Leak Testing Gas Components and Systems
LEAK TESTING BASICS
“Everything leaks,” McLaren says in a webinar on leak testing of cryogenic and industrial gas components and systems. “What matters is how much and where.” He offered some basic principles for leak testing.
“The overriding principle is: Test it like you use it,” says McLaren. If the component or system will be operated pressurized, leak test it pressurized. If it is intended to operate under vacuum, test it that way.
If you pressurize a vacuum system or pull a vacuum in a system intended to operate under pressure, this may induce leaks that would not occur under normal operating conditions and you can get incorrect results.
Another factor in “test it like you use it” is to test at the system or component’s operating pressure, if possible. “Sometimes that can be problematic,” McLaren says. Even if testing needs to be done at a lower pressure, “it might simply be a matter of tightening up your specification and looking for a smaller leak.”
In some applications, it is clearly necessary to measure the size of leaks, how much gas is flowing. In other applications, their manufacturing process simply calls for a determination of excessive leakage, a pass/fail approach. Pass/fail testing likely takes less time than measurement does.
Another consideration in developing a leak test strategy is the time needed for the whole testing procedure: test cycle, response and clean-up. Instrumentation, method and test configuration depend on what data are required and how much time is available for testing.
VISUALIZING LEAKAGE RATES
To get a sense of what a leak looks like—how much gas is actually leaking at different leakage rates—McLaren suggests imagining a cube 1 centimeter (about 0.4 in.) on a side, about the size of a sugar cube. That’s 1 cubic centimeter (cc), about 0.06 cubic in. That amount leaking out of a pipe in one second is one cc/sec.
McLaren considers helium, specifically, but the principle is the same for any gas. Table 1 shows in the left column the size of a leak in cc/sec, from large to very small. The right column shows how long it would take at that rate for 1 cc to leak. (A standard cc of gas is one cc of volume at one atmosphere pressure and 0°C (32°F).)
METHODS FOR MEASURING LEAKAGE IN GAS SYSTEMS AND COMPONENTS
Visual indicators: This method is similar to the common experience of placing a bicycle inner tube in water to see where the air bubbles out, indicating a leak. Visual indication depends on a material that changes somehow when subject to a gas leak. Reactive liquids change color as they are leaked through. Special soap solutions produce foamy bubbles when gas from a leak flows through them.
This is an inexpensive process. However, it provides only very rough leak detection, as it has limited sensitivity, detecting leaks down to perhaps 0.001 (1x10-3) cc/sec, McLaren says.
The disadvantages of the visual indicator method relate to the material used and the dependence on the operator. The indicators are wet; at best, they require cleaning after use and at worst can contaminate the system being tested. Also, leak detection depends on the operator to apply the solution to the leaking area.
Pressure/vacuum decay: In this method, first pressurize or produce a vacuum in the system and then record what happens to the pressure over time. It is a good way to test a system for its overall leakage, but does nothing to pinpoint the location of a leak. Its sensitivity is only slightly better than the visual indicator method, perhaps detecting 1x10-4 (0.0001) cc/sec leakage, McLaren says. The relative cost is more than the visual indicator method, but still relatively low. Different technologies can be used to track the pressure change during the test.
Acoustic leak detection: This method uses special microphones to detect the ultrasonic sound produced by gas escaping through small leaks. Good for long-distance leak detection, the acoustic method uses directional microphones one can point at overhead compressed air piping, for example. The sensitivity is “fairly crude,” McLaren says, “but good for certain maintenance applications.” The sensitivity ranges lower than even the visual approach, around 1x10-2 (0.02) cc/sec leakage, he says. The cost is more than either of the preceding methods.
Tracer gas leak detection: This approach uses a gas such as helium to show where gas leaks into or out of a test part or system. A specialized instrument (see next section) is used to detect the presence of helium.
For a pressurized system, helium is injected inside the test part and a “sniffer” connected to the helium detector senses helium where it leaks out. For a vacuum system, helium is “sprayed” near the likely leak locations. If it is detected inside the part or system, that indicates a leak at the sprayed location.
This approach is more expensive than the others described, McLaren says, but goes many orders of magnitude beyond their leak-detection capability and can pinpoint the location of leaks. Its speed, accuracy and sensitivity may offset the cost. This method can detect very small leaks, down to 1x10-11 (0.00000000001) cc/sec, he says.
WHY USE HELIUM AS A TRACER GAS?
Helium has a great many advantages as a tracer gas. It is non-toxic and inert, so it harms neither people nor the equipment. Because of its size as a very small atom, it flows easily through small leaks. Its presence in air is only a trace (approximately 5 parts/million), McLaren says, so helium leaks can be clearly differentiated from the background concentration in air.
Two types of devices are used to detect the helium:
Mass spectrometers are used in testing both pressure and vacuum modes and come in a range of configurations for different leak testing applications.
Selective ion pump detectors are small units useful in maintenance applications and where ultra-portability is needed in pressurized applications.
HELIUM LEAK DETECTION, OUTSIDE-IN
Where leakage is from the outside into a vessel, chamber or pipe—the pressure inside is lower than that of the surroundings. Here, helium leak detection involves spraying helium in different areas where leaks may occur and monitoring when helium is detected in the interior of the test part.
To run a test, pressurized helium is regulated down to low pressure and fed to a spray gun that dispenses the helium. An operator sprays the helium carefully near possible leak areas, such as flanges, seals or welds on the part being tested. Meanwhile, the helium detector monitors how much, if any, helium is drawn inside the part when the spray gun is dispensing helium in each location.
This outside-in test does not measure the total leak rate. However, it does measure and identify individual leak locations. It is dependent on having a skilled the operator spraying the helium in the right places in the right amounts for effective leak measurement.
HELIUM LEAK DETECTION, INSIDE-OUT
To detect leakage from a pressurized system or component, helium is released inside the test part. An operator passes a sniffer probe over and around the part being tested and can find the exact location of the leaks by detecting the helium leaking out.
This inside-out measurement on pressurized systems or parts identifies the leak location. It does not measure the rate of leakage for the individual leaks or the system as a whole.
Both of these procedures depend on the operator’s skill in sniffing or spraying helium in the right places and in a fashion that coordinates with the readings on the helium detector.
ADAPTIBLE TO VARIED APPLICATIONS
McLaren gives examples of how helium leak detection is used with systems large and small. Huge cryogenic storage containers, for example, may require repeated pumpdown/vent cycles to remove moisture from their insulation before leak testing. A long, narrow piping system under vacuum needed to have a slight flow of nitrogen provided to help propel the helium to the detector connected at the end of the pipe, so it could be sensed more quickly. The principles remain the same; the applications may require some accommodation to get the best results most quickly.
In applications that need and can take advantage of its capabilities, helium leak testing provides performance, speed and flexibility, McLaren says. For other types of applications, choose one of the other leak testing methods that can meet the system or component’s requirements.
Barbara Donohue is web editor of VALVE Magazine
One of the founding members of VMA’s Education & Training (E&T) Committee, Ed Holtgraver (QTRCO), remembers the excitement created by the first Valve Basics Seminar held in Houston a decade ago.
There’s a lot to know about valves in order to select the right ones and use them effectively.
More than 200 people from across the valve industry participated virtually in the VMA Valve Forum and Valve Basics event, which included 4 days of presentations on technical, manufacturing, management and marketing, valves 101, and valve repair.