480v Fault at heat pump blows phone lines

Status
Not open for further replies.
L

labigriver

Member
My building has a 277/480v service with solidly grounded neutral at the padmount xfmr (by the utility) and also grounded at the panel by ground rod. This service feeds 277v lighting load, 480v pump motors, 480v hoists, and two 10ton 480v heat pumps. It also serves a three phase 277/480 to 120/208 transformer which feeds office area with 120 volt receptacles, lighting, and kitchen area, etc. The 120/208 transformer neutral (X0) is solidly grounded by ground rod and the neutral and gnd bus in the 120/208 panel are bonded together. This morning we found the surge protectors on all three incoming phone lines blown. Phone company stated the boards in their distribution center were blown out. We found two wires on the 480v fan motor in one of the heat pumps burned really bad and wires melted from terminals but voltage still present on the wires. About this same time someone inside the building connected two gnd points (by coincidence) together with a condutor and drew a very large arc which in turn blew one of the fuses in the heat pump. The heat pumps have only terminals for the three phase conductors and an equipment ground, not a neutral. Is this correct and if so could someone please explain the theory of what happens with a phase to case gnd fault at the heat pump location? Is it possible the heat pumps were designed for a 480v delta instead of what we have? Any assistance greatly appreciated.
 
Are you referring to the bond in the 480 panel? I forgot to mention building is metal with emt conduit from 480 panel to fused disconnect mounted on outside wall to feed heat pump. From fused disconnect conduit is pvc to heat pump. Would this make any difference?
 
This morning we found the surge protectors on all three incoming phone lines blown. Phone company stated the boards in their distribution center were blown out.
Did you loose any office equipment or have any trips on 120/208 CB's and GFI's? I assume the phone lines did not supply a phone system such as Merlin or the like, or this system if installed is undamaged.

We found two wires on the 480v fan motor in one of the heat pumps burned really bad and wires melted from terminals but voltage still present on the wires.
Did the O/L's trip and someone reset them? Did the motor MEG out ok? Was the wire sized correctly?

About this same time someone inside the building connected two gnd points (by coincidence) together with a condutor and drew a very large arc which in turn blew one of the fuses in the heat pump.
Do you happen to have taken a Amp reading on the ground compared to the grounded conductor?

The heat pumps have only terminals for the three phase conductors and an equipment ground, not a neutral. Is this correct and if so could someone please explain the theory of what happens with a phase to case gnd fault at the heat pump location?
A dead short such as in a motor, phase to ground, should trip the CB, and is potential danger to reset without prior investigation. If your connections were overheating prior you may have experanced O/L trips prior. If 2 of 3 phases burnt open at the same time the both O/L and CB may not have opened the circuit. I have seen this caused due to a outside forces such as a forklift load moving a conduit, pulling wires loose. {Sorry little theory}

Is it possible the heat pumps were designed for a 480v delta instead of what we have? Any assistance greatly appreciated.
If heat pumps were designed for delta they have 3 Ph connections and a ground. I think you stated this. A "Y" heat pump would have 3 Phase connections, a grounded conductor (N) connection, and ground. Its ok to connect a Delta load such as a motor to a Y service, just DON'T use a (N)...
 
No office equipment damage and no breaker trips on the 120/208 panel. Have not megggered motor or taken amp readings as of yet. Will start troubleshooting in a couple of hours when we can do what we wish with the branch circuits. The phone line protectors are on the phone patch panel at phone line entrance to building and connected from each line to gnd rod below patch panel. Phone company says we had a rise in that gnd potential which cased our protectors to trip and damaged their equipment. I will be checking wire size via code book but I believe they are correct. No one was in building overnight when we think this occured. Came in and found phones not working and the fan on one of the heat pumps not running. Hence we started looking at the heat pump.
 
is this close?

is this close?

check the ground wire at the heat pump back to the source and i bet you find that the connection was not good, so when the fan motor failed i did not open the ocpd.

i speculate that the fan motor failed and caused the primary fault conditon,
then the persons in the bldg working on the grounds, with the potential from the fault on the ground wire they either disconnected the grounds or connnected them causing the fault on the phone lines.
 
labigriver said:
About this same time someone inside the building connected two gnd points (by coincidence) together with a condutor and drew a very large arc...
Click to expand...

Theres the clue; one of those ground points isn't. Solve that and the rest will fall into place.

Be careful; a ground that isn't a ground may bite you unexpectedly!
 
dbuckley said:
Theres the clue; one of those ground points isn't. Solve that and the rest will fall into place.

Be careful; a ground that isn't a ground may bite you unexpectedly!
Click to expand...

Excellent responses, thanks everyone. We found the 480v panel gnd bar and neutral were not bonded. We then measured the resistance from the 480v neutral at the padmount xfmr to the gnd rods (and gnd bars) at both panels and found about 12.5 ohms. If I am thinking correctly the 277v faulted phase could have only "pushed" roughly 22.5 amps back to the padmount service transfomer through the earth resistance to the xfmr neutral. This is below the 30amp fuses on the motor circuit in the heat pump. When the employee touched the two gnd's he was making a circuit from the equipment gnd directly to the padmount neutral bypassing the earth resistance and allowing full fault current to flow. At this time a 30amp fuse in the heat pump blew. Before this the gnd potential in the building rose to the faulted phase voltage and blew the phone system trying to push 277v fault current through the earth. Does this sound reasonable to everyone or am I off the mark here?
 
labigriver said:
Excellent responses, thanks everyone. We found the 480v panel gnd bar and neutral were not bonded. We then measured the resistance from the 480v neutral at the padmount xfmr to the gnd rods (and gnd bars) at both panels and found about 12.5 ohms. If I am thinking correctly the 277v faulted phase could have only "pushed" roughly 22.5 amps back to the padmount service transfomer through the earth resistance to the xfmr neutral. This is below the 30amp fuses on the motor circuit in the heat pump. When the employee touched the two gnd's he was making a circuit from the equipment gnd directly to the padmount neutral bypassing the earth resistance and allowing full fault current to flow. At this time a 30amp fuse in the heat pump blew. Before this the gnd potential in the building rose to the faulted phase voltage and blew the phone system trying to push 277v fault current through the earth. Does this sound reasonable to everyone or am I off the mark here?
Click to expand...
Sounds very reasonable, Had a problem a few years ago on a brand new install where cabinet in a furnace control panel was not grounded. One phase of a burner magement controller shorted to ground through a bad MOV. Made everything else in the panel 208 instead of 117. Let the magic smoke out of a lot of stuff. Fortunately no one touched the panel and a real ground at the same time.
 
Please check the grounds at the heatpump

Please check the grounds at the heatpump

The item that concerns me most is the part where when two grounds were touched togeather with a jumper an arc was produced. If the motor had 2 phases open and didn't trip CB or blow any fuses prior to this event. The path of the single phase in the motor returning through ground is concerning. VERIFY its ground is connected all the way to the panel. If the motor megs good, You may have larger issues.
 
VinceS said:
The item that concerns me most is the part where when two grounds were touched together with a jumper an arc was produced. If the motor had 2 phases open and didn't trip CB or blow any fuses prior to this event. The path of the single phase in the motor returning through ground is concerning. VERIFY its ground is connected all the way to the panel. If the motor megs good, You may have larger issues.
Click to expand...


See thread #9, The gnd is connected all the way to the panel but at that point it stopped since the gnd to neutral bonding was not in place in the panel. When the two "gnds" were touched the employee was actually making the same connection as the bonding jumper would have been if it were in place in the service entrance panel as it should have been. He was essentially making a connection from the 480v service entrance panel neutral to the gnd bus in the panel, which is where the equipment gnd was connected, this was a much better path for current than having to go through the earth resistance.
 
Yes, I understand... but

Yes, I understand... but

I understand what could happen when the main bonding is missing. Are you no longer concerned how your 480VAC motor opened 2 of three connections without tripping, a CB, Fuse, or O/L, or are you attributing the missing bond as the answer to the motor issue also?
 
Status
Not open for further replies.
Top