Flange leakag: PEQ: allowable stress

I have a pipe stress spec that states "while using equivalent pressure method if leakage is detected the same may be rechecked with 1.5 x allowable for operating case".

My question is where and how do you change the allowable for the recheck?

The flange rating data is stored in two files found in the SYSTEM folder - e.g., ASME-2009.FLG and ASME-2009.G.
Change one line in the FLG file to use 1.5 times the rating. In the line following "* CONVERSION FACTORS -- TEMP SCALE, TEMP OFFSET, PRESSURE" you see the line "1.000000 0.000000 1.000000"; change that to "1.000000 0.000000 0.666667"
I suggest you save this in a separate set - e.g., ASME-2009_times_1.5.FLG and ASME-2009_times_1.5.G.
Verify proper use...

Thanks Dave,
I hate sound stupid, but where do I find the SYSTEM folder?

easiest?
On CAESAR II main menu, select Utilities. There you will see "open System folder".

found it. set the file to open with Caesar II Application. but after opening it goes back to the utilities window.

open with Notepad

Made the changes. both were saved as a text file.???

So, how do i select this particular set of allowables for the flange check. it didn't show up in the "read from file"?

get rid of that .txt extension so that those two new files have the same extensions as the originals. If located in SYSTEM, CAESAR II will add your new filename to the available list of tables when you are selecting tables inside CAESAR II.

Dear Sir,

We have performed Flange Leakage analysis applying Peq method and CAESAR II predicted failure. We took out the forces and moments and calculated the flange stress using the Kellog Method, that is used by C II. The calculation gave exactly the same value given in C II output. We observed from the calculation that more than 92% of stress comes from design pressure and only 8% is contributed by Axial force and moments. By changing the pipe routing, we could not achieve reduction in stress. When we discussed the situation with customer, he advised us to use NC method.

When NC method was used on the same flange joint, it predicted much lower stress (10% approx) than what was predicted by Peq method. This is strange. Two methods can not predict results with so much of variation. Minor variation may exist.

My question is:
Did any body face similar situation? If so what shall we do? What went wrong with our input? We will be thankful, if somebody can through light on this.

Regards
S. RAGHAVA CHARY

My opinion on NC method is in http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=showflat&Number=53645
By the other hand, PEQ fails in establishing a limit of the method; at the first approach it seems logical to consider the "rating pressure" as limit for "total pressure" however we cannot prove that this limit gives leakage. In my opinion a good improvement of a limit for the "total pressure" evaluated by PEQ would be a formula using Blick theory, in line with NV-RP-D101
Quote
There are also variants of the “Pressure Equivalent Method” in use by some Engineering Companies. One of these variants is the “Blick Theory” from the 1950’s. Refer also “CASTI Guidebook to ASME B31.3 Process Piping”, chapter 2. A piping system designed according to ASME B31.3 shall have the flanged connection hydrostatically tested to 1.5 x design pressure and it has therefore been a practise among many large Engineering Companies to allow the total equivalent pressure including external bending moments and axial forces to reach a level of 1.5 times the rated flange pressure at temperature.
Quote

In brackets, the two methods you mentioned do not predict stress in flange, so in fact we cannot say "results with so much of variation" because we haven't a common reference. But, of course, I understand what are you referring to.

The Peq method, especially if you are not applying any additional factor on the allowable, is known to be very conservative. NC-3658.3 method on the other hand in my opinion is unconservative, for some situations it will tell you the flange is fine at a bending moment that has yielded the pipe. I am not surprised you have found an instance where the results from the 2 methods are out by a factor of 10. If using NC-3658.3, I would only accept the flange if the result was less than 30% of allowable. Otherwise I would go to the more detailed flange stress / leakage calculations within Caesar.

Dave,

Do you mean that NC-3658.3 qualified the flange but piping adjacent to flange is yielded (Sy status)?
And is the rule "30% of allowable of NC-3658.3" validated by other method more accurate (as FEA)?

Yes, there are flanges / pressures / bending moment combinations that will pass NC 3658.3, but the bending moment is high enough to cause yielding in the adjacent pipe.

The 30% of NC-3658.3 just roughly gets the allowable bending moment down to a point where flanges pass Caesar's ASME Sec VIII flange stress and leakage calculations.

Thank you for your feedback.
About the method; I think it is useful to study the original work, which can be downloaded at www.ornl.gov/info/reports/1976/3445600189702.pdf
In fact, Mr. Rodabaugh (in B31.3 2012 list of B31 MECHANICAL DESIGN TECHNICAL COMMITTEE appears to be still active as Honorary Member, Consultant) applied there a "new" method to calculate stress in flanges, which is correct in my opinion.

Thank you all guys. Your questions and feedbacks have just helped me to resolve some flange leakage problems...