LDS WD-1 - 13
LDS WD-1 - 13
LDS WD-1 - 13
accuracy of geophones depends highly on the proficiency and the experience of the op-
erator, and it also might fail to detect some leak classes. Furthermore, the exact location
of the pipeline to be assessed must be marked so that the operator would know where
to put the device. The examination renders the area above the pipe unusable in case it
is a street or highly utilized area. Similar to Geophones, Hydrophones try to listen to
leaks by sometimes being placed in the system and rarely on the surface of the ground.
Hydrophones can be more accurate than geophones in detecting leaks but they require
more training than geophones to operate, and they are approximately seven times more
expensive than geophones (United States Environmental Protection Agency 2009). To
detect a leak, these devices rely on the high-frequency acoustic signals sent by the re-
lease of pressurized fluids, to detect leak existence and leak locations. Sound frequen-
cies are then amplified and filtered at 1 kHz using a preamplifier to remove high-
frequency noises that are not related to the network. By measuring the time delay be-
tween two detection instants between two given listeners the leak can be pinpointed by
relating propagation speed within the medium with time and distance.
Another listening approach would be the use of listening sticks which are extended
earpieces. This approach is highly dependent on the ability of the operator to listen and
distinguish leak sounds properly. This approach is most suited for metallic pipelines be-
tween 75 mm and 250 mm in diameter and having preferably a pressure equal to or
higher than 10 m or 15 psi. The accuracy of listening rods is independent of material
type and thus it can be used in a versatile manner yet the overall hearing experience is
deterred by the existence of external noise (Hamilton and Charalambous 2013).
The aforementioned technologies are versatile and provide fast on-site leak detection
in suspected areas for any type of transmission pipelines. On the other hand, they rely
heavily on human senses and interpretation and are susceptible to inaccuracies due to
external mechanical noise. Therefore, such models can be improved further by redu-
cing the impact of noise using advanced signal analysis and filtering in addition to pro-
cessing automation which eliminates the dependency on human listening and replaces
it with computer analysis (Hamilton and Charalambous 2013).