High pressure design considerations 11

[GD_P]

Hello,

Please help me to get rid of below problem & Sorry for the long post.
We design vessels & piping for pressure max 10 bar & temp 350 deg C. But our next project demands for 100 bar design pressure. So I need your advice (based on industry practice) regarding flanged connection for piping & vessel. Till now we never used gaskets other than non-metallic type & flanges other than slip on type raised face. So we have many questions such as:
Can we use threaded coupling (B16.11) (sealing using thread) or adapter where taper metal to metal surface is used for sealing?
Which type of flange should be used, since B16.5 code doesn't restrict any specific flange?
Can we use non standard flange (non 16.5) with PTFE O ring (off course within its temp limits) instead of gasket to avoid blow out?
Which gasket should be used non-metallic (PTFE, Graphoil etc ), semimetallic, metallic or ring join type?

I understand there can not be one line simple solution for this, but I will have to go through many articles, literatures to arrive at a conclusion. So could you share any guidelines where I will get the answers for this. Thank you for being patient.
Regards,

GD_P

 

[Jay_]

You are missing one important detail.
What is this line is used for?
And where is that line located?
There is specific codes for such designs that includes the answers of all your concerns and more.
Such as ASME B31.4 B31.3 etc.
they state which flanges, if couplings are acceptable and specify the material and everything you would need to know, but you would need some software to verify the elasticity of that pipe if it’s required.


Detailing is a hobby,

 

[SnTMan]

GD_P, further detail would be helpful, such as service, materials and methods of construction and size, esp. diameter of vessel. Exact design practices might vary, but in general:

At your stated pressure & temp I'd avoid threaded fittings in anything over about 1 NPS. Standard B16.5 weld-neck flanges can be useful both for nozzle and girth flanges, assuming given the sizes and materials that a suitable rating is available. Custom design flanges could be used as well. I'd recommend a metallic gasket with perhaps soft facing such as grooved metal, graphite covered (GMGC) type. A spiral wound gasket with retaining rings, or RTJ type likely would work as well. O-ring gaskets if manufacturing capabilities are up to it.

Design choices will be influenced by feasibility / desirability of using standard piping components over custom designed components.

Regards,

Mike

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

 

[GD_P]

Thank you Jay & Mike for your replies.
The service will be lethal or category-M fluid service. Actually few connections are 1/4" & 1/8" tubes in size, so threaded connections will be feasible here. Also I went through ASME B31.3 but it doesn't restrict use of threaded connections up to 1" size for category M fluid service. But as the temperature (350 deg C) & pressure (100 bar) are high, we are are not sure whether they will achieve leak tightness or not. Do share your opinion.
Regards,

GD_P

 

[MJCronin]

So, you are designing and building a pressure vessel for "service will be high pressure lethal or category-M fluid service?

....And

1) You don't seem to have a specific design code that you are following (Does B31.3 apply or not ?)

2) I cannot tell if this is a pipe or vessel

3) You don't know if your component will be leak tight since you are trying flanges you are unfamiliar with

4) We do not know anything about vessel size, materials, NDE, pressure testing etc..

5) You have never done this before and the component has TEN TIMES the pressure of anything previously fabricated

6) You seem to be putting a lot of faith on the advice of internet strangers ....

Is that all correct ???



MJCronin
Sr. Process Engineer

 

[XL83NL]

Michael, if I could give more than 1 star, you’d deserve it.

GD_P; from the snippets you posted, your work seems similar to what my company does. But it doesn’t seem appropriate to give you sound advice in this, rather than promoting to leave this over to someone experienced with this matter.

Huub
- You never get what you expect, you only get what you inspect.

 

[GD_P]

I apologise for the inadequate information.
Code of constructions - ASME Sec VIII Div 1 and ASME B31.3 (for the piping associated to vessel)
Material - SS316
It is small vessel with size DN 40. It's a trial equipment which will handle multiple fluids depending on the customer, which may include Lethal / category M fluid service.
We will not directly implement the connection into the vessel. We will demonstrate its suitability for operating conditions with some other trial fluid on a prototype. Hence I am currently looking for information for the prototype.
Regards,

GD_P

 

[Jay_]

Asme b31.3 is a process piping code.
So the vessel should actually be located in a plant or refinery of some sort.
Lethal fluid has very specific requirements in the ASME SEC 8 code, and also in B31.3 especially regarding testing and fabrication as well as for choosing the right fittings and connection.
A vessel to contain lethal substances is always more expensive.
And why SS316, may i ask?

Detailing is a hobby,

 

[r6155]

@ GD_P
Sorry, but this job is not for you. You must do special training.
Lethal service requires extreme responsibility and experience.

Regards

 

[r6155]

For lethal service ( 100 bar and 350C ) all joints with piping should be by welding, not flanges.
Inspection opening should be flanged with Weld-Ring Gasket (see Flexitallic).

Regards

 

[XL83NL]

@r6155; Lethal service isn't defined for piping. For B31.3 it's Category M Fluid Service (short Cat M). For Cat M, flanged joint are permitted; see M308.2.

Huub
- You never get what you expect, you only get what you inspect.

 

[r6155]

@ XL83NL
Excuse me, I was not clear. I talk about the vessel and the connections to the piping.
Sorry again.

Regards

 

[SnTMan]

316 material. Lethal service. Threaded connection. Bad combination, you ask me.

Regards,

Mike

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

 

[moltenmetal]

GD_P- you list yourself as a structural engineer. As a chemical engineer, if I was asked to design a structure which, if it failed, would result in people dying, and the conditions involved high seismic and wind loadings etc., I would refuse that work and refer them to a competent structural engineer.

Yes, your "vessel" is small- but if it might contain hazardous substances at very high pressure and high temperature, you are clearly out of your depth to even begin looking at what is permitted or required in designing it. Category M (lethal service) implies that even a leakage event might lead to serious injury or death.

You need to hand this task off to someone competent to do the work. That you lack the judgment to realize that yourself, without a bunch of strangers on the Internet telling you so, is not inspiring confidence. In fact, it positively terrifies me for what your "customer" may encounter if you persist.

Please back away from this and find something else to do.

 

[LittleInch]

GD_P,

I think you've confused the daylights out of everyone here so maybe best to go back and define exactly what your design is and what your questions are.

100 bar is a lot different to 10 bar.

100 bar at 350C for any ASME B16.5 material is almost certainly class 900.

For this service any flanges are probably best RTJ type to prevent blow out.
Use of welded connections is really required unless impossible to do. Threaded and slip on are for low pressure (class 150 or class 300 at most) connections.

If your only connections to this rather small "vessel" is 14" and 1/8" tubing, then contact vendors such as swagelock for their capabilities. There are thousands of such connections worldwide.

However any connection never likes being made and broken then remade.

As a few others say above, this needs to be properly reviewed by competent experienced engineers before manufacture.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[r6155]

@ LittleInch
"For this service any flanges are probably best RTJ type to prevent blow out".

For this service (lethal) the best gasket is to prevent LEAKS.

Regards

 

[GD_P]

Thank you all for your cautionary posts which are based on your understanding or practical experience about the criticality of application or possible implications of uncertainty. But Honestly speaking, I really felt demoralized after reading the posts. I accept that I may not be competent to perform the design for CRITICAL conditions mentioned, but does it mean I should never do or learn it. I never said that I wanted a ready solution but I was expecting some guidelines / specifications / literatures reference, if any. Certainly this will be checked by the competent personnel, but applying the conventional design used for 10 bar and sending it for approval makes no sense to me. How will this add value to my knowledge?
Regarding my qualification, I am basically a mechanical engineer but additionally I have also done special training in structural engineering (adequate to suit the job profile). So in the beginning I approached eng tips for enhancing my structural knowledge and hence listed myself as a structural. I tried to edit it, but I was unable to do so.
I was specifically talking about any GUIDELINES for 1) vessel nozzle flange types & pipe flange type & gasket for above mentioned design conditions (Weld neck will be best solution for this & for gasket we will discuss this with supplier)
2) for small nozzles (NPS ⅛" & ¼"), way to connect nozzle to vessel i.e., threaded connections (we will discuss this with special suppliers like swagelock)
Regarding the combination of Lethal & SS316, this has been decided by the process engineer. So can not comment on this.
At last I will take your comments with a positive intent. Thank you once again.....


GD_P

 

[SnTMan]

GD_P, I made my last post with the problem in mind of galling in austenitic stainless steels used in threaded connections, including ordinary nuts and bolts. This can happen even before leak tightness is achieved. You are then truly screwed stymied. Lethal service, doubly so.

Not saying you can't use threaded connections, but this potential serious problem must be considered.

Best of luck.

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

 

[georgeverghese]

The first port of call for piping design is the Plant Owners piping class data for the service intended, and other related general piping design guidelines. If you dont have this from the Plant Owner, ask the design contractor. If they dont have this either, its best to keep well away from this task - there are just too many pitfalls whatever you select, almost.

 

[BJI]

100 bar is not considered high pressure and the vessel and nozzle sizes are quite small, so concerns with gasket blow out and connection type are not really warranted, only leak tightness and maintenance should really be considered. This seems like a pretty common design for small laboratory type vessels. On actual high pressure vessels in the 700-1000 bar range, just about everything is threaded. For stainless steel swageloc fittings galling is definitely a concern, as SnTMan pointed out. You could also check out the Parker catalogues for threaded fittings easily capable of this relatively low pressure rating.

If you do go down the flanged approach, then I would go for a gasket with soft facing, requiring a lower gasket stress to maintain a seal. RTJ gaskets can be more problematic to seal and should not be required in this diameter and temperature range. They are also less suitable for joints that need to be broken on a regular basis. RTJs can be successfully swapped out for kammprofile gaskets in larger diameter class 900 flanges so I would steer away from hard faced gaskets if the service conditions permit.

I definitely agree the design should be reviewed and stamped by someone experienced if you don't have it in-house. There is constructive criticism above that should be taken on-board but if you think anything overstepped the mark then you can always report a post.