The standard bubble and static pressure decay put to the test
06/01/2003
By Mike Fitzpatrick and Ken Goldstein, Ph.D.
 Mike Fitzpatrick |
Last month, we discussed high-purity gas systems and how that even extremely small leaks can present problems. We looked at the phenomena of partial pressure differences to explain how gas phase impurities can "backflow" from the low pressure, external side into the high pressure, high purity side.
 Ken Goldstein, Ph.D. |
Because of this effect, very small leaks in systems that contain high-purity gases can cause significant contamination. Consequently, systems designers and installers go to great lengths to eliminate the possibility of leaks. Extensive leak-checking procedures are conducted prior to putting these systems into service.
Now, let's look at two common leak-check methods and examine why they are probably not appropriate for use with high-purity systems—although they work well in most piping.
The standard bubble test
Most of us are familiar with plumbers leak-checking water or natural gas lines in our homes and offices. After completing the initial installation, they seal off the system while opening all the internal valves; and then they pressurize the systems with compressed air or nitrogen, typically to some pressure a few pounds per square inch (psi) above atmospheric ambient.
Immediately after, they "paint" all joints and connections (potential leak sites) with a "soap and water" solution or some commercial variant. Finally, they walk around and observe the painted joints while looking for visible bubbles, which indicate the presence of leaks.
After wiping the joint clean, the leak is repaired and then re-tested. All in all, this is a simple and straightforward process. The materials required to perform the test are inexpensive, non-toxic and readily available. This method has the significant advantage of not only telling us that there is a leak but also telling us precisely where it is located.
Unfortunately, the bubble test is neither very accurate in terms of measuring the actual leak rate, nor very sensitive in terms of being able to detect the small leaks that are of concern in high-purity systems. As an example, consider a "leak" in a system that leaks at the rate of 10-9 atm.-cc/sec, or one billionth of a cubic centimeter per second at one atmosphere of differential pressure between the system interior and the external ambient.
Admittedly, this is a small leak, but eliminating leaks of this magnitude is a common requirement for high-purity installations. Now let's see how long it would take for this leak to form a visible bubble using our soap solution.
Of course, the answer is: "it depends." For starters, it depends on what we mean by a visible bubble. Let's be conservative and say that a visible bubble is at least 1/16" in diameter or 1/32" radius. This is a fairly conservative number, as you would have to be fairly close to the line to see this bubble. It's highly unlikely you'd detect it from six feet away.
Assuming a sphere, the volume of our bubble is ~ 0.000128 cubic inches, or around 0.0021 cubic centimeters. If we assume that there is one atmosphere of differential pressure (another conservative estimate), then the leak rate from our opening would be 10-9 cc/sec, or 0.000000001 cubic centimeters per second.
Dividing our bubble volume by the leak rate gives us a time of approximately 2.1 million seconds, or a bit more than 24 days to form a visible bubble. Long before this, our soap solution would have dried out, causing our bubble to collapse—even if we did not mind waiting three and a half weeks for our bubble to form. So, while the bubble test is simple, inexpensive and intuitively appealing, it cannot be used for high-purity systems.
The static pressure decay test
An alternative to the bubble test is the static pressure decay test. In this procedure, we close off the system being tested, open all the internal isolation valves and pressurize the system to anywhere between six and ten times atmospheric pressure.
Using a pressure gauge to monitor the system, we look for a pressure drop of some percentage over a period of hours or days to indicate the presence of a leak. Although conceptually simple, this test is full of complicating factors. First, selecting a percentage pressure drop that would indicate an unacceptable leak tends to be rather arbitrary. We also find that pressure gauges are seldom accurate. Measuring a pressure drop of 0.1 percent is well beyond the capabilities of most commonly used instruments.
Additionally, our measurement parameter—pressure—is linearly related to the temperature of the system being tested. So, if any portion of the system changes temperature during the testing interval, it will affect our pressure and require some form of temperature compensation. Varying sunlight, clouds and time of day will affect the system's temperature.
While this test may inform us that we have a leak, it is not very useful in helping us to locate it. And lastly, the static pressure decay test suffers from the same fundamental shortcoming as the bubble test. It is just not sensitive enough to detect the presence of the small leaks (on the order of 10-9 atm.cc/sec) that our testing must be capable of finding. Although the static pressure decay test is often used to qualify high-purity systems, it is used only to verify the overall physical integrity of the system: "We took the system up to 150 psig and it didn't come apart."
So, what tests do we use to identify leaks in high-purity systems? Join us next month when we will go overboard on inboard and outboard helium leak testing.
Michael A. Fitzpatrick has participated in the design and construction of semiconductor facilities for over 24 years and is a Senior Member of the Institute of Environmental Sciences and Technology (IEST). Mike can be reached at [email protected]. Ken Goldstein is principal of Cleanroom Consultants Inc. in Phoenix, Ariz., and is a member of the CleanRooms Editorial Advisory Board. He can be reached at [email protected].