Well, one very important variable you left out was the shielding gas. Your settings at your previous shop are pretty sound for 100% CO2 and the settings at your second shop could work with Argon mixes, but would be very difficult to operate with CO2. Also, the 21.5V and 180-190 wfs would be pretty sloppy with Argon, tons of spatter, wire fighting back to tip and even some hollow bead porosity in rashes with the excessive voltage for that gas. My guess is that you are running different shielding gases - for those reasons. Is there a Welding Procedure? If so, I would get with management and talk to whoever qualified it to discuss techniques used. The 19.5V and 300 wfs are not adequate parameters for use with CO2. I wouldn't recommend those for Argon mixes either, but I don't know the wall thickness nor position or rolling speed if it's rolled, or the root opening. It could be adequate. I have to believe the wall thickness is fairly thick given the SAW process for fills and cap. Even so, in my experience a 17.5 to 18V and 200 to 240 wfs with 85% Argon works well on heavy wall roots. As far as the nozzle and tip goes, it's all in relation to electrode extension (the amount of wire protruding from the tip to the welding arc). You want this "stick-out" to measure around 3/8". This will promote stability in the short-circuit arc and amperage at a given wire feed speed. If you are using a small tapered nozzle then the tip could be recessed. If you were using a large orifice nozzle then you would want the tip to be flush or even sometimes extending out beyond the nozzle, because you can't reach as far into the bevel since the large nozzle will be resting on the groove faces. A recessed tip and large nozzle combo using short circuit transfer are begging for incomplete fusion and porosity. Good luck
The difference that you mentioned was 70S-2 vs 70S-6.
I've compared 70S-2 to S-3 and the S-2 did not wash anywhere near as well as S-3.
Haven't done the same with S-6, but I'd start there, go back to what worked.
The last # (6,3,2 etc.) is representative of the amount of silicon in the wire. Higher amounts of silicon improve the weldability for dirtier material, ie welding over rust mill scale etc. Higher # = more silicon. Can make a difference but I would guess it is not the main culprit.
Gas can play an important factor as can voltage and WFS.
Higher voltage less penetration and wider bead
Lower voltage more penetration and narrower bead
Higher WFS more penetration
Lower WFS less penetration
This is why it is important to dial the two in to give the best of both worlds.
Dial these two in for the type of gas being used and you should see good results. Just a matter of a few trials to set the parameters.