Computer rooms - EPO's

Status
Not open for further replies.

mshields

Senior Member
Location
Boston, MA
Just checking in

Just checking in

Driving back with my family from xmas celebrations and geek that I am, decided to check in to see what more had been said about EPO's, data centers, etc and I find that the string has gone to hell. Just kidding. Merry Christmas everybody! By the way my wife is driving.

Mike
 

Tori

Senior Member
Location
Maryland
Just wanted to add, as I was doing some demo work at on a dat center floor I was surprised to find that one on the recepticle cir.s was going into a jbox under the floor in the elec rm also had two epo cir.s in it also.

this is a recipie for dissaster I thought to myself
 

steve66

Senior Member
Location
Illinois
Occupation
Engineer
Tori said:
The gas fire suppresion you speak of , what type is it ?
I do remember a gas of of old but not the name -that was disscontiued in my area because of the danger to personel
Most of the data centers I work in have a dry type sprinkler -so that a fire triggering the epo would shut down everything so the water would'nt ruin it-hot it will off it won't .
Most of the raised floors in the newer buildings we have wired are 3 ft. raised, some of the old ones we have added power to have 2 ft. raised.
I am about to wire an expansion to a data center so I find this thread very interesting


There are new types of gas fire suppression systems that aren't as dangerous as the old gases. Check with your local fire suppression contractor.
 

engy

Senior Member
Location
Minnesota
How do fire marshals react to No EPO's in a datacenter?

A while back, in a different thread, it was mentioned that fire departments kinda like a quick means of disconnecting power to datacenters...
 

bhsrnd

Senior Member
Location
Fort Worth, TX
mshields said:
Or is it Intergen? Does it work the same as FM200?


FM200 works by rapidly absorbing heat from the source whereas Intergen works by reducing oxygen levels to around 15% or so. Both are NOT Halon nor are they in any way related to that product. Halon was discontinued in December of 1993 because of its negative environmental impact and potential health hazards to humans. Sites where Halon systems were installed prior to 1993 were allowed to remain unless some circumstance required the replacement of the system or local codes/laws required replacement.

The MSDS for both FM200 and Intergen indicate some health risk but it is minor. Anyone who stands in a room and purposely inhales that stuff deserves whatever they get healthwise. In a data center environment there is typically a three-step alarm process that occurs before the gaseous agent is dumped allowing for ample evacuation time.

And for useless information sake FM200 is a brand name copyrighted to Great Lakes Chemical Corporation and Intergen is a brand name copyrighted by Ansul Inc..
 

bhsrnd

Senior Member
Location
Fort Worth, TX
engy said:
How do fire marshals react to No EPO's in a datacenter?

A while back, in a different thread, it was mentioned that fire departments kinda like a quick means of disconnecting power to datacenters...

In additon, the codes where we are require that all air movement cease in the event of a fire alarm to prevent fueling the fire and the spread of smoke. This means that our EPO not only sheds power to the PDUs but also the CRAC units. However, a series of events has to occur before the EPO is actuated. The EPO system actuates when all else has failed to rectify the situation. By this time one would see that the gaseous agent failed to suppress the ignition source which means there's a fire which has actuated the sprinklers. At this point we have bigger fish to fry than worrying about our EPO system actuating.

However, this is not to say that equipment (CRAC units and the like) nowadays cannot be connected directly to an output on the fire alarm system and allow it to trigger the shutdown of specific equipment unlike an EPO system where everything is turned off.

Before dismissing Article 645 I would check with local fire and building codes and ensure you wouldn't put yourself or the customer in violation by not following certain parts of said article in relation to EPO systems.
 
Last edited:

mshields

Senior Member
Location
Boston, MA
Intergen/Ansul

Intergen/Ansul

Vaughn,

Thanks for the input. Tell me, are "Ansul" systems we see put in, in kitchen hoods, actually Intergen systems?

Thanks,
 

mshields

Senior Member
Location
Boston, MA
With regard to fire departments requiring a quick disconnect in Data Centers

With regard to fire departments requiring a quick disconnect in Data Centers

I believe the concern your fire department would have is having liquid tight flex under the floor. In that instance, you've got the potential for combustible gasses to be circulated within the room. If your wiring in the floor plenum is in accordance with NEC 300, and say you've used MC cable under the floor, then the hazard posed by a data center is no greater than it would be in any room.

Mike
 

mshields

Senior Member
Location
Boston, MA
Article 400

Article 400

Doesn't Article 400 prohibit flexible cords in plenums. Is it one of the leniencies of 645 to allow this? Or is it implied somehow given that you are allowed to put boxes under the floor that are not secured in place?

In short, how do we get away with a typical application where your NEMA plugs are in boxes sitting on the floor of a data center. The question applies to a 645 data center or a non 645 data center?

thanks,

Mike
 

bhsrnd

Senior Member
Location
Fort Worth, TX
dereckbc said:
The rest goes overhead on cable racks centered over the equipment line-ups.

I urge you to spring for a copy of NEBS. It has tons of eviromental, lighting, line-up spacing, equipment racks, and space allocation info.


Allow me to preface this by saying that I have not read NEBS and know about it only what I have read on Telcordia's website.

Being in the IT industry since 1997 I am cautious on the applicability today of any literature from the 1990's that discuss data center design. I say this because of the fact that the thinking behind data center design has changed drastically over the last ten years. The way a data center would have been designed in 1996 when NEBS was last updated (according to Telcordia; formerly Bellcore) is not the way it would be designed in 2007. In addition, from what I gather, NEBS is geared towards telecom facilities and Central Offices (COs) which are not data centers but rather telephone switching stations and the like used by telco providers like Verizon. However, I'm sure that good information can be found in NEBS and applied to data centers.
 

mshields

Senior Member
Location
Boston, MA
Nebs

Nebs

Trying getting a copy of the standard! You need to pay 1500 dollars for an Enterprise wide version of it. They do however have a group to which you can pose questions. I'll share whatever input I get if and when I hear back from them.

Mike
 

tmillard

Member
FM200 and EPO's

FM200 and EPO's

I spent some time reading all the posts on data centers and EPO's and have a quick question. If you discharge an FM200 or similar system, do you have to shut the power off to all the equipment in the room? I working on a computer room in a hospital and they have a telemetry rack in there. They don't want to shut power off to that rack. I've read NEC 645 and NFPA 2001 Clean Agent Fire Extinguishing Systems and can't find where it says you have to disconnect the power.

Thanks...

Tom
 

jjs

Member
Location
Puryear, TN, USA
high density data centers are not able to be cooled by underfloor cooling alone. In rack cooling is required. Blade servers generate way too much heat per SF to cool properly with in floor only. You get hotspots in the rack itself. Data center design actually goes in cycles, so something in that 1990 document will eventually come back in to mainstream.

For example, there is talk about going to DC power for datacenters again. They are suppose to be more energy efficient.

Also liquid cooled computers such as the old mainframes were are becoming more popular as power densities go up.

The last data center I did had in rack cooling and all overhead elec and communication. But the floor height of the space was too low to even consider a raised floor. Raised floors in my opinion just get really clogged full of cables that are very difficult to trace.

Unless client requests EPO, avoid them like the plague. EC adds a circuit to a panel and accidently shorts the EPO signal wire that was run in the wrong place and now the whole room goes down.
 

jjs

Member
Location
Puryear, TN, USA
FM200 you have to leave the room. Inergen you can stay without adverse effects for quite some time. Inergen reduces oxygen level but replaces the oxygen with another gas which fools your body into not needing as much oxygen. I was at a inergen display where they put a volunteer in a booth and give him a lit candle. they close the door and turn the inergen system on. the candle goes out and the guy stays alive.
 

bhsrnd

Senior Member
Location
Fort Worth, TX
tmillard said:
I spent some time reading all the posts on data centers and EPO's and have a quick question. If you discharge an FM200 or similar system, do you have to shut the power off to all the equipment in the room?

As you've probably already read, 645 is an OPTIONAL section of the NEC unless the IT Equipment Room was designed under the auspices of Article 645. In that case, yes, ALL power would have to be shed upon actuation of the EPO system. See 645.10 and 645.11. If all the conditions in 645.4 are met, then it is possible that Article 645 applies.

Here is another thread over EPO systems:
http://www.mikeholt.com/code_forum/showthread.php?t=82814


jjs said:
FM200 you have to leave the room. Inergen you can stay without adverse effects for quite some time. Inergen reduces oxygen level but replaces the oxygen with another gas which fools your body into not needing as much oxygen. I was at a inergen display where they put a volunteer in a booth and give him a lit candle. they close the door and turn the inergen system on. the candle goes out and the guy stays alive.

In either case you will want to leave the room. See post #47 in this thread for my earlier response to FM200 and Intergen. Neither poses immediate health risks but I wouldn't recommend long term exposure to either agent.
 
Last edited:

tmillard

Member
EPO and Life Safety Equipment

EPO and Life Safety Equipment

I read 645 and I do not believe it applies to my application

I then looked at NFPA 2001 which deals with 'Clean Agent Fire Extinguishing Systems'. No where in this section does it state that you have to disconnect power if you release the agent.

I then moved on to NFPA 75 which deals with 'Protection of Information Technology Equipment'. Section 8.4.2.1 states "The power to all electronic equipment shall be disconnected upon activation of a gaseous agent total flooding system, unless the risk considerations outlined in Chapter 4 indicate the need for continuous power.'

Chapter 4, section 4.1 Risk Factors states "The following factors shall be considered where determining the need for protecting the environment, equipment, function, programming , records, and supplies:

1. Life safety aspects of the function (e.g. process controls, air traffic controls.)
2. Fire threat of the installation to occupants or exposed property
3. Economic loss from loss of function or loss of records
4. Economic loss from value of equipment"

So for my health care application, it looks like I can leave the power to the telemetry rack on based on #1, Life Safety aspects.
 
Status
Not open for further replies.
Top