I am curious on what people think of this arrangement and perhaps they have seen it before. This is a 1750kVA, 12.47kV - 575V, pad mount transformer with a circuit breaker for the secondary side mounted inside the cabinet. So to reset the breaker or if one wants to read amperage or other info off the display, the transformer doors need to be opened. The incident energy level is around 28 cal/cm^2, so one would have to suit up to open the doors.

I have attached a picture of the set up.

Neat, but I have never seen a setup like that before.

A few questions about your IE calculation:
Are you calculating the incident energy at the load side of the breaker, line side, or transformer bushings?
What gap distance did you use?
What is the XF primary protection mechanism / type?

In short I am surprised the worst case IE calculation is only 28 c/c2.

Also curious do those feeders run into a cable vault?

I have not seen this type of set up. I do not allow my technicians to open pad mount transformers energized. We install the circuit breaker in a weather proof enclosure onto the side of the transformer. I know that most of the time we have >25cal/cm^2 even on the line side of the breaker, but with their PPE on and the covers safely and correctly installed, they can operate the circuit breaker without having to be exposed to any other hazard from having the transfromer open. We just installed a new pad mount at one of our stations this week with a current limiting circuit breaker. It has 28cal/cm^2 on the line side and only 4cal/cm^2 on the load side that feeds our MCC's in the plant.

Typo on my end, it should be 128 cal/cm2.

The gap distance used was 32 mm for <600V equipment.

Primary protection is a 100T fuse link.

The feeders run into a vault below the pad mount transformer.

Need for risk assessment here. What can happen by opening the door...what interaction is happening by opening a sheet metal hinged door with no electrical components, not even controls mounted on it? How can an arc flash even hppen here? IEEE 1584 is no longer sufficient alone to meet the risk assesament requirement of 70E-2015, nor of 1910.269 as of 2015.

True on the risk assessment. Items to consider is bare lugs on breaker exposed and phase spacing close, so what happens when opening the doors scatters mice up the cable to across the breaker lugs? Far fetched but in realm of possibility, I think.

Can you somehow cover the bare lugs, or at least discourage the possibility of a rodent bridging between phases?

The idea of a rodent causing a problem when opening a cabinet door has been an ongoing area of discussion at my utility. OSHA's 1910.269 Appendix E, Assessment Guidelines, says the following,

"The Occupational Safety and Health Administration will consider an employee exposed to electric-arc hazards if there is a reasonable likelihood that an electric arc will occur in the employee's work area, in other words, if the probability of such an event is higher than it is for the normal operation of enclosed equipment."

To me, the key to this statement is "if the probability of such an event is higher". Many people seem to think that the probability of an event is increased because a rodent is scared by the opening of a door. Personally, I'm not convinced that this is true and believe that maybe the probability is the same or even reduced. A rodent inside a cabinet would have already explored the entire contents and been at risk of a problem on a daily basis while the door was closed, so I don't believe the probability goes up. It might even be less since a scattering rodent is likely to retreat down, away from energized parts.

Most definitions of acceptable risk are about 1 in 100,000 to 1 in a million. This is the level where even signge or other simple precautions are unnecessary. Tolerable risk is around 1 in 10,000 to 100,000. This is from the AiChe CCPS LOPA Handbook. Using IEEE 493 data for breakers I get about 1 in 10,000 to 100,000 for arcing faults in general. For disconnects it is about 1-2 orders of magnitude lower. There are lots of factors that go into this to get an actual likelihood such as if arcing happens only during operation of the breaker but no hard data. ESFI and other studies place the average reported incident rate for arc flash at around 1 in 100,000 so clearly arcing fault does not automatically mean arc flash injury, and this is an average mixed in with lots of very at risk tasks (live work). This is while operating an apparatus which intentionally arcs, hopefully in a controlled manner, during de-energizing or re-energizing, which may (and has at my site) contain rodents or arachnids especially at certain times of the year. Just as with raptors and powerlines they can and sometimes do initiate an arc but are not limited to during operation. Depending on equipment design and especially system voltage, they can and do self clear so none of this matters, but with a 10,000 man operation spread over 3 countries we get about 1-2 arc flash injuries about every 10 years over the last 15 years (fuzzy records before then) which is in line with ESFI data. None of the incidents so far going back about 25 years involved rodents. And we run mostly metal enclosed gear (few if any phase barriers) so it is not inherently anti-rodemt. So in real world environments (mostly mining) with relatively high infestation rates, I am not seeing an unusually high arc flash rate. Granted this is anecdotal and hearsay on some internet discussion forum.

BBA inc. has done that type energy calculation and risk assessment in wind farms applications. It’s used to lower the energy level when the operator needs to open the wind turbine infeed cabinet. The breaker in the padmount is used as the primary protection for that cabinet. For the padmount secondary side energy calculation, the gap we used is the breaker gap distance, typically 32 mm and the primary side fuse protection will be the protective device interrupting the arc fault. PPE will be at 4 or >4 so a special risk assessment is made for the operator that will need to lock the breaker off or reset it. In the assessment an upstream disconnect switch (internally in the primary side of the transformer) is used. Furthermore, a plexiglass around the breaker is added for protection. We don’t recommend operating the breaker with the transformer being energized.

Daniel Lebeau, eng.

A breaker in the secondary compartment looks intriguing, but like others have said there would still exist a large amount of potential incident energy between the transformer secondary bushings and the line side of the circuit breaker.

Has anyone looked at using a switch in the primary compartment activated by a relay using CTs installed on the secondary side of the transformer? It would seem like this would provide a better solution than either a secondary circuit breaker or primary fuses. You wouldn't have to worry about the problem mentioned above, and it could be set to trip faster than primary fuses since transformer magnetizing inrush doesn't have to be accounted for (CTs on secondary side).

Cooper has a product marketed as their VFI Transformer which looks like it could be setup this way. Has anyone had any experience using this scheme to reduce arc flash hazard levels?

Any other thoughts as to why a solution like this would or wouldn't work?

Thanks!

It works. The scheme is called a virtual breaker. Use bushing CTs so the high incident area is essentially nil. I used an SEL 651 but separatebreakers work, too. Reclosers/breakers for 15-35 kv run around $18-$25k from Elastimold, Tavrida, and occasionally Cooper. Fuse cutouts run $10k installed so there is a price bump but not excessively high. Ifit is low voltage, you'd need a breaker with a shunt trip input and either a battery backup or capacitive trip device.