The first draft of 70E-2018 uses the term "risk assessment" to be essentially what we could call a field-based risk assessment. This is also a "toolbox talk", "job briefing", or similar titles led by the person in charge who may be a foreman or a lead man on the crew and is performed by the crew doing the work. It is also sometimes called a pre-task plan, "Take 2", JSA, JHA, or many other words or terms. It can range from a simple form that has three columns labelled "task, hazards, prevention" or something similar, or may be a check list of generic hazard terms (least effective in my opinion) or it can be something much more detailed with pages of work procedures attached.

Usually what happens with these is that the first couple times there is a major impact because of lack of familiarity. A discussion about the hazards as seen by the workers ensues although it is usually involving those things unique to the job and skips over the ones that are "typical" for a given job site. But after a few times of doing this exercise it becomes a pencil-whipping exercise and quickly loses any value in recognizing critical changes to conditions which may be precursors to an accident. At best I could make a very small case for the new worker in that this could be considered a more formal knowledge transfer exercise. To date even though this seems to be a new and popular concept in the safety community, I am a little dubious based on actual experience with the implementation. I've seen no documented research or proof that these things actually work other than the gleeful claims of safety personnel during the first month of implementation.

The second type is much more formal. There are a lot of methods for finding hazards (what-if, structured what-if, HAZOP, matrix methods e.g. for chemical reactions, FMEA's focussed on safety). These are usually performed by a team that includes field personnel as well as various subject matter experts that do the exercise in an office setting, usually generically without having an actual work order in hand. Various studies have shown that the success rate of these exercises in identifying hazards is between 60 and 90%, partly dependent on the nature of the task analyzed. The hazard identification step is followed with either informal (e.g. HAZOP) or formal (e.g. LOPA, ANSI TR3, ANSI RIA, IEC 61511, fault-trees, Markov modelling) methods for identifying mitigation (if any) and its effectiveness. It is required for PSM compliance. I haven't seen statistics on the success rate but most practitioners seem to point to hazard recognition as the weak link rather than mitigation steps.

Taking maintenance and construction activities into consideration, there are a number of field conditions that can be forseen by the formal (office) method but can't identify the specific situation on a case-by-case basis. Thus what will come out of the formal study is a procedure with a list of inspections to be performed. The results of those inspections direct the work down different procedures depending on the outcome. For instance an inspection may be required to identify "burn marks". If any are found, arc flash PPE may be required or the power disconnecting device may be changed for a particular task while if none are found, PPE may not be required. It may appear at first glance that this is really the same as the first approach but that is not the case because the hazards and the plan of action were determined ahead of time.

As summarized in at least one public comment it is very clear that when OSHA refers to a risk assessment, they mean the formal version and the responsibility is placed on management. Field risk assessments are the reverse case...placing sole responsibility for identifying and mitigating hazards on the workers. So it seems that 70E-2018 in the current draft is at odds with the intent of OSHA regulations as I see it. However my reading of the changes to 1910.269 in the 2014/2015 update clearly point to doing field inspections. This is not really in the scope of 70E since generally generation/transmission/distribution would be more in the scope of NESC but it is an interesting new direction nonetheless.

My questions are:
1. Has anyone seen any formal studies and/or documented evidence of the efficacy of field risk assessments, other than anecdotal evidence (flavor of the month club) where most of the gains are short lived due to simply raising awareness? I've looked everywhere and come up empty handed.
2. Am I reading it correctly that OSHA supports and almost mandates formal risk assessments and that they recognize but do not require or even strongly suggest field risk assessments?
3. Knowing that there are a myriad of environmental factors which may indicate a potential for a hazardous condition (such as burn marks), there seems to be a strong case to be made for the field risk assessment but can the formal version still at least identify any required inspections?