ASME B16.5 Flange Ratings 1

[johnnymist2003]

Hi all,

Philosophically (gees, a big word for an engineer speaking.......

Assume I have a Group 1 (say SA-105) Cl. 150 flange, at 100 degF, operating at 285 psig (this corresponds exactly with the pressure-temperature rating in ASME B16.5). I also have an external moment and a radial force acting on the flange (eg. from piping).

According to norm, one would convert the radial force and moment acting on the flange into an equivalent pressure, and then add it onto the design pressure. It is also the norm to compare the "new" design pressure with the pressure-temperature rating in ASME B16.5 to ensure that the tabulated value is not exceeded. Assuming the operating pressure = design pressure, the "new" design pressure of the flange will be greater than the tabulated value of the pressure-temperature rating, indicating that the flange is no longer suitable for use. In my opinion, this indication is not necessarily true. In some of the older refinery plants, the piping was never designed using say "caeser" software, thus the piping loads may not have been accurately calculated, yet the flanges, which may have been there for years already, are still leakage/problem free. Does anyone have any other feelings on this issue? Do we go ahead and carte blanche condemn piping because the so-called "norm" of calculating equivalent pressure + design pressure and comparing to the ASME B16.5 value now indicates that the rated pressure is exceeded ? I think not.

Obviously this also begs the question "what about the existing vessel flanges attaching to the piping ?"

Your comments please.
Thank you

 

[jte]

johnnymist-

"In some of the older refinery plants, the piping was never designed using say "caeser" software, thus the piping loads may not have been accurately calculated, yet the flanges, which may have been there for years already, are still leakage/problem free."

Please let me know which refinery you've seen which did not have an engineering/maintenance force that was not familiar with wire wrapping/peening or clamping flanges. Such a happy place would be a good place to work.

OTOH... Yes, for the most part there will be some extra margin in the flanges. I'd take (actually, I have taken)two approaches to your question: For new construction, include the equivalent pressure due to external loads in your design pressure. No need to build in problems.

For existing piping (rerates and/or modifications) determine the consequence of a flange leak. If a leaking flange results in dripping some cooling tower water to grade I won't be too concerned. If a leaking flange is in 1500 psi hydrogen service I'd be very conservative. Most refinery piping will fall somewhere in between. If I'm looking at leaking diesel and lower, I'd warn the unit owners that they may experience a flange leak under the proposed conditions and we may have to clamp the flange. If possible, I'd break the flange pair and put in a GMGC (kammprofile) gasket and lube and torque new studs and nuts. This will significantly reduce the probability of failure. However... an expert witness could still tear you up in court so don't forget to CYA and make it someone else's decision. "If you neglect external loads, it may leak. The probability of leakage will be reduced by... My recommendation is to increase the flange rating."

jt

 

[TomBarsh]

good comments jte.

A number of other factors to consider:

1- The external loads may result with an equivalent design pressure exceeding the pressure-temperature rating. But there is a service factor or safety factor that was applied to determine those ratings. So the end result is that even though the flange rating has been exceeded there is still some amount of "reserve" capacity remaining, but we've crept into the safety factor. Based on this fact we don't need to blindly condemn the piping (flanges) but understand that by doing so we don't have the safety factor that we may think.

2- The flange loads from piping may not be calculated accurately. It may be that the loads the flanges have actually been subject to over the years may be less than the calculated loads. There is a lot of fudge room in the pipe program assumptions.

3- The flange pressure-temperature ratings are based on several possible failure modes for the flange and connection (these are discussed in Annex D of ASME B16.5). But when a standard flange is analyzed using methods of ASME Section VIII Div 1 Appendix 2 there is a wild variation in rated MAWP for these flanges. When analyzed for MAWP by Appendix 2 some standard flanges fail to meet their published pressure-temperature rating, while others far exceed the published rating.

My own opinion based on observation of MAWP calculations of standard flanges per Appendix 2 is that "stockier" flanges (smaller pipe size flanges with a high class rating) will exceed their rating from B16.5 but that "skinny" flanges (larger pipe size with a low class rating) will not meet their rating. This issue may have an impact on the decision whether or not to go to a higher class flange rating.

 

[tothepoint]

Right on TomBarsh. ASME/ANSI flanges were not designed using section 8 div 2 rules, and their code calculated MAWP do not relate consistently to the B16.5 PxT ratings.

I don't know that I would use the section 8 equivalent pressure adder when working/designing with B16.5 flanges. Instead, always move up to a higher flange class if your PxT is close to the B16.5 allowables- especially on NPS 3, 8 and 12 x 150.

 

[johnnymist2003]

Hi All,

thank you all for the valuable posts. I agree whole-heartedly with all the sentiments. jte, yes, in critical service applications, to be more conservative is the safest route, and yes, maintenance teams are aware of the problems, and solutions, to leaking flanges (including those you list). TomBarsh, you hit the nail on the head with what you said. I believe it is nonsensical to "condemn" the piping flanges because the "theoretical" pressure-temperature ratings have been exceeded. The question really is how do you quantify and "prove" whether a flange is acceptable or not, without doing a proof test on the line ? It's a touchy area, on the one hand you need to be safe, on the other hand you need to be aware of the economic impact of increasing flange ratings. Personally, I'd go the route of increasing the rating on critical applications, and then argue with the bean-counters afterwards. But that still does not solve the quantification problem.
John

 

[TomBarsh]

John,

Safety is paramount (consider the recent 'meltdown' at a refinery near Houston that made the national news with a number of deaths). But there is no sense in blindly throwing money at replacing every possible flange. As you say, address the critical applications first, then work on from there as time and money permit.

Regarding the problem of quantifying...if you have specific cases of installed piping and flanges and you're concerned if the flanges are safe or not, then I suggest contacting Paulin Research Group (

http://www.paulin.com).

They have some new FEA software specifically for analyzing flanges for leakage, stresses, etc. and they can provide consulting for specific problems.

(no, I do not work for PRG)

 

[reichertc]

A little late to this thread but...

In piping systems in jurisdictions where I have worked, rerates and modifications are typically permitted to be designed to the original code of construction. This has permitted me to use alternate allowable stresses for piping design. I think I recall having a discussion with someone in our pressure vessel group about them being able to do the same thing.

B31.3 has a great little clause in it that allows qualification of a fitting or component based upon "past successful service experience". On the piping side at least, that allows a "like for like" replacement to be put into service. Of course if the reason that you are replacing something is because it has failed, then it probably wouldn't be appropriate to call it successful...