Hydrostatic TP exceeds B16.5 flange rating

[NKBeresford]

I have an Alloy 825 pipe system -
DT= 654°F; DP = 2820psig.
TP = 1.5*(2820)*(23.3/17.484) rounded up to next 25 psig increment or 5650 psig.

ASME B16.5 Class 1500 flange rating @ 100°F equals 3750psig. (Table F2-3.8)
Per Para 2.6 - I can system test these flanges to 1.5*3750 = 5625psig.

B31.3 Para 345.4.2 (c) states that the TP cannot produce a nominal pressure stress or longitudinal stress in excess of the yield strength at test temp.
How do I determine this?
What is my test pressure for this line?

 

[BigInch]

It would have to be 5625 psig. I wouldn't test it yet. You probably are contractually obligated to bring up a design deficiency notification, or a design clarification request, if not both. If it is only tested to 5625, then the design pressure must be reduced, so that will have to work itself back into the PFD, PIDs & drawings and op manuals. The way to do that is usually initiated by a DCR or DDF.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[jte]

I'd take a different approach from BigInch. B16.5 does not limit the hydrotest pressure to 1.5*rating, it allows tests to exceed it but declines responsibility:

Flanged joints and flanged fittings may be subjected to system hydrostatic tests at a pressure of 1.5 times the 38°C (100°F) rating rounded off to the next higher 1 bar(25 psi) increment. Testing at any higher pressure is the responsibility of the user, taking into account the requirements of the applicable code or regulation.

As an engineer, I'd be comfortable going 1/2% higher than the B16.5 hydrotest statement and testing to 5650 psi which is "taking into account the requirements of the applicable code..." What's the required tolerance on the pressure gages? And... you'll go over the hydrotest pressure by a few psi anyway to ensure that you are, and stay, above the nominal hydrotest pressure for the duration of the test.

As for determining the hoop stress and longitudinal stress... ummm... if ya gotta ask, you really need to hire a consultant.

jt

 

[BigInch]

Jte, I don't see any harm in your suggestion as 1/2% higher is not usually going to present any problem. The real problem as I see it is the Responsibility, Authority and Documentation Issue... and these days it seems 50% of a project is documentation and change management in accordance with ISO 9000 tracing issues.

Since most hydrotests I have been associated with were made by the vessel fabricator or some construction company, I would not want them taking responsibility for deciding what test pressure to use on my vessels, nor do I want them to decide my allowable operating pressures.

So, we must ask two questions,

1.) Does the person making a hydrotest have the "authority" to decide what design pressure will be permitted. By limiting his test pressure, he is effectively making that decision?

2.) Does the person making a hydrotest have the authority to accept "responsibility as owner" and exceed the provisions of the design code on that basis alone?

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[BigInch]

and... given the nature of the question, I doubt it.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[NKBeresford]

Thank you for your replys.

We are still in the design phase. Nothing has been designed or purchased so I do have time. Installation estimate - Spring 2008.

Yes - I do understand that B16.5 allows me to exceed the 1.5* rating.

Yes - I did calculate the hoop and longitudinal stresses and I would like your opinion.

Hoop Stress = (Internal Pressure)(OD)/(2*thickness)
for Sch 140 NPS 14 Pipe
Hoop Stress = (5650)(14)/(2*1.25)= 31640 psi
Hoop Stress w/ Mill Tol = 31640/.875 = 36160 psi
Yield Strength of Alloy 825 is 35 ksi
So... considering the mill tolerances I cannot use 5650 for my test pressure.

Do I need to consider the longitudinal stress since I'm looking at the flanged joint in particular?

Longitudinal stress is 1/2 the hoop so even a 5625 test pressure would exceed the 35 ksi yield strength.

Honestly, I just need to reduce my DP from 2820 to 2814 to make this work. Most importantly I'd like to know what to do in the future if this comes up again.

 

[BigInch]

I would say you can't test to anything higher than 34999.

B31.3 should take precedence over 16.5, as that's the operational design spec. 16.5 is for production run item, before its been shipped, rusted, trimmed, beveled, gouged, repaired, undercut, beat into position, welded, etc.

Usually 0 psi longitudinal stress is considered the critical case, as if there is any longitudinal stress during a hydrotest, its typically tension caused from end caps or closed valves longitudinally stretching the pipe. According to a Mohr's circle analysis, axial tension reduces the maximum shear stress; axial compression load increases maximum shear stress, and compression axial loads usually only occur from restrained temperature expansions. If your pipe is flexible in regard to temperature caused expansion stresses during the test, you would not be adding to the maximum shear stress, so it is conservative to ignore any axial tension loads you may have. I have never considered axial tension loads to help a hydrotest stress calculation, because of the somewhat unpredictable nature of determining flexibility and restraint provided by soil (as I am usually doing pipelines).

In the future, the design pressure should be determined by evaluating the allowed test pressure of all components that will be exposed to test pressures. US pipeline regulations and the B31.4 and B31.8 codes actually determine the design pressure by dividing the documented test pressure by the appropriate area class factor, so test pressure levels are critical in the MAOP determination.

I haven't seen the new EN codes yet and I am curious to know if they provide for different test pressure levels depending on an area classification or fabrication factor.

If you reduce the test pressure, be sure to file the appropriate design notification form, so any repercussions to allowed operating pressure can be determined, if any.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[dcasto]

you only need to test to 4230 psia. B31.3 does not require a strenght test, you have qualified the strenght by design and testing of componets before assembly. If you want to do a strenght test, 25 psi isn't going to hurt anything.

Think about this one. I had a heater with 1 1/4% CrMo and the heater was designed with 600# flanges for 300 psi at 900 degrees the reactor was 304 stainless with 300# flanges and the long piping run between them was 105CS with 900# flanges, what would be the test pressure for the assembled unit?

 

[BigInch]

dcastro, As for your question, your test pressure must be F x Design Pressure, and all materials used and all vessels installed in the system should withstand and have proved or have been tested to at least F x Design Pressure.

Agreed that 25 psi may not hurt anything, but whether it "hurts" anything or not is irrelevant to the code.

From the OP,

B31.3 Para 345.4.2 (c) states that the TP cannot produce a nominal pressure stress or longitudinal stress in excess of the yield strength at test temp.

So +1 psi is a violation any way you look at it.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[dcasto]

If you use the design formulas under 31.3, there is no way to get close to the yield strenght of the pipe, there is nearly a 4 to 1 safety factor.

 

[jte]

dcasto-

That 4:1 "safety" factor is a design margin against ultimate tensile strength... not yield. I'm not as familiar with B31.3 but with VIII-1 its easy to find common steels at common temperatures which are governed by 2/3 yield and not UTS/3.5. So its not that unusual to get close to yield when hydrotesting.

jt

 

[rconner]

I noticed that there was a mention in the initial post of this thread of a "flange". I suspect most posts of this thread thus far have appeared to address stress level in the steel of the pipe or flange, but I was just curious since I don't deal with such high pressures very often myself if there needs to be any consideration of the specific gasketing means used in any flanged joints?