Nosetackle,
The only similar situation I have encountered is after "buttering" you perform the PWHT prior to joining to the carbon steel pipe.
We had a piece of CA6NM (cast stainless steel) that we buttered with 410 (if my memory is correct), carried out PWHT and then after PWHT we attached a piece of 316 s/s.
Hope that helps,
Regards,
Shane
By js55 
Date 06-30-2010 11:55
Edited 06-30-2010 12:00
nosetackle,
Transition pieces do not eliminate carbon migration. The only way to minimize carbon migration is to use a nickel based filler. And even then you will have a carbon denuded zone and a carbon enriched zone right along the fusion line.
What you are proposing is extremely conservative and IMO unecesary. Butter the P91 with a nickel alloy. run your P91 cook. Then weld out the carbon steel with the Nickel alloy as well. This also has the distinct advantage under Section IX of simplifying the qualification because you are using the same filler for the butter as you are the final deposit. You just have to qual with and without a cook.
If you use a transition piece P91 to P22: if you use B9 filler the carbon will still migrate from the P22 to the B9 filler denuding the P22 of carbon along the fusion line and enriching an already hard B9 weld deposit with even higher carbon. If you use B3 filler, which would be my preference, the carbon will move from the B3 filler to the P91 base metal. You will still have the same mechanical phenomena but now you have softened the B3 filler which will still be stronger than the P22 base metal, and you will harden the P91 base metal, but it can handle it better than B9 filler can.
After reading this, my question is why would you ever want to allow carbon steel to be welded to P91? I would only expect this to happen at a point in the piping system where the temperature drops below 800 degrees F, since carbon steel should not be used above that temperature. Perhaps this is at a point where a bypass line attaches to a condenser downstream of attemperator sprays, or a similar application. If this is the case, carbon migration is not going to be an issue with that low of an operating temperature. Either the buttering layer or the entire weld will need to be postweld heat treated at the P91 temperature range of 1350-1425 degrees F. This would be above the lower transformation temperature of 1335 F for carbon steels, and would cause some grain growth in the carbon steel. However, if the the pressure is low, which is typical at a condenser connection, and notch toughness is not a requirement, the grain growth would not be expected to affect the performance of the weld. So if the procedure qualification passed the ASME IX requirements, and the design temperature was below 800 F, and the pressure is "low", it might be an acceptable case. However, I would not put a procedure like this out for general use, because first thing you know someone that does not understand the limitations would be using it to attach fitup lugs, carbon steel pipe supports, HP drain lines or other in other cases where carbon steel should not be used on P91.
I have not seen a code yet that would allow it, or a welding procedure qualified for it. Every time we have went P91 to Carbon steel, we do it with a 12" spool piece between the two. Even at 12" you are probably pushing the limits of the Carbon Steel during PWHT.
The codes really don't provide any rules that would prohibit a carbon steel to P91 weld as long as it passed ASME IX qualification. Our typical practice on a HP or IP bypass line is to make the change at an elbow, with P91 to a WP22 elbow, then the WP22 elbow to the carbon steel pipe. Of course this has to be at a point where the temperature is low enough to use carbon steel.
i think to use wp22 elbow is a brilliant idea. it will minimize all problems... unfortunately in my case the elbow is CS. thats why i am thinking about transition piece between CS and P91.
It was quite common in the early days of Grade 91. Not so much these days. If you use a nickel alloy filler you pretty much eliminate carbon migration which exists in pretty much every dissimilar if there's carbon available to move around.
If memory serves, the first step in the olden days was welding Grade 91 to carbon by using a buttering layer(some morons (not this moron) tried using Ni alloys as a direct weldment without butter and keeping the cook to 1300 which did little for the 91 but it met code (I did have customers try and tell me it was OK because they didn't want to pay for a butter-this was before that new fangled stuff about lowering transformation and Ni content etc was common knowledge-though I believe Oak Ridge/DOE had data going back to the 60's on that when they were developing this stuff for nukes). Then this was abandoned for the the transition piece considered as a conservative approach. Then that was abandoned for the double transition. Really not sure why that would be necesary. I'm with Marty. Just don't do it. But I ain't no powerhouse deesigner. If you don't like a dissimilar flange it. But thats one of them thar deesigner deesizions.
Why it is done I honestly don't know. But fabricators will build to the designers design.
Oh, and one other thing, if you butter the 91 and then weld the CS to the butter layer you need to utilized longitudinal bends or your bend will look like a broken finger because the HAZ of the CS will take all the elongation.
I've done transveres with it and been successful but you wouldn't want to bet on it.
MBsims,
you are right it is piping connection point between HP bypass valve and condenser, design pressure 15 barg and design temperature 556C. exact connection point is p91 pipe to CS 90deg elbow.
