Leakage test with dry air

[TheLucifer]

Hi Colleagues,
We are trying to design a Water/Waste System to an aircraft and ground test phase is coming soon. After the installation of system to aircraft, ground test team will test the system with pressurized dry air for leakages. Test duration is 15 minutes and pressure will be measured after this duration to see if system leaks or not.

Ground test team warned me that they are used to have a permitted leakage rate, which does not count as installation failure. They say that since all the system connections are not welded, there will be negligible leakages from connections, couplings, even from ball valves when dry air is used.

On the other hand, our supplier insists that pressure drop during the test should be zero. I beleive that ground test team is right, but I can't calculate a leakage rate for them. Is there a rule of thumb for this? If you know, can you suggest me any references/books that I can find more information?

 

[BigInch]

Assuming you want to be able to accept some leakage....

If they test with air, but the actual system will be handling liquid, or some 2 phase and only a liquid leak is important, you could develop a method based on the difference between theoretical gas leak rates vs liquid leak rates. Given their different characteristics, a measurable leak rate for a gas during a leak test would correspond to a much lower leak rate of a liquid emanating during actual use. As long as the liquid leak rate predicted from a gas leak test is less than some "allowable" liquid leak rate (whatever that might be I don't know), you could say it passes the test.

http://virtualpipeline.spaces.msn.com

"We can't solve problems by using the same kind of thinking we used when we created them." -Albert Einstein

 

[TheLucifer]

Dear BigInch,
You're right: our medium is water, hence a liquid. My theoretical leak rate for water is zero. the problem is, I don't know how to calculate a theoretical gas leak rate.

Test group says that in every connection, valve etc. that are designed for water, there are extremely small gaps, through which water cannot flow but dry air can. They estimated a standard leak rate of 12 cm3/min for three differend subsystems, which seems illogical to me since our subsystems have different scales, different numbers of equipment, so should have different leak rates. I'm trying to find leak rate of each connection/valve/motor etc. for dry air for a range of pressures and I thought that maybe there existis a simpler rule/equation/database that I don't know but you experts use.

haven't you ever encountered some calculation like this?

 

[BigInch]

Whenever I've seen anybody try to compare leak rates, the calculations are equated or correlated to a test medium leakage through some assumed orifice of set diameter.

12 cm3/min looks like a lot of leakage. I think that such a leak rate is not valid for joint and fitting leakage testing.

Actually for a 15 minute test, I would't expect ANY leakage caused by anything other than for a connection not properly completed at all. Is that what the test is for?

http://virtualpipeline.spaces.msn.com

"We can't solve problems by using the same kind of thinking we used when we created them." -Albert Einstein

 

[TheLucifer]

Definitely. Equipment will already be tested after manufacturing for performance but complete system can only be installed on aircraft, so we want to be sure that nothing leaks due to wrong installation.

 

[BigInch]

In that case, I'd go with the supplier's recommendation for zero leakage. I have the impression that if you can see a leak in 15 minutes, air or liquid, its a bum installation no matter what the allowable is.

If you set up an air test on one end of the system at some test pressure, how much flow would you get out of the farthest end (at atomospheric outlet pressure), using something like an average diameter of tubing in the system? For lack of anything better, try something like 1/100th of that flowrate as an allowable to prove all connections were made or not made. What's that flowrate? How does it compare with the 12 cm3/m?

http://virtualpipeline.spaces.msn.com

"We can't solve problems by using the same kind of thinking we used when we created them." -Albert Einstein

 

[mfitzs]

Just my 2 cents here. Remember "EVERYTHING" leaks. It is just what is the leak rate. Just for an example here...If I take a lets say pneumatic coupling cap one end off and energize the other end with compressed air to 100 psi I would watch the coupling and time 15 seconds and did not achieve a 3mm bubble I would say it was a passing sample. This would be a 10 to the -3. You may want to check out the Cincinati test site for some other info.

 

[eyec]

out of my area but, if the complete system is tested and passes prior to install, why not just bubble check the tie-in connections?

however, you need to specify any leak rate contracturally first.

also, in my experience with air tests, temperature fluctuations during testing can cause psi drops without any leaks.

 

[unclesyd]

Here is a similar system to one that we use in conjunction with He test to test the integrity of pipe and tube assemblies. You can use this type system to easily establish testing parameters required by your system.
It takes care of all the problems with straight pressure decay.

http://www.atcinc.net/how-it-works.asp#pressure-operation

 

[c1defence]

Check out the tests for valves.
API Specification 6D, January 2002
ISO 14313:1999,
See Annex C
And ISO 5208, Industrial valves—Pressure testing of valves.
Also see API 598
Table 5 Maximum Allowable Leakage Rates for Closure Tests
a) For the liquid test, 1 millimeter is considered equivalent to 16 drops.
b) There shall be no leakage for the minimum specified test duration (see Table 4). For liquid test, O drop means no visible leakage per minimum specified test
duration. For gas test, O bubble means less than 1 bubble per minimum specified test duration.
c)The maximum permissible leakage rate shall be O. 18 cubic inch (3 cubic centimeters) per minute per inch of nominal pipe size.
d) The maximum permissible leakage rate shall be 1.5 standard cubic feet (0.042 cubic meter) of gas per hour per inch of nominal pipe siz