LOTO "exceptions"

[Jraef]

I have a project in which there is a dispute brewing among those of us working on it. I am the integrator, someone else has the installation contract.

10 identical machines in a room, hard wired (as opposed to plug-in). Each one has a 480V motor and a 480V 1 phase heater in it, plus various and sundry sensor circuits. Because of the heat, all of the controls for each machine go to a central control panel just outside the door of this room, housing the motor starters, breakers, contactors for the heaters, PLC, HMI and power supply. I designed in a Main Fusible Disconnect with a flange mount handle (to meet NFPA79 just in case). That is the only disconnect accessible from the outside of the box. Each motor starter and heater circuit has it's own breaker, but they are not going through the enclosure door.

The issue in dispute is over LOTO disconnects for each machine. The electrician installing it seems to believe that each motor and each heater in each machine must have a separate local LOTO disconnect, because they are not "within sight" of the control panel that has the lockable main disconnect.

My viewpoint is this; if the owner accepts that when someone wants to work on one machine ALL 10 WILL HAVE TO BE LOCKED OFF, then procedurally, having the one LOTO disconnect on the main is OK as long as he also labels each machine properly (proper label wording is another sub-issue I'll need to address).

Normally I wouldn't sweat it and just go along with the disconnects, we all know that's the safest way. But the big bone of contention here is that there are over 100 machines in the project, so 2 additional LOTO disconnects on each machine, no matter how cheaply I can get them, are going to amount to a big installed cost item and since it is already way over budget it is potentially a project killer!

The idea of using the breakers in the control panel has come up, but having 20 through-the-door rotary disconnect handles is extremely unpleasant to me. I've been the poor schmuck who has had to try to get a bunch of handles like that all lined up to close the door, I wouldn't wish that (or design that) on anyone. LOTO clips for the breakers inside the box was another option offered up, but I contend that Arc Flash rules now make that very impractical, if not impossible to implement; someone would have to don the PPE "bunny suit" just to install and then again remove any padlock and we all know THAT is not going to happen a second time.

Thoughts?

 

[eric9822]

I agree with you, a single lockable disconnect that disconnects power to all the machines provides the required means to isolate power and apply a lock. If that is the only economical solution and the owner agrees to it then I don't see an issue from a compliance standpoint. From a practical standpoint I think the first time a single machine needs to be serviced there will be a lot of resistance to shutting down the other nine. This could cause people to take shortcuts they would not normally take if a more convenient method of lock out had been provided. I also agree that 20 rotary disconnects on a door would be a nightmare, I have panels that have multiple rotary disconnects and I curse the designer everytime I have to deal with it. Is it possible to install barriers inside the panel to conceal energized components and allow entry into the panel without having to wear bulky arc flash PPE? If so you could use LOTO clips on the breakers.

 

[augie47]

I'm a bit uneasy with the wording " if the owner accepts".... too many will "accept" if it saves a dollar.

For my own liability, I would not be comfortable unless they could demonstrate that the installation conformed to:

(b) In industrial installations, with written safety procedures, where conditions of maintenance and supervision ensure that only qualified persons service the equipment

and that they were familiar with NFPA70-E

 

[don_resqcapt19]

I think 430.102(A) requires a disconnect for each machine. The general rule requires a disconnect for each motor controller and an exception permits a single disconnect for multiple controllers that are part of a single machine. I don't see that exception as permitting a single disconnect for controllers for multiple machines. Also when that exception is used, both the disconnects and the controllers must be within sight of the machine.
What about door mounted rotary disconnect switches....actual switches, not door mounted operators for panel mounted breakers?

 

[Jraef]

What about door mounted rotary disconnect switches....actual switches, not door mounted operators for panel mounted breakers?

That means running power conductors to the door, something else I am loathe to do. I've never liked running 480V across a hinge, I've seen too many 120V circuits get damaged and arc against it, the extra umph behind a 480V circuit makes me more nervous. But along the same line, maybe I'll arrange to put those on an interior false front, then put a window over the whole group. I'll have to think about that W.R.T. 70E though. I've done things like that before, but not since 70E came up.

Good points all, thanks for the feedback. I'm going to look at 430.102A in more detail.

 

[zog]

Is it possible to install barriers inside the panel to conceal energized components and allow entry into the panel without having to wear bulky arc flash PPE? If so you could use LOTO clips on the breakers.

Barriers would not remove the need from wearing arc flash PPE.

 

[eric9822]

Barriers would not remove the need from wearing arc flash PPE.

Why not? If nothing is exposed why would PPE be required?

 

[zog]

Why not? If nothing is exposed why would PPE be required?

No shock protection PPE is required but arc flash PPE is still required as long as the equipment is energized and you are interacting with it, exposed live parts or no exposed live parts, it does not make a difference unless the enclosure is arc flash rated.

The 70E is very clear about this issue. You may want to look at it again.

 

[mtfallsmikey]

That means running power conductors to the door, something else I am loathe to do. I've never liked running 480V across a hinge, I've seen too many 120V circuits get damaged and arc against it, the extra umph behind a 480V circuit makes me more nervous. But along the same line, maybe I'll arrange to put those on an interior false front, then put a window over the whole group. I'll have to think about that W.R.T. 70E though. I've done things like that before, but not since 70E came up.

Good points all, thanks for the feedback. I'm going to look at 430.102A in more detail.

That would make for a good arc blast/flash video!

 

[foqnc]

What about some kind of 'key switch' by the machine, which the operator/maintenance can remove when working on the machine, and wire the switch to kill the control power for that machine. When done he plugs the Key back in. Each key would need to be different for each machine.
It is not an ideal setup, but would prevent having to shut down all 10 machines at the same time.

 

[steve66]

Can't you get the control panel down to a Cat 0 or a Cat 1?? What's the largest HP motor you have?

As far as running the 480 across the hinges, I wouldn't have a problem if the wires have already gone through the branch circuit OCP, assuming we are talking about fairly small motors.

Or is there any way to run the 480V wires to the door sleeved in sealtight??

Steve

 

[Jraef]

The final decision was to go with disconnects at each machine after all. I got them to changed the heaters to 277V so I could get it down to one power wire, now I am having them use a 4 pole disconnect with a Neutral terminal for the heater.

Thanks for all the input everyone.

 

[skeshesh]

I'm happy with your final response since I do believe that the code leaves little wiggle room for your situation. That being said, if coordination studies have been conducted and a proper arc flash study (small note on "proper": I work under a very knowledgeable engineer that has shown me that many engineers conducting arc flash studies ignore contributing factors that in many cases requires the engineer to tool with the software to adjust for true levels of incident energy) have been a part of the project I see little technical value in this approach if the owner is willing to shut down all relevant machines. Personally, during design I show an individual switch for each machine in early submittals, so by the time the final submission is at hand the owner and his people have seen the design and cost estimates several times. If they have a problem with it they bring it up and we sit down for a conversation. Else the contractor bids accordingly so even if there is a question about eliminating some devices in the future during construction it doesn't effect the cost scope of the project.