Flange Face Machining

[stoker123]

Steam cutting of a raised face 1500 lb 20 inch feedwater flange set has led to grinding the raised face 1/4" below B16.5 dimensions (raised face side). System design code is B31.1. System pressure is 1250 psig and temperature is 427 deg. F.
Schedule is pushing to not weld build back up. Machining will re-level face. Assuming no metal removal from the bolted faces (no degredation of strength structurally), is the risk only the potential to leak during operation with the reduced raised face? System flexibility will be checked for effect of 1/4 " make up. Is there any additional structural check that need be done?

 

[racookpe1978]

Go back to first principles:

The purpose of a raised face RF flange is twofold:
(1)when you make up the flange and gasket, the bottom two bolts/studs make a place to "hold" the gasket while you insert the rest of the bolts, and the RF makes it much easier to separate the faces in the field during disassembly.
(2) the smaller RF area concentrates the clamping force between the two flanges in a smaller area.

Thickness of the RF - as long it is not ground flush !!!!! -is not critical to its sealing capacity. You must be able to verify- (before applying pressure) - that the flange has has not "topped out." That is, make sure the outer rim of the flange has not touched the outer rim of the mating flange. I don't think that will happen realistically. Flanges are pretty stiff.

Flatness IS critical.
Perpendicularity IS critical.

Flange thickness (compresson and bending resistance, sealing strength) does not seem to be affected by grinding down the RF - in this case.

Retests? A hydro at least. You are making (machining) an entirely "new" steam seal surface. This is NOT re-bolting up the same bolted flange that was taken apart.

I would not walk up to that flange with steam pressure behind it to look and see if the machinists actually ground it flat enough until after a full pressure test with a cold fluid. I'd would not tell any pipefitter to do so either.

If your 'scheduliar" can't find time for hydro test....tell him to be the one who walks past that flange the first time and checks it for leaking....

Any one see any problems in my logic?

 

[racookpe1978]

Part II.

Is anybody thinking they can do this machining by hand? You'll need (at a minimum) a portable flange-face machine tool, set time, and time to break the joint and get the (good) flange bent out of your way. That too will take time.

Seems to me that replacing the flange isn't that long a job (2-3 shifts) and will be less threatening than machining down a steam-cut flange.

RFSO or RFWN as the original flange? You could set up and prep and weld out a RFSO on a 12" stub of pipe in the shop, prep its back end. While the shop is working on the new stub and RFSO, another crew is cutting 12" (with weld prep, gasket thickness, and root gap considered) off the existing pipe and weld prepping the root. Do the shop NDT, pass it.

From the shop, take the new stub topside and make a simple 1/2 shift butt weld. NDE, Xray, and you're ready for the same hydro you'd need anyway. Practically before they can take the flange face machien up to the job site.

 

[Shmulik]

stoker:
ASME B16.5 (6.3) particularly regards to permitted flat face flanges that the raised face has been removed without reducing pressure-temperature ratings:
6.3.1 : The thickness of a class 150 or 300 flange from which the raised face has been removed shall be no less than the applicable dimension C of tables 9, 10, 12, 13 minus 0.06 in.
6.3.2 : The thickness of a flange of class 400 or higher from which the raised face has been removed shall be no less than the applicable dimension C of tables 15, 16, 18, 19, 21, 22, 24, 25, 27, 28 .

 

[racookpe1978]

Granted. That works for new construction - when neither flange is in place. Or repairs, when the FF flange is already welded up and only a RF is available as its mate.

---

But fully removing the RF in this case (from the upstream flange, for example) would mean the downstream flange is STILL a RF.

So you'd have a FF(cut down)-gasket-RF(original) sequence - and that can't seal.

You can only seal reliably with either a RF-gasket-RF or FF-gasket-FF.

 

[MJCronin]

Post in the "Piping and fluid mechanics" forum