Underfloor receptacles and flex cords in data centers

Status
Not open for further replies.
mshields

mshields

Senior Member
Location
Boston, MA
Article 400.8 prohibit flexible cords in floors. Is it one of the leniencies of 645 to allow this (I don't see it)? Or is it implied given that you are allowed to put receptacles under the floor per 645.5(D)(2)?

In general how do we get away with a typical 645 application where your NEMA plugs are in boxes resting on the structural floor. i.e. beneath the raised floor.

Isn't the allowance of receptacles under the floor per 645 redundant in that they are allowed regardless? i.e. The problem in all scenario's is installing flexible cords such that the receptacles are useful.


Mike
 
400.8 prohibits any type of flexible cord beneath the raised floor whereas 645.5(5) allows for Type DP cable to reside below the floor in an IT Equip. Rm.. However, I can tell you that all power cords shipped with our servers are not DP rated. They are mostly Type S... cords. I don't forsee us paying x-amount of dollars to replace every single cord in our data center.

I'm sure I know the answer but who should be responsible for the right cord Type? The consumer or the supplier?? I did some poking around online and have yet to find a supplier of Type DP-1 or 2 power cords.


Good article on ECMWeb by Mr. Holt himself:
http://ecmweb.com/nec/electric_exploring_requirements_information/

IAEI Article:
http://www.iaei.org/subscriber/magazine/01_a/burke.htm

UL Standard for Type DP cable:
http://ulstandardsinfonet.ul.com/scopes/1690.html
 
Thanks Vaughn

Thanks Vaughn

At least 645 gives you rules for which you can put flexible cord under the floor (however impractical they might be). But many in this forum contend that Article 645 is not mandatory (and I would agree based on their arguments and code citings). But if you're not going by article 645, how do you accomplish distribution from your below floor cabling without violating 400.8?

Mike
 
Articles you posted

Articles you posted

Gave them a cursory look and I see that, particularly the one by Mike Holt, provides the counter argument in favor of 645. I look forward to more thoroughly reviewing these over the weekend (what a geek I am).

Thanks,

Mike
 
Mike,
If your not complying with 645, then you are correct in saying that a cord cannot go through a hole in the raised floor to reach a receptacle below the raised floor.
I have had a lot of luck having the branch circuit brought up through the raised floor in MC cable and the receptacle either mounted flush to the raised floor, on top of the raised floor or to the frame of the IT equipment rack, depending on the application.

Almost all raised floor manufacturers sell a raised floor tile that has a device box mounted flush in it if you like it pre-made.
 
BTW, there are a lot of DP cord manufacturers out there.
Go to
http://database.ul.com/cgi-bin/XYV/template/LISEXT/1FRAME/index.htm
and type in EMRB in the UL category code

There are 43 companies with listed DP product.
The plug strip folks need to buy some and make it an option.
Most of those 43 companies require a minimum order because the DP cable is special, but it would seem like a good idea for a plug strip company to be the first on the block to make one, but I can't figure out why they don't do it.
 
Last edited:
I agree. I don't understand why DP type cords aren't offered at time of server purchase. I understand that only data center environments with a raised floor would fall into using DP type power cords but these were introduced to the NEC in 1996. Seems like a long time for no one to press the issue or maybe that's why....No one has pressed the issue.

How often, if any, does an inspector of any type enter a data center and slap the owner's hands for non-compliance of DP cords? APC and other major PDU distributors don't even use DP cords on their rack mount PDUs. I guess either don't know or don't care? Maybe the cost is too high?
 
I've spoken with a major plug strip manufacturer (I don't call them PDU's as they are just fancy plug strips) R&D. My impression of the discussions were that they just don't want to invest the $$$ & effort for something that they don't think will be enforced to recoup the upfront R&D. That is just my opinion of the discussions, not the party line.

IMHO A real PDU is on the distribution portion of the system not the branch circuit end.
 
Most of the data centers I have installed and or work in that have a raised floor, we used a sealtight or flex up into the rack for the L5-20R or L6-30R. I hate raised floors. They are always a mess and you can never get in the tile you need because the rack is sitting on it. Some of the older raised floor installations are not able to cool the enviroment do to the blade servers that have been getting installed. The lastest data center I worked in refused to put in the raised floor...Plus the EPO issues we half all read about!
 
Safety provided by EPO for fire fighters

Safety provided by EPO for fire fighters

In the article by Mike Holt (to which there is a link in this thread) he talks about the importance of an EPO for the safety of fire fighters. His contention is that the signage required by Article 700 and 701 is simply not enough. Some compelling evidence has been given in this thread for not using EPO's and for them not being required by the NEC. But what about this serious safety concern that Mike Raises. Any thoughts on alternative safety measure in lieu of an EPO?

Mike
 
Mike,
But what about this serious safety concern that Mike Raises.
Click to expand...
I don't really see it as a real safety issue for firefighters. First if the room is so involved that hose streams are needed, it is not real likley that any of the equipment is still energized. Even if it is, fires in energized electrical equipment can be safely fought with water from a fog nozzle up to about 100,000 volts. If this is a concern of the resonding department, it should be addressed in their "pre-plan" for the occupancy.
Don
 
Maybe not the greatest way to argue ...... but
If it was so dangerous, then why are there no EPO requirements for switchgear rooms or industrial operations?
Seems not to be that much of a danger as compared to regular fires.
 
Switchgear rooms

Switchgear rooms

To play devils advocate Ron, a switchgear room can be de-energized via the main disconnect. As you know, A UPS doesn't work that way.

Then again, if you have dc control power, you've still got DC power to the gear.
 
On a different note....

I attended Eaton's "Power Quality Tour" in Sept. 2006 and it was a very informative week to say the least. Highly recommended! The most interesting was the last day where the big Q&A session took place. Me being in the IT world and everyone else in the room being strictly EE's or contractors of some sort, it was enlightening to me the disconnect between the "electrical world" and the "IT world".

I say the latter not to degrade either discipline but rather to make a point about how important power is in a data center. I don't think anyone would argue that there is lack of knowledge on both sides of the table when it comes to the power distribution needs of a data center. The last couple threads over Article 645 makes this apparent in one form or fashion.

Just my thoughts :).
 
Mike,
To play devils advocate Ron, a switchgear room can be de-energized via the main disconnect.
Click to expand...
Not really, the unprotected main bus feed from the utility is still hot.
Don
 
It seems like all the trade mags have a different spin on this issue. I believe the AHJ will have the final say weather or not the EPO is required. I agree with you on the safety issue for the fire crew, but they generally know if they are going to a data center or not. Most data centers have a Pre-action (air then water) or a FM200 system installs. The man issues I see for having the EPO is nuisance trips or upset employees. Maybe the fire Dept could have a switch to dump everything when they get on site. After they access the fire.
 
I once went to a data center where the AHJ insisted that there be an EPO at the main entrance to the building.
It was funny seeing this mushroom switch on the wall near the "check-in" desk, knowing that if there was a disgruntled worker, they could shutdown the entire datacenter on the way out :(
The design engineers should have forced a code interpretation from the state and not just did what the AHJ wanted on a whim.
 
ron said:
I once went to a data center where the AHJ insisted that there be an EPO at the main entrance to the building.
It was funny seeing this mushroom switch on the wall near the "check-in" desk, knowing that if there was a disgruntled worker, they could shutdown the entire datacenter on the way out :(
The design engineers should have forced a code interpretation from the state and not just did what the AHJ wanted on a whim.
Click to expand...
Had the AHJ at King of Prussia and San Jose requiring an EPO in about the same situation. A year after the sites were built we had our day in court and won. Removed the EPO's shortly thereafter. The moral I learned is th eAHJ does not have the authority to require 645 application.
 
Does'nt the EPO protect the equipment also ?
I know it is a safety devise but it also protects the equipment from being doused by water when energised.
The amount of personel in the actual data center is minimal.
The owner has a ton of money invested in equipment and the epo protects this as well
 
Status
Not open for further replies.
Top