Help needed for a Hydro-Test 2

[iam42]

I am "trying" to Hydro a 42" Pipe spool with 600# Flanges. We are now on our third set of gaskets and our second set of Blind Flanges.

We have been able to achieve a pressure of between 1500 and 1800lbs before it springs a leak at the joint.

Does anyone have any tips on Hydro-testing large bore spools.

I have to admit I have very little experiance of this size.

Thanks

 

[zdas04]

"Springs a leak at the joint" is colorful, but not very informative. Are you having gasket leaks or material failure? What kind of gaskets are you using?

David

 

[iam42]

We are using Spiral Wound Gaskets and the leak is from the gasket joint.

Thanks

 

[jte]

I'd suggest trying CMGC (corrugated metal graphite coated) gaskets such as Graphonic from Garlock. Guess its a bit obvious, but torque the nuts...

Just to verify - the blind flange is one which has bolt holes in it - not a slip blind which gets sandwiched between two flanges, right?

jt

 

[btrueblood]

For any hydrotest over 150 psig, we usually have an O-ring groove machined into a set of blind flanges. Pick a ring OD that fits within the expected gasket sealing area, and size the groove using the Parker O-ring handbook. For the high pressures you are talking about, I would look at getting a dished or domed cover, and possibly one with extra reinforcement (e.g. more thicker) to reduce the amount of bending you will expect at the test pressure (this assumes that the flanges/covers are for test purposes only, and not a part of the spool).

 

[iam42]

It is a temporary blind flange with holes.

 

[ApC2Kp]

iam42,

Check the bolting. Spiral wound gaskets usually require high strength ASTM A193 Gr B7 alloy steel stud bolts. Ordinary steel A307 Gr B machine bolts could yield before the spiral wound gaskets are seated, or hydrotest forces could cause yielding of bolts.
You mention the spool is 42". What is the flange standard that is used for the dimensions? What is the thickness of the flange? Thickness of the blind flange? Number and dia of bolts? These could be 2.5" dia bolts, and how are bolts being torqued? Are these flanges weld neck or slip-on type? (ASME B16.5 has dimensions up to 24", so we need to know flange joint design)

Check that the gasket spiral winding area is comnpatible with the flange face dimensions. Some spiral wound gasket designs include an inner compression ring along with outer compression ring to prevent spiral winding inner bulging failure during compression by flanges. Use gasket vendor's torque specifications, torque by 2 or 3 stages usng crossing pattern on bolt circle to reach the final torque.

Check gasket sealing surfaces for flatness and the machined finish (usually 125-250 rms spiral pattern serrations). The welding heat could distort slip-on flanges so that flatness of sealing face is 'cupped'. Total flatness over the 42" dia must be limited to 0.020" or less? Mis-handling could damage the serrated finish and leave a groove or scratch across face for a leak path.

 

[jte]

It is a temporary blind flange with holes.

... and (as queried above) which standard is the thickness determined by? Or have you run hand calc's to determine the required thickness?

Guess what I'm suspecting at the moment is that the temporary blind flange may be bending enough under pressure to both rotate the flange as well as scuff the gasket radially as it "sucks in" the perimeter as the center bulges out. Although restrained by the bolt holes, there will be some play between the holes and the studs. I suffered from this effect with some specialized testing I was overseeing a few years ago. Especially if you have several uses for a flange, I'd consider using a regular flange mated to a 2:1 SE head. Might be more expensive up front but less likely to see the problems a flat blind will when sealing.

jt

 

[unclesyd]

ApC2Kp beat me to the bolt question.

Is this a machined raised face blind or a flat plate blind?

You can rent a proper blind from either one of the following companies.

http://www.usaindustries.com/pipe_blinds.htm#blindflanges

http://www.twincoindustrial.com/flanges.html

 

[iam42]

The 42" Flange is a RFWN.

The blind is a rented RF Blind.

We are using ASTM A193 Gr B7 bolts.

Thanks

 

[unclesyd]

28-2 1/2" studs torqued to what value?

How are you torquing the studs?

Are you using the proper torquing pattern?

 

[BillBirch]

For that size flange you absolutely must use hydraulic tightening equipment, e.g hydratight. Wht did you change the blind flanges - trial and error, or a genuine problem>
Regards,
Bill

 

[iam42]

Bill,

The original Blinds had slight scratches on the face. We are torquing to 5000ft/lb.

 

[unclesyd]

Your troque value is way too low.

My tables give:

dry .223 15,331 ft lbs

lub .167 11,481 ft lbs

These values give a clamp load of 330,000 lbs.

 

[unclesyd]

A little clarification for my last post.

The torque values given are essentially the maximum usable values based on 95% of yield strength. You can calculate the needed values for your test which will be less that what I posted or you can use these values to give the maximum available clamp load.

Watch your friction factor on these studs.

 

[Cockroach]

This is a typical problem, gasket leakage. One obvious issue is that you are using a gasket for the ANSI Class 1500 pressure rated spool, typically these are ring joint designs. The gland and ring area provide you with a mechanical seal on the surface of the donut when loaded with nominal torque along the bolting diameter studs.

What you can do, not sure if it has been suggested, is apply a gasket sealant to both surfaces prior to bolting it up. This will help to remove micro leak paths.

Also, leaks are highly dependant on exactly how you torque the fasteners. It is best to use the star pattern approach after the nuts are put on hand tight. My practice is to put minor pressure to the assembly, say 150 psi, remove or neutalize the system and then redo the torque following a different star pattern along the perimeter of the flange. Nuts do back off under pressure and I would expect this to be a factor with your spool.

Also, rent a torque machine to get uniform compression of the blind on the RFWN RF of the spool.

Good luck with it.

Kenneth J Hueston, PEng
Principal
Sturni-Hueston Engineering Inc
Edmonton, Alberta Canada