Hi All,

Sometimes we receive jobs where we have to rerate the piping at higher temperature or pressure. In some situations the required code hydrotest pressure becomes higher than previous hydrotest pressure and a rehydrotest is needed.

In such situations we check if pipe is Ok for new hydrotest pressure. We check the thickness of pipe for hydrotest pressure using the equation given in B31.3 but using 90% of yield stress as allowable. We also use the same equation to calculate the required thickness for design conditions.

In one such situation we have Wc less than 1 due to very high temperature. We checked the thickness for the design pressure and temperature using Wc less than 1 as required by the code. While checking the pipe thickness for hydrotest I am not sure whether we should use Wc or not. One argument is that since piping has been in operation for many years, the weld strength may have gone down and therefore we should use Wc in our calculations.

To summarise should the Wc factor be used when checking the suitability of a pipe in the hydrotest case?

Could any senior member comment on this please?

Thanks and regards

MW

Just a few points...
B31.3 "is not intended for the ... testing .. of piping that has been placed in service". (300(c)(2)).
The Required Leak Test shall be performed "prior to initial operation" (345.1)
"The test pressure may be reduced to the maximum pressure that will not exceed (longitudinal stress) yield strength at test temperature". (345.2.1(a))
Minimum test gage pressure = 1.5(internal design gage pressure)[(allowable stress at test)/(allowable stress at design temperature)] (345.4.2)
"The designer is responsible for the application of weld joint strength reduction factors to welds other than longitudinal and spiral (e.g., circumferential)." (302.3.5(e)) - this is a change for 2008. Wc is the CAESAR II term for the circumferential weld strength.

and my own opinion -
I assume your test temperature will NOT be in the creep range. So, even if you apply Wc, Wc=1 when you calculate the longitudinal stress at test condtions.

Hi Dave,

Yes the test temperature is ambiant but since the pipe has been used for many years at around 800C (creep range)the weld strength may have gone down.

To check whether test pressure calculated from 345.4.2 would cause yield in the pipe or not, we have to use equation 3a. Now do we need to take W less than 1 in this calculation.

Since we take account of the reduction in allowable strength while calculating the test pressure using 345.4.2, one could argue that if the pipe is to be used in creep range, the effect of weld strength reduction should also be taken into account. Although code does not say anything about this but what would be a good and safe engineering practice?

Let me know your opinion about this.

MW

You point to equation 3a. That is internal pressure design thickness equation. It set thickness based on hoop stress. It incorprates creep through the terms Y and W. Wc does not affect this calculation. (Neither does the CAESAR II term Wl.) But (3a) is thickness, not stress.

The Code (345.2.1(a)) says the "nominal pressure stress" should not exceed yield strength at test temperature. What does nominal pressure stress mean? (That term is also used in Fig. 323.2.2B.) To me, nominal means a simple hoop stress calculation such as PD/2t.

No where does this W (CAESAR II's Wc and Wl)come into play in calculating stress. In B31.3-2006, W was used on the other side of the equation to reduce the limit of calculated stress due to sustained loads (302.3.5(c)). W was removed from 302.3.5(c) in teh 2008 Edition.

My opinion? Based on what I know, I would not consider W in your evaluation of test pressure. But I am not a practicing piping engineer.

This may or may not help...
The hydrostatic leak test is a "leak test". It is one type of leak test. (Others being pneumatic leak test, initial service leak test and sensitive leak test.) A leak test is used to "ensure tightness" (345.1) It is not a proof test - meaning, if we survive this test, the system will not burst. Only the hydrostatic leak test appears similar to a proof test.

Hi Dave,

Sorry for causing the confusion. I should have used W instead of Wc. The factor W I reffered to is Weld Joint Strength Reduction factor which is used in equation 3a.

Here what I am trying to do.

1. Calculate hydrostatic pressure needed using 1.5*Design Pressure*SH/SC

2. Check using equation 3a if the pipe thickness is OK for hydrostatic pressure. I used this equation to calculate stress by substituting values of wall thickness, hydrostatic pressure, W, Y and E.

3. Compare calculated stress with yield to see if hydrostatic presure is OK.

4. My question is should I use W=1 in step 2 or W=0.5 as the pipe will be operating at 815C?
If I use W-0.5 I will have to reduce the hydrostatic pressure as the stress in the pipe exceeds the yield strength.

I hope I am able to explain my question clearly.

Thanks and regards

MW,

Your Step 1 should be Test Pressure=1.5*Design Pressure*ST/S, where ST/S is SC/SH, not SH/SC.

Like I said earlier, (345.2.1(a)) "nominal pressure stress" should not exceed yield strength at test temperature. Apparently you using the pressure wall thickness as your Step 2 check. This equation does not use yield stress. I figure that, with your approach, you should be using yield stress (or 90% of yield?) in place of S in equation (3a). Likewise, since you are checking the TEST CONDITION, W should be 1.0 as you are not at a creep temperature that may reduce the strength of the weld material.

Unfortunately, the Code does not offer a stress calculation that uses W. Instead of increasing calculated stress (as does an SIF), the W reduces the limit to stress (SL
One problem here is that you are to reproduce the severity of the design stress state. That's tied up in that SC/SH (or ST/S) term in Step 1. But the definition of S is "allowable stress value at component design temperature (see Table A-1)". Now, before there was a W, the allowable stress value was simply SH - the basic allowable stress at design temperature. In B31.3-2006, the longitudinal stress due to sustained loads (SL) was limited to W*SH. Is that the "allowable stress value"? But now the 2008 Edition removed that W from W*SH, so, once again, I see no reference to W in the allowable stress value.

Another problem is that the Code now says Wc (using CAESAR II terms) is the responsibility of the designer. The Code wants you to use W (CAESAR II's Wl) in the thickness (or burst) calculations but leaves it up to you for bending (CAESAR II's Wc).

The Code is not a cookbook.