Old Australian steel, MDMT calc, AS vs API579
Old Australian steel, MDMT calc, AS vs API579
(OP)
Hello colleagues.
We have three 1960 package D type boilers. The steam and mud drum are made from AS B58 gr A as rolled plate, circa 1958. They were post weld heat treated at the time. The steam drum is about 77mm (~3")thick.
We have reason to doubt the early 90's determined MDMT / MAT for the drums (there was a plant wide B.F. review at the time).
Here's the dilemma.
AS B58 grA from '50's/'60's is known to be crappy steel and Australia's current pressure vessel code doesn't even mention it. It would however have to be considered as a curve A steel based on AS1210's MAT curves however.
It would also have to be considered as a curve A steel on the API579 / ASME VIII MAT curves.
AS1210's MAT curves go up to 40mm (~1.57") governing thickness. Extrapolation is not recommended. It gives credit for PWHT via a second set of curves which actually goes up to 100mm thickness. Though it sounds like I should be able to use the second curves, I think the curves really reflect modern steel practices and not the open hearth methods of the 50's and so I'm reluctant to apply the curves.
My wariness is further enhanced when I read in API579 (section 3) that you can't take PWHT credit if the governing thickness is above 1.5".
So given the 40mm limitation of the AS1210 MAT curves, I think I'm forced to use API579. Does anyone violently disagree?
So using API579 and not taking credit for PWHT, I'm left with MAT of 111F (44degC).
This is significantly higher than the listed MAT that we currently have for the drums and the documentation for setting the MAT in the past is unfortunately ambiguous.
Now here's an interesting piece of info... in the past the boilers have been hydrotested at the lower figure (past practices didn't apply a margin over the MAT... scary really) and the boilers survived. If I can cobble together proof, I think I will have a case to grandfather the boilers to a MAT below API579 figure. Is this reasonable?
Alternatively, I could arrange a controlled temperature hydrotest and take hydrotest temperature credits for setting a new MAT. Given that the boiler has well survived hydros well below the API579 figure in the past, what thoughts/objections rise in mind if I was to target a hydrotest at the 111degF/44degC figure??
Some other questions this scenario has thrown up:
- Just how conservative is the brittle fracture side of our codes? By rights, our boilers should have failed brittly years ago.
- Why does API579 put a thickness limit on PWHT credits? There are soak times based on thickness, so why limit PWHT credit to 1.5" or less??
Thanks in advance, I look forward to the discussion.
Cheers
Rob
We have three 1960 package D type boilers. The steam and mud drum are made from AS B58 gr A as rolled plate, circa 1958. They were post weld heat treated at the time. The steam drum is about 77mm (~3")thick.
We have reason to doubt the early 90's determined MDMT / MAT for the drums (there was a plant wide B.F. review at the time).
Here's the dilemma.
AS B58 grA from '50's/'60's is known to be crappy steel and Australia's current pressure vessel code doesn't even mention it. It would however have to be considered as a curve A steel based on AS1210's MAT curves however.
It would also have to be considered as a curve A steel on the API579 / ASME VIII MAT curves.
AS1210's MAT curves go up to 40mm (~1.57") governing thickness. Extrapolation is not recommended. It gives credit for PWHT via a second set of curves which actually goes up to 100mm thickness. Though it sounds like I should be able to use the second curves, I think the curves really reflect modern steel practices and not the open hearth methods of the 50's and so I'm reluctant to apply the curves.
My wariness is further enhanced when I read in API579 (section 3) that you can't take PWHT credit if the governing thickness is above 1.5".
So given the 40mm limitation of the AS1210 MAT curves, I think I'm forced to use API579. Does anyone violently disagree?
So using API579 and not taking credit for PWHT, I'm left with MAT of 111F (44degC).
This is significantly higher than the listed MAT that we currently have for the drums and the documentation for setting the MAT in the past is unfortunately ambiguous.
Now here's an interesting piece of info... in the past the boilers have been hydrotested at the lower figure (past practices didn't apply a margin over the MAT... scary really) and the boilers survived. If I can cobble together proof, I think I will have a case to grandfather the boilers to a MAT below API579 figure. Is this reasonable?
Alternatively, I could arrange a controlled temperature hydrotest and take hydrotest temperature credits for setting a new MAT. Given that the boiler has well survived hydros well below the API579 figure in the past, what thoughts/objections rise in mind if I was to target a hydrotest at the 111degF/44degC figure??
Some other questions this scenario has thrown up:
- Just how conservative is the brittle fracture side of our codes? By rights, our boilers should have failed brittly years ago.
- Why does API579 put a thickness limit on PWHT credits? There are soak times based on thickness, so why limit PWHT credit to 1.5" or less??
Thanks in advance, I look forward to the discussion.
Cheers
Rob
- - - - - - - - - - - - - - -
"Life! No one get's out of it alive."
"The trick is to grow up without growing old..."





RE: Old Australian steel, MDMT calc, AS vs API579
Before you go any further, have you run this by the local regulatory body for boiler safety in AU, and as a minimum your Inspection Agency that insures the boilers? If not, you had better do this before you go down a path that is not approved for use.
Here are my thoughts on your questions;
Agree
Disagree. Just because you have hydrotested the drums below the MAT and they survived (no brittle fracture) doesn't provide a green light to go to metal temperatures below MAT and API 579 for this material. You have demonstrated that the drum metal has low notch toughness, and I would suggest you apply the API 579 curve as the lower limit.
Agree
You need to discuss this with someone that actively attends AS Boiler Code meetings. From ASME B&PV Code, they do not address brittle fracture of boiler drums because after the initial hydrotest, ASME does not care about subsequent pressure tests with water. Also, one obvious reason for ASME sticking their head in the sand on this one is that boilers operate well above the DBTT of lower and steam drum steels and as such, is not a concern to most committee members.
Because of the thickness constraint. In other words, any pre-existing defects or cracks would be treated in plane strain condition, which assumes very high stress intensity with little plastic deformation at the crack tip. PWHT doest NOT improve thickness constraint.
RE: Old Australian steel, MDMT calc, AS vs API579
The regime that's run in Victoria Australia is essentially self managed via a safety case, with the regulator retaining audit rights at any time. There is no suitable regulatory Chief inspector type person to bounce these things off. Believe me, we are discussing this scenario in detail within the technical folk of our organisation and need to develop a rigorous robust solution that would withstand external technical review.
Well that's a very interesting response. Why have you disagreed with the grandfathering proposal based on previous hydro's but agreed with the path of conducting a new controlled hydro?
But more interestingly, are you disagreeing with the philosophy of grandfathering?
Exxon Mobil (we used to be a part of them) had guidelines on grandfathering equipment that had seen colder than MAT service. There were many precaution and qualifications, but it boiled down to carefully understanding the operating conditions. In this case we've survived a 1.5xdesign pressure hydrotest in the past at a shell temperature below the recommend MAT of API579. To me this is a compelling grandfathering argument. Infact, the more I think about it, given that API579 is not willing to give MAT margin credit for the stress relief, I'm coming to a conservative position that I'm unwilling to accept the 17degC MAT credit hydrotest margin. It would seem to be a suitably conservative position that if the drum survived hydrotesting at a certain temperature, then we know it's at least good for that temperature and would be expected to happily withstand that temperature at design pressure.
As to the question of how conservative our codes are in regards to brittle failure, this was not a question directed at ASME I. It was directed to pressure service in general. I have a hunk of supposedly crap steel in the shape of a pressure vessel that has survived a hydro test at temperature reportedly 14degrees C BELOW the MAT curve of API579. Have I got a good example of this old steel? Did stress relief infact make all the difference? Are the B.F. charts overly conservative? [In my time I've had modern CS steels charpied at temperatures below MAT curve values and have come up trumps every time... this question has cross my mind often over the years]
- - - - - - - - - - - - - - -
"Life! No one get's out of it alive."
"The trick is to grow up without growing old..."
RE: Old Australian steel, MDMT calc, AS vs API579
I just think you stand a better chance of using the controlled hydro approach in this situation because the key word here is "controlled".
No. Grandfathering is a case-by-case approach, and my opinion is that for hydrotesting this does not apply due to the complex variables associated with brittle fracture.
Good questions with no absolute answers. The solution is to remove a sample from the drum and run impact tests. One apprach is to use the Punch test for determination of DBTT. This would provide DBTT data that you could defend regarding hydrotesting temperatures in the future.
RE: Old Australian steel, MDMT calc, AS vs API579
I've since learned that our circa 1960 AS B58 plate material was closest or the equivalent of A201 grB material.
If you had a 3" walled vessel and carried out a controlled hydro at a specified temperature, would you be happy to apply the 17degC margin reduction that API579 allows. i.e., MAT = T(hydro) - 17DegC ??
- - - - - - - - - - - - - - -
"Life! No one get's out of it alive."
"The trick is to grow up without growing old..."
RE: Old Australian steel, MDMT calc, AS vs API579
Reference to API 579-ASME FFS-1: Fitness For Services 2007?
Regards
Leonard Thill
RE: Old Australian steel, MDMT calc, AS vs API579
For anyone interested, I've marked up the MAT curve and attached it as a file.
For the record, we are going to hydro one of the boilers.
Cheers
Rob
- - - - - - - - - - - - - - -
"Life! No one get's out of it alive."
"The trick is to grow up without growing old..."