Doesn't this seem like a bad idea?

[doubleE]

So just making sure I understand this correctly, it's in violation of 210.7(B) because there is no way to simultaneously disconnect both the circuits that supply the receptacle? Would some sort of shunt trip breakers get around this?

 

[weressl]

I'm doing work for a client designing a facility with UPS backup.
This is what the client wants to do:

They are basically suggesting having one panelboard fed directly from the utility, and one fed from the UPS, then run wire from both and connect them to the same outlet, for "redundancy" as they put it. Does this seem like a bad idea to anyone else but me?

As I understand it, they'd have to make damn sure they connected the same phases (since they're 3-phase panelboards). But even then, if the power from the UPS inverter leads or lags the power coming from the other panelboard, even a little bit, wouldn't that be disasterous?

I don't consider this a bad idea. It is outright stupid.

 

[doubleE]

I don't consider this a bad idea. It is outright stupid.

Thank you. I'm trying find a nicer way of telling the client this :smile:

 

[76nemo]

Thank you. I'm trying find a nicer way of telling the client this :smile:

"It's just outright stupid,......SIR"???

 

[charlietuna]

If your UPS is functioning as designed why do you have a second power source applied to this receptacle?

 

[dbuckley]

Yeah, +1 on the stupid.

However, as noted above, all is not lost; you could add a second receptacle everyplace there is one receptacle, with the second feed serving the second receptical, and ensure that every device in the data centre has dual power inlets and plug one inlet to each receptacle.

This provides continuity of operation through UPS failure, or just maintenance of the UPS fed distribution board. Ok theres a bit of risk involved with just running on the raw mains (what if the supply failed whilst the UPS is out for maintenance?) but it may be quite a small risk.

This configuration approaches industry best practice, in that the primary power feed is concurrently maintainable with ther data centre continuing to operate. Though in most cases the second feed has a UPS all of its own too.

 

[Ibarra]

Redundancy

Redundancy

Double,

You are right; in a sense a UPS is an inherently redundant system (Double Conversion Online UPS). A UPS has got an internal static switch that automatically switch to utility (less than 2 micro-seconds) when necessary (failure of internal components of UPS or short-circuit). Additionally, some UPSs have got a manual switch (also called maintenance bypass) to bypass the UPS electronics. This design is aimed to keep your load up the most of the time (typically 99.9%). If your load needs a higher availability (99.999) you can parallel additional UPSs. These UPSs have got (obviously) a technology to share the load and sync their output.

Having said that, I close my comment with Laszlo?s words,

It is outright stupid!

Ric

 

[don_resqcapt19]

Charlie,
There is no question what so ever, in my opinion, that a UPS is an Article 700, 701 or 702 system. The article that applies would be based on the loads served.

 

[doubleE]

Yeah, +1 on the stupid.

However, as noted above, all is not lost; you could add a second receptacle everyplace there is one receptacle, with the second feed serving the second receptical, and ensure that every device in the data centre has dual power inlets and plug one inlet to each receptacle.

That's something they had suggested. In some places, particularly under the LAN equipment racks, they will have two receptacles as you describe. The rack equipment then has a rack ATS that will be supplied by both the UPS and the mains. This seems like an extra switching step and overkill to me, and I see no benefit over just having the one receptacle backed up by the UPS.

 

[jim dungar]

Unless the UPS is a very expensive unit it does not output a true sine wave, so its output could never be synced with the utility.

But, two 20A breakers interlocked so that only one could be on at a time would provide the "bypass" redundancy found in larger UPS installations. Backfeed two breakers into a small panel and interlock them like you do for a generator.

 

[doubleE]

I appreciate all the input on this, especially the possible code violations. That info will come in handy. Once this system is up and running and proven to work, it will be widely implemented, so I want to get it right this first time. As it is, I'm almost embarrassed to have my name associated with the design. It just doesn't make sense to me. I really hope I've just misinterpreted the client's intentions with this system. I think it's just a case of too many non-electrical cooks in the kitchen.

 

[dbuckley]

and I see no benefit over just having the one receptacle backed up by the UPS.

Concurrent maintainability!!! - You can switch off the UPS output to work on the distribution panel without dropping service and without needing to work live. These are big benefits.

Unless the UPS is a very expensive unit it does not output a true sine wave, so its output could never be synced with the utility.

UPS outputs are always sync'd to the utility because UPSs have really poor fault clearing capability, so they need to be able to static switch back to utility power to clear downstream faults. If the UPS and utility aren't in sync then static switching isn't going to be pretty.

 

[jim dungar]

UPS outputs are always sync'd to the utility because UPSs have really poor fault clearing capability, so they need to be able to static switch back to utility power to clear downstream faults. If the UPS and utility aren't in sync then static switching isn't going to be pretty.

Only UPS's that have static bypasses are synced with the utility. Most small UPS's do not contain bypass circuitry.

 

[doubleE]

For your information this is the UPS that will be used. There will be a maintenance bypass installed as well.

 

[76teach]

During a utility outage the UPS would back feed into the non UPS panels via the receptacle as well. Which 20 amp breaker trips first?

 

[Doug Peter]

Does this not back feed one panel from the other and this is not allowed or safe?

 

[dbuckley]

Only UPS's that have static bypasses are synced with the utility. Most small UPS's do not contain bypass circuitry.

Most small UPSs are not double conversion and thus must be mains synced.

Seriously - I've never seen or heard of a UPS that doesn't sync to the incoming mains from toys that kick round on the floor at home, to 2MW serious modular UPS systems.

Thats not to say they dont exist - maybe they do, but I've never seen one. But you couldn't use a UPS that doesn't sync in any sort of situation where you want any form of bypass static, relay or manual. That must be a very small market segment.

 

[steelersman]

UPS systems that have battery backup are standby systems


I would also cite 210.7(B)Multiple circuits.

Don't all UPS systems have battery backup? I thought that was what makes them "uninterruptible power supplies".

 

[don_resqcapt19]

Most small UPSs are not double conversion and thus must be mains synced.
Seriously - I've never seen or heard of a UPS that doesn't sync to the incoming mains from toys that kick round on the floor at home, to 2MW serious modular UPS systems. ...

Why would they be "synced"? They don't parallel their output with the utility. They close a switch to provide battery power to the loads from the inverter after the sense the loss of utility power. The do not have a "closed" transition back to the utility power after it has been restored.

 

[jim dungar]

Thats not to say they dont exist - maybe they do, but I've never seen one. But you couldn't use a UPS that doesn't sync in any sort of situation where you want any form of bypass static, relay or manual. That must be a very small market segment.

I guess I am thinking the OP would be a personal/small office/network server style. There are many smaller units that are used in stand-alone applications and are not maintanable while energized.