Electrical conduits go through IT room

[weressl]

Can you provide more than your say so as that is not what the NEC currently requires

Did I say anything about the NEC?

I suggest you play the lottery as you are the only one that has posted that they have seen it happen at all ... never mind several times.:grin: Or maybe you were the cause? :grin:

The more likely cause would have been you or someone like you; the 'electrician'.

You would think that a properly engineered and installed system would have overcurrent and ground fault protection that would open when there was a fault instead of counting on a person being there to hit an EPO ...... but hey I am just an electrician so what do I know?

If you are correct then why would we NOW have arc-fault protection in homes? Because you can protect the dickens out of anything and Murphy still proves superior....

 

[iwire]

Did I say anything about the NEC?

No ... which is odd as I think the OP was looking for NEC information.

The more likely cause would have been you or someone like you; the 'electrician'.

No not me, I have worked in many large date centers, the only time an EPO tripped in my presence Liebert admitted it was an issue with their PDU.

A false trip that cost the software company that owned the room a ton of money.

If you are correct then why would we NOW have arc-fault protection in homes? Because you can protect the dickens out of anything and Murphy still proves superior....

We were not talking about 20 amp branch circuits, I thought we were talking about a 400 amp feeder to an IT room?

Normally I see something like that with GFP but maybe the engineers are better up here. :grin:

 

[roger]

but then again I hang around installations that are engineered and installed properly.:roll:

So you are saying that all those instances mentioned in the earlier posts were due to bad engineering, (as installers we locate the EPO's where they are shown on the drawings) hmmmm, I guess there are alot of subpar engineers designing IT spaces.

Roger

 

[barclayd]

EPO started with IBM,

Speaking of IBM - Many years ago I worked at the IBM factory in Boulder. The EPO "Design Standard" actually switched the neutral in the shunt trip circuit.

Now back to the Bob & Laszlo show.

db

 

[iwire]

Now back to the Bob & Laszlo show.

There is room in the billing for you as well.

 

[weressl]

We were not talking about 20 amp branch circuits, I thought we were talking about a 400 amp feeder to an IT room?

Normally I see something like that with GFP but maybe the engineers are better up here. :grin:

Unless it is a switchgear type setup you seldom see standard 400A MCB's equipped with GFP. If they have separate ground sensing, the trips are rarely the electronic, accumulative type that provide Arcing Fault Protection.

 

[weressl]

So you are saying that all those instances mentioned in the earlier posts were due to bad engineering, (as installers we locate the EPO's where they are shown on the drawings) hmmmm, I guess there are alot of subpar engineers designing IT spaces.

Roger

Selective posting syndrome again. What I said "engineered and installed". Data centers are routinely handled by people, both engineers and installers, who are not specialized in the field nor do they ahve operational experience on what works and what doesn't. Engineers design systems that look and work well on paper, but not in real life. The majority of electricians installing large new systems like that have no 'living' experience with those installed systems, even if they are involved in service calls.

 

[weressl]

Speaking of IBM - Many years ago I worked at the IBM factory in Boulder. The EPO "Design Standard" actually switched the neutral in the shunt trip circuit.

Now back to the Bob & Laszlo show.

db

Well designed EPO uses UPS power. In this case "when it absolutely, positively has to trip, use UPS" and have an active 'incomplete circuit' monitoring circuit.

 

[dbuckley]

Data centers are routinely handled by people, both engineers and installers, who are not specialized in the field nor do they ahve operational experience on what works and what doesn't.

+1 to that!

Junk data centres must outnumber decent facilities dozens or hundreds to one.

Having said that, where I work now (and not in an electrical capacity at all!) there is a recent vintage fairly well designed data centre. its well enough designed so that there is never the need for a shutdown the centre to work on stuff, and there is never a need to work live on anything; it's electrical system is "concurrently maintainable".

There are (or were) three static switches installed as part of the design, one biggie, which failed, and took out the equipment it protected with it, and two little ones, one of which died the first time it was used in anger, and the second is not reliable in terms of power delivery.

This data centre was designed by people who actually do know what they are talking about, and have many data centres under their belts (including several in the City of London financial district), and even with all that experience they can still get bitten.