Aspe - Cold Water
Aspe - Cold Water
Aspe - Cold Water
5
Cold-Water
Systems
s×w
(16 × 19.1) 20 (1.26) 10.4 (71.76) 16 (1.00) 6.1 (42.19) 10 (0.63) 1.0 (6.9) 6
w (19.1) 30 (1.89) 10.6 (73.13) 24 (1.51) 6.9 (47.61) 15 (0.95) 1.05 (7.24) 6
1 (25.4) 50 (3.15) 9.3 (64.14) 40 (2.52) 6.3 (43.47) 25 (1.58) 1.0 (6.9) 6
1½ (38.1) 100 (6.30) 11.3 (77.10) 80 (5.05) 8.6 (59.34) 50 (3.15) 0.9 (6.21) 6
2 (50.8) 160 (10.08) 10.4 (71.76) 128 (8.08) 6.5 (44.85) 80 (5.04) 0.5 (3.45) 6
3 (76.2) 300 (18.93) 13.1 (90.39) 240 (15.14) 8.3 (57.27) 150 (9.46) 1.1 (7.59) 3
2 (30) 160 (10.08) 9.2 (63.48) 128 (8.07) 6.1 (42.09) 80 (5.04) 2.6 (17.94) 3
3 (80) 320 (20.19) 13.4 (92.46) 250 (15.77) 8.9 (61.36) 160 (10.08) 4.2 (28.98) 5
4 (100) 500 (31.54) 9.6 (66.24) 400 (25.23) 6.3 (43.47) 250 (15.77) 3.5 (24.15) 5
6 (150) 1000 (63.09) 9.4 (64.86) 800 (50.46) 5.8 (40.02) 500 (31.54) 2.5 (17.25) 4
8 (203) 1600 (100.94) 12.0 (82.8) 1280 (80.75) 7.8 (53.82) 800 (50.46) 4.0 (27.60) 3
2 (50) 160 (10.08) 4.5 (31.05) 128 (7.57) 2.8 (19.32) 80 (5.04) 1.0 (6.9) 5
3 (80) 350 (22.37) 4.6 (31.74) 280 (17.66) 3.0 (20.69) 175 (11.04) 1.2 (8.3) 4
4 (100) 600 (37.85) 3.5 (24.15) 480 (30.28) 2.1 (14.5) 300 (18.93) 0.8 (5.5) 4
6 (150) 1250 (78.86) 3.5 (24.15) 1000 (69.09) 2.0 (13.8) 625 (39.43) 0.7 (4.9) 4
Figure 5-1 Friction Loss of Head Chart, Coefficient of Flow (C) = 140 (derived from the Hazen and Williams Formula)
ASPE Data Book — Volume 2
Chapter 5 — Cold-Water Systems
Figure 5-1 (M) Friction Loss of Head Chart, Coefficient of flow (C) = 140 (derived from the Hazen and Williams Formula)
119
120 ASPE Data Book — Volume 2
This formula is most accurate for the flow of quently used fixtures should not be reflected in
water in pipes larger than 2 in. (5 cm) and at the total demand.
velocities less than 10 fps (3 m/s).
After the designer has determined which fix-
Equation 5-1 yields accurate results only tures to include in the water demand calcula-
when the kinematic viscosity of the liquid is about tion, the maximum demand can be obtained.
1.1 centistokes, which is the case of water at Fixture unit (fu) values for each fixture can be
60°F (15.6°C). However, the kinematic viscosity assigned by using Table 5-5 and a total fu value
of water varies with temperature, from 1.8 can be obtained by adding the fu values of all
centistokes at 32°F (0°C) to 0.29 centistokes at water-using fixtures with a normal domestic di-
212°F (100°C); therefore, the tables are subject versity. The total fu value can be converted into
to this error, which may increase the friction loss a gpm (L/s) flow rate by using Table 5-6 or Fig-
by as much as 20% at 32°F (0°C) and decrease it ures 5-2 or 5-3, each of which includes a diver-
by as much as 20% at 212°F (100°C). Values of sity factor.
C, for various types of pipe, are shown in Table
The demand flow rates of all constant-use
5-4, together with the corresponding multipliers
fixtures must be added to this flow rate. A con-
that should apply to the values of the head loss, f.
stant-use fixture uses water continuously and
Figure 5-1 shows the friction loss of head does not have normal diversity. Air-conditioning
chart, C = 140, derived from the Hazen-Williams cooling towers, booster pumps, commercial laun-
formula (Equation 5-1). Figure 5-2 illustrates the dry or dishwashing equipment, lawn sprinklers,
conversion of fixture units to gallons per minute and industrial processes are examples of con-
(liters per second). stant-use fixtures. Any such equipment must be
figured separately and added to the gpm (L/s)
Factors Affecting Sizing flow rate obtained from the conversion of all fix-
ture units. This combined figure is the peak de-
The three factors affecting the sizing of a water mand flow rate for the project. (Note: Fixtures
line are the demand flow rate (gpm) (L/s), the that are timed to operate during “off” hours
velocity (fps) (m/s), and the pressure available should not be added.)
for friction loss. The fixture-unit listings in Table 5-5 are for
Demand The first factor, flow rate, is the water the total water consumption of the fixture. For
demand of the system, in gpm (L/s). There is a the purposes of sizing either the hot or cold-wa-
vast difference in the water demand flow rates of ter line, the fixture-unit loading for a fixture that
flush valves in different types of occupancy. For uses both hot and cold water would be 75% of
example, ten water closets with flush valves in the total value. The 75% figure applies only to
an apartment building may have a demand flow fixtures served by hot and cold water. It does
rate of 60 gpm (3.8 L/s), while ten water closets not apply to single-service fixtures, such as wa-
with flush valves in a public school may have a ter closets, urinals, and dishwashers.
demand flow rate of 90 gpm (5.7 L/s). The judg-
Velocity The second factor affecting the sizing
ment and experience of the designer plays an
of a water line is velocity. A maximum velocity of
important part in accommodating such differ- 15 fps (4.6 m/s), which is suggested by some
ences in the design of water systems. model plumbing codes, is much too high for many
Another problem encountered in establish- installations. A velocity above 6 or 7 fps (1.8 or
ing flow rates is the practice of counting fixtures 2.1 m/s) normally creates noise. Also, depend-
that are not normally in use. For example, a ser- ing on the piping material used and the tem-
vice sink in an office building is normally used perature, hardness, and pH of the water,
only by the janitors at night; therefore, it should velocities above 4 fps (1.2 m/s) can cause ero-
not be counted as a fixture in the total demand. sion of the piping material.
Hose bibbs are other fixtures that should not be
Another justification for lower velocities in a
figured at 100% of their number. For example,
system is water hammer. Water hammer is the
the systems of large buildings may have many
pounding force created by the sudden starting
hose bibbs installed but only a few will be oper-
or stopping of water flow, which can be caused
ated simultaneously. Individual branch lines by quick-opening or closing valves. The impact
should be sized to handle all the fixtures on the of water hammer is directly proportional to the
branch; however, the presence of these infre-
change in velocity and is equal to approximately
60 times the velocity change. For instance, if
122 ASPE Data Book — Volume 2
water traveling at 15 fps (4.6 m/s) is stopped sure. If the maximum pressure is above 80 psi,
suddenly, the increase in pressure within the and a pressure-regulating device is installed, the
pipe line will be approximately 900 psi (6205.3 pressure regulator will introduce an additional loss
kPa). This increased pressure can do consider- in the piping system when the water system is at
able damage to piping systems and connected minimum pressure. The water pressure should
equipment. be determined from a fire-hydrant flow test, which
is taken as close to the site as possible and in-
Pressure The third factor affecting the sizing of a cludes static and residual pressures at a flow rate.
water line is the pressure available for friction
loss. The first step in ascertaining pressure avail- Many model plumbing codes state that, if a
able for friction loss is determining (from the lo- pressure-regulating device is installed, the avail-
cal water department) the maximum and able pressure must be considered as 80% of the
minimum water pressures and flow rate to be reduced pressure setting. Spring-operated, pres-
encountered at the project site. The maximum sure-regulating devices have a fall-off pressure
and minimum pressures may be nearly the same that is below the system pressure setting. Many
or they may vary greatly; care must be taken to engineers design a system incorporating the fall-
handle the high pressure as well as the low pres- off pressure of the equipment they are using;
Table 5-4 Surface Roughness Coefficient (C) Values for Various Types of Pipe
Values of C
Range Average Value Value Commonly
(High = Best, smooth, well-laid for Good, Used for
Type of Pipe Low = Poor or corroded) Clean, New Pipe Design Purposes
Table 5-5 Demand Weight of Fixtures, however, the 80% factor is a rule of thumb that
in Fixture Unitsa should not apply to an engineered system.
If the available water pressure at a project
Weight Minimum
site is high enough to require the use of a
(fixture units)c Connections,
in. (mm) pressure-regulating device, the pressure-regulat-
ing valve is considered the starting point of the
Fixture Typeb Cold Hot
system for the purposes of calculation.
Private Public Water Water
The next step in obtaining the pressure avail-
Bathtubd 2 4 2 (13) 2 (13) able for friction loss is to determine the residual
pressure required at the governing fixture or
Bedpan washer — 10 1 (25) — appliance (not necessarily the farthest fixture).
Bidet 2 4 2 (13) 2 (13) “Residual pressure” is the pressure required at
the fixture for it to operate properly with water
Combination sink flowing. Normally, but not always, 8 psi (55.2
and tray 3 — 2 (13) 2 (13) kPa) is required for a flush-tank system and 15
Dental unit or cuspidor — 1 a (10) — psi (103.4 kPa) is required for a flush-valve sys-
tem. Some flush-valve fixtures require 20 or 25
Dental lavatory 1 2 2 (13) 2 (13) psi (137.9 or 172.4 kPa); some water closets re-
Drinking fountain 1 2 a (10) — quire 40 psi (275.8 kPa); commercial dishwash-
ers require 20 or 25 psi (137.9 or 172.4 kPa). It
Kitchen sink 2 4 2 (13) 2 (13) is evident, then, that the residual pressure
Lavatory 1 2 a (10) a (10) should be figured as the actual pressure needed
at the governing fixture.
Laundry tray (1 or 2
compartments) 2 4 2 (13) 2 (13) The third step is to determine the static pres-
sure loss required to reach the governing fixture
Shower, each headd 2 4 2 (13) 2 (13) or appliance. The static loss (or gain) is figured
Sink, service 2 4 2 (13) 2 (13) at 0.433 psi/ft (9.8 kPa/m) of elevation differ-
ence, above or below the water main. The differ-
Urinal, pedestal — 10 1 (25) — ence in elevation is usually a pressure loss to
Urinal (wall lip) — 5 2 (13) — the system, as fixtures are normally at a higher
elevation than the source. If the fixture is lower
Urinal stall — 5 w (20) — than the source, there will be an increase in pres-
Urinal with flush tank — 3 — — sure and the static pressure is added to the ini-
tial pressure.
Wash sink, circular or
multiple (each set of Another pressure loss is created by the wa-
faucets) — 2 2 (13) 2 (13) ter meter. This loss of pressure, for a disc type
meter, can be determined from Figure 5-4 or from
Water closet: the manufacturer’s flow charts. The flow is de-
Flush valve 6 10 1 (25) — termined from charts indicating the total flow
rate, in gpm (L/s), the size and type of the meter,
Tank 3 5 a (10) — and the pressure drop for the corresponding flow.
The loss is given in pounds per square inch (psi)
a For supply outlets likely to impose continuous demands, esti-
and kilopascals (kPa). The selection of meter size
mate the continuous supply separately and add to the total demand
for fixtures. is very important in the final sizing of the piping
b For fixtures not listed, weights may be assumed by comparing system and is one variable the designer can con-
the fixture to a listed one then using water in similar quantities and trol. Many other factors, such as the height of
at similar rates. the building, city water pressure, and require-
c The given weights are for the total demand of fixtures with both ments for backflow protection or water treatment,
hot and cold-water supplies. The weights for maximum separate are dictated by codes or by the particular situa-
demands may be taken as 75% of the listed demand for the sup-
ply. tion. The designer must review the system very
d A shower over a bathtub does not add a fixture unit to the group. closely prior to the selection of a meter size. Usu-
ally, the larger the meter, the higher the initial
installation price and monthly charge. On the
124 ASPE Data Book — Volume 2
Table 5-6 Conversions—Gallons per Minute (Liters per Second) to Fixture Units
other hand, a larger meter may mean a smaller- in available water supply pressures as an area
sized piping system, which might prove to be incurs growth.
more economical in the long run. These two fac-
As previously determined, the governing ap-
tors are evaluated by the designer and economic
pliance in the example in Figure 5-5 is the dish-
considerations guide the selection. Furthermore,
washer. For the same example, assume that the
if a system does not have ample pressure, a
minimum incoming water pressure is 60 psi
means of preserving the available pressure is to
(413.7 kPa). To determine the pressure available
use a larger meter, thereby decreasing pressure
for friction, start with 60 psi (413.7 kPa) and
loss. This fact may well enable the designer to
subtract 3 psi (20.7 kPa) for the meter loss, 10
eliminate the use of a water-pressure booster
psi 69 kPa) for the softener, 5 psi (34.5 kPa) for
system, thereby substantially reducing the
the water-heater coil, and 25 psi (172.4 kPa) re-
plumbing system costs.
sidual for the dishwasher. This leaves a remain-
The last step is to determine the other pres- der of 17 psi (117.2 kPa), which is the pressure
sure losses encountered between the meter and available for friction. The losses for the backflow
the governing fixture. These could be caused by preventer and the static do not occur on the line
a water softener, a backflow preventer, a filter, between the meter and the governing fixture or
or any other device that creates a pressure loss appliance; therefore, they are not included in the
in the system. calculations at this time. Only losses that occur
on the line between the meter and the governing
The “governing fixture” or appliance is the
fixture or appliance are to be included in the ini-
device that has the highest total when the re-
tial calculations to determine the pressure avail-
sidual pressure, static pressure, and all other
able for friction. The other losses will enter into
pressure losses are added. Take, for example,
subsequent calculations.
the system shown in Figure 5-5. To find the gov-
erning fixture or appliance, determine which After obtaining the pressure available for fric-
device requires the most pressure. Knowing that tion, the next step is to calculate the “average
the meter loss is the same for all parts of the pressure drop.” This is the pressure available for
system, it can be temporarily ignored. Going from friction divided by the equivalent length of the
the meter to the flush-valve water closet, there run. The quotient is multiplied by l00 to obtain
are 15 psi (103.4 kPa) residual and no static loss an answer in terms of loss in psi/100 ft (kPa/
for a total of 15 psi (103.4 kPa). As a total going l00 m). In determining the equivalent length of
through the backflow preventer, there are 16 psi run, an allowance must be made for fittings. This
(110.3 kPa) residual and 8.66 psi (59.7 kPa) static can be determined from Table 5-7 or by adding a
for a total loss of 24.66 psi (170 kPa). Going to percentage to the developed length. The average
the dishwasher, there is a total of 40 psi (275.8 pressure drop is an average loss over the system
kPa)—25 psi (172.4 kPa) residual plus 5 psi (34.5 and should be used only as a guide in sizing
kPa) loss through the water heater plus 10 psi piping.
(69 kPa) loss through the softener. Therefore, the
Part of the system can be designed to exceed
dishwasher is the governing fixture, for it has
the average pressure drop, while another part is
the highest total when the residual, static, and
designed to be less than the average. The aver-
other losses are added.
age pressure drop can be exceeded—as long as
Summarizing the steps, all the system needs the total pressure available for friction is not
or losses are subtracted from the minimum wa- exceeded. The average pressure drop calculation,
ter pressure. The remainder is the pressure avail- which is made initially, pertains only to the line
able for friction, defined as the total energy (or from the meter to the governing fixture or appli-
force) available to push the water through the ance. Care should be taken to account for the
pipes to the governing fixture or appliance. How average pressure drop calculations for the other
this force is used is up to the designer, who may lines. The branches off the main line should be
decide to use it evenly over the entire system, as sized on a different pressure-loss basis, or the
an average pressure loss, or unevenly over the branches closest to the meter may take pres-
system. In designing the system, as long as the sure away from the farthest branches. Table 5-8
designer does not exceed the pressure available shows typical flow and pressure required during
for friction, the system will work. A certain flow for various fixtures.
amount of pressure may be held in reserve, how-
ever, to allow for aging of the piping or decreases
126 ASPE Data Book — Volume 2
precisely sized pipe is obtained to give a pres- each fixture is used up as friction loss, it tends
sure loss (due to friction) of exactly 10 psi/100 ft to cause more water than necessary to flow
(226.2 kPa/l00 m), the pressure loss in this sec- through the branches to use the excess avail-
tion is 1 psi (6.9 kPa) and the pressure for fric- able pressure.
tion at point B is 9 psi (62.1 kPa). In section
Method B illustrates the ideal system. All the
K–L, at point L, there is 0 pressure left for fric-
available frictional pressure in each of the
tion. This is the governing fixture.
branches is used. In actual practice, this method
The next tabulation illustrates the sizing of can not be utilized. The average pressure loss in
branches (using a different friction-loss basis each section is very high, far higher than is nor-
than was used for the main). mally accepted in modern construction. Many
engineers and designers would be concerned with
10 psi (69 kPa) available for friction loss; long-
the high pressure loss as well as with the high
est run: A–L, 100 ft (30.5 m); average pressure
velocity shown by this example.
drop: (10 × 100)/100 = 10 psi/100 ft (226.2 kPa/
100 m). Method C is a modified header system. The
main was sized on the average pressure drop of
Method A uses the same average pressure
the system and the branches sized on their al-
loss in the branches as was used in the line to
lowable frictional pressure drop. At section M–J,
the governing fixture. The pressure available for
the total allowable pressure drop over the entire
friction at the end of each branch is not 0. At
system (point A to point M) is 10 psi (69 kPa).
point M, it is 1 psi (6.9 kPa); at point R, it is 5 psi
Point M has an equivalent length of 90 ft (27.4
(34.5 kPa); and at point U, it reaches a maxi-
m) from point A. This gives an average pressure
mum of 8 psi (55.2 kPa). Unless the pressure to
1.26
1.89
2.52
3.15
3.78
5.04
6.30
12.6
18.9
25.2
37.8
50.4
63.0
20 137.9
Pressure Loss, pounds per inch squared
16 110.3
/2"
3"
"
"
2"
6"
5/8
3/4
4"
1-1
1"
3 20.7
2 13.8
1 6.9
4 5 6 7 8 9 10 20 30 40 50 60 80 100 200 300 400 600 800 1000
Table 5-7 Allowance for Friction Loss in Valves and Threaded Fittings
Equivalent Length of Pipe for Various Fittings (ft)
Diameter 90° 45° Coupling
of Fitting Standard Standard Standard or Straight Gate Globe Angle
(in.) Elbow Elbow T 90° Run of T Valve Valve Valve
Note: Allowances based on nonrecessed threaded fittings. Use ½ the allowances for recessed threaded fittings or streamline solder fittings.
Table 5-7 (M) Allowance for Friction Loss in Valves and Threaded Fittings
Equivalent Length of Pipe for Various Fittings (m)
Diameter 90° 45° Coupling
of Fitting Standard Standard Standard or Straight Gate Globe Angle
(mm) Elbow Elbow T 90° Run of T Valve Valve Valve
Note: Allowances based on nonrecessed threaded fittings. Use ½ the allowances for recessed threaded fittings or streamline solder fittings.
Chapter 5 — Cold-Water Systems 129
Table 5-8 Flow and Pressure Required for Various Fixtures during Flow
Fixture Pressure, psi (kPa)a Flow, gpm (L/s)
METHOD A
Developed Developed Friction Friction Loss Total Pressure Pressure at End
Length in Section, Length from Loss, psi/100 ft in Section, Loss from Friction, of Section for
Section ft (m) Point A, ft (m) (kPa/100 m) psi (kPa) psi (kPa) Friction, psi (kPa)
METHOD B
Developed Developed Pressure Friction Friction Pressure
Length in Section, Length from at Start, Loss, psi/100 ft in Section, at End,
Section ft (m) Point A, ft (m) psi (kPa) (kPa/100 m) psi (kPa) psi (kPa)
METHOD C
Friction Friction Pressure Friction Friction Pressure
Loss, psi/100 ft in Section, at End, Loss, psi/100 ft in Section, at End,
Section (kPa/100 m) psi (kPa) psi (kPa) (kPa/100 m) psi (kPa) psi (kPa)
M–J 20 (452.4) 2 (13.8) 0 (0) 11.1 (251.1) 1.1 (7.6) 0.90 (6.2)
N–H 30 (678.6) 3 (20.7) 0 (0) 12.5 (282.8) 1.25 (8.6) 1.75 (12.1)
P–G 40 (904.8) 4 (27.6) 0 (0) 14.3 (323.5) 1.43 (9.9) 2.57 (17.7)
Q–F 50 (1131) 5 (34.5) 0 (0) 16.6 (375.5) 1.66 (11.5) 3.34 (23)
R–E 60 (1357.2) 6 (41.4) 0 (0) 20 (452.4) 2 (13.8) 4 (27.6)
S–D 70 (1583.5) 7 (48.3) 0 (0) 25 (565.5) 2.5 (17.2) 4.5 (31)
T–C 80 (1809.7) 8 (55.2) 0 (0) 33.3 (753.3) 3.33 (23) 4.66 (32.1)
U–B 90 (2035.9) 9 (62.1) 0 (0) 50 (1131) 5 (34.5) 4 (27.6)
drop of 11.1 psi (7.6 kPa) and an unused fric- erning fixture or appliance.
tional pressure of 0.9 psi (6.2 kPa). By going 4. The static pressure loss to get to the govern-
through all the branches in the same manner,
ing fixture or appliance.
one can see that the unused frictional pressure
varies from 0.9 psi (6.2 kPa) to a maximum of 5. The meter loss.
4.66 psi (32.1 kPa). These pressures are far less 6. Other losses between the meter and the gov-
than those resulting from Method A and the av- erning fixture or appliance.
erage pressure drops are far less than those re-
sulting from Method B. Consequently, Method C 7. The pressure available for friction.
is the one most widely used by designers. In ac- 8. The average pressure drop from the meter to
tual practice, it is not necessary to calculate the the governing fixture or appliance.
average pressure drop for each branch; usually,
the branches are close together and the changes 9. The average pressure drop for the other sys-
in the average pressure drop are very small. tems.
10. The size of the line from the meter to the
The last step is to take advantage of all avail-
governing fixture or appliance.
able pressure. For example, a water heater could
be located on the roof of a building. If the water 11. The size of the branch line.
system was designed to have a residual pres- For the convenience of the designer in sizing
sure on the roof of 15 psi (103.4 kPa), then the water systems, the following tables and figures
hot water piping system can be sized with a static are provided:
pressure gain available, to be used for friction
loss in the hot water piping. Another example of • Table 5-9. Water pipe sizing, fixture units vs.
utilizing available pressure is an installation with psi/100 ft (kPa/100 m), Type L copper tub-
a combination of flush valves and flush-tank ing.
water closets sized on the basis of a flush-valve • Table 5-10. Water pipe sizing, fixture units
system having a residual pressure of 15 psi vs. psi/100 ft (kPa/l00 m), galvanized, fairly
(103.4 kPa). Within this system, the branches rough pipe.
that have only flush-tank fixtures have an addi-
tional 7 psi (48.3 kPa) of pressure, which can be • Figure 5-7. Pipe sizing data, copper tubing,
used for friction. The 7 psi (48.3 kPa) is the dif- smooth pipe.
ference between the 15 psi (103.4 kPa) and 8 psi • Figure 5-8. Pipe sizing data, fairly smooth
(55.2 kPa) residual pressures. pipe.
• Figure 5-9. Pipe sizing data, fairly rough pipe.
Velocity Method Another method designers
use to size water piping is the velocity method. • Figure 5-10. Pipe sizing data, rough pipe.
The average pressure drop available for friction
is calculated and, if it is greater than 7 or 8 psi/
100 ft (158.4 or 181 kPa/100 m), the lines are
WATER HAMMER
sized on the basis of a 5 or 6-fps (1.5 or 1.8 m/s)
velocity. In this method, the main line is conser- “Water hammer” is the term used to define the
vatively sized and the short branches may slightly destructive forces, pounding noises, and vibra-
exceed the average pressure drop. However, the tions that develop in a piping system when a
total pressure drop of the system does not ex- column of noncompressible liquid (water) flow-
ceed the allowable pressure loss for friction. ing through a pipeline at a given pressure and
velocity is stopped abruptly. The surge pressure
(or pressure wave) generated at the point of im-
Summary pact or stoppage travels back and forth through
the piping system until the destructive energy is
The following items must be determined and cal-
dissipated in the piping system. This violent ac-
culated when sizing a system:
tion accounts for the piping noise and vibration.
1. The maximum flow rate of the system.
The common cause of shock is the quick clos-
2. The maximum and minimum water pressure ing of electrical, pneumatic, spring-loaded valves
in the main. or devices, as well as the quick, hand closure of
3. The residual pressure required at the gov- valves or fixture trim. The valve closure time is
132 ASPE Data Book — Volume 2
directly related to the intensity of the surge System Protection and Control
pressure.
Water hammer arresters prolong the life and
service of piping, valves, fittings, trim, equipment,
Shock Intensity
apparatus, and other devices that are part of, or
Quick valve closure may be defined as a closure connected to, a water-distribution system.
time equal to or less than 2L/a seconds, where To reduce shock pressure and confine its
“L” is the length of pipe (ft) (m) from the point of action to the section of piping in which it oc-
closure to the point of relief (the point of relief is curs, a suitable means of control must be pro-
usually a larger pipe riser or main or a water vided to absorb and dissipate the energy causing
tank), and “a” is the velocity of propagation of the shock. Water hammer arresters provide a
elastic vibration in the pipe (fps). The expression diaphragm that moves with the pressure fluc-
“2L/a” is the time interval required for the pres- tuations, absorbing the shock wave. Air or an-
sure wave to travel from the point of closure to other gas is the most effective medium to use for
the relief point and back to the point of closure. this purpose since it is highly compressible,
Maximum pressure rise can be calculated by thereby offering the maximum displacement
the following, known as Joukowsky’s formula: cushion for absorbing the shock.
Equation 5-2 Air chambers The air chamber has been uti-
lized for controlling shock for many years. The
wav unit consists of a capped piece of pipe having
Pr =
144g the same diameter as the line it serves; its length
where ranges from 12 in. to 24 ft (304.8–609.6 mm).
The air chamber is constructed in several differ-
Pr = Pressure rise above flow pressure, psi ent shapes.
(kPa)
Figure 5-11 shows a few examples of air
w = Specific weight of liquid, lb/ft3 (kg/m3)
chambers. Plain air chambers, Figure 5-11(a) and
a = Velocity of pressure wave, fps (4000– (b), are generally placed on the supply lines to
4500 average for water) (m/s [1219– fixtures or equipment. A standpipe type of air
1372 average]) chamber, Figure 5-11(c), is generally placed on
v = Change in flow velocity, fps (m/s) a piping main. A rechargeable type of air cham-
ber, Figure 5-11(d), is generally placed at the end
g = Acceleration due to gravity, 32 ft/s2
of a branch line or on a piping main.
(10 m/s2)
The air chambers shown are made of pipe
This action produces a pressure rise of ap- and fittings. Unless devices are of the correct
proximately 60 times the change in velocity. En- size and contain a prescribed volume of air, how-
gineers generally employ a velocity between 5 and
ever, they cannot be regarded as suitable even
10 fps (1.5 and 3.1 m/s), which may produce a
for the temporary control of shock.
shock pressure of 300–600 psi (2068–4137 kPa).
The resultant water-hammer shock wave trav- Most valves and fittings used in plumbing
els back and forth in the piping, between the water-distribution systems are designed and con-
point of quick closure and the point of relief, at a structed for normal maximum working pressures
rate of 4000–4500 fps (1219–1372 m/s). of 150 psig (1034 kPa). Therefore, unless an air
chamber can reduce shock pressures to some
Although noise is generally associated with
degree less than 250 psig (1724 kPa), serious
the occurrence of water hammer, water hammer
damage to the valves, fittings, and other compo-
can occur without audible sound. Quick closure
nents of the piping system may result. The com-
always creates some degree of shock—with or monly used air chamber, even when correctly
without noise. Therefore, the absence of noise sized, controls shocks only temporarily after its
does not indicate that water hammer or shock is
initial installation.
nonexistent in a water-distribution system.
Although a correctly sized air chamber tem-
porarily controls shock to within safe limits of
pressure, its performance is effective only while
Chapter 5 — Cold-Water Systems 133
Table 5-9 Water Pipe Sizing—Fixture Units vs. psi/100 ft (kPa/100 m),
Type L Copper Tubing
15 69 11 81 285
1.0 0 2 6 12 21 58 155 3.2 1 6 14 26 50 183 421
(22.6) (72.4)
17 73 12 87 309
1.1 0 2 7 13 22 62 170 3.4 1 6 15 28 52 194 441
(24.9) (76.9)
20 82 13 95 336
1.2 0 3 7 14 23 67 185 3.6 1 6 15 29 55 205 460
(27.2) (81.4)
23 91 14 102 365
1.3 0 3 7 15 24 74 199 3.8 1 6 16 30 57 215 479
(29.4) (86)
26 100 15 106 390
1.4 0 3 8 15 25 81 213 4.0 1 6 16 31 58 225 500
(31.7) (90.5)
28 109 16 116 410
1.5 0 3 8 16 27 86 226 4.2 1 7 16 32 61 236 517
(33.9) (95)
31 120 18 124 430
1.6 0 3 8 17 28 93 241 4.4 1 7 17 34 63 245 533
(36.2) (99.5)
33 130 5 20 131 448
1.7 0 4 9 17 30 98 252 4.6 2 7 18 35 65 253 549
(38.5) (104.1)
36 140 6 21 139 466
1.8 0 4 9 18 31 105 264 4.8 2 7 19 36 68 263 564
(40.7) (108.6)
39 150 6 22 145 484
1.9 0 4 10 19 32 111 277 5.0 2 7 19 37 72 271 580
(43) (113.1)
42 161 7 24 153 504
2.0 0 4 10 20 33 115 287 5.2 2 8 19 38 75 280 597
(45.2) (117.6)
6 48 183 7 25 163 526
2.2 0 4 11 21 36 127 312 5.4 2 8 20 40 79 289 614
(49.8) (122.2)
7 53 205 8 26 171 *549
2.4 1 4 12 22 39 138 337 5.6 2 8 20 42 83 298 630
(54.3) (126.7)
8 59 225 8 27 177 *570
2.6 1 4 12 23 42 150 360 5.8 2 8 21 43 85 306 646
(58.8) (131.2)
9 66 245 9 29 185 *591
2.8 1 5 13 24 45 160 380 6.0 2 8 21 44 88 314 662
(63.3) (135.7)
10 74 265 9 30 199 *610
3.0 1 5 13 25 47 171 401 6.2 2 9 22 45 92 323 676
(67.9) (140.3)
(Continued)
134 ASPE Data Book — Volume 2
(Continued)
Chapter 5 — Cold-Water Systems 135
Table 5-10 Water pipe sizing fixture units versus psi/100 ft. (kPa/100 m),
Galvanized fairly-rough pipe
Pipe Size, Pipe Size,
in. (mm) in. (mm)
Pressure Pressure
Loss, ½ ¾ 1 1¼ 1½ 2 2½ Loss, ½ ¾ 1 1¼ 1½ 2 2½
psi/100 ft (12.7) (19.1) (25.4) (31.7) (38.1) (50.8) (63.5) psi/100 ft (12.7) (19.1) (25.4) (31.7) (38.1) (50.8) (63.5)
(kPa/100 m) Fixture Unitsa (kPa/100 m) Fixture Unitsa
8 37 40 162
1.0 0 1 4 8 16 42 107 3.2 0 3 9 19 35 112 288
(22.6) (72.4)
9 42 6 43 174
1.1 0 1 5 9 17 45 115 3.4 0 3 10 20 36 118 302
(24.9) (76.9)
11 46 7 46 186
1.2 0 1 5 10 19 48 124 3.6 0 4 10 20 38 123 315
(27.2) (81.4)
12 51 7 49 198
1.3 0 1 6 11 20 51 133 3.8 0 4 11 21 40 129 329
(29.4) (86)
13 55 8 52 210
1.4 0 2 6 11 20 54 143 4.0 1 4 11 21 42 135 343
(31.7) (90.5)
14 62 9 54 221
1.5 0 2 6 12 21 56 153 4.2 1 4 12 22 43 141 356
(33.9) (95)
15 67 10 58 238
1.6 0 2 6 12 22 58 162 4.4 1 5 12 23 45 147 369
(36.2) (99.5)
16 74 10 62 345
1.7 0 2 6 12 23 60 171 4.6 1 5 12 23 46 153 380
(38.5) (104.1)
18 80 10 66 256
1.8 0 2 6 13 23 63 180 4.8 1 5 12 24 48 160 391
(40.7) (108.6)
20 85 11 71 265
1.9 0 2 7 13 24 66 189 5.0 1 5 13 24 49 167 403
(43) (113.1)
22 90 12 75 278
2.0 0 3 7 14 25 70 190 5.2 1 6 13 25 51 174 415
(45.2) (117.6)
25 102 13 79 290
2.2 0 3 7 15 26 77 215 5.4 1 6 13 26 52 180 426
(49.8) (122.2)
27 112 13 82 302
2.4 0 3 7 15 28 85 231 5.6 1 6 14 27 54 185 436
(54.3) (126.7)
30 124 14 85 314
2.6 0 3 8 16 30 92 245 5.8 1 6 14 27 55 191 446
(58.8) (131.2)
33 136 14 89 329
2.8 0 3 8 17 32 99 259 6.0 1 6 15 28 56 197 455
(63.3) (135.7)
36 148 15 93 343
3.0 0 3 9 18 33 105 275 6.2 1 6 15 29 57 202 465
(67.9) (140.3)
(Continued)
Chapter 5 — Cold-Water Systems 137
(Continued)
138 ASPE Data Book — Volume 2
it retains its initial charge of air. Air-chamber In most installations where there are several
requirements are shown in Table 5-11. fixtures, usually only one fixture valve will be closed
The air charge can be depleted during the at a time. Occasionally, however, two or more fix-
flow cycle since water is drawn from all direc- ture valves may be closed at the same instant.
tions during flow. Moreover, the entrapped air is Table 5-12, “Sizing and Selection of Water-Hammer
also diminished by turbulence. During this pro-
cess the water absorbs the air, and as the unit
becomes waterlogged, it loses its ability to ab-
sorb shock.
Arresters,” takes into consideration all design fac- To prevent the harboring of Legionella
tors, including simultaneous usage, pipe size, pneumophila, bellows containing rubber should
length, flow, pressure, and velocity. not be used.
• Pressure vacuum breakers (PVB). vacuum breaker should be used only to isolate a
• Check valves with vent port (CVB). severe hazard if area isolation is provided. Where
a moderate hazard exists, a double-check valve
The theory of backflow and back-siphonage assembly, or pressure or atmospheric vacuum
and the devices for their prevention are described breaker may be used. Where a minor hazard ex-
in Volume 4, Chapter 9, of the ASPE Data Book ists, a pressure or atmospheric vacuum breaker
(forthcoming). Refer to local codes and standards or check valves with vent port (no test cocks)
before making selections. may need to be installed.
Toxicity and probability of occurrence illus-
Assessment of Hazard
trate the relationship between assessment of
The correct application of devices depends on the hazard and application of devices. Because of
correct assessment of the degree of hazard, on the subjective nature of assessing hazard, such
whether back pressure or back-siphonage will oc- illustrations cannot be used as a strict guide,
cur, and on knowledge of the operation of various providing a fixed answer for all circumstances.
types of approved backflow-prevention device. Instead, past experience and local code require-
ments must also be used as guides. Such past
In applying the recommendations outlined experience was the basis of Tables 5-13 and 5-14.
in this section, three degrees of hazard must be
considered: severe, moderate, and minor. They The requirement of protection increases as a
are defined as follows: function of both an increase in the probability
that backflow or back-siphonage will occur and
1. Severe. A cross connection or probable cross an increase in the toxicity or possible toxicity of
connection involving any substance in suffi- a potential source of contamination. Where it is
cient concentration to cause death or spread highly probable that backflow or back-siphon-
disease or illness or containing any substance age will occur, say from a standpipe in a tall
that has a high probability of causing such apartment building, the need for a backflow-pre-
an effect. vention device is low if the hazard of the poten-
2. Moderate. A cross connection or probable tial source of contamination (sinks, water closets,
cross connection involving any substance etc.) becoming toxic is very low. The converse is
that has a low probability of becoming a se- also true, however, where a known health haz-
vere hazard and would constitute a nuisance ard exists, the tendency is to be conservative
or be aesthetically objectionable if introduced when selecting a backflow-prevention device
into the domestic water supply. (RPBD used in place of DVC). The risk factor for
a health hazard is usually of greater concern than
3. Minor. An existing connection, or a high prob- the probability of backflow or back-siphonage in
ability of a connection being made, between the selection of a device.
the domestic water pipe and any pipe, equip-
ment, vat, or tank intended for carrying or
holding potable water that has a low prob- Premise Isolation
ability of becoming contaminated with any
In addition to installing backflow-prevention de-
substance.
vices at the source of potential contamination, it
The application of backflow and back-sipho- may be necessary, or required by code, to install
nage prevention devices is related to the prob- a backflow-prevention device on the water-ser-
ability of contamination as well as the recognition vice pipe to isolate an entire area or premise.
of an existing health hazard. For the assessment This additional protection for the purveyor’s wa-
of probability, consideration must be given to the ter system is warranted if the potential health
possibility of changes being made to piping, im- hazard is severe, or if a high probability exists
proper use of equipment, negligence of the cus- that piping within a premise will be changed. If
tomer, etc. inspection on private property is restricted, the
only protection for the purveyor’s water system
Where a severe hazard exists, an air-gap
is the installation of a backflow-prevention de-
separation or a reduced-pressure-principle,
vice on the water-service pipe.
backflow-prevention device should be used be-
cause these two devices offer the highest known Whenever possible, in-plant isolation is pre-
degree of protection. An atmospheric or pressure ferred over premise isolation because it protects
146 ASPE Data Book — Volume 2
aWhere a higher hazard exists (due to toxicity or health hazard), additional area protection with RPBD is required. See Table 5-14 for
additional information.
Chapter 5 — Cold-Water Systems 147
Table 5-14 Guide to the Assessment of Facility Hazard and Application of Devices—
Containment of Premise
Recommended Device
Description of Premise Assessment ot Hazard on Water-Service Pipe
Table 5-15 Minimum Flow Rates and Size of Minimum Area of RPBD
capacity. Water then flows from the tank to the taining the water level in this way ensures a rela-
waterlines by gravity action. tively constant water pressure regardless of
whether demand is at a low or peak condition.
Three approaches may be used to determine
The following piping connections are required at
tank capacity for a building:
the tank:
1. Rule of thumb. An arbitrary tank capacity
• Water supply to the tank.
equal to 30 times pump capacity (gpm) (L/s)
is recommended by some authorities. This • Water supply to the system.
criterion theoretically provides a building • Overflow line.
with a 30-min emergency reserve supply of
• Tank drain.
water in case of power failure or disruption
of the source of water supply. The locations of these connections on the
tank are illustrated in Figure 5-14. The system
2. Empirical. With this method, the quantity of shown is also equipped with fire-standpipe and
water required for emergency conditions is
sprinkler connections to meet local code require-
arbitrarily fixed. Based on this determina-
ments. The tank connections shown in Figure
tion, the length of time needed for pumping
5-14 provide the required water supply for each
the water before safe shutdown can be esti-
system, with the sprinkler reserve at the bot-
mated. tom, the fire-standpipe reserve at the next level,
3. Cycling of pumps. The capacity of the tank and the water storage at the top. Piping connec-
is sized so that cycling of pumps will not oc- tions to the standpipe and sprinkler systems
cur more than 6 times per hour. This trans- should be fitted with bronze strainers within the
lates to 5 min off, 5 min on. The fewer the tank to prevent any debris from entering those
cycles per hour, the less the wear and tear systems.
on motors and the less maintenance required.
Level controls are installed in the tank to start
Reducing the number of cycles, however, will
and stop pumps at low and high levels. The level
produce greater fluctuations in tank-water
control can be a float switch, pressure switch,
reserve.
electric prober, or any other acceptable device.
Selecting a tank that provides a large water
surface relative to its capacity makes it possible Tanks should be equipped with both high and
to withdraw a considerable volume of water with- low-level alarms. The low-level alarm indicates
out appreciably lowering the liquid level. Main- that the pumps are not keeping up with demand.
The high-level alarm warns that water has Dead-end service The type of service in which
reached the overflow level and is spilling to waste. the PRV is required to close bottle-tight when
When storage tanks are used for gravity feed, there is no demand on the system.
consideration must be given to the weight of the
Fall-off The amount that pressure is decreased
tank and water so proper support can be
from set pressure to meet demand. The amount
provided.
of fall-off depends on the quantity of flow—the
greater the flow, the greater the fall-off. A fall-off
Booster-Pump System of 20 psi (137.9 kPa) is considered to be the maxi-
mum allowable fall-off.
There are two ways to make a continuously run
system deliver a relatively constant system pres- No-flow pressure The pressure maintained in
sure under varying load conditions. One way is the system when the PRV is shut tight so that
to use a constant-speed pump with a pressure- high pressure at the inlet of the valve is not per-
regulating valve in the discharge piping. The mitted to enter the system.
other way is to vary the speed of the pump shaft
Reduced-flow pressure The pressure main-
either at the motor or in the coupling.
tained at the PRV outlet when water is flowing.
A variety of booster-pump systems are cur- The no-flow (closed), set-point pressure of a PRV
rently in use, with more being introduced all the is always higher than the reduced-flow (open)
time. Detailed information on the design criteria pressure. A PRV that is set to open at 45 psi
and operational characteristics of water-pressure (310.3 kPa) pressure (no-flow) would deliver a
boosting systems is given in the ASPE Pumps reduced-flow pressure of 30 psi (206.8 kPa) at
and Pump Systems Handbook. peak demand if a 15 psi (103.4 kPa) fall-off had
been selected. Then the reduced-flow pressure
at peak flow would be 30 psi (206.8 kPa).
EXCESS WATER PRESSURE
Response The capability of a PRV to respond
One of the main sources of trouble in a water- to change in outlet pressure.
distribution system is excessive pressure. Un- Sensitivity The ability of a PRV to sense a
less a piece of equipment, fixture, or operation change in pressure. If the valve is too sensitive
requires a specified high pressure, a water sys- and quick to respond, the results are over-con-
tem should not exceed a maximum of 80 psi trol and a hunting effect. Not enough sensitivity
(551.6 kPa) (check local code). To ensure this, a results in operation that is sluggish and great
pressure-regulating valve (PRV) should be in- variations in the outlet pressure.
stalled.
Set pressure That pressure, at the outlet of
The purpose of a pressure-regulating valve the PRV, at which the valve will start to open.
is to reduce water pressure from higher, supply-
main pressures to desirable and adequate flow Types of pressure-regulating valve All pres-
pressures when water is required at fixtures, sure-regulating valves fall into the following gen-
appliances, or equipment. eral categories:
• Single-seated—direct-operated or pilot-oper-
Pressure-Regulating Valves ated.
• Double-seated—direct-operated or pilot-oper-
Definitions The following are definitions of ated.
terms used in discussing, sizing, and ordering
pressure-regulating valves: Single-seated pressure-regulating valves are
used for dead-end service and when the flow to
Accuracy The degree of fall-off in the outlet be regulated is intermittent. For dead-end ser-
pressure from the set pressure at full-flow ca- vice, the valve must be able to shut tight and
pacity. Also, the capability of producing the same not permit the passage of any water when there
results for repetitive operations with identical is no demand. Double-seated PRVs are used for
conditions of flow. continuous-flow conditions. They are not suited
for dead-end service and should never be used
for this purpose.
Chapter 5 — Cold-Water Systems 153
Direct-operated PRVs tend to have a reduc- mance. Also, this type of installation reduces the
tion of the outlet pressure in direct proportion velocity of flow (there’s less pressure drop across
with the increase of the flow rate. Pilot-operated two regulators than across one), providing longer
PRVs will maintain a close fluctuation of the valve life.
outlet pressure independent of the flow rate as-
Selection of PRVs and pressure settings is
suming that the valve was sized properly.
fairly simple. The first PRV could reduce from
Sizing, selection, and installation Initial cost, 250 to 150 lb (1723.7 to 1034.2 kPa) and the
maintenance cost, and specific project require- second from 150 to approximately 50 lb (1034.2
ments regarding flow rates and pressure should to 344.7 kPa) or there could be some similar di-
determine which PRV is recommended for a par- vision. PRV size can be selected according to the
ticular application. manufacturer’s capacity tables if it is remem-
bered that each PRV should exceed the total ca-
Sizing and selection of a pressure-regulating pacity of the system.
valve can be performed after the following crite-
ria are estimated: inlet pressure, outlet pressure, Where there is a wide variation of demand
and capacity (flow rate). “Inlet pressure” is the requirements and where it is vital to maintain a
maximum pressure expected upstream of the continuous water supply as well as provide
regulating valve. “Outlet pressure” is the pres- greater capacity, “parallel installation” is recom-
sure required downstream of the regulating valve. mended. Parallel installation is the use of two or
For large-capacity systems, which may also ex- more smaller size pressure-regulating valves
perience periods of low flow, or when extreme serving a larger size supply-pipe main. This type
pressure reductions are expected, it is not ad- of installation should be employed wherever there
visable to have only one regulating valve. is a wide variation of reduced-pressure require-
ments and where it is vital to maintain a con-
A PRV sized to accommodate both small and tinuous water supply. It also has the advantage
large flows has, in general, a high noise level of providing increased capacity beyond that pro-
during operation. In addition, small flows will vided by a single valve where needed. Multiple
produce wire-drawing of the seat and possible installation improves valve performance for
chatter. widely variable demands and permits the ser-
In addition to having economic advantages, vicing of an individual valve without the com-
the proper application of pressure-regulating plete shutdown of the line, thus preventing costly
valves can greatly influence the overall perfor- shutdowns.
mance of the system. Under most circumstances, For a two-valve parallel installation, the to-
a good application can increase system perfor- tal capacity of the valves should equal or exceed
mance, reduce operating costs, and ensure a the capacity required by the system. One valve
longer life expectancy for regulators. should be set at 10 psi (69.0 kPa) higher delivery
For example, where initial pressures exceed pressure than the other. For example, assume
200 psi (1379.0 kPa) or where there is a wide that the system requires 400 gpm (25.2 L/s) and
variation between the initial pressure and the the reduced-flow pressure required is 50 psi
reduced pressure, or where the initial pressure (344.7 kPa). Select two valves, each rated at 200
varies considerably, “two-stage reduction” is ben- gpm (12.6 L/s), with one valve set at 50 psi (344.7
eficial. Two-stage reduction is the use of two PRVs kPa) and the other valve set 10 psi (69.0 kPa)
to reduce high service pressure proportionately higher at 60 psi (413.7 kPa). Thus, when low
and to eliminate an extremely wide variance be- volume is required, the higher-set valve oper-
tween the initial and reduced pressure. It is rec- ates alone. When a larger volume is demanded,
ommended where the initial pressure is 200 lb both valves open, delivering full-line capacity.
(1379.0 kPa) or more and where the ratio of ini- Another possible choice is to install two PRV
tial to reduced pressure is more than 4 to 1 (e.g., combinations of different sizes. This is practical
200 to 50 lb [1379.0 to 344.7 kPa]), or where the on larger installations where supply lines are 2
initial pressure fluctuates greatly. The advan- in. (50 mm) and larger and where there are fre-
tage of this installation is that neither valve is quent periods of small demand. The smaller PRV
subjected to an excessive range of pressure re- would have the 10-psi (69.0-kPa) higher delivery
ductions. This seems to stabilize the final reduced pressure and thus operate alone to satisfy small
pressure, ensuring close and accurate perfor- demands, such as urinals and drinking foun-
154 ASPE Data Book — Volume 2
tains. When a larger volume is demanded, the as follows (or as required by the administrative
main PRV would open to satisfy the system de- authority):
mand. For example, take an apartment building
1. Cleaning and disinfection applies to both hot
requiring 300 gpm (18.9 L/s) at 60 psi (413.7
and cold, domestic (potable) water systems
kPa). The selection might be a 4-in. (100-mm)
and should be performed after all pipes,
PRV rated for 240 gpm (15.1 L/s) (80% of total
valves, fixtures, and other components of the
maximum flow rate) and set at 60 psi (413.7 kPa)
systems are installed, tested, and ready for
and a 1½-in. (40-mm) PRV rated for 60 gpm (3.8
operation.
L/s) and set at 70 psi (472.7 kPa).
2. All domestic yard, hot and cold-water piping
Manufacturers have tables indicating recom-
should be thoroughly flushed with clean, po-
mended capacities and valve sizes for use in par-
table water prior to disinfection to remove
allel installations.
dirt and other contaminants. Screens of fau-
cets and strainers should be removed before
TESTING, CLEANING, AND flushing and reinstalled after completion of
disinfection.
DISINFECTION OF DOMESTIC,
WATER-SUPPLY SYSTEMS 3. Disinfection should be done using chlorine,
either gas or liquid. Calcium or sodium hy-
Testing pochlorite or another approved disinfectant
may be used.
Prior to disinfection, connection to faucets and
4. A service cock should be provided and lo-
equipment, and installation of pipe insulation,
cated at the water-service entrance. The dis-
the domestic water system should be hydrostati-
infecting agent should be injected into and
cally tested for leakage. A typical test for interior
through the system from this cock only.
piping is accomplished by capping all system
openings, filling the system with water, and then 5. The disinfecting agent should be injected by
pumping a static head into the system at a mini- a proportioning pump or device through the
mum of 1½ times the working pressure (100 psi service cock slowly and continuously at an
[689.5 kPa] minimum) for a period of not less even rate. During disinfection, flow of the dis-
than 2 hours. The aforementioned test require- infecting agent into the main connected to
ments are acceptable to most inspectors, but note the public water supply is not permitted.
that 80 psi (551.6 kPa) is the maximum pres- 6. All sectional valves should be opened during
sure allowed by most designs and codes. disinfection. All outlets should be fully opened
Under conditions where systems are subject at least twice during injection and the re-
to freezing, and with the approval of the author- sidual checked with orthotolidin solution.
ity having jurisdiction, an air test may be sub- 7. If chlorine is used, when the chlorine residual
stituted for the water test. This can be concentration, calculated on the volume of
accomplished by connecting an air compressor water the piping will contain, indicates not
to the system, bringing the system up to 40 psi less than 50 parts per million (ppm) or milli-
(275.8 kPa), checking for leaks with liquid soap, grams per liter (mg/L) at all outlets, then all
repairing any leaks, and then subjecting the sys- valves should be closed and secured.
tem to a minimum of 1½ times the working pres-
sure (100 psi [689.5 kPa] minimum) for a 8. The residual chlorine should be retained in
minimum of 2 hours. the piping systems for a period of not less
than 24 hours.
Any equipment that may be damaged by
these tests should be disconnected from the 9. After the retention, the residual should be
system. not less than 5 ppm. If less, then the pro-
cess should be repeated as described above.
Cleaning and Disinfecting 10. If satisfactory, then all fixtures should be
flushed with clean, potable water until re-
New or repaired potable water systems shall be sidual chlorine by orthotolidin test is not
cleaned and disinfected prior to use whenever greater than that of the incoming water sup-
required by the administrative authority. The ply (this may be zero).
method to be followed should be per AWWA or
Chapter 5 — Cold-Water Systems 155
11. All work and certification of performance NOTE: It should be understood that local code
should be performed by approved applica- requirements, if more stringent than above sug-
tors or qualified personnel with chemical and gested procedures, shall be included in the speci-
laboratory experience. Certification of perfor- fications.
mance should indicate:
• Name and location of the job and date REFERENCES
when disinfection was performed.
• Material used for disinfection. 1. American Water Works Association (AWWA).
AWWA cross connection control manual. New York.
• Retention period of disinfectant in pip-
ing system. 2. AWWA. AWWA standard for disinfecting water
mains, AWWA C601.
• Ppm (mg/L) chlorine during retention.
3. AWWA. AWWA standard for disinfection of water
• Ppm (mg/L) chlorine after flushing. storage facilities, AWWAD105.
• Statement that disinfection was per- 4. AWWA. Standard for hypochlorites, AWWA B300,
formed as specified. AWWA M22.
• Signature and address of company/per- 5. AWWA. Standard for liquid chlorine, AWWAB301.
son performing disinfection.
6. Manas, V.T. National plumbing code illustrated
12. Upon completion of final flushing (after re- handbook. New York: McGraw-Hill.
tention period) the contractor should obtain 7. n.a. 1978. Piping systems fundamentals and ap-
a minimum of one water sample from each plication. Plant Engineer Magazine.
hot and cold-water line and submit samples
to a state/province and/or local, approved 8. US Department of Commerce, National Bureau
of Standards. BMS 65, Methods of estimating
laboratory. Samples should be taken from loads in plumbing systems, by R.B. Hunter.
faucets located at the highest floor and fur- Washington, DC.
thest from the meter or main water supply.
The laboratory report should show the fol- 9. US Department of Commerce, National Bureau
of Standards. BMS 66, Plumbing manual. Wash-
lowing: ington, DC.
• Name and address of approved labora- 10. US Department of Commerce, National Bureau
tory testing the sample. of Standards. BMS 79, Water distributing sys-
• Name and location of job and date the tems for buildings, by R.B. Hunter. Washington,
DC.
samples were obtained.
• The coliform organism count. An accept- 11. White, George Clifford. 1972. Handbook of chlo-
rination. New York: Van Nostrand Reinhold.
able test shall show the absence of
coliform organisms. (Some codes require
an acceptable test for 2 consecutive days.)
• Any other tests required by local code
authorities.
13. If analysis does not satisfy the above mini-
mum requirements, the disinfection proce-
dure must be repeated.
14. Before acceptance of the systems, the con-
tractor should submit to the architect (engi-
neer) for his review 3 copies of the laboratory
report and 3 copies of the certification of per-
formance as specified above.
15. Under no circumstances should the contrac-
tor permit the use of any portion of domestic
water systems until they are properly disin-
fected, flushed, and certified.