AF Safety concerns: 150kVA pad mount xfmr boosting 208 => 480 in Data Center

[wyreman]

Some concerns about a 20x25 room with one exit and a
150kVA pad mount xfmr boosting 208 => 480 in Data Center
https://1drv.ms/f/s!AkuU8mYuf3eEgV_krr3OkdkHYYEi

We only have one exit door in the server room.
Server room is classified as S2 Low Hazard Storage because the xfmr, but the batteries on the UPS could bring it up to H3 - which would require less than 25' to the exit door.

What is the actual Arc Flash danger here?
There is no present danger in my opinion, but there is a question of whether this is intrinsically unsafe to have the xfmr in a room that may have unqualified people in it.
If the doors on that xfmr are not opened, is there a danger to anyone in the room?
I've put my hands on the effects of AF when it happened with the doors shut, and its usually holes burned in the cabinet, but the doors didn't get blown off or anything.
I've got to call square d and ask them

The next OCP before the xfmr is a 400a main disconnect. The xfmr is tapped from that bus.
The only load on the 480 [secondary side] is a pair of intermittently used 90a 3p cbs

square d 35149-17232-023 150t3hb instruction manual three phase insulated transformer

Watchdog® Low Temperature Rise
• Designed for low energy losses at loads greater than 50% of
nameplate rating.
• Extra long life expectancy using 220 oC insulation system
designed for full load operation at a maximum temperature rise
of 115 oC or 80 oC instead of 150 oC.
• Continuous emergency overload capability of 15% on 115 oC
rise and 30% on 80 oC rise.
NOTE: These products are obsolete and can no
longer be ordered. The information in this table
is for reference only.

 

[winnie]

Speaking outside of my training/experience level:

IMHO the transformer itself presents no more arc flash hazard than any other utilization equipment at that voltage and power level. IMHO there will only be a significant arc flash hazard when the equipment is being worked on...and it should not be worked on while energized.

But to actually evaluate the hazard you will need to know the available current on the feeder supplying the transformer.

-Jon

 

[wyreman]

It's a 240 V 400 amp
Cutler Hammer
three-phase large mainframe
molded case Breaker
pretty old I have to go and look at the label

 

[jim dungar]

NFPA70E pretty much requires that your company perform a Risk Analysis. For instance; the 'task table' methodology now starts with a Yes or No table for PPE being required.

There are several methods available for you to calculate what amount of Incident Energy might result from an arcing fault event. But this is simply the hazard. It
You must also evaluate the 'likelyhood' or possibility that the hazard occurs.

It it probable you will reach the same conclusion that many others have - there is no inherent danger in simply being near the majority of enclosed 480V equipment.

 

[wyreman]

If I have any concerns regarding this, the only concern I really have is that when a lot of current is flowing that main switchgear there is some little buzzing noise could be from loose busbars
I really want to turn it off
take it all apart
Turn it back on
run at Hard and at least do a FLIR camera test on it
but nobody ever wants to turn it off.

It's just the buzzing
I put my hand on it when I was buzzing and try to see if I can silence the vibration
But nothing that was accessible in the lower part of the cabinet was vibrating

 

[Chamuit]

I don't think I have ever heard a "quiet" transformer.

Cute Data Center, BTW.

 

[GoldDigger]

I don't think I have ever heard a "quiet" transformer.

Sort of by definition???

 

[Chamuit]

Sort of by definition???

:slaphead:

 

[wyreman]

I don't think I have ever heard a "quiet" transformer.

Cute Data Center, BTW.

Hey it's only 20 x 25!
&
That was back when they were just moving in.
And it's not the transformer that buzzing
Humming
it's the busbars at the main that feed the transformer
They sound a little strained
Have A great weekend!

 

[Chamuit]

Hey it's only 20 x 25!
&
That was back when they were just moving in.

And it's not the transformer that buzzing Humming
it's the busbars at the main that feed the transformer
They sound a little strained

Okay, makes more sense now. I have very limited experience with bus duct, so I'll go back to lurking. Although, have you put your ear to connection points, so to speak, to localize the noise?

Have A great weekend!

:bye:

 

[ron]

If the equipment covers are on, in of itself NFPA 70E would consider there to be no arc flash danger.

Although not directly related, I would re-evaluate the Occupancy groups. A transformer does not make a S-2 occupancy, and UPS batteries do not make a H-3 occupancy.

If that occupancy labeling were taken to the next level, then office space with garbage cans would take on an H-3 category.

 

[wyreman]

But really in the worst case scenario
What's the threshold boundary for a
480 V 150 kVA transformer anyway
Even with all the dead fronts
the doors open

18 inches ??

in arc flash training there were
different circles and radius
This is No 12 KV
Must just be at most arms length 36"

 

[wyreman]

boost buck

boost buck

while some were snickering:dunce: I had missed that not every xfmr can boost as well as buck
as per sQd

It seems your customer is reverse feeding/backfeeding the transformer, and we don't recommend that. Backfeeding causes very high excitation inrush, making coordination to breakers or fuses difficult without exceeding the limits set by the National Electrical Code.

For a 208v - 480wye application use EX150T212H.

he said that would generate lots of heat and noise and trouble - not recommended

 

[wbdvt]

What is the actual Arc Flash danger here?

In the normal operating state, there is no arc flash danger. An arc danger would exist when work is being done that could cause an flash. Examples are removing covers from energized equipment, testing with a multimeter, etc.

I guess I am a little confused what the concern is on having a transformer there, humming is normal for a transformer as well as some heat. Looking the pictures, I would be more concerned with the housekeeping in an electrical room

 

[wyreman]

This is a common xfmr
rated for a 480 primary and 208 secondary.
Back feeding it to 150kVA will cause
Extra heat, humming and coordination trouble


The arc flash question would be
if the insulation is sufficient
on the 208v rated assembly

 

[wyreman]

anybody in the market for a slightly used 150kVA xfmr ?

anybody in the market for a slightly used 150kVA xfmr ?

anyway, that was put in by someone else and I'm glad we spotted it, and are getting rid of it...
anybody in the market for a slightly used 150kVA buck xfmr that was boosting?

got to move it out of the network room because it generates too much heat.
I am thinking about taking the opportunity to get rid of some of the 480 load as well -
redistribute some of that load at 208
It may help with the sequence of work


Also have to get a mitsubishi a/c unit in the IDF room - the big issue is the condensate
... but first thing is to move that xfmr.

 

[GoldDigger]

It was a step down transformer that was stepping up. No buck or boost involved.

mobile

 

[wyreman]

we used to boost these up and down
up to 480v for a triplex run and then
down to 208 for the point of use on portable bldg.

step down xfmrs on both sides - no problem for years
we even had spare 4160's

what is your opinion on the insulation on the low volt side here being a 480 secondary?
probably work fine for years...

just not what the manufacturer recommends tho

 

[winnie]

As far as conductor insulation is concerned, it pretty much doesn't matter if you use the transformer for step up or step down. The 208V side is being used at 208V, and the 480V side is being used for 480V.

The problems come in with other features of the design.

1) Grounding. In most situations you want to use a grounded electrical system. The 'standard' approach for 3 phase systems is to ground the neutral of a wye secondary. Being used in reverse, your transformer doesn't have a wye neutral, so grounding is less obvious...and might require more expensive switchgear or cause problems with equipment.

2) Primary side 'taps'. Transformers are usually designed with 'taps' on the primary coils, to permit adjustment for different supply voltages. The transformer might be nominally 480V HV, but have taps for 456V, 468V, 480V, 492V and 504V. This lets you get proper magnetic saturation and proper output voltage over a range of input voltages. When the transformer is reversed, these 'taps' end up on the secondary side. If your supply voltage is high, then you might over-saturate the transformer core.

3) Inrush current. Transformers are designed so that the primary has more 'leakage inductance' which acts to limit the surge of current just as the transformer starts up. When you swap primary and secondary you don't get this design feature, and will see higher proportional inrush current, which might do things like trip breakers when you first energize the transformer. For smaller transformers this often is not a problem, and specialized switchgear can 'pre-charge' a transformer to get around this sort of problem...but it is easier to simply have a proper step up transformer.

All of the above can frequently be tolerated or accommodated with proper design; if I was in a rush I would not hesitate to use a 480V delta to 208V wye transformer 'in reverse' to get 480V from 208V. But if I had time to get the correct transformer that would be my very strong preference.

To expand on GoldDIgger's terminology issue: buck and boost are 'terms of the art' that specifically describe transformer connections used to make small changes in voltage where the transformer secondary is placed in series with the supply voltage. See http://www.emerson.com/resource/blo...9d2737c515229fc9a7/shdcat-buck-boost-data.pdf for example

In 'normal English' you are boosting 208V to 480V. But in 'transformer English' you do not have a 'boost' configuration. Instead you have an isolating transformer connected to step up voltage

-Jon

 

[wyreman]

for the grounding... how should the backfed xfmr be grounded?
I havent opened it yet, but there is a rod driven and a small EGC I'll follow

also, the here is a video of the sound of the backfed load on the main gear