Machine Safety

[Timbert]

Get yourself a copy of NFPA 79 (you can view it online for free at NFPA.org but you should get a copy as this guy is gonna become your best buddy). Chapter 6 Protection from Electric Shock is what you want to read.

6.2.3 Enclosure Interlocking. When required by 5.3.1.4, each disconnecting means mounted with or adjacent to a control enclosure that contains live parts operating at 50 volts ac (rms value) or 60 volts dc or more shall be mechanically or electrically interlocked, or both, with the control enclosure doors so that none of the doors open unless the power is disconnected. Interlocking shall be reactivated automatically when all the doors are closed.

Now, there is no defined you will use a x style limited switch, connected to y type disconnect device monitored by a z. There are just way too many possible solutions to this problem.

You could also look into IEC 62061, the international machine safety standard. This standard includes methods for analyzing how 'safe' a particular solution is.

I would also try to educate the customer on what is required i.e. at 48vdc this type of interlocking is not required. It will add expense and complexity with little additional safety benefit and might even lull them into a false sense of security. If you overuse a complex safety technique for extra low voltage, when someone opens up a cabinet with MV in it they will think "business as usual." Just an opinion, if the customer insists then go right ahead. But there are appropriate levels of safety. I would not use a trapped key to keep someone out of 48vdc.

 

[CEDEng]

Excellent - that's what I was looking for. For all the folks who said "there is no standard," - is this not what you thought I meant (I'm serious, not being a smarta$$, I realize it sounds snarky, but not intended as such)? And now that you know there is such a thing - do you implement it? Do you see it implemented?

OK - so it says the door cannot open unless the power is disconnected. And it's not just for above 600V, but 50V.

So, while the very act of opening the door can (somehow) cause power to disappear, a better bet (and more...to the letter...) would be that power is OFF before the door CAN be opened. (I give in to the fact that a rotary disconnect/lock is - technically - disconnecting the power before the door opens.)

So the Kirk Key plan (or similar) is really the winner and coverall. The power must be off to release the key that allows the door to open.

Does this all jive now? Sorry for the confusion, if I caused any - but thanks for the excellent info.

 

[just the cowboy]

original question

original question

In the op question you a re talking about removable panels or doors. Doors and panels are machine guards in my mind and would just use a door interlock safety switchs such as http://ab.rockwellautomation.com/Sensors-Switches/Safety-Interlock-Switches that prevent machine operation. For maintenance to work you need a lockout/tagout program. I have seen where maintenance can bypass a door switch for inspection of moving parts, by following a written policy but are not recommended.

 

[pfalcon]

It's the "ordinary" I am trying to define!

It's not actually >600V. I used that, because I want to secure 48V as if it were 600V. This is in response to a mandate from our largest customer - that all enclosure doors and panels (>48V) be interlocked and protected in the "standard" way. Which is escaping me! What standard? What way? ... !

Standard Practice in US >> NFPA79:2012:Chapter 6 Protection from Electrical Hazards

There are other available standards that will provide comparable answers; this one's in plain english.

You might want to start with:

When a qualified person, using appropriate work practices, needs to enter an enclosure that does not have a disconnect, one of the following conditions shall be met:
(1) The use of a key or tool shall be required for opening the enclosure.
(2) An enclosure door shall be permitted to be opened without the use of a key or a tool and without disconnection of live parts only when all live parts inside are separately enclosed or guarded such that there cannot be any direct contact with live parts.

 

[pfalcon]

Excellent - that's what I was looking for. For all the folks who said "there is no standard," - is this not what you thought I meant (I'm serious, not being a smarta$$, I realize it sounds snarky, but not intended as such)? And now that you know there is such a thing - do you implement it? Do you see it implemented?

A lot of these guys are residential rather than tied to a manufacturing floor so NFPA79 isn't commonly discussed. We use this standard extensively and demand it from our machine providers.

OK - so it says the door cannot open unless the power is disconnected. And it's not just for above 600V, but 50V.

Well, 6.2 says >50VAC or >60VDC

So, while the very act of opening the door can (somehow) cause power to disappear, a better bet (and more...to the letter...) would be that power is OFF before the door CAN be opened. (I give in to the fact that a rotary disconnect/lock is - technically - disconnecting the power before the door opens.)

If you're looking at a rotary disconnect then you need to look at NFPA79:2012:6.2.3 Enclosure Interlocking.
The disconnect is mechanically preventing the door from opening while in the ON position or will electronically kill the machine if the door is opened while in the ON position (pick 1).

Additionally to qualify for proper OSHA (NFPA70E) lockout requirements the lockout point must be located where the lock cannot be removed without disassembling the disconnect. So a rotary must go through the side wall or the lockout point must be on the inner disconnect mechanism - not the part that swings away with the door. Kits are sold for all major brands.

Additionally to qualify for proper NFPA79 troubleshooting requirements the disconnect must be easily operable. This would be during troubleshooting when the disconnect has been bypassed so you can take voltage readings. The guy has to be able to kill the power. Kits are sold for all major brands.

 

[pfalcon]

In the op question you a re talking about removable panels or doors. Doors and panels are machine guards in my mind and would just use a door interlock safety switchs such as http://ab.rockwellautomation.com/Sensors-Switches/Safety-Interlock-Switches that prevent machine operation. For maintenance to work you need a lockout/tagout program. I have seen where maintenance can bypass a door switch for inspection of moving parts, by following a written policy but are not recommended.

Interlock safety switches can be used for access doors and panels that are easily removeable - just not for the electrical enclosure itself. Those that require tools for disassembly do not require switches.

 

[CEDEng]

Thanks, again, for the feedback.

Regarding LOTO - I can provide the hardware - but that is just the first step in the (customer's) process. It is inadequate to protect the doors and panels in the manner required.

Regarding the standard - pretty clear-cut then, it seems - Power must be off to open the doors.

I find 6.2.3.1 interesting - "means shall be permitted to allow qualified people access without removing power" - what do you think they had in mind there?

 

[petersonra]

Thanks, again, for the feedback.

Regarding LOTO - I can provide the hardware - but that is just the first step in the (customer's) process. It is inadequate to protect the doors and panels in the manner required.

You need to understand the difference between a guard that is normally used by the operator in some fashion and a guard that is not used by the operator. If the operator has to interact with the guard in some way as part of the operation of the machine, it is appropriate to protect the operator in some way. One way might be to have some kind of safety rated switches on the guard that trip the estop when the guard is not in place.

However, just tripping the estop might not be appropriate either. If a guard can be opened while the machine is running in a way that might introduce a hazard to the operator (think a machine that cannot spin down fast enough to prevent moving parts presenting a hazard to an operator), just tripping the estop is not an adequate safety measure.

A guard that is not normally used by the operator (such as an access panel only used by maintenance personnel) would likely not require a switch of any kind.

Regarding the standard - pretty clear-cut then, it seems - Power must be off to open the doors.

I find 6.2.3.1 interesting - "means shall be permitted to allow qualified people access without removing power" - what do you think they had in mind there?

It is not required that you turn off power to open the door. Only that it be moderately inconvenient to do so, such as requiring a tool to do so. Disconnect switch or CB handles generally have a means by which they can be defeated with some kind of tool.

 

[iceworm]

... I find 6.2.3.1 interesting - "means shall be permitted to allow qualified people access without removing power" - what do you think they had in mind there?

You mean besides:

... Yes, us t-shooters, as well as being nefarious, are sneaky, underhanded, and devious. Yes we will get around it as needed. ...

Most all low voltage (<600V) door mounted disconnect switch operators have a screwdriver operated defeat that allows the door to be opened with the power still on.

ice

 

[iceworm]

... I want to secure 48V as if it were 600V. This is in response to a mandate from our largest customer - that all enclosure doors and panels (>48V) be interlocked and protected in the "standard" way. ...

I think you have the concepts correct. If the customer mandates "all >48V" then that is what you do. The third party specs (such as NFPA) are pretty meaningless - except as a guide for how to acomplish, a method to establish the "Standard".

ice

 

[CEDEng]

From my original post, "this is not referring to operator safety, or moving parts and guarding. I am referring to the "rear door" of large machines and panels."

So - while I do understand the importance and complexity of operator safety, that's a fish for a different barrel today.

I've worked with many disconnects that allow the power to be "cheated" once the door is open. I have not used one (with a screwdriver or otherwise) that allowed power to remain on without interruption while the door opens. I want one though!!

"A guard that is not normally used by the operator (such as an access panel only used by maintenance personnel) would likely not require a switch of any kind."

This seems to exactly contradict 6.2.3 (stated above a few posts...), no?

 

[petersonra]

I've worked with many disconnects that allow the power to be "cheated" once the door is open. I have not used one (with a screwdriver or otherwise) that allowed power to remain on without interruption while the door opens. I want one though!!

I have never seen one that does not allow you to open the door with the power on if you use a screw driver to operate the defeater mechanism while opening the door.

"A guard that is not normally used by the operator (such as an access panel only used by maintenance personnel) would likely not require a switch of any kind."

This seems to exactly contradict 6.2.3 (stated above a few posts...), no?

There is a difference between an electrical enclosure door and a machine access panel. A door is meant to be opened and thus if there are hazardous voltages to be found behind it, it should have a disconnect switch preventing easy access to the hazardous voltage. An access panel is meant to be opened only under unusual circumstances and requires more effort to do so, which tends to discourage the average untrained person from doing so.

 

[CEDEng]

Well, I think we've beat this horse into submission!

Thanks, all, for the excellent information that I always find on this forum.

And for your patience as I feel my way through strange territory.

Be safe,

C.

 

[pfalcon]

Sorry, was on vacation.

...
Regarding LOTO - I can provide the hardware - but that is just the first step in the (customer's) process. It is inadequate to protect the doors and panels in the manner required.
Regarding the standard - pretty clear-cut then, it seems - Power must be off to open the doors.
I find 6.2.3.1 interesting - "means shall be permitted to allow qualified people access without removing power" - what do you think they had in mind there?

6.2.3.1 is used to address the need to perform live troubleshooting within an electrical enclosure. A common tool is used to bypass the interlocked disconnect (screwdriver most common) so that the faulted state is not reset by cycling the disconnect.

I think you have the concepts correct. If the customer mandates "all >48V" then that is what you do. The third party specs (such as NFPA) are pretty meaningless - except as a guide for how to acomplish, a method to establish the "Standard".

ice

OSHA requires national consensus standards to be used in many areas including electrical for industrial machinery. "Third party specs" such as NFPA are a vital defense against OSHA penalties. Should you fail to adopt a standard for your company then OSHA will rule according to the standard of their choice - which might be very ugly for your company. For industrial machinery in the US that currently means they will presume compliance with NFPA79 unless you direct them to a different standard. In short, the standards are used by OSHA to determine "recognized hazards".

From my original post, "this is not referring to operator safety, or moving parts and guarding. I am referring to the "rear door" of large machines and panels."

So - while I do understand the importance and complexity of operator safety, that's a fish for a different barrel today.

I've worked with many disconnects that allow the power to be "cheated" once the door is open. I have not used one (with a screwdriver or otherwise) that allowed power to remain on without interruption while the door opens. I want one though!!

"A guard that is not normally used by the operator (such as an access panel only used by maintenance personnel) would likely not require a switch of any kind."

This seems to exactly contradict 6.2.3 (stated above a few posts...), no?

A number of things determine the requirements for guards. So from above:
Maintenance access only - no operator access.
A "rear door" implies this isn't access to an industrial enclosure but rather access to a machine envelope.

NFPA79:2012:6.2.2 pertains to classic industrial enclosures. That would help explain why 6.2.3 seemed contradictory. Different rules apply for accessing the envelope.
Generally access to the machine envelope requires one of two conditions:
(1) Access that requires tools to disassemble the panels.
-or-
(2) Safety interlocks that either:
(a) Hold the door shut until safe
-or-
(b) Bring the machine to a safe state before a person can be at risk.