robsalv
Mechanical
- Aug 8, 2002
- 311
I have a old 3" (77mm) thick vessel made from circa 1958 AS B58 Australian steel, which I believe is equivalent to A201 grB. The vessel was stress relieved at the time and 100% radiographed.
According to the API579 fig3.4M MAT curve, the steel has a MAT of 44degC (111degF). Since API579 doesn't allow any credit for PWHT for thicknesses above 38mm thick, 44degC would appear to be the MAT for this vessel.
Anecdotal evidence suggests that the vessel has survived multiple pressure tests and at least one hydrotest at well below this temperature.
We recently conducted a 1.5xDP hydrotest at 44degC - I did not apply the +17degC recommended margin as I was confident about the quality of the anecdotal evidence.
I now have a documented 1.5xDP hydro on record at 44degC. So my first query is: given that circa 1958 steels were open hearth and their quality (grain size, carbon addition, inclusions etc) is possibly suspect, would you be willing to apply the 17degC MAT hydro credit that API579 allows and call the MAT 44 - 17 = 27degC??? If yes, why is this margin acceptable for a thickness of steel that API579 excludes PWHT benefits?
(As an aside, do you think that the curves accurately reflect MAT for old crappy steels? Are some curve A steel inherently better than others?)
Second consideration: This vessel is actually the steam drum of a boiler. The mud drum is made of identical material, was similarly constructed and is 2.75" (~70mm) thick. During the hydro it recorded a temperature of 28degC (water fill is through the steam drum). According to the MAT curve, it's MAT is 42.5degC (108.5degF), i.e. 1.5degC less than the steam drum.
According to API579, the mud drum could actually have the hydro margin applied and the MAT could be set at 28 - 17 = 11degC (52degF). Frankly, I feel very uncomfortable in my engineering gut about this. It just seems too low for old thick walled steel. Thankfully this is a boiler so it's operational CET means that discussions of MAT are largely academic - however, leaving that to one side, my conservative side is happy to call the hydrotest temperature the MAT for the drum - i.e., no credit/margins applied.
This brings up an interesting question since the drums expereinced different temperature during the hydrotest (again, the mud drum's temperature was perceived to be OK given past anecdotal evidence and that it has a smaller ID)
Given that the materials and construction are the same and the thicknesses are similar, could you mount an engineeringly supportable case to set the MAT of the steam drum at 29.5degC, i.e., the mud drum's recorded hydro temperature, PLUS the MAT curve positive delta for the steam drum of 1.5degC?? Is that a misuse of the MAT curves? I guess I'm asking in another way, in setting a MAT for the boiler, can any credit be taken for the lower mud drum temperature?
If the hydro credit margin does apply for the mud drum and 11degC is allowable, does that mean the steam drum could be allocated a MAT of 11 +1.5degC = 12.5degC ???
I'm sure this query will have set off some controversial thoughts. Let's discuss them.
Bottom line, since the steam drum is the thickest component, it sets the global MAT for the boiler. What MAT should I set?
* 44degC (based on the MAT curve figure alone - forget the hydro, we were lucky the boiler survived, move on)
* 27degC (based on the steam drum hydro minus the allowable hydro credit)
* 29.5degC (being the mud drum's hydro temperature plus the MAT curve delta)
* 12.5 degC (being the mud drum's hydro temperature minus the hydro credit, plus the MAT curve delta)
????????????????
- - - - - - - - - - - - - - -
"Life! No one get's out of it alive."
"The trick is to grow up without growing old..."
According to the API579 fig3.4M MAT curve, the steel has a MAT of 44degC (111degF). Since API579 doesn't allow any credit for PWHT for thicknesses above 38mm thick, 44degC would appear to be the MAT for this vessel.
Anecdotal evidence suggests that the vessel has survived multiple pressure tests and at least one hydrotest at well below this temperature.
We recently conducted a 1.5xDP hydrotest at 44degC - I did not apply the +17degC recommended margin as I was confident about the quality of the anecdotal evidence.
I now have a documented 1.5xDP hydro on record at 44degC. So my first query is: given that circa 1958 steels were open hearth and their quality (grain size, carbon addition, inclusions etc) is possibly suspect, would you be willing to apply the 17degC MAT hydro credit that API579 allows and call the MAT 44 - 17 = 27degC??? If yes, why is this margin acceptable for a thickness of steel that API579 excludes PWHT benefits?
(As an aside, do you think that the curves accurately reflect MAT for old crappy steels? Are some curve A steel inherently better than others?)
Second consideration: This vessel is actually the steam drum of a boiler. The mud drum is made of identical material, was similarly constructed and is 2.75" (~70mm) thick. During the hydro it recorded a temperature of 28degC (water fill is through the steam drum). According to the MAT curve, it's MAT is 42.5degC (108.5degF), i.e. 1.5degC less than the steam drum.
According to API579, the mud drum could actually have the hydro margin applied and the MAT could be set at 28 - 17 = 11degC (52degF). Frankly, I feel very uncomfortable in my engineering gut about this. It just seems too low for old thick walled steel. Thankfully this is a boiler so it's operational CET means that discussions of MAT are largely academic - however, leaving that to one side, my conservative side is happy to call the hydrotest temperature the MAT for the drum - i.e., no credit/margins applied.
This brings up an interesting question since the drums expereinced different temperature during the hydrotest (again, the mud drum's temperature was perceived to be OK given past anecdotal evidence and that it has a smaller ID)
Given that the materials and construction are the same and the thicknesses are similar, could you mount an engineeringly supportable case to set the MAT of the steam drum at 29.5degC, i.e., the mud drum's recorded hydro temperature, PLUS the MAT curve positive delta for the steam drum of 1.5degC?? Is that a misuse of the MAT curves? I guess I'm asking in another way, in setting a MAT for the boiler, can any credit be taken for the lower mud drum temperature?
If the hydro credit margin does apply for the mud drum and 11degC is allowable, does that mean the steam drum could be allocated a MAT of 11 +1.5degC = 12.5degC ???
I'm sure this query will have set off some controversial thoughts. Let's discuss them.
Bottom line, since the steam drum is the thickest component, it sets the global MAT for the boiler. What MAT should I set?
* 44degC (based on the MAT curve figure alone - forget the hydro, we were lucky the boiler survived, move on)
* 27degC (based on the steam drum hydro minus the allowable hydro credit)
* 29.5degC (being the mud drum's hydro temperature plus the MAT curve delta)
* 12.5 degC (being the mud drum's hydro temperature minus the hydro credit, plus the MAT curve delta)
????????????????
- - - - - - - - - - - - - - -
"Life! No one get's out of it alive."
"The trick is to grow up without growing old..."
