Who designed the weld and the joint detail? How is it stressed and loaded, and how highly is it stressed? Do they really need a full pen. weld, or would smaller bevels or even two fillets work? If this is a flange to web weld, it seems that many people mistakenly assume that since a rolled WF beam is solid steel at this flange/web joint, all plate girders or built-up girders must be too. This is rarely the case. The fact is that many (some percentage) of the larger rolled sections have rolled-in defects (inclusions, piping and the like) at this very location, similar to what would occur if you didn’t backgouge the double vee groove before welding the second side. Ask them why they don’t want you to backgouge before the second weld, it just may not be needed, but then the full bevels probably aren’t either. I’m not saying this complies with AWS, or advocating doing this, but with SAW the penetration is good enough that the second side weld probably boils any problems out of the root of the first side weld, when the root face is small.
If they really need a full pen. weld, then backgouging is more than likely needed. This might be the case on a flange plate splice which is highly loaded, perpendicular to the potential defects. But, then you have the wrong weld joint designation. And, I would most likely UT this weld after completion.