Ungrounded Delta anomaly

[rsnell22]

130413-2207 EDT

rsnell22:

If you don't expect 280 V from each leg to ground, then what do you expect?

.

I know for sure that the secondary of the service transformer is a 3 wire delta, no XO, no system ground.
So, I expected the delta to be floating as it is at the point of origination.
The same thing I've seen for years on other ungrounded delta systems.

 

[ATSman]

While not common, I have seen ungrounded systems that had enough capacitive coupling that there was enough current to pull in a solenoid voltage tester when testing voltage to ground.

I second that comment.

 

[gar]

130414-1437 EDT

rsnell22:

So what voltage do you normally see from line to ground for an ungrounded delta? Is it 260, 240, 220, etc.? Is it balanced, or severely unbalanced? If you look at it with a high impedance meter, and then lower the impedance to possibly 10k ohms, what is the change?

Since you read voltages that would imply a wye source, and apparently not much change with meter loading, then does not this imply a wye source, or relatively balanced loads connected in a wye fashion with the midpoint connected to ground? The loads might be stray capacitance, or leakage loads, or some thing else that has a low enough impedance that meter loading does not change the voltage much.

You mentioned a transformer to supply of 208/120 wye. Could this have a wye primary with the center point sufficiently connected to ground so that the meter voltages to ground remained close to balanced with only the meter load.

What you don't have is a center tapped 480 secondary with the center tap connected to ground.

.

 

[hurk27]

Duly noted. However, if the first fault is never cleared, as sometimes happens, what happens later on when there is a second fault?

In an ungrounded delta the first fault is not supposed to be cleared, a second fault will cause a OCPD to open, this is the whole idea of not having a production line or operation from crashing down because of a fault

Ungrounded deltas are not safer and should have a ground detection system.

I don't see an ungrounded delta as a safety problem as some do, you still have to respect there might be a voltage between a phase conductor and the grounding system just like in a grounded system, and if a phase does go to ground then you have nothing more then a grounded system, most of us who work on these systems will not touch a phase conductor and the EGC, so how is it any more dangerous then a grounded system? the grounding system will still have 0 volts potential to Earth even after a phase faults to ground just like in a grounded system but even less since there is no grounding path back to the transformer for current to cause a voltage drop on it.

But if the utility makes the mistake of bonding the X0 or a corner at the transformer without requiring a grounded conductor to be brought to the service, you now have a difference of potential through Earth that if a phase conductor was to fault to the systems grounding it would now have the potential of the bond at the transformer, Earth would never clear the fault, and the grounding of the building would be at this potential.

There will be capacitive coupling without the solid ground, so reading 277 (or thereabouts) isn't unheard of and is to be expected.

Agree

RUS has no specs on ungrounded deltas.

Who's RUS?

 

[rsnell22]

It has been my experience that with a properly maintained ungrounded Delta that there is virtually nothing to ground other than whatever overall system capacitance might create. And with those, a solenoid type tester bled it right down to zip.

But I've also seen high and low voltages to ground on some of them, due to various problems. One was tracked down to a bad motor, another a skinned wire. But on all of them the voltage to ground wasn't even across all three phases, like I'm seeing on this one. And all the others had at least ground indicator lamps.

Maybe it has to do with the 208Y120 stepdown transformer. That is the wild card here, in my best guess.

Thanks for all the good input so far.

 

[hurk27]

It has been my experience that with a properly maintained ungrounded Delta that there is virtually nothing to ground other than whatever overall system capacitance might create. And with those, a solenoid type tester bled it right down to zip.

But I've also seen high and low voltages to ground on some of them, due to various problems. One was tracked down to a bad motor, another a skinned wire. But on all of them the voltage to ground wasn't even across all three phases, like I'm seeing on this one. And all the others had at least ground indicator lamps.

Maybe it has to do with the 208Y120 stepdown transformer. That is the wild card here, in my best guess.

Thanks for all the good input so far.

True in most cases you would have only capacitive couplings and if these are not long runs should have very little current, but one more food for thought, I had just remembered an installation where we were getting a solid WYE reference and was pulling our hair out as to where it was coming from, we had turned off all SDS's that might had a WYE primary and still had the voltage, it wasn't until we shut down an electric drier that we lost this WYE voltage, when we opened it up we found it had 277 volt elements that were connected in a WYE and for some reason they bonded the center point to ground ( yes the manufacture had told them to wire it like this), this gave the whole system a reference to a WYE X0 even though there was none, any equal load resistive or inductive that is connected into a WYE will produce a WYE reference if the center point is also bonded to the grounding system.

 

[jim dungar]

...this gave the whole system a reference to a WYE X0 even though there was none, any equal load resistive or inductive that is connected into a WYE will produce a WYE reference if the center point is also bonded to the grounding system.

For the curious, they can look up articles on 'artifical neutrals'.

 

[jumper]

For the curious, they can look up articles on 'artifical neutrals'.

Ok, I am curious and will do. Thanks:thumbsup:

 

[Hv&Lv]

Who's RUS?

Rural Utilities Service, a subdivision of the USDA. Cooperatives are governed by this organization if they borrow funds. This is also the organization that says the NESC is mandatory if you borrow from them.

 

[hurk27]

Rural Utilities Service, a subdivision of the USDA. Cooperatives are governed by this organization if they borrow funds. This is also the organization that says the NESC is mandatory if you borrow from them.

A'hhh a CO-OP we have one here to our south, Kankakee Valley

I find it strange they don't have any spects on ungrounded systems as even in the C0-OP areas they still have industrial installations that can or will have ungrounded deltas?

And having the misunderstanding of the danger of replacing a transformer bank at one of these facilities and grounding the secondary without supplying a grounded conductor (making it a grounded supply) to the facility can and will put lives in jeopardy.

Are you saying that the NESC requires that all transformer secondaries are required to have the secondaries bonded even if the transformer is supplying a facility that is using an ungrounded system? if so this is a grave error on the NESC part?

Because by doing so you are putting the Earth as a part of the current path from the building back to the transformer and while you might drop 75% of the voltage at the electrode point touching the building metal siding or lets say an A/C unit that is over 26 feet away from the grounding electrode you will see the full potential of the phase to ground voltage to Earth.

If the secondary is left isolated then there is no such path and the grounding electrodes at the building will be the reference to earth for the lightning strike path as well as the reference for the buildings service even if a first fault were to occur second fault will be a phase to phase fault which will open the OCPD

Also another problem is any other services such as cable or telephone, by having a transformer reference to the MGN at the pole the cable and phone will take the fault path of a phase to ground fault in the building, this will place the fault current on the shield of these conductors which can cause a fire, again if the transformer is left isolated there is no such path for the fault current.

Sorry for repeating myself on this but it hit a very big red flag with me with your statement of bonding of a replacement transformer on a ungrounded service.

 

[kwired]

I think the OP has a wye connected element somewhere with the mid point bonded, either in the source or a load somewhere.

 

[Hv&Lv]

A'hhh a CO-OP we have one here to our south, Kankakee Valley

I find it strange they don't have any spects on ungrounded systems as even in the C0-OP areas they still have industrial installations that can or will have ungrounded deltas?



Sorry for repeating myself on this but it hit a very big red flag with me with your statement of bonding of a replacement transformer on a ungrounded service.

I enjoy reading your objections, it helps me to understand your concerns. I take offense to very little. Look at this and see if you interpret it the same way I do.
From the NESC, section 9, rule 092.

B.Alternating Current Systems That Are to Be Grounded

1.750 V and below
The point of the grounding connection on a wye-connected three-phase four-wire system, or on a single-phase three-wire system, shall be the neutral conductor. On other one-, two-, or three-phase systems with an associated lighting circuit or circuits, the point of grounding connection shall be on the common circuit conductor associated with the lighting circuits.

The point of grounding connection on a three-phase three-wire system, whether derived from a delta-connected or an ungrounded wye-connected transformer installation not used for lighting, may be any of the circuit conductors, or it may be a separately derived neutral.

The grounding connections shall be made at the source, and at the line side of all service equipment.

 

[hurk27]

I enjoy reading your objections, it helps me to understand your concerns. I take offense to very little. Look at this and see if you interpret it the same way I do.
From the NESC, section 9, rule 092.

B.Alternating Current Systems That Are to Be Grounded

1.750 V and below
The point of the grounding connection on a wye-connected three-phase four-wire system, or on a single-phase three-wire system, shall be the neutral conductor. On other one-, two-, or three-phase systems with an associated lighting circuit or circuits, the point of grounding connection shall be on the common circuit conductor associated with the lighting circuits.

The point of grounding connection on a three-phase three-wire system, whether derived from a delta-connected or an ungrounded wye-connected transformer installation not used for lighting, may be any of the circuit conductors, or it may be a separately derived neutral.

The grounding connections shall be made at the source, and at the line side of all service equipment.

I see this is from "B.Alternating Current Systems That Are to Be Grounded" do they have a section for B.Alternating Current Systems That Are not required to be grounded?

Because who ever wrote the above code had to have their head stuck in the sand, you can not ground the supply transformer secondary of an ungrounded system without creating a dangerous voltage potential to Earth on all the grounding of the building if a first fault were to happen, keep in mind that a first fault is supposed to give the operators time to shut down the equipment at a more convenient time this could be a week later, which doesn't matter, having a difference of voltage potential on the grounding and Earth for even one second is un acceptable because someone could be touching something grounded and Earth when it happens.

I'm not trying to give any offense, just trying to point out something that would be very dangerous and deadly if it was allowed to happen.

This same problem happens when we lose a neutral to a house that has no water pipe to another service, except we are talking about 120 volts not 480, still dangerous just the same, the difference is we are using the neutral to maintain 120 volt balance for 120 volt loads, so most times the voltage will be in the range of 10-30 volts unless the service is very unbalanced.

 

[mivey]

I enjoy reading your objections, it helps me to understand your concerns. I take offense to very little. Look at this and see if you interpret it the same way I do.
From the NESC, section 9, rule 092.

B.Alternating Current Systems That Are to Be Grounded

1.750 V and below
The point of the grounding connection on a wye-connected three-phase four-wire system, or on a single-phase three-wire system, shall be the neutral conductor. On other one-, two-, or three-phase systems with an associated lighting circuit or circuits, the point of grounding connection shall be on the common circuit conductor associated with the lighting circuits.

The point of grounding connection on a three-phase three-wire system, whether derived from a delta-connected or an ungrounded wye-connected transformer installation not used for lighting, may be any of the circuit conductors, or it may be a separately derived neutral.

The grounding connections shall be made at the source, and at the line side of all service equipment.

IF the system is required to be grounded, then that section of the NESC identifies the grounding point. It does not say that the system has to be grounded.

092. Point of connection of grounding conductor...
A. Direct-current systems that are required to be grounded...

B. Alternating current systems that are required to be grounded...

C. Messenger wires and guys
1. Messenger wires
Messenger wires required to be grounded shall be connected to grounding conductors at poles or structures at maximum intervals as listed below:...

E. Fences
Fences that are required to be grounded by other parts of this Code shall...

And I will refer you to

314. Grounding of circuits and equipment
C. Circuits
1. Neutrals
Primary neutrals, secondary and service neutrals, and common neutrals shall be effectively grounded.
EXCEPTION: Circuits designed for ground-fault detection and impedance current-limiting devices.

Which says you do not have to ground neutrals on systems designed for ground fault detection.

 

[mivey]

Because who ever wrote the above code had to have their head stuck in the sand.

Nope. The code was just mis-interpreted.

 

[rsnell22]

Earlier I referred to a 208Y120 dry step down transformer with the three wire delta as it's high side.

Today I had time to go have a look briefly at the transformer. The nameplate show it as a 3 wire delta on the 480 side, and the typical 208Y120 on the secondary.

So there's still a wild card somewhere in the 480 system.

 

[hurk27]

Earlier I referred to a 208Y120 dry step down transformer with the three wire delta as it's high side.

Today I had time to go have a look briefly at the transformer. The nameplate show it as a 3 wire delta on the 480 side, and the typical 208Y120 on the secondary.

So there's still a wild card somewhere in the 480 system.

Well if loading the phase conductors to the grounding system requires a bit of loading to force an imbalance then I would say it would have to be a WYE load of some kind, about the only way I would know to find such a load would be turning off loads until you loose the Wye balance which is how we found the drier with three 277 volt heater elements tied into a WYE and center point bonded to the grounding system, the bad thing was that if there was ever a first fault to ground by a phase conductor two of the elements would have seen the full 480 volts which would have either taken out its OCPDs or the elements, we unbonded the WYE point and everything was fine, the manufacture was also notified of their instructional mistake.

I still don't see where this should have caused a problem with the circuit board, most electronics are damaged by over voltage or heat or both, proper design of such an expensive board should include over temp shut down and onboard SPD's to prevent voltage spikes from damaging components, had a lightning strike been ruled out? even a near field event can product current into I/O wiring that is not properly shielded, knowing what failed on the board is the first clue to what might have caused the damage, I have done some work at finding damage causes in electronics and there are some tell tale things that can point to a cause if the person who is doing the research knows what to look for.

like if we find I/O circuit damage then I look for shielding bonded at both ends for parallel current problems where input or output devices might be across areas fed from two or more SDS transformers, if this pans out that its not the case then I suspect a near field lightning strike or LEMP that can induce circulatory current into these conductors, if we see power supply component damage then I look closely at the supply source and possible of inductive loads that may produce transients into the power supply system?

May I ask the purpose of this board if it doesn't reveal any trade secrets? you said this is for a chiller which leads me to believe it might be control system to regulate control over temp and flow of the material which could be water or hydraulic fluid or some other medium, this would lead me to believe you have some 4-20ma I/O's as well as maybe some TC's that could be susceptible to induced voltages if the conductors are long, also does it have some VFD motor control and is there proper filtering in the input and output of these VFD's? as VFD's can cause voltage spikes many times the supply voltage.
And as I said before most VFD's do not like being on an ungrounded system.

The above is just a small example of what it may take to find the cause of the board failure, I have dealt in the past with many pieces of equipment that the engineer of it just didn't have an open enough mind to cover all aspects that may cause premature failures as every installation can be different each having its own unique set of design problems that the engineer never took into account at the time of the system design.

A fix could be as simple as having a transformer installed just for this chiller so its secondary can be grounded, not a cheap option but one never the less.

just trying to give you some ideas to work with.

 

[mivey]

So there's still a wild card somewhere in the 480 system.

See gar's post #7:

The next test might be to put a somewhat larger resistive load than just the meter from one leg to ground and observe the voltage change from that leg to ground. Basically checking the source impedance. If there is some ground fault sensing of some sort stay below that current level.

 

[Jraef]

Could be that your chiller, or something else in the system, has surge protection that is configured in a Wye to ground and nobody paid attention to the fact that it was an ungrounded delta system. Usually the MOVs fail open, but sometimes on low level surges they short, in which case the surge protector becomes a defacto Wye point for your entire system. Happens quite a bit with VFDs, especially those of European design where they never have ungrounded delta systems. Usually there are instructions in the manual to remove the ground connection in that case, but often times people like HVAC technicians have no idea what ungrounded delta means, so they don't do it.

I would start by opening up the disconnect to the chiller, then test the line side voltage and see if you still have the anomaly. If not, it's in the chiller circuit. If it's still there, look for VFDs in the HVAC system and do the same with them. Cheap test, albeit it could be a lot of time if the system is big.

 

[bure961]

We have a delta ungroundedservice with ground detection sysem(simple pilot lites.We also have a standby gen that has a (i believe) wye feed . This could be problems?