Minimum design metal temperature 2

[ChemE2912]

We are reviewing a vessel design where during depressurizing the fluid temperature can reach a temperature of -46 °C @ atm. pressure. The vessel material is not low temperature carbon steel. In some specification we found following "When depressurizing results in temperatures lower than the minimum design metal temperature (MDMT) for full pressure, repressurization shall not be done until the system warms up to the minimum design temperature for full pressure". From the previous quotation we can infer that a CS vessel can reach temperatures below -29 °C (temp. limit) @ atm pressure without fragility failure. Is such quotation true? Does somebody know which frabication code (i. e. ASME) support such statement?

 

[Metalguy]

CS will always become brittle at some low temp.--depends on the comp. and processing history. But even with CS at a low temp. and "brittle as glass", it may not crack-as you have seen. But it WILL crack if impacted by something of sufficient force. Remember the Titanic.

 

[Carburize]

I think BS 5500 covered this area.

 

[SMF1964]

"Engineering Materials and Their Applications" by Flinn/Trajan, p 532, shows what can happen if carbon steel is loaded to below its transition temperature. That temperature range can be low or high - the only way to know for sure is actual material tests, which were probably done to generate your MDMT. The requirement your documentation provided is there because below the MDMT the material may be 'brittle as glass'. Any defects present in the material while it is then pressurized can be at critical flaw size and catastrophic failure will occur, with or without impact loading.

 

[VeryPicky]

The, so called, "not low temperature carbon steel" I understand to be the material like SA 516 Gr. 70, which as per ASME VIII Div 1 falls under curve B of UCS-66.

This type of material may be used at MDMT of -20F (-29C) without additional impact testing. The thickness, however is limited to 0.394" (10mm).
In general, the material operating below the code designated curve is not certain to fail. There is always a safety factor built into the engineering. The values and conditions given by the code are influenced by a number of conditions and variables. Special consideration shall be given to the vessels that were already in service to ensure the process medium does not have adverse effect on the material impact properties (eg hydrogen service).

Pressurising and depressurising at the temperature lower than MDMT at can be dangerous for the integrity of the vessel since the both these processes are unstable by nature.

To upgrade the "non-low temperature CS" material it is required to recertify the material to the "low temperature material" as per code by typically looking at the steel making process, chemical and mechanical properties, impact testing, heat treatment etc.
Vessel should not be pressurised below its MDMT, and unless there is a proper start-up/pressurising procedure in place and followed.

 

[ChemE2912]

Hi All,

Some vessels are SA 516 Gr 70 and impact tests were performed.
But the flanges are SA-105 and necks are SA-106 Gr. B, as far as we know
the material used for flanges and flanges necks are not suitable for low
temperature services. We think now that the flanges/necks could be a weak
point if the vessel temperature decreases below -29 °C (-20 F). Are we
right?

For these vessels, is it acceptable to depressurize and reach temperatures below -29 °C, as long as we ensure that repressurization does not take place until the minimum metal design temperature (or higher) is reached??

Thanks!

 

[VeryPicky]

The decision is yours. Pressurising is more critical than depressurising.
I would be very cautios, however, if I were to depressurise the vessel at temperatures below MDMT for the reasons described in my previous reply. The re-pressurising procedure as you described sounds good. Obviously you need to specify the details in your procedure. You could start filling the vessel with the fluid provided the pressure does not rise. Remember - the Charpy impact values can be very low.

Good luck and I would like to hear about what procedure you decided to follow and how did it go.

 

[ChemE2912]

Thank you! Jaceksan et. al.

Is there any Code and/or Standard that could support this idea of not repressurizing the vessel until we are above the minimum design temperature of the material?

We have heard this from several people, but would like to know if we could have something more "formal", prior to preparing a procedure.

Best regards.

 

[march1971]

I think visiting this site can be beneficial to you:

http://www.key-to-steel.com/Article/Art61.htm

 

[JoeTank]

The concept of not streesing a carbon steel vessel or tank while it it is below its brittle fracture transition temperature is supported by several vessel and tank codes.

ASME VIII and API-510 bothe require that hydrotesting be performed at a temperature above the transition temperature. ASME VIII requires an added margin of 30 degrees F (ie, test at 30 degrees above the transition temperature). API-510 requres a temperature margin that varies by plate thickness (+10F for thicknesses of 2 inches and less, +30F for thicknesses abovce 2 inches).

Back in the 1980s I worked with BS-2654 a bit. I believe that the tank code required that the hydrotest be conducted at a test temperature that was above the transition temperature. Please allow for the fact that my memory may not be entirely correct on the British tank standard.

I trust that these examples illustrate the point that loading a structure when it is below its transition temperature is not a good practice.

Regards,
Steve Braune
Tank Industry Consultants

Steve Braune
Tank Industry Consultants

http://www.tankindustry.com

 

[VeryPicky]

Steve, I think you just touched on a very important subject, which is the hydrotest. Unfortunately I never worked to BS codes, so I can not quote them. The ASME I, a boiler section of the ASME does not allow the hydrotest to be performed in temperatures below 70°F (21°C) (metal temperature).
I always recommend for all pressurised equipment to be pressurised at the temperature above the MDMT.
The transition temperature is somewhat difficult to establish. A proper procedure would require a number of impact tests to be performed at various temperatures. I always go with the ASME VIII UCS-66 for the MDMT and transition temperatures for pressure vessel applications.

 

[liberoSimulation]

From reading all above discussions, I'm still not clear about the standards or practices of the design basis of the pressure vessels or towers.
I mean which factor of the design determines the material selection of the equipment.
If the contractor has told that LTCS is good enough for a service where the MDMT can be reached up to -100C, do you accept his material selection of LTCS while sudden depressizing might occur due to PC valve failure of operator malfunction of one of control valves to lower pressure source.

I think this needs much understanding prior making some not practical assumptions.

Cheers

 

[metengr]

The actual selection, and not determination, of minimum design metal temperature (MDMT) is by the designer of the component and not from Codes or Standards. For example, ASME B&PV code does not recommend any MDMT, all the code does is assure that if you select a material for low temperature service, proper testing is or has been performed to demonstrate that the material exhibits adequate toughness at or above a minimum design metal temperature (MDMT).

For proper selection of MDMT, the designer of the pressure vessel or tower must take into account all aspects of low temperature service conditions. Once a MDMT is selected by the designer, the code is referenced to assure proper material selection for safe service.

 

[matweld1]

thank you for all but the questions were :Is such quotation true? Does somebody know which frabication code (i. e. ASME) support such statement?
You will not find any code or standards given any precisition about the MDMT for the CS except for the material that is low temperature designation CS for exemple ASTM A 333 Gr.1.....6.or equivelent for plat or flange.
I have a discution this evening with process engineer and he was stonish to fing such material like API 5L Gr.B used for the low temperature -43ºC for propane liquide.
My answer is during the charpy test I find that the value of the energy (J)is multiple 10 than the value calculate or given in the specification of the material.

 

[liberoSimulation]

To metengr
Your explanation is totally true but you miss something that the designer is the service provider of the equipment and me as a client.
Do you think the contractor (designer of the process and who is selecting the material based on the operating temperature of the service with some margin) will consider sudden depresszing in the process that will create low temp. and that is the MDMT which the equilibrium temp. of the service at atmosphere pressure, in our case, depresszing to flare through failure of PC valves or operator malfunction..etc?

Theoretically agree with all, but practically that is another story!!

Cheers

 

[metengr]

Yes, they should have taken this into consideration!! I have seen and heard of this problem before. You, as the owner, are cautioned to either select an experienced engineering firm that can establish some type of design criteria and service conditions or provide this information yourself to the contractor – this is your responsibility. The contractor can take this information and build you a vessel using the rules of the code of construction. Many times, clients will short cut and use a contractor that can supposedly design and build. There is really not much you can do other than in the future buyer beware!

Codes and Standards are intended to provide guidance for safe operation based on following a set of rules for construction. Codes and Standards are NOT intended to substitute for engineering experience. You as the client ultimately assumes engineering responsibility for proper selection of materials and service conditions. In the future, you may want to hire an engineering firm that can provide the necessary design criteria based on your input.

 

[liberoSimulation]

Thanks metengr for the advice
I'm still looking for feedback from the members.

Cheers