Modifications after Hydrostatic/Pneumatic Test 2

[AndyChE]

We have a piping system being built in accordance with B31.3. Due to the layout, we must remove a flanged spool to get our test apparatus in. After we test the piping and replace the spool do we need to retest the two flanged joints? We have similar cases with union joints.

My thinking is that we could consider it a minor addition to replace the spool and waive retesting based on 345.2.6 (with the owner's approval, of course).

Is this common industry practice? We don't have any welds to make the golden weld and we did not design in any test ports for the leak test. The fluid service category is M, but since M345 states that 345 applies in it's entirety I would think we could still do this.

Any thoughts are appreciated.

 

[metengr]

Yes, I would agree.

 

[jte]

If I understand your question correctly, it sounds as though the pipe and the removable spool have already been hydrotested for Code compliance. Now the spool is being removed and reinstalled, and the issue is whether you need to retest to verify the integrity of the two flanged joints. In this case you do not need a "code hydro"; a "tightness test" or "in service leak test" will do. ASME PCC-2, Repair of Pressure Equipment and Piping has an article (5.1) which deals with pressure testing in the post-construction environment. It is available for $95 at

http://catalog.asme.org/Codes/PrintBook/PCC2_2006_Repair_Pressure.cfm,

the 2008 edition is anticipated to be available in late summer, with a substantial increase in the volume of contents. The flow chart in Fig. 1 in Article 5.1 is pretty useful in helping to determine what type of test to perform.

jt

 

[AndyChE]

Thank you both for your replies.

jte - The spool has not been hydrotested for code compliance. We need to break into the system to get our apparatus in to do the code test (it is actually pneumatic). So we will have two flanged joints that are not tested. In fact, there are additional joints that are disassembled and realigned for the test set-up so we will have multiple untested joints.

I realize this is not the best practice, but unless it is a code violation I can't get anything done about it.

The plant is not yet in service so this would not be maintenance (which is exempt from B31.3).

 

[jte]

Ok, so you will need to Code hydro the spools. Not a big deal. They'll just be hydro'd separately from the rest of the piping. Take the spool, get a couple of blind flanges and hydrotest the spool. I'll leave the discussion on the safety aspects of pneumatic testing to another post - you'll find plenty on eng-tips with a quick search. If the spool has conveniently located connections then use them for your hydrotest connections. Otherwise, you can put a 1/2" or 3/4" connection on the blind flange and attach your "testing spool" to it which will consist of appropriate valves, a pressure gage, and connection to your hydro pump. You'll need a connection top and bottom to drain the water or kerosene or whatever. Avoid pneumatic testing unless you have a good engineering reason not to. Start another thread if you need to discuss why.

Remember - for a code pressure test, you're hydrotesting the welds and gross structural integrity of the pipe and fittings, not the bolted flanged connections.

jt

 

[AndyChE]

That is indeed what we will do (except pneumatically). There are no vessels in line and by my estimation there is only about 15 ft³ of piping at 110 psig, so the amount of stored energy is relatively low (but I have read several posts in regards to pneumatic testing and the dangers).

I must humbly disagree with your interpretation of the code test. You are not testing gross structural integrity, but rather checking for leaks. 345.1 states that “Prior to initial operation … each piping system shall be tested to ensure tightness…” Also, you are required to check everything, not just welds and fittings. 345.2.2 (a) states that “… all joints and connections shall be examined for leaks.” At least this is my interpretation.

Thanks again for your thoughts.

 

[jte]

Test away! In my experience, it is not at all unusual to 1.5* test spools, bolt 'em up and tightness/leak test them at a lower pressure. Would you really expect a flanged connection to a large vessel to be hydrotested after it is made up? Many vessels are not built to withstand a 1.5* test in the vertical position, and don't forget about the foundation or structure...

only about 15 ft³ of piping at 110 psig, so the amount of stored energy is relatively low

Yep, only equivalent to around 30 lbs of TNT (unless I ran the numbers wrong). Min distance for personnel should be 155' and at that perimeter you can expect glass windows to fail. At a bit below 100' you can expect eardrums to rupture; ~50' you can expect lung damage.

jt

 

[AndyChE]

All vessels are blinded off and the final connection doesn't have to be tested per code. I would prefer a low pressure test after everything is bolted up, but they will only do code requirements (contract issues). I will do a full test later when we do other testing.

I don't have ASME PCC-2 and a copy isn't available at our library, but here is the equation I always use based on adiabatic expansion:

E = (k)/(k-1) * (PV) * [1 - (P/atm)^((k-1)/(k))]

Where:
P = test pressure (Pa)
V = volume (m³)
k = ratio of specific heats (1.4)
atm = atmospheric pressure (101325 Pa)
1 kg TNT = 4184 kJ

Running those numbers I get less than half a pound of TNT.