Not sure, but if you completely remove all prior weld and HAZ and ensure the parent metal is not overtempered during the multiple PWHT processes ( by adding thermocouples directly under the induction heating coils) I think at least 3 weld cycles could be allowed.
If you do not remove all weld metal and HAZ metal, then the remnant HAZ becomes "reheated" and add'l carbide precipitation or aging occurs to the old HAZ and adjacent material- if the region that is over-aged extends beyond perhaps 2t ( t= wall thickness) then there may be concern that type IV creep damage would impact the life of the joint( I'm assuming we are talking about operating above 800 F).
The axial stress imposed on a butt weld is only 50% of the circumferential stress that the parent metal is exposed to, so the ( 50% ) weaker butt weld HAZ is normally permitted under the assumption that the damaged HAZ is thin- but once the damage or aged material exceeds a significant fraction of the wall thickness , then that assumption might not be valid.
A larger issue with the newer materials ( P91 et al)is that the PWHT that occurred prior to circa 2009 likely did not include monitoring thermocouples directly under the induction heating coil- that zone may have been overtempered. It has since been found that the temperature difference between the under-coil zone and the weld zone thermocouples could be as much as 100F- it is roughly proportional to the rate of pipe heat up ( F/min) times the square of the distance between the monitoring t/c's and the center of the induction coil zone. If you make that same mistake multiple times, it doesn't get better.
First off, there is no 'fixed' number or rule of thumb. The simple answer is it depends on the base material and if post weld heat treatment (PWHT) is required. For carbon steel with no PWHT requirement, you can weld repair as many times as necessary.
For other materials that are higher alloy, heat treated and air hardenable you need to qualify a weld procedure with some idea as to the number of PWHT cycles one would anticipate during fabrication and service. This way you can evaluate the weld region and base material based on ‘x’ number of total PWHT cycles. Again, there is no magic number of times, and you will not find this in most Codes and Standards.
