Equivalent Leak Detection - Hydrostatic vs. Aerostatic (nitrogen)

[voytski]

Okay, this sounds like an off-the-wall question. I've seen similar topics discussed before but they haven't led me to any answers. My company is doing leak testing on jetboxes (not sure of the details on them so please don't ask). Normally a hydrostatic test is done at a certain pressure to find any leaks/defects in the component. Some of our customers want a nitrogen leak test done.
What I need to figure out is at what pressure the nitrogen should be applied to cause an equivalent amount of leakage as the water test. The theory behind this is that a gas (nitrogen) will leak more readily than water and can therefore be applied at a lower pressure to reveal the same defects. If anyone has any ideas, I welcome the help. Thanks!

 

[unclesyd]

Was your hydrostatic test used as a proof test or as just a leak test?

 

[voytski]

It is a leak test. In the given application, it is extremely important that water does not leak from the apparatus. Noramlly they are tested to 250 psig with water. However, some of our customers require a nitrogen leak test and we need to figure out how to meet those standards.

 

[DLANDISSR]

Is the question really how much does it leak or is it does it leak?
If the question is does it leak then no calculations are needed.
Presure up the device with N2 and let it sit for 24hrs. Keep the temperature constant or errors will be introduced.
There should be no pressure drop in 24 hrs. You need sensative pressure measuring instruments.

Be aware that a catastrophic failure using water will result in some water squiring out, buy a catastrophic failure with N2 at 250PSIG will result in explosive forces being released. You should have containment if you are going to use N2 and an explosion proof barrier for personnel to hid behind while conducting the test.

Piping codes require 1.5 times the operating pressure for hydrostatic tests but only 1.1 times the operating pressure for N2 tests.

Also a gas that can be easily detected with a gas leak detector may be useful.

 

[BigInch]

The real problem I see is defining exactly what leaked volume of N2 will equal a given leaked volume of H2O.

It would also appear that if you were looking a maximum leakage rate through a maximum hole size, a corresponding standard might be more readily defined.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[Thealanator]

You will want to keep the dangers of asphyxiation in mind if these boxes are tested inside a building. Can the boxes be tested outside?

 

[voytski]

Thanks for the responses, guys. I apologize that it is kind of an abstract question. The practical question here is "does it leak?" Obviously we do not want to match the 250 psig of water with 250 psig of nitrogen. We are assuming that the applied water pressure will show leaks down to a certain size, and that another given (lower) pressure of nitrogen will show leaks down to the same size. I guess my question then expands to: How do I find the minimum defect size that the water test will find, and how I do I find the equivalent nitrogen pressure that will push nitrogen through that same size defect? I understand that this is a complex question. If it helps, the jetbox is made of cast copper.

I will be out for the rest of the night, so I will hopefully be able to get back in the morning.

 

[BigInch]

Look at your hydrotest requirements now and start with those conditions for water.

Make a series of calcs,
assuming leak diameters, calculate flow (or resulting pressure drop) across such a leak diameter for both H2O and then the N2 during a certain assumed time interval and you'll have an idea of what "equivalent volumes" might represent.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[BillBirch]

The is a leak test being used more in the gas industry using a mivture of nitrogen and helium. The helium is in a fixed concentration within the nitrogen that is above the naturally occurring ambient concentration. A gas detector "counts" the helium atoms, thus quantifying the leakage rate. Is this the technique that you have in mind?

 

[unclesyd]

For certain leaks N2/Air will flow 50 times the volume of water. So if you are just looking to see if you have any leaks and not quantify same N2 at 50 psig would go a long way at finding them.

Here is a leak rate comparison for gases including water vapor, not liquid.

Using N2/Air how are you going to check for leaks, like underwater, bubble fluid, pressure decay, mass flow, or other?

What size leak are you expecting?

 

[JR97]

Voytski,

Here is my 2 cents worth. I don't believe the molecular size difference between H2O and N2 will be significant when compared to the test parameters. As Dlandissr eluded, the time at pressure and the resolution of the pressure sensing equipment is what will control the sensitivity of the test.

How long did you hold the hydrostatic pressure at 250 psig? and what decrease in pressure was readable on your pressure gage?

Based on the answers to those questions you can calculate the effective leak hole diameter and length. Then use the leak hole diameter and length from the hydrostatic test to define what you need for the N2 test. You then work the equations using the desired pressure and how small a pressure decrease you can read to determine the required hold time at pressure. Higher pressure = shorter hold time. Smaller graduation on gage = shorter hold time.

You can find the formulas in ANSI N14.5.

The problem with a pressure drop test is as Dlandissr stated, the temperature of the N2, and of the container must be kept constant. If your test piece is outside or not in airconditioned room, changes in part and gas temperature will throw off the pressures.

JR97

 

[voytski]

Thanks again for the responses. Let's assume, for the nature of simplicity, that the test conditions are controlled, all safety measures considered, and the detecting method is not important. Consider the test items to be two pressurized vessels, one filled with water and the other nitrogen. The leak is a tiny void in the chamber wall which provides a certain resistance to fluid flow. A certain water pressure is required to overcome this flow resistance. What I need to calculate is the minimum void diameter based on pressure for water, and then a minimum pressure for nitrogen based on that diameter. Assume a short time for the test, perhaps 20 - 30 minutes.

 

[JR97]

Voytski,

You are on the right path. Again I recommend ANSI N14.5. It has all the necessary equations (which get quite involved). First calculate the leak rate you can detect during the hydrostatic test based on test volume, hold time, and gage resolution. Next equate that leak rate to the gas pressure drop equation which is based on test volume, hold time, test pressure, gage resolution, and change in temperature.

That will result in series of solutions for different test pressures and hold times.

JR97

 

[voytski]

Thank you, I will check out ANSI N14.5 and see what I can figure out.

 

[voytski]

FYI - I could not get ahold of the ANSI formulas, but I did find some good resources in case anyone else has a similar question:

http://www.cincinnati-test.com/downloads/pdf/app_bulletin/ab120a.pdf

http://www.atcinc.net/faq.asp#liquids

http://www.lacotech.com/products/997300658_leaktechref.pdf