Gasket Selection and Minimum Required Bolt Preload 3

[LargeTuna]

Hello All,

Although my background is in engineering, I'm definitely "green" when it comes to pressure vessels, so I thought I'd reach out to the experts for a little guidance. Basically, I have been tasked with getting a very large horizontal pressure vessel ready for re-certification. I'm mostly looking at replacing all of the blind flanges, gaskets, nuts, bolts, etc... The vessel is very large (~30k gallons) and is only actually filled and pressurized (with water @ 150 psi) during testing. Most of the flanges already welded to the vessel are 2" 150# SA 105 RF, although a lot of the blinds I have been pulling off are actually FF. So my first question is, shouldn't RF flanges also be mated with other RF flanges? Also, is it recommended to only use ring-type gaskets with RF flanges instead of full face gaskets? The only reason I'm considering full face is that they seem like they will be easier to align and keep in place during assembly. Also, how do I go about determining what type of bolting configuration to use (studs, bolts, lock washers, etc..)? Is the normal process to select the type of gasket that will best suit your system's needs first, then to calculate the minimum preload required to seat that gasket along with containing the internal pressure (Wm1)?

Thanks in advance. Any and all information is appreciated!

 

[racookpe1978]

FIRST! Go read all of the gasket requirements, definitions, and their required minimum AND maximum bolt torques for each size pipe flange (which is what you are using) at the Flexitallic site.

http://flexitallic.com/

review the references, it will explain many items you've not thought about. Yet.

Why have you decided to replace old blind flanges - other than the RF/FF criteria?

 

[LargeTuna]

Thank you for your quick reply, racookpe1978.

I appreciate the link as well. I've seen a lot of similar information on other companies' sites (Lamons, etc...), but I had not run across Flexitallic until now. So I suppose the first step is to actually determine the specific gasket material/manufacturer then either just torque within their suggested limits, or obtain a more precise range by calculating Wm1 then determining the torque(the upper torque limit for A 193 B7 seems to be a standard ~120 ft-lbs to achieve no more than 60ksi bolt stress). I would like (probably need) some of my own calculations/analysis to back up the minimum applied gasket stress and bolt preload/torque for our system parameters. I have yet to see anything in the ASME codes that covers the use of lock washers, which I believe may help retain bolt stress due to our cyclic operating conditions). I think I did see something in the ASME PCC-1 that recommended through hardened washers on both sides of the joint for better torque translation. (I don't have access to ASME BPVC, but I can get some of the other documents through our tech library).

What are some of the items you were referring to that I have yet to consider?

Basically the majority of the blinds that are currently on the tank were all fabricated in house and unmarked. I believe they are all carbon steel, but I think they are also thinner than the minimum required ASME thickness for 2" 150# as well, and they are experiencing surface corrosion. We need all bolts, nuts, flanges properly marked and identified to pass our inspection.

 

[racookpe1978]

Did you not answer part of your own question? 8<)

the majority of the blinds that are currently on the tank were all fabricated in house and unmarked. I believe they are all carbon steel, but I think they are also thinner than the minimum required ASME thickness for 2" 150# as well, and they are experiencing surface corrosion.

Cleanliness (no scratches nor residue on the gasket faces) and makeup of the joints are important. Torque methodology, training (should be a routine, but you have to check, and keep checking, and let the crew know you are going to keep checking) for first torque, intermediate torque and star pattern of bolting, and final torqueing.

 

[TGS4]

which I believe may help retain bolt stress due to our cyclic operating conditions

This caught my attention, as it is some thing that I am often brought in post-facto to troubleshoot. Is your service cyclic in temperature or pressure or both? Cyclic service has different factors for BFJ than "normal" service.

 

[MJCronin]

Lock washers are not used on metallic pressure vessels.

Follow the tightening requirements of PCC-1 and you will have no problems.

As stated above, be careful of gasket faces

MJCronin
Sr. Process Engineer
Venture Engineering & Construction

http://www.VentureEngr.com

 

[LargeTuna]

racookpe1978,

I was just responding to your question about why I'm replacing all the flanges. The bottom line is that all our current blinds (~16) need to be replaced, it's just a matter of making sure I am getting the right ones. I understand the torque methodology and sequence, I'm more concerned about defining an accurate value of minimum torque to safely seal the flanges.

TGS4,

Well, the PV is actually cyclic in both temperature and pressure. Temperature control ranges from room (~70°F) to about (100°F), and pressure ranges from 0 psig to 150 psig. The tank (PV) is used to simulate hydrostatic pressure up to approximately 300 ft below sea level.

MJCronin,

Thank you for the info, and I'll be sure to follow the PCC-1 procedures when installing, but I would still like to calculate a ballpark torque range for a little more accuracy. That's really my main objective right now. I understand the recommended torque ranges will more than likely suffice, but I would still like the peace of mind that goes along with actually approximating a unique value based on all of our specific variables.

 

[gasketguru]

You may want to look at the sealing industry notes from either the Fluid Sealing Association or the European Sealing Association depending on whether you are in the USA or Europe (fluidsealing.com or europeansealing.com) - the major manufacturers such as mentioned above are represented here and these associations offer literature and guidance notes on gasket fitting and usage.

If you have flange faces in a poor condition then a soft material such as expanded graphite or even an expanded PTFE might be needed to accommodate some imperfections if compressed fibre is too firm. The ASME PCC-1 document tends to aim high in terms of bolting and is more geared towards spiral-wound gaskets. - You could try an EN1591 calculation for greater load requirement accuracy (the traditional ASME calculation often under-calculates especially on gas applications). Note of course that torque tightening is not that accurate as friction can vary considerably - you are measuring turning power not bolt load. However you are only talking about class 150 here, so pressures are not that high anyway - note that 3" and 8" are perhaps the worst flanges for bolt load to gasket area ratio.

Most ESA and FSA member companies have good technical support and can probably assist you locally with material selection

 

[Duwe6]

Where possible, use spiral-wound gaskets [any brand]. Then use the torque values from the Flexitallic catalog. Follow ASME PCC-1. "So simple, it's almost hard"

 

[ricklts]

You might also want to be aware that most 150# flanges lack sufficient strength to properly maintain the gasket/bolt loads required for normal spiral wound gaskets. The manufacturers have all developed a 'low stress' version, with a much lower winding density to accommodate this reduced load capability. If you had a critical joint, you might look into one of the KAMMPROFILE designs.

Be aware the over tightening and plastic flange rotation is very hard to avoid on most 150# flanges. A FF gasket surface, with the proper sheet gasket, or one of the corrugated designs might be a solution to that potential problem.

 

[XL83NL]

The manufacturers have all developed a 'low stress' version, with a much lower winding density to accommodate this reduced load capability

Here's what Warren Brown replied to me when I asked him a question on the 'advantage' of low stress SPWND's.

Basically the low stress gaskets offer no advantage over standard gaskets. None of the leakage tests performed on them that I have seen show that they seal any better at low bolt levels than a standard gasket (so the basic premise of them is flawed) and they are at a disadvantage at higher bolt stress levels in that you may compress them down to the rings. In some cases (or for some designs) you may also be able to buckle the windings, which means that you will get a lot more relaxation by comparison to a standard gasket.

You should be able to ask the manufacturers for their EN13555 or ROTT leakage test data for both the standard and low stress – plot the points on the same graph and see if there is any advantage (let me know if there is!).

I am not sure that everyone knows the issues with them, PVP can be interesting like that, a lot of people sit there nodding, but if you had of asked them before they heard it said, they probably wouldn’t have known…

The trend now is actually towards higher density spiral wound gaskets and higher bolt loads, so low stress are definitely losing market share.

 

[ricklts]

The only advantage a low stress gasket has is that the 'm and y' values do not over stress the flange upon make up at higher bolt loads. If you want a leak free joint at low bolt loads, use a metal-to-metal seal like one of the clamp connectors do. Any plasticly deformed flange face-oriented seal, either SW, RJ, Kammprofile, or Jacketed, will leak due to cyclic pressure/thermal/external loading which approach the upper design limits. Most of these leakages are due to some form of relaxation(I want to say creep, but that's not the proper term) in either the gasket or the bolting. When dealing with flanges with limited bolt stress capacity, you walk a fairly fine line between not enough and too much stress in the joint. The low stress gaskets are an attempt to mitigate some of those problems. Changing to higher density gaskets and higher bolt stresses will only work if you replace the standard ANSI 150# or ASME/API low pressure class flanges with those of higher bolt load/gasket load capacities. And that will cost more money, another driver for the low stress gasket design.

 

[chicopee]

I hope that your 30K gallon PV is cribbed at its center in addition to its end supports when you fill it with water.

 

[LargeTuna]

Hey guys, thanks for all the input.

I'm more than likely going to go with Garlock 5500s for my flanges. I contacted their engineering department just to verify that ring gaskets and full face gaskets can be used interchangably with RF flanges, and they confirmed. I kind of prefer full face gaskets with the bolt holes for alignment purposes since most of my flanges are all oriented vertically, and they will also help keep foreign matter from getting inbetween the flanges...just in case anybody was still wondering, and since I never really received a definitive answer on this thread. Garlock also has quite a bit of information and resources on their site pertaining to gaskets, and they are very easy to contact if anybody here has any future questions related to gasket determination.

 

[Duwe6]

Full-face tends to be a Pain-in-the-A$$ to install, 'Real World', unless the flange is small enough to easily manhandle, or has a lifting-lug welded to it. Most mechanics/boilermakers/pipefitters prefer ring-shaped gaskets vs. full-face.

And I have to take exception to rickets "... most 150# flanges lack sufficient strength to properly maintain the gasket/bolt loads required for normal spiral wound gaskets." Spiral-wound "Flex" gaskets are industry standard in refinerys and petrochem plants for Class 150 flanges, at least in the USA. I havn't run any flange calc's*, I've just used, and seen used, Flex's for 32 years. They work, and work well. To include drip-tight hydrostatic tests run at "Full Flange Hydro" pressure; i.e. 1.5 x 70°F MAWP of the flange. For A-105 c/s Class 150# flanges, Full Flange hydro is 420 psi. For 304/316L s/s, it is 345 psi. Flex gaskets and B-7 studs, using torque values from the Flexitallic literature.

*Aerodynamicists in the 1950's (?) ran calc's on bumblebees. Proved that they were unable to fly. Took a lot of improvements in theory and computing 'horsepower' [IIRC - 1995] to be able to get calc's on bees that agreed that they actually could fly.

 

[ricklts]

good on you.

DUWE6, your exception is noted, but I worked for Flexitallic for ten years as part of their field service group, and I can tell you that most of the larger (10" and above) 150# RF flanges in nearly every major petrochem facility in the US is warped due to overloading of the bolts. This is the reason Flexitallic began to make their low stress version. At one Houston refinery, the plant had to replace nearly seventy 10" and larger 150# flanged joints when the deformation created a 1/8" stand-off of the gasket face before the bolts were installed.

 

[Duwe6]

SIDEBAR: Willing to bet a paycheck that those warped large 150# flanges at those refineries were tightened to "good'n damn tite" greatly in excess of the recommended torque value. I've personally seen it, especially from muscular young 'shutdown hands' in a rush -- faster to wrench it down HARD than to go get a torque wrench and use the spec value.

 

[SnTMan]

Once you are clamped firmly against the centering ring(s) (if so equipped), further bolt load does nothing to effect a seal. I'll betcha

Regards,

Mike

 

[TGS4]

Actually, SnTMan, I've written a number of papers about that topic - and the only way that the calculations match with experience is to also assume that the flange seals on the metal centering ring. I would be willing to bet that a large portion of Class 150 flanges seal only on the centering ring. and not the spiral windings (due to the flanges cupping and unloading the windings).

 

[EngAddict]

Surely the seal would be held by the outer winding portion, not the centring ring. Do you have a a reference link for the papers TGS4?

Under high external loading and cyclic operation I would think additional bolt loading would be beneficial to maintain the seal.