density of contributors and the capabilities of
mechanical cleaning equipment. A minimum requirement would include one of these facilities lo-
cated on the main line wherever a branch main is to connect, as shown in figure 1-10, or wherever
pipe sizes change. State and local regulatory officials will necessarily be the final arbiters in all cases.
The problem of bypassing between valve locations relates to the system's physical design. Be-
cause some main line repairs are likely to require significant time to complete, the question of emer-
gency overflow capacity again arises. The primary design question involves the relationship between
the isolated (down) portion of the system and the upstream portion. Why do the upstream users
need service when the isolated users do not? Two things that affect the validity of this question (but
do not eliminate it) are: Isolated users can more easily be contacted to request reduced use of water
during the outage and major repairs may take longer than the normal excess holding capacity can
accommodate. In answering the question of bypassing, the designer should give strong consideration
to the maximum repair operation, which would probably involve locating, excavating, and repairing
a main break, as it compares to the emergency overflow capacity of the contributary units. If, after
the comparison is made, the repair time does appear to be excessive, some form of bypassing should
be considered. If the expected frequency of such occurrences is low, the bypassing arrangement
considered should require minimal investment; if high, a more sophisticated approach may be re-
quired.
Two bypassing arrangements that have been proposed involve temporary hose connections and
parallel mains. Voell discussed the use of temporary firehose connections. The only modifications
required to implement this approach would involve the use of tees immediately upstream and down-
stream of a main line ball or gate valve, with a ball or gate valve and threaded fitting attached to
each tee stem in the main line valve box. The provision of parallel lines at the time of construction
is a relatively expensive solution. However, when the hydraulic load factor is low at the time of
construction, one alternative technique that will be discussed later is the use of parallel lines of dif-
ferent sizes. If this approach were adopted, the existence of these parallel lines might offer a ready
solution to the bypass problem.
Another item relating to the maintenance of the system is the terminal cleanout provision. A
cleanout facility should be provided at the ends of each branch of a pressure sewer system. Figure
1-22 shows a typical design of one of these facilities. It should be noted that watertightness is not
necessarily required for these valve boxes, but local conditions such as high ground water levels and
poor soil drainage should be weighed in making this determination.
Another design consideration is that of thrust blocking. As in water transmission practice, the
designer must consider the need for thrust blocking or anchoring any bends, plugs, or caps by using
standard techniques of calculation. The Grandview Lake system used these anchoring techniques on
plugs, caps, and bends exceeding 22.5° .14 Thrust blocks must not obstruct access to joints or other
fittings of the system.
New housing developments represent a typical example of the load factor problem. When the
population served at the time of construction is significantly less than the projected maximum, the
hydraulic design may be extremely difficult, especially for GP systems where critical velocities are
required. Obviously, if the ultimate number of contributors is twice the construction period num-
ber, velocities in the pressure main will also be related by a similar ratio. The entire aspect of hy-
draulic design is dealt with later, but some suggestions on the problem of varying load factors will
be discussed here. One of the reasons for recommending the branched layout shown in figure Ml is
that smaller branch mains can be used efficiently by a community developer. Although the practice
can be difficult, it is not uncommon for sections to be fully or at least greatly developed before new
sections are opened for house construction. This approach offers many logistical advantages to a
builder during the construction period.
33
-------
Meter box
Figure 1-22. Terminal cleanout.
Voell has discussed the use of dual lines, whereby one pressure sewer is sized for present needs
and a second line is constructed for later development.33 The pitfalls of this approach are manifold
in redundant hookups and ancillary needs, which can translate into excessive costs as pointed out
by Rees and Hendricks.13 However, the earlier discussion on the need for main line contingency
bypassing represents a possible advantage for this approach.
Probably the most widely accepted approach is the use of flushing to provide scouring velocity
for minimization of grease and solids accumulations. These flushing units should discharge a volume
of liquid that is at least equal to the total internal volume of the pressurized branch being flushed.
Although Bowles1 and Voell33 suggest using flushing devices that employ potable water supplied
with approved air breaks, this approach negates one of the advantages of pressure sewers, that is,
reduced sewage volumes. The Grandview Lake design used 1,000-gallon (3.8-m3) holding tanks
equipped with end suction and 0.75-hp (0.56-kW) centrifugal pumps to flush the pressure mains.14
These pumps were actuated by timers and pumping was stopped by low-level float switches. The
flushing liquid was septic tank effluent. This sytem, therefore, did not increase the quantity of
wastewater to be treated from the pressurized system. Also, the timers can be set to provide flush-
ing during minimum flow periods (between 12 a.m. and 7 a.m.).
If initial flows are significantly less than the design flows, the use of flushing units appears to
be the most advantageous design procedure, especially for GP systems. Considerations in their de-
sign are: the volume of main to be flushed, required scouring velocities, the number of household
sources required to provide the necessary volume, and proper pump selection to obtain the neces-
sary flows at the projected TDH.
The hydraulic design of the pressure main has been discussed by a number of authors.2'8'1 *•>
19,22,32,34,35,40 Tne original work for the ASCE project was done by Hobbs.40 His work
determined the relationship between sewage characteristics and carrying velocity for pressure
34
-------
sewers. He found that the minimum scouring and maximum redepositing velocities for 2- to 8-inch
(5.1- to 20.3-cm) plastic pipe are nearly identical and can be estimated by the equation,
Vs = V3/2
where
Vs = minimum scouring velocity, ft/s
d = inside diameter of pipe, inches
The critical material in the sewage tested was found to be sand. Although some evidence of a
sand size versus Vs relationship was evident, insufficient data were taken to define it. The sewage
tested had grease concentrations that ranged from 15 to 365 mg/1. Tests were conducted in their
entirety without lengthy no-flow periods.
Carcich et al.2 noted that the above criterion refers only to prevention of solids-settling. In the
Albany study, the accumulations (which amounted to as much as 40 percent of the pipe cross-
sectional area) were all located in the crown of the pipe and consisted primarily of grease. In ex-
plaining the problem, it was noted that during periods of inoperation (V = 0), grease accumulation
at the crown is inevitable. These periods allow the release of gases that float the solids that combine
with the grease at the crown to create a solidified mass of substantial strength and durability, which
could be highly resistant to dislodgement. The crown-oriented mass creates greater flow resistance
in the pipe.
The problem of grease accumulations affects the hydraulic design problem in two ways: The
friction coefficient for the pipe will be different from its nominal "clean water" value, and scouring
velocities are thereby required to minimize the effects of these accumulations. The magnitude of
the grease problem is obviously greater in systems using GP units than in those using STEP systems
because grit, grease, and SS are removed in the septic tank.
The Albany project was designed on the following assumptions:
• The maximum number of GP units operating simultaneously would be all 12 in the sys-
tem.
• The minimum number of units routinely operating would be four.
• Four units would then provide a flow greater than the minimum scouring velocity.8
The study determined that the system was hydraulically overdesigned, because the following
frequencies were actually obtained:
• Two simultaneous GP operations 20 times per day
• Three simultaneous operations at least once per day
• Four simultaneous operations about once every 14 days2-8
Two major GP system design decisions were modified on the basis of the Albany data. First, a
minimum velocity of 2 ft/s (0.61 m/s) is required in all pipe sizes normally used in GP pressure sys-
tems.2' 8 Second, on the basis of these data and other information, the E/One Corporation pro-
duced the design table shown in table I-4.22 There also appeared to be an inverse relationship be-
tween the number of users of a particular section of pipe and the amount of grease that accumu-
35
-------
Table 1-4.—Simultaneous operat t of grinder pump units
Grinder pump units
1
2-3
4.9
10-18
19-30
31-50
51-80
81-113
114-146
147-179
180-212
213-245
246-278
279-31 1
312-344
Maximum
operating
simulta-
neously
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
lated in each pipe size,2-8 reinforcing the theory that no-flow periods allow grease accumulations
to develop. However, grease accumulation in the pressure laterals did not follow this pattern.8
The Phoenixville GP system design was tested with a computer program that used the Hazen-
Williams formula along with a mathematical expression for the H-Q relationship of the GP units.11
This program checked all possible operating conditions to determine the headlosses and velocities
that could occur at various locations in the system. An increase of the discharge head of one GP,
from 81 feet (24.7 metres) at the start of the study to 123 feet (37.5 metres) at the end indicated
that some constriction had developed in the system.
The Grandview Lake design was based on the design engineer's peak water demand curve. A
value of 80 percent of the peak water demand was used to determine the maximum simultaneous
use of home units. A Hazen-Williams C factor of 130 was used to estimate pipe friction losses.3 As a
result of the Grandview Lake experience, the engineer is designing new STEP systems with a peak
flow of 70 percent of their peak water demand rate and a minimum velocity of 2 ft/s (0.61 m/s)
instead of the 1 ft/s (0.30 m/s) minimum used at Grandview Lake.19
A hydraulic design procedure for GP systems has been recommended by the E/One Corpor-
ation.22 After a preliminary layout of the branches and pressurization facilities, a tabular analysis
of the system is made. Table 1-5 represents this type of analysis for the GP system design shown in
figure 1-10. The maximum number of units operating simultaneously is estimated by using table 1-4,
and the maximum flows are assumed to be 11 gal/min (0.69 1/s) per GP unit. From these data, the
maximum velocity is obtained by using the pipe cross-sectional area. The velocity relates to specific
headlosses per unit length of pipe (available from the company based on a Hazen-Williams C factor
of 150). Flanigan and Cudnik35 present a strong case for the use of C = 140 to allow for grease and
other deposits on the inside of the pipe. The accumulated frictional headlosses at this maximum
flow condition are determined, starting at the discharge point. The E/One procedure then super-
imposes the static head difference between the highest elevation in the system between the pumps
aS. M. Rees, SIECO, Inc., personal communication.
36
-------
Table \-5.-Pressure sewer system PVC, SDR-21 pipe schedule arid branch analysis sheet
Branch no.
1
2
1+2
3
4
5
4+5
6
3+6
7
No. of
pumps
3
6
2
3
5
0
3
6
2
3
5
o
0
Cumu-
lative
total
3
9
11
3
8
19
19
3
9
11
3
8
19
19
38
38
Maxi-
mum
on
2
3
4
2
3
5
2
3
4
2
3
5
6
Maxi-
mum
flow,
gal/
mm
22
33
44
22
33
55
22
33
44
22
33
55
66
Size,
in
2
2
2
2
2
2
2
2
2
2
2
2
3
Maxi-
mum
veloc-
ity,
ft/s
2 03
3.05
4.07
2 03
3.05
5 08
2 03
3.05
4.07
2 03
3.05
5 08
2 81
Length,
ft
60
500
180
80
450
300
60
500
180
80
450
300
800
Fric-
tion
loss.
ft/100
ft
0 79
1.68
2.85
79
1.68
4 31
79
1.68
2.85
79
1.68
4 31
91
Fric-
tion
loss,
total
ft
0 48
8.40
5.12
79
7.58
12 90
48
8.40
5.12
63
7.58
12 90
7 28
Head-
loss,
total
ft
34 18
33.70
25.30
28 39
27.76
20 18
34 is
33.70
25.30
28 39
27.76
20 18
7 28
Max.
eleva-
tion
main,
ft
785
785
785
785
785
785
785
785
785
785
785
785
785
Pump
site
eleva-
tion.
ft
735
738
748
765
760
751
744
746
755
770
768
758
755
Eleva-
tion
differ-
ence.
ft
50
47
37
20
25
34
41
39
30
15
17
27
30
Maxi-
mum
total
head.
ft
84 18
80.70
62.30
48 39
52.76
54 18
75 18
72.70
55.30
43 39
44.76
47 18
37 28
CO
-------
in question and the discharge point. The sum of the dynamic and static heads should then be made
approximately equal to 81 feet (24.7 metres) by the E/One method. One possible difficulty is the
use of a Hazen-Williams C factor of 150. Although nonspecific on the subject, the E/One tables
suggest that pipe sized to accommodate a minimum velocity of about 1.8 ft/s (0.55 m/s) be used.
Flanigan and Cudnik recommend that velocities ranging from 2 to 5 ft/s (0.61 to 1.52 m/s) be
used and that flushing be provided. They also offer a series of tables of suggested design flows based
on number of units, occupants per dwelling unit, and level of affluence to occupants. As in the
E/One approach, the design flows are expected to occur once or twice per day. As noted earlier, a C
factor of 140 is recommended at this time, with a possible revision upward with more experience.35
Sanson19 has indicated the latest approach of SIECO, Inc., which relates to the company's
experience with peak water supply demands. Pressure mains are sized on the basis of a sewage flow
of 70 percent of the peak water demand. This assumption is based on a pump that pumps 10 gal/
min (0.63 1/s) at a design head of 80 feet (24.4 metres). A minimum velocity of 2 ft/s (0.61 m/s) is
also required. The use of a C factor of 130 or less has been indicated.3
Figure 1-23 is a plot of the foregoing design flows for various sizes of pressure sewer systems.
The figure includes the recommendations of Sanson19 and E/One22 and the maximum and mini-
mum recommendations of Flanigan and Cudnik.34 This last report recommended eight levels of
design, but only the two extremes are shown in the figure. These extremes represent daily flows of
400 and 175 gallons (0.61 and 1.53 m3) per household. Sanson's curve is based on 200 gal/d (9.76
m3/d) per connection, while the E/One curve is said to be based on peak flows obtained in the Al-
bany project and other existing pressure sewer systems.22 In terms of peak to average flow, the
Flanigan and Cudnik curves for smaller systems depend on the capacity versus TDH curves of the
pumping units chosen. For example, the peak flow predicted by wastewater diurnal flow equations
is usually less than the author's suggested design peak flow. The reasons relate to the possibility of
simultaneous operation of pumping units with much greater flow capacity. Because this assumption
is also the basis of the E/One curve, it may explain the similarity of the E/One and Flanigan and
Cudnik curves for systems of 20 or fewer homes. Sanson's curve,19 however, is based on the fact
that centrifugal STEP's will seek their own equilibrium condition at all times and will not affect
sewer design beyond the peak flow assumption as a percent of peak water demand. This peak flow
assumption is based on the H-Q curve of the pump at the assumed maximum pressure of the system.
For the curve plotted in figure 1-23, a pump capable of pumping 10 gal/min (0.63 1/s) at 81 feet
(24.7 metres) of TDH was assumed along with a maximum peak household water demand of 15
gal/min (9.95 1/s) to obtain a (10/15) (100) s 70 percent factor to be applied to the peak water
demand curve. Bowne3 2 has chosen the Flanigan and Cudnik curve, which is based on 215 gal/d
(0.81 m3/d), for his design. The generally lower peak flows reflected by Sanson's curve take into
consideration the flow-smoothing capabilities of centrifugal pumps, that is, the ability of these
pumps to adjust to prevailing hydraulic conditions.
Once design flows have been chosen, the pipe sizes can be determined by using the Hazen-
Williams formula. At this time, with minimum experience in the use of these systems, a maximum C
factor of 140 seems prudent to provide some safety factor for flows in excess of the design flow and
to allow for grease or other accumulations on the inner walls of the pipe. Other design checks
should include the load factor and possible need for flushing. Another factor to consider is the
headloss because of valves and fittings in the system. The Albany report estimated that the losses
caused by valves and fittings were 32 percent of the total friction loss.8 This reinforces the need to
use a more conservative C factor and also implies that designers should incorporate additional safety
factors when considering critical conditions. Hydraulic friction loss data for PVC valves and fittings
are becoming generally available for use as part of the design calculations.
Bowne offers a simplified design procedure for use in STEP systems with centrifugal pump-
ing.32 After determining the number of homes to be served by the system, a peak flow is estab-
aS. M. Rees, SIECO, Inc., personal communication.
38
-------
280
240
200
CO
CO
q
d
X
c
160
120
co
LU
Q
80
40
10
100
1,000
NUMBER OF SERVICE CONNECTIONS
Figure I-23. Suggested design flow.
lished based on one of the above sources or a recognized equivalent. A profile of the proposed sys-
tem is then prepared, and hydraulic gradelines corresponding to various pipe sizes are plotted, as
shown in figure 1-24.
Because a reasonable approximation of pump characteristics is already known based on eco-
nomics, pressure limitations, and other factors, any pipe size that indicates an excessive TDH re-
quirement (difference in elevation between the hydraulic gradeline and ground or sewer profile) is
39
-------
200
180
r 160
<
>
ui
140
120
100
0
With pressure
sustaining valve
Without pressure
sustaining valve
4 6 8 10 12 14 16
PIPELINE DISTANCE FROM DISCHARGE POINT, thousands of feet
Figure I-24. Pipe sizing procedure.
18
20
sequentially discarded until a satisfactory pipe size is found. In figure 1-24 the 3-inch (7.62-cm) pipe
requires greater pumping capacity than is feasible and the 6-inch (15.24-cm) size is excessive from
the standpoint of cost and low velocities. Close examination of the system profile for this example
indicates the need for a pressure-sustaining valve and possibly one or more air-release valves to pre-
vent problems caused by air pockets. With the introduction of the pressure-sustaining valve, a new
dynamic hydraulic gradeline results and is plotted, as shown in figure 1-24. Individual pump charac-
teristics can then be tested for sufficiency by the elevation difference between the proposed pump
elevation and the elevation of the dynamic hydraulic gradeline at the mainline station where the
pump lateral intersects. To make this test, the modified H-Q curve of the proposed pump, which
includes service line losses for various sizes and lengths of discharge pipe, is plotted, as shown in
figure 1-25. The previously noted head requirement can then be located on the H-Q diagram at the
design flow to determine the adequacy of the pump and the most suitable pipe size.
The type of plastic pipe chosen can significantly affect the design and economics of the pres-
sure sewer system. PVC pipe has been used almost exclusively. The pressure sewer mains at Albany
were PVC Type I, Schedule 40, with PVC, DWV fittings, and the joints were solvent welded.8 At
Phoenixville, PVC Type I, SDR-26 piping was used with PVC Schedule 40 fittings.11 At Grandview
Lake PVC, SDR-26 pipe (solvent welded) and PVC fittings were used. The Miami systems use PVC,
SDR-26 pipes with slip-ring joints, as do the Priest Lake installations.6>f PVC, SDR-26 pipe with
solvent welded joints is used in the Weatherby Lake pressure sewer system. Flanigan and Cudnik
recommend the use of PVC, SDR-26 piping in all systems whose pumping heads do not exceed 90
eH. Schmidt, General Development Utilities Company, personal communication.
f C. W. Rose, Farmer's Home Administration, personal communication.
40
-------
50
40
30
£
Q
ill
I
^20
10
01
Standard
pump curve
Pump curve with
100 feet of service
line of size shown
11/4 inches
I
20
40
CAPACITY, gal/min
60
80
Figure I-25. Pump and service line testing procedure.
feet (30.5 metres).35 E/One recommends, in order, SDR-21, Schedule 40 pipes and SDR-26.22 The
pressure rating of SDR-26 pipe is 160 psi (1,110 kPa); SDR-21 is rated at 200 psi (1,380 kPa), while
Schedule 40 pipe may vary. Schedule 40 pipes of 2 and 3 inches (5.1 and 7.6 cm) are rated at 277 and
263 psi (1,920 and 1,820 kPa), respectively.34 All pressure ratings are at a temperature of 73°F
(22.8°C) and are generally reduced at higher temperatures to the extent that PVC is not recom-
41
-------
mended above 150° F (65.6° C). The higher pressure-rated pipe recommended by E/One may be
related to its GP's ability to operate at very high pressures. The safety factor between pressure rat-
ings of SDR-26 pipe and system design pressures in almost all cases exceeds 4. Because the other
recommended PVC pipes all have greater pressure ratings, their safety factors are larger. Because a
safety factor of 4 is common for water supply systems where water hammer conditions are more
likely, the safety factor for all PVC pipes discussed appears adequate.
As a matter of interest, one system has been designed with polyethylene pipe in northern
Michigan. This system is a combination pressure-gravity design that uses GP's. The details of how
the polyethylene pipe was adapted to the pressure system are not available at this time, but 40-foot
(12.2-metre) lengths of polyethylene pipe will be fused together through the use of heat and pres-
sure before being laid in the trench.17
Some concern has been expressed about the shallow depth of pressure mains and their in-
creased susceptibility to damage by excavating equipment. Bowne has suggested that markers be set
along the pipe route warning of its presence.32 He further suggests burial of a copper wire with the
pipe for easy location by a cable finder and the institution of a permit system for excavation to
minimize such accidents. Accurate "as built" drawings are a necessity in all cases. Leckman34 has
raised the related question of differentiation between water and sewer lines made of PVC. He also
suggests markings where applicable and other methods such as standardized relative locations of
sewer and water lines. The Grandview Lake system uses brown-colored PVC pipe to simplify differ-
entiation, while the Miami systems use green PVC pipe in one location and red striped pipe in an-
other.e> 14 This type of color coding is required by the Pennsylvania Department of Environmental
Resources.4 *•
CONSTRUCTION CONSIDERATIONS
The small pressure sewer system at Albany was installed by a plumbing contractor. A tempor-
ary variance was obtained from the Plumbing Code of the City of Albany to use PVC, DWV pipe.
The system was pressure-tested at 80 psig (553 kPa) before backfilling took place. A number of
leaks were found and repaired at that time. These leaks were attributed to the plumbers' lack of
familiarity with the use of PVC. After the leaks were repaired, the pipe was covered with 18 inches
(45.7 cm) of sand.8
The Phoenixville system was installed by a general contractor. The trenching machine used
allowed construction in a trench of less than 4 inches (10.2 cm) in width at an average pipe depth of
2.5 feet (0.76 metre). In areas where rock removal was required, a backhoe was used. Where the
pipe crossed under public or private roads, it was encased in a 4-inch (10.2-cm) asbestos cement
pipe to protect it from traffic or vehicle loads. Upon completion the pressure system was leak
checked at 100 psig (692 kPa) with potable water before backfilling.
The system at Grandview Lake also was installed by a general contractor. Where rock was en-
countered, 5 inches (12.7 cm) of sand bedding were required.3 Several 8-inch (20.3-cm) layers of
granular fill with tamping were required where pipes passed under roads. The normal pipe depth was
3 feet (0.92 metre) below the ground surface. Joints were solvent welded and the solvent was
allowed to set up before "snaking" the pipe into the trench. After the pipe was laid in the trench it
was pressure tested at 100 psig (692 kPa) for at least 30 minutes after all excess air had been ex-
pelled. After being in place for about 2 weeks, the pipe was leak tested at 150 psig (1,040 kPa). The
pipe was backfilled only when the temperature was below 65° F (18.3° C).14
aS. M. Rees, SIECO, Inc., personal communication
eH. Schmidt, General Development Utilities Company, personal communication.
42
-------
Bowne has emphasized the need for care during construction to avoid scoring PVC pipe, which
could result in strength reduction.32 He also noted that bedding and backfill requirements are less
stringent when compared to conventional pipe because of the reduced brittleness of PVC at moder-
ate temperatures. His preference for pipe protection and ease of installation under roadways is to
encase the plastic pipe in a steel pipe. His specifications are illustrated in figure I-26.1
The British Standards Institution (BSI) provides an excellent set of guidelines for plastic pipe
application.42'43 In relation to construction or pipe laying procedures, the BSI specifies:
• Trench width should be equal to or greater than the sum of the outside diameter of the
pipe plus 12 inches (30 cm).
• Depth of bedding below the pipe barrel should be no less than 4 inches (10 cm).
• Bedding material should not exceed 0.5 inch (1.0 cm) in size.
• Sidefilling around the pipe should be done in 3-inch (7.5-cm) layers, compacting each
layer by hand.
• Pipes should be partially backfilled (leaving joints exposed for inspection) before pressure
testing.
• Pressure testing at 1.5 times the maximum working pressure at the point of maximum
stress should be done for 1 hour.
30 inches
minimum
4 inches'
< Toning wire
•(Pipe zone backfill
Approved sand, maximum
Maximum particle size < 0.5 inch
Figure 1-26. Pipe trench and backfill.
. C. Bowne, Douglas County (Oregon) Road Department, personal communication.
43
-------
• A successful pressure test must not lose more water than 0.24 gal/1,000 ft of length per
in of nominal diameter per day per 43.4 psi of test pressure (3 I/km per 2.5 cm/d per 300
kPa).
• No construction should be undertaken at temperatures below 14° F (-10° C), because of
reduced impact strength of PVC pipe in cold weather.
• Scoring of pipes by dragging them over rough ground or other careless handling must be
avoided.
For any pressure sewer installation, one of the desired advantages, which may be maximized, is
the ease of construction. Where rock is absent and progress is unimpaired by natural or manmade
barriers, trenches can be dug in a very short time by mobile trenching devices. In difficult areas,
backhoes may be required, although the Bend system used a special "rock-saw" trenching device for
main construction.11 In any case, the rate and ease of construction, compared to conventional grav-
ity systems, are obvious. For example, at Weatherby Lake the contractor laid 900 feet of 2- and
3-inch (5.1- and 7.6-cm) main on the first day of work.16
In any established area where pressure sewers are installed, care must be taken to avoid damag-
ing existing water, electrical, and gas service lines. At Grandview Lake, where the homeowners' rec-
oDections were the only guide to the location of service lines, an average of 1.3 existing service lines
was cut during pressure sewer service line installation at each home. The maximum for one home
was seven.14 Ideally, "as built" drawings should be obtained, but these often do not exist in rural
areas.
OPERATION AND MAINTENANCE
Obviously, any system that uses numerous pressurization facilities and other more sophisti-
cated mechanical equipment will require a significant amount of operation and maintenance. The
institutional concepts alluded to earlier for the pressurization facilities must also take into account
the needs of the entire pressure sewer. These have generally been described in the design considera-
tions for the pressure main. For the most part, operation and maintenance can be divided between
on-lot needs (pressurization units and service lines) and main line needs.
The Albany project started with 12 E/One prototype units, most of which were replaced dur-
ing the study by modified units.8 The number of service calls then fell off sharply. Most of the 44
malfunctions reported were caused by faulty pressure sensors. Because the modified units had im-
proved pressure-sensing tubes, only 5 of the 44 malfunctions noted involved modified units. The mal-
functions took the form of excessive noise (because of their in-house location), continuous motor
operation and nonfunctioning units resulting in overflows. An operation ratio based on any greater-
than-15-minute malfunction was calculated. The operation ratio consisted of the number of days when
no malfunction occurred over the total service days. Because it is a measurement of reliability, the
ratio provides a meaningful account of expected service requirements. The operation ratio over the
entire project period varied from 0.90 to 0.99 at the 12 homes. A separate operation ratio was cal-
culated to be 0.995 for the modified units. A number somewhat less than this would probably be
more accurate because all the startup difficulties that would normally be expected in the system
were not included in this latter calculation. The corresponding downtime, defined as the hours out
of service over the total hours of possible service, was 2.69 percent for the prototype units and 0.27
percent for the modified GP units. The difference was primarily because of the improved pressure
sensors of the modified units.
hL. R. Clark and J. E. Eblen, C & G Engineering, personal communication.
44
-------
Power consumption at Albany was measured for 2 of the 12 units.8 Monthly power consump-
tion averaged 10.2 and 5.3 kWh for the units, which also averaged 28.5 and 16.0 min/d of opera-
tion. Because the wastewater flows were not measured at each home, the operational time could be
converted to an approximate flow per home by multiplying the operating time by an average GP
pumping rate. If 15 gal/min (0.94 1/s) is assumed as an average, the operating times represent about
427 and 240 gallons (1.62 and 0.91 m3) per day, respectively, and the power required per unit vol-
ume is 0.795 and 0.736 Wh/gal (0.210 and 0.195 kWh/m3), respectively. A conservative average
would then appear to be about 0.8 Wh/gal (0.212 kWh/m3) for the GP units and conditions of this
study.
The Phoenixville system experienced startup difficulties, primarily caused by oversights during
installation.11 During the course of the study, problems with faulty GP discharge line materials,
causing electrical short circuiting, a faulty circuit breaker, and faulty grinder assemblies were re-
corded. If the operating ratio analysis used in the Albany report8 had been applied to the Phoenix-
ville data, the operating ratios for units 1 through 5 would have been 1.0,1.0, 0.975, 0.97, and
0.97, respectively. The average operating time per day for GP 3 was 33 minutes at an average dis-
charge pressure of 11 psig (76 kPa), while drawing an average of 12.5 A. For unit 4, the valves were
59 minutes at 44 psig (304 kPa) and 14.0 A. Therefore, by converting pressure to flow by the H-Q
curve and calculating kilowatt-hours from the voltage, amperes and time of operation, the daily
flows and power requirements can be calculated to be 461 and 590 gallons (1.74 and 2.23 m3) and
0.37 and 1.51 kWh for units 3 and 4, respectively. These convert to 0.803 and 2.56 Wh/gal (0.214
and 0.677 kWh/m3), respectively. These unit costs are in close agreement with the Albany data.
The Grandview Lake operation and maintenance information is extensive.14 Of the three
major commercial GP units used, (E/One, Hydromatic, and Tulsa), the E/One unit required the
lowest number of service calls per number of units installed, while the Hydromatic had the next
lowest, and the Tulsa had the highest. The data are, however, only approximate on the E/One and
Hydromatic units because service calls were often made directly to the private maintenance services
for each of these units and the number of units was continually changing as the number of connec-
tions grew. The calls reported to the engineer were classified according to the nature of the mal-
function, as shown in table 1-6.14 The E/One unit was found to fail because foreign particles scored
the metal rotor and excessive delay in thermal overload activation.14 The overload was caused by
excessive air in the pressure sewer line. Maintenance difficulties were compounded by the weight of
the unit, which generally required more than one person to install or remove it. The Hydromatic
unit was often found to leak at its quick-disconnect fittings because of excessive pressures in the
system. The ball check valves also were found to require substantial maintenance, as did the float
switch that controlled operation.14 Some float switch problems occurred because of grease accumu-
lation brought on by a lack of sufficient swirling action in the tank. The lack of an antisiphon de-
vice also caused some problems.14 Although lighter than the E/One, this GP was also difficult to
install and remove. A number of design problems ,/ere cited that resulted in the increased rate of
service calls for these units.14 The report recommended redesign and upgrading of the units used in
the project. Both the E/One and Hydromatic GP units have been modified since the study.
One of the difficulties encountered in analyzing the Grandview Lake reportb is that the dy-
namic nature of the system precluded determination of precise operation ratios, downtimes, and so
forth. It is fair to say that the experimental nature of this large installation produced better practi-
cal engineering data than the Phoenixville or Albany projects, but less mechanical-electrical per-
formance information. The number of commercial unit service calls for the size of installation in-
volved (shown in table 1-6), excluding an unsuccessful noncommercial GP unit, appears to be quite
low, despite any data shortcomings. The E/One Corporation indicates that 19 service calls were
made by its representative during 11,800 unit days of possible operation in the Grandview system.b
This would yield an operation ratio of 0.9984, assuming the numbers are accurate.
' Jl. P. Farrell, Environment/One Corporation, personal communication.
45
-------
siiimary of maintenance
[2-year period]
Cause
Pump failure
Grinder failure
Piping failure (within tank)
Electrical failure (excluding controls)
Control failure
Piping failure (outside tank)
Infiltration/inflow of water
Collection system malfunction
Improper installation
Miscellaneous
Total
Maximum number of units
Locally
manu-
factured
unit
52
25
41
66
23
11
56
7
9
81
371
27
E/One
unit3
4
0
4
3
0
2
1
1
2
0
17
15
Hydro-
matic
unit3
8
1
7
2
10
8
0
2
2
12
52
28
Tulsa
unit
2
2
1
0
2
0
0
4
0
5
16
2
Maintenance of the E/One units was done primarily by the manufacturer's representative. The Hydromatic units were serviced
by the manufacturer's field personnel as malfunctions were reported to the factory. Therefore, the figures listed above were based on
the field notes taken by the engineer's maintenance crew and may not include all of the service calls by others.
The Horseshoe Bay (Lake Lyndon Baines Johnson) GP installation has also experienced some
operation and maintenance problems.15 These have been identified as excessive wear and failure of
the GP stators because of grit particles in the sewage and improper ventilation of pump motors.
Similar grit problems were noted during the early stages of a similar GP system at Point Venture
(Lake Travis).15
The Miami systems represent the longest history of operation and maintenance of all the pres-
sure sewer systems. Thus far, the experience with these pressure sewer systems indicates that the
operation and maintenance costs are the same as for gravity systems. Currently, an annual preven-
tive maintenance inspection is employed. This inspection consists of the following:
• The pump is removed from the holding tank, inspected for corrosion and suction plate
condition, and cleaned (if necessary).
• The check valve and gate valve are inspected for proper functioning.
• The pump is returned to its operating position and tested.
A two-man crew normally requires 30 minutes to complete such a preventive maintenance proce-
dure. This preventive maintenance program, therefore, amounts to 1 man-hour per year. In addi-
tion, replacement of diaphragm switches and refurbishing of brass disconnect fittings are included in
this program every other year. On one of these systems, which contains 26 STEP units, one emer-
gency repair has been reported in 3 years of operation.6
eH. Schmidt, General Development Utilities Company, personal communication.
46
-------
Routine maintenance of the Priest Lake systems is reported to have resulted in service calls to
about 8 percent of the STEP units (approximately 500 total units in systems) in the first year and
only 2 percent in the second year.32 During the third year the service calls averaged about five per
week, with an average service time of 30 minutes per call. One man services both Priest Lake sys-
tems, routinely inspects and pumps out the system septic tanks, and operates the treatment facili-
ties. Experience has indicated that this individual can remove 30 pumps per day, replacing the im-
pellers and returning them to service.32
The combined experiences of several STEP systems and some raw sewage pumping systems
indicate that a 10-year life can be expected for the submersible sump pumps that have been used
thus far before rebuilding is required.6'32 One of the major causes for servicing of STEP systems is
the buildup of iron sulfide on the pump impeller after lengthy periods of inactivity, such as in a
vacation home.'
Operation and maintenance of pressure mains includes periodic cleaning, repairing of leaks,
and major replacement of sections. Because of the short length of the Albany project's main, no
maintenance was reported.8 At Phoenixville, a 2-day shutdown was incurred when an air-release
valve was damaged by heavy equipment during routine snow removal.11 At Grandview Lake, a
number of service calls were required because the tapping tool, used to connect individual home
services to the pressure main, was improperly used, resulting in. leaks.14 Additional leaks and breaks
were caused by heavy equipment, earthslides, and improper installation. Line cleanouts were neces-
sary during the early stages of operation because of low flows and concomitant solids buildups. On
four occasions, supplementary flushing of the lines was accomplished with lake water. Maintenance
of the air-release valves was necessary and was somewhat more difficult for the automatic valves
because no pressurized water source was available at the valve locations. Some odors were reported
when these valves were actuated. Earth shifts at one location in the Grandview Lake system caused
repeated breaks in the pressure main. The eventual solution to this maintenance problem was effec-
ted by replacing the PVC pipe with a looped section of flexible pipe.14 No maintenance of the
Miami or Priest Lake systems' mains has been necessary.e>32
In STEP systems, periodic septic tank pumping is considered an operation and maintenance
requirement. The cost of pumping varies with geographic location, but regular surveillance of solids
buildup in these tanks is required to prevent significant reductions in their grease and solids removal
capabilities. Pumping is not normally required at intervals of less than 3 years6 and may greatly ex-
ceed this length of service. For instance, the Miami STEP systems have not required any pumping of
septic tanks in 5.5 years of service.6 Because the scum and sludge accumulation is highly variable,
annual inspection of septic tanks is initially recommended until sufficient data are generated to
determine a proper interval for pumping individual tanks.
VARIATIONS
It is obvious that any technology as new as the one described in this seminar is not limited to
the design and construction methods discussed herein. One example that is conspicuous by its ab-
sence is the use of a single collection tank and pressurization device for more than one dwelling
unit. Rose3>f has long been an advocate of this approach and the Bend system employs one three-
home and three two-home installations. The potential savings in cost are obvious because the num-
ber of pumps or GP units needed is reduced. Rees has expressed concern about the possible prob-
lems resulting from allocation of operation and maintenance costs among contributors that could
make such an approach difficult to implement.3 Another variation that could occur relates to the
aS. M. Rees, SIECO, Inc., personal communication.
eH. Schmidt, General Development Utilities Company, personal communication.
fC. W. Rose, Farmer's Home Administration, personal communication.
'W. C. Bowne, Douglas County (Oregon) Road Department, personal communication.
47
-------
inclusion of multifamily or recreational/commercial contributors. Although the Phoenixville system
appeared to handle a small apartment building without difficulty,11 some changes in design, such as
holding tank size and pipe sizing changes based on specific flow patterns, will probably be necessary.
CODES
Information on State and local codes that restrict or govern the use of pressure sewer technol-
ogy must be obtained from the responsible agencies. A partial list of States that undoubtedly have
permanent or temporary regulations for pressure sewer installations appeared in the section on
previous experience. States not listed should be contacted before planning begins.
Leckman34 contacted 11 Illinois local government units to determine whether 1.25-inch
(3.18-cm) polyethylene pipe would be allowed for transmitting sewage. All 11 local authorities indi-
cated that it would be in violation of their plumbing codes. When asked about the possibility of
granting conditional or special permission to use a GP system, 9 out of 11 authorities said they
would refuse permission. Leckman also analyzed the 1971 Ten State Standards44 as they applied to
pressure sewer systems. Serious conflicts were found in chapters 20 and 30 concerning design con-
siderations, such as per capita flow, minimum sewer size and slope, sewer alignment, pump open-
ings, wet well requirements, emergency operation, and minimum sewer velocity.
WASTEWATER CHARACTERIZATION AND TREATMENT
As noted earlier, wastewater from an individual home is more concentrated than normal muni-
cipal wastewater. Some data from pressure sewer systems have been obtained. At Albany, a thor-
ough sampling and analysis program produced the results summarized in table 1-7. These data are
reasonably consistent with those in table 1-1.
The Phoenixville and Grandview Lake systems were sampled (generally grab samples) only a
few times; in contrast, more than 50 composite samples were analyzed at Albany.11'14 The con-
centrations of pollutants analyzed were generally diluted by table 1-1 standards. The Phoenixville
data confirmed the absence of dissolved oxygen, as would be expected in a pressure sewer.11 Be-
cause the Grandview Lake system used both GP and STEP units, the combined wastewater does not
necessarily follow the pattern indicated in Table 1-1. Twenty-four-hour composites yielded SS, 5-
day BOD (BOD5) and chemical oxygen demand (COD) concentrations ranging from 80 to 265
mg/1,100 to 310 mg/1, and 230 to 462 mg/1, respectively.14
Table 1-7.—Albany wastewater characterization
[mg/1]
Parameter
BODS
COD
TSS
TKN
Total phosphorus
Concentration
Mean
330
855
310
80
15.9
81
Range
216-504
570-1,450
138-468
41-144
7.2-49.3
31-140
7.1-8.7
48
-------
The GP-pressure sewer effluent at the Horseshoe Bay project is treated by advanced treatment
methods. The treatment sequence involves an activated sludge system followed by chemical precipi-
tation and filtration. The final effluent is then used to irrigate a nearby golf course. Occasional odor
problems in the treatment plant lift station have been controlled by the use of odor-masking com-
pounds.15
The amenability of pressure sewer wastewater to treatment is of primary concern. Hobbs in-
vestigated the effects on sewage solids, both from grinding and comminution.40 He found no appar-
ent difference in required solids transport velocities, but noted that the grinder did appear to yield
finer solids than the comminutor. If this condition were significant, the possibility of reduced sedi-
mentation efficiency in the primary clarifier could be significant in the design of treatment facili-
ties.
At Albany, hydrogen sulfide odors were detected at the discharge of the pressure main, and
daily grab samples analyzed for sulfide showed concentrations of up to 2.5 mg/1.8 Some method of
freshening the sewage when it reached the treatment plant was suggested. Also, settleability tests
were made on the sewage from the pressure sewer and compared to some local residential sewage
collected in gravity sewers. Although the details of the comparison were not presented, the data
show that SS removal at equivalent overflow rates is somewhat less for the pressure sewer waste-
water, when expressed as a percent. Because of its high initial strength, the primary effluent result-
ing from the GP pressure sewer wastewater was significantly stronger in terms of SS and organic
matter, representing a higher loading to subsequent biological treatment processes. The conse-
quences are increased sludge production in the biological reactor and a need to maintain a higher
mixed liquor SS concentration for equivalent treatment. These considerations are not likely to off-
set the advantage of treating lower flow volumes. Because the collection system represents from 75
to 90 percent of the total cost of conventional wastewater management in smaller communities, a
50-percent savings in the collection system cost could be offset only by doubling the treatment
cost.
The special characteristics of wastewater from pressure sewer systems require that the de-
signers of treatment facilities use their best engineering judgment in matching the type of treatment
and specific modifications with the application, at least until further data are obtained. If, however,
the pressure sewer terminates by discharge to a gravity interceptor, the effects of these special char-
acteristics on the ultimate treatment facility will be a function of the relative contributions of the
pressure and gravity systems. If the pressure sewer contribution is locally or totally significant, sul-
fide control methods may be needed at the pressure main connection to the gravity interceptor. If
this is the case, the use of control methods described by Pomeroy45 are applicable.
In the case of a fully pressurized system, two effects of the pressure sewer wastewater on nor-
mal treatment systems must be considered—the tremendous variation in flow and the anaerobic
condition. Because a completely pressurized sewer will normally be located in a relatively small
community, the treatment system employed should be consistent with normal small-flow methods.
Likely choices, therefore, would include trickling filters, oxidation ditches, activated sludge package
plants, and lagoons, with possible land application of effluents during all or part of the year. Treat-
ment systems can be divided into lagoons and conventional biological processes.
The Grandview Lake pressure sewer system used lagoons with land disposal of the effluent.14
The Grandview Lake system had some difficulties from a physical standpoint (levee failure), but
runoff from the effluent-irrigated and nonirrigated hay fields was not significantly different in qual-
ity. The Priest Lake design incorporates two-stage lagoons with supplemental aeration followed by
land spreading. Primarily because of a high evaporation rate, however, no effluent has yet been pro-
duced for land spreading. Bowne has also suggested facultative lagoons followed by slow sand filtra-
tion as a means of obtaining a high quality effluent.1
'W. C. Bowne, Douglas County (Oregon) Road Department, personal communication.
49
-------
Facultative lagoons, with or wilhout mechanical surface aeration, represent a convenient
method of handling both the noxious gases potentially emitted from pressure sewer wastewater and
the wide flow variations. Inlets should be located near the bottom of the lagoon to allow proper
dispersion of the influent into the anaerobic bottom area. Designers must consider the relative
strengths of GP wastewater and septic tank effluent compared to normal domestic loading rates. In
the case of GP-pressure systems, any sludge allowances normally made in depth dimensioning might
be increased to allow for higher influent solids and sludge accumulation rates. In the case of STEP
systems, these allowances may be reduced.
When conventional biological systems are to be used, the problems of extreme flow variation
and hydrogen sulfide odor potential must be accommodated in the design. Introducing anaerobic
wastewater directly into the aeration compartment of a system of the activated sludge type appears
attractive, but high sulfide concentrations in wastewater can encourage the growth of filamentous
organisms in such a system.49 However, the STEP system near Port Charlotte is using this approach
with no ill effects. No odors have been reported around the extended aeration unit, and effluent
BOD5 and SS concentrations have averaged 6.9 and 14.3 mg/1 over 3 years of operation.6
Other methods of sulfide control include U-tube aeration, chlorination, and ozonation.45 The
latter two methods are expensive and somewhat inconsistent with smaller treatment systems, while
the first results in substantial headlosses. Any oxygen addition method would be most suitable at
the discharge end of a pressure sewer because of the air-binding hydraulic problems discussed ear-
lier.
Of all the aspects of pressure sewer technology, treatment problems are possibly the least
studied, primarily because very few systems have been totally pressurized. Most have constituted a
small portion of an overall sewer system and have not demonstrated major effects on treatment.
However, the few totally pressurized systems from which information is available have reported no
major difficulties in treatment.6-14'*
One additional point should be stressed. There may be a major difference in treatment require-
ments between STEP systems and GP systems. In the former case, relatively weak wastewater in
terms of BOD5 and SS must be treated, while additional maintenance in the form of septic tank
pumping is required. In the latter case, very concentrated wastewater in terms of BOD5 and SS
must be treated. The trade-offs must be weighed by the designer. Further experience with both
systems will expedite design selections.
COSTS
Capital Costs
No new technology is valuable unless its total costs are competitive with those of existing tech-
nology in a significant number of situations in engineering practice. Situations where these systems
have been found economically superior to conventional sewerage include hilly terrain, outcropping
rocky areas, and low areas with unfavorable on-site disposal capability. Other favorable conditions
also exist, such as high ground water regions and low population density communities. For practical
purposes, these areas generally fall into the following categories:
• Low density areas unsuited for on-site disposal
eH. Schmidt, General Development Utilities Company, personal communication.
'W. C. Bowne, Douglas County (Oregon) Road Department, personal communication.
50
-------
• Geological conditions unsuited for normal excavation
• Undulating or hilly terrain
The practicality of pressure sewers is not limited to any of these categories nor are pressure
sewers the exclusive answer in such categories.
Some information is available on the relative economics of pressure sewers as compared to
existing technology. Often these comparisons are "broad brush" in nature and do not represent
itemized cost comparisons. All information on costs is instructive, however, when the high cost of
sewerage is considered.
Numerous citations in the literature document the excessive cost of installing conventional
sewers in low density areas. Voell has estimated that it would cost $2.6 million for conventional
sewers at the Central Chautauqua Lake Sewer District of New York, as compared with an estimate for
a GP-pressure sewer system of $1.2 million. Grandview Lake sewerage estimates of $3,000 per lot
and $10,000 per existing home were reduced to less than $2,000 per home for the pressure sewer
system.32 An estimate of $100,000 for a 10-house sewer system in Saratoga was rejected in favor of
a cost of $20,000 for a pressure sewer-GP system. Conventional sewerage estimates for two com-
munities at Priest Lake were about $12 million with treatment.f A STEP-pressure sewer system was
subsequently built for less than $1 million. Bowne20 has presented a present worth comparison of a
conventional gravity system with a STEP-pressure sewer system using average costs in his region of
Oregon. On a total cost basis for this rural-suburban area, Bowne has estimated present worth costs
of $4.7 million for conventional sewers and $2.4 million for a STEP-pressure system. A list of seven
unidentified municipalities receiving GP-pressure sewer and conventional sewer estimates has been
provided by E/One and is shown in table 1-8. The two locations where the highest percent of savings
are noted (3 and 5) involve low-lying areas below existing sewer grades.
To determine the limits of pressure sewer application, it may be assumed that low density
housing areas, where soil conditions are favorable, will be better served by on-site wastewater treat-
ment and disposal systems. This assumption leads to the question, At what level of housing density
Table \-8.-Costcomparisons
Location
1
2
3
4
5
6 .
7
Connec-
tions
285
30
9
309
10
320
100
Engineer-
ing News
Record
index
1700
1895
2014
1753
1753
1895
2098
Grinder
pump-
pressure
sewer,
dol per
connec-
tion
1,930
2,800
2 222
3240
1,653
2 709
1,360
Conven-
tional
sewer,
dol
per
connec-
tion
4,570
4,667
10000
6,176
10,000
4088
2,350
Savings
for
grinder
pump-
pressure
sewer.
percent
48
40
78
49
83
34
42
NOTE: All estimates by consulting engineering firm except 2 and 5, which are bid prices for grinder-pump/pressure sewer
systems. Conversion of estimates to current costs requires use of present Engineering News Record index.
fC. W. Rose, Farmer's Home Administration, personal communication.
51
-------
would conventional sewer technology, rather than pressure sewers, better serve the population?
Although computation of this density would be a most valuable piece of information from a theo-
retical point of view, it would be highly subjective and would vary with each assumption made or
existing physical condition for any potential application. It is probable that the experience gained in
each new application in the coming years will permit reasonably reliable definitions of the limits of
applicability of pressure sewers.
The type of data required includes capital costs in various regions of the country, various cli-
mate conditions, and various soil and other geological conditions. For example, the cost of pumps,
pipe, valves, labor, and so forth, varies by region. The requirement that pipes be buried below the
frostline makes pressure sewers in Florida less expensive than in Minnesota, if all other conditions
are the same. In rocky areas, a cost comparison between pressure sewers and conventional sewers
will be more dramatic than in areas where more favorable soil conditions exist. Although some in-
formation has been generated on capital costs, a great deal more is necessary to allow reasonable
estimation of the limits of pressure sewer applicability.
Data are also needed on operation and maintenance requirements in terms of manpower, skills,
and costs. This aspect of the total picture will remain the least-definable of the entire equation until
information on these requirements becomes available. Only well-documented experience from a
number of installations over a significant period of time will permit solid estimates of these factors.
To initiate the task outlined above, a series of data points can be found in the literature, but a
considerable variation in physical constraints exists for each. Also, some bid prices may reflect ad-
justable profit margins; for example, if a large profit was figured for one category of work, a differ-
ent category bid may reflect smaller profits to complement the other. Also, it should be noted that
all cost estimates, bid prices, and actual costs can vary significantly on any job. All costs are related
to the month and year estimated or incurred and to geographic location.
In determining capital costs, a number of subelements must be considered, such as engineering,
valve boxes, fittings, cleanouts, flushing arrangements, testing. Some of these items may be difficult
to estimate; therefore, a few basic facts will be presented with expanded information based on tradi-
tional practice. One primary cost category is the pressure main cost per lineal foot. At Phoenix-
ville11 the cost of PVC pipe, excavation, installation, lateral tie-ins, and restoration for 2,800 feet
(854 metres) of main was $2.00 per foot ($6.56 per metre). Additional costs of rock removal, res-
toration of streets and driveways, load protection, an automatic air-release valve ($350), and a ter-
minal cleanout ($350) resulted in a total pressure sewer cost of $2.82 per foot ($9.25 per metre).
Pressurization facilities and service connections amounted to $2,050, including $900 per GP. In this
total, service line costs of $2.50 per foot ($8.20 per metre), circuit breaker costs of $60 each, and
power cable connection costs of $3.00 per foot ($9.84 per metre) were incurred. The overall (Janu-
ary 1971) capital cost for the pressure system was $19,020. Dividing this total among the users is
difficult because of the presence of apartment units. The costs, however, can be allocated as $595
per person, $1,270 per dwelling unit, $2,720 per structure, or $3,810 per GP. The engineering fee
for design and supervision of construction was $2,100 and the legal charges were $2,500. This addi-
tional $4,600 increases the above unit costs to $864 per person, $1,842 per dwelling unit, $3,946
per structure, and $5,530 per GP, respectively.
Early bids on the Grandview Lake project were rejected because they were almost twice the
engineering estimates. Subsequent bids were acceptable. The final cost of the 28,352-foot (8,640-
metre) pressure main was $35,491, or about $1.25 per foot ($4.11 per metre). This total includes
blacktop road repair, manual ($125 each) and automatic ($200 each) air-release valves, gate valves
and boxes ($100 each), and a small vacuum collection station. The total on-lot costs are difficult to
establish from the information available, but some of the data are useful. The bid price for 6,600
feet (2,014 metres) of 1-inch (2.54-cm) service line was 60 cents per foot ($2.07 per metre); for 360
feet (110 metres) of 1.5-inch (3.81-cm) line, it was 95 cents per foot ($3.12 per metre). The cost of
the curb cock is shown to be $20. In addition, a 1971 report on the project showed that on-lot in-
52
-------
stalled costs for the various GP and STEP units ranged from $1,000 to $1,500, assuming 150 feet
(45.8 metres) of service line and no electrical hookup.12 The original costs at Grandview reflected
the use of homemade GP units, which were later replaced with commercial ones.14 Engineering and
legal fees amounted to $23,384 at 1969 rates.
Several costs have been accumulated on the various equipment components of a pressure sewer
system. Installed prices for PVC pipe, for example, have ranged from $1.00 to $3.00 per lineal foot
($3.28 to $9.84 per metre) in sizes up to 6-inches (15.2-metres) nominal diameter. 1:L>a'f In diffi-
cult locations, where more expensive methods of installation than simple trenching are required,
these costs have risen as high as $6.00 to $9.00 per lineal foot ($19.68 to $29.52 per metre).I8,d,h
Carcich et al.2 and Bowne20 indicate that costs could be as high as $15.00 per lineal foot ($49.20
per metre) for some areas.
Main line accessories, such as air-release valves, have been installed at a cost of $120 to $350
cach.11*3^ Cleanouts have been estimated to cost from $150 to $400 each.2'11'18 Valve boxes
cost between $100 and $900 each, depending on their design and construction conditions.a>d
GP's are estimated to cost from $1,000 to $2,000 each, with an additional $300 to $700 in-
stallation cost per unit.2 STEP's cost about $200 to $400. Depending on the amount of ancillary
equipment (alarms, valves, sensors, switches, tankage, removal'mechanism), STEP units may be install-
ed for $1,000 to $2,000 each.20>h Bowne breaks down his estimate for a septic tank that must be
replaced in an existing home to $450 for replacement, $150 for the pump vault, $250 for the
pump, $150 for electrical work, and $400 for the balance. This results in a grand total of $1,325 for
the on-lot facilities.20
Operation and Maintenance Costs
Operation and maintenance costs are generally unknown for pressure sewer systems. For all
systems, the cost of main repairs and cleaning must be added to the operation and maintenance
needs of the pressurization facility. Leckman34 has estimated the maintenance cost on GP units to
be $4 to $8 per month, and Dounoucos46 has estimated a GP maintenance cost of 1.4 to 2.0 per-
cent of the on-lot capital cost per year. Bowne relates that GP service contracts have been instituted
at rates of $48 and $60 per year.20 The Priest Lake system operator indicates that effluent pumps
can be rebuilt for $50 to $100 each.' Replacement motors cost less than $100, while seals, bearings,
and capacitors cost about $7, $5, and $9, respectively.20 Bowne has estimated that on-lot systems
will require an operation and maintenance cost of $50 per year, and that pressure main operation
and maintenance will cost $100 per year per mile ($62 per year per kilometre).20
Septic tank cleaning generally costs $30 to $50 and is required at 3- to 5-year intervals to pro-
tect the grease and solids removal capability of the tank. This assumption is based on traditional
septic tank practice, where pumping is performed to protect the subsurface disposal field from po-
tential clogging caused by wholesale unloading of grease and solids. In the case of STEP systems, the
results of septic tank failure may not be as severe when occurring at an isolated location, but simul-
taneous failures of several tanks might result in serious problems because of fouling of pumps, con-
trols, or pressure mains. Schmidt surveyed 12 septic tanks in the Miami STEP systems and discovered
that accumulations of sludge and scum were significantly less than earlier U.S. Public Health Service
studies.27-29,6 Bowne, like Schmidt, suggests a 10-year interval between pumpings, tempered ini-
tially by yearly inspections to determine individual site accumulation rates for developing rational
"S. M. Rees, SIECO, Inc., personal communication.
dj. Schultz, Becher-Hoppe Engineers, Inc., personal communication.
eH. Schmidt, General Development Utilities Company, personal communication.
*C. W. Rose, Farmer's Home Administration, personal communication.
hL. R. Clark and J. E. Eblen, C & G Engineering, personal communication.
*W. C. Bowne, Douglas County (Oregon) Road Department, personal communication.
53
-------
pumping schedules.20-6 The cost of inspection and pumping must be included in the operation and
maintenance cost estimate for STEP systems.
Operation and maintenance costs on a pressure sewer main are difficult to assess but are likely
to be less than a conventional gravity system. Based on the data of Smith and Eilers,5 an updated
average cost for gravity sewer operation and maintenance is probably between 7 and 8 cents per
year per foot (23 and 26 cents per year per metre) of pipe, which converts to about $400 per year
per mile ($248 per year per kilometre). Bowne used actual operation and maintenance costs for
rural water supply systems to obtain a pressure main operation and maintenance estimate of $100
per hour per mile ($62 per year per kilometre).20 Because burial depths of water mains and pres-
sure sewer mains, as well as many of their other physical features, are quite similar, this analogy is
probably the best estimate available at this time.
A few bits of data may be of use in checking some of the above assumptions. These data in-
clude a 0.0033 service ratio for the E/One GP units at Albany.8 In terms of days per year requiring
service calls, this corresponds to 1.2. At Grandview Lake, E/One GP units required 19 service calls
in 11,800 unit days of operation,15 reflecting a service ratio of 0.0016 or about 0.6 service days
per year. Hendricks and Rees present similar numbers for the same project and GP unit.14 Their
information indicates that the Hydromatic and Tulsa GP units indicate approximately 1 and 4 serv-
ice days per year, respectively. From the above data and the factory improvements and modifica-
tions that have occurred in the interim, it would seem prudent to assume a conservative service re-
quirement for the GP units of about 1 day per year.
The Priest Lake STEP systems had problems with 8 percent and 2 percent of the pumps (total
approximately 500), respectively, during the first 2 years of operation.20 Much of the first year's
operation and maintenance was because the supplier provided improper impellers on the pumps.
During the summer season of the third year when all units were in operation, approximately five
service calls per week were required. The operation and maintenance charge per home is $5 per
month for these systems.
The Miami STEP systems are serviced by the developer for the same monthly charge that is
levied on gravity system contributors.6 This includes about 1 man-hour per year of preventive main-
tenance on each STEP unit. One-half hour of operation and maintenance was required on the Big
Bend system during the first month of operation.11
Some Oregon systems are using submersible sump pumps of the same type as those employed
in STEP systems for pumping raw sewage.20-11 The operation and maintenance experience for one
installation is said to be five service calls per month for a system of about 150 pumping units.8
The consensus, from the foregoing experiences, is that an effective preventive maintenance and
overall operation and maintenance program for STEP systems would involve a yearly inspection of
each pressurization facility (septic tank, pump, sensors, valves, and so forth) and about 0.5 service
call per year. Pressure main operation and maintenance should be less than that required for GP
pressure mains because most of the problem-causing material (fibers, grease, and so forth) is re-
moved by the septic tank.
The power cost of GP or STEP units can be estimated. The Albany and Phoenixville GP infor-
mation conservatively indicates the need for about 1 Wh/gal (0.264 kWh/m3). The power cost can
be estimated by multiplying this figure by the number of occupants per home and the average
bR. P. Farrell, Environment/One Corporation, personal communication.
eH. Schmidt, General Development Utilities Company, personal communication.
gj. Ward, personal communication.
hL. R. Clark and J. E. Eblen, C & G Engineering, personal communication.
54
-------
wastewater flow generated daily per capita. For example, if a GP were to be used by a four-person
family with an average wastewater flow of 50 gallons per capita per day (0.19 m3 per capita per
day), the monthly power cost would be about 15 cents on the basis of 2.5 cents per kilowatt-hour.
Bowne estimates that the power cost for 0.33 hp (0.25 kW) STEPs would be about 10 cents per
month.32 A conservative estimate for both STEP and GP units would be about 20 cents per month,
or about the same as an electric coffeemaker.b Actual costs for each installation will depend on the
number of people served, the cost of electricity, and the specific pressurization device chosen.
Estimating Procedures
The system depicted in figure 1-10 can be used as a simplified example, assuming a scale of 1
inch = 300 feet (1 cm = 36 metres). Additional assumptions are: PVC mains of 2-inch (5.1-cm)
nominal diameter, except for a 3-inch (7.6-cm) interceptor (branch 7); service lines of 1.25-inch
(3.2-cm) nominal diameter PVC; and GP units.
Table 1-9 shows assumed unit costs and a rough estimate of the capital cost of the system. The
total of $104,850 represents a cost per house of $2,760.
Two things are shown in this example: the economical nature of the pressure sewer and the
high cost of GP's. It is because of the latter factors that the use of STEP'S is being investigated and
tried by many.
In many rural areas sewers are required because poor soil conditions have precluded the con-
tinued use of the original ST-SAS. If this were the case in the example location, the data developed
by Bowne indicate that STEP units can be substituted for the GP units at a cost of about $1,000
per installation as compared to $2,000 for the GP installation. This substitution in the foregoing
example would reduce the cost per home to about $1,760.
Approximate operation and maintenance costs for the example installation appear in table 1-10.
The total of $2,161 per year amounts to $56.87 per home, or a monthly cost of about $4.74. The
amortized capital cost must be added to this amount to get the total monthly cost.
Table \-9.-Assumed unit costs and estimated capital cost of typical pressure sewer installation3
Component
3-inch PVC pipe
2-inch PVC pipe
1 25-inch PVC pipe
Service line connections
GP units including electrical hookup
Cleanouts with manual air release valves
Total
Unit cost
$4 per foot
$3 per foot
$2 per foot
$35 each
$2 000 each
$500 each
Quantity
800 feet
3,140 feet
5,700 feet
38 each
38 each
7 each
Esti-
mated
system
cost,
dollars
3,200
9,420
11,400
1,330
76,000
3,500
104,850
aShown in figure 1-10.
bR. P. Farrell, Environment/One Corporation, personal communication.
55
-------
Table 1-10. —Operation and maintenance costs of typical pressure sewer installation*
Component
Pipe
GP
Power
Total
Unit cost
$100 per mile per year
$54 each per year
$2 40 each year
Quantity
9 640 feet
38 each
38 each
Approxi-
mate
cost per
year,
dollars
18
2052
91
2 161
aShown in figure 1-10.
For this example, no engineering or legal fees or other additional costs will be considered.
Therefore, amortization of the $104,850 capital cost over 20 years at 5-7/8 percent interest (muni-
cipal rate) yields an annual cost of $238 per home, or $19.83 per month. The total monthly cost
per home is then $24.57. Because these systems are eligible for EPA Construction Grant funding,
the cost per home could be reduced to a fraction of that amount.
From the discussion of the operation and maintenance costs for GP and STEP systems, there is
not enough evidence available at this time to justify a difference in the operation and maintenance
cost estimates for these two types of pressure systems. Therefore, the substitution of STEP units for
GP units in the above example yields a monthly cost per home of $4.74 for operation and mainten-
ance, in addition to the amortization cost of the STEP system computed on the same basis as the
GP system. The amortization of the $66,850 capital cost yields an annual cost of $151 per home, or
$12.65 per month. The total monthly cost for the STEP system is, therefore, $17.39. Grant eligibil-
ity is the same for both systems, except that new septic tanks required for the STEP approach are
not eligible.
The example is admittedly crude, but it gives some idea of the cost-estimating procedures nec-
essary to evaluate proposed pressure sewer systems. Additional factors will have to be evaluated to
properly accomplish such an estimate in a real situation.
CONCLUSIONS
Pressure sewer systems are a viable alternative technology and should be considered in any
cost/effective analysis of alternative wastewater management systems in rural communities.
Pressure sewers offer many advantages over conventional gravity sewers in areas where:
• Population density is low.
• Severe rocky conditions exist.
• High ground water or unstable soils prevail.
• Undulating terrain predominates.
The most serious impediment to wider adoption of pressure sewer technology is the present
lack of comprehensive long-term operation and maintenance data and treatment information.
56
-------
Lower capital costs and significantly shorter construction times are inherent in pressure sewer
technology, as compared to conventional methods.
Pressure sewers should only be considered with properly conceived management arrangements.
Failure to do so could seriously limit the effectiveness of this technology.
Two major types of pressure sewer system designs are available: GP systems and STEP systems.
The relative merits of the systems should be weighed by the engineer in his cost/effective
evaluation.
REFERENCES
!E. J. Bowles, "Pressurized Waste Water Collection," presented at the Water Pollution Control
Federation Conference, Atlanta, Ga., 1972.
2I. G. Carcich, L. F. Hetling, and R. P. Farrell, "Pressure Sewer Demonstration,"./. Environ.,
Eng. Div., Am. Soc. Civ. Eng., 100,1, 25-40,1974.
3C. W. Rose, "Rural Wastes: Ideas Needed," Water Wastes Eng., 9, 2, 46-47,1972.
4P. L. Gainey and T. H. Lord, Microbiology of Water and Sewage, Prentice-Hall, Inc., Engle-
wood Cliffs, N.J., 1957.
5R. Smith and R. G. Eilers, Cost to the Consumer for Collection and Treatment, U.S. Environ-
mental Protection Agency, Report No. 17090-07/70,1970.
6Manual of Septic-Tank Practice, U.S. Public Health Service, Publication No. 526,1969.
7 American Society of Civil Engineers, Combined Sewer Separation Using Pressure Sewers, U.S.
Department of the Interior, Federal Water Pollution Control Administration, FWPCA Report No.
ORD-4,1969.
81. G. Carcich, L. J. Hetling, and R. P. Farrell, A Pressure Sewer Demonstration, U.S. Environ-
mental Protection Agency, Report No. R2-72-091,1972.
9M. A. Clift, "Experience with Pressure Sewerage," J. Sanit., Eng. Div., Am. Soc. Civ. Eng.,
94, 5, 849-865,1968.
10I. G. Carcich, L. J. Hetling, and R. P. Farrell, "The Pressure Sewer: A New Alternative to
Gravity Sewers," Civ. Eng., 44, 5, 50-53,1974.
11G. Mekosh and D. Ramos, Pressure Sewer Demonstration at the Borough of Phoenixville,
Pennsylvania, U.S. Environmental Protection Agency, Report No. R2-73-270,1973.
12G. F. Hendricks, "Pressure Sewage System and Treatment at Grandview Lake, Indiana,"
presented to the American Society of Agricultural Engineers Convention, Pullman, Wash., 1971.
13S. M. Rees and G. F. Hendricks, Grandview Lake Sewage Research and Demonstration
Project—Annual Report, prepared for the U.S. Environmental Protection Agency, 1971.
14G. F. Hendricks and S. M. Rees, Economical Residential Pressure Sewage System with No
Effluent, U.S. Environmental Protection Agency, Technology Series, Report No. EPA-
600/2-75-072, December 1975.
57
-------
15D. M. Cochrane, "Pressurized Sewer Systems: Regulatory Agency's Viewpoint," presented
at the Water Pollution Control Federation Conference, Denver, Colo., October 1974.
16G. C. Gray, "Environmental Constraints Challenge Designers of Shoreline Community near
Kansas City, Missouri," Prof. Eng., 45, 6, 1975.
17T. C. Williams, "Plastic Pipe and Pressure Sewers Mark Expansion," Water Wastes Eng., 12,
11,85,1975.
18 Alfred Crew Consulting Engineers, Inc., Preliminary Report on Water Pollution Control
System-Siek Road Area (Addendum No. 2), prepared for Borough of Kinnelon, N.J., 1974.
19R. L. Sanson, "Design Procedure for a Rural Pressure Sewer System," Public Works, 10,
86-87, 1973.
20W. C. Bowne, Glide-Idleyld Park Sewerage Study, Douglas County, Ore., Report, 1975.
21 Detail Drawings and Specifications SA-2021B-137 through SA-2021B-144, Environment/
One Corporation, 1973.
22Design Handbook for Low Pressure Sewer Systems, Environment/One Corporation, 1973.
23E. R. Bennett and K. D. Linstedt, Individual Home Wastewater Characterization and Treat-
ment, Colorado State Environmental Resources Center, Completion Report No. 66, July 1975.
24M. Witt, R. Siegrist, and W. C. Boyle, "Rural Household Wastewater Characterization,"
Home Sewage Disposal, Proc. Am. Soc. Ag. Eng., 175, 79-88,1975.
25E. E. Jones, Jr., "Domestic Water Use in Individual Homes and Hydraulic Loading of and
Discharge from Septic Tanks," Home Sewage Disposal, Proc. Am. Soc. Ag. Eng., 175, 89-103,1975.
26Metcalf and Eddy, Inc., Wastewater Engineering, McGraw Hill, New York, 1972.
27S. R. Weibel, C. P. Straub, and J. R. Thoman, Studies on Household Sewage Disposal Sys-
tems—Part I, U.S. Public Health Service Publication, 1949.
28T. W. Bendixen, M. Berk, J. P. Sheehy, and S. R. Weibel, Studies on Household Sewage Dis-
posal Systems—Part II, U.S. Public Health Service, Publication, 1950.
29S. R. Weibel, T. W. Bendixen, and J. B. Coulter, Studies on Household Sewage Disposal Sys-
tems—Part III, U.S. Public Health Service, Publication No. 397,1954.
30J. W. Patterson, R. A. Minear, and T. K. Nedoed, Septic Tanks and the Environment, pre-
pared for Illinois Institute for Environmental Quality, 1971.
31 Specifications for Hydromatic Model CSPG-150A, Hydromatic Pump Company, 1973.
32W. C. Bowne, Pressure Sewer Systems, prepared for Douglas County, Ore., 1974.
33 A. T. Voell, Investigation of Pressure Sewer System and Advanced Waste Treatment, pre-
pared for Chatauqua County, N.Y., Health Department, 1972.
34 J. Leckman, Pressurized Sewer Collection Systems, prepared for Illinois Institute of Envi-
ronmental Quality, 1972.
58
-------
35L. J. Flanigan and R. A. Cudnik, Review and Considerations for the Design of Pressure
Sewer Systems, Hydromatic Pump Division, Weil-McLain Company, 1974.
36 Wafer Flow Characteristics of Thermoplastic Pipe, Plastics Pipe Institute, Technical Report
No. PPI-TR14, 1971.
37Report on the Performance of Grinder Pump Model Farrell 210, National Sanitation Foun-
dation Testing Laboratory, Report No. C-9-3,1973.
38J. C. Kent, The Entrainment of Air by Water Flowing in Circular Conduits with Downgrade
Slopes, master's thesis, University of California, 1952.
39R. P. Farrell, Operation of Grinder Pumps and Pressure Sewer Systems under Abnormal
Conditions, information bulletin from Environment/One Corporation, 1974.
40M. F. Hobbs, Relationship of Sewage Characteristics to Carrying Velocity, Report No. R-
2598 to Am. Soc. of Chem. Eng., 1967.
41 Sewerage Manual, Pennsylvania Dept. of Environmental Resources, Bureau of Water Quality
Management, Publication No. 1, 3d ed., Harrisburg, Pa., 1976.
4 2Code of Practice for Plastics Pipework: Part 1—General Principles and Choice of Material,
British Standards Institution, Publication CP 312 (Gr B), London, England, 1973.
43Code of Practice for Plastics Pipework: Part 2—Unplasticized PVC Pipework for the Convey-
ance of Liquids under Pressure, British Standards Institution, Publication CP 312 (Gr B), London,
England, 1973.
44Great Lakes—Upper Mississippi River Board of State Sanitary Engineers, Recommended
Standards for Sewage Works, Health Education Service, Albany, N.Y., 1971.
45R. D. Pomeroy, Process Design Manual for Sulfide Control in Sanitary Sewerage Systems,
U.S. Environmental Protection Agency, Technology Transfer Publication, 1974.
46 A. Dounoucos, "Sanitary System Construction Costs Turn Engineering Attention to Alter-
nate Solutions," Prof. Eng., 44, 8,1974.
59
-------
Part II
VACUUM SEWERS
BACKGROUND
Vacuum sewers are one of the alternatives in sewer systems for smaller communities, land de-
velopments, and rural areas. The advantages of these systems may include substantial reductions in
water use, material costs, pipe size, excavation costs, and treatment expenses, and a potential for
overall cost/effectiveness.
Vacuum systems depend on a central vacuum source constantly maintaining 15 to 25 inches of
mercury on small-diameter collection mains (fig. II-l). A gravity vacuum interface valve separates
atmospheric pressure from the vacuum in the mains. The valve can be either in the home sanitary
sewer service line or in a vacuum toilet. When the interface valve opens, a volume of sewage enters
the main, followed by a volume of atmospheric air. After a preset interval, the valve closes. The
packet of liquid, called a slug, is propelled into the main by the differential pressure of vacuum in
the main and the higher atmospheric pressure air behind the slug. After a distance, the slug breaks
down by shear and gravitational forces, allowing the higher pressure air behind the slug to slip past
the liquid. With no differential pressure across it, the liquid then flows to the lowest local elevation,
and vacuum is restored to the interface valve for the subsequent operation. When the next upstream
interface valve operates, identical actions occur, with that slug breaking down and air rushing across
the second slug. That air then impacts the first slug and forces it further down the system. After a
number of operations, the first slug arrives at the central vacuum source. When sufficient liquid
volume accumulates in the collection tank at the central vacuum source, a transfer device, such as a
sewage pump, delivers the accumulated sewage to a treatment plant (fig. II-2).
Vacuum
pump
Transport
pockets
To treatment facility—J
Figure 11-1. Elements of a vacuum sewage system.
61
-------
River
Force main
D,
D,
Collection
station
rO
Cleanout-
Cff
U
( /r
LineB-'
rf
D
rf
D
Tn
a
D?
LT
Line A
V
-'
rf
CP
I—T
Cleanout"
• House
I I
PLAN
Collection station-
CROSS SECTION
Figure 11-2. Typical layout—vacuum sewer system.
Vacuum sewage collection systems have been patented in the United States since 1888, when
Adrian LeMarquand invented a system of sewage collection by barometrical depression.1 The first
commercial applications of such systems were by the Liljendahl Corporation (now known as Elec-
trolux) of Sweden in 1959.2 Currently, several companies in the United States are actively market-
ing vacuum equipment for residential systems in this country. There are significant differences
among the designs of the four types of currently installed systems covered in this paper: Liljendahl-
Electrolux, Colt Envirovac, AIR VAC, and Vac-Q-Tec.
62
-------
The Liljendahl-Electrolux system was introduced to this hemisphere in the Bahamas (fig. II-3)
between 1965 and 1970. This concept uses separate black and gray water collection mains. Black
water refers to toilet wastes and gray water includes all other domestic wastewater. Black water is
generated at a vacuum toilet (fig. II-4), which discharges about 3 pints per flush. The gray water
generated in the home is discharged into the system by a specially designed vacuum valve. The
wastewater is transported through separate black and gray water vacuum mains to vacuum collec-
tion stations for disposal. About 1,600 vacuum toilets are components of 14 separate systems in the
Bahamas. The 90-percent reduction in toilet wastewater volume was a definite consideration in the
selection of these systems for critically water-short Nassau.
The first residential vacuum collection system in the United States was designed by Vac-Q-Tec
and serves the Lake of the Woods development near Fredericksburg, Va. Vac-Q-Tec has designed
several other residential vacuum sewage collection systems, all in use by private developers. This
system uses concepts of the Liljendahl system but has many important differences.3 The Vac-Q-Tec
system requires no inside vacuum toilets or vacuum plumbing and has combined black and gray
water vacuum collection mains. The system includes a 750-gallon storage tank at the homesite;
pneumatically operated, electrically controlled vacuum valves; 4-inch polyvinyl chloride (PVC) col-
lection lines; and 13 vacuum collection stations.
The Colt Envirovac system, depicted in figure II-5, is the direct descendant of the Liljendahl-
Electrolux system.4 Colt is currently marketing self-contained vacuum sewage comfort stations, as
well as preengineered community vacuum collection systems. The Colt system at South Seas Planta-
tion near Fort Myers, Fla., serves 33 residences, with separate building plumbing for gray and black
wastewater. A gray water valve serves each residence. Black water piping from the vacuum toilet
joins the gray water vacuum piping immediately downstream of the gray water valve. The system
then functions as a single pipe network to the vacuum collection station.
The AIRVAC Company markets a pneumatically controlled and operated vacuum valve for
combined gray and black water systems (fig. II-6). A design manual illustrating the use of the AIR-
Vacuum
pump
Black water
collection
tank
Buffer volume—^/
jr*i
Gray water
collection
tank
Sewage-^
pumps
Gray water vacuum main
Transport pockets
-To treatment facilities
Figure 11-3. Liljendahl-Electrolux vacuum sewer collection system—Nassau.
63
-------
Vacuum toilet
Flushing mechanism
Discharge valve'
Figure 11-4. Vacuum toilet.
.Vacuum main
L
Gray
water
valve
-Transport pocket
-Vacuum main
Figure 11-5. Colt Envirovac vacuum sewage collection system.
To treatment
facilities
64
-------
Atmospheric air
Vacuum
Atmospheric
air
Check valve
Flow
Figure 11-6. AIR VAC 3-inch valve.
VAC product is currently available. AIR VAC has seven operating residential systems throughout the
United States, with about 1,000 valves installed. The AIRVAC system (fig. II-7) uses conventional
gravity household plumbing, with wastewater discharging into their 3-inch valve. The vacuum valve
is located in a valve pit. This valve starts its operation cycle when it senses approximately 10 gallons
of accumulated sewage, admitting that liquid and a quantity of air to the mains.5 Sewage travels
through 3-inch, 4-inch, or 6-inch mains to a vacuum collection station.
65
-------
Vacuum—'
valve
-Transport pocket
-Vacuum main
Figure 11-7. AIRVAC vacuum sewage collection system.
SYSTEM PARAMETERS
Sewage
pump
To treatment facilities
General
A comparison was made of the basic system parameters of each of the four types of vacuum
sewage collection installations. The comparison was based on manufacturers' promotional literature,
design guidelines, and site visits to 18 systems throughout the Western Hemisphere. The major dif-
ferences among these collection systems are shown in table II-I. The water-saving feature of the
Electrolux and Colt systems is reported to be as much as 27 percent of the total in a domestic appli-
cation with the use of vacuum toilets.4'6'7 AIRVAC and Vac-Q-Tec systems can be altered to
accommodate these water-saving devices.
Vacuum Valves
Vacuum toilets are flushed after each use, while vacuum valves operate automatically, based on
the volume of gray or gray and black water behind the valve. When a predetermined volume has
accumulated, the valve opens, provided there is adequate vacuum available. Atmospheric air forces
the sewage into the mains, which is followed by a volume of air. The valve is actuated by a pneu-
matic controller in all systems except the Vac-Q-Tec. Vac-Q-Tec's valve operation requires a sepa-
rate electrical power source at each valve site to control valve operation.
The Vac-Q-Tec's gravity-vacuum interface valve assembly is unique in that it requires an exter-
nal electrical power source. The valve's position can be monitored and the valve operated from the
collection station through an extra set of contacts in the controller. A separate cycling mode, called
Auto-Scan, offers added flexibility. This mode locks out the accumulated volume-cycle command
from each valve, and sequentially operates each valve during low-flow periods. Additional operating
and skilled electronics technicians are required to maintain these more complex systems.
The capability for shaving peak flows is possible with the Vac-Q-Tec system when additional
controls are added to the base system. During high-flow periods, wastewater is stored in the 750-
66
-------
Table 11-1 .—Vacuum collection system parameters
System type
Electrolux
Colt
Vac-Q-Tec
AIR VAC ...
House piping
Black and gray separate
Black and gray separate
Conventional plumbing
Conventional plumbing
Valve type
Black vacuum toilets;
gray, pneumatic
valves, 2 in
Black vacuum toilets;
gray, pneumatic
valves, 2 and 3 in
Electrically actuated
pneumatic valve
Pneumatic valve, 3 in
Discharge
volume
Black, 3 pt
Gray, 10 gal
Black, 3 pt
Gray, 10 gal
75-1 00 gal
10-15 gal
Piping profile
Set configuration with
traps
Set configuration with
traps
Parallels terrain with
traps
Set configuration with
traps
Cleanouts
200-250 ft
200-250 ft
No
Yes
Collection line
Black, T/2 and 2 in;
gray, 2 and 3 in;
PVC, solvent weld
Single main, 3, 4, and
6 in; PVC, special
"O" ring
Single main, 4 in;
PVC
Single main, 3, 4, and
6 in; PVC, solvent
weld
O
-------
gallon residential tank and is released later during low-flow periods. All other systems must be de-
signed to handle peak flows.
The amount of water entering the system with each valve operation varies with each manufac-
turer and the type of gravity vacuum interface valve. The vacuum toilet admits approximately 0.3
to 0.4 gallon per flush, and the pneumatically controlled valves admit 10 to 15 gallons per cycle.
U.S. Navy research8 has reported that good transport characteristics are found with sufficient inlet
air and small enough slug loadings for the available pressure differential to overcome the liquid's
inertia. This results in rapid slug breakdown, reestablishing vacuum quickly at upstream valves.
Piping Systems
Piping profiles vary in each system. The manufacturer recommends differing profiles, depend-
ing on uphill, downhill, or level terrain. Vacuum reformation traps are located where the designer
wishes to reform a slug of water for transport purposes. Traps also are used to gain elevation by
raising the mains closer to the ground surface or to conform to terrain variations.
PVC pipe is common to all systems. Solvent-weld and O-ring joints have been used; when
O-ring is used, however, it must have a joint designed to seal against vacuum. Studies showed in
some systems initial savings in capital expenditures from the use of low-cost, smaller diameter PVC
pipes rather than gravity sewer mains.7 Construction cost savings were also realized by not having to
shore deep, sandy trenches or blast deep rock trenches to install gravity sewers.
Collection Stations
Collection station (often called vacuum central) design parameters, shown in table II-2, vary
with each manufacturer. Electrolux and Colt vary their use of vacuum reserve tanks with each in-
stallation, while AIRVAC and Vac-Q-Tec use vacuum reserve tanks between the receiving tanks and
their vacuum pumps. The use of vacuum reserve tanks extends vacuum pump life by reducing pump
cycling.
Vacuum pump construction has been both sliding vane and liquid ring. Pros and cons are nu-
merous on the use of each, and no standard has yet emerged. Liquid ring pumps, however, have
been used more frequently in vacuum sewer applications. The contents in the vacuum collection
tank must be transferred to a treatment facility after sufficient volume has been collected. Nonclog
Table 11-2.—Vacuum collection station parameters
System type
Electrolux . . .
Colt
Vac-Q-Tec . . .
AIRVAC
Separate black and
by installation.
Common receiving v
One receiving vessel
One receiving vessel
Receiving tank
gray water vessels. Reserve tank use varies
essels. Reserve tank use varies.
plus reserve tank.
plus reserve tank.
Receiving tank
evacuation device
Sewage pumps
Sewage pumps
Pneumatic ejectors
Sewage pumps
Valve
moni-
toring
and
con-
trol
capa-
bility
No
No
Yes
No
68
-------
sewage pumps with sufficient net positive suction head (NPSH) to overcome tank vacuum are gener-
ally used. NPSH refers to the total suction lift in feet measured at the suction nozzle less the vapor
pressure of the liquid in feet. It is important to use pumps whose shaft seals close against vacuum;
otherwise, system vacuum will be depleted during low-flow periods.
VACUUM SEWAGE CHARACTERISTICS
In a contractual study for EPA, a sampling and monitoring program was initiated to determine
sewage flow and strength characteristics from AIRVAC's Mathews site. This level site, with persis-
tent high ground water, offered an excellent baseline study as the newly completed vacuum system
served residences and businesses, while an existing gravity system served the central portion of the
town. Wastewater flow and strength characteristics were studied for a continuous 7-day period.
Monitoring was accomplished by Manning flow meters, level recorders, and automatic samplers lo-
cated at the influent splitter box of a 100,000-gal/d contact stabilization treatment plant.
Mathews
Gravity wastewater flows exhibited a characteristic diurnal flow pattern (fig. II-8). This com-
posite of 7 days' flow data showed vacuum wastewater flows lagged the gravity flow pattern by up
to 2 hours, depending on the time of day. Flows from the vacuum system showed greater hourly
flow fluctuations than gravity flows because of the storage and intermittent discharge features of
the vacuum station. The midnight to 6 a.m. time period points to possible reduced infiltration of
the vacuum system over the gravity system.
Dissolved oxygen (DO) of the gravity sewage, collected from sources close to the monitoring
station, exhibited an uncharacteristically high 6.0 to 7.0 mg/1, possibly because of closeness to the
200 n
o
^ 150
_J
Q
UJ
(D
S 10°
LU
o
DC
50'^
0
Vacuum
flows
•4-
12p.m. 4a.m. 8a.m. 12m. 4p.m. 8p.m.
Figure II-8. 7-day composite wastewater flows at Mathews treatment plant.
12p.m.
69
-------
source (+ 300 feet) and turbulence from pumping immediately before the sampling point. Vacuum
sewage consistently maintained a DO in the range of 0.6 to 1.0 mg/1. The low DO readings could
have been caused by the collection system vacuum conditions removing DO. From dye tests per-
formed during a morning high usage period, discharges from a vacuum valve 6,000 feet from the
collection station reached the treatment plant in 10 hours. In 2 hours the flow introduced by a
valve 1,500 feet from the collection station appeared at the plant. The sewage residence time in the
receiving tank at the collection station is 1 hour or less. This time allows for a reduction in the re-
ceiving tank of the sewage's oxygen content.
Biochemical oxygen demand (BOD) tests on both gravity and vacuum sewage composite sam-
ples were run for the 7-day monitoring period. Vacuum sewage exhibited a 5-day BOD (BOD5) of
136 mg/1, while gravity sewage contained 76 mg/1. The lower BOD in gravity sewage can be attri-
buted to infiltration in the old gravity mains located in the high ground water area. Reaction con-
stants for biological degradation of raw sewage, or K values, were determined for both gravity and
vacuum sewage. Gravity-collected sewage exhibited a K value of 0.176, while vacuum sewage
yielded a K value of 0.168. From these limited data, no substantial difference can be seen in the
treatment of these two wastewaters.
Vacuum-collected sewage was observed to be more homogeneous, with solids much more fine-
ly divided than gravity sewage, a consequence of the turbulence during vacuum transport.
South Seas Plantation
Sewage samples from Colt's South Seas Plantation vacuum system were also collected over a
7-day period. With vacuum toilets, a correspondingly higher BOD5 would be expected. The BOD5
from the 33 residential connections was 371 mg/1. During the test period, a substantial savings in
water was realized by the Colt system using vacuum toilets.
COST COMPARISONS
The life cycle cost of the vacuum sewage collection system was compared to the alternatively
bid gravity system for Mathews, shown in table II-3. This comparison is not intended to be typical
but represents one system where alternative bids were taken and data were available. This value
engineering analysis9 brings all costs to the present worth position and then amortizes these costs
over an assumed period—20 years in this case.
Initial capital expenditures are bid prices or installed costs and are in the present worth form.
These costs are amortized over the bond issue life at 6 percent, which converts these figures to
annual capital recovery cost amounts. The costs for replacing equipment included rewinding motors
in the seven pumping stations, which would be required in the gravity alternative at 10 years and 20
years. The identical time periods are selected for replacement of vacuum valve assemblies and man-
holes. No allowance has been made for price increases because of inflation. Operating experience in
the United States is limited, and data on experienced replacement periods are not available. The
replacement costs for each alternative are brought back to their present worth based on the year of
their replacement and then amortized by the capital recovery annuity over the life of the project.
Annual operation and maintenance costs for each system are based on actual previous year
expenditures for the vacuum system or from similar costs incurred in other gravity-serviced com-
munities. An additional amount is added to the annual cost of the gravity system to account for
treatment of infiltration at a rate of 100 gal/in diameter per mile per day. The annual difference
70
-------
Table \\-3.-Life-cyclecost analysis, Mathews, Va., sewer system
[Dollars]
Costs
System
Gravity
Vacuum
Initial:
Instant contract:
Base 555,325 270,773
Interface:
Pump stations 88,267
Vacuum collection station 105,028
Other (collateral) 0 0
Total 643,592 375,801
Replacement (life cycle):
Year 10 at 6 percent 6,000 60,062
Present worth of future replacement cost (0.5584) 3,350 33,538
Year 20 at 6 percent 6,000 60,062
Present worth of future replacement cost (0.3118) 1,870 18,727
Life-cycle:
Annual owning and operating:
Capital recovery of total initial cost amortized at 6-percent, 20-year initial
factor (0.08718) 56,108 32,762
Capital recovery of present worth of replacement cost at 6 percent:
Year 10 292 2,923
Year 20 163 1,632
Annual cost:
Maintenance 4,101 7,235
Operation 1,769 2,452
Infiltration at 100 gal/in diameter per mile per day 1,009
Total 63,442 47,004
Annual difference 16,438
Present worth of annual difference 188,543
(fig. II-9) shows a savings of almost $16,438 per year for the vacuum system, which is equivalent to
a life cycle present worth of $188,543. The cost differential between this vacuum system and its
gravity alternative would be less if an inflation component were incorporated into the replacement
cost expenditures.
Recreational and second-home developers may see a significant initial cost advantage. Valve
assemblies need to be added to the basic piping and collection station only when an owner decides
to build on his property. The cost of the valve is then paid by the homeowner. This cost is usually
in excess of the connection fee cost assessed by municipalities.
71
-------
tu
a
jo 40
o
13
TJ
CO
v>
O
V)
O
O
Operation,
maintenance,
and replacement
10
Gravity Vacuum
Figure II-9. Annual operation and maintenance costs, Mathews treatment plant.
CRITIQUE OF EXISTING SYSTEMS
Earlier vacuum sewage collection systems were often plagued with consistent operational prob-
lems. Although this situation has improved with successive generations of systems, some problems
still exist.
Deficiencies in vacuum systems can be broadly defined into three areas: system design, com-
ponent reliability, and lack of operation and maintenance guidance.
System Experiences
Although vacuum systems provide the means for installing economical sewer systems in prob-
lem areas, early systems were installed without sufficient prototype or field testing of equipment
components. As a result, there were several operational problems. Significant field experience has
72
-------
provided an opportunity for upgrading the original systems, and most problems have now been
eliminated.
Problem Areas
Early U.S. systems, particularly Vac-Q-Tec's, were installed without a thorough investigation
of two-phase transport characteristics. As a result, these systems experienced significant problems in
transporting sewage from as few as 10 percent of the design population10 because of improperly
planned vacuum main profiles, too large slug volumes, and insufficient air-admittance techniques.
Vac-Q-Tec's early systems, particularly at Lake of the Woods, located sensitive electronic valve
control equipment in 55-gallon drums adjacent to the sewage holding tanks. These drums and elec-
tronics control boxes often corroded, causing numerous and continued malfunctions. Electrical
control wires were buried alongside the sewage mains, which resulted in signal failures. The system's
complex electronics, which includes monitoring and controlling valve operation from the collection
station, proved a significant drawback. Highly skilled electronics technicians were required to keep
the system in operating condition. Aware of these problems, current management is undertaking a
replacement program aimed at simplifying the system and correcting existing problems.
Several AIRVAC and Colt systems were found to lack components now generally accepted as
minimum design standards. These items include standby power and system malfunction reporting
devices. The lack of standby power in the Colt system has caused valve boots to rupture when
power outages occurred. An excessive amount of liquid built up behind the valves during outages.
When power was restored and the valves cycled, there was not enough time to discharge the higher
than normal column of water instead of air. The momentum of sewage resulted in the ruptured
valve boots. With the addition of standby power, these problems should be alleviated.
The use of weak materials in early AIRVAC valve manholes caused problems. Manholes made
of tar-impregnated paper deformed when placed in unstable soil or areas subject to vehicle traffic.
Consequently, the transite cover bolt holes would not align with the manhole bolt clips, preventing
adequate fastening and resulting in damage in some cases. Improved manhole materials are now
recommended, such as the spun-wrapped fiberglass valve pit, counterweighted to prevent flotation
with a cast-iron manhole cover, and should eliminate these problems. A breather tube extension
above potential water levels and controller modifications have also minimized past reliability prob-
lems.
Additional problems have resulted from the use of manholes without bottoms in high ground
water areas. During high ground water periods, standing water was able to enter the sensitive sensor-
controller pneumatic circuit, causing valves to continually cycle and deplete system vacuum. These
valve reliability problems were evident in Eastpoint, Fla., and Mathews systems. Reliability has been
improved through recent modifications.
Designers must be aware of a public health hazard that may exist if a house on a vacuum sys-
tem has a vent stack smaller than a 7.62-cm (3-inch) bore. When the valve operates, it may evacuate
those water traps, allowing sewer gas from the local holding tank to enter the house. A 7.62- to
10.16-cm (3- to 4-inch) vent stack, installed on the gravity sewer lateral adjacent to the house wall,
will eliminate this problem.
Valve failure also can cause failure of a system or a large portion of a system. If a valve fails in
an open position or cycles continuously, available vacuum in the system may fall below acceptable
levels.
OPERATION AND MAINTENANCE
Because of the complexity of vacuum equipment, operating personnel must be properly
trained to maintain a vacuum sewer system. Some early installations suffered for lack of proper
73
-------
operation and maintenance manuals and other aids that would have assisted the operators in coping
with this new technology. Manufacturers are now recognizing this need and are reacting accordingly
with improved technical assistance and operation and maintenance manuals to assist system oper-
ating personnel.
Operation and Maintenance Tasks
Operation and maintenance tasks for vacuum systems can be divided into normal operation,
preventive maintenance tasks, and breakdown or emergency operation and maintenance tasks. Each
of these headings can be further divided into tasks related to vacuum valves and to the collection
station.
Valves. Depending on a system's emergency breakdown history, some periodic valve inspection
is required. As a starting point, semimonthly inspection and manual operation of each valve is sug-
gested. An experienced operator learns the sounds a well-functioning valve and controller make and
can use this tool as a preventive maintenance device. The breather lines in the AIR VAC valve should
be inspected for liquid accumulation, which, if found, should be removed. Yearly maintenance in-
cludes an exterior cleaning of AIR VAC's valve breather cap and Colt's sensor controller mechanism.
In hard-water areas, AIR VAC suggests that valves be removed and overhauled every 3 to 4
years for scale removal. At 6-year intervals, new seals and valve seats are recommended.
A card file listing each valve location and any preventive or breakdown maintenance performed
will identify problem locations. This procedure is consistent with good management practice.
Collection Stations. Specific components will vary from station to station, as will the required
operation and maintenance details. Some maintenance procedures common to all vacuum systems
include a daily record of pump running hours, ammeter readings, and oil levels. Weekly procedures
include checking battery terminals and battery condition of the standby generator, exercising the
standby generator, blowing down sight glass of the collection tank (if present), and checking mech-
anical seal pressurizers on sewage discharge pumps (which prevent loss of system vacuum). Yearly
preventive maintenance might include inspection of check valves on sewage discharge and gas evacu-
ation lines.
Breakdown Maintenance
Vacuum system malfunctions occur in one of three places: the valve, the piping system, or the
collection station. Malfunctions in the collection station are usually the result of a pump, motor, or
electrical control breakdown and will not be discussed here.
If a valve malfunctions in the closed position, identification of the broken device is simplified
as the homeowner will experience a backup of sewage in or near his house. A complaint call invari-
ably follows. Replacing the sensor is usually the quickest solution to this problem.
If the system experiences low vacuum, characterized by a low-vacuum relay energizing an
alarm system or by vacuum pumps running excessively with vacuum below normal, a vacuum leak
has occurred. A vacuum leak is possible from either a break in the vacuum transport piping or from
a malfunctioning valve. Breaks in the transport piping usually have been the result of underground
construction (e.g., by a telephone or gas company) in the area cutting the vacuum line rather than a
passive piping system failure. Valve malfunctions that result in low-system vacuum occur either
when a valve sticks in the open position (also very rare), when an AIRVAC valve continually cycles,
or when a Colt valve's boot ruptures. The AIRVAC problem is caused by accumulated moisture in
the sensor lines and is a more common occurrence. Successive generations of controller designs with
74
-------
preventive maintenance may provide a satisfactory solution to this problem. System vacuum loss
from a Colt valve has been reported after rupture of the valve boot in a situation described earlier.
Programed boot replacement with mandatory standby power should prove effective in eliminating
this problem.
An outline procedure for locating the source of vacuum failure has been documented by AIR-
VAC as follows:5
• When a low-vacuum condition occurs in a system, isolate, in turn, each incoming line to
the collection tank to identify the problem line.
• Close off the line with low vacuum. Open remaining lines to clear sewage from them.
• Allow vacuum in operational lines to reach maximum vacuum; then close valves on all
incoming valves.
• Open line with problem. Sometimes high vacuum applied quickly may correct malfunc-
tion. Leave line open to the collection tank.
• Starting at the collection tank, go to the first system isolation valve on problem line. Con-
nect vacuum gage to vacuum valve prior to isolation valve. Close isolation valve and ob-
serve if vacuum builds up. If it does not, problem is between collection tank and isolation
valve. If vacuum rises, repeat process on next isolation valve. Before reopening each isola-
tion valve, allow vacuum to build up in nonproblem sections of sewer to clear that sec-
tion's sewage.
• After isolating problem section, check each valve pit to locate malfunction. Often this can
be accomplished by driving to each pit and listening from the vehicle window.
• After locating the malfunctioning valve, pump accumulated water and remove any debris
from valve manhole. The manufacturer's emergency maintenance procedures should then
be followed.
• If no valves are malfunctioning, check for underground construction that could have
caused a break in the transport piping.
• If construction activity did not cause the leak, isolate leak by plugging vacuum main with
test balloons at selected cleanout point locations.
• After plugging a small segment, inspect segment by walking the line to audibly determine
the location of the underground break.
• Repair pipe section following specific pipe manufacturer's repair procedures.
Based on the Mathews system, it is estimated that 4 hours per connection per year should be
allocated to operation and preventive maintenance. Breakdown maintenance will require time in-
puts in addition to preventive maintenance tasks. Total system operation and maintenance time
required was found to range from about 4 hours per connection per year in systems with few prob-
lems to over 30 hours per connection per year in systems with significant problems.
At AIRVAC's Plainville, Ind., location, an attempt was made to determine the time necessary
to locate a failed valve. A valve was caused to fail at a location unknown to maintenance personnel,
who located the failed valve and placed the system back in operation after only 21 minutes had
elapsed. No sewage backups or service interruption occurred during this short time period. A key
75
-------
component in continual operation is an effective alarm system, coupled with constantly available
maintenance personnel.
CONCLUSIONS
Vacuum sewer systems in new and existing communities offer potential cost/effeotiveness
through:
• Lower construction costs from smaller diameter mains installed closer to the ground sur-
face
• Decreased infiltration/inflow
• Reduced water consumption with use of vacuum toilets
• Ease of installation
Vacuum sewer systems are relatively new in the United States. Each system has provided infor-
mation and operating experience that have generally improved subsequent system designs. System
reliability, costs, design, and applicability to a particular site should be evaluated by the design pro-
fessional before selection of a sewage transport system.
Contractors have reported that small-diameter PVC vacuum sewers laid close to the ground
surface have been installed more quickly and at less cost under the following conditions:
• High ground water areas in permeable soil
• Rocky areas
While vacuum sewers contain many of the same advantages as the concurrently developed pres-
sure sewer systems, some apparent differences are worth mentioning. The advantages of vacuum
sewers over pressure systems can be called the three C's—conservation, centralization, contamina-
tion.
• Conservation—with the water-saving feature of the vacuum toilet, water conservation is
possible.
• Centralization—because the motive force of a vacuum system depends on vacuum pumps
operating from a central source, power outages would not affect a vacuum collection
station equipped with a standby power source. It would be impractical, however, to pro-
vide standby power to each pump unit in a pressure sewer system.
• Contamination—vacuum systems are subject to infiltration during pipe leak or break,
which is undesirable and expensive. Pressure systems in the same situation will force con-
taminated sewage into the soil. This feature is especially important in systems serving
marinas, ship facilities, or warm climate systems where piping may be exposed. Leaking
sewer lines in these instances would be a health hazard. Water mains might be laid closer
to vacuum lines as compared to pressure or gravity systems, with a substantially lower
risk of contamination.
76
-------
RECOMMENDATIONS
The consultant must consider, at a minimum, the following items when selecting and designing
a sewage transport system:
• Application in specific terrain
• System reliability
• Operation and maintenance requirements
• System life
• Standby equipment
• Alarm systems
• Emergency operating procedures during partial or total system failure
• Standby power requirements
• Cost analyses
• Simplicity of operation
• Recommendations of manufacturers
All the factors in this section should be evaluated by consulting engineers and appropriate gov-
ernment agencies before the selection and design of a vacuum transport system.
REFERENCES
1 A. LeMarquand, "Sewerage or Drainage of Houses, Towns, or Districts, and Apparatus There-
for," Patent No. 377681, Feb. 7,1888.
2"Water, Our Most Precious Possession," Liljendahl Vacuum Company, Ltd., Stockholm,
Sweden, undated.
3B. C. Burns et al., "Method and Apparatus for Conveying Sewage," Patent No. 3,730,884,
May 1,1973.
^Envirovac Technical Information, Colt Industries, Beloit, Wis., undated.
^Design Criteria Manual, AIRVAC, The Vacuum Sewer Systems, Rochester, Ind., May 1976.
6D. W. Averil and G. W. Heinke, "Vacuum Sewer Systems," report prepared for the Northern
Science Group of the Canadian Department of Indian Affairs and Northern Development, Jan.
1973.
71. A. Cooper and J. W. Rezek, "Vacuum Sewer System Overview," presented at the 49th An-
nual Water Pollution Control Federation Conference, Minneapolis, Minn., Oct. 3-8,1976.
77
-------
8E. P. Skillman, "Characteristics of Vacuum Wastewater Transfer Systems," presented at the
American Society of Mechanical Engineers Conference on Environmental Systems, July 1976.
9A. J. Dell 'isola, Value Engineering in the Construction Industry, Construction Publishing
Co., Inc., New York, 1974.
10J. W. Rezek and LA. Cooper, "Preliminary Report on Vacuum Sewer System at Lake of the
Woods, Va., Volume I: Hydraulic Field Test Program," report prepared for Lake of the Woods Serv-
ice Company and Utilities, Inc., Rezek, Henry, Meisenheimer and Gende, Inc., Libertyville, 111.,
Nov. 1974.
78
-------
Appendix A
VACUUM SEWAGE FLOW CHARACTERISTICS
FLOW REGIMES
Vacuum sewers exhibit characteristic two-phase flow regimes. The air (gaseous phase) and
sewage (liquid phase) occur in segregated, intermittent, or distributed flow, as shown in figure A-l.
Segregated flow includes stratified flow, wavy flow, and annular flow, and can be identified by
a continuity of liquid and gas flows for a long length of pipe.
Intermittent flow is characterized by intermittent segments of liquid and gas, such as in plug or
slug flow.
Distributive flow more nearly approaches a homogeneous fluid than other flow regimes. It
exists only momentarily in vacuum sewer systems when a slug breaks down. Mist and bubble flow
are forms of this regime.
Baker developed a correlation for predicting flow regimes from air-water data, as shown in
Figure A-2.1 Other investigators, however, have reported problems in the accuracy of these predic-
tions.2 Baker's abscissa is defined by a viscosity compensation parameter:
*=(73/a)[/x/(62.3/p,)2]l/3 (1)
which is found with the other parameters in the section on nomenclature.
Because this flow map is valid only under steady state conditions, predictions using this tech-
nique may not be typical, as consistently variable flow rates are the norm rather than the exception
in vacuum sewer systems.3 One author has pointed out that with a nonvarying flow, it may take
weeks to achieve steady state conditions.2
SLIP
In hydraulic tests at Lake of the Woods, Va., vacuum sewer system, Brockmeier4 added vari-
ous volumes of air through a 2.54-cm (1-inch) opening (fig. A-3), after adding 284 litres (75 gallons)
of liquid to the system. When adequate vacuum was available (usually > 25.4 cm, or > 10 inches of
mercury), a rate of 1,135 1/min (40 f3/min) air entered the mains. Skillman5 found an inlet orifice
between 3.18 and 10.16 cm (1.25 and 4 inches) in diameter will allow 1.4 times the rate of air as a
2.54-cm (1-inch) orifice under an initial 50.8-cm (20-inch) vacuum. Using this correlation, 1,590
1/min (56 f3/min) of air would enter through a larger sized orifice.
Applying this correlation to an AIRVAC system that allows about 38 litres (10 gallons) to
enter in 3-5 seconds of the 10-second valve cycle, an air/liquid ratio of 5:1 enters the mains. Under
an average vacuum of 38 cm (15 inches), this ratio becomes 10:1 inside the main. Thus, in order for
10 volumes of air to be transported through the same conduit as 1 volume of liquid, the air must be
79
-------
SEGREGATED FLOW
Stratified
Wavy
Annular
INTERMITTENT FLOW
Slug
DISTRIBUTIVE FLOW
///////////////////^^^^^
Bubble
. .
^
Mist
Figure A-1. Two-phase flow regimes in vacuum sewers.
80
-------
10'
I 1 I I [ |lr<| 1 I I j Mll| 1 I I | I Ml| 7 I I | lllij 1 ' « | "I
I mil i i i I mil i i i 11 ml
10
1.0 10 100 1,000
Figure A-2. Flow pattern region according to Baker.1
10,000
27
*•'
co
0 ° 15
12
>- 9
<
DC
| 6
LL
DC 3
0
30 60 90 120 150 180 210 240 270 300
AIR ADMITTANCE TIME, seconds
Figure A-3. Air-flow rate into vacuum sewer system after sewage discharge for various air admittance times.4
81
-------
flowing faster than the liquid. The gaseous phase seems to slip by the liquid phase. The slip velocity
is the difference in the phase velocities, or
Vs = Vt-Vg (2)
The above analysis agrees with previous Navy research6 that found liquid velocity in the range
of 1.83-3.05 m/s (6-10 ft/s) and gas velocity of 9-15 m/s (30 ft/s). Solids, generally neglected in
these analyses, flowed at 1.22 m/s (4 ft/s).
FRICTION HEADLOSS
Early vacuum sewers may have accounted for friction loss via an equalizing two-phase friction
factor, 0tp. Averill and Heinke7 developed a <£tp based on a homogeneous model, where
-(ap/BzJftp = 0tp[-(dp/dz)ffullpipe] (3)
or the two-phase flow friction factor is some safety factor greater than unity times the full pipe
friction factor.
AIR VAC
AIRVAC presents an even simpler model of headloss by multiplying full pipe flow by a derived
2.75 safety factor. This number is obtained from an assumed 2:1 air to liquid ratio in the pipe. The
liquid, therefore, is flowing at three times the velocity as full pipe flow to deliver the same liquid
flow rate. The Hazen-Williams formula raises the velocity to the 1.85 power, whereas the Darcy-
Weisbach formula squares the velocity. The average of the velocities raised to the respective powers
is 7.5 and 9. Since only one-third of the pipe diameter is said to be wetted, 7.5 + 9/2 x 1/3 = 2.75 is
the average two-phase friction factor applied to the full pipe flow friction factor.
Mechanical Energy Balance
A more elegant and reliable analysis of friction headloss in vacuum systems was developed by
Dukler,8 using a similarity analysis verified by other researchers' experimental data.2
The general equation for pressure gradient in constant slip, two-phase flow is:
dp/dz =- - [(7y)cg + apcs(g/gc)] /(I - ACCS) (4)
The pressure gradient equation for dynamic headloss in vacuum sewers is a function of three
distinct terms.
• Friction term, (r^)cs
• Inclined flow term, apcs(gc/g) where a is valid from +10° to -10°
• Acceleration term, (1 - ACCS)
Friction. The friction term (r^)cs is evaluated by:
) (5)
82
-------
with
Recs = PcsDVns/»ns (6)
f0 = {2 log [Recs/(4.5223 log Recs - 3.8215)] }~2 (7)
pcs = P/(X2AR,) + pg(l - X)2/(l - Rt) (8)
^cs = Vtl + Vsg (9)
Mcs = n/X+/Ug(l-X) (10)
/OS = "(Wo (ID
o(X) = l-(lnX/$) (12)
f = 1.281 + 0.4781 (In X) + 0.444 (In X)2 + 0.094 (In X)3 + 0.00843 (In X)4 (13)
The friction factor, fcs, can be evaluated from the constant slip Reynolds number using the
smooth tube friction factor (C > 150) and a standard Moody Diagram.
If alternative friction factors are desired, a correction can be applied as follows:
l/roi/2 = -2 log [(e/3.7£) + 2.51/(Recsf0i/2)] (14)
The friction term is then a function of viscosity; density; Reynolds number; Euler number;
flowing volume holdup, X; and the in situ volumetric holdup, Rt.
The following volume holdup, X, is the ratio of liquid volumetric flow rate to the total volu-
metric flow, or the ratio of the liquid superficial velocity to the total superficial velocity. The super-
ficial velocity of either phase is calculated by assuming the pipe is occupied by only one phase and
dividing that phase's flow by the pipe's cross-sectional area.
X = Qll(Ql + Qg) = Vsl/(Vsl + Ysg) (15)
The in situ volumetric holdup, R{, is a more difficult concept to grasp. While in homogeneous
distribution flow, Rt = X, intermittent and segregated flow regimes, as seen in vacuum sewers, do
not result in this equality and a separate estimate of R{ is necessary. The in situ volumetric holdup
is a key variable in the analysis of two-phase flow and is termed Rt for liquid and Rg for gas (which
equals 1 - R^). R{ is the fraction of a pipe element that is occupied by the liquid for some pipe
length. RI is then the average over both length and cross-section in slip flow, as opposed to X vary-
ing only with cross-sectional area.
Hughmark9 developed a holdup correlation, as presented by DeGance and Atherton,2 by solv-
ing the following equation for Rt:
F = R,-l+K(l-\) = 0 (16)
where K is a function of 6 .
83
-------
If
G! = 0.642 yns0.5Gf0.1667jD0.0417/ysl0.25 (17)
then
6 = C1/[^(M,-Mf)+Mg]0-1667 (18)
By taking the function derivative
= ! + (!- A)(3.K/35)(3S/3.R,) (19)
Ou, -ng)HRi(Hi -Hg) +Mg]°-1667 (20)
For 6 < 10:
K= -0.16367 + 0.310378 -0.035256 2 +0.0013665 2 (21)
and for 8 > 10:
K = 0.75545 + 0.00035.855 - (0.1436 x 10-4)82 (22)
For the derivative, if 6 < 10:
3K/3S = 0.31037-0.070506 +0.004106 2 (23)
and for 8 > 10:
atf/as = -0.003585 + (0.2872xlO-4)8 (24)
Because
38 ISRt = C^ - ^/[R^, - ng) + pg] °'1667 (25)
and
3F/3JS, = ! + (!- A)(3£/36)(38/3fl,) (26)
then
(27)
Successive iterations are necessary to obtain a more accurate estimate of the holdup correla-
tion, Rj. Successive iterations yielding Rt. and Rj. to two significant figures are satisfactory.
Inclined Flow. The inclined flow term, apcs(gc/g), relates the angle of incline and the constant-
slip flow density. Little, if any, decrease in friction is experienced when dealing with vacuum sewers
having minimum slopes laid in flat terrain. At slopes from 1° to 10° , progressively greater theoretical
effects are experienced. At slopes exceeding 10° inclined or declined, the accuracy of this predictor
diminishes rapidly. Because most lift in gravity sewers is installed at 45° , headless here must be
empirical and is counted solely as elevation loss as frictional distances are small.
84
-------
Acceleration. The acceleration term, 1 - ACCS, is most evident during mist flow regimes, de-
scribed by high values of Re and low values of Rt. As can be seen from the design examples, the
acceleration terms in the range of application to vacuum sewers is negligible.
The headloss from the acceleration term reflects changes in velocity because of slugs breaking
down and falling to a lower elevation trap; and liquid level in the trap building up, finally reaching
the crest of the pipe, and then being hurtled along as the differential pressure acts across a full face
of liquid. The acceleration term loss is a calculation of a change in two-phase kinetic energy.
Depending on initial assumptions, two expressions for acceleration can be used. The first equa-
tion is the simpler of the two, yet equation 29 is generally thought to be more accurate based on
the nature of holdup.
ACCS . GgZ/tegg^l-R,)] (28)
or
ACCS = - ^G,VtlIRt) + (GJV.JRa)(l - R,/Re)/gcp (29)
In design problems, both equations are usually found to have virtually identical effects on
headloss.
NOMENCLATURE
AC Acceleration term, defined by equations 28, 29
Ap Cross-sectional area of conduit, ft2
Cj Parameter, equation 17
D Inside diameter of conduit, ft
Eutp Two-phase Euler number
f Friction factor, one-phase flow
fcs Friction factor, two-phase constant slip
f0 Friction factor, defined by equation 7
F Parameter, equation 16
Fr Froude number, Vns2/gcD
/tp Friction factor, two-phase flow
g Local acceleration, ft/(s)(s)
gc Gravitational constant, 32.174 (Ib.) mass x (ft)/(lb) force (s)(s)
Gg Gas mass flux, lb/(ft2 )(s)
G, Liquid mass flux, lb/(ft2 )(s)
Gt Total superficial mass flux, Ib/(ft2)(s); (Wt/Ap)
i Any given point in a conduit
K Parameter:
if
5<10, K = 0.1637-0.310376 + 0.35255 2 - 0.0013666 3
if
5 > 10, K = 0.75545 - 0.003586 + 0.1436 x !Q-46 2
where
6 =
QI, Qg Volumetric flow rate of liquid or gas
Retp Two-phase Reynolds number, DGt/[Rtiii + (1 - -
£ Universal gas constant, 1,545 x (Ib) force (ft)/(lb/mole)(° R)
85
-------
T
t
V
Inplace gas holdup
Inplace liquid holdup
Temperature
Time
Volume
Velocity of gas, ft/s
Velocity of liquid, ft/s
Slip velocity, ft/s
Superficial liquid velocity, ft/s
Superficial gas velocity, ft/s
Liquid mass flow rate, Ib/hr
Gas mass flow rate, Ib/hr
Total mass flow rate, Ib/hr
Conduit length, ft
Pressure gradient (total pressure gradient), Ib/(ft3)/ft
Greek Letters
a Conduit slope, Sine 9
a(X) Defined in equation 12
6 Parameter, equation 18
e Absolute roughness, equation 14
X Flowing volume-holdup of liquid
| Parameter, defined in equation 13
It Viscosity (Ib) (s)/ft2
p Density, lb/ft3
0 AIRVAC's two-phase friction factor
a Surface tension, dyn/cm2
* Correlating parameter, defined in equation 1
Tf Partial derivative of pressure with respect to Z of frictional contributions
Subscripts
c
cs
CT
DP
f
8
/
ns
RT
Sg
S
t
tp
VP
Cycle
Constant slip
Collection tank
Discharge pump
Frictional
Gas
Liquid
No slip
Reserve tank
Superficial gas
Superficial liquid
Total
Two phase
Vacuum pump
REFERENCES
1O. Baker, "Experiences with Two-Phase Pipelines," presented at the joint meeting of the
Canadian Natural Gas Processing Association and the Natural Gasoline Association of America, Cal-
gary, Canada, Sept. 15,1960.
86
-------
2 A. E. DeGance and R. W. Atherton, "Chemical Engineering Aspects of Two-Phase Flow,"
Chem. Eng., 8-part series, 1970-71.
3 J. W. Rezek and I. A. Cooper, "Preliminary Report on Vacuum Sewer System at Lake of the
Woods, Va., Volume I: Hydraulic Field Test Program," report prepared for Lake of the Woods Serv-
ice Company and Utilities, Inc., Rezek, Henry, Meisenheimer and Gende, Inc., Libertyville, 111., Nov.
1974.
4 C. R. Brockmeier, C. R. Brockmeier Consulting Engineers, Los Angeles, Calif., unpublished
test data, 1975.
5E. P. Skillman, Civil Engineering Laboratory, Naval Construction Battalion Center, Port
Hueneme, Calif., unpublished test data, 1975.
6E. P. Skillman, "Characteristics of Vacuum Wastewater Transfer Systems," presented at the
American Society of Mechanical Engineers Conference on Environmental Systems, July 1976.
7D. W. Averil and G. W. Heinke, "Vacuum Sewer Systems," report prepared for the Northern
Science Group of the Canadian Department of Indian Affairs and Northern Development, Jan.
1973.
8A. E. Dukler et al., "Frictional Pressure Drop in Two-Phase Flow: (A) Comparison of Exist-
ing Correlations for Pressure Loss and Holdup; (B) An Approach Through Similarity Analysis," Am.
Inst. Chem. Eng., J., 10, 38-43, (A); 44-51, (B).
9G. A. Hughmark, "Holdup in Gas Liquid Flow," Chem. Eng. Prog., 58, 62-65, 1962.
87
-------
Appendix B
DESIGN EXAMPLE
This design example is based on serving a small rural town located in a flat sandy area with
consistently high ground water (fig. B-l). Lines are initially laid out to serve the town from a central
location, one line east (line A) and the other west (line B). Line sizes are chosen based on serving
the peak period. Two homes' gravity lateral will use one valve that will connect to the vacuum main.
Traps are placed at a maximum of every 300 feet to reform slugs or where line elevation drops 1
foot below the invert after the previous trap. Because the trap drops 6 inches before being raised 1.5
feet, at a 45° angle, each trap requires 1.5 feet of lift to gain 1 foot of elevation. This assumption is
conservative, based on Electrolux's1 data indicating only 1/2-metre headloss is experienced by each
1 metre static lift. Their recovery of 1/2 metre may be caused by a partial mixing of sewage and air,
lowering average density and, therefore, energy required to lift the liquid.
The lines are sloped from trap to trap, based on flowing 0.61 m/s (2 ft/s) at 0.7 full pipe flow.
The velocity chosen allows for suspension of solids, while the 0.7 full pipe flow at maximum flow
periods still allows sufficient void volume across the top of the liquid to allow for transfer of air.
The transfer of air is the mechanism for reestablishing the local vacuum gradient along the pipeline.
After traps are laid out, the line is arbitrarily divided into segments. Headloss for each segment
is calculated, first via the mechanical energy balance, then by AIRVAC's method as a checking pro-
cedure. Significant variations should be investigated.
Collection station design considerations for size of collection tank, reserve vacuum tank,
vacuum pumps, and sewage force main pumps are presented. Formulas derived by AIRVAC2 appear
workable and are used.
BASE CONDITIONS
Line length = 1,000 ft (2 lines)
40 homes, 3.5 persons per home, 75 gallons per capita per day
Design at peak flow, four times average
Divide line into four segments
Assumes entire system under 1/2 atm vacuum (15 inches)
Segment
A
B
C
D
Length
250
250
250
250
Homes
40
30
20
10
Maximum Q
29
22
15
8
Pipe size
4 inches
3 inches
3 inches
3 inches
Slope
0.0030
.0055
.0055
.0055
PIPING PROFILE
Traps are every 300 feet or drop of 1-foot elevation in level terrain. Starting at end of Segment
D(10 + 00):
Trap 1 at station 10 + 00 = 1 ft/0.0055 ft/ft = station 8 + 18
88
-------
River
100 feet
''
Dj
q
°l
•
^
JL7 "
Cleanout
d
D-
D1
pr
D,
q
LineB-^
D1
D
Cc
st;
)llect
ition
Cleanout
Cf
D
ion
"5
i
°\
a
1
Cleanout
to
a
tf
D
f°
P
tf
D
1,000 feet
1°
P
TJ
a
Collection station
250 feet 250 feet
CROSS SECTION
250 feet
250 feet
- Figure B-1. Vacuum sewer system design example.
Trap 2 at station 8 + 18 = 1 ft/0.0055 ft/ft = station 6 + 36
Trap 3 at station 6 + 36 = 1 ft/0.0055 ft/ft = station 4 + 54
Trap 4 at station 4 + 54 = 1 ft/0.0055 ft/ft = station 2 + 72
89
-------
Trap 5 at (slope changes at 2 + 50 to 0.0030) drop in (2 + 72 - 2 + 50) = 22 ft; 0.0055 x 22
0.12; 1 - 0.12 = 0.88
2 + 50 - 0.88/0.0030 = 250 - 293.3 = 43 feet
Therefore, trap 4 is last trap before collection tank.
Elevation loss per trap = 1.5 ft
4 traps at 1.5 ft = 6.0
Final trap lifts 1.0 ft =1.0
Total static /ILS
= 7.0ft
FLOW CONDITIONS
Parameter
D, ft
Ap.ft2
7,°F
P. Ib/ft2
a. sin 0
Q/, gal/min
Qg, gal/min
Vsi, ft/s
V ft/s
v
P/Jb/ft3
Pg. Ib/ft3
M/, cp
M cp
M/,, Ib/hr
W Ib/hr
X
Gf, Ib/ft2 /s
Gfff Ib/ft2/s
G,, Ib/ft2/s
A
0.33
.0873
68.0
1,058.4
0.003
29.2
291.7
.76
7.61
8.37
62.4
.0752
1.008
.01794
14,612
176.0
.0907
48.04
.57
47.47
B
0.25
.0491
68.0
1,058.4
.0055
21.9
218.8
.99
9.94
10.93
62.4
.0752
1.008
.01794
10,959
132.0
.0907
62.75
.75
62.00
C
0.25
.0491
68.0
1,058.4
.0055
14.6
145.8
.66
6.63
7.29
62.4
.0752
1.008
.01794
7,306
88.0
.0907
41.83
.50
41.33
D
0.25
.0491
68.0
1,058.4
.0055
7.3
72.9
.33
3.31
3.64
62.4
.0752
1.008
.01794
3,653
44.0
.0907
20.92
.25
20.67
Segment A
RI (in place holdup):
Assume Rt = 1.0, then
G! = 0.642 Vns0-5Gf°-1667I>0-0417/ysl0-25
Cj = 3.6219
6 = COi-/)*]0-1667-3.6171
90
-------
6
K
K
< 10
= -0.16367+0.310376 - 0.035256 2 + 0.0013665 3
= 0.5624
= 0.31037 - 0.070506 + 0.00416 2 = 0.1090
F = Rt - 1 + K(l - A) = 1 - 1 + 0.5624 (1 - 0.0907) = 0.5114
b8/bR} = CjCjU/-Mg)/[-R/(M;-A«^)+Mg]0-1667 = 3.374
= 1 + (1 -X)(3X/36)(36/3B/) = 1.3344
.) = 1 - 0.5114/(1.3344) = 0.6168
Successive iterations of the above process are repeated using R,-+, until R/. and Rt. are
sufficiently close at two significant figures.
Iteration
1
2
3
4
5
6
7
8
9
10
11
*/
1.0
.6168
.8339
.6916
.8337
.7367
.8028
.7578
.7884
.7676
.7817
F
0.5114
.1559
.4025
.2244
.3557
.2650
.3270
.2853
.3136
.2944
dF/dR,
1.3344
1.3428
1.2053
1.3492
1.3509
.3478
.3500
.3486
.3494
.3488
Iterations 10 and 11 are sufficiently close to use a two-significant figure R{ correlation number
of 0.77.
The dynamic pressure loss for segment A can now be calculated as follows:
(bp/bz)ftp = -[(iy)C8 + apcs(g/gc)] /(I - ACCS)
where:
PCS = /°/(X^/-R/) + Po(l ~ ^)2/(l - RI)
pcs = 62.4 (0.09072/0.77) + 0.0752 (1 - 0.0907)2/(1 - 0.77)
'cs
= 0.9370
+ Hg (1 - X) = 1-008 (0.0907) + 0.0752 (1 - 0.0907) = MCS = 0.1598
Re
ucs rcs-^ ' ns"*ns
Recs = 1,488 (0.9370)(0.33)(8.37)/(0.1590) = 2.41 x 104
91
-------
f0 = {2 log [Recs/(4.5223 log Recs - 3.8215)] }-2
= {2 log [2.4 x 104/(4.5223 log 2.41 x 104 - 3.8215)] }-2 = 0.0248
a(X) = l-{ln
£ = 1.281 + 0.4781 (In A) + 0.444 (In X)2 + 0.094 (In X)3 + 0.00843 (In X)4
| = 1.6713
a(X) = 1 - In 0.0907/1.6713 = 2.4361
/cs = a(\)f0 = 2.4361(0.0248) = 0.0603
= (0.0603)(8.37)2(0.9370)/2(32.174)(0.33) = 0.1864
ACCS =
ACCS = (47.47)0.76/0.77 + (0.57)(7.61)/(0.23)(1 - 0.77/0.23)/(32.174 x 10584)
ACCS = 0.000076
However, in this range a more reflective acceleration term would be :
ACCS = Gg2/PggcPRg
ACCS = (0.57)2/[(0.0752)(32.174)(1058.4)(1 -0.77)]
AC0
-------
3F/3R, =1.3387
R;. = 0.59, after iterations, use R[ = 0.75
pcs = 0.9332
A,cs =0.1598
Recs = 2.31 x 104
f0 = 0.0250
| = 1.6713
a(X) = 2.4361
fcs = 0.0607
(rf)cs = 0.0609(10.93)2 (0.9332)/2(32.174)(0.25) = 0.4220
ACCS = Gg2/pggcpRg = (0.75)2/(0.752)(32.174)(1058.4)(0.25) = 0.0009
= - 0.4172 Ib/ft2/ft = - 6.69 ft/1,000 ft
Segment C
G! = 3.3823
6 = 3.3778
K = 0.5352
= 0.1190
= 0.4867
=3.3442
dF/dRt =1.3619
Rj. = 0.64, after iterations, use 0.78
PCS
^cs
Recs
fo
= 0.7748
= 0.1598
= 1.46 x 104
= 0.0280
93
-------
$ = 1.6713
a(X) = 2.4361
/c, = 0.0682
(7y)CB = 0.2196
ACCS =9.9xlO-5
9p/9z = - 0.2142 Ib/ft2/ft = - 3.44 ft/1,000 ft
Segment D
G! = 0.642(3.64)°-5(20.92)0-1667(0.25)°-0417/(0.33)0-25
G! = 2.5321
6 = 2.5288
K = 0.4179
9#/95 = 0.1583
F = 0.3800
95/9.R; =2.5036
9F/9E, =1.3604
E, + = 0.7207, after iterations use 0.83
pcs = 0.9842
Mes = 0.1598
Recs = 8.33 x 103
f0 = 0.0324
| = 1.6713
o(X) = 2.4361
fcs = 0.789
(T^,, = 0.0640
ACCS = 0.0003
9p/9z = 0.0586 Ib/ft2/ft = 0.94 ft/1,000 ft
94
-------
DYNAMIC HEADLOSS {/?LD)
By AIR VAC calculation:
Segment
A
B
C
D
Maximum Q
29.2
21.9
13.6
7.3
Pipe
size
4-in
3-in
3-in
3-in
Pipe
''LD
0.62
1.86
1.00
0.39
Two-phase
factor
x 2.75
x 2.75
x 2.75
x 2.75
Two-phase
/7LD, ft/1, 000 ft
1.71
5.12
2.75
1.07
By mechanical energy balance:
Segment
A
B
C
D
/?LD, 1,000ft 1,000fta
2.94 x 0.280 = 0.82
6.69x0.310=2.05
3.44x0.280= .96
0.94x0.280= .26
Total Dynamic/?|_o 4.09 ft
a250 ft per segment + 30 ft for each trap.
By AIRVAC (hLD = 2.75 x full pipe bLD):
Segment
A
B
C
D
''LD.
1.71
5.12
2.75
1.07
1,
x
x
X
X
000ft
0.280
0.310
0.280
0.280
1
,000
fta
= 0.48
=
=
=
Total dynamic /7Lo
1.59
0.77
0.30
3.14
ft
Total Headless (/?LT)
3250 ft per segment + 30 ft for each trap.
Total fcLT = /iLs - ftLD + ftLy, not to exceed 18 ft
(static) (dynamic) (valve)
(ftLV = 5 f t reserved for valve operation)
(30)
Mechanical Energy Balance
ftLT = 7.0 + 4.09 + 5.0 = 16.09 ft
AIRVAC/iLT = 7.0 + 3.14 + 5.0 = 15.14 ft
95
-------
Therefore, the vacuum piping system is functional because both the AIR VAC and mechanical
energy balance equations yield a headless of less than 18.0 feet.
DISCHARGE PUMP
Discharge pumps shall be sized to handle 120 percent of the design peak of the maximum sew-
age flow with the largest pump out of service and with a minimum size of 80 gal/min each. Assume
each pump would be the same size. Treatment plant sizing should reflect discharge pump sizing
requirements.
QDP = 1.2 x Qmax = 1.2 x 29.1 = 69.8 (31)
Thus, 69.8 gal/min for two pumps would require that each pump be sized at 80 gal/min each.
AIR VAC suggests 25 feet of head should be added to the design point to account for collection
tank vacuum. Sufficient net positive suction head must also be available, as described earlier.
Vacuum Pump Capacity
Vacuum pump sizing allows for withdrawal of the volume of air in the mains, based on peak
flow conditions with a 100 percent safety factor with the largest unit out of service. An allowance
for valve sensor leakage in the AIR VAC system of 0.25 f3/min per valve must be added.
Vacuum pump sizing should also consider the length of pump running time. Research in this
area addressed by Skillman3 showed optimum performance is affected by both vacuum pump size
and vacuum reserve. A compromise should be reached between a large-sized pump cycling frequent-
ly and a minimum-sized pump that may not be capable of maintaining a satisfactory system vacuum
during high-flow periods. A reasonable design equation is presented below:
QVP = 2(Q, + Qg) x 1 ft3/7.48 gal + 0.25 x No. valves (32)
Because two lines (lines A and B) enter the station, the sum of the air flows must be
considered:
QVP = 2 x 2(291.7 + 29.2) x 1/7.48 + 0.25 x 40 = 181.6 ft3/min
Because 181.6 ft3/min is required with the largest pump out of service, use three pumps rated
at 91 ft3/min at 45.7 cm (18 inches mercury). Pump curves should be analyzed for shutoff vacuum
and free-flow conditions.
Collection Tank Volume
While most pump station design manuals suggest the minimum time between cycles should be
as low as 10 minutes,4 AIR VAC suggests 30 minutes at half the average daily flow.
The minimum time between cycles, or 30 minutes, is the sum of the filling time plus the
pumping time, or
cycle time = filling time + pumping time (33)
Because the collection tank's operating volume is a maximum of 65 percent of the total collec-
tion tank volume, the following calculation shows the required size, with a minimum of 400
gallons:
96
-------
yCT = tc/[0.65(l/Qmin + 1/QDP - Qmin)] (34)
yCT = 30/[0.65(l/7.3 - 1/80 - 7.3)] = 307.6 gal
Therefore, use a 400-gallon tank.
Reserve Tank Volume
AIRVAC calculates reserve tank volume from two equations. Their first equation relates a
total volume necessary to bring the vacuum up from 16 to 20 inches in t minutes, with t usually 1
or 1.5, and the previously calculated vacuum pump capacity,
Vt = 3QVP x t (35)
or Vt = 3x181.6x1 = 544.80
The second equation determines the reserve tank volume, with a minimum size of 400 gallons.
For this equation it is assumed that one-third of the piping is occupied by the gaseous phase:
VRT = vt ~ !/3 ^Piping - 0.35(FCT) (36)
VRT = 544.80 - 1/3(877.4) - 0.35(400) = 112.3 gal
The minimum-sized tank recommended is 400 gallons; the size of the reserve tank in this
example, therefore, is 400 gallons.
Auxiliary Power
In order to size standby generator sets, an analysis of the continuity of electrical service should
be undertaken. If outage times are significant and frequent, full load power should be recom-
mended. If infrequent, short outages occur, generator set sizing should be sized to operate one vac-
uum pump and one discharge pump. If local standards dictate more conservative requirements, they
should be followed.
REFERENCES
1"Electrolux Vacu-Flow System for Nash Sewerage," report prepared for the Borough of Nash,
England, by Electrolux Corporation, Stockholm, Sweden, Feb. 1976.
2Design Criteria Manual, AIRVAC, The Vacuum Sewer Systems, Rochester, Ind., May 1976.
3E. P. Skillman, "Characteristics of Vacuum Wastewater Transfer Systems," presented at the
American Society of Mechanical Engineers Conference on Environmental Systems, July 1976.
4"Lift Stations Engineering Manual," Clow Corporation, Waste Treatment Division, Florence,
Ky., undated.
97
-------
METRIC CONVERSION TABLES
Description
Length
Area
Volume
Mass
Force
Moment or
torque
Flow (volumetric)
Description
Precipitation,
run-off,
evaporation
Flow
Discharges or
abstractions,
yields
Usage of water
Unit
meter
kilometer
millimeter
micrometer or
micron
square meter
square kilometer
square millimeter
hectare
cubic meter
litre
kilogram
gram
milligram
tonne
newton
newton meter
cubic meter
per second
liter per second
Unit
millimeter
cubic meter
per second
liter per second
cubic meter
per day
cubic meter
per year
liter per person
per day
Recommended tlmts
Symbol Comments
m Basil SI unit
km
mm
urn or i*
m2
km'
mm2
ha The hectare 00,000
m2} is a recognized
multiple unit and will
remain in interna-
tional use
m3
1
kg Basic SI unit
g
mg
t 1 tonne = 1,000 kg
N The newton is that
force that produces
an acceleration of
1 m/s2 m a mass
of 1 kg
N m The meter is mea
sured perpendicular
to the line of action
of the force N
Not a joule
m3/s
l/s
Application of Units
Symbol Comments
mm For meteorological
purposes, it may be
convenient to meas
sure precipitation m
terms of mus/unit
area (kg/m2)
1 mm of ram =
1 kg/m2
m3/s
t/s
m3/d 1 l/s = 864m3/d
m3/year
I/person/
day
Recommended (/m/s
Customary
Equivalents*
3937m = 3281 ft =
1 094 yd
06214 mi
003937 in
3937X 105m= 1 X 104 A
10 76 sq ft = 1 196sqyd
03861 sqmi 247 1 acres
0 001550 sqm
2471 acres
3531 cuft = 1 308cuyd
1 057 qt = 02642 gal =
08107 X 104 acre ft
2 205 Ib
003527oz= t543gr
001 543 gr
09842 ton (long) -
1 102 ton (short)
0 2248 Ib
= 7 233 poundais
07375lb ft
23 73 poundal ft
1 5 850 gptn -
2,119cfm
1585gpm
Description
Velocity
linear
angular
Viscosity
Pressure or
stress
Temperature
Work, energy.
quantity of heat
Power
Unit
meter per
second
millimeter
per second
kilometers
per second
radians per
second
pascal second
centipoise
newton per
square meter
or pascal
kilo newton per
square meter
or kilopascal
bar
Celsius (centigrade)
Kelvin tabs )
(oule
kiloioule
watt
kilowatt
loule per second
Symbol Comments
m/s
mrn/s
km/s
rad/s
Pas
2
N/m2
or
Pa
kN/m2
or
kPa
bar
'C
°K
J 1 joule = 1 N m
where meters are
measured along
the line of action
of force N.
kJ
W 1 watt * 1 J/s
kW
J/s
Customary
Equivalents*
3281 tps
0003281 fps
2,237 mph
9 549 rpm
0 6722 poundal(s)/sq ft
1 450 X 10 7 Reyn (p)
00001450 Ib/sq in
0 14507 Ib/sq in
14 50 Ib/sq in
(°F-32)/1 8
°C + 273 2
2778X 10 7
kwhr -
3725X 10 7
hp hi - 0 7376
fHb = 9478X
10^ Btu
2778X 10^ kwhr
44 25 ft Ibs/min
1 341 hp
3412Btu/hr
Application of Unit*
Customary
Equivalents"
3S31cfi
1S85gpm
0 1835gpm
264 2 gal/year
02642gcpd
Description
Density
Concentration
BOD loading
Hydraulic load
per unit area,
e g., filtration
rates
Air supply
Optical units
Unit
kilogram per
cubic meter
milligram per
liter (water)
kilogram per
cubic meter
per day
cubic meter
per square meter
per day
cubic meter or
liter of free air
per second
lumen per
square meter
Symbol Comments
kg/m3 The density of water
under standard
conditions is 1,000
kg/m3 or l,000g/l
or 1 g/ml
mg/l
kg/m3/d
m3/m2/d If this is converted
to a velocity, it
should be expressed
in mm/s (1m m/s =
864m3/m2/day)
m3/s
l/s
lumen/m2
Customary
Equivalents*
0 06242 Ib/cu ft
1 ppm
006242 Ib/cu ft/day
3281 cu ft/sq ft/day
0 09294 ft candle/sq ft
•Miles are U S statute, qt ind gal are U S liquid, and 01 and Ib are avoirdupois
*US GOVERNMENT PRINTING OFFICE 1977—757-140/6603
------- |