Hydrostatic test pressure boundary problem

[tlewis]

I have a 10" 150 psig steam line that runs between two boilers on a large university campus (transfer line). One section of the pipe midway between the boilers is in a walkable tunnel. The pipe enters the west wall of the tunnel, turns north for 60' then then turns east and exits the east wall. We plan to cut an elbow (turning east) and install a "tee" to run a new 10" line down the tunnel to the north. The university wants to take only a minimal (24 - 48 hr) shutdown to cut the elbow and install the tee, which is feasible with a pre fab tee, but my concern is the pressure boundary of the hydrostatic test.

Relative to the tee, there are three boundaries for the pressure test:

1) NORTH: this will be a new north isolation valve on the north end of the tee. New 10" piping will run north of the tee, and we will pressure test it after the tee is installed.
2) EAST: when we cut the elbow, the east side of the tee will be connected to existing pipe with an existing east isolation valve within 3 feet.
3) SOUTH: will cut the other end of the elbow, and install tee with new south isolation valve. HOWEVER, the new weld on the other (south) side of the new south valve in the existing pipe will be outside the boundary established by the new south & north valves and the existing east valve. To make things worse, there is no isolation valve in the 10" line from the tunnel back to the steam boiler (well over 1,000+ feet).

So the quesion is: if you have a new weld in a 150 psig steam line 5' away from an isolation valve and over 1,000' away from the other isolation valve, how would one pressure test the new weld? Hydro test seems impossible due to length and short shutdown duration. Pneumatic also seems impossible, and possibly dangerous. Do pipefitters have an inflatable "pig" that can be placed in the pipe for the hydrotest, and then deflated and pulled out later?

Any ideas or input appreciated!

 

[davefitz]

I think the code allows you to substitute 100% radiographic exam for hydrotest.

If you are adding another 10" line, I assume the pipeline flexibility analysis has be completed that shows no issues with flexibility under all operating scenarios.

 

[stewbaby]

Very possible to use the test plug senario. Reference such vendors as

http://www.expansionseal.com/index.htm

We've used these type plugs with great success, but typically we just dye penetrate test steam and condensate piping welds and then perform a service test...but with this much pipe / volume / and exposure the test plug looks like the way to go.

Good luck in the new year! Stew

 

[bvi]

You did not state the Code of constuction. Presuming it to be ASME B31.1, you are permitted to opt, as the owner, for an initial service leak test when other tests are not practicable. See para. 137.7. I suggest, in addition, you hydrotest the new piping prior to tie in, and radiograph the tie in joints. You would probably rather not find the leaks, if there are any, at the time you start up the system.

 

[tlewis]

Thank you gentlemen for your responses on the last day of the year. Happy New Year to you all.

Parts of the campus distribution system are now 100 years old (although, fortunately, not the part I want to modify), and they have not committed to a code for construction (very little new piping in last 20 years). However, we are recommending that all new work be designed and installed to B31.3, not B31.1, with an ALGOR pipe package doing the piping analyses.

Thanks for the tip on the pipe plugs. They will prove useful in a system with very few isolation valves (that is another future system upgrade). We will evaluate time and money (isn't that always the case?) of the radiography versus the pipe plug, with time being the dominant factor.

By the way, we sucessfully modelled the entire campus steam systems (they have 4 different distribution pressures) with the University of Kentucky's STEAM2000 program, and the results of that study are now being implemented to enhance total system operability and reliability. If you have a large steam network to analyze, you might want to check out this program (

http://www.kypipe.com)

 

[Sharik]

Actually you cannot arbtitrarily substitute 100% RT or UT for a hydrostatic or pneumatic test in ASME B31.3. This Code is very specific about when you can substitute NDE for the pressure test:

1) There is a danger of cold brittle fracture

2) There is a possibility of damage to internal refractory or linings.

Both of which would probably not apply in case described. You must also realize the Codes are written for new construction and not for repairs/alterations to existing piping systems (but all repair/alteration work is to be done to the original Code of fabrication, as far as possible).

Legally then, you would have to test as much of this steam line as required between block-in points in order to test the new welded elbow installation. This may not be practical.

In real life we have gotten around situations like this by making a presentation to the local Authorized Inspector. This presentation must detail the reasons why a hydrostatic or pneumatic test is not feasible (reasons of cost or convenience are never accepted by the AI) and detail the amount and types of alternate tests to be performed.

Usually we perform 100% RT or 100% UT, 100% MT or 100% PT, we do as much visual inspections as possible of weld internal and external and we leave the welds exposed (uninsulated and unpainted) for a service test.

The main thing is, you must have the acceptance of the Authorized Inspector as this is a deviation from the strict, literal requirements of the Code.

 

[bvi]

If you are trying to comply with B31.3, the answer is different. Note that I would expect that there is no authorized inspector, only the owner's inspector, involved.

For a B31.3 system, the tie in joints are not considered within the scope of the code. You can turn to API 570, which covers alterations to existing piping, for that tie in joint. That Code gives an alternative of inspection in leiu of leak testing for the tie in joint.

Alternatively, if you are simply looking for good practice guidance, ASME B31.3 gives an option for inspection of closure joints in leiu of leak testing.

Alternatively, if the system is 150# steam or less, you can call it a Category D Fluid Service and simply do an initial service leak test.

So, with a B31.3 system, you end up with roughly equivalent choices as with B31.1, but you get there a different way. I would have the same advice, hydrotest the system except the closure and radiograph the closure joints, if testing the closure joints (e.g. using plugs) is not practical.

 

[unclesyd]

I see no reason that a hydrotest is required on the weld in your case. As stated above I would have a long talk with your jurisdictional inspector or campus engineer. An RT and another NDT method should satisfy anyone unless you are have having welding problems. We have not tested the closure weld numerous times for various reasons and have always gotten a wavier from the AI.

If for some means your are required to test the weld it can be tested by itself with specially designed plug for test individual welds. You should be able to rent a plug for just one test. The first Url is a rental place and the second is one maker of the special plug.

Remember to furnish coffee and donuts when you make the inquiry about the hydrotest.

Good Luck

http://www.improvedconstructionmethods.com/index.html

http://www.sealfastproducts.com/pipejointtester.htm

 

[TBP]

I'm having a tough time imagining a campus steam heating system falling under B31.3, especially when B31.1 specifically includes "Distict Heating and Steam Distribution Systems" (122.14). I've never heard of a university campus with their own inspector, particularly one that hasn't done much piping for the last 20 years.

A meeting with the inspector for the jurisdiction is the way to go before too much work of any kind gets done, because I'll bet there's a way of dealing with this situation that's not too onerous.

 

[tlewis]

You are correct, we will be using B31.1, not B31.3 (my industrial background is showing through).....