Feedback on UG-44 (b) Practice From One of Flange Manufacturers 6

Worms. De-canned while you wait

The problem with sloppy work is that the supply FAR EXCEEDS the demand

Forgive me if I missed anything, that article was quite difficult to read. Not sure what the relevance is to flange manufacturers, or the committee covering B16.5, the standard flange dimensions won't be changing. I think the premise of your argument is inaccurate, the moment factors were evaluated specifically on the capacity of the standard flanges, they are not required to comply with an unrelated design method, such as Appendix 2 calculations.

Fugitive emissions continue to be a major problem for bolted joints, a leakage based approach like the UG-44(b) method is a step in the right direction. IMHO the likely the next step would be to remove conservatism from the current moment factors, separate out material types and allow the moment factor to vary by size and class.

Perhaps a review of the current approach to using standard tabulated nozzle loads is required. Nothing wrong with conservative loads for strength design of nozzle junctions, but designing above operating for leakage based assessment of bolted joints is unnecessary.

@BJI ,
The main point I cited the feedback from the manufacturer here is the troubles made by the different requirement from UG-44 (b) and B16.5. The manufacturer of pressure vessel use UG-44 (b) as additional to B16.5, but piping engineers only use B16.5. As for whether B16.5 how to consider the loads for falnges, you may refer to <Backgroud of ANSI B16.5 Pressure-Temperature Rating> E.C. Rodabauch. Personally I think it should be B16.5 committee responsibility that will be much better and suitable than UG-44 (b) from ASME VIII-1 committee.
Another point, it is rare case that exert an axial tensile force and moment to flange joint (make the flange or piping as a structure/equipment support) in addition to the piping system with internal pressure, thermal load and boundary point constraints. In such case, normally it is not allowed, it may consider the way of UG-44 (b)to evaluate the B16.5 falnges. So far I have not seen such case in same rating piping line equip with different rating flanges whatever what reason is.
The methods to design a flange are greatly different in between ASME VIII-1 and B16.5. We could not use the way of ASME APP-2 to think B16.5.
Regards,

@BJI,
I disagree to use the nozzle external loads table from client specification or analysis results by Caser II from piping system for UG-44 (b). That is for the local stresses analysis at the junction of the nozzle neck to shell, not the flanges sealing point.
Regards,

ASME B16.5 covers pressure-temperature ratings only, it doesn't provide any allowance for external loads on bolted joints. Both vessel and piping designers apply B16.5 for pressure ratings. Both vessel and piping designers may choose what method they use for assessment of external loads on bolted joints, but they must be considered regardless.
I think it is appropriate for the overarching design code to apply limits appropriate to their respective design. That may allow use the of components covered by other codes and standards, within the limits permitted by those codes, or it may apply additional restrictions. ASME B16.5 provides no such limits for external loads. Refer to the below quote from B16.5:

B16.5 said:

2.5.1 General. Use of flanged joints at either high or low temperatures shall take into consideration the risk of joint leakage due to forces and moments developed in the connected piping or equipment. Provisions in paras. 2.5.2 and 2.5.3 are included as advisory with the aim of lessening these risks.

Click to expand...

B16.5 is not a design method for flanges. There is no requirement to apply a design method to a standard flange. There is a requirement to consider external loads on standard flanges. There are a number of method available to achieve this and UG-44(b) is one of them.

I think the main issue arises from using standard loads at the design stage when actual loads are not known. Why would CII operating loads extracted at the flange sealing face not be appropriate for assessment using UG-44(b)?

@BJI,
Nozzle external loads derive from internal pressure, which has already considered in flange design.

@BJI,
Please read <Backgroud of ANSI B16.5 Pressure-Temperature Rating> E.C. Rodabauch to know the formula to get the thickness of the flange for B16.5.

BJI said:

IMHO the likely the next step would be to remove conservatism from the current moment factors, separate out material types and allow the moment factor to vary by size and class.

Click to expand...

Exactly (except for the separate out material types). And the conservatism is related only to the size aspect. There is only a small design margin in those moment factors.

ASME B16.5 flanges are standard flanges designed for internal pressure only, and have an "unstated" margin for external loads. The Rodabaugh paper that mechengineer referred to, indeed has some calculations and graphs showing pipe bending stress limited by leakage of B16.5 flanged joints. Class 150 shows moment capacity well below the attached piping limit, and Class 300 shows on the lower range of the limit. So, ostensibly, there needs to be a method of demonstrating that the loads applied to the flange (nozzle in the case of UG-44(b)) is within acceptable limits. The method in UG-44(b), from the Brown PVP2013-97814 paper, is the method adopted there - the only reason for limitations on applicability is that the Brown paper only checked WN flanges. For all other flanges, the engineer is on their own.

mechengineer said:

Nozzle external loads derive from internal pressure, which has already considered in flange design.

Click to expand...

Incorrect - they come from the layout of the attached piping, the restraints, temperature, etc.

In general, my preference is (in order):
1) Reduce the piping loads. Sometimes a small re-routing can reduce the loads substantially. If the "standard" loads are too high, then reduce them, as well.
2) Check using another method - I recommend ASME Section III, Division 1, Subsection NCD, NCD-3658.3 using Level A Service Limits.
3) Increase the flange rating. Do NOT make the mistake of applying either standard loads or calculated loads using the higher Class - the attached piping isn't changing, just the vessel mating flange. If you end up increasing the rating more than one class, then you are doing something horribly wrong.

@TGS4,
Yes, the nozzle loads (one constrain point) derive from internal pressure, thermal load and boundary constraints on the integrated piping system by Caesar II analysis, the internal pressure is one of them.
Regards

Yes the internal pressure is one aspect, but it is not included in any external loads. So, your earlier statement that nozzle loads derive from internal pressure is incorrect insofar as it is incomplete. Your most recent statement is correct.

@TGS4. Thus, the point is that should not use the external nozzle/piping load table (for WRC107/297 local stress analysis on shell and nozzle neck) applying to UG-44 (b). The analysis points are different, one is the junction of shell and nozzle, and another is the flange sealing joint. As I know that currently most of EPC engineers and manufacturers are going wrong way that use Nozzle External Load Table from specifications (or API 660 Table 2) applying to UG-44 (b). And it is failed for all small size nozzle flange joints with the rating provided by pipng, then request to reduce the nozzles loads from piping. Taking a long time to discuss and clarify. That is totally incorrect loads used and waste engineering man-hours meaninglessly.
In view of above, I hope you may feedback to the relevant committee to provide a problem example for using UG-44 (b). That will be greatly help and guide engineers to understand UG-44 (b) application. Thank you very much in advance.

The load application point aspect is reality a red herring. The differences are miniscule, at best.

The ASME Code Committee is currently reviewing a proposal to reduce the conservatism in the Table UG-44-1. And permit other methods. The conservatism decrease will switch a large number of flanges from unacceptable to acceptable based on API 660 loads. Except for the really lousy flanges: 3" Class 150, 8" Class 150, 8" Class 1500.