STATUS OF PRESSURE SEWER TECHNOLOGY
BY
JAMES F, KREISSL
the.
Age.nc.ij
Technology
Pe6-tgn Seminasi
Small flow.
Sanitary Engineer, Urban Systems Management Section, Systems & Engineering
filiation Branch, Wastewater Research Division, EPA, MERL Cincinnati, Ohio
-------
STATUS OF PRESSURE SEWER TECHNOLOGY
J. F. Kreissl
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 solutions described in
the literature and the mid-19th century, when municipal sewerage enjoyed a
rebirth in Europe and the United States, a development that has continued to
the present. Between the time the first large American system was installed
in Brooklyn, New York, (1857) 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 presently stands at about 71 percent.
A recent review of nearly 300 facilities'plans for rural communities in
the United States produced the relationship shown in Figure 1 (5). Monthly
charges much above $20 are considered excessive in rural areas where median
incomes are generally significantly lower than in urban areas. Since most on-
site wastewater disposal systems would cost significantly less than this on a
monthly basis, the on-site approach has been generally employed in these areas.
Difficulties have arisen in areas where conventional on-site systems have failed
due to unfavorable soil conditions. Typically, the result from this condition
has almost invariably been a recommendation to sewer the community. Implementation
of this recommendation has been dependent on the financial status of the community,
availability of Federal grants and public attitude. Without enjoining the already
burdensome discussions on the merits and demerits of the grant programs and
centralized collection and treatment systems, it suffices 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 1 clearly
illustrates the relationship between cost and population density, which is
primarily explained by the greater length of sewer per contributor, greater
problems with grade resulting in more lift stations or excessively deep sewers
(see Figure 2), and regulations which limit the smallest sewer pipe diameter.
Essentially due to the above economics, the primary form of wastewater
treatment and disposal in rural areas has been the septic tank-soil absorption
system (ST-SAS). Prior to the passage of the Norris-Rayburn Act during the
depression of the 1930s few rural areas had the electricity necessary to
provide for water carriage of human wastes. However, as the rural electrification
program took effect in the following decade, two major effects resulted. First,
the rural population obtained electricity and upgraded standards of living.
-------
40
30
w
0
o20
0
U
10
2 4 6 8 10 12
POPULATION DENSITY-PERSONS/ACRE
FIGURE 1. MONTHLY COST OF GRAVITY SEWERS
14
-------
2 8-32
24-28
20-24
LU
LU
. 16-20
I-
14-16
0
1’ 1
10-12
8-10
<8
0 5 10 15 20 25 30 35 40 45 50 55 60
CONSTRUCTION COST (DOLLARS PER LINEAR FOOT)
FIGURE 2. Co ST OF SEWER CONSTRUCTION
-------
4
including pressurized water supplies and water carriage of wastes. Second,
urban dwellers emigrated to previously rural environs where they could enjoy
the best of both societies. The disposal of wastewaters generated in these
unsewered areas was best accomplished by ST-SAS systems, as shown in Figure
3 (7). Developers of these areas also found advantages in these systems
because costs were directly related to the dwelling and offered a minimum
of post-construction responsibility. Unfortunately, many of these systems
have failed due to faulty design and construction, unsuitable soil conditions,
and owner negligence. However, present estimates indicate that 15 to 20
million septic tank systems still exist in the United States, serving more than
one-quarter of the population.
Unfortunately, many situations have come about in recent years which
defy solution by using either of the traditional alternatives described
above, and the results of attempting to apply either technology in these
situations have been unsatisfactory to all involved. For example, building
moratoria have been imposed which prevent the development of highly desirable
land parcels, conventional sewers have been installed at tremendous cost to
the homeowners serviced, and ST-SAS have been constructed which cannot function
properly, therefore, contaminating the very environment which made the site
so attractive originally. The problem has become so acute that the 92nd Congress
specifically directed the Environmental Protection Agency (EPA) in Section 104(q)(1)
of PL 92-500 to “.... conduct a comprehensive program of research and investi-
gation. . . .eliminating pollution from sewage in rural and other areas where
collection of sewage in conventional, coninunity—wide sewage collection systems
is impractical, uneconomical or otherwise infeasible, or where soil conditions
or other factors preclude the use of septic tank and drainage field systems.”
-------
OOF TERMINAL
LAID ON
COMPAC
EARTH
PLUMBING FIXTURES TO BE
PROPERLY TRAPPED AND
VENTED
ABSORPTION
FIELD
GE
HOUSE
DRAIN
HOUSE
SEPTIC
TANK
NON
PER F
TILE
TILE
LINES
FIGURE 3. TYPICAL ON-SITE SYSTEM
-------
6
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 did not
emerge until the latter part of the 1960s. 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 (8). This study was performed by the American Society of
Civil Engineers for the Federal Water Pollution Control Administration and
Included numerous research studies on such topics as household wastewater
generation patterns, critical velocities of flow, alternate system layouts,
and prototype grinder-pump 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 grinder-pump
pressure sewer systems at Albany, New York; Phoenixville, Pennsylvania; and
Grandview Lake, Indiana. Other coninunities have also utilized 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 and design criteria
for pressure sewer systems.
A number of advantages of pressure sewers have been presented in the
literature (1—3, 6, 8—12). These benefits are primarily related to installation
costs and inherent system characteristics. Since these systems all employ
small-diameter plastic pipes buried just below the frost penetration depth,
their installation costs can be quite low when compared to conventional gravity
systems in low density areas. Other site conditions which enhance this cost
differential include hilly terrain, rock out-cropping, and high water tables.
Since the 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
which occurs in most gravity systems.
As with any technology, certain disadvantages also exist. The disadvan-
tages of pressure sewers include high operation and maintenance (O/M) costs
relating 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 wastewaters.
It may, therefore, require a higher level of treatment to satisfy effluent
requirements. A wastewater will also be devoid of oxygen.
-------
7
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 4. The user inputs to the pressure main follow a generally direct
route to treatment facility or to a gravity sewer, depending on the
application. The primary purpose of this type of design is to minimize
sewage retention time in the sewer.
The two major types of pressure sewer systems are the grinder-pump
(GP) system and the septic tank effluent pumping (STEP) system. These
are depicted in Figures 5 and 6. From these figures it is obvious that
the major differences between these 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.
-------
WATER MAIN
FIGURE 4. PRESSURE SEWER
VS. WATER MAIN
PRESSURE SEWER
00
-------
EXIST I N G
GRAVITY
SEWAGE
PIPING
SEWER
PVC PIPING
DRAINAGE FIELD
STORAGE
TANK
EXISTING
SEPTIC TANK
OVERFLOW LEVEL SENSOR
ON-OFF LEVEL SENSOR
FIGURE 5. TYPICAL PUMP-GRINDER INSTALLATION
-------
JUNCTION BOX AND
HIGH
ALARM
2” PLASTIC PIPE I OR ELIC.
PVC PLASTIC MAIN
1—1/4” PLASTIC
EXISTING OR NEW
SEPTIC TANK
BALL OR GATE VALV
E
24” CONC. PIPE WITH FLOOR & LID
1/3 HP SUMP PUMP
FIGURE 6. TYPICAL STEP SYSTEM
-------
11
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 Clift (9) in 1968, but the
system described did not employ the techniques and materials which are
now considered standard design practice. In order to serve 42 customers
in low-lying areas of Radcliff, Kentucky, it would have been necessary
to finance S 3 ,170 per connection, while the use of the prototype pressure
sewer system cc st only $1,346 per connection. This prototype design
employed pneumatic ejector units at each connection which discharged in
to a 3-inch (7.6 cm) cast iron lateral and a 4-inch (10.2 cm) cast iron
main which emptied into a gravity sewer. Even though mechanical and
electrical problems were encountered which eventually caused abandonment
of the system, Clift reported that during the first three years of operation
no odors or blockage of pressure lines occurred (9). However, severe
corrosion was encountered and found to be the primary cause of the electrical
and mechanical component shortcomings.
Clift also performed preliminary estimates on similar prototype
pressure sewers for two other locations. In one case, 120 lots of 280
lots around a lake were considered well-suited to sewering with conventional
gravity sewers, while the more inaccessible lots were perceived to be
served with pressure sewerage. In the second case a similar hybrid design
approach was estimated to save 5.5 percent in capital cost (9).
The most highly-instrumented study on pressure sewers was performed on
a group of twelve homes in Albany, New York (2, 6, 10). Each dwelling was
equipped with a comercial grinder-pump (GP) and connected by laterals to
a pressure main which emptied into a gravity sewer, as shown in Figure 7.
The system operated well after the original prototype GP units were replaced
with improved models. The pressure main had been oversized to allow for
all units operating simultaneously. Subsequent accumulations of grease
and fibrous materials reduced some pipe cross-sectional areas by as much as
40 percent. Valuable information was reported for design and construction
methods and on the operational characteristics and maintenance requirements
of the GP units. Monthly power costs of 10 to 27 /home were incurred based
on a rate of 2.3 t/kwh. Wastewater from the pressure sewer was characterized
and found to be more concentrated than normal municipal wastes, ostensibly
due to the lack of sewer infiltration.
Another relatively short-term (6 months) study of a pressure sewer
system with grinder—pumps was made at Phoenixville, Pennsylvania (11).
This system, as shown in Figure 8, was 2800 feet (854m) long and discharged
into a gravity sewer more than 60 feet (l8.3m) above the farthest GP
location. Another unique feature of this system was the inclu:ion of
multiple-family dwellings serviced by a single GP. Data reported on
construction costs were excellent. Some indirect evidence of pipe cross-sectional
-------
8 7 6 2 1
-- - _CQ NTROL BOX
M.H.
FIGURE 7. ALBANY SYSTEM
-------
EXISTING
GRAVITY
SEWER
1
U.’
140
3000
2 APIS.
SINGLE HOME
r
SINGLE
LOCATION OF PUMP STORAGE GRINDERS •
1 APARTMENTS
ROUTE 113
APTS
MANHOLE
ROAD SURFACE
110
0 500 1000 1500 2000 2500
FIGURE 8. PRESSURE SEWER SYSTEM - PHOENIXVILLE
-------
14
area reductions, similar to the Albany, was also noted. The GP units
employed were similar to the Albany units, and their operation resulted
In a monthly power cost from 11 to 25t on a per capita basis.
The Grandview Lake, Indiana, pressure sewer system (12, 13, 32, 33,
47) was much larger in size (it served 93 homes). The need for this
system was related to an engineering estimate for conventional sewerage
of $3,000/lot or $10,000/existing home due to unfavorable terrain and the
resultant 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 m) of pressure main, six automatic and seven manual air relief
valves, and treatment by stabilization pond with effluent land spreading
were en loyed in the Grandview system, as illustrated in Figure 9. Grease
problems plagued the system by causing faulty operation of automatic air
release valves and by promoting deposits on flow measuring devices at
the plant. The installed cost of pressurization equipment and ancillary
on-site components varied from $1000 to $1500 per home. A contingency
provision for potential on-lot overflows during equipment or electrical
outages was included In the system design. Existing soil adsorption
systems were used whenever possible. Where these were not available, a
small (2-day capacity) gravel-filled absorption bed was provided. Generally,
one—inch (2.5 cm) service connections were employed 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, 31, 35, 38, 47, 52, 53). Bowles (1)
and Cochran (31) describe an installation at Horseshoe Bay on Lake Lyndon
Baines Johnson, Texas. As many as 4,000 connections are planned for this
development with about 106,000 feet (32,300 m) of a 2 to 12 inch (5.1 to
30.5 cm) pressure main. As about 200 GP units were in operation (31).
Equipment problems relating to installation and design have been experienced,
but corrections have been made, and the system is now functioning satisfactorily.
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 Venture on Lake Travis.
This GP installation also suffered from initial problems resulting from
construction activity, but was functioning in an acceptable manner in 1974.
Two more large GP systems and two smaller septic tank effluent pumping
systems have been approved in the State of Texas (31).
Gray has reported on the circumstances which led to the design and
construction of a GP pressure sewer system at Weatherby Lake, Missouri (52).
The system presently serves 330 homes and is expected to ultimately 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 consists of 309 GP units, 35,000 feet (10.7 km) of
pressure main and 37,100 feet (11.3 km) of service lines (PVC SDR-26 with
gasketed joints), 42 air relief valves, and 24 flushing and cleanout
-------
•TArC OF INO1ANA
LOCATION MAP
LA(00N CELLS
FIGURE 9. GRANDVIEW LAKE-A SEWAGE RESEARCH
AND DEMONSTRATION PROJECT
LANC
INDI414 jl
PVC PIPE (SIZ AS SHOWN)
-------
16
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, New York; Clifton Park, New York; and Kinnelon, New Jersey (10,
50). A GP pressure sewer project on Madeline Island, Wisconsin, is being
built to serve a recreational development (38).
The most noteworthy septic tank effluent pumping-pressure sewer
installations are located in Florida and Idaho. General Development Utilities,
Inc., of Miami, Flordia, has installed two large systems; one in Port
Charlotte and the other in Port St. Lucie. The Port St. Lucie pressure
sewer (125 homes), buried at a depth of 2 feet (0.61 m), discharges into
a gravity sewer, while the smaller (26 homes) one at Port Charlotte discharges
into an extended aeration treatment plant. The Port Charlotte system is the
oldest, having been in operation since August of 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 cu m) septic tanks
which pretreat the wastewater (16).
Two separate pressure sewer installations located at Coolin and Kalispell
Bay on 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 employed 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 (27). Although some initial problems resulted from improper
impellers being 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 septic tank effluent
pumpings systems for two Indiana conuiiunities (14). Additional pressure
systems employing septic tank effluent pumping (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 centri-
fugal pumps to pressurize raw sewage directly from the source. One system
services houseboats ( ‘5OO), while the other services a private housing
development (“150 homes). These systems are operating without excessive
O/M requirements, despite the higher potential for 0/M with this design (46,
55).
Two major manufacturers of pressurization equipment have supplied
information on present and future installations of pressure sewer systems
(35, 47). The States listed in Table 1 represent those that have approved
at least one project which is either being designed, constructed or operated,
according to these sources.
-------
17
TABLE 1
PRESSURE SEWER PROJECT LOCATIONS
Arkansas Kentucky North Carolina Vermont
California Michigan Ohio Virginia
Delaware Mississippi Oregon Washington
Florida Missouri Pennsylvania West Virginia
Idaho Nebraska South Carolina Wisconsin
Illinois New Jersey South Dakota
Indiana New York Texas
TABLE 2
HOUSEHOLD WASTEWATER CHARACTERIZATION
HOUSEHOLD WASTES (19) TYPICAL MUNICIPAL (21)
PARAMETER (UNITS) W.O. Grinder W. Grinder Medium Strength
BOO 5 , mg/i 415 465 200
TSS, mg/i 296 394 200
VSS, mg/i 222 309 150
TKN, mg/i 51 52 40
NH 3 -N, mg/i ii 10 25
TP, mg/i 33 32 10
Grease, mg/i 123 129 100
-------
18
SYSTEM DESCRI PTION
A pressure sewer system consists of two major elements: (1) the
on—site or pressurization facility; and (2) the primary conduit or
pressurized sewer main. Probably the widest divergence of opinion
exists on the proper design and equipment selection for the
pressurization facility. Opinion varies both because 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 the household wastes are collected in the building
drain and conveyed therein to the pretreatment and/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 sewer. The two major alternative
systems, which are illustrated in Figures 5 and 6 employ a pressuriza-
tion device which is located below ground in a manhole or access hole
to collect the household wastes by gravity discharge. Grinder pumps
can also be installed in the basement of a home to provide easier
access for maintenance and greater protection from vandalism (41).
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 (8). A typical example
of a pressure sewer flow diagram is illustrated in Figure 10 (17).
-------
19
LEGEND
O PRESSURIZATION DEVICE
— COMBINATION CLEANOUT,
MANUAL AIR RELIEF AND
FLUSHING STATION
o BRANCH NUMBER
O EXISTING MANHOLE
— PRESSURE SEWER
GRAVITY SEWER
—-- CONTOUR LINE
FLOW DIRECTION
FIGURE 10. TYPICAL PRESSURE SEWER LAYOUT
-------
20
DESIGN ALTERNATIVES - PRESSURIZATION FACILITIES
Septic Tank Effluent Pumping
As previously noted, household wastewaters are collected by the building
drain and transported from the home via the building sewer to a septic tank.
A few investigators have characterized household wastewaters and per capita
flows (18) (19) (20). Mean flows have been found to vary from 43 to 50 gal/
cap/day (0.16 to 0.19 cu rn/cap/day). Table 2 represents the results of the
most extensive of these studies (19) for homes with and without garbage
grinding arrayed against typical municipal wastewater analyses for the
same parameters (21). Generally, the wastewater generated 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 sewerage systems.
Significant treatment occurs in a septic tank. Primarily, the septic
tank serves as a settleable solids and grease removal device. Heavy solids
settle during the multi-day nominal detention period, while grease and
other floatables collect in the scum layer. A cutaway view of a septic
tank is shown in Figure 11. Anaerobic biological activity occurs on a
sporadic basis, and this causes some liquefaction of accumulated solids.
This digestive action produces gas which 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
(including all the grit), and 50 to 80 percent of the BOO (22-25). 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 separation.
Suspended solids removal may be temporarily reduced or negated during
extended hot sumer periods due to increased anaerobic digestion and
resulting gas production and mixing from rising bubbles. BOO removal
is higher than that which is normally credited to primary sedimentation.
Typical septic tank effluent may have the following analysis:
BOO 5 , 100-180 mg/i
SS, 50-75 mg/i
Grease, 10-20 mg/i
The septic tank effluent then flows to a receiving tank, as depicted
in Figure 6, which houses the pressurization device, control sensors, and
valves required for a septic tank effluent pumping (STEP) system. The heart
-------
INSPECTION
___— OUTLET
FLOATING SCUM
/ /
/
LIQUID
I ‘ - ..‘
/ DIGESTING SLUDGE
7/////////// /
FIGURE B. SEPTIC TANK CUTAWAY
-------
22
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, Florida, area. The pumps
employed at Port Charlotte and Port St. Lucie, Florida, 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 12. The same pumps
are also employed in two STEP systems at Prfest Lake, Idaho although some
0.5 and 1 HP (0.37 and 0.75 kW) are also included in these systems for
locations where higher heads were required. The U.S. EPA demonstration
project at Bend, Oregon, employs similar pumps manufactured 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 purported
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 employed. The cost of these 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
alternatives. Design decisions for effluent holding tank installations
include material of construction, diameter, and working levels for the
tank, pump choice, and ancillary needs.
The effluent holding tank can be constructed of any material suited
for septic tank use. This includes properly-cured precast or cast-in-
place reinforced concrete and steel tanks meeting Ccnn erical Standard
177-62 of the U.S. Department of Coninerce with proper anti-corrosion
coatings (7). The Albany project report (6) indicated the epoxy-coated
steel tanks underwent severe corrosion during that study. Molded fiberglass,
reinforced polyester resin (FRP) tanks were found to be quite acceptable
on that project. The Phoenixville and Bend study used concrete tanks with
no apparent difficulty (ll)(15), while the Grandview Lake project (12) used
FRP, precast concrete and steel tanks with no mention of corrosion problems.
The Miami systems employ fiberglass tanks and the Priest Lake systems
utilize steel tanks with a litumastic coating (16). 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 13 (15). For single-
family dwellings, the ASCE report (8) determined that a minimum of 30 gallons
(114 1) of net storage capacity was required for a 12 gpm (0.76 1/sec)
discharge rate. The concern for storage capacity relates to the submersible
-------
30
E
20
0
C’)
0
0
4
w
-J
4
I-
0
I-
20 40 60 80
DISCHARGE, gpm (lgpml5.85I/s)
FIGURE 12. HEAD-DISCHARGE CURVE FOR SP33A PUMP
-------
24
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 1) in two 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 1) storage capacity was used at Phoenixville. Other
considerations for required tank diameter include the provision for adequate
access for repair personnel to remove and replace the defective pump or other
malfunctioning item. If the unit is located within three feet of the surface,
required diameter may be reduced to 24 to 30 inches (0.61 to 0.76 m), but
small diameter units have been troublesome at one GP location where grease
accumulations interferred with float switches (32). In locations where soil
conditions do. not provide adequate bearing strength or high groundwater levels
occur, a concrete pad or collar may be required as shown in Figure 13.
The working levels in a tank are the levels at which the pump originates
and terminates operation (see Figure 13). The volume of the tank between
these levels is considered as 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 1) while at Phoenixville monthly averages varied
from 33 to 137 seconds per 20 gallon (76 1) working volume cycle (6) (11).
These times are a function of the pump characteristics, 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 design (Model GP 210)
indicates a 10 to 14 inch (25 to 35 cm) differential which corresponds
to about 20 to 28 gallons (77 to 106 1) per cycle for their standard tank (17).
The Hydromatic GP design (Model CSPG-150A) indicates for a 6 inch (15 cm)
differential which would correspond to approximately 12 gallons (45 1) in a
similar tank (28). The Miami STEP systems employ 11 and 24 inch (28 and 61 cm)
differentials, which correspond to 22 and 47 gallons (83 and 178 1), respectively
(16). 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 presently not quantified,
but implications as to the relative merits of each can be derived in other
sections of this text. Presently, it appears that manufacturer’s standard
designs control the issue.
A similar statement can be made about the type of pump control switches
employed for STEP systems. Many pump manufacturers offer “packages” which
may include level control switches, control panels, wiring, simplified
maintenance systems, etc. Although several types of control switches exist
only two types have been employed 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 employed in a privately owned system
-------
25
CONTROL
BOX
CONCRETE
COLLAR
GUIDE RAIL
FLOAT SWITCHES
FIGURE 13. TYPICAL EFFLUENT PUMP CHAMBER
-------
26
near Bend, Oregon. In the Albany project one inch (2.5 cm) tube openings
were rejected for GP units due to grease buildup which caused them to
malfunction. After replacement with pressure sensing tubes having three
inch (7.6 cm) openings, no further problems occurred (6). No pressure
sensing tube malfunctions have been noted in either the Phoenixville
or Grandview Lake projects (11, 32). The Miami area STEP systems
employ diaphragm type pressure switches, and it is reported that these
devices become the major source of maintenance problems after two years
of service due to reduced elasticity of the diaphragms (16).
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 non-corrosive 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 it
operation. Several forms of mercury switches have been used in pressure
systems, but usually the switch is either attached directly to the
pump housing or is suspended from a stationary point above the liquid. These
type of controls are also standard equipment for several pump and GP
malfunctions. Some difficulties were experienced with these controls 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 employ mercury float switches with
negligible problems reported.
Some of the ancillary factors that must be considered are the
tank location, depth, and covering, electrical connections, warning
signals, and contingency items. Since the effluent holding tank 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 may require greater physical
separation of these tanks, however. Since 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 water-tight manner by gaskets and/or grooves
and be sufficiently above the ground to prevent entrance of normal surface
runoff. They should be made decorative as possible without imparing their
accessibility. Tradeoffs 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
which merits consideration is the use of covers which incorporate locks
which require 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 reconinended for
-------
27
both the pump and control circuits which should be wired separately, i.e.,
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 available. Outdoor locations must be
designed to thwart would be vandals. The choice of pump must be compatible
with the available electrical service, e.g., 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 employed, 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 in case the pump must be removed for servicing, but
must also be completely waterttght.
The primary contingency concerns of the designer are the possibility
of a power failure or the ability of the system to handle a pump malfunction.
It has been noted (17) that, during the period from 1968 to 1972, the 187
power outages recorded in the United States lasted as follows:
%of Total Outages Duration
53% < lhr.
81 <2hr.
89 <3hr.
95 <5hr.
97 <9hr.
The three percent of greater 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 impact
the total effect of the tragedy. Since nine hours appears 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 cu m)
would occur in that period considering the probability of reduced water
usage during the power outage. Since septic tanks usually have 6 to 12 inches
(0.16 to 0.33 m) of freeboard, a rectangular 1000 gallon (3.8 cu m) tank could
hold anywhere from 100 to 200 gallons (0.33 to 0.76 cu m) excluding the
capacities of the effluent holding tank and house sewer. Also, the loss of power
in rural areas which are served by individual wells and cisterns essentially
eliminates any possibility for wastewater generation since water supplies
become inaccessible. Since the system appears to handle power outages, the
primary potential difficulty appears to be malfunctioning mechanical components.
-------
28
The time involved between deteniiination of a malfunction by the alarm
light and the arrival of a repair crew are a function of the institutional
approach. of the sewer district. The approach, in turn, would be influenced
by factors such as the prior existence of soil absorption fields, the
size of existing septic tanks and the number of system contributors. For
example, if soil absorption beds were previously in use, an overflow from
the effluent holding tank to the bed could be sufficient to permit a normal
five day work week for repair personnel. Also, a larger septic tank with its
increased storage capacity above its normal water level would allow a somewhat
more generous response time than a smaller one would. A larger number of
contributors would justify having a larger repair staff employed by the
authority; if the number is smaller, a contract servicing arrangment with a
private firm might be more advantageous.
Rose (3) has posed the question of who should purchase, install and
maintain pressurization facilities. The unanimous opinion of several authors
(l)(26)(29)(30)(31) has 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, etc.).
Bowles (1) and Cochrane (31) have recoulBended 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
for a service charge upon request. Voell indicates that users prefer to
have a utility arrangement whereby regular maintenance and repair would be
performed by utility employees (29). 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 (30). All sources reconTnend 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 reliability, the number of units in the system, and
the rate of repair. For example, if a system of 100 pumps having 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 following situations:
p (1 pump fail) = 0.370
p (2 pumps fail) = 0.185
p (3 pumps fail) = 0.061
p (4 pumps fail) = 0.015
-------
29
The authority would then have to make a decision on the number of spare
pumps on the basis of this type of analysis. 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 condttions, 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 septic tank effluent pumping system appears to require no auxiliary
holding capacity. If, however, an existing soil absorption system is
available, the minor cost of installing an emergency overflow drain from
the effluent pumping tank to that system could be a prudent investment.
The major concern in using an existing soil field is infiltration during
wet periods which could result in a reverse flow from the field to the
effluent pumping tank via the overflow drain (32, 33).
The pump chosen for pumping septic tank effluent should be selected
on the basis of reasonable cost, reliability, proper head versus discharge
characteristics, and compatibility with the application. 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 penetration, e.g., submersible
or water—tight properties. If the above criteria are satisfied, the pump
choice becomes one of economics and proper performance characteristics.
In this type of application a centrifugal pump is probably the most economic
selection.
Although proper pump selection is a matter which is well discussed in the
literature (8)(34)(35), a few additional items should be taken into consider-
ation. 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.
Since little design information is available on this specific application of
multiple pumping units, the methods of analysis discussed by Metcalf and
Eddy (35) and Flanigan and Cudnik (34) should be helpful to the designer.
The original ASCE report (8) indicated that the maximum economical curb
pressure head should be equivalent to 69 feet (21 m) of water and that the
minimum pressurization unit discharge pressure head be equivalent to 0 to 11.5
feet (0 to 3.5 m) of water. Therefore, a maximum discharge pressure head
of about 81 feet of water (24.7 m) 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 restrict the capabilities
of the system. However, when the conditions of a particular site do not
demand it, there is no reason not to use less restrictive pressurization
criteria.
As previously noted, the most popular pump for septic tank effluent
pumping has been the low—head submersible sump pump. The primary reasons
-------
30
for this type of pump 1 s popularity have been low cost, generally flat
terrain applications, and availability of parts. Unfortunately, until
very recently no commercial pumps were available which were specifically
designed for this type of application. A need exists for alternative
units which 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 14 is one of two types of pneumatic ejectors
that have been developed expressly for pumping septic tank effluents;
these are manufactured by Clow Corporation and Franklin Research Company.
Both units require an air compressor to impart discharge pressures.
Design methods for selection of septic tank effluent pumps 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 which must be satisfied are primarily
related to system size, pipe sizes and lengths, the probability of
simultaneous pumping, and growth characteristics of development. Most
of these will be discussed with the pressure main and service line design
alternatives. The methods employed to determine the applicability of a given
pump have been outlined by Bowne (26), Metcalf and Eddy (35) and
Flanigan and Cudnik(34). Starting from the head-discharge curve shown in
Figure 15, the operating point of each pump can be found by (all of this
section in English units):
1. Determining the static head on the pump,
H = ht - h
where: H = static head, ft
ht = elevation of discharge point of pressure main, ft
h = elevation of pump, ft
2. Determining the approximate dynamic head due to pipe friction
and other pipe constrictions such as valves, bends, elbows, and
other fittings. The pipe friction losses are usually computed
by use of the Hazen-Williams formula,
H = 3.023 jv\ 1.852
Fl d 1167 V /
-------
COMPRESSOR
VENT
DISCHARGE
FIGURE 14. PNEUMATIC EJECTOR(COURTESY OF
CLOW CORPORATION)
SEPTIC TANK
CHECK VALVES
-------
I
0
4
w
I
V
4
z
0
-J
4
0
I-
FIGURE 15.
H 2
H
H 5
CAPACITY, 0
OPERATING POINT OF PUMP
WITH CHARACTERISTIC H=f(Q)
ANDA PIPELINE WITH
CHARACTERISTIC Hs+HL
°2°1
-------
33
where: HF 1 = pipe friction head, ft
d = pipe diameter, ft
V = velocity of flow, ft/sec
C = Hazen-Williams coefficient
while losses in valves and fittings are computed by use of the
following formula,
HF 2 = K V
2g
where: HF 2 = fitting friction head, ft
g = gravitational constant (32 ft/sec 2 )
K — fitting coefficient
Values of K for each type of fitting can be found in various hydraulics text-
books. The value of the Hazen-Williams coefficient “C” has been subject to
interpretation. The Plastic Pipe Institute indicates that tests in several
laboratories of new and used thermoplastic pipe resulted in “C” values ranging
from 155 to 165 and recomends the use of a conservative value of 150 (36).
The Albany and Phoenixville designs also used this value. Flanigan and Cudnik
indicate that it is proper for clean water applications, but due to grease
and other interfering matter present in wastewater, they reconinend the use
of 140. They note that this conservative value should permit easier operation
of the system during periods of stress and, if found to be overly conservative
with experience, it can be revised upward. The Grandview Lake system was
designed using a “C” value of 130 (33).
3. Applying the above information at various values of “Q” (discharge,
gpm) will yield the total dynamic head by:
TDH = H + HF 1 = HF 2
Since H and H are functions of discharge, the TDH is represented by a non-
linear creasi curve in Figure 15, where H 1 = H 11 + H ,. The intersection
of these curves is called the “operating point.” 1h norfn l pump selection
design practice, the pump would be chosen which has its optimum efficiency
at this point. However, since these pumps operate under varying conditions
of TDH in a pressure sewer system and since the cost of inefficient operation
is negligible in this type of operation, this requirement is not very
important.
-------
34
4. This analysis is used to “plug in ’ the information of a tentative
system design to the extreme cases, e.g., the pumps requiring the greatest
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 li’ne losses. The maximum and minimum TDH analyses
determine the variation in flows which 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 16 is assumed for all pumps in the system.
Line and fitting losses related to the service lateral which 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 16(c). Then each pump
must be referenced to a single location on the mainline, usually the point
where the pump closest to the discharge end of the main enters the mainline.
This point or station is shown in Figure 17.
This referencing can be accomplished by a reiterative series of
combinations, i.e., 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 Figure 17, it is
station #4 on the mainline). 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 mainline connection of the closest pump (station
#2 in this case) by converting the static and mainline friction heads
between the two stations. The major steps in the referencing process
are shown graphically in Figure 16. This simplified example assumes that
the pump and mainline station are at the same elevation. In part A,
the service discharge line losses are subtracted 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 part B, the elevation difference and
piping losses between stations #1 and #2 are shown and are applied to the
modified curve from part A to get the new H-Q curve for pump #1 at station
#2. In part 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 under pressure main design.
Bowne (26) has proposed using a much-simplified pump selection method,
which takes advantage of a centrifugal pump’s flexibility. Essentially, he
establishes the hydraulic grade line for the system and determines the
difference in elevation between the pump level and the hydraulic grade line.
Then, knowing the length and size of the service line, the modified pump H-Q
-------
35
H
01
FIGURE 16. MULTIPLE PUMP OPERATING ANALYSIS
-------
FLOW
FIGURE 17. REFERENCE POINT FOR MULTIPLE
PUMP OPERATING ANALYSIS
REFERENCE POINT
4
1 2 3
-------
37
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 obtai’ned. The number of simultaneous operati’ons which may reasonably
be expected will be dtscussed later, as will pipe sizes and other mainline
design factors which may tmpact pump selection. One of the advantages of
centrifugal pumps is that they can operate 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 (1.12 kW) pump immersed in 70 gallons (265 1)
of water for 4 hours raised the temperature of the water to 122°F (500C).
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.
Grinder Pumps
Much of the preceding analysis provided for septic tank effluent pumping
(STEP) stations is applicable to grinder-pump (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 which 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 presently exists or for a multi-family dwelling which
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 the resultant freedom from being 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 10,
for ease of maintenance and less-severe operation conditions. One other
significant difference is the fact that more data are presently available
for GP systems than STEP systems.
Design techniques will vary somewhat for GP installations, as compared
to septic tank effluent pumping 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 previously noted, the concept of
a grinder-pump for sanitary sewage transmission was integral to the original
ASCE study (8). As part of that study, General Electric Company developed
a comercial 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 Environment-One Corporation,
which became a pioneer in GP - pressure sewer development. Since that time,
Environment-One (E/One) and the Hydromatic Pump Division of Weil-McLain Company
have become the leading suppliers of GP units. There are other firms in
competition with them, such as Robbins & Meyers, Toran, Peabody-Barnes and
Empo—Corriell (no attempt has been made to compile a complete list, since any such
attempt would only be accurate at the time it was compiled). E/One and Hydromatic
-------
38
units represent the two major GP design choices, i.e., a progressing-
cavity (semi -posi ti ve-di spi acement) pumping element and a centrifugal
pumping element, respectively.
The H-Q curves for the basic single home models of each manufacturer
presented in Figure 18 differ markedly, as would be expected. The E/One
pump has been shown capable of operation above the 81 foot (24.7 m) design
limit for a considerable number of cycles by the National Sanitation
Foundation (37). The extreme condition of operation would occur ininediately
following restoration of power after a prolonged outage. Although it is
un1ikely that normal wastewater generation patterns would exist during
such an outage, the assumption is that all or a significantly large number
of the total units could be at or above their discharge or actuation levels.
Therefore, at the instant power is restored it could be assumed that all
units would coninence to discharge. However, since the resultant TDH would
be greater than both maximum heads. Therefore, a sequential pumpout would
likely occur. The sequence would initially permit discharge from those
units which pump against the least TDH, e.g., in the case of a “flat”
system, the units closest to the discharge point. The other units in the
system which cannot discharge due to excessive TDH must wait their turn
(thus, the sequential pumpout). During the period of excess TOM 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 1) of water
to 122°F (50°C) (34). Since all the units in a small co mnunity system would
likely be emptied in less time than this, no difficulties should be experienced.
The progressing-cavity design, when pumping against excessive TDH, employs a
thermal overload protector with automatic-reset capability. Since this type
of unit can pump at destructive pressures when unchecked, the thermal over-
load 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
Figurel13 it is apparent that the centrifugal unit will pump a given quantity
of sewage in a significantly shorter time at any TDH below 88 ft (26.8 m) of
water. This provides a higher velocity in the discharge system and reduces
the probability that simultaneous pump operations will occur.
Each manufacturer offers a package which 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 anti-
siphon 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 coninercial packages.
-------
39
0, gpm
FIGURE 18. GP CHARACTERISTICS
E/ONE (MODEL FARRELL 210)
U-
I
a
I-
I
I-
90
60
:30
0
HYDROMATIC(MODEL SPG 150)
0
10 20 30 40
-------
40
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 malfunctions.
GP holding tanks are not normally large enough to provide proper
storage of raw sewage. Solutions that have been proposed to handle
this contingency include the one used at Grandview Lake, which is shown
in Figure 19. These absorption pits and beds were installed when an
existing septic tank effluent soil absorption system was not available.
The specifications required that the bed have a minimum volume of
500 cu ft (14.2 cu m) 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 point 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 (33). Another design
proposed by Schultz (38) is shown in Figure 20. This 200 gallon (0.76 cu m)
holding 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 since the top of the tank is at least 2 inches
(5.1 cm) below the level of the lowest house drain. When power and/or
system operation is restored, the emergency tank drains by gravity into
the GP tank. The Phoenixville and Bend systems also employ existing
septic tank systems for emergency overflows (11, 15).
Leckman has discussed various other alternatives, such as standby
power, water service termination, other holding tank designs, and
interconnection with an adjacent GP unit (30). He estimated 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 activities to prevent overflow conditions.
Holding tanks could be conveniently located and pijiiped out by septic tank
pumpers, i.e., interim holding tank systems. If GP units were close
enough, interconnecting overflows are possible during periods when
malfunctioning of one unit occurs.
It appears essential that GP systems have some type of contingency
arrangement. This is based on the assumption that public health
considerations obviate the possibility of raw sewage contamination of
lawns, basements, etc. Confronted with this problem, the designer must
choose one of the above, or develop a new means of storage for such
emergencies. The choice should be influenced by the institutional
(servicing) arrangement, fluctuation of groundwater levels in the area
and available existing facilities. Also, the contingency approach should
minimize the cost to the homeowners.
-------
Absorption Pit Detail X- Section
5’ + JDR - (70 SLOPE UP)
SEPTIC
— O B E L
UI II r fli III fl Y II( I il f
nvric rurirzi u ui &i
Plan View
NOTE: SEPTIC TANKS ARE NOT INSTALLED WITH HOME UNITS THAT GRIND
AND PUMP RAW SEWAGE
SEPTIC TAN
ORCE MAIN
VICE LINE
Typical P ofile of a New Residence Installation
•ABSORPTION BED
FIGURE 19. EMERGENCY OVERFLOW DESIGNS
-------
EXIST. GROUND
I
4” VALVE W/BOX
VENT CAP
4”CLEANOUT
NOTE: ALL HOLES IN VAULT WALLS
SHALL BE SEALED AND
WATERTIGHT SOLVENT
WELDS MAY BE UTILIZED.
200 GALLON
EMERGENCY HOLDING
TANK
:
:
:
4’-O”
J
NOTE: THE TOP OF THE HOLDING
TANK SHALL BE AT LEAST TWO
INCHES BELOW THE LOWEST
FLOOR LEVEL OF THE
DWELLING.
FIGURE 20. EMERGENCY TANK DESIGN
r —
—
TO OP UNIT Y 1 ’1 4” I I
SLOPE 1/8” PER FT.
I
BUILDING
4 SEWER
-------
43
DESIGN ALTERNATIVES - SYSTEM
Service Connecti ons
The service connection 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) i.d. 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 pressurization facilities. Under these conditions the
minimum discharge pipe velocity was greater than 2 ft/sec (0.61 m/sec) and
headloss was less than 4 ft/TOO ft (4 cm/rn) of pipe length. The Albany
project employed 1.25 in (3.18 cm) polyvinyl chloride (PVC), Type I,
schedule 40 service lines with PVC-DWV fittings (6). The DWV fittings were
used due to their smoother transition properties when compared to schedule
40 or 80 fittings. At the end of this project grease accumulations which
resulted in reductions of as high as 40 percent of the lateral (service line)
cross-sectional area were found. The Phoenixville system employed 1.5 in
(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), poly-
propylene, 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 (32).
Most of the Grandview Lake service lines used 1.0-inch (2.54 cm) piping.
Bowles (1) noted that the Lake Lyndon BaThes Johnson Municipal Utility
District project uses PVC service lines from 1.25 to 2.0 inches (3.18 to
5.08 cm) in nominal diameter. The septic tank effluent pumping system at
Priest Lake, Idaho, incorporates 1.5 inch (3.81 c i i ’) PVC -SDR-26 service
laterals (27). The Miami systems employ either 1.25 or 1.5 inch (2.54 or
3.81 cm) PVC-SDR-26 service lines, while the Bend system employs only 1.25
inch (2.54 cm) PVC-SDR-26 laterals (16, 15).
The original ASCE study report (8) indicated that polyethylene pipe
had the advantage of being “plowed in” for quick sewering. Essentially,
“plowing in” refers to a system in which trenching, feeding of coiled
tubing or pipe, and backfilling are accomplished in a single operation.
Despite this advantage and polyethylene’s lower cost, it has not been used
in any pressure sewer systems. The reason appears to be that pressure-
resistant fittings are not available for polyethylene. Therefore, PVC has
been almost exclusively used for pressure sewer mains and service connections,
despite its greater cost and its inadaptability to “plowing” methods. It is
not inconceivable that future development will allow the use of polyethylene,
however.
-------
44
The need for check and gate valves in service connections is obvious.
Since the main is under pressure at all times, and 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 its downstream side to prevent the accumulation of solids which
impeded normal operation of the flapper. Their project report (6) reconi ended
that a one foot (0.30 m) run of horizontal pipe be used, while Leckman (32)
recomended two feet (0.61 m). The Phoenixville system employed 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 degree seating from the vertical.
This latter design was reported to be superior (32). The need for a
horizontal run following GP discharge check valves was also documented,
along with a preference for swing check valves over the ball check type (32).
The Miami systems employ a single check valve and gate valve on the
discharge line of the pressurization 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 have been fabricated
of brass or plastic. All of the above types of valves have been trouble-
free for as long as five years. Serious problems have been noted in these
systems with valves made of steel or iron due to corrosion (iron sulfide),
which caused failure within two years (16).
At least two manufacturers now offer flexible rubber-like check valves
for pressure sewer systems. 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 determinant 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
removed, redundancy of on—off valves on the service line, and economic
tradeoffs for the initial cost of the on-lot facilities (16, 26).
-------
45
It is likely that the valve would be located within a holding tank of
sufficient size where a curb valve is also employed for the service
line.
Some designers prefer redundancy of check and on-off (gate or ball)
valves, while others feel that such redundancy is unnecessary (16, 26).
The use of two on-off valves has been more coni ion 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 a redundancy in isolation from the main. The use of two check
valves is less clear. If such a design is chosen, however, it could be
accomplished by two separate check valves or a dual-check valve in a
single housing.
In order to connect each service lateral to the pressure main at
Albany and Phoenixville, sanitary tees with 45 degree elbows for 1.25 inch
(3.18 cm) connections were used (6, 11). The Grandview Lake system
employed curb cock-tapping saddle connections to the main, as shown in
Figure 21 (32). 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 m) lengths of PVC pipe in laying the pressure
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 coniiionly available 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 connection for his lateral. Variations on these two alternative are
manifold, and the economic tradeoffs 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 Avenue, New York,
New York 10017.
As noted previously, the size of service connection lines are normally
between 1 and 2 inches (2.5 and 5.1 cm) in diameter. However, the designer
should determine his proper size based primarily on the characteristics
of the pressurization pumps used in his particular system, i.e., a pipe size
which is proper for a 15 gpm (O.95-l/sec) pump may not be proper for a 30 gpm
(l.89-l/sec) pump. The choice, which is based on the tradeoff between head-
loss (which includes service line length) and minimum scouring velocity,
becomes more difficult with increasing system size. For small 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 laterals may be necessary. Greater flexibility
in pipe sizing is available with septic tank effluent pumping systems because
-------
PVC THREADED SEWER CLEAN-OUT
PIPE CLASS SDR2I 1.0. 4.072”
3 PVC PIPE CLASS SDR26 0.0. 4.000”
TO HOME UNIT
BRtCKS / STONES / CONCRETE PAD
FIGURE 21. CURB VALVE AND
RISER INSTALLATION
MAIN LINE
30”
PVC CURB VALVE
I”SERVICE LINE
TAPPING SADDLE
W/ CURB VALVE FOR
PRESSURE TAP
-------
47
of less stringent minimum velocity concerns. When the complete system
hydraulics are analyzed, the decision can be made in context with all
other hydraulic considerations, including friction losses due to valves
and fittings. In the Albany study approximately 32 percent of the total
friction loss was computed to be due to these items (6).
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 (8). 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 due to fittings. One potential difficulty
encountered by eliminating the loops is that of pressure main repair,
which will be discussed later. By employing a single main with unidirect-
ional 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 Environment-One design handbook also recommends
the use of separate sections or branches, where possible, to minimize
difficulties that result from occasional main repairs (17). An example
of this approach is shown in Figure 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.
Since the pressure sewer systems at Albany and Phoenixville
pumped against a static head, i.e., the terminal level of the pressure
main was at a higher elevation that the remainder of the system, maintaining
a positive pressure in the main was no problem. Difficulties can occur
if a positive pressure cannot be maintained due to high points in the system.
Several investigators (8, 26, 34) have recommended positive pressure maintenance
at all points within a system, but such a conclusion must be weighed against
a possible combined pressure gravity system where the latter is feasible. If
a system were located on flat terrain, the use of a standpipe could accomplish
the need for a positive pressure gradient. The above investigators all
describe the use of pressure-sustaining and pressure—control valves to
provide sufficient backpressure or artificial head to prevent draining of
portions of the line and the associated problems of siphoning and flow
impairment. If undulating terrain or long downhill runs are involved,
use of these devices may be necessary if a combined system is not feasible.
There is no reported experience with these devices, but the ASCE report (8)
and Bowne (26) describe their function in greater detail.
The Grandview Lake, Phoenixville, Radcliff and Priest Lake systems
employed air-release valves and the Environment-One design handbook (17),
the ASCE report (8), and Flanigan and Cudnik (34) recomend them. However,
-------
48
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 soruces of air
in the sewer include insufficient purging after main filling and testing,
malfunctioning pumps, or the release of air which 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 multi—bubble or foam—like” 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 component of the buoyant forces long enough to carry the bubble
in the direction of flow (with minimum slippage) beyond the next low point
in the system. If the duration of necessary flow is insufficient, the
bubble will return to its original location. Kent has determined the
relationship between slope, velocity and bubble size (39). According to
his studies, at a 10-degree downslope, a bubble 3 inches (7.6 cm) long
requires a velocity of almost 1.2 ft/sec (0.36 m/sec) 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 due to the air could be as high as 50 ft/l000 ft
(1.5 to 15 rn/rn) of pipe length. Flanigan and Cudnik (34) and Farrell (40)
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 requirement for sufficient duration is quite difficult to achieve since
pressurization units rarely operate for much more than a minute or two.
If sufficient velocity were present in the main to move the bubble at a
velocity of 1 ft/sec (0.30 m/sec), a one minute operation would only move
the bubble 60 feet (18.3 m). Since each pumping generally contains about 15
gallons (56.9 1) of sewage, the internal volumes of three coninon PVC pipes
would correlate to the following displacement distances (1 ft 0.305 m):
Nominal Size (in) Schedule 40 tft) SDR-2l (ft) SDR-26 (f: )
1.25 193 163 157
1.5 142 124 120
2.0 86 77 76
3.0 39 37 35
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 (39). By using
some engineering judgement combined with calculated velocities, displacement
-------
49
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 (26).
When it is deter, ined 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 essentially vertical risers attached
via 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 employed to release trapped air. Both types require an access way
for required operation and maintenance.
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. Since the branched or
dendriform design of a pressure sewer system has already been shown desirable,
all homes upstream of the shutoff point will be without service. Two
questions which must then be answered are:
(1) How quickly can repairs be made?
(2) Can the shutdown area be bypassed?
The design aspects of these questions are:
(1) What mainline ancillary facilities are required?
(2) What would be an optimum spacing?
The question of mainline 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 project, however, the desirability of using mainline gate
valves to isolate sections for repair was noted by the engineer (32). The
National Sanitation Foundation sewer layout described in the ASCE report (8)
suggested 600 ft (183 m) intervals between valve and cleanout facilities,
while Leckman (30) suggested a maximum distance of 400 feet (122 m). The
typical inline cleanout facility is shown in Figure 22. Two items should
be noted regarding this figure. The first is the fact that this design does
not provide for a temporary by-pass to minimize the number of units out of
operation and the second is the obviously reduced space and depth require-
ments of the valve box, as compared to a conventional manhole. This latter
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
-------
METER BOX
BALL OR
GATE
U,
0
FIGURE 22.
IN LINE CLEANOUT
-------
51
400 to 600 ft (122 to 183 m) 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 located on the
mainline wherever a branch main is to connect, as shown in Figure 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. Since some mainline repairs are likely to
require significant time to complete, the question of emergency overflow
capacity again arises. The primary design question involves the relation-
ship 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 which affect the validity of this question (but do not
eliminate it) are: (1) the fact that the isolated users can more easily
be contacted in order to request reduced use of water during the outage and
(2) the possibility that major repairs will take longer to make than the
normal excess holding capacity can acconii odate. 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 required.
Two bypassing arrangements that have been proposed involve temporary
hose connections and parallel mains. Voell discussed the use of fire hose
connections which could be employed on a temporary basis. The only modifi-
cations required to implement this approach would involve the use of tees
imediately upstream and downstream of a mainline ball or gate valve, with
a ball or gate valve and threaded fitting attached to each tee stem in
the mainline 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
which will be discussed later is the use of parallel lines of different
sizes. If this approach were adopted, the existence of these parallel
lines may offer a ready solution to the bypass problem.
Another item relating to the maintenance of the system is the terminal
cleanout provision. At the ends of each branch of a pressure sewer system
a cleanout facility should be provided. Figure 23 shows a typical design of
one of these facilities. It should be noted that water-tightness is not
necessarily required for these valve boxes, but local conditions such as high
groundwater levels and poor soil drainage should be weighed in making this
determination.
-------
METER BOX
FIGURE
23. TERMINAL CLEANOUT
BALL OR
GATE VALVE
/
/
-------
53
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 employed these anchoring
techniques on plugs, caps and bends exceeding 22.5° (32). 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 contri-
butors is twice the construction period number, velocities in the
pressure main will also be related by a similar ratio. The entire aspect
of hydraulic design is dealt with later, but some suggestions on the
problem of varying load factors will be discussed here. One of the
reasons for recomending the branched layout shown in Figure 11 is the
fact that smaller branch mains can be used efficiently by a community
developer. Although this practice can be difficult, it is not uncommon
that sections are fully or, at least, greatly developed before new
sections are opened for house construction. This offers many logistical
advantages to a builder during the construction period.
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 (29). 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 previous discussion on the need for mainline 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 Bowles (1) and Voell (29)
suggest using flushing devices that employ potable water supplied with
approved air breaks, this approach negates one of the advantages of
pressure sewers, i.e., reduced sewage volumes. The Grandview Lake design
employed 1,000 gallon (3.8 cu m) holding tanks equipped with end suction,
0.75 hp (0.56 kW) centrifugal pumps to flush the pressure mains (32).
These pumps were actuated by timers and pumping was stopped by low-level
float switches. The flushing liquid was septic tank effluent. This
system, therefore, did not increase the quantity of wastewater to be
treated from the pressurized system. Also, the timers can be set to
provide flushing during minimum flow periods (between 12 a.m.. and 7 a.m.).
-------
54
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 design
are: (1) the volume of main to be flushed; (2) required scouring velocities;
(3) the number of household sources required to provide the necessary
volume; and (4) proper pump selection to obtain the necessary flows at
the projected total dynamic head.
The hydraulic design of the pressure main has been discussed by a
number of authors (2, 6, 11, 14, 17, 26, 30, 34, 43). The original work
for the ASCE project was done by Hobbs (43). His work determined the
relationship that exists between sewage characteristics and carrying
velocity for pressure 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:
V 5 =
where V = minimum scouring velocity, ft/sec
d = inside diameter of pipe, in
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 was taken to define it. The sewage employed had grease concentrations
that ranged from 15 to 365 mg/i. 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 explaining this 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 which float the solids which combine with the grease
at the crown to create a solidified mass of substantial strength and durability
which could be highly resistant to dislodgement. This crown-oriented mass
creates greater flow resistance in the pipe.
The problem of grease accumulations affects the hydraulic design problem
in two ways: (1) the friction coefficient for the pipe will be different
than its nominal “clean water” value; and (2) scouring velocities are
thereby required to minimize the effects of these accumulations. The
magnitude of the grease problem is obviously greater in systems employing
GP units than in those using septic tank effluent pumping because grit, grease,
and suspended solids are removed In the septic tank.
-------
55
The Albany project was designed on the following assumptions:
(1) The maximum number of GP units operating simultaneously
would be all 12 in the system.
(2) The minimum number of units routinely operating would be
four.
(3) Four units would then provide a flow greater than the
minimum scouring velocity (6).
The study determined that the system was hydraulically over-
designed since the following frequencies were actually obtained:
(1) Two simultaneous GP operations were attained 20 times
per day.
(2) Three simultaneous operations were attained at least
once per day.
(3) Four simultaneous operations occurred about once every
14 days (2, 6).
Two major GP system design decisions were modified on the basis of the
Albany data. First, a minimum velocity of 2 ft/sec (0.61 m/sec) is required
in all pipe sizes normally employed in GP pressure systems (2)(6). Second,
on the basis of these data and other information, the Environment-One Corp-
oration produced the design table shown in Table 3 (17). There also
appeared to be an inverse relationship between the number of users of a
particular section of pipe and the amount of grease that accumulated in
each pipe size (2)(6). This reinforces the theory that no-flow periods
allow grease accumulations to develop. However, grease accumulation in
the pressure laterals did not follow this pattern (6).
The Phoenixville GP system design was tested with a computer program
which used the Hazen-Willimas 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 ft (24.7 m) at the start of the study
to 123 ft (37.5 m) 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 usage of home units. A Hazen-
Williams “C-factor” of 130 was used to estimate pipe friction losses (33).
As a result of the Grandview Lake experience, the engineer is designing
-------
56
TABLE 3
SIMULTANEOUS OPERATION OF GP UNITS
MAXIMUM OPERATING
SYSTEM GP UNITS SIMULTANEOUSLY
1 1
2-3 2
4-9 3
10-18 4
19-30 5
31-50 6
51-80 7
81-113 8
114-146 9
147 — 179 10
180 - 212 11
213 — 245 12
246 - 278 13
279 — 311 14
312 - 344 15
-------
57
new septic tank effluent pumping systems with a peak flow of 70 percent
of their peak water demand rate and a minimum velocity of 2 ft/sec
(O.6lm/sec) instead of the 1 ft/sec (O.3Oni/sec minimum used at Grandview
Lake (14).
A hydraulic design procedure for GP systems has been recommended by
Environment-One Corporation (17). After a preliminary layout of the
branches and pressurization facilities, a tabular analysis of the system
is made. Table 4 represents this type of analysis for the GP system
design shown in Figure 10. The maximum number of units operating
simultaneously is estimated by using Table 3, and the maximum flows
are assumed to be 11 gal/mm (0.69 l/sec) 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 Cudnik (34) present a strong case for the use of C140 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
superimposes the static head difference between the highest elevation in
the system between the pumps 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 m) by the E/One method. One possible difficulty
is the use of a Hazen-Williams “C-factor” of 150. Although non-specific
on the subject, the E/One tables intimate that pipe sized to accommodate
a minimum velocity of about 1.8 ft/sec (0.55 m/sec).
Flanigan and Cudnik recommend that velocities ranging from 2 to 5 ft/sec
(0.61 to 1.52 m/sec) 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 one to two times
per day. As previously noted, a “C-factor” of 140 is recommended at this
time, with possible revision upward with more experience (34).
Sanson (14) 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 which pumps
10 gal/mm (0.63 1/sec) at a design head of 80 feet (24.4 m). A minimum
velocity of 2 ft/sec (0.61 m/sec) is also required. The use of a “C-factor”
of 130 or less has been indicated (33).
Figure 24 is a plot of the above design flows for various sizes of
pressure sewer systems. The figure includes the recommendations of Sanson (14)
and Environment-One (17) and the maximum and minimum recommendations of
Flanigan and Cudnik (34). This last report recommended eight levels of
design, but only the two extremes are shown on the figure. These extremes
-------
TABLE 4
PRELIMINARY PRESSURE SEWER SYSTEM PIPE SCHEDULE AND BRANCH ANALYSIS SHEET
Pipe: SDR 21 PVC
(1) (2)
(3) (4) (5)
(6) (7) (8) (9) ( 10 ) (11)
(12) (13) (14) (15)
3
0
19
5
4
3
55
3
2
6
2
9
5.03
2
22
3
300
11
2
4.31
33
4
5
• 12.90
2.03
2
3
44
60
Br.
No.
No. of
Pumps
Accum. Maximum
Total on
Maximum
Flow
(a,m)
Size
(Inches)
Maximum
Veloci
ty (fps)
Length
(ft.)
Friction
Loss ft/
100 ft.
Friction
Loss
(total)
Accum.
Heedloip
(ft.)
Max. Ely.
Main
(ft.)
Pump
Site Ely.
(ft.)
th. .
01ff.
(ft.)
Max. Tot
Head
(ft.)
1
3
3
2
22
2
2.03 -
60
.79
.48
34.18
785
735
50
84.18
6
9
3
-
33
2
3.05
500
—
1.68
8.40
33.70
785
738
47
80.70
2
11
4
44
2
4.07
180
2.85
5.12
25.30
785
748
37
62.30
2
3
— 5
3
8
2
3 —
22
33
2
2
2.03
3.05
80
450
.79
1.68
.79 -
7.58
28.39
27.76
785
765
760
20
25
48.39
52.76
1+2
19
7
0
38
6
66
3
2.81
800
.91
7.28
7.28
785
755
30
37.28
20.18
3
2
3.05
5
.79
2
785
4+5
500
8
c i,
4.07
.48
3
1.68
22
751
180
34.18
19
8.40
33
2
2.85
34
785
6
33.70
54.18
0
2.03
2
5.12
3+6
744
785
25.30
80
19
3.05
5
41
.79
746
785
459
38
75.18
.63
55
1.68
39
755
28.39
72.70
2
7.58
30
785
55.30
27.76
5.08
785
300
770
768
4.31
15
43.39
12.90
17
20.18
44.76
785
758
27
47.18
-------
59
NUMBER OF SERVICE CONNECTIONS
FIGURE 24. SUGGESTED DESIGN FLOWS
‘a
I ’
C )
•iO
0
0
0
E
0.
0
-I
U-
z
0
w
0
0
Lu
I ’ ,
Lu
0
0
D
280
240
200
160
120
80
40
0
10
100
1000
-------
60
represent daily flows of 400 and 175 gallons (0.61 and 1.53 cu m) per
household. Sanson’s curve is based on 200 gal/day (9.76 cii rn/day) per
connection while the E/One curve is said to be based on peak flows
obtained in the Albany project and other existing pressure sewer systems (17).
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. Since 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 (14), however, is based on the fact that centrifugal septic
tank effluent pumps 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 24, a pump capable of pumping 10 gpm
(0.63 1/sec) at 81 feet (24.7 m) of TDH was assumed along with a maximum
peak household water demand of 15 gpm (9.95 1/sec) to obtain a (10/15)
(100) 70 percent factor to be applied to the peak water demand curve.
Bowne (26) has chosen the Flanigan and Cudnik curve which is based on
215 gal/day (0.81 cu rn/day) for his design. The generally lower peak flows
reflected by Sanson’s curve take into consideration the flow smoothing
capabilities of centrifugal pumps, i.e., 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 which should be addressed is the headloss due to
valves and fittings in the system. The Albany report estimated that the
losses due to valves and fittings were 32 percent of the total friction
loss (6). 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 septic tank
effluent systems with centrifugal pumping (26). After determining the
ultimate number of homes to be served by the system, a peak flow is
established based on one of the above sources or a recognized equivalent.
A profile of the proposed system is then prepared, and hydraulic grade lines
corresponding to various pipe sizes are plotted, as shown in Figure 25.
-------
8 10,000 12 14 16 18 20,000
PIPELINE DISTANCE FROM DISCHARGE POINT, Ft.
FIGURE 25.
PIPE SIZING PROCEDURE
z
0
I-
4
>
LU
-a
LU
200
180
160
140
120
100
0 2 4 6
-------
62
Since a reasonable approximation of pump characteristics is already
known based on economics, pressure limitations, etc., any pipe size which
indicates an excessive TDH requirement (difference in elevation between
the hydraulic grade line and ground or sewer profile) is sequentially
discarded until a satisfactory pipe size is found. In Figure 25 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 relief valves to prevent problems due to air pockets.
With the introduction of the pressure sustaining valve, a new dynamic
hydraulic grade line results and is plotted, as shown in Figure 25.
Individual pump characteristics can then be tested for sufficiency by
the elevation difference between the proposed pump elevation and the
elevation of the dynamic hydraulic grade line at the mainline station
where the pump lateral intersects. To accomplish this, 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 26. The previously noted head requirement can then be
located on this H-Q diagram at the design flow to detennine the
adequacy of the pump and the most suitable pipe size.
The type of plastic pipe chosen can appreciably affect the design
and economics of the pressure sewer system. PVC pipe has almost
exclusively been employed. The pressure sewer mains at Albany were
PVC Type I Schedule 40 with PVC-DWV fittings, and the joints were
solvent welded (6). At Phoenixville, PVC Type I SDR-26 piping was used
with PVC Schedule 40 fittins (11). At Grandview Lake PVC SDR-26 pipe
(solvent-welded) and PVC fittings were used. The Miami systems employ
PVC SDR-26 pipes with slip-ring joints, as do the Priest Lake installations
(16, 27). SDR-26 PVC pipe with solvent welded joints is used in the
Weatherby Lake, Missouri pressure sewer system. Flanigan and Cudnik
recommend the use of SDR-26 PVC piping in all systems whose pumping
heads do not exceed 90 feet (30.5 m) (34). Environment/One recommends,
in order, SDR-21, Schedule 40 and SDR-26 (17). The pressure rating of
SDR—26 pipe is 160 psi (1110 kPa) and SDR-2l is rated at 200 psi
(1380 kPa), while Schedule 40 pipe may vary. Two and three inch (5.1
and 7.6 cm) Schedule 40 pipes are rated at 277 and 263 psi (1920 and
1820 kPa), re pective1y (34). All pressure ratings are at a temperature
of 73°F (22.8’ C) and are generally reduced at higher temperature, to the
extent that PVC is not recommended above 150°F (65.6°C). The higher-
pressure-rated pipe recommended by Environment/One may be related to
their GP’s ability to operate at very high pressures. The safety
factor between pressure ratings of SDR-26 pipe and system design pressures
in almost all cases exceeds four. Since the other recommended PVC pipes
all have greater pressure ratings, their safety factors are larger.
Since a safety factor of four is common for water supply systems where
water hammer conditions are more likely, the safety factor for all PVC
pipes discussed appears adequate.
-------
63
50
40
.30
LA
0
4
w
I
20
10
0
CAPACITY, G.P.M.
FIGURE 26. PUMP AND SERVICE LINE
TESTING PROCEDURE
0 20 40 60 80
-------
64
As a matter of interest, one system has been designed with poly-
ethylene pipe in northern Michigan. This system is a combination
pressure-gravity design which employs grinder pumps. The details of
how the polyethylene (PE) pipe was adapted to the pressure system are
not available at this time, but 40 foot (12.2 m) lengths of PE pipe
will be fused together through the use of heat and pressure prior to
being laid in the trench (53).
Some concern has been expressed about the shallow depth of pressure
mains and their increased susceptibility to damage by excavating equipment.
It has been suggested by Bowne that markers be set along the pipe route
warning of its presence (26). 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 any and all cases. The
related question of differentiation between water and sewer lines of PVC
has been raised by Leckman (30). He also suggests markings where applicable
and other methods such as standardized relative locations of sewer and
water lines. The Grandview Lake system used brown colored PVC pipe to
simplify differentiation, while the Miami systems employ green PVC pipe
in one location and red striped pipe in another (16, 32). This type of
color coding is required by the Pennsylvania Department of Environmental
Resources (54).
-------
65
CONSTRUCTION CONSIDERATIONS
At Albany a plumbing contractor installed the small pressure sewer
system. A temporary 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’ unfamiliarity with the use of PVC. After the leaks were
repaired, the pipe was covered with 18 inches (45.7 cm) of sand (6).
The Phoenixville system was installed by a general contractor.
The trenching machine employed allowed construction in a trench of less
than four inches (10.2 cm) in width at an average pipe depth of 2.5 feet
(0.76 m). 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 four 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.
At Grandview Lake the system was also installed by a general
contractor. Where rock was encountered, 5 inch (12.7 cm) of sand bedding
was required (33). Several 8 inch (20.3 cm) layers of granular fill
with tamping were required where pipes passed under roads. The normal
pipe depth was three feet (0.92 m) below the ground surface. Joints
were solvent-welded and the solvent was allowed to set up prior to
“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 expelled. After about two weeks
in place, the pipe was leak-tested at 150 psig (1040 kPa). The pipe was
backfilled only when the temperature was below 65°F (l8.3°C)(32).
Bowne has emphasized the need for care during construction to avoid
scoring PVC pipe, which could result in strength reduction (26). He also
noted that bedding and backfill requirements are less stringent when
compared to conventional pipe due to the reduced brittleness of PVC at
moderate temperatures. His preference for pipe protection under roadways
is to encase the plastic pipe in a steel pipe for protection during service
and for ease of installation. His specifications are illustrated in
Figure 27 (42).
The British Standards Institution (BSI) provides an excellent set
of guidelines for plastic pipe application (44, 45). In relation to
construction or pipe laying procedures, the BSI specifies:
(1) The trench width should be equal to or greater than the sum of
the outside diameter of the pipe plus 12 inches (30 cm).
(2) The depth of bedding below the pipe barrel should be no
less than four inches (10 cm).
-------
66
TONING WIRE
PIPE ZONE BACKFILL
L APPROVED SAND
2. MAX. PARTICLE SIZE O.5in.
FIGURE 27. PIPE TRENCH& BACKFILL
30”
MIN.
4”
-------
67
(3) The bedding material should not exceed 0.5 inches (1.0 cm)
in size.
(4) Sidefilling around the pipe should be done in three inch
(7.5 cm) layers, with compaction of each layer by hand.
(5) Pipes should be partially backfilled (leaving joints exposed
for inspection) before pressure testing.
(6) Pressure testing at 1.5 times the maximum working pressure
at the point of maximum stress should be done for a period of one hour.
(7) A successful pressure test must not lose more water than 0.24
gal/bOO ft of length/in, of nominal dia./day/43.4 psi of test pressure
(3 l/km/2.5 cm/day/300 kPa).
(8) Due to reduced impact strength of PVC pipe in cold weather
no construction should be undertaken at temperatures below 14°F (_100C).
(9) 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 man-made 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 utilized a
special “rock-saw” trenching device for main construction (15). In any
case, the rate and ease of construction when compared to conventional
gravity systems are obvious. For example, at Weatherby Lake, Missouri,
the contractor laid 900 feet of 2 and 3 inch (5.1 and 7.6 cm) main on the
first day of work (52).
Care must be taken in any established area where pressure sewers are
installed to avoid damaging existing water, electrical and gas service
lines. At Grandview Lake, where the homeowners’ recollections were the
only guide to the location of service lines, an average of 1.3 existing
service lines were cut during pressure sewer service line installation
at each home. The maximum for one home was seven (32). Ideally, “as
built” drawings should be obtained, but these often do not exist in
rural areas.
-------
68
OPERATION AND MAINTENANCE
Obviously, any system which employs numerous pressurization
facilities and other more sophisticated mechanical equipment will
require a significant amount of operation and maintenance (O/M).
The institutional concepts alluded to earlier for the pressurization
facilities must also taken into account the needs of the entire
pressure sewer. These have generally been described in the design
considerations for the pressure main. For the most part, operation
and maintenance can be divided between on-lot needs (pressurization
units and service lines) and mainline needs.
At Albany the project started with twelve E/One prototype units,
most of which were replaced during the study by modified units (6).
The number of service calls then fell off sharply. Most of the 44
malfunctions reported were due to faulty pressure-sensors. Since the
modified units had improved pressure-sensing tubes, only 5 of the 44
malfunctions noted involved modified units. The malfunctions took the
form of excessive noise (due to their in-house location), continuous
motor operation, and non-functioning units resulting in overflows.
An “operation ratio” based on any greater-than-l5-minute malfunction
was calculated. This operation ratio consists of the number of days
when no malfunction occurred over the total service days. Since this
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 twelve homes. A separate
operation ratio was calculated to be G. 995 for the modified uni ts. A
number somewhat less than this would probably be more accurate since all
the startup difficulties which would normally be expected in the system
were not included in this latter calculation. The corresponding “down
time,” defined as the hours out of service over the total hours of possible
service, expressed as a percentage, was 2.69 percent for the prototype
units and 0.27 percent for the modified GP units. The difference was
primarily due to the improved pressure sensors of the modified units.
Power consumption at Albany was measured for two of the 12 units
(6). Monthly power consumption averaged 10.2 and 5.3 kWh for the units,
which also averaged 28.5 and 16.0 minutes/day of operation. Since 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 gallons/mm
(0.94 1/sec) is assumed as an average, the operating times represent
about 427 and 240 gallons (1.62 and 0.91 cu m) per day, respectively,
and the power required per unit volume is 0.795 and 0.736 Wh/gal (0.210
and 0.195 kWh/cu m), respectively. A conservative average would then
appear to be about 0.8 Wh/gal (0.212 kWh/cu m) for the GP units and
conditions of this study.
-------
69
The Phoenixville system experienced startup difficulties, primarily
caused by oversights during installation (11). During the course of the
study, problems were recorded with faulty GP discharge line materials
causing electrical short circuiting, a faulty circuit breaker, and
faulty grinder assemblies. If the operating ratio analysis used in the
Albany report (6) had been applied to the Phoenixville 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 discharge pressure of 11 psig (76 kPa),
while drawing an average of 12.5 amps. For unit #4, the valves were 59
minutes at 44 psig (304 kPa) and 14.0 amps. Therefore, by converting
pressure to flow via the H-Q curve and calculating kWh from the voltage,
amps and time of operation, the daily flows and power requirements can
be calculated to be 461 and 590 gallons (1.74 and 2.23 cu m) 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/cu m), respectively. These unit
costs are in close agreement with the Albany data.
The Grandview Lake operation and maintenance information is extensive (32).
Of the three major conniiercial GP units employed (E-l, Hydromatic, and Tulsa),
the E/One unit required the fewest 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 because the number of units was continually
changing as the number of connections grew. The calls reported to the
engineer were classified according to the nature of the malfunction, as
shown in Table 5 (32). The E/One unit was found to fail because foreign
particles scored the metal rotor and excessive delay in thermal overload
activation (32). 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 which
controlled operation (32). Some float switch problems occurred because
of grease accumulation due to a lack of sufficient swirling action in the
tank. The lack of an anti-siphon device also caused some problems (32).
Although lighter than the E-l, this GP was also difficult to install and
remove. A number of design problems were cited which resulted in the
increased rate of service calls for these units (32). The report recommended
redesign and upgrading of the units employed on this project. Both the E/One
and Hydromatic GP units have been modified since the time of this study.
One of the difficulties encountered in analyzing the Grandview Lake
report (47) is that the dynamic nature of the system precluded determination
of precise operation ratios, down times, etc. It is fair to say that the
experimental nature of this large installation produced better practical
-------
70
TABLE 5
SUMMARY OF MAINTENANCE FREQUENCY
(2 Year Period)
Locally Hydro-
Manufactured E-l mati Tulsa
Cause Unit Unit — Unit
1. Pump Failure 52 4 8 2
2. Grinder Failure 25 0 1 2
3. Piping Failure (Within Tank) 41 4 7 1
4. ElectrIcal Failure
(Excluding Controls) 66 3 2 0
5. Control Failure 23 0 10 2
6. Piping Failure (Outside Tank) 11 2 8 0
7. InfIltration/Inflow of Water 56 1 0 0
8. Collection System Malfunction 7 1 2 4
9. Improper Installation 9 2 2 0
10. Miscellaneous 81 0 12 5
Totals 371 17 52 16
Maximum Number of Units 27 15 28 2
aMaintenance of the E-l units was done primarily by the manufacturer t 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
englneer t s maintenance crew and may not include all of the service calls
by others.
-------
71
engineering data than Phoenixville or Albany, but less mechanical—electrical
performance information. The number of coniiiercial unit service calls shown
in Table 5 for the size of installation involved, excluding an unsuccessful
nontco erc1a1 GP unit, appears to be quite low, despite any data short-
comings. The E/One company indicates that 19 service calls were made by
their representative during 11,800 unit days of possible operation in the
Grandview system (47). This would yield an operation ratio of 0.9984,
assuming the numbers are accurate.
The Horseshoe Bay (Lake Lyndon Baines Johnson) GP installation has
also experienced some O&M problems (31). These have been identified as
excessive wear and failure of the GP stators due to 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) (31).
The Miami systems represent the longest history of 0/M of all the
pressure sewer systems. Thusfar, the experience with these pressure sewer
systems indicates that the 0/M costs are the same as for gravity systems.
Presently, an annual preventive maintenance inspection is employed. This
inspection consists of the following:
(1) The pump is removed from the holding tank, inspected for
corrosion and suction plate condition, and cleaned (if
necessary).
(2) The check valve and gate valve are inspected for proper
functioning.
(3) The pump is returned to its operating position and tested.
A two—man crew normally requires 30 minutes to complete such a preventive
maintenance procedure. Therefore, this preventive maintenance program
amounts to one man hour per year. In addition, replacement of diaphragm
switches and refurbishing of brass disconnect fittings is included in this
program every other year. On one of these systems which contains 26 STEP
units, one emergency repair has been reported in three years of operation (16).
Routine maintenance of the Priest Lake systems is reported to have
resulted in service calls to about eight percent of the STEP units ( 500 total
units in systems) in the first year and only two percent in the second year
(26). During the third year the service calls averaged about five per week,
with an average service time requirement of 30 minutes per call. One man
services both Priest Lake systems, routinely inspects and pumps out the
system septic tanks, and operates the treatment facilities. Experience
has indicated that this individual can remove 30 pumps per day, replacing
the impellers and returning them to service (26).
-------
72
The combined experiences of several STEP systems and some raw sewage
pumping systems indicates that a 10 year life can be expected for the
submersible sump pumps that have been thusfar employed before rebuilding
is required (16, 26). One of the major causes for servicing of STEP
systems is the buildup of iron sulfide on the pump impellor after
lengthy periods of inactivity, e.g., vacation homes (42).
Operation and Maintenance of pressure mains includes periodic cleaning,
repairing of leaks, and major replacement of sections. Due to the short
length of the Albany main, no maintenance was reported (6). At Phoenixville,
a two-day shutdown was incurred when an air-relief valve was damaged during
routine snow removal by heavy equipment (11). At Grandview Lake, a number
of service calls were required because of improper use of the tapping tool
used to connect individual home services to the pressure main was improperly
used, resulting in leaks (32). Additional leaks and breaks were caused by
heavy equipment, earthslides, and improper installation. Line cleanouts
were necessistated during the early stages of operation due to low flows
and concomitant solids builduos. 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 since 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 system caused repeated
breaks in the pressure main. The eventual solution to this maintenance
problem was effected by replacement of the PVC with a looped section of
flexible pipe (32). No maintenance of the Miami or Priest Lake system
mains has been necessary (16, 26).
In systems employing septic tanks prior to pumping, periodic septic
tank pumping is considered an 0&M 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. Pumpiu9 is not normally required at
intervals of less than three years (7), and may greatly exceed this length
of service. For instance, the Miami STEP systems have not required any
pumping of septic tanks in 5.5 years of service (16). Since 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 herein
is not limited to the design and construction methods discussed herein.
One example which is conspicuous by its absence is the use of a single
collection tank and pressurization device for more than one dwelling unit.
Rose (3, 27) has long been an advocate of this approach and the Bend
system employs one three home and three two home installations. The
-------
73
potential cost savings are obvious because the number of pumps or GP units
needed is reduced. Rees has expressed concern about the possible problems
resulting from allocation of operation and maintenance costs among
contributors which could make such an approach difficult to implement (33).
Another variation which could occur relates to the inclusion of multi-
family 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.
-------
74
CODES
Information on state and local codes which restrict or govern the
use of pressure sewer technology must be obtained from the responsible
agencies. Table I is a partial list of states that undoubtedly have
permanent or temporary regulations for pressure sewer installations.
States not listed should be contacted before planning begins.
Leckman (30) contacted 11 Illinois local government units to determine
the allowability of using 1.25 inch (3.18 cm) polyethylene pipe for trans-
mitting sewage. All 11 local authorities indicated 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 of 11
answered in the negative. Leckman also analyzed the 1971 Ten State
Standards (48) as they applied to pressure sewer systems. Serious conflicts
were found in Chapters 20 and 30 concerning design considerations such as
per capita flow, minimum sewer size and slope, sewer alignment, pump
openings, wet well requirements, emergency operation and minimum sewer
velocity.
-------
75
WASTEWATER CHARACTERIZATION AND TREATMENT
As previously noted, the wastewater from an individual home is more
concentrated than normal municipal wastewater. Some data from pressure
sewer systems have been obtained. At Albany, a thorough sampling and
analysis program produced the results summarized in Table 6. These data
are reasonably consistent with those of Table 2.
The Phoenixvjlle and Grandview Lake systems were sampled (generally
grab samples) only a few times, in contrast to more than 50 composite
samples analyzed at Albany (11, 32). The concentrations of pollutants
analyzed were generally dilute by Table 2 standards. The Phoenixville
data did confirm the absence of dissolved oxygen, as would be expected
in a pressure sewer (11). Since the Grandview Lake system used both GP
and STEP units, the combined wastewater does not necessarily follow the
pattern indicated by Table 2. Twenty-four hour composites yielded SS,
BOO 5 , and COD concentrations ranging from 80 to 265 mg/i, 100 to 310 mg/i,
and 230 to 462 mg/i, respectively (32).
The GP—pressure sewer effluent at the Horseshoe Bay, Texas, project
is treated by advanced treatment methods. The treatment sequence involves
an activated sludge system followed by chemical precipitation 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 compounds (31).
The amenability of pressure sewer wastewater to treatment is of primary
concern. Hobbs investigated the effects on sewage solids both from grinding
and conRninution (43). He found no apparent 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 sedimentation efficiency in the primary clarifier
could be significant in the design of treatment facilities.
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/i (6). 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 do show that suspended solids removal at
equivalent overflow rates is somewhat less for the pressure sewer wastewater,
when expressed as a percent. Due to its high initial strength, the primary
effluent resulting 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 consequences of this are
increased sludge production in the biological reactor and/or a need to maintain
a higher mixed liquor SS concentration for equivalent treatment. These
-------
76
TABLE 6
ALBANY WASTEWATER CHARACTERIZATION
CONCENTRATION (MG/i)
PARAMETER MEAN RANGE
BOD 5 330 216 - 504
COD 855 570 - 1,450
TSS 310 138 - 468
TKN 80 41-144
TP 15.9 7.2 - 49.3
Grease 81 31 - 140
pH (units) 7.1 - 8.7
-------
77
considerations are not likely to offset the advantage of treating lower
flow volumes. Since 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 the wastewaters from pressure sewer
systems require that the designers of treatment facilities use their best
engineering judgement 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 characteristics
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, sulfide 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
in Reference 49 are applicable.
In the case of a fully pressurized system, two effects of the pressure
sewer wastewater on normal treatment systems must be considered, i.e., the
tremendous variation in flow and the anaerobic condition. Since 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. Therefore, likely choices 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.
Treatment systems can be divided into lagoons and conventional biological
processes.
The Grandview Lake pressure sewer system employed lagoons with land
disposal of the effluent (32). The Grandview system had some difficulties
from a physical standpoint (levee failure), but runoff from the effluent
irrigated and non irrigated hay fields was not significantly different
in quality. The Priest Lake design incorporates two-stage lagoons with
supplemental aeration followed by land spreading. Due primarily to a high
evaporation rate, however, no effluent has yet been produced for land
spreading. Bowne has also suggested facultative lagoons followed by slow
sand filtration as a means of obtaining a high quality effluent (42).
Facultative lagoons with or without 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.
-------
78
When conventional biological systems are to be used, the problems of
extreme flow variation and hydrogen sulfide odor potential must be
accoimrndated in the design. The introduction of anaerobic wastewaters
directly into the aeration compartment of an activated sludge-type
system appears attractive, but high sulfide concentrations in wastewater
can encourage the growth of filamentous organisms in these systems (49).
However, the STEP system near Port Charlotte, Florida, is doing so with
no ill effects; no odors have been reported around the extended aeration
unit, and effluent BOD 1 Z and SS concentrations have averaged 6.9 and 14.3
mg/i over three years 5f operation (16).
Other methods of sulfide control include U-tube aeration, chlorination,
and ozonation (49). The latter two are expensive and somewhat inconsistent
with smaller treatment systems, while the first method results in substantial
headlosses. Any oxygen addition method would be most suitable at the
discharge end of a pressure sewer due to the air-binding hydraulic
problems discussed previously.
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 (16, 32, 42).
One additional point should be stressed. There may be a major
difference in treatment requirements between septic tank effluent pumping
systems and GP systems. In the former case, a relatively weak wastewater
in terms of BOD and SS must be treated, while additional maintenance in
the form of sep ic tank pumping is required. In the latter case, a very
concentrated wastewater in terms of BOD and SS must be treated. The
tradeoffs must be weighed by the design r. Further experience with both
systems will expedite design selections.
-------
79
COSTS
Capital Costs
No new technology is valuable unless its total costs are competitive
with those of existing technology in a significant number of siutations
encountered 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
groundwater regions and low population density coimiunities. For practical
purposes, these areas generally fall into the following categories:
(1) Low density areas unsuited for on-site disposal
(2) Geological conditions unsuited for normal excavation
(3) Undulating or hilly terrain
The practicality of pressure sewers is not limited to any of these
categories nor are they 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.
However, all information on costs is instructive 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 to an estimate for
a GP-pressure sewer system of 1.2 million dollars. 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). A
Saratoga, New York, estimate of $100,000 for sewering 10 houses was
rejected in favor of a cost of $20,000 for a pressure sewer-GP system.
Conventional sewerage estimates for two comunities at Priest Lake, Idaho,
were about $12,000,000 with treatment (27). A septic tank effluent pumping-
pressure sewer system was subsequently built for less than $1,000,000.
Bowne (46) has presented a present-worth comparison of a conventional gravity
system with a septic tank effluent pumping-pressure sewer system using average
costs in his region of Oregon. On a total cost basis for this rural-suburban
area, he 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 7. The
two locations where the highest percent savings are noted (Nos. 3 and 5)
involve low-lying areas below existing sewer grades.
-------
80
To determine the limits of applicability of pressure sewers, it may
be assumed that low density housing areas where soil conditions are favorable
will be best served by on-site wastewater treatment and disposal systems.
This assumption leads to the question of what housing density constitutes
the point where conventional sewer technology becomes best suited to the
population. Although computation of this density would be a most valuable
piece of information from a theoretical point of view, it would be highly
subjective and would vary with each assumption made or physical condition
which would exist 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 include capital costs in various regions of
the country, for various climates and for various soil and other geological
conditions. For example, the cost of pumps, pipe, valves, labor, etc., vary
by region. The requirement that pipes be buried below the frost line 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 information 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. Until information on these requirements
becomes available, this aspect of the total picture will remain the least-
definable aspect of the entire equation. Only well-documented experience
over a significant period of time from a number of installations 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 adjustable profit margins, e.g.,
if a large profit was figured for one category of work, a different category
bid may reflect smaller profits to complement the other. Also, it should be
noted that 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 detennining capital costs, a number of subelements must be considered,
such as engineering, valve boxes, fittings, cleanouts, flushing arrangements,
testing, etc. Some of these items may be difficult to estimate. Therefore,
a few basic facts will be presented with expanded information based on
traditional practice. One primary cost category is the pressure main cost
per lineal foot. At Phoenixville (11) the cost of PVC pipe, excavation,
installation, lateral tie-ins, and restoration for 2800 feet (854 m) of main
was $2.00/ft ($6.56/rn). Additional costs of rock removal, restoration of
-------
TABLE 7
COST COMPARISONS
LOCATION ENR GP-PRESSURE CONVENTIONAL SAVINGS FOR
NO. CONNECTIONS INDEX SEWER ($/CONN) SEWER ($/CONN) GP-PS (% )
1 285 1700 1,930 4,570 48
2 30 1895 2,800 4,667 40
3 9 2014 2,222 10,000 78
4 309 1753 3,240 6,176 49
5 10 1753 1,653 10,000 83
6 320 1895 2,709 4,088 34
7 100 2098 1,360 2,350 42
NOTES: 1. All estimates by consulting engineering firm except no. 2 and no. 5, which are bid
prices for GP-PS systems.
2. Conversion of estimates to current costs requires use of present ENR Index.
-------
82
streets and driveways, load protection, an automatic air-relief valve
($350), and a terminal cleanout ($350) resulted in a total pressure
sewer cost of $2.82/ft ($9.25/rn). Pressurization facilities and service
connections amounted to $2,050, including $900 per GP. In this total,
service line costs of $2.50/ft ($8.20/rn), circuit breaker costs of $60
each, and power cable connection costs of $3.00/ft ($9.84/rn) were
incurred. The overall (Jan. 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. However, the costs can be allocated as
$595/person,$l ,270/dwelling unit, $2,720/structure or $3,8l0/GP. The
engineering fee for design and supervision of construction was $2,100 and
the legal charges were $2,500. This additional $4,600 increases the
above unit costs to $864/person, $1842/dwelling unit, $3946/structure and
$5530/GP, respectively.
Early bids on the Grandview Lake project were rejected because they
were almost twice that of the engineering estimates. Subsequent bids
were acceptable. The final cost of the 28,352 foot (8,640 m) pressure
main was $35,491 or about $1.25/ft ($4.11/rn). This total includes blacktop
road repair, manual ($125 each) and automatic ($200 each) air-relief 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 pieces of information are useful. The bid price for
6,600 feet (2,014 m) of 1 inch (2.54 cm) service line was $0.60/ft ($2.07 m),
and for 360 feet (110 m) of 1.5 inch (3.81 cm) line it was $0.95/ft ($3.12/rn).
Also, the cost of the curb cock is shown to be $20. In addition, a 1971
report on the project showed that on-lot installed costs for the various
GP and septic tank effluent pumping units ranged from $1000 to $1500,
assuming 150 feet (45.8 m) of service line and no electrical hookup (12).
The original costs at Grandview reflected the use of home-made GP units,
which were later replaced with coninercial ones (32). Engineering and
legal fees amounted to $23,384 at 1969 rates.
Several costs have been accumulated on the various items of equipment
that make up 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/rn)
in sizes up to 6 inches (15.2 m) nominal diameter (11, 27, 33). In difficult
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/m)(15, 38, 50). Carcich, et al, (2) and Bowne (46)
indicate that costs could be as high as $15.00 per lineal foot ($49.20/rn)
for some areas.
Mainline accessories such as air relief valves have been installed at
a cost of $120 to $350 each (11, 27, 33). Cleanouts have been estimated to
cost from $150 to $400 each (2, 11, 50). Valve boxes cost between $100 and
$900 each, depending on their design and construction conditions (33, 38).
-------
83
Grinder-pumps are estimated to cost from $1,000 to $2,000 each, with
an additional $300 to $700 installation cost per unit (2). Septic tank
effluent pumps cost about $200 to $400. Depending on the amount of
ancillary equipment (alarms, valves, sensors, switches, tankage, removal
mechanism, etc.), these units may be installed for $1,000 to $2,000 each
(15, 46). Bowne breaks down his estimate for an existing home whose septic
tank must be replaced 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 (46).
0/M Costs
Operation and maintenance (O/M) costs are generally unknown for
pressure sewer systems. For all systems, the cost of main repairs and
cleaning must be added to the OIM needs of the pressurization facility.
Leckman (30) has estimated the maintenance cost on GP units to be $4 to
$8 per month, and Dounoucos (51) has estimated a GP maintenance cost of
1.4 to 2.0 percent of the on-lot capital cost per year. Bowne relates
that GP service contracts have been instituted at rates of $48 and $60/year
(46). The Priest Lake system operator indicates that effluent pumps can be
rebuilt for $50 to $100 each (42). Replacement motors cost less than $100,
while seals, bearings, and capacitors cost about $7, $5, and $9, respectively
(46). Bowne has estimated that on-lots systems will require an O/M cost
of $50/yr, and that pressure main O/M will cost $100/yr per mile ($62/yr per
km)(46).
Septic tank cleaning is generally found to cost $30 to $50 can be
assumed to be required at 3 to 5 year intervals to protect the grease and
solids removal capability of the tank. This assumption is based on
traditional septic tank practice, where pumping is performed in order to
protect the subsurface disposal field from potential clogging due to whole-
sale 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 simultaneous failures of several tanks might result in serious
problems due to fonling of pumps, controls or pressure mains. Schmidt
surveyed 12 septic tanks in the Miami STEP systems to discover that
accumulations of sludge and scum were significantly less than earlier
USPHS studies (16, 22—24). Bowne, like Schmidt, suggests a ten-year
interval between pumpings, tempered initially by yearly inspections to
determine individual site accumulation rates for developing rational
pumping schedules (16, 46). The cost of inspection and pumping must be
included in the O/M cost estimate for STEP systems.
O/M 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 O/M is
probably between 7 and 8 cents per year per foot (23 and 26 cents/yr/rn) of
pipe, which converts to about $400/yr per mile ($248/yr/km). Bowne used
-------
84
actual 0/M costs for rural water supply systems to obtain a pressure main
0/M estimate of $100/hr/mi ($62/yr/km)(46). Since burial depths of
water mains and pressure 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 include a 0.0033 service ratio for the E/One GP units at Albany (6).
In terms of days/year requiring service calls, this corresponds to 1.2.
At Grandview Lake, E/One GP units required service calls 19 times in 11,800
unit days of operation (47). This reflects a service ratio of 0.0016 or
about 0.6 service days/year. Hendricks and Rees present similar numbers
for the same project and GP unit (32). Their information indicates that
the Hydromatic and Tulsa GP units indicate approximately 1 and 4 service
days/year, respectively. From the above data and the factory improvements
and modifications which have occurred in the interim, it would seem prudent
to assume a conservative service requirement for the GP units of about one
day per year.
The Priest Lake STEP systems experienced problems with 8 percent and
2 percent of the pumps (total ‘ . 500) during the first 2 years of operation,
respectively (46). Much of the first year 0/M was due to the supplier
providing improper impellers on the pumps. During the summer season (when
all units were in operation) of the third year approximately 5 service calls
per week were experienced. The 0/M 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 (16). This
includes about 1 man hour/yr of preventive maintenance on each STEP unit.
One-half hour of 0/M was required on the Bend system during the first month
of operation (15).
Some Oregon systems are using submersible sump pumps of the same type
as employed in STEP systems for pumping raw sewage (15, 46). The O/M
experience for one installation is purported to be 5 service calls per month
for a system of about 150 pumping units (55).
The consensus of the above experiences is that an effective preventive
maintenance and overall 0/M program for STEP systems would involve a yearly
inspection of each pressurization facility (septic tank, pump, sensors, valves,
etc.) and about 0.5 service call per year. Pressure main 0/M should be
less than required for GP pressure mains since most of the problematic
material (fibers, grease, etc.) is removed by the septic tank.
The p ier cost of GP or STEP units can be estimated. The Albany and
PhoenixvilIe GP information conservatively indicate the need for about
one Wh/gal (0.264 kWh/cu m). The power cost can be estimated by multiplying
-------
85
this figure by the number of occupants per home and the average wastewater
flow generated per capita daily. For example, if a GP were to be used by
a four—person family with an average wastewater flow of 50 gal/capita/day
(0.19 cu rn/capita/day), the monthly power cost would be about 15 cents on
the basis of 2.5 /kWh. Bowne estimates that the power cost for 0.33 hp
(0.25 kW) septic tank effluent pumps would be about 10 cents per month (26).
A conservative estimate for both STEP and GP units would be about 2O /mo.,
or about the same as an electric coffee maker (47). Actual costs for each
installation will depend on the number of people served, the cost of
electricity, and the specific pressurization device chosen.
-------
86
DESIGN EXAMPLE
For a simplified example, the system depicted in Figure 10 can be
used, assuming a scale of 1 in. = 300 ft (1 cm 36 m). Additional
assumptions are that: (1) Pvc mains of 2 in. (5.1 cm) nominal diameter,
except for a 3 in. (7.6 cm) interceptor (branch no. 7); (2) service
lines of 1.25 in. (3.2 cm) nominal diameter PVC; and GP units. Unit
Installed costs are assumed to be:
3 in. PVC pipe @ $4.00/ft
2 in. PVC pipe @ $3.00/ft
1.25 in. PVC pipe @ $2.00/ft
Service Line Connections @ $35.00 each
GP units, including
electrical hookup @ $2,000 each
Cleanouts with
Manual Air-Relief
Valves @ $500 each
A rough estimate of the capital cost of this system is:
3 in. pipe (800 ft) = $ 3,200
2 in. pipe (3,140 ft) = $ 9,420
1.25 in. pipe (5,700 ft) = $11,400
Connections (38) = $ 1,330
GP units (38) = $ 76,000
Cleanouts & Valves (7) = $ 3,500
TOTAL $104,850
This represents a cost per home of about $2,760.
Two things are vividly shown in this example: the economical nature of
the pressure sewer and the high cost of grinder-pumps. It is because of this
latter fact that the use of septic tank effluent pumps is being investigated
and tried by many.
In many rural areas sewers are being required because poor soil conditions
have obviated the continued use of the original septic tank-soil absorption
system. If this were the case in the example location, the data developed
by Bowne indicates 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 previous example would reduce the
cost per home to about $1,760.
-------
87
The O&M costs for the example installation would approximately be:
Pipe: 9,640 ft @ $100/mi/yr = $18/yr
GP: 38 units @ $54/yr = $2,052/yr
Power: 38 units @ $2.40/yr = $91/yr
TOTAL $2,161/yr
This amounts to an 0/Fl cost per home of $56.87 or a monthly cost of
about $4.74. To this the amortized capital cost must be added to get the
total monthly cost.
For this example, no engineering, 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 (municipal rate) yields an annual
cost of $238 per home, or $19.83/month. The total monthly cost per home
is then $24.57. Since these systems are eligible for U. S. EPA Construction
Grant funding, the cost per home could be reduced to a fraction of the above
amount.
From the previous discussion of the O/M costs for GP and STEP systems,
there is not enough evidence available at this time to justify a difference
in the O/M 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 O/M plus 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/home, or $12.65!
month. Therefore, the total monthly cost for the STEP system is then
$17.39. Grant eligibility is the same for both systems, with the exception
that new septic tanks required for the STEP approach are not eligible.
The foregoing example is admittedly crude, but it gives some idea of
the cost estimating procedures necessary to evaluate proposed pressure sewer
systems. Additional factors will have to be evaluated in order to properly
accomplish such an estimate in a real situation.
-------
88
CONCLUSIONS
1. Pressure sewer systems are a viable alternative technology and should be
considered in any cost-effective analysis of alternative wastewater management
systems in rural coimiunities.
2. Pressure sewers offer many advantages over conventional gravity sewers in
areas where:
a. Population-density is low
b. Severe rocky conditions exist
c. High groundwater or unstable soils prevail
d. Undulating terrain predominates
3. The most serious impediment to wider pressure sewer technology adoption
is the present lack of comprehensive long-term operation and maintenance (O/M)
data and treatment i niormati on.
4. Lower capital costs and significantly shorter construction times are
inherent in the pressure sewer technology, as compared to conventional methods.
5. Pressure sewers should only be considered with properly conceived manage-
ment arrangements. Failure to do so could seriously limit the effectiveness
of this technology.
6. Two major types of pressure sewer system designs are available, i.e.,
grinder-pump (GP) systems and septic tank effluent pumping (STEP) systems.
The relative merits of each of these systems should be weighed by the engineer
in his cost—effective evaluation.
-------
REFERENCES
1. Bowles, E. J., “Pressurized Waste Water Collection,” paper presented
at WPCF Conference, Atlanta, Georgia (1972).
2. Carcich, I. G,, Hetling, L. F., and Farrell, R. P., “Pressure Sewer
DemonstratIon,” Jour. Envir. Eng. Div. - ASCE , 100, No. 1, pp 25-40
3. Rose, C. W., “Rural Wastes: Ideas Needed,” Water and Wastes Engineerjj ,
9, No. 2, pp 46-47 (1972).
4. Gainey, P. L. and Lord, T. H., Microbiology of Water and Sewag ,
Prentice-Hall, Inc., Englewood Cliffs, N. J. (1957).
5. Smith, R., and Eilers, R. G., Cost to the Consumer for Collection and
Treatment , USEPA Report No. 17090—07/70 (1970).
6. Carcich, I. G., Hetling, L. J., and Farrell, R. P., A Pressure Sewer
Demonstration , USEPA Report No. R2-72-091 (1972).
7. Manual of Septic-Tank Practice , USPHS Publication No. 526 (1969).
8. American Society of Civil Engineers, Combined Sewer Separation Usij
Pressure Sewers , USD1, FWPCA Report No. ORD-4 (1969).
9. Clift, M. A., “Experience with Pressure Sewerage,” Jour. Sanitary E .
Div. - ASCE , 94, No. 5, pp 849-865 (1968).
10. Carcich, I. G., Hetling, L. J., and Farrell, R. P., “ The Pressure
Sewer: A New Alternative to Gravity Sewers,” Civil Engineering , 44,
No. 5, pp 50-53 (1974).
11. Mekosh, G. and Ramos, D., Pressure Sewer Demonstration at the Boroyg
of Phoenixville, Pennsylvania, USEPA Report No. R2-73-270 (1973) .
12. Hendricks, G. F., “Pressure Sewage System and Treatment at Grandview
Lake, Indiana,” paper presented to ASAE Convention, Pullman, Washington,
(1971).
13. Rees, S. M. and Hendricks, G. F., Grandview Lake Sewage Research and
Demonstration Project - Annual Report , prepared for USEPA (1971).
14. Sanson, R. L., “Design Procedure for a Rural Pressure Sewer System,”
Public Works , No. 10, pp 86-87 (1973).
15. Clark, L. K. and Eblen, J. E., C&G Engineering, personal convuunication.
-------
16. Schmidt, H., General Development Utilities Company, personal coimiunications.
17. Design Handbook for Low Pressure Sewer Systems , Environment/One Corporation
(1973).
18. Bennett, E. R., and Linstedt, K. D., Individual Home Wastewater Character-
ization and Treatment , Cob. St. Envir. Resources Center Completion Report
No. 66 (July, 1975).
19. Witt, M., Siegrist, R., and oy1e, W. C. “Rural Household Wastewater
Characterization,” Home Sewage Disposal , ASAE Publication Proc. - 175,
79-88.
20. Jones, E. E., Jr., “Domestic Water Use in Individual Homes and Hydraulic
Loading of and Discharge from Septic Tanks,” Home Sewage Disposal , ASAE
Publication Proc.-l75, 89—103 (1975).
21. Metcalf and Eddy, Inc., Wastewater Engineerin _ g , McGraw Hill, New York (1972).
22. Weibel, S. R., Straub, C. P., and Thoman, J. R., Studies on Household
Sewag e Disposal Systems - Part I , U. S. Public Health Service Publication
(1949).
23. Bendixen, T. W., Berk, M., Sheehy, J. P., and Weibel, S. R., Studies on
Household Sewage Disposal Systems - Part II , U. S. Public Health Service
Publication (1950).
24. Weibel, S. R., Bendixen, 1. W., and Coulter, J. B., Studies on Household
Sewage Disposal Systems - Part III , U. S. Public Health Service Publication
No. 397 (1954).
25. Patterson, J. W., Minear, R. A., and Nedoed, T. K., Septic Tanks and the
Environment , Report to Illinois Institute for Environmental Quality (1971).
26. Bowne, W. C., Pressure Sewer Systems , report presented to Douglas County,
Oregon (1974).
27. Rose, C. W., Farmer’s Home Administration, personal communication.
28. Specifications for Hydromatic Model CSPG-150A , Hydromatic Pump Company (1973).
29. Voell, A. T., Investigation of Pressure Sewer System and Advanced Waste
Treatment, report to Chatauqua County, N. V., Health Department (1972) .
30. Leckman, J., Pressurized Sewer Collection Systems , report prepared for
Illinois InstTtute of Environmental Quality (1972) .
31. Cochrane, D. M., “Pressurized Sewer Systems: Regulatory Agency’s Viewpoint,”
paper presented at WPCF Conference, Denver, Colorado (October, 1974).
-------
32. Hendricks, G. F., and Rees, S. M.., Economical Residential Pressure Sewage
System with No Effluent , USEPA Environmental Protection Technology Series
Report No. EPA-600/2-75—072 (December, 1975).
33. Rees, S. M., SIECO, Inc., personal communication.
34. Flanigari, L. J. and Cudnik, R. A., Review and Considerations for the
Design of Pressure Sewer Systems , Hydromatic Pump Division, Weil-McLain
Company (1974).
35. Langford, R. E., Peabody Barnes Co., personal communications.
36. Water Flow Characteristics of Thermoplastic Pipe , Plastics Pipe Institute
Technical Report No. PPI-TR14 (1971).
37. Report on the Performance of Grinder Pump Model Farrell 210 , National
Sanitation Foundation Testing Laboratory Report No. C-9-3 (1973) .
38. Schultz, J., Becher-Floppe Engineers, Inc., personal communication.
39. Kent, J. C., The Entrainment of Air by Water Flowing in Circular Conduits
with Downgrade Slopes , thesis, University of California (1952).
40. Farrell, R. P., Operation of Grinder Pumps and Pressure Sewer Systems
under Abnormal Conditions , information bulletin from Environment/One
Corporation (1974).
41. Detail Drawings and Specifications SA-2021B-137 through SA-2O2lB-l44
Envi ronment/One Corporation fl 973).
42. Bowne, W. C., Douglas County Road Department (Oregon), personal communication.
43. Hobbs, M. F., Relationship of Sewage Characteristics to Carrying Velocity ,
Report No. R-2598 to ASCE (1967).
44. Code of Practice for Plastics Pipework: Part 1 - General Principles and
Choice of Material , British Standards Institution Publication CP312 (Gr 8)
London (1973).
45. Code of Practice for Plastics Pipework: Part 2 - Unpiasticized PVC Pipe—
work for the Conveyance of Liquids under Pressure , British Standards
Institution Publication CP 312 (Gr 7), London (1973).
46. Bowne, W. C., Glide-Idleyld Park Sewerage St y , Douglas County, Oregon
Report (1975).
47. Farrell, R. P., Envirorui ent/One Corporation, personal communication.
-------
48. Great Lakes - Upper Mississippi River Board of State Sanitary
Engineers, Reconinended Standards for Sewage Works , Health Education
Service, Albany, N. Y. (1971).
49. Pomeroy, R. D., Process Design Manual for Sulfide Control in Sanitary
Sewerage Systems , USEPA, Technology Transfer Publication (1974).
50. Alfred Crew Consulting Engineers, Inc., Preliminary Report on Water
Pollution Control System—Siek Road Area (Addendum No. 2) , report
prepared for Borough of Kinnelon, N. J. (1974) .
51. Dounoucos, A., “Sanitary System Construction Costs Turn Engineering
Attention to Alternate Solutions,” Professional Engineer , 44, No. 8
(1974).
52. Gray, G. C., “Environmental Constraints Challenge Designers of Shoreline
Coninunity near Kansas City, Missouri,” Professional Engineer , 45, No. 6,
(1975).
53. Williams, 1. C., “Plastic Pipe and Pressure Sewers Mark Expansion,”
Water and Wastes Engineering , 12, 11, 85 (1975).
54. Sewerage Manual , Pennsylvania Dept. of Environmental Resources, Bureau
of Water Quality Management Publication No. 1 (3rd Ed), Harrisburg, PA,
(1976).
55. Ward, J., personal coninunication.
------- |