I have a Switchgear breaker that is not an electrically operated breaker. It has a remote racking and a remote closer, but you have to charge the breaker springs manually. Question is do you need to don your PPE to charge the springs?

It depends.Clearly charging is part of operating as is pushing the close or trip button. IT has nothing to do with maintenance which is a totally separate category.

70E gives you two choices for how to do things. You can either follow the table-based risk assessment approach or you can do the engineered approach where you do your own risk assessment. If the former, then you follow the tables to the letter. First you decide whether or not PPE is needed as per the 2015/2018 edition with the table. Look specifically at normal operation. Do you meet all the criteria there? There is one that you might not. A lot of breakers (ICCB and ANSI) have an outer cover/door which is really just a piece of sheet metal which protects against even inadvertent contact. With the door open the "business" end of many breakers is blocked by the breaker itself in such a way that it is protected against inadvertent contact so this qualifies as guarded. This is important because the definition of arc flash (see the definition section) refers to INADVERTENT contact. I think we can agree that idiots can always find a way to bypass any and all guarding in such a way that no amount of guarding will ever stop them, and the worst ones are safety specialists. So by design the breaker really isn't exposed in these cases so you might still meet the "normal operation" criteria and thus the answer is NO PPE needed.

Second approach is the engineered approach. In this case you are precluded from "using" the table or mixing and matching. But you are free to adopt portions of 70E including the tables or the table in 1910.269 Annexes that works very similar. In this case the 1910.269 criteria is whether or not personnel are inside out outside the minimum approach distance (MAD) which is the term used in IEEE 516 which for some strange reason 70E chose to rename as the "restricted approach boundary". The distances are identical. If the equipment is not exposed (it is insulated, isolated, or GUARDED), then it is not exposed so the boundary does not exist. Even if we accept that the boundary does exist if you are outside the restricted approach boundary with body parts and tools again we end up at the same place...no PPE needed. BUT in this case you really need to do a proper risk assessment.

I would recommend the AiChE LOPA (Layers of Protection Analysis) standard for performing this kind of analysis of electrical equipment. Most risk assessment standards (e.g. ANSI B11, NFPA 70E Annex F) have the problem that they are designed around moving mechanical equipment where exposure risks are measured in minutes or hours, and where failure rates of equipment are fairly high whereas electrical equipment failure rates are typically 1 in 100,000 or better and exposures are maybe once a year or less. This matches up much more with chemical plant incident rates and equipment so this is why I suggest the AiChE standard over one where you kind of get a lot of "defaults" that don't work in reality. The method listed in Annex F of 70E is actually an attempt to use an ANSI B11 style methodology and utterly fails when trying to analyze electrical hazards because the incident rates and reliability just don't match up to what it is designed for. If you'd like, it is fairly readable...just attempt to use it once and notice also that there are no specific criteria for many of the terms such as likelihood of "remote" leaving the end user purely guessing which is not something that should be happening in a safety standard.