Are tubesheet gasket grooves necessary? If so, when and where and how?

[Keith84]

Hello everyone,

The quick question:

Is it common or even standard practice to machine flange-face-style gasket grooves into a tubesheet, so that the gasket seats and seals similarly on its tubesheet side and its flange side (since the flange usually has gasket grooves)? If so, how's it usually done?

The longer question and background:

I have a question about the necessity, character, and machining of gasket grooves in tubesheets. I'm a chemical engineer, but I've designed and had machined dozens of tubesheets over the years for shell-and-tubes. Most of the fluids in the exchangers have either been natural gas, crude, refrigerants, water, glycol/water, and other fairly benign stuff at reasonably low pressures (maybe up to 500 psi for exchangers we've directly built). I've done BEU-type, sandwiched tubesheets and BEM-type, fixed tubesheets. The thing is that I've always just specified flat surfaces on the tubesheets where the gaskets sit and seal (with a flange-matching raised face on the fixed tubesheets, but still smooth). We've used simple custom-cut, sheet-type gaskets on both sides of the tubesheet, and have never failed a Section VIII, Div 1 pressure test or even had any flange-to-tubesheet leaks doing it this way. HOWEVER...

We're starting to get jobs that have a lot of hydrogen (~75 mol% in some places) in the process lines and some of them are at high pressures, e.g. 2200 psi, and high temperatures, e.g. 900-1200 F. This has brought up a lingering concern I've had about our not machining gasket grooves into the tubesheet, since it seems like the additional grip would help contain these more extreme process fluids better.

a) Is this commonly done or always done or not all that commonly done? If it's a grey area, are there guidelines or rules of thumb or standards out there that specify where gasket grooving the tubesheets becomes necessary?

b) More of a machining question if tubesheet gasket grooves are necessary: What's the common way of machining them? B16.5 and B16.47 paragraph 6.4.5.3 (Other Flange Facings) describes the flange standard, which I think I'd want to emulate on the tubesheet. However, it merely says "serrated...finish having a resultant surface finish from 3.2 um to 6.3 um average roughness...[with a] cutting tool... [of] 1.5 mm or larger radius...45-55 grooves/in." We have a three-axis vertical mill we intend to do these on. Is there tooling designed to cut multiple grooves simultaneously, or does one buy a very small tool and cut one groove at a time? If the latter, what would that tool be called, so I can navigate the many gigantic machine tool catalogs I have? I ask so specifically about the machining, because we've talked to a lot of machinists in the past who don't have a clue what I'm talking about or how to make it happen when I try to describe this gasket face I think we might need on the tubesheets.

I hope I've laid this out clearly enough, but definitely let me know if I need to provide more information. I've had this question for years, but now I need to get an answer to it, or certain others in the company are going to give the go-ahead to start machining a bunch of tubesheets using the old design, and that worries me with the current process conditions I described above. If nothing else, I really need thoughts, references, standards, etc. on whether the tubesheet gasket grooves are necessary, and the machining aspect of them may require a separate thread in the machining forum. Thank you all for any help you can give me.

 

[Duwe6]

Tooling is available to 'comb' in the gasket surface 'phonograph' finish. It is usually specified as "125 to 250 µ-inches". And yes, a smooth finish has exactly ZERO chance of being even marginally acceptable at 2200 psi.

Personally, I would be considering ring-joint flanges and RTJ gaskets for that pressure of H2.

 

[SnTMan]

Keith84, your terminology "flange-face-style gasket grooves" is not clear to me. Several configurations are possible.

1) Unconfined gasket: No machined grooves, steps, turns or whatever you wish to call them need be present on either the flange face or tubesheet face. If single pass tubeside no pass partition or pass rib is present. There is no mechanical means to prevent extrusion of the ring portion of the gasket. If multi-pass tubeside, no mechanical means is present to prevent extrusion of the pass rib from the pass partition. Note that a normal hydrotest will not detect this condition. Sheet gaskets are the worst about this.

2) Tongue and groove: Normally a raised tongue is present on the flange face, with a mating groove on the tubesheet. The ring portion of the gasket is confined on both OD and ID. If multi-pass tubeside, the pass partition is flush with the tongue on the flange face and a mating groove is machined in the tubesheet. The gasket pass rib is captive.

3) Confined gasket: Normally a raised face on the flange confines the gasket on the OD. A raised face on the tubesheet may or may not truly confine the gasket on the ID, however extrusion to the ID is not likely. If multi-pass tubeside, a groove is machined in the tubesheet raised face, confining the pass rib.

As to finishes, it is common for the ring portion gasket surfaces to be lathe-turned, providing a spiral or concentric finish easily using either single-point or multi-point tools. The pass partition may be turned in the same operation, however tool marks will cross the partition. Any groove will likely be milled in a separate operation. Again, tool marks will normally run across the groove. A spiral finish is rarely if ever provided on the partition or groove in my experience. I am not aware of tooling that would provide a spiral or concentric finish of either the ring or straight portions on a mill. To be clear, on a mill the cutting tool rotates, on a lathe the workpiece rotates.

Guidelines for whether a gasketed joint should be confined or not are found in TEMA, and varies based on TEMA class.

I hinted above, external leaks and leaks across a pass partiton are different animals, the latter can only practically be detected by its effect on performance. Also none of these configurations assures adequate sealing on its own.

See TEMA F-3, R-6.5, CB-6.5.

Hydrogen is tough, but several types of metallic gaskets can be used sucessfully. RTJ might be best, but incorporating a pass rib is a little difficult, although with careful design and manufacturing it could be done. So-called Kammprofile gaskets with suitable facings are considered to be among the lowest leak rates available. Suggest you contact a major manufacturer such as Flexitallic about gasket selection and joint design. They are there to help.

Long post, hope you find at least some of this useful

Regards,

Mike

 

[Keith84]

I'm sorry it's taken a couple of days to respond. I appreciate both of you guys for answering. And, Mike, I'm always ecstatic when someone's generous enough to give me a long, helpful answer like you've done.

I'll guess I'll give a little more background on how we've done our heat exchangers up to this point. We've always used standard B16.5/47 raised-face flanges for the body flanges. This has been true for the BEM or BEU types we've built. On the BEMs, the gasket has either been sheet-type or spiral-wound and kept centered by the bolts. Definitely an unconfined type according to TEMA CB-6.5. For the BEUs, we've always had to use sheet gaskets to get the partition side gasket custom cut. We weld the partition plate into the bonnet right up to the flange face and grind the exposed edge it as smooth and level as possible to seal against the gasket. For the U-tube tubesheets, I machine 1/8" deep grooves on each side to hold the 1/8"-thick shell- and bonnet-side gaskets in place while we insert the bundle and sandwich it between the shell- and bonnet-side body flanges. Every surface on all tubesheets is machined smooth, though; there are no 125-250 u-inch finishes as Duwe6 described on the tubesheets as we've currently built them.

For this project, I figured we'd at the very least have to go with Kammprofile gaskets on the partition side of any BEUs we build. I've definitely seen RTJ flanges used in high pressure hydrogen applications before, but, Mike, as you point out this would be difficult for the partition side of a BEU. I've also always worried about that edge of the partition plate in the bonnet that seals against the gasket separating the two sides; it always seemed to me that it might need a machined finish equivalent to the flange face it was welded to but with the grooves running parallel to the partition. I'm not sure how I'd machine that at the moment. though.

I knew this was going to be a whole different ballgame from our previous exchangers, so is there anything out there, in addition to just valuable experience, that makes recommendations for a proper heat-exchanger design for hydrogen, especially high-pressure hydrogen? I've got other questions, too, such as whether to weld the tubes or roll them, and where the line is drawn there. I figured I'd stick to tubesheet questions with this thread to not make it overly broad. For the tubesheets, I think I'll definitely put a flange facing where the gasket seals against it, though now I need to also check out the RTJ option. The customer did not specify a specific TEMA class, and now I'm wondering which I should design to. I notice that R-6.5 REQUIRES confined gaskets whereas CB doesn't.

Since we have touched the topic, another question I've had in the past is whether it's acceptable to use the same RTJ gasket for hydrotest and for final assembly in a pressure vessel. They're certainly not cheap, but I don't want to cut corners when I'm dealing with hydrogen. Really, I could use some documents or guidance just discussing when to actually switch over to RTJ flanges and gaskets. From what I've seen, the ASME Code doesn't seem to require them at any time, but I've heard anecdotally that many people just go RTJ when the flange rating gets up to Class 900 or 1500, depending on who you talk to. That may be yet another question to ask in a different thread, but if it's already been asked, and someone could point me to the proper answer, I'd appreciate it.

Thanks again, guys.

 

[SnTMan]

Keith84, TEMA considers a spiral wound with rings as confined, regardless of flange facing. These gaskets are often used in multi-pass tubesides by the addition of metal jacketed pass ribs welded to the inner ring. Any half-way decent gasket shop can do this.

Assembly should always be a concern of the designer. Often better to make custom flanges that located the gasket during.

I would not be too concerned about finish of the pass plates. Obviously, a machined surface is preferable to a ground one. However, the pressure differential across the plate, and therfore the gasket pass rib is usually farily small as compared to design pressure. All you're trying to do is prevent bypass between the tubeside passes.

Kammprofiles should be useful in your service, however I would still suggest you discuss your proposed designs, services, etc with an applications engineer at a major gasket manufacturer. It's free

Regards,

Mike