Page 217 - Water Loss Control

190 Cha pte r T w e l v e

however, in the majority of situations, on-site testing is more economical in terms of
time and resources. From a technical standpoint, the piping configuration surrounding
the meter can have an appreciable impact on the meter’s accuracy, and such impacts
can be detected and evaluated when the testing is conducted on-site.
Both on-site and test bench testing of large
meters rely on a large volume of water being passed
On-site testing of large through the meter being tested. When tested on a
meters is often the preferred test bench in the shop, water is passed into a tank of
known volume. In conducting on-site testing, the
method as the customer site
flow registered by the meter being tested is com-
is tested for suitability as well
pared to a meter that has been previously calibrated
as the meter for accuracy.
and known to be accurate. The two meters are con-
nected in series, and the test water is discharged to
waste. Since the calibrated meter is not 100% accurate on all flows, it may be necessary
to adjust for its accuracy variance at different rates of flow, in order to ensure proper test
results. One very important point to remember in field-testing is that both meters must
be full of water and under positive pressure with all air removed. The control valve for
regulating flow, therefore, should always be on the discharge side of the calibrated meter.
A valve on the inlet side of the meter being tested or one located between two meters for
controlling rates of flow should not be used, as inaccurate results may occur.
One acceptable method of maintaining proper performance for certain types of
larger meters is to replace the operating components and assemblies while leaving the
meter body in place. For such meters it is also recommended that an on-site meter accu-
racy test be conducted at the time of installation to confirm that the composite metering
unit is functioning as designed. If the measurement and registration functions are
within one integral assembly, no accuracy tests are required at the time of installation,
and the entire unit must be tested at the regular maintenance intervals.
Some larger meters have built-in test plugs while others do not. For installations
requiring test outlets, these can be fabricated in a number of ways. Service saddles
and reducing tees are the most frequently used approaches. These need to be installed
according to the recommendations of the meter manufacturer and located so that the
connecting hose to the on-site tester is correctly located downstream to the meter. To
facilitate periodic testing of the meter, it is suggested that, as part of the original
installation process, a short length of pipe be permanently attached to the test outlet,
along with a shut-off valve, which can be locked into position. These features will
allow for quick, efficient testing at regular intervals throughout the life of the large
meter.
The piping configuration around the meter must include valves to positively isolate
the meter, while still maintaining an adequate flow to the end user through temporary
or permanent bypass piping. If either of the isolation valves fails to seal tightly, an inac-
curate test result may occur. Similarly, if leakage occurs at either of the valves or at the
meter connections, the integrity of the accuracy test may be compromised. The lower
the test flow rate, the higher the significance of any such leaks.
Large meter settings are relatively expensive and require considerable preliminary
planning. These meters are heavy and removal of the meters for servicing or testing is
costly and time-consuming. Therefore on-site testing of large meters is the preferred
method in many instances. When small meters are rotated out of service, the water
supply to the customer property is halted for the typically brief period of time that is
needed to remove the old meter and install the new meter. Such outages are usually