It distributes the load reaction over a larger surface of the pipe to avoid local stress issues. In severely loaded applications, this may avoid immediate failure due to punching shear.
It is also common practice in alloy piping when you plan to weld the support to the pipe. Where PWHT is necessary the pad can be welded onto the pipe in the shop and given a proper PWHT, then when the support is welded in the field the quality of the field PWHT will not affect the pressure containment boundary.
I had this discussion with a young engineer in our office the other day. He thought this sort of support should be modeled as an intersection. I pointed out that we have not punched a hole in the run pipe, so the effects of the localized load are much less severe. Still, we are applying a force that will tend to ovalize the pipe, reducing its Ixx about the axis perpendicular to the load. So there is going to be some effect.
I suspect someone, somewhere, has done some research on the matter, but maybe not. Rather than mess around with a lot of esoteric analysis, I'd rather just use a load pad and save the worry for important stuff.
_________________________
CraigB