Dear Cesar,
In case the runs are corresponding to the same scoring but just executed in different jobs you can simply use the command line tool usbsuw which you can find in pathtofluka/src/tools/. In case you combine different contributions over diverse runs, e.g. using two qualitatively different source terms, you need to use usbscw, as mentioned by @horvathd in this post.
Hope this helps, don’t hesitate to reach out again if something is not clear.
Andreas