NGR question

[Ingenieur]

Inspected a mining machine today
unusual set up
motors are 480/3
but the machine is built to run on 480 or 575
The machine has a voltage selector switch
when in 480 mode there is a direct connection to the supply
when 575 is selected the supply is fed to an auto xfmr bucked to 480 then fed to the motors

both the 480 and 575 supply feeds are wye with a 15 A NGR

when hooked up for 575 operation what will the ground fault current be?

thanks

 

[petersonra]

might well depend on the impedance of the xfmrs.

 

[Ingenieur]

might well depend on the impedance of the xfmrs.

don't think so
ngr ohms >>> xfmr Z
ngr limiting factor

 

[ActionDave]

ngr limiting factor

Then you have answered your own question, have you not?

 

[mbrooke]

Are you saying you want to know fault current if something faults on the secondary of the 575->480 auto transformer?

Post a pic (diagram) of the setup and the point of interest.

 

[gray.one]

The resistors will not interact. Only one source feeds the fault.


Sent from my iPhone using Tapatalk

 

[Smart $]

...
both the 480 and 575 supply feeds are wye with a 15 A NGR
...

15 amp or 15 ohms?

I thought the resistor of NGR systems are generally spec'd to limit fault current to 5 to 10A.

 

[Ingenieur]

The resistors will not interact. Only one source feeds the fault.


Sent from my iPhone using Tapatalk

only 1 is in the ckt at a time
ie, only 1 source connected

what I came up with
480 ngr = 277/15 = 18.5 Ohm
575 331/15 = 22.1 Ohm

When connected to the 575 only 480 is available for the fault
480/sqrt3 / 22.1 = 12.5 A
for the 480 and 575 protection is set at 7.5 A
should be no issue
the 480 (no at) fault is the regular 15 A

I plan on inducing a fault to verify next Wed
just wanted a second opinion lol
basically v ratio 480/575 x 15 = 12.5 A since ngr values are equal

 

[Ingenieur]

15 amp or 15 ohms?

I thought the resistor of NGR systems are generally spec'd to limit fault current to 5 to 10A.

our .law stipulates 15 A for any voltage 480 up to 12470
our ground monitor trips at >3.5 Ohm loop R
frame v upon fault = 52v, survivable with a <15 cycle trip

 

[mbrooke]

I plan on inducing a fault to verify next Wed
just wanted a second opinion lol
basically v ratio 480/575 x 15 = 12.5 A since ngr values are equal

I am sure you know... but i will say it anyways... please do this from a distance. Nothing worse than an MOV or marginal insulation suddenly breaking down from a fault somewhere else giving you generous fault current.

 

[Ingenieur]

I am sure you know... but i will say it anyways... please do this from a distance. Nothing worse than an MOV or marginal insulation suddenly breaking down from a fault somewhere else giving you generous fault current.

I hear ya
We have test equip, ppe and a procedure
we routinely do it, hundreds of times
We do ph-gnd thru the ngr so i is limilted to 15 A or less
We use a remote contactor with a fast acting 25 A fuse (2 in series)
never sustained more than 1/4 sec, the relays drop a uv cb on the feeder
we have feeder protection and have never has a ph-ph fault (yet)
safety is the priority

 

[rian0201]

cant wait for the result. connect a datalog with waveform capture and have it posted here.. hehe


Sent from my iPad using Tapatalk

 

[Ingenieur]

cant wait for the result. connect a datalog with waveform capture and have it posted here.. hehe


Sent from my iPad using Tapatalk

not as dramatic as it sounds
we use a recording clamp on amp meter
very smooth since x/r very low with a >18 Ohm resistor

 

[mbrooke]

not as dramatic as it sounds
we use a recording clamp on amp meter
very smooth since x/r very low with a >18 Ohm resistor

You size the resistor 125% of the normal capacitive phase to ground current (current which would otherwise appear during a fault without the resistor)?

 

[mbrooke]

I hear ya
We have test equip, ppe and a procedure
we routinely do it, hundreds of times
We do ph-gnd thru the ngr so i is limilted to 15 A or less
We use a remote contactor with a fast acting 25 A fuse (2 in series)
never sustained more than 1/4 sec, the relays drop a uv cb on the feeder
we have feeder protection and have never has a ph-ph fault (yet)
safety is the priority

Sounds good then Do you trip or hold on a ground fault?

 

[Ingenieur]

You size the resistor 125% of the normal capacitive phase to ground current (current which would otherwise appear during a fault without the resistor)?

The capacitive current is typically less than an amp
we size the resistor to limit frame potential under a ground fault
our std is 15 A
loop R is monitored to be 3.5 Ohm
if it exceeds this the ckt trips
so frame v is 50 v when a ground fault occurs
if a 1000 Ohm person is a parallel fault path he sees 50 mA until tripping
according to Daziel curve for fibrillation threshhold
i = 175/sqrt(t)
i mA
t sec
175 varies with body mass, 175 is avg man
or t = (175/i)^2
if 50 mA t = 12 sec
we allow a max delay 1/2 sec

we trip at 50% so he only sees half (25 v or 25 mA)

 

[Ingenieur]

Sounds good then Do you trip or hold on a ground fault?

trip
but sometimes we'll hold to get the max a drive can put out
some drives at low freq <5 Hz can only supply 10 A of fault i
a sec or two max, dial a td into the relays
the equip has 6 level of sensing (7 counting the drive)
i and v on ngr
i on branch
3 x i on machine (in, out, gnd)
2 points of interruption, main cb and branch cb
we do the same testing on the vfd dc bus

 

[mbrooke]

The capacitive current is typically less than an amp
we size the resistor to limit frame potential under a ground fault
our std is 15 A
loop R is monitored to be 3.5 Ohm
if it exceeds this the ckt trips
so frame v is 50 v when a ground fault occurs
if a 1000 Ohm person is a parallel fault path he sees 50 mA until tripping
according to Daziel curve for fibrillation threshhold
i = 175/sqrt(t)
i mA
t sec
175 varies with body mass, 175 is avg man
or t = (175/i)^2
if 50 mA t = 12 sec
we allow a max delay 1/2 sec

we trip at 50% so he only sees half (25 v or 25 mA)

This is why we like having you here :happyyes:


Can you elaborate on what you mean by frame potential? Or how this concept works? I think I am following but slightly unsure. Wouldn't a lower value resistor increase frame potential?

 

[mbrooke]

trip
but sometimes we'll hold to get the max a drive can put out
some drives at low freq <5 Hz can only supply 10 A of fault i
a sec or two max, dial a td into the relays
the equip has 6 level of sensing (7 counting the drive)
i and v on ngr
i on branch
3 x i on machine (in, out, gnd)
2 points of interruption, main cb and branch cb
we do the same testing on the vfd dc bus

Sounds like a dream job in heaven :happyyes:

 

[Ingenieur]

This is why we like having you here :happyyes:


Can you elaborate on what you mean by frame potential? Or how this concept works? I think I am following but slightly unsure. Wouldn't a lower value resistor increase frame potential?

assume a machine with 3 ph wye (or delta w/zig-zag derived neut) plus egc
a phase faults to the frame of the machine
a fault i returns to the X0
with no ngr only limitation is conductor and xfmr Z
frame rises to ph-n v (bad for guy touching machine)
all vdrop is across the egc, vn on the frame dropping to 0 at the X0

Now interpose an ngr in the ngr-X0 path
Fault current is limited by the ngr
so now the v drop is divided across the egc and ngr and fault i << less than no ngr
if vph is 480 then vn 277
total loop = ngr + egc = 18.5 + 2.5 (for example) = 21 Ohm
i fault = 277/21 = 13.2 A
across ngr 18.5 x 13.2 = 244
on frame 277 - 244 = 33 dropped along egc up to ngr(simple kvl loop)
Much smaller shock potential
all done for personnel protection
these faults are common in mining since cables (no conduit) is used