Minimum Design Metal Temperature (MDMT)

[liberoSimulation]

Does any one have an idea about the basis of MDMT of an equipment?
Is the basis depressing the equipment to atmospheric pressure assuming control valve failure or opening PSV or draining the equipment with cold liquid rapidly or de-pressurzing the equipment with cold material before drying the equipment?
Many cases that lead to design the equipment MDMT to be based on atmospheric pressure.
I know that by this basis low temperature material shall be used for such equipment that fall in one of these cases.

Regards

 

[IFRs]

Most metals get increasingly brittle as their temperature lowers. Is this pertinent in your situation?

 

[StoneCold]

liberosimulation
The MDMT for carbon steel equipement is -20F. That is the cutoff. If you need an MDMT that is lower than that then you have to switch metallurgy. In piping A333 can be used in low temperature service. I believe the carbon steel used for pressure vessels at low temperatures containes 9% chrome and is referred to as "Nine chrome". I believe in most cases it is cheaper to switch to stainless and take the U66 exemption in the ASME code than to use the 9% chrome but you will have to check that out for your self.
As you stated rapid depressurization can set the requirement for a low MDMT but there are also other conditions like chilling a heat exchanger or reactor that can cause the need for a low MDMT.

Does that help?

StoneCold

 

[liberoSimulation]

Thank you both of you for the reply

Let me explain the situation further: i need to check the material selection of a column operating separating ethylene/ethane with top and bottom temperatures of -26 and -3 C respectively and average pressure of 22 bara.

Is low temp. CS (LTCS) with MDMT of -49 C suitable as a material for the column?

But in case the pressure gets down below 12 bara or the worst case atmospheric pressure for any reason, LTCS can not stand as MDMT would be about -100 C. Is this valid argument for choosing SS material? Is this called for in PIP?

I hope this explains the subject well.

Regards

 

[SJones]

thread330-82042

ASME VIII Div I UCS 66 and Figure UCS 66.1. It's all about the stress experienced at the low temperature.

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[Longinthetooth]

One other thing to be careful with is, if you have a plant crash and the liquids in the upper section dump into the bottom. The BP in the bottom will plunge and if the tower is depressured the temperaure could go very low there. Having said that our C2/C2= splitter is low temperature CS. Good for -46C.

 

[liberoSimulation]

geordie87
The example of c2/c2 splitter you gave is similar situation I am in to evaluate currently.
The column top temperature is -26 C and bottom at -3 C , with operating pressure at about 22 barg as explained before.
The question: is LTCS the suitable material selection for the column? What if the column is depressuized to 1 bara as MDMT would reach about - 100C (I guess)and MDMT for LTCS is -49 C?

Regards to all

 

[StoneCold]

liberosimulation
While the flashing liquid equilibrium temperature may be -100C the thermal mass of the system will probably keep the temperature from going that low. Maybe some one else with more C2 splitter experience can chime in on this.

StoneCold

 

[Longinthetooth]

I'm dredging this from memory (dont have my books here). I think one of our engineering providers guideline was to make the minimum design temperature equal to the BP at half the operating pressure (see Steve jones comment above). I think the reason was the reduced stress the tower would see at half pressure. If you didnt do something like this, every light ends tower would be SS throughout.

Regards,

 

[sshep]

Typical construction is lower temp carbon steel (i.e. SA-516). Obviously the temp of 14.7 psia boiling ethylene, ethane, etc is well below the temp limit of these steels. The golden rule of dealing with this type material in this type carbon steel (min temp -50F) pipe and vessels is to always pressure up first with warm vapors of similar composition before introducing any LPG.

Also remember that it is partial pressure of hydrocarbon, not total pressure that determines flashing. I know a few engineers who have found out the hard way that pressuring up with N2 before introducing a liquified hydrocarbon gains you nothing in terms of supressing flashing.

best wishes, sshep

 

[liberoSimulation]

I have found the following definition of MDMT or LPFT :lowest possible fluid temperature"
LPFT is the flashing adiabatic temperature of the fluid a atmospheric pressure.
By this definition and the calculated LPFT, the material of equipment shell and associated parts can be selected based on the LPFT value even though it is lower than -100C.

Any objection in this basis!!

Regards

 

[Longinthetooth]

There are several papers from the Ethylene Producers conference that deal with this. Amongst them are:
"Assessment of cold service pressure vessels for risk of brittle fracture"- J.E. Mclaughlin et al - 1995
"A program for analysing the potential for low temperature brittle failure in eqipment"- J. Dugas and D. Damin- 1996
""Auto refrigeration/brittle fracture of existing olefins plants - identification of potential excusions" - Ralph E. king- 2004.
There are more but the last one is really good.

 

[liberoSimulation]

geordie87
Can you help in getting these papers as I am interest to know if something related to my issue?

Regards

 

[Longinthetooth]

Try this link

http://www.ethyleneweb.com/

If you cannot get direct access, there is a contact at the top

 

[liberoSimulation]

Just to give an update for the tough question I have posted here.
The dynamic simulation has been done with the case of sudden depresszing of each process system to the flare through the pressure controller valves provided for the process.
The results were similar to what I have had a concern in which MDMT can reach much lower than the selected MDMT of the selected materials of the service.
Therefore, the calculations were done versus time and it has been found that a time of 15 to 25 minutes is the time taken for the temperature to go much below the MDMT of the selected materials.
Now, is it acceptable to rely on the operator to close the PC valve before MDMT can go much below within the time given of 15 to 20 minutes?

Cheers

 

[SJones]

Look at the stresses the material will experience when cold and see if they exceed the ASME limits. The triangle you are dealing with is: stress, defect size, fracture toughness. Stress will be the only parameter that you have a handle on if you are simply selecting materials according to their ASME low temperature limits without performing fracture toughness testing.

Steve Jones
Materials & Corrosion Engineer

http://www.pdo.co.om/pdo/

 

[unclesyd]

[blue]"Now, is it acceptable to rely on the operator to close the PC valve before MDMT can go much below within the time given of 15 to 20 minutes?"[/blue]

Wasn't one of the initial premisses that this valve could fail to operate properly thus causing the problem?

I don't think this would fly under most jurisdictions. I would never want to rely on such a scenerio as I can think of many reasons for this not to work as planned.

 

[liberoSimulation]

unclesyd
What you meant in your reply?
Can you explain a bit your view as to understand your point clearly?

Cheers

 

[unclesyd]

A couple of points:

If this PC valve fails and causes the problem it is taken out of the picture in controlling the release, unless there is a manual override on the valve actuator.

The loss of instrument air may cause to the PC valve to become inoperable without manual override on the valve actuator.

Depending on your control type, the loss of power in the control room could result in the PC valve being inoperable remotely.

We handle similar situations where we use isolation valves in series with the PC valves and in the case of PSV failures we will have dual PSV's on a manually operated diverter valve.

We try never to depend only on the operator's action to control a process excursion. It is our goal to design the control scheme to mitigate the problem prior to operator's intervention.

 

[liberoSimulation]

Thank you for the good exploitation
The situation now I'm evaluating is in which there is PC valve with isolation valves for individual system.
If that PC fails open or it was open fully by any error, the estimated time to reach MDMT for the equipment materials fall in that system will take about 20 minutes and by then if no action taken by operator in closing the valve or its isolation valves, the temperature will go much lower than the MDMT of the selected materials.

As I asked before, is it practical in selecting adequate material with the lowest possible temperature material to satisfy MDMT regardless of the actions to be taken by the operator even if the time is about 20 minutes or the estimated time is enough for the operator to take an action?

Any suggestion or advice is highly appreciated.