hello everyone.

Just curious, B16.5 states that flanged joints may develop leakage problems at temps above 200 C. How about pipe sizes? Does this statement include "all sizes"?

Need clarifications.

Regards,

can u tell me the page no in B16.5 where it tells so ????????????

What B16.5 does say is
[2.4.2]"When used above 400"F, Class 150 flanged joints may develop leakage unless care is taken to avoid imposing severe external loads and or severe thermal gradients. For other classes, similar consideration should be given above 750°F."
I can't comment on that, but the implication is certainly all sizes, and by extension all gasket types too, which doesn't seem quite right somehow.
It is well known however that many B16.5 flanges 'fail' when run through a Div II app3 flange calculation. I've recently looked at [I think it was] a 3" 150# flange where the bolts 'failed' under seating stress requirements for a SS spiral wound gasket. We all know they work however!
Oh the joys of flange leakage predictions! Must get on with doing that EN1591 calc sheet sometime.................................

Ahoy there Captain Kenny,

I agree with you. I have found many standard B16.5 flanges that fail seating stress limits to ASME VIII div 1 App 2, but are ok in the operating case. It seems like an inconsistency in the ASME code. Use of a spiral wound gasket with high a seating stress doesn't help the numbers either.

Bolts may fail due to the very conservative (and not mandatory) allowable stress of around 25ksi for alloy steel, where actually the bolts are often tightened nearer to 50% yield, or around 50ksi on site.

A calc sheet (MS Excel ?) for EN1591 .... Besides being immensely complex, written by committee and poorly explained in both EN1591 and EN13445, the method has two iterative loops nested inside one another. Worse if it's a lap joint flange. Good luck with that one.

Cheers.

Most flanged joints leak. Anyone that's ever spent a day walking around a benzene unit can attest to that. Improperly installed joints (incorrect bolt torque values or even incorrect sequencing of the tightening of the bolts are common problems) are frequntly to blame, but even proper installations can and do leak.

For spiral-wound gaskets, there are two parallel open passages from the pressure side to ambient. They are very small, and have immense L/d ratios, but they're there and they do leak unless the process fluid is quite viscous.

For elastomeric gaskets, the problem is worse, since those materials typically take a set after very few hours of service. Most plastics and elastomers are subject to severe creep deformation at room temperature under loads of the magnitude found in gaskets.

Information is available from the gasket manufacturers. It's not uncommon for flanged joints in LP steam service using PTFE gaskets to have to be retightened a couple of times each year. Talk to your local gasket sales rep. He will undoubtedly have access to photos of numerous types of gasket failure.

And "flange leakage" calculations are at best an inexact science. Attempting to do more that two or three of these rigorously in one lifetime would leave little time for anything else, like eating and sleeping. So we use a lot of approximations. Try to trace the hsitory of the derivation of the "Gasket factor 'm'" or the Design Seating Stress 'y' as shown in Mandatory Appendix 2 of the BPVC, for example. Start by reading Note (1) for table 2-5.1 and then spend a couple of hours researching how they came about.

So I'm not surprised to find that "flange leakage" calculations turn out to be incorrect now and then. Actually, I regard it as a minor miracle that they ever work at all. Not that I have any better tool to use; I do them from time to time and have learned to present favorable results to clients with a straight face and deliver the usual disclaimers for cases where the calculation fails.

I don't do flange leakage calcs and I don't do bolt torque calcs. Waste of time. They are office bound engineer's dreams. Reality on site is 5lb hammer and strong spanner but sometimes no spanner. When customers ask for torque values I tell them to get them from their gasket supplier as "they know best".

I agree with the views of all of you and "hate" to do a Flange leakage analysis ( and thereby adding excesssive flexibility to minimize bending moments and/or relocate the valves to convert high bending to torsional moments for thermal based loadings)based on the existing ASME ( Sec VIII Div 1 ,2 and Sec III i.e. except EN1591 rules). Somehow, most consultancies are hell bent on this requirement for typically ( > 900 Lb )high ratings.Yes, a spreadsheet for EN1591 is indeed a dream for all of us, but I don't think it will happen in near future, because of the highly involved methodology of this standard.I believe that lesser importance is given to the practical aspect of a solution to Flange leakage and too much is concentrated on theoretical qualifications.

Flange calculations are sometimes essential.

EN1591 / EN13445 has an advantage over ASME in that it handles a flange including a lap joint flange as a connected assembly of flexible components, like a reduced FE calc. ASME basically ignores the aspect of flexibility, and can cause big problems with high yield materials.

There is a piece of software available that will do both ASME and EN1591 / EN13445 flange calculations, and has a quick builder to assemble the huge amount of data needed for the latter. Visit www.pipemill.com.

The need to address flange leakage seems to be very variable across different fields from what I can see. I'm doing offshore stuff at the moment and almost every hydrocarbon, or blowdown line in the North Sea is now required to withstand some sort of nominal blast load [see some the recent posts on this].
Looking at the FABIG advice on this matter it is readily apparent that flanges and connections are THE critical components within any system and the system will not behave in a ductile manner absorbing the blast energy if the joints within it behave in essentially a 'brittle' manner. Given that almost 90% of brownfield modifications offshore are of flanged / hubbed construction because of the cost and implications of welding offshore and you can see that suddenly performing some sort of flange assessment is not just some 'office bound' engineers dream but a potential nightmare. Regardless of the methodology used, doing some sort of assessment is always going to be better than doing nothing and trusting 'judgement' or worse ignoring the issue completely. It is also sobering to note that the HSE also found that for mechanical failures in static offshore processing equipment the failures were largely from leaking gaskets (22%) [Offshore Technology Report 1999/064], and that is just in normal operation!
Personally I do a quick check on every flange in the system [takes a few minutes per run once the spreadsheet has been set up], using the immensely simple equivalent pressure and NC3658 methods. They may be simple and flawed in many many ways, but they are easy to implement and do not encourage pushing the performance of the flange because of that simplicity. If flanges need to be relocated or increased in rating so be it. Using more detailed methods may help justify the odd highly loaded flange set, but the work load is always much higher.
Of course it helps if you specify controlled tightening procedures on these flanges as is now pretty well universally specified in the UK offshore sector [see HSE research report 253 "Piping systems integrity: Management review"] rather a relying on a bear with a big spanner.

Capn. Kenny,

I agree with you. I would use NC3658 with some caution personally 'cos it takes little account of the physical make-up of the joint, and kind of assumes that ASME III level surveillance will be there. Similarly I reckon the Kellogg eq. pres. is fine as a rough check, but if it fails, think again before making piping changes, and do a Kellogg + ASME VIII. App 2.

I recall an Offshore Technology report that suggested 'under-bolting' was responsible for many 'fugitive emissions' and some fires in North Sea developments. Probably nearer the truth is an absence of any bending moment check or poor alignment to start with.

It is not only different industry to look at but also average contractor/fitter on field. If you can control the erection, maintenance and the application is critical then I would look at the calcs and issue detailed instructions. But then I would also look at hydraulic bolt tensioners instead of torque values. Most of the time I cannot contol the workmanship so I try to design simple and robust.