DC Disconnect

[Canton]

Hello All,

I have a 600KVA UPS that we are installing. It is going to have (6) Remote battery cabinets about 15 feet away. Each battery cabinet is going to have a 200 amp factory installed disconnect (So there will be (6) DC Breakers in parallel feeding the DC bus in the UPS). There is an option to install a Main DC disconnect 1200 amp between the batteries and the UPS but they have chosen not to.

My question is this: Is it code compliant to not have a single DC breaker/Disconnect feeding the UPS? I see it as an accident waiting to happen. In order to work on the UPS you would have to open all (6) DC breakers to de-energize the DC input into the UPS.

Any thoughts?

Thanks!

 

[dereckbc]

I have installed many a medium and large UPS and have only used disconnects on just a few jobs where the local jurisdiction demanded they be used.

Really bad things happen when you disconnect large battery banks, or I should say when you reconnect. Unless the battery bank you reconnect is equalized to the rest of the batteries in service you are going to burn up your disconnect. You will weld the contacts together. To reconnect you have to have a portable rectifier/charger to bring up the battery voltage to match the plant voltage. A disconnect current rating of say 200 amps is meaningless as the EQ current can run into the thousands of amps supplied by the other batteries as there is nothing to limit current other than multiple pairs of 750 MCM cable connecting them together.

You can also damage the disconnects when opening if there is a load current flowing from the batteries as DC will pull an arc where as AC does not because of the Zero Crossover Sequence.

So for these reasons we do not use disconnects unless it is required by the operator or AHJ.

 

[GoldDigger]

Any thoughts?

Thanks!

As a practical matter, disconnecting one battery bank out of six under load will not have a great effect on the other banks.
But if you have only, say, two of the six connected, they will be supplying more current than they are designed for, and when their individual breakers trip you have lost your backup power.

Ideally the individual bank disconnect will only be operated when the UPS is not drawing power from the battery bank, but you may not have that luxury for a dual conversion, always on UPS.
Otherwise you could operate the bypass and then deal with the unloaded battery bus.
It is true that, except for a fault isolation incident, it may not critical to have a single disconnect switch (or single disconnect breaker).

 

[Canton]

I have installed many a medium and large UPS and have only used disconnects on just a few jobs where the local jurisdiction demanded they be used.

Really bad things happen when you disconnect large battery banks, or I should say when you reconnect. Unless the battery bank you reconnect is equalized to the rest of the batteries in service you are going to burn up your disconnect. You will weld the contacts together. To reconnect you have to have a portable rectifier/charger to bring up the battery voltage to match the plant voltage. A disconnect current rating of say 200 amps is meaningless as the EQ current can run into the thousands of amps supplied by the other batteries as there is nothing to limit current other than multiple pairs of 750 MCM cable connecting them together.

You can also damage the disconnects when opening if there is a load current flowing from the batteries as DC will pull an arc where as AC does not because of the Zero Crossover Sequence.

So for these reasons we do not use disconnects unless it is required by the operator or AHJ.

Thank you for your response. Let me clarify a few things. The "disconnects" I said are really DC rated 3 pole breakers, capable of disconnecting under load. (Not a knife switch or other standard disconnect). They are also shunt trip breakers and could be opened in the event of EPO activation or fire alarm.

My concern is there are (6) breakers to open if you want to de-energize the DC bus in the UPS to work on it safely. There was an option to have a 1200 amp DC rated 3 pole breaker installed between the battery cabinets and the UPS (shuntable also) so you would only have to open one breaker to kill the DC bus. As it stands now you have to open all (6) DC breakers, (1) Main UPS Input, and (1) By-Pass Input....(8) breakers in total to work on the UPS de-energized.

Should there not be (1) DC breaker to de-energize all DC input energy?

Thanks

 

[Canton]

As a practical matter, disconnecting one battery bank out of six under load will not have a great effect on the other banks.
But if you have only, say, two of the six connected, they will be supplying more current than they are designed for, and when their individual breakers trip you have lost your backup power.

Ideally the individual bank disconnect will only be operated when the UPS is not drawing power from the battery bank, but you may not have that luxury for a dual conversion, always on UPS.
Otherwise you could operate the bypass and then deal with the unloaded battery bus.
It is true that, except for a fault isolation incident, it may not critical to have a single disconnect switch (or single disconnect breaker).

This UPS is a dual Input model. The expansion with the (6) battery cabinets was for increased utility outage time to protect the critical load and for future increase of the critical loads. This UPS can have the batteries disconnected while running.....but yes the critical load will not be protected by the DC supply....it would then look for the By-pass input.....but the By-pass input and Main input are from the same source (because they must be synced), so you would loose the load.


My only concern was from a safety standpoint of having one breaker to disconnect the DC input for maintenance....not (6) breakers as is the scenario now.

Thank you for any input and or code references (if any)!

 

[iceworm]

Slow poster - did not see post 5

... My concern is there are (6) breakers to open if you want to de-energize the DC bus in the UPS to work on it safely.

... Should there not be (1) DC breaker to de-energize all DC input energy?

Give us some context. As in:

What is driving this?
Is there some reason you think there needs to be an additional CB to isolate all six banks?​

Are you thinking:​

A regulatory issue?
Non-regulatory safety issue?
An operational/design issue?
​

Added after seeing post 5
Expand on the safety issue. Regulatory, Non-regulatory? Why is One main safer?

I'm not against what you are advocating. I'm curious why you think it is better/safer.


ice

 

[Canton]

Slow poster - did not see post 5


Give us some context. As in:

What is driving this?
Is there some reason you think there needs to be an additional CB to isolate all six banks?​

Are you thinking:​

A regulatory issue?
Non-regulatory safety issue?
An operational/design issue?
​

Added after seeing post 5
Expand on the safety issue. Regulatory, Non-regulatory? Why is One main safer?

I'm not against what you are advocating. I'm curious why you think it is better/safer.


ice

Just from the stand point that someone might only turn off 5 of the 6 disconnects. Or they may go to the first cabinet and see the first disconnect and assume it is the Main for all (6). What I was really looking for was a Code Requirement that would force us to put one in....as always everything is about $$$ and safety comes second.

Thanks for your opinion on anything you can offer!

 

[iceworm]

Just from the stand point that someone might only turn off 5 of the 6 disconnects. Or they may go to the first cabinet and see the first disconnect and assume it is the Main for all (6). ...

I suspect you have a tough sell. Essentially you have to convince the management that:

y'all (not just you) are untrained and not knowledgable on the system. The crews are being sent out to work on the system with no knowledge. The energy isolation must be readily apparent to one without access to system configuration information. There is no work permit system insisting on insuring/verifing energy isolation prior to starting work.
​

I'm not saying this is true - just what it sounds like from here.

This is wrong on a whole lot of levels. Management for not insisting on training concerning energy isolation. The crews for going to work with out specific knowledge to insure energy isolation. What you are describing - you guys aren't qualified. Before you get upset with me - I'm not saying you guys are not knowledgable about your craft, I'm certain you are. Rather you are not trained up on the specific equipment isolation requirements.

Again, I am neither for nor against the additional disconnect. It appears to be an operational/design issue.. It could even be a load break 1200A disconnect? That is a question - not a statement. I don't know if 1200A DC load break disconnects are readily available.

As for the disconnect, the biggest difference I see is there are 3 keys/tags in the lock box instead of eight. That does not appear to be as important as the crew having the knowledge necessary for insuring energy isolation.

Good luck on getting this resolved. I am never in favor of the crews not being fully knowledgable.

ice

 

[Canton]

I suspect you have a tough sell. Essentially you have to convince the management that:

y'all (not just you) are untrained and not knowledgable on the system. The crews are being sent out to work on the system with no knowledge. The energy isolation must be readily apparent to one without access to system configuration information. There is no work permit system insisting on insuring/verifing energy isolation prior to starting work.
​

I'm not saying this is true - just what it sounds like from here.

This is wrong on a whole lot of levels. Management for not insisting on training concerning energy isolation. The crews for going to work with out specific knowledge to insure energy isolation. What you are describing - you guys aren't qualified. Before you get upset with me - I'm not saying you guys are not knowledgable about your craft, I'm certain you are. Rather you are not trained up on the specific equipment isolation requirements.

Again, I am neither for nor against the additional disconnect. It appears to be an operational/design issue.. It could even be a load break 1200A disconnect? That is a question - not a statement. I don't know if 1200A DC load break disconnects are readily available.

As for the disconnect, the biggest difference I see is there are 3 keys/tags in the lock box instead of eight. That does not appear to be as important as the crew having the knowledge necessary for insuring energy isolation.

Good luck on getting this resolved. I am never in favor of the crews not being fully knowledgable.

ice

Wow...Well....I have to say you are a little off here on your analysis of the situation. All the Electricians at this government facility carry a Masters License and are knowledgeable and qualified to work on UPS systems. There are many UPS systems in this facility that we work on. Every single one is different...from single or dual input, maintenance by-passes or other multiple redundancies built in. This one, the one in question is the odd ball without the Main DC disconnect unlike the others, thus the reasoning for my questioning.

Even with a Energy Isolation Plan, and LOTO procedure in place accidents can still happen. If we can mitigate any of these by incorporating another disconnect in there...why not.

I don't understand how you can process the information and the question of an additional Main DC battery breaker/disconnect for a UPS system, and come to the conclusion that the crews installing it are untrained and unqualified and not knowledgeable....???

If you have nothing of value to offer, then do not respond to the Question.

 

[Smart $]

Thank you for your response. Let me clarify a few things. The "disconnects" I said are really DC rated 3 pole breakers, capable of disconnecting under load. (Not a knife switch or other standard disconnect). They are also shunt trip breakers and could be opened in the event of EPO activation or fire alarm.

My concern is there are (6) breakers to open if you want to de-energize the DC bus in the UPS to work on it safely. There was an option to have a 1200 amp DC rated 3 pole breaker installed between the battery cabinets and the UPS (shuntable also) so you would only have to open one breaker to kill the DC bus. As it stands now you have to open all (6) DC breakers, (1) Main UPS Input, and (1) By-Pass Input....(8) breakers in total to work on the UPS de-energized.

Should there not be (1) DC breaker to de-energize all DC input energy?

Thanks

Just from the stand point that someone might only turn off 5 of the 6 disconnects. Or they may go to the first cabinet and see the first disconnect and assume it is the Main for all (6). What I was really looking for was a Code Requirement that would force us to put one in....as always everything is about $$$ and safety comes second.

Thanks for your opinion on anything you can offer!

...

Even with a Energy Isolation Plan, and LOTO procedure in place accidents can still happen. If we can mitigate any of these by incorporating another disconnect in there...why not.

...

First, I know of no requirement for a single disconnecting means of the DC supply. Just to be certain, I tried to find one and could not.

While I understand your concern, I do not believe one DC disconnect will reduce accident potential to any substantial degree. It also adds another level of complexity to the system. The one advantage I do see is disconnecting all six banks simultaneously vs. one at a time. That should only make a difference when disconnecting under load... which shouldn't occur when safe switching procedures are documented and adhered to.

With all six disconnects being shunt trip, what about tripping them all simultaneously through an auxiliary contact in the Main UPS input disconnect? Designing UPS is not my forte, but I cannot think of any reason to maintain battery connection when the AC input is disconnected (other than testing purposes, during which the shunt trip circuit can be defeated).

 

[iceworm]

Wow...Well....I have to say you are a little off here on your analysis of the situation. All the Electricians at this government facility carry a Masters License and are knowledgeable and qualified to work on UPS systems. ....

Of course I'm off on the analysis. Look at what you said:

Just from the stand point that someone might only turn off 5 of the 6 disconnects. Or they may go to the first cabinet and see the first disconnect and assume it is the Main for all (6). ...

Does this meet the definition of qualified? Having a Masters License does not make one qualified. Having knowledge of the internal workings of the UPS does not make one qualified. I think you guys are highly qualified. I suspect you are using this as a straw argument to get help with a manager that does not want to spend the money.

...Even with a Energy Isolation Plan, and LOTO procedure in place accidents can still happen. If we can mitigate any of these by incorporating another disconnect in there...why not..

Every single one is different...from single or dual input, maintenance by-passes or other multiple redundancies built in. This one, the one in question is the odd ball without the Main DC disconnect unlike the others, ...

Okay - then tell your management you want the disconnect so that this one is less of an oddball. You wish to mittigate the accidents that can happen even with EI and LOT procedures in place. With two additional isolation points, you probably can't even use the Single Source isolation procedures.

... I don't understand how you can process the information and the question of an additional Main DC battery breaker/disconnect for a UPS system, and come to the conclusion that the crews installing it are untrained and unqualified and not knowledgeable. ....

Actually what I said was, " Rather you are not trained up on the specific equipment isolation requirements." I meant that in an NFPA 70E definition of "qualified " sense - cause that is essentially what you said, "someone might only turn off 5 of the 6 disconnects ...."

I'm not telling you y'all are dumb. I'm not telling you y'all are not excellect at your craft. I'm telling you that you need a better argument. Half of what I do is to get the techs the installation they want/need to do their jobs easier. Typically they start out with, "I don't like this ... Here's what I want .... I think it is a code violation to ... " Typically that doesn't pry any money loose. Short of a regulatory issue, Faster, Safer, less money over life, less maintenance over life gets me the money. And I would never start out with, "We might not find the right breaker ...."

If you have nothing of value to offer, then do not respond to the Question.

I don't have anything to offer that you want to hear and you are correct that makes it of no value. Per your direction, I'll stop. I wish I could have been of help.

ice

 

[dereckbc]

Thank you for your response. Let me clarify a few things. The "disconnects" I said are really DC rated 3 pole breakers, capable of disconnecting under load.

Disconnecting under load is not the problem, it is reconnecting a single string that has been removed for maintenance or what ever reason that is the problem.

OK you said an EPO is required? Is that correct?

If that is a true statement and I were designing the system I would only use 1 Main Disconnect to take all 6 strings off line. As I said the problem with Disconnects is not a disconnect, it is a reconnect. Having a disconnect makes it too easy to remove a battery string off-line, be forgotten about, then when noticed too easy to close back in and destroy the disconnect because of the huge amount of EQ current which cannot be controlled when you close it back in.

FWIW I have worked with Telecom sector for 35 years and my specialty is battery plants, UPS, AC distribution, Emergency, and Power Protection. We avoid battery disconnects like the plague as it greatly increases the chances for an outage and equipment damage from operator error.

If you are in a plant with trained power plant electricians, disconnecting a batery string for service should be no issue for experienced personnel as all they have to do is remove the jumpers from the Charge Buss, and they had better know to use a rectifier/charger to bring the disconnected battery back up to plant operating voltage before reconnecting the jumpers to the charge buss. A mechanical disconnect makes things too easy to screw up out of ignorance.

 

[Canton]

Disconnecting under load is not the problem, it is reconnecting a single string that has been removed for maintenance or what ever reason that is the problem.

OK you said an EPO is required? Is that correct?

If that is a true statement and I were designing the system I would only use 1 Main Disconnect to take all 6 strings off line. As I said the problem with Disconnects is not a disconnect, it is a reconnect. Having a disconnect makes it too easy to remove a battery string off-line, be forgotten about, then when noticed too easy to close back in and destroy the disconnect because of the huge amount of EQ current which cannot be controlled when you close it back in.

FWIW I have worked with Telecom sector for 35 years and my specialty is battery plants, UPS, AC distribution, Emergency, and Power Protection. We avoid battery disconnects like the plague as it greatly increases the chances for an outage and equipment damage from operator error.

If you are in a plant with trained power plant electricians, disconnecting a batery string for service should be no issue for experienced personnel as all they have to do is remove the jumpers from the Charge Buss, and they had better know to use a rectifier/charger to bring the disconnected battery back up to plant operating voltage before reconnecting the jumpers to the charge buss. A mechanical disconnect makes things too easy to screw up out of ignorance.

Yes, they will have a EPO shunt capability.

There are no Disconnects only DC rated Breakers for isolation.

The system has already been designed and released and equipment will ship within 2 weeks. The disconnection and reconnection of the battery strings is not a concern. I am aware and see your concern with them. The disconnection of the DC power would only be for maintenance on the UPS. Under normal conditions all the batteries would be charged, equalized and at the same ready potential to protect the critical load. This inquiry for a code compliant installation is in reference to the installation of a MAIN DC Breaker Disconnect that would only be for the purpose of disconnecting the DC potential for Maintenance on the UPS.

The Main DC Breaker I am proposing would not have anything to do with shutting down any one, or combination of the six battery stringers in question as they are all in parallel to the UPS (or Main DC Breaker if ever installed) This would only be to isolate the DC power to the UPS. This would have nothing to do with maintenance to the batteries, this would not even enable you to work on the individual battery stringers, you would have to open the individual DC breaker in that cabinet for that stringer.

 

[Canton]

First, I know of no requirement for a single disconnecting means of the DC supply. Just to be certain, I tried to find one and could not.

While I understand your concern, I do not believe one DC disconnect will reduce accident potential to any substantial degree. It also adds another level of complexity to the system. The one advantage I do see is disconnecting all six banks simultaneously vs. one at a time. That should only make a difference when disconnecting under load... which shouldn't occur when safe switching procedures are documented and adhered to.

With all six disconnects being shunt trip, what about tripping them all simultaneously through an auxiliary contact in the Main UPS input disconnect? Designing UPS is not my forte, but I cannot think of any reason to maintain battery connection when the AC input is disconnected (other than testing purposes, during which the shunt trip circuit can be defeated).

Smart$, thank you for your response. I have used your advise in the past and always appreciate your insight.

There have been multiple times when it has been needed to eliminate all sources of power from a UPS for maintenance, and just last week to replace an entire UPS unit. This system is different as compared to the others in the facility with respect to the Battery Disconnects/Breakers. This is why the extra securitization. Safety is always a concern for us. I have discussed your idea with others about using the shunt capabilities and I like the idea. The only draw back that we talked about was having to go back and manually reclose those six breakers. As others have pointed out this leads to other potential issues.

If our client does not want to pay for the additional Main DC breaker, we are along with other documented procedures, going to place a placard on the UPS noting the additional DC disconnect procedures. My initial inquires where only to find a code reference that would force the additional breaker. I was unable to find anything, and I'm confident it does not exist if you were unable to find anything either.

Thank you for your time!

 

[dereckbc]

This inquiry for a code compliant installation is in reference to the installation of a MAIN DC Breaker

OK let's make it easy. No code requirement unless it is local requirement, or specified by the client. Are you taking the lienancies of 645? If so the EPO assuming operates the 6 other smaller disconnects meets the requirement.

 

[Canton]

OK let's make it easy. No code requirement unless it is local requirement, or specified by the client. Are you taking the lienancies of 645? If so the EPO assuming operates the 6 other smaller disconnects meets the requirement.

Correct, Per 645.10 and 645.11, the EPO will disconnect all (6) Battery Cabinets. I was hoping I could find some other requirement to force a Main Breaker into the install.

Thank you for your time and input.