MAWP of vessel vs flange rating 4

[SNORGY]

A vessel has been specified with MAWP = 740 psig. This is coincident with a 300# ANSI flange rating of A-105-N below 38 C.

Vessel fabricator states that the vessel (horizontal, 1829 mm OD) cannot be registered with MAWP = 740 psig because, if liquid is in the vessel, the flanges at the bottom might be exposed to a pressure of 743 psig. Fabricator wants to nameplate to 735 psig to align completely with ASME VIII Division 1 (UG-21), or else state on the nameplate that liquids shall never be in the vessel.

It's a separator...so...liquids will get into the vessel.

The Code indeed reads that way.

I have not come across this, ever.

Do we de-rate?

So then, what do I do with 300# ANSI piping in the pipe rack that might have to go from the low elevation to a high elevation? De-rate all that piping to 725 psig?

Tired, don't want to argue, just want to move forward.

 

[georgeverghese]

But what is the upper design temperature for this vessel (and also for the connected 300lb piping)?; 38degC sounds a bit low for most places in summer.

 

[SNORGY]

Thanks georgeverghese. The temperature is OK.

I read the Code definition of MAWP, the legal - from lawyers' perspective - definition of MAWP, the applicable local Regulations in force, and the inescapable conclusion is that the nameplate rating must be lowered; the fabricator is correct.

It's not even a point worth debating, in retrospect, since the design of the rest of the facilities associated with this one vessel is such that I could go as low as 500 psig and things would still work.

The nameplate MAWP will be lowered.

 

[SnTMan]

SNORGY, this is not a totaly uncommon situation. One way it gets addressed is to set the design temperture for the offending flange just enough lower that the static head is acommodated. Drawings, U-1's etc. so state.

But at 38 C, is would not seem possible in your case. Lowering vessel MAWP probably your only way.

Regards,

Mike

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

 

[SNORGY]

Thanks, SnTMan.

I came into the office early this morning on my Saturday and communicated to the fabricator - along with an apology - to lower the nameplate MAWP. This was on my conscience. So, the issue has been resolved, and while I was irritated by it, the rules are what they are; the fabricator has no choice but to follow them.

 

[SnTMan]

All's well that ends

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

 

[jtseng123]

This kind of problem normally stems from Process engineer that they set design pressure equal to flange rating without looking at the actual operating pressure to set a good design pressure that can be way below flange rating.

It will also cause flange problem that when equivalent pressure from piping loads adding to MAWP may way exceed flange rating that potential leaks will occur. Process engineers need to understand how mechanical is designing the system.

 

[marty007]

I see this all the time.

Datasheets come in with design pressures equal to maximum flange ratings, and liquid head has been completely ignored.

There are two solutions, both of which I've used in the past:
[ol 1]
[li]Decrease design pressure.[/li]
[li]Increase flange rating of any flange subjected to liquid pressure.[/li]
[/ol]

Either will work, I just always hope the customer will go for option #1. Option #2 works, but it creates headaches for the customer as they now have a higher class flange that can be overlooked when it comes to piping.

Cheers,
Marty

 

[jte]

This is precisely why I prefer to see two sets of design conditions.

The first set establishes what conditions are necessary (including reasonable margins) for the process to work. These process design conditions are set by the person who would be familiar with these needs, the process engineer.

The mechanical engineer then optimizes these process conditions and develops mechanical design conditions which meet or exceed the process requirements.

I have a simple agreement which I frequently propose: I will not pretend to be a process engineer if the process engineer agrees to not pretend to be a mechanical engineer.

This is not all that different than having two sets of design conditions for things like utility steam: The mechanical engineer's design conditions are conservative from their perspective, and take the max possible pressure and temperature at the source. In contrast, the process engineer must use for their design the lowest reasonable pressure and temperature at the equipment which is consuming the steam. Each is conservative in their own way, and neither should set design conditions for the other.

 

[SNORGY]

It's actually the mechanical engineer who missed on this one.

In any event, my vision is different. I believe the person drawing the P&ID should be the same person as the person filling out the Line List (person #1) from which is spawned the data sheet by person #2. Person #1 should know enough about both process and mechanical to get it right so that Person #2 can't go wrong.

 

[moltenmetal]

Setting design limits is an art. It is very easy to trip across natural boundaries which have cost associated with them, like the temperature limits of particular materials of construction or the P/T limit of a flange class, if you set your limits too arbitrarily or too conservatively.

Another example is defining a flange-limited design but forgetting that there will be stainless flanged instruments in a carbon steel line. Bumping everything up a flange class unnecessarily is a huge headache and can be a significant cost as well.

A chem eng (a "process eng" is just a subspecialist in my opinion) working in equipment/piping systems design needs to know enough mech eng to be able to navigate through problems like this. Too rigid a division of labour results in a total mess.