We see the dramatic headlines: "RAIL CAR
EXPLODES", "OSHA INVESTIGATES CHLORINE GAS LEAKAGE", "EPA
REQUIRES UNEARTHING OF GAS STATION FUEL TANK", "LNG TANKER
FAILS COAST GUARD CERTIFICATION". Government agencies have
become tougher and legal liability claims grow higher. Yet,
for every headline, there are probably a hundred such leaks
that have not yet come to the public's attention.
Why are so many containers
leaking? With today's concern for resource conservation and
public safety, it cannot be because no one cares.
A major reason is the very
size and construction complexity of these vessels make them
very difficult and expensive to test. Each vessel has
special problems; it is double walled, it's leaking on the
bottom which is firmly in the ground, it leaks only
intermittently, the source of the leak is distantly removed
from it's indication.
Another possible reason for
so many large tank leaks lies in the variety of possible
leak testing methods available. These methods have evolved
partially because many of the codes have failed to keep up
with current needs and technologies. Because of the variety
of available methods, and the lack of adequate information
on these methods, the NDT engineer may be frustrated in
trying to find the appropriate method for his application.
All too often he ends up relying on a leak testing method
too cumbersome for the job. NDT engineers sometimes give up
on leak testing altogether, relying on the better known flaw
detection techniques to find both flaws and leaks.
The Alaskan Pipeline, is a
dramatic illustration of the need for leakage testing. The
pipeline was not leak tested, instead the most sophisticated
methods of traditional flaw testing were relied upon to
solve both problems - leaks and flaws. However, once the
pipeline went into operation, leaks developed immediately,
some large enough to cause the shutdown of the pipeline. The
Alaska Pipeline is not the only project that has suffered,
sometimes at great expense, from a failure to understand
that flaw testing is not a substitute for leak testing.
Disasters which can cause death and great loss dramatize the
need for a better understanding. As a result of such
problems there is a developing interest in leakage testing.
Concepts & Terms
Leakage testing defined in its simplest form is a branch of
nondestructive testing used for the detection (location or
measurement) of fluid leakage in either a pressurized or an
evacuated system. The word "leak"
refers to the physical hole and not-to the quantity of gas
or liquid flowing through the hole. In other words, leaks
are flaws which-affect the safety or performance of a system
or which result- in environmental contamination or energy
loss. The word "leakage"
refers to the flow of a-fluid through a leak without regard
to the physical size of the hole. Leakage typically occurs
as a result of a pressure differential across the hole,
however capillary effects can also be a cause of leakage.
When fluid flows through a small leak the rate of flow
depends upon the geometry of the leak, the nature of the
leaking fluid and the prevailing pressure and temperature.
The characteristics of a leak
are often referred to as the conductance of the leak.
Because the leak hole can usually not be seen or measured,
the quantity used to describe the leak size is the
conductance or leakage rate of a given fluid through the
leak under given conditions. The leakage rate used as a
measure of leak size must have dimensions equivalent to
pressure, temperature, time, and volume.
Users and manufacturers are often unable to
communicate effectively with each other or to make
comparisons between various methods. It is difficult enough
to think of a hole so small it cannot be seen by x-ray and
must be defined by a mass flow through it, without having
the added complication of a great variety of different
measures used to define it. The most commonly used measure
now is the standard cubic centimeters per second. What does
a standard cm3/sec. look like? A leak of 1x1O-4
std. cm3/sec. is equivalent to the loss of 1 cm3
of air over a 3 hour period -- this does not sound like a
relatively large leak. However, it is equivalent to losing a
pound of freon over 27 years. Besides these two equivalents,
there are many more leakage flow measures in use. Some of
the more common are torr liters per second, bubble time,
kilograms per year, standard liters per day, micron liters
per second, and micron cubic feet per hour. The latest to
appear is the standard international unit pascal-cubic-meters
per second. S.I. units have been adopted by almost all of
the technical organizations such as N.B.S., ASTM and ASNT
Equivalent Leak Size
should the test be?
Nothing can ever be completely free of
leakage. Every container always has some leakage, even if
those leaks are so minute that it would take 320 years for a
cubic centimeter of gas to leak out. The choice of which
method to use revolves around two questions:
- What should be the standard of leak tightness be?
- How can this standard be met most economically and
reliably, so as not to dramatically increase the cost of the
The increasing cost of
finding smaller leaks must be balanced with to the
functioning of the unit over its useful life. Leakage tight
therefore has no meaning except in relation to the substance
which is to be contained, its normal operating conditions,
and the objectives with respect to safety, contamination,
and reliability. Leakage tight is the practical leakage
which is acceptable under normal circumstances, e.g.,
clearly a gravel truck need not be free of water leakage.