Air Cooled Condenser vacuum testing

[AndreChE]

Dear all

We have recently switched 3 complete cells of our Air Cooled Condenser system.
The old ones were cut and new ones were welded.

The test that is being done (before plant startup) is a positive leak test.
We pressurize the system up to 500-700mbar (design pressure) and check pressure loss.

However, since it is a vacuum system we would like to test under vacuum conditions before plant startup.

Is there any procedure, API or recommended practice for that?


Best regards,
AndreChE

 

[gr2vessels]

No, there is no specific testing instructions for vacuum testing. It is normally assumed the strength test to be a pressure test, followed by a leak test to check your gaskets and other bolted joints. The test will monitor the vacuum level over a period of time. However, operating temperature and other conditions might influence the joints integrity and accordingly the level operating vacuum.
Cheers,
gr2vessels

 

[Duwe6]

Use shaving cream [soap foam] to find any leaks that the "watch the gauge" test may find.

 

[rhg]

See ASME V Article 10 Leak Testing.
Household soap or detergents are not permitted as substitutes for bubble testing solutions.

Regards

rhg

 

[Duwe6]

Ya can't Bubble Test for vacuum. You have to look for the shaving cream disappearing to find a vacuum leak.

 

[rmw]

Helium Leak Test. Google it. Tried and true.

rmw

 

[gr2vessels]

AndreChE,
You have to take the advice with a grain of salt and pepper;- The leak test is not an integrity test of the equipment, it is only a leak test. You have to judge for yourself the consequences of air leaking into the system, also how critical is the leaking itself (you need to quantify the leak, to determine if you can live with it or not). The most onerous test is probably the helium leak test, if you have located the joint which leaks. However, it is a very expensive exercise and not really relevant, since any joint sealing water will almost certainly leak for helium (small molecular size), no matter how much you tighten the joint. Also, an internal light pressure test (bubble test) is not representative, since the vacuum test may fail where the pressure leak test is OK. As for Duwe's shaving cream, it is practical if you can afford the ingress of soap in the process flow...First of all, you do your homework selecting the best gasket suitable for vacuum conditions, without breaking the flange bolts or bending the flanges.
Cheers,
gr2vessels

 

[AndreChE]

No gaskets since the piping is welded.

We did X-ray 100%, penetrating liquids test and pressure test at 500 mbar with air.
It was accepted by authorities.

Vacuum test to be done in startup of the ACC

AndreChE

 

[Duwe6]

To save money, and conserve a non-renewable natural resource -- helium -- a lot of refineries have gone to a 5% Hydrogen / 95% N2 testing gas. It works very well, is fairly cheap, renewable, and is nonflammable at 5% or lower concentrations.

Only disadvantage is that a Mass Spectrometer calibrated and set up for He will require a complete new setup and cal for H2.

The above notwithstanding; AndreChE you have done plenty already. Either the PT - dye penetrant testing - or the hydro would both have found any leaks. Don't overthink the problem - a full 'hard' vacuum is only (-)14.7 psig.

 

[stanweld]

You could perform a vacuum box, bubble leak test on the welds but that will not prove that the system has been properly stiffened to prevent collapse. From a leak detection standpoint, you've done enough.

 

[EdStainless]

Or use SF6 as a tracer gas. This is common in high voltage and in HVAC.
Similar equipment to He testing.
Call a testing company such as Conco Sysetms.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[Duwe6]

SF6= Mol Weight of 146, but is easy to detect.
H2 = Mol Weight of 2, and will find ANYTHING resembling a leak. Moderatly easy to detect.

 

[EdStainless]

Duwe6, That is true for molecular leak conditions, but you are talking 10e-6 or lower.
Most commercial applications are in the 10e-4 range and involve viscous flow.
For viscous flow in a leak there is very little differences between gases and detection becomes the critical issue.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[rhg]

Inspections and tests must be considered during design of equipments, not after fabrication.
Regards

rhg

 

[rmw]

I would hope you don't have too many flanged joints in an air cooled condenser.

Helium leak testing has been around since dirt was invented. I just recommended reading about it. I didn't say it would be cheap or easy. In cases where I have seen it used very effectively around water cooled steam condensers, the turbines were generally inside of buildings or in the vicinity of large structures at ground level.

With your A/C condenser being up in the sky, you would have to develop a plan to miminize the amount of helium (or any other of the leak test gasses mentioned) in the vicinity of known joints. Doing it when the wind doesn't just kill you would be mandatory.

If you are leak testing just to check the integrity of initial construction and installation joints, that probably would suffice because you can use a targeted approach.

If on the other hand, you were leak testing trying to find a leak of undetermined origin and all the welded and flanged joint areas had been checked, the massive size of an A/C condenser might make a general search prohibitive.

I once saw an air leak so difficult to find that the staff finally resorted to just spraying Helium under the belly of a large steam turbine (almost 1000 MW) and lo and behold after several wild sprays they got a hit and then with some more targeted spraying, narrowed it down to a 3/4" pipe plug left out of a turbine ventilator valve housing.

It might or might not be the answer for you.

rmw

 

[gr2vessels]

Very interesting topic, I might learn a thing or two. Is there any benefit of performing an onerous leak test after the mandatory pressure test on a pressure vessel?
Best regards,
gr2vessels

 

[stanweld]

gr2vessels,
Possibly so if the vessel is in toxic service and the nozzle/manway flanges have been opened during installation to check for leak tightness.