31.3 and 31.4 test 1

prens11us · Jul 10, 2006

Hello all,

we have a 4500 meter 48" pipe line. 3000 meter of it is designed in accordance with asme B31.3 and 1500 meter of it is designed in accordance with asme B31.4. the line is continuous line.

31.3 states test pressure of 1.5 X Desgin pressure.
31.4 states test pressure of 1.25 X Desgin pressure.

Which criteria should be used.

Or is there another way of soving this problem.

thanks in advance.

ApC2Kp · Jul 11, 2006

prensllus,

Test pressure, whether 1.5X or 1.25X, could be viewed as a minimum test pressure. You are allowed to test to a higher pressure, but should check to not exceed 90% of yield stress (SMYS, not allowable stress SA) for pipe wall and other components subjected to test pressure. The corrosion allowance in the wall thickness could be included in the available wall thickness for check of not exceeding the yield stress in pipe wall during testing.

The large 48" diameter of your pipe probably does not have much extra wall thickness for over-pressure. If the same pipe wall thickness has been used for both the B31.3 and B31.4 sections of pipeline, then I would suggest testing entire pipeline to the 1.5X design pressure of B31.3 code.

prens11us · Jul 12, 2006

thanks apc2kp

BigInch · Jul 15, 2006

As apc2kp noted, test pressures are a minimum, you can test higher.

I do not know of a reason to limit test pressure to 90% SMYS. apc2kp, if you have a reference, I'd appreciate knowing what it is. I personally have tested to 110% in unusual circumstances involving extreme elevation changes.

NOTE! You might not be able to test at 1.5 X through both segments if there is a significant elevation change. Since your line is so short, its not likely you have a problem, unless you're going up a 3000 meter riser to the topsides of an offshore platform. Hey, its a possibility, right?

Therefore, before I give any tips, I would want to see a list of,

Distance from start, Code, Design Pressure, Elevation, Pipe Mfg Spec, Pipe Grade, Pipe Wall Thickness.

Include any valves, fittings, flanges and their ratings too.

http://virtualpipeline.spaces.msn.com

ApC2Kp · Jul 18, 2006

To BigInch,

Both the B31.1 and B31.3 piping codes use full yield as the limit for the 'not to exceed' stress value. I realize that the 90% of yield stress is a company standard procedure to limit test pressures. It might include a 10% pressure build up factor for the safety relief valve on test pump.

Testing to 110% of yield stress certainly does not guarantee any failure. Most materials are produced such that the minimum strength required will always be achieved, to avoid reject scrap. The situation is apporoaching the use of Probabilistic Design methods to take into consideration average material properties plus variations. Perhaps in the past many materials were produced with strengths testing out at 10-20% more than the required minimum values. Maybe more modern manufacturing processes (Six Sigma, etc.) may have allowed materials to be produced consistently to a 3-5% extra strength over the required minimum strength. Is it still be possible to test to 110% of SMYS? Most likely yes.

If one had to push the limits of a pressure test, then I suppose that 1) corrosion allowance would be backed out of the deduction from wall thickness, 2) the measured actual wall thickness could be used for calculation instead of nominal wall minus mill tolerance, 3) actual yield strength from MTR could b used instead of SMYS of code book, 4) the weld efficiency factor would be considered.