Strength-temperature relation of ASTM A387 Grade 22 1

[phex]

I'm looking for a relation of maximum yield strength over temperature and maximum tensile strength over temperature for A387 Grade 22 Class 1 & Class 2 pressure vessel plate steel under presence of hydrogen.
I need this because we need to know the minimum save temperature to apply pressure to a reactor for sulphiding the catalyst. I also asked the manufacturer, but the response is too d**n slow.
As additional information, the wall thickness is 77 mm and we need 70 barg pressure. Our information was that you will need at least 150 °C, but I need to verify if 120 °C is enough.

Every scrap of information is greatly appreciated. Thanks for your help,
chris

 

[metengr]

phex;
Are you looking specifically for SA 387 plate steel tensile data obtained at elevated temperature in an atmosphere containing hydrogen gas? There is published data showing the response of yield and tensile strengths for this material as a function of service temperature. I personally have not seen mechanical property data published for atmospheres containing partial pressures of hydrogen.

I would believe that the response of tensile and yield strength data at elevated temperature will be independent of this environment. The environment would affect crack propagation rate once a defect is present. Hydrogen gas at elevated temperature and at 1 atm will do nothing to affect the mechanical properties of the steel. It is hydrogen atoms formed as a by-product of chemical reactions in the vessel or from corrosion of the vessel wall that results in some form of hydrogen damage.

 

[phex]

Thanks for your reply, metengr. The vessel in question is a deep desulphurization reactor, meaning it is operating at high temperatures and pressures (about 400 °C and 70 barg). So one of the prerequisities for the MoC would be high temperature resistance (which is what the material is supposed to have).
The other point is, that the diffusion of hydrogen into the material is obviously temperature dependent also. At least, this is what the manufacturer thinks, as he made the statement that to operate at pressures above 25% of working pressure, we need to have at least 150 °C wall temperatur on the outside surface. When I looked at the yield- and tensile strength over temperature curves under normal circumstances, this material behaves just as expected: the higher the temperature, the lower the respective strength. So I am at a miss as to how the manufacturer came to the above mentioned premise. The only possible answer in my opinion would be the presence of hydrogen, which undoubtly has an influence on material strength, though normally only over time, not instantaneous.

If there is another cause for this, I would be very grateful for hints. Understanding the reason for this premise would enable us to adjust the parameters according to our needs.

Once again, thanks for all kinds of suggestions,
chris

 

[phex]

Another thing came to my mind, and I would like to hear some opinions on that.

Is it possible that this requirement is to be read like this: Once the reactor has been heated up to operating temperature for the first time, the structure of the metal changes. This change in the structure results in the fact that AFTER THE FIRST HEAT-UP you shouldn't operate under full pressure at lower temperatures, because due to this change in structure, the low temperature embrittlement is noticeable at higher temperatures than with "un-heated" metal (eg. already at +20 °C instead of -10 °C).

Comments?

Thanks for bothering,
chris

 

[metengr]

phex;
According to your stated service temperature of 400 deg C (752 deg F), I would consider this to be very suitable for SA 387 Grade 22 (a 2.25% Cr - 1% Mo alloy steel) plate. For you to alter the microstructure of this steel would require service temperatures in excess of 1000 deg F for a prolonged periods of time, even in a hydrogen environment. The Mo in this steel is a good carbide former, which would reduce the tendency to form methane pockets in this steel resulting in hydrogen attack. As far as strength, the Mo also increases elevated temperature strength.

The only thing I could think of for recommending 150 deg C external vessel wall temperature during initial operation at higher then design operating pressure is to prevent harmful thermal gradients in the vessel wall that could be additive to the primary membrane stresses. Other than this, the issue of hydrogen charging is a mute point because above 400 deg F, hydrogen atoms diffuse readily thru steels. So, I don't believe you would embrittle this material on start-up.

 

[phex]

Thanks again, metengr. It seems the only way to get the gist behind this requirement is waiting for the manufacturer to answer our request. Let's hope it won't take too long...

 

[GenB]

I see no problems with the material but I think the welds are to be paid special attention, also the tests as
PWHT and harness tests are needed.

This is a perfect question for

http://www.nacecorrosionnetwork.com

Regards,
ER

 

[phex]

Thanks for your comments, generalblr.

How can I access the site you mentioned? I always get a login request. Is there an url where one can register or is it kind of "invite-only"?

PWHT has been done twice, once after the reactor was manufactured and the second one after the weld-plating (correct term? There is a plating on the inside to protect the wall applied by welding) has been done. These things have been kept in mind.

Bye for now,
chris

 

[GBENARD]

We used an API standard for the design of our reform gas boilers a few years ago.
"API 941, Steels for Hydrogen Service at Elevated Temperatures and Pressures in Petroleum Refineries and Petrochemical Plants, Fifth Edition (1997, January). Supplement 1 (1998, April 1) also available."
Also known as nelson's curve.
You have a simplified version in Perry's chemical engineer handbook. At first sight, 22 seems immune at 400°C even with a 70 bar H2 partial pressure (better to check with last revision of API, it refers to long term operating cases).
Hope it helps

 

[phex]

Thanks for your comment, but I was looking for data at low temperatures especially after it has been heated up to operating temperature and cooled down again.