NEC 517 violation

[mbrooke]

Why would code reject this?

 

[mbrooke]

But be ok with this?

 

[roger]

Why would an engineer or designer submit such sorry prints for bids or comments? :roll:

Roger

 

[Russs57]

No division of service between essential, life safety, critical, and equipment branches?

 

[mbrooke]

Why would an engineer or designer submit such sorry prints for bids or comments? :roll:

Roger

They are mine, rough mouse sketches, drawn up in an hour. Mostly for discussion purposes before going "official"

 

[mbrooke]

No division of service between essential, life safety, critical, and equipment branches?

Bingo.

But- why would you need division if two independent systems are going to be in place? Failure of either will not effect the other.

 

[topgone]

Lemme guess:
Because the electrical industry is still practicing "segregation"!

 

[mbrooke]

Lemme guess:
Because the electrical industry is still practicing "segregation"!

Codes reasoning, as best as I can figure is that should fault happen on one branch it does not effect that other. However, by having two separate systems as seen here no common mode or simultaneous failure can take place either. Yet 517 wants me to believe this does not meet the practical safeguards of life and property.

 

[romex jockey]

Is there a specific 517 ref for this?

~RJ~

 

[steve66]

Bingo.

But- why would you need division if two independent systems are going to be in place? Failure of either will not effect the other.

But failure of half the receptacles or lights in an OR is still a big issue. Doctors have to take time figuring out where to move plugs, and how to work with partial lighting.

(One OR once lost emergency power, but still had power to all normal power receptacles. Of course everything was plugged into the EM receptacles, so the doctors never even realized that the white receptacles still had power. )

Add that to the fact that large HVAC equipment and elevators are much more likely to fault, or to have large current surges, which in turn could trip an upstream breaker. And all in all, the less equipment and wiring you have on the output of an ATS, the more reliable the system is going to be.

However, I do agree that with all the newer requirements for selective coordination, this shouldn't be the issue it used to be. If an AHU or condenser trips a circuit breaker, it theory it would only be the branch breaker serving that unit.

However, there are still a lot of older systems that may not be coordinated because they were installed before the NEC requirements were added.

 

[mbrooke]

But failure of half the receptacles or lights in an OR is still a big issue. Doctors have to take time figuring out where to move plugs, and how to work with partial lighting.

(One OR once lost emergency power, but still had power to all normal power receptacles. Of course everything was plugged into the EM receptacles, so the doctors never even realized that the white receptacles still had power. )

How does the current code address this issue?

 

[steve66]

How does the current code address this issue?

IMO, the code only addresses the issue indirectly. It does this by allowing the configuration in your second post, but not the one in the first post.

The less you have on a critical or life safety branch, the less likely a fault will occur and take out the entire system.

I think it would be a good idea to give doctors and staff drills where power outages are simulated. But I don't know of any facilities that do that, and its obviously outside the scope of the NEC or any building codes.

 

[mbrooke]

IMO, the code only addresses the issue indirectly. It does this by allowing the configuration in your second post, but not the one in the first post.

It addresses it by mandating a certain percentage of receptacles be on the normal branch or a second critical branch. The idea is that if critical branch failed with utility power available, doctors and nurses can still transfer everything to the normal branch.

The less you have on a critical or life safety branch, the less likely a fault will occur and take out the entire system.

Not if you have selective coordination I'd think.

I think it would be a good idea to give doctors and staff drills where power outages are simulated. But I don't know of any facilities that do that, and its obviously outside the scope of the NEC or any building codes.

I know ICUs in some hospital practice blackout / med gas out protocols where they run down what to do and how to go about mechanically ventilating patients.

 

[Russs57]

Mbrooke, I'll admit to not knowing why you are asking questions of the good members of this forum that you already know the answers to. You come across as wanting to "reinvent the wheel" and /or proving that NEC/AHJ/NFPA are wrong and you are right. Good luck with all of that.

To me, you seem more interested in the cheapest solution. From where I sit, as a maintenance engineer in a hospital, I would think your "due diligence" should be aimed at providing a system most likely to insure delivery of power in the face of the worst calamity. Perhaps I have an unrealistic expectation or have misread your intent? Regardless, on to the designs you have proposed.

In your first design you have a 4,000 amp main feeding a pair of 2,500 breakers. IIRC all three of these breakers will require ground fault protection on them. You are not required to have ground fault protection on the 2,500 amp generator breakers. You may NOT have ground fault protection on the load side of any ATS so none of the breakers in the East or West switchgear may have that protection. Now what happens if you have a serious ground fault (call it a bolted connection) on the HVAC load on the East/Orange branch. At best you only trip the 2,500 amp breaker feeding normal power to the East/Orange switchgear (my experience has shown you will trip all normal power breakers). So now your East generator starts, ATS transfers, and you re-energize your bolted fault on your HVAC load. What happens? You might like to think the generator delivers nominal voltage to all loads until the 400 amp HVAC breaker clears the fault (and that severe under voltage to sensitve medical equipment will cause no harm). You might like to think nothing could explode or ignite before that 400 amp breaker trips. Remember that 400 amp breaker will carry 2,000 to 4,000 amps for a cycle or two at least. If impedance of wire limits current at the fault it may take a good deal longer to clear. Next thing you know that generator breaker trips and you just lost half of your life safety, critical, and equipment power to your entire facility. Don't think individual HVAC OPC devices will save you when some idiot just core drilled through your HVAC feeder. Only you can decide what design choices you can live and sleep with.

If you want to please me simply replace the purple and orange panels feeding the OR iso panels with breakers, have said breakers feed an ATS, have that ATS feed the OR iso panels. Now each iso panel has normal and emergency as an option on each pole of that ATS.

I'll stop now as I'm doubting you care what I think. If I'm wrong I'll be glad to comment on what I see as weak points in design two.

 

[topgone]

Mbrooke, I'll admit to not knowing why you are asking questions of the good members of this forum that you already know the answers to. You come across as wanting to "reinvent the wheel" and /or proving that NEC/AHJ/NFPA are wrong and you are right. Good luck with all of that.

To me, you seem more interested in the cheapest solution. From where I sit, as a maintenance engineer in a hospital, I would think your "due diligence" should be aimed at providing a system most likely to insure delivery of power in the face of the worst calamity. Perhaps I have an unrealistic expectation or have misread your intent? Regardless, on to the designs you have proposed.

In your first design you have a 4,000 amp main feeding a pair of 2,500 breakers. IIRC all three of these breakers will require ground fault protection on them. You are not required to have ground fault protection on the 2,500 amp generator breakers. You may NOT have ground fault protection on the load side of any ATS so none of the breakers in the East or West switchgear may have that protection. Now what happens if you have a serious ground fault (call it a bolted connection) on the HVAC load on the East/Orange branch. At best you only trip the 2,500 amp breaker feeding normal power to the East/Orange switchgear (my experience has shown you will trip all normal power breakers). So now your East generator starts, ATS transfers, and you re-energize your bolted fault on your HVAC load. What happens? You might like to think the generator delivers nominal voltage to all loads until the 400 amp HVAC breaker clears the fault (and that severe under voltage to sensitve medical equipment will cause no harm). You might like to think nothing could explode or ignite before that 400 amp breaker trips. Remember that 400 amp breaker will carry 2,000 to 4,000 amps for a cycle or two at least. If impedance of wire limits current at the fault it may take a good deal longer to clear. Next thing you know that generator breaker trips and you just lost half of your life safety, critical, and equipment power to your entire facility. Don't think individual HVAC OPC devices will save you when some idiot just core drilled through your HVAC feeder. Only you can decide what design choices you can live and sleep with.

If you want to please me simply replace the purple and orange panels feeding the OR iso panels with breakers, have said breakers feed an ATS, have that ATS feed the OR iso panels. Now each iso panel has normal and emergency as an option on each pole of that ATS.

I'll stop now as I'm doubting you care what I think. If I'm wrong I'll be glad to comment on what I see as weak points in design two.

Yeah, you're right! As much as possible, we design things to fit every requirement our clients need, We must be careful to follow what the regs stipulate. It must be made clear that the regs are "minimum" requirements also. But, a more careful investigation on every design could be best served if designers listen to others' point of view, IMHO. Just saying though!:thumbsup:

 

[mbrooke]

Mbrooke, I'll admit to not knowing why you are asking questions of the good members of this forum that you already know the answers to. You come across as wanting to "reinvent the wheel" and /or proving that NEC/AHJ/NFPA are wrong and you are right. Good luck with all of that.

Perhaps more of wanting to know the reasoning of the code. I can think of many designs which are against code but function the same of not better. Of course, I am open to being proven wrong which is why I ask.

To me, you seem more interested in the cheapest solution. From where I sit, as a maintenance engineer in a hospital, I would think your "due diligence" should be aimed at providing a system most likely to insure delivery of power in the face of the worst calamity. Perhaps I have an unrealistic expectation or have misread your intent? Regardless, on to the designs you have proposed.

Honestly, yes. The cheapest solution means more money can be spent on things like gensets for the same budget as but one example. IMO not just hardening, but diversity also assures good outcomes in the face of calamity.

In your first design you have a 4,000 amp main feeding a pair of 2,500 breakers. IIRC all three of these breakers will require ground fault protection on them. You are not required to have ground fault protection on the 2,500 amp generator breakers. You may NOT have ground fault protection on the load side of any ATS so none of the breakers in the East or West switchgear may have that protection. Now what happens if you have a serious ground fault (call it a bolted connection) on the HVAC load on the East/Orange branch. At best you only trip the 2,500 amp breaker feeding normal power to the East/Orange switchgear (my experience has shown you will trip all normal power breakers).

Not if you coordinate the ground fault. The 400amp HVAC fuse should trip first.

Curiosity: what would happen if you had no ground fault protection?

So now your East generator starts, ATS transfers, and you re-energize your bolted fault on your HVAC load. What happens? You might like to think the generator delivers nominal voltage to all loads until the 400 amp HVAC breaker clears the fault (and that severe under voltage to sensitve medical equipment will cause no harm).

If medical equipment can survive a 10 second outage, why can't it survive a voltage dip? I mean what happens if the utility voltage dips for several seconds?

You might like to think nothing could explode or ignite before that 400 amp breaker trips. Remember that 400 amp breaker will carry 2,000 to 4,000 amps for a cycle or two at least. If impedance of wire limits current at the fault it may take a good deal longer to clear. Next thing you know that generator breaker trips and you just lost half of your life safety, critical, and equipment power to your entire facility. Don't think individual HVAC OPC devices will save you when some idiot just core drilled through your HVAC feeder. Only you can decide what design choices you can live and sleep with.

I'd think the 400amp HVAC fuse will blow before the generator's fuse or breaker if coordinated correctly.

If you want to please me simply replace the purple and orange panels feeding the OR iso panels with breakers, have said breakers feed an ATS, have that ATS feed the OR iso panels. Now each iso panel has normal and emergency as an option on each pole of that ATS.

I'll stop now as I'm doubting you care what I think. If I'm wrong I'll be glad to comment on what I see as weak points in design two.

[/QUOTE]


Are you kidding me? Of course I care, hence why I ask I am open to critique.

Can I ask... why breakers?

The 225 amp panels have breakers in them.

 

[Russs57]

Design two critique and/or suggestions for improvement. Why I share your viewpoint about some codes (not to mention AHJ), I feel we have no choice but to adhere to them.

First, I am accustomed to seeing a “non-interruptible type” POCO supply to a hospital. So give me at least two POCO feeders with two transformers and two POCO ATS’s. Any one feeder and transformer most be capable of carrying the entire connected load.

Moving on to the main switchgear, I’d suggest breaking it down into two 2,000 amp sections. Each section shall have two main breakers (from the two different POCO transformers), and a tie breaker. Call the mains 1A, 1B and the tie 1C. Give me analog panel meters with a rotary switch so I can quickly see all load voltages and currents on breakers 1A and 1B. Make breakers 1A, 1B, and 1C draw out style. Needless to say 1A and 1B shall each be capable of supplying the entire connected load. The other section would be the same with main breakers 2A, 2B and tie breaker 2C. You will need ground fault protection on those breakers. I would not want fuses. At this point you could keep your pair of 800 amp breakers feeding NDP and SDP. They will need ground fault protection. Personally I’d give consideration to doing away with NDP and SDP and moving all loads back to the main switchgear. This way a ground fault on an 800 amp breaker isn’t taking down half my facility. Sure that is ten circuits that have to be brought all the way back and the switchgear has to have more breaker spaces. But 200-250 amp MCCB’s have about 1/3 the failure rate of the larger 800 amp breakers. Plus a ground fault on any of the 200-250 amp breakers is a much more tolerable event.

On the essential side I would require two generators, each capable of carrying the entire connected load. The generator parallel switchgear shall have generator main breakers and a tie breaker. Those shall be draw out style. Individual breakers for the ATS’s can be MCCB’s but thought should be given to ease of replacement. I would not have ground fault protection at this level. Provisions shall be made to run with either a split bus or generators in parallel. Necessary controls shall be in place to automatically and manually parallel the generators. An indicator panel shall be in place at this switchgear showing the position of each ATS. I’d prefer incandescent lamps switched thru ATS cam switches using the generator battery. The indicator panel should have a “push to test” feature so all lamps can be tested. I would want a rotary switch so I can pick which generator battery is being used for the ATS generator run command. I can’t tell you the number of times I have seen batteries changed out with no thought to disabling every ATS in a facility. Oh, and I want dual batteries and dual starters on each generator.

I would want my ATS’s to be as close to the load as possible. I would rather see you run a pair of normal power feeders to switchgear near the critical emergency power loads (i.e. OR rooms). I’d like that switchgear to have one main from section 1, one main from section 2, and tie breaker. Make them draw out style breakers. Panel meters please. As near as possible to that gear I’d want an essential power board with dual feeders from the generator parallel board. Same setup, draw out breakers with a tie breaker. By dual feeders it is assumed one feeder comes from section 1 (or generator 1) and the other comes from section 2 (or generator 2). The intent is that upon a notification of loss of power that one guy could quickly see it he has lost a normal power feeder or an emergency power feeder. He would be able to see this and restore power in mere seconds by opening a source breaker and closing a tie breaker. Then he can head to the main switchgear/parallel gear and reset a breaker. I would want all my breakers feeding individual ATS’s to have ground fault protection. IMHO I’d rather have stand alone ground fault modules and shunt trip coils. It is a lot less costly to have a couple of modules and coils in stock than a bunch of LSIG breakers. By the same token make the draw out breakers the same frame size so they can be used in multiple locations simply by adjusting trip settings. I’d want a bunch of little ATS’s in the same room as the essential gear, as in one for each and every iso, life safety, and critical panel. It would be nice for the equipment loads as well, especially the elevators and MCC. The one place I would use fuses is at the end of the line, like before every motor. I would use NEMA rated motor starters with electronic overload relays. It cuts down on the need to stock heater elements and guys putting the wrong size ones in.

So there you are! I could spend a lot more of the customers money and require 1,200 RPM prime rated gen sets with 96 hours of fuel (and of course I’d want dual tanks and dual STP’s on each tank and four separate pairs of fuel lines). And of course, given my name, I favor Russelectric gear. No doubt Siemens will ruin my stuff.

 

[mbrooke]

I'll type more, but three questions:

1) Why avoid fuses?

2) While code, what would happen without ground fault protection?

3) Why paralleling emergency switch-gear? I see it as expensive and less reliable. A gimmick if you will.

 

[Russs57]

Take a quick look at the following links. I am making the assumption that fused mains equates to "bolted pressure contact switch". In certain applications, say an industrial plant where planned outages are not a problem and you have a qualified staff, I think they are the right choice. Just not for a hospital with staff that isn't as knowledgeable and an administration bent on constant 24/7/365 operation.

For example, are you aware that at least one manufacturer of bolted pressure contact switches requires that if it ever opens a fault it may not be returned to service without maintenance? Pretty much all of them require annual maintenance. That is going to require a complete shutdown of all POCO feeders in your designs. Good luck on getting any hospital to obey that.

http://apps.geindustrial.com/publib...OARDS?TNR=White Papers|GEAR-OR-BOARDS|generic

https://www.munichre.com/site/hsb/g...News/In-The-News/2016/main-switch-failure.pdf

Circuit breakers offer much longer service intervals. You don't have guys hunting for the proper fuse and needing to know where a torque wrench is. I have seen folks harm these switches and themselves but not knowing how to charge springs. Some people get in a downright panic mode and just can't think clearly. Keep in mind these switches are good for thousands of no load operations but relativity few full load operations. If they have ground fault protection they will be subjected to more full load breaks than is desired. Staff might not be knowledgeable enough to understand the need to open all downstream devices before closing. I have seen them fail and it isn't pretty. They tend to be "custom built" so it takes a lot of time for factory engineers to come in and re-engineer a solution or rebuild them on-site. Throw in portable generators, interruption of service cost, etc, and you are going to be looking at nearly a million bucks+ per incident. I'll gladly take the draw out breaker that I can talk most any idiot thru changing out.

Paralleling generator gear allows me the means to effectively double the capacity of my generators on loads that can be a delayed transfer, such as the equipment branch. I would never sign off on a single generator so the ability to parallel isn't that much of an extra cost.

You might be able to avoid ground fault requirements under the "less than 150 volts to ground" loophole but I don't think this is in the best interest of anyone. Well maybe it is in the best interest of bean counters that think catastrophic faults never happen and people are expendable commodities.

I mean my God people. EE's design far more robust electrical system for data and telecommunications sites than they do hospitals. Medical errors are now the third leading cause of death and it isn't just heart attacks and bacteria. System failures are a part of this. Damn the cost and the minimal requirements of code. Do what is right and just don't even offer them less than what is right. If every EE did so the field would be even. Sadly, one of these days, almost every one of us will end up in a hospital and be dependent on its ability to keep us alive. Lets do our part to make sure they have what it needs in an electrical infrastructure to keep us alive.

 

[roger]

I mean my God people.

You should change it to "I mean my God OP".

Roger