In 2015 NFPA 70E, Table 130.7(C)(15)(A)(a), Task for insulated cable examination with manipulation of cable, it is stated arc flash PPE is required. Is there any where in 70E defines manipulation?
I have a project to replace 15kV old lead insulated cable. Each set of cable represents a circuit to a motor unit, and I have a total of 10 sets. Each set of circuit cable is laid on a cable tray, and each cable tray is about 36 inches apart. Our plan to replace the cable is one circuit at a time while the other circuits are energized. In this situation, should we need any PPE?

70E is talking about energized work where for instance you are digging into a cable tray to inspect a cable. If you've ever done this, you probably know what a scary and potentially hazardous situation this is if simply moving a defective cable arcs it.

From your description it sounds like everything in the tray would be de-energized. What interaction with energized cables is going to happen?

Thank you for your comments and information. I would assume "Manipulating an insulated cable" is for moving or splicing energized cable.
In my situation, the circuit cable that will be replaced is certainly de-energized. However, contractors who are working will be in a close proximity to an energized circuit which is also insulated. The concern is making sure the contractor is cutting the right cable, and these cables are 750 mcm.

750 MCM? Sounds like somebody had it in for the electricians. So that added, what triple the cost both in cable and installation and 25% more ampacity over another 350? Typical idiot contract engineering.

Why not use non contact voltage detectors and/or cable tracers for added protection? Also the reason meter probes are spikes is so you can push them through the jacket to get a reading. If you want something more remote, Honeywell has a hydraulic cable spiking tool that punctures the cable with a steel spike to remotely ground it as verification. Its more common in Europe than U.S. practices. When looking at this obviously the first method is non contact so no interaction that can cause an arc is taking place. The second is similar because you can de-energize and test at a marked location where mixing up cables cannot happen. The third is a but more risky except nothing is really ever exposed so again, no PPE needed. Cable spiking remotely can cause an arcing fault but is more of a bolted fault by design and is done remotely so again no PPE needed. Following 70E LOTO if you are not sure if the correct cable is de-energized the standard procedure is to test for absence of voltage so above 1000 V noncontact voltage readings are the definitive test. Under 1000 V direct contact which would be using meter probes to penetrate the insulation is the correct method. Spiking satisfies grounding if this is an absolute must. Cable tracers and using noncontact probes in a low voltage scenario improve confidence in identification.

There are also utility cable cutters that are hydraulic and insulated for hit stick use. Its not cheap but very effective. This could be an alternative and could be used as "energized" work. Justification would be that absolute identification is not possible and exposure has been eliminated or reduced to as low as practical. I've used these cutters without the remote attachment. You need something like these if you are cutting 750 mcm anyway.

Finally you could just put in new tray and pull new wire until you have bypassed all the old stuff, then demo it all at once. Trays are very inexpensive overall.