where the pipeline initially started out at the be-
CONSTRUCTION
ginning of the street. During the blasting and ex-
cavation we tried to maintain at least this depth
General physical layout
to give us an extra buffer against the cold surface
Labossiere Street is a dead-end hillside street
temperatures.
on the northwest side of Berlin. Our test section
The numerical model predicted that the pipe
started at the intersection of Sixth Avenue and
would get very close to 0C but would not go
continued up the roughly 17% slope of Labossiere
below it by using our final design from Second
to its end at the base of a mountain of bare rock
Avenue: a 1.5-cm (6-in.)-thick extruded polysty-
that projects upwards from the end of the street.
rene insulation with a total shield width of 1.2 m
The water line serving the users on the street is
(4 ft) with the pipe resting at the 1.5-m (5-ft)-deep
a dead-end line that runs about 128 m (420 ft) up
level.
the west side of the street.
Several points should be made about our de-
signs based upon the numerical simulations:
Pipe construction
1. The temperatures used at the model surface
The shield design called for a 15.2-cm (6-in.)-
are air temperatures from a very cold year. The
thick layer of insulation in an inverted U around
temperatures used at the model surface should
the pipe. The sides of the U were 61 cm (2 ft) high,
be ground-surface temperatures, which tend to
with the bottom of the legs even with or slightly
be warmer overall than the air temperatures be-
below the bottom of the pipe. The total width of
cause of the effect of solar energy. This study will
the shield was 1.2 m (4 ft). The shield was con-
structed of 5.1-cm (2-in.)-thick, 4- 8-ft SSE boards
help to quantify this effect.
2. The ledge is modeled as if it were solid un-
of Foamular 250 extruded polystyrene. This
broken rock. Natural ledge is frequently broken
material meets the specification requirements
and can have areas of water flow and other anoma-
of ASTM C578, Type IV. The boards have lines
lies. We expect the effective thermal conductivity
scored at 40.6, 61, and 81.3 cm (16-in., 24-in. and
of the natural ledge to be less than our modeled
32-in.) spacing across their width, which make
ledge.
them easy to break at the job site and leaves a
3. Our failure criteria could be much too con-
clean even line where they break. Consequently,
servative. We "turn off" the water flow and as-
the 61-cm (2-ft)-high sides of the shield could
sume that the pipe fails if afterwards the 0C iso-
easily be obtained from the 4- 8-ft. (1.2- 2.4-m)
therm touches it. In reality there is probably some
boards. Table 2 lists insulation data.
water flowing frequently within the pipe, which
The new 8-in. ductile iron pipe construction
would bring a little heat into the shield area.
started with a T connection into the existing wa-
4. A great uncertainty exists in the water tem-
ter line at Sixth Avenue and proceeded up
perature boundary that we applied to the pipe.
Labossiere Street. We did not start insulating the
As mentioned above, we used the water tempera-
line from this T because of the presence of a storm
tures at the treatment and filter plants for our
drain line that ran down Sixth Avenue and crossed
pipe boundary temperatures. This may or may
over the new water line just up from this connec-
not be correct, depending upon how the water
tion, as shown in Figure 6. We were concerned
temperature changes as it travels through the dis-
that cold air would be flowing through the
tribution system.
underdrain and would cool down the water pipe
5. Another concern in this study is the effect
where they cross, so we installed a thermocouple
that Berlin's very cold water temperatures will
to monitor the temperature at the bottom of the
have within an insulated system. Will the cold
drain pipe. The stations shown in Figure 7 start
temperatures quickly remove the stored heat
where the storm drain crosses the water line. The
within the insulation shield, much faster than
would have occurred had the heat loss only
been caused by heat conduction to the sur-
Table 2. Insulation data.
rounding ground? One of the ways an insu-
lated system is effective is by preventing the
Thermal conductivity
Density
Specific heat
(kg/m3)
rapid loss of this stored energy to the sur-
(J/hr m C)
(J/kg C)
rounding ground. The cold pipe tempera-
Foamular 250
93.46
28.84
1339.8
tures could cancel out this benefit.
7
Previous Page
