Has anybody been involved in recalculating arc flash after a significant reconductoring job? For a job that includes new relay settings, there could be a significant enough change to alter fault currents in nearby lines. How many tiers from your newly conductored line do you think you should recalculate arc flash hazards? I would estimate performing calculations until the value does not result in a change of HRC level. Any thoughts on this?

Yes, it's pretty common, but there are a host of factors that go into it.

Generally if you are only changing wiring then it USUALLY won't result in a big change unless there's a big change in impedance, and it depends on lengths, too. Close to a transformer where impedances are usually very small, "significantly" increasing bus length (in percentage terms) can have a big impact. And that's the big difference...what's the percentage change, and is the cable/bus itself the major driver in terms of impedance? If you purposely order transformers with high %Z, as is commonly done for large industrial plants with say 50+ MVA transformers where fault currents are approaching the interrupting limits of the breakers/fuses, it is not uncommon to see %Z well above ANSI. In these systems cable/bus impedance may well have no effect at all.

Once you start messing with overcurrent protection though, frequently this affects other downstream calculations because it manipulates the time-current curve. It does not however affect any other electrical properties so voltage and fault currents remain the same.

Any time you add/remove inductance downstream such as adding/removing motors, GENERALLY it will affect upstream values because it changes the fault current. Same thing happens with changing transformers. This affects upstream values. Usually the change is fairly small but if it pushes things past a "threshold" value (if for instance something was already borderline), it can result in a significant change to upstream values, and this can propagate all the way to the assumed utility bus. You can tell by looking at the change to the short circuit and arcing fault currents whether or not this change matters significantly.