Here below is some experiences about the stress distribution around a reinforced nozzle based on the FE study.
Pad or repad, often used to increase the available area as per UG-37, is essentially introduced to redistribute the local bending stress due to the internal pressure or piping loading. However, as some codes recommended, the size needs to be limited. The reason is, increasing the width within some limits, the peak stress is relocated from the inner conner of nozzle to shell, to the pad and shell (in some cases, the peak stress moves to the edge of the pad), and the stress magnitude drops.
However, beyond the limit, there is little impact on the stress magnitude.
So I think that's why the size of pad needs to be limited.