Volume of Nitrogen for Leak Testing

Enough to fill the pipe at 20 Barg.

Lyn

I voted your answer as good answer. I think with this kinda answer u have become a guru in this forum. :) Long way to go mate.

When I first read BARG I thought the OP and you made a misspelling. A quick search demonstrated I was wrong. My question now is why should one use a relative instead of an absolute measuring value???

Changes in atmospheric pressure (measured in millibar = 1/1000 of a BAR) are almost insignificant at 20 BAR ? Wouldn´t it make more sense to test the pipe at 20,5 or 21 BAR and forget atmospheric pressure?

What's wrong with barg? It's used all the time, as is psig. Assuming the outside of the pipe is at atmospheric pressure (and if it isn't presumably it was taken account of in coming up with the test pressure figure) the differential pressure is 20 bar, which can be used to calculate wall tensile stress.

20 barg is 290 PSI; 2 MPa, as measured by a gauge.

And when there is a leak even more dependend on how big the leak!

That is a simple matter of PVT calculation. Dont forget the T!

Also do not forget to make a HAZOP.

Stay Safe!

http://www.pipeflowcalculations.com/

If we told you how much in standard (or normal) m³, would that help? Can you relate that to the number of cylinders you need to supply?

Codemaster, Yes. I wil relate that amount to number of cylinders required which are at 3000 Psig.

OK, in that case volume of nitrogen in standard m³ is about 20 x volume of the pipe, as others have said. I make pipe volume about 5.3m³, depending on pipe schedule hence actual ID (which I'm sure you could have worked out ).

Ahem, this is a leak test. Depending on the size of the leak, if it can be found, how long it will take to repair and retest this could take many multiples of the approximately 100 liters of STP nitrogen I calculated to get this pipe to pressure once.

Oh, why does a metric length of pipe have an SAE inner diameter?

Why does a metric length of pipe have an SAE inner diameter? Probably because it's a 10" NB pipe. Metric pipe has different dimensions

Quite.

It would be rather less if the thing were tested with water to a higher pressure beforehand.

Go get one of these....

When it runs out,,,,go get another one.....(repeat as necessary)

20 x the volume of the pipe.

Well, almost. 21 times the volume of the pipe.

The 'first' volume fills an empty, evacuated pipe to atmospheric pressure, 0 barg.

Each additional 'volume' adds 1barg.

Why bother to evacuate? Do you have any idea at all how long one would have to pump a pipe this long out from any one point on the pipe? This idea is just as preposterous as the idea of pressurizing a 100 meter pipe with an odorless, colorless gas that is difficult to detect. Particularly when one pressurizes to dangerous pressure levels to find this leak.

Quite, OP didn't mention it, and only reason I can think of is to have no oxygen during the test, which seems an unlikely requirement, and could be done by purging with extra nitrogen, as somebody suggested.

Probably nitrogen chosen as cheaper than getting a 20barg compressor to site, and (maybe) cheaper than bottled compressed air. That's assuming he goes ahead with pneumatic testing after seeing all the comments!

BTW in #10 you said approx 100 liters, should be 100 m³.

Actually that was a failed attempt to make a point of mixing measurement unit systems. Some characters fell somewhere into the bit bucket. My answer was supposed to say over 100 kilo liters. Remember that 1 meter cubed is 1000 liters. Instead I (like many others) made a scaling error when switching between different measurement systems.

No problem, we're used to SI units over here now. But I remember in the 70s I knew guys who to convert m³ to litre would multiply by 220 UK gal/m³ then by 4.54 litre/gal!!

Agreed, if starting at 0 psia.

As much as it takes to blow it sky high if it ruptures?

8,582 cu ft @ 1 bar is required

At dia. = 10", r = 5" therefore Area = r sq * pi = 1.31 ft sq

Length = 100 meters = 328 ft

Area * length = volume 1.31 ft sq * 328 ft = 429 cu ft @ 1 bar

429 cu ft @ 1 bar * 20 bar = 8,582 cu ft

8,582 cu ft at 1 bar = volume needed (rough calculations only)

It is a very poor procedure to test with a gas such as Nitrogen. A hydraulic test using water is much better, safer, and will show leaks much sooner than a gas. Repeat- MUCH SAFER! Go through the trouble of getting the water out after testing instead of picking bodies up during the test.

Good Luck, Old Salt

Actually, quite a few leak tests are done pneumatically, such as submerged bubble tests for air unit "coils"; however, they are typically at considerably less than credible rupture pressures.

I get a kick out of people who go ape sh¡t over any sort of pneumatic test whatsoever.

Not sure where you get 1.31 ft² from. Must be < 1 ft² as a 10" square has area < 1 ft². I make it ~ 0.55 ft² and volume 180ft³ or 0.53 m³ (depending on actual ID). Not a case of using dia instead of radius as difference is 2.4 x, not 4.

I believe you have a math error. 5 inches (5") squared is 25 squared inches. Multiply this by 3.1 (about pi) and this comes to 80 square inches. A square foot is 144 square inches.

Service fluid is "Jet Fuel" and there is a possibility that fuel may vapourize. So this test is not intended for checking the integrity of welds rather for testing the leaks which may occur in flange joints. (As I hear from people around.)

Contact your local inert gas supplier. They usually offer that service, calculation of quantity. Nitrogen is cheap but depending on distance from site - the bank set up and trucking may be pricy. If this piping is above grade, consider adding high point vents and low point drains and switching to hydrostatic testing. Afterwards, seal weld the drains and vents, purge piping and apply a nitrogen blanket until commissioning and putting into service.

It's not obvious to me what nitrogen test you can perform that is not just as good as testing with compressed air, which of course is 78% nitrogen anyway. Is a little oxygen going to dramatically affect your test results? If you are planning a bubble test, you can use dry air instead of nitrogen.

Another idea would be to fill your system about 95% full of water and pressure test the remaining head-space with compressed air or nitrogen. It would be a bit safer and use a lot less gas. Remember, all that gas will need to be vented off after the test is complete. Make sure you don't vent that quantity of nitrogen into an enclosed space. It will displace the oxygen and somebody will be asphyxiated. Also, if, for whatever reason, you must fill that entire volume with compressed gas, pay close attention to the temperature during the test. As the gas cools, the gas will contract and the pressure will drop off. You must compensate for this or you will think you have a leak.

I assume you are asking how much liquid nitrogen you need to do the job.

Essentially the volume of your pipe. About 179 cubic feet (@ 20 barg).

Because the liquid-to-gas expansion ratio of nitrogen is 1:694 at 20 °C (68 °F)

That calculates to .258 cubic feet of liquid nitrogen expands to 179 cubic feet at atmospheric pressure. But you need to go to 290 psi...

where

p_a = atmospheric pressure (14.7 psia, 101.325 kPa)

V_a = volume of the gas at atmospheric pressure (cubic feet, cubic meter)

p_c = compressed pressure (psi, kPa)

V_c = volume of the gas at compressed pressure (cubic feet, cubic meter)

Volume of free gas in a Storage Volume

The amount of free gas at atmospheric pressure in a given volume as a cylinder storage can be calculated my modifying (1) as:

V_a = p_c V_c / p_a

290 (psi) * 179 (cubic feet) / 14.7 (psia) = 3531.3 cubic feet of gas at atmospheric pressure, Or 19.73 times as much at atmosperic pressure.

SO... .258 cubic feet of liquid N * 19.73 (from above) is about 5.1.cubic feet of liquid nitrogen

Of course temperature can change these things I think most of my calculations are for 15 degrees C. AND I'm not sure if your test requires you to purge the pipe of air.

Would that be after it has been pressed to 30 Barg with water? After all, water leaks are visible and nitrogen leaks are not, and one wouldn't want to press a pipe to a pressure that it cannot withstand unless spectacular catastrophic self-disasembly is a requirement.

May be going to use pressure decay testing. A sniffer is a possibility, Pressure decay may be better.

...assuming that the <...pipe...> can withstand the pressure and doesn't spectacularly self-disassemble.

Please do a hydraulic test to a higher pressure before introducing a compressible gas. It may well be an insurance assessment requirement in any case.

Not only are water leaks visible but since the water is not compressible the leak is not normally dangerous.

The standard procedure for the testing of new sections of natural gas pipelines is to block and/or blind the ends of the test section. Next the line is filled with water and all higher sections have any air removed. For a 1,000psi operations the line will be tested to something around 1,400 psi for a designated period of time. If no leaks are found it is pumped out, pigged out with several different types of pigs, any water not removable with the pigs is cleaned out by personnel inside the pipe (proper confined space procedures of course). It is then resealed and put into operation.

If the pressure test should fail, depending upon the location of the leak, enough or all the water is removed, the fix made and the whole testing repeated.

What are the advantages of water for hydraulic testing instead of gases for pressure testing? The water is not compressible! When a leak occurs with water the 1,000cu ft of water wants to become 1,000.00001 cu ft of water. When the gas leaks at 20 barg leaks it wants to and will become 20 times its volume, 20,000cu ft real quick with a lot of force behind it. For a reasonably sized hole this makes a blow-out on a tire seem slow and quiet. Also a water test does not have the temperature fluctuations using gases do.

The hydraulic pressure testing is much safer, both to the piping, operating personnel and the observers. Why take the risk? Does anyone want to be the one who was standing too close and got a 10 lbs chunk of metal imbedded in their head or torso?

All should try to remember when they were near a general usage compressed pipe or hose (how about a 3/4" supplying a jack-hammer) and it broke. That ain't nothin' compared to this system leaking rapidly.

I have observed several hydraulic tests of 36" and 42" gas pipe lines and always felt confident that nothing would happen to me if there was a leak.

Also, sorry about the calculation mistakes. Case of too late in the evening, a crappy calculator that was exposed to fine carbon dust and most of all mind not engaged enough (slipping clutch I guess)!

Good Luck, Old Salt

Why <...nitrogen...>?

Water is practically incompressible, can be seen, and is relatively safe; <...nitrogen...> cannot be seen, or smelled, and could throw the pipe many metres in any directiuon, which makes it "notrogen".

Any particular reason you're posting on a thread from 2014?