Ungrounded plant

[pintofstout]

I hope I'm ok to be posting here. I'm an HVAC tech. I have been reading till my eyes bleed but so much of what I'm reading is above my head. I would really appreciate some customized help or a link so something I can understand. Honestly, reading about grounds in the stickies only destroyed what I already understood.

I have been working in a facillity and noticed I had a-b=460 b-c=460 c-d=460, but ground-a=10 ground-b=460 ground-c=460. It was like I was reading phase a right off the grounding lug of the chiller. I found a fan on a package unit with a phase to ground short, all returned to normal. This was my first experience with an ungrounded facility. In the interest of safety, nevermind due diligence for my customer, I need to understand more.

My questions

Initially, with the breaker and disconnect off, I read continuity (70k) on the line side of the disconnect, all 3 phases. What am I reading there?

This facility is fed by 3 substations that are grounded. During a phase to ground fault, the grounded phase is readable everywhere in the plant it seems. what prevents the phase from finding a ground elsewhere in the plant, specifically a person who may be in contact with something that is grounded?

What is the final destination of "spent" current (assuming you will let me get away with that). I hate to admit this but I always thought earth would provide a clear drain for current to move to. This forum killed that but I'm struggling to properly retool this fundamental.

If you guys know of any good reading that might help me I'd really appreciate it.

Thank You
Brian

 

[George Stolz]

Sure looks like you found a ground fault on the A phase.

All generated power is looking to return to its source via the other phases. Half of what leaves A looks for B, and half looks for C. With an ungrounded system, there is no connection between the electrical system conductors and the exposed metal of the system. When there is a ground fault the system becomes "corner grounded".

The continuity you read were the normal loads, more than likely.

 

[GoldDigger]

But since you described measuring this on the line side of the open disconnect (but on the load side of the open breaker??), those loads would have to be between the breaker and the disconnect. Possibly controls?
One thing that has not been mentioned is that the NEC requires an ungrounded system, such as the one you have, be equipped with a ground detector. This will notify you that a first fault has taken place so that it can be repaired before a second fault either creates a hazard or shuts the system down.

Tapatalk!

 

[petersonra]

Initially, with the breaker and disconnect off, I read continuity (70k) on the line side of the disconnect, all 3 phases. What am I reading there?

70k Ohms? L-L? L-G? did it take a while to settle in to the reading? Might just be capacitance in the wiring.

This facility is fed by 3 substations that are grounded. During a phase to ground fault, the grounded phase is readable everywhere in the plant it seems. what prevents the phase from finding a ground elsewhere in the plant, specifically a person who may be in contact with something that is grounded?

I don't know what you mean by this. If the secondary side of the substation transformers are grounded and you got a L-G fault downstream, something should have tripped somewhere.

What is the final destination of "spent" current (assuming you will let me get away with that). I hate to admit this but I always thought earth would provide a clear drain for current to move to. This forum killed that but I'm struggling to properly retool this fundamental.

Current always flows back to whence it came.

Some of what you are describing is hard to follow. It sounds like you have an ungrounded 480V delta electrical system and somehow a ground fault was introduced at a fan.

The reason ungrounded systems are used is so if this happens the system stays on. However, the code requires ground detectors on these type of systems so that you can tell if a ground fault occurs. usually it is three lights that are dim most of the time but if you have a ground fault two of them go bright and the third one goes off.

 

[pintofstout]

Thanks guys. I was implicitly taught that if electricity cannot return to its source, it can flow freely to ground. Cannot believe I've been so wrong for 17 years. So if the source is not grounded, then power will not flow to ground? Correct?

I'm not the only HVAC guy with this misconception. In fact I was challenged on the notion this morning already.

This system does have ground fault detection. When I alerted the engineer running the show, he said they had been looking for the fault all morning, they were aware of it.

As far as the line side continuity- the controls were on the other side of the open disconnect and any other loads were on the other side of the open disconnect. Should have simply been three wires sitting in conduit. Wonder if someone tapped another load in on that breaker.

I spoke to an electrician who had been involved in searching for the corner fault, and asked him about the continuity. He said well obviously there's continuity, this is an ungrounded system.

I didn't press, he wasn't giving me that opening.

 

[pintofstout]

Petersonra, I just noticed that you commented within my quote. Thank you, resistance readings were all taken l-g and did not settle or seem to charge like a cap would. Very clear continuity.

 

[jumper]

Thanks guys. I was implicitly taught that if electricity cannot return to its source, it can flow freely to ground. Cannot believe I've been so wrong for 17 years. So if the source is not grounded, then power will not flow to ground? Correct?

I'm not the only HVAC guy with this misconception. In fact I was challenged on the notion this morning already.

Correct, electricity/current does not flow to ground/earth/dirt.

The fact that you are willing to learn is a good step forward.

The difference between grounding and bonding is a concept that is misunderstood by many electricians and others in this field, not only HVAC techs by any means.

 

[texie]

I hope I'm ok to be posting here. I'm an HVAC tech. I have been reading till my eyes bleed but so much of what I'm reading is above my head. I would really appreciate some customized help or a link so something I can understand. Honestly, reading about grounds in the stickies only destroyed what I already understood.

I have been working in a facillity and noticed I had a-b=460 b-c=460 c-d=460, but ground-a=10 ground-b=460 ground-c=460. It was like I was reading phase a right off the grounding lug of the chiller. I found a fan on a package unit with a phase to ground short, all returned to normal. This was my first experience with an ungrounded facility. In the interest of safety, nevermind due diligence for my customer, I need to understand more.

My questions

Initially, with the breaker and disconnect off, I read continuity (70k) on the line side of the disconnect, all 3 phases. What am I reading there?

This facility is fed by 3 substations that are grounded. During a phase to ground fault, the grounded phase is readable everywhere in the plant it seems. what prevents the phase from finding a ground elsewhere in the plant, specifically a person who may be in contact with something that is grounded?

What is the final destination of "spent" current (assuming you will let me get away with that). I hate to admit this but I always thought earth would provide a clear drain for current to move to. This forum killed that but I'm struggling to properly retool this fundamental.

If you guys know of any good reading that might help me I'd really appreciate it.

Thank You
Brian

While I agree with the others that it sounds like you have an ungrounded delta supply that became corner grounded by the bad motor, one of your statements troubles me. Hopefully you are in error about the building supply being grounded at the substation. For an ungrounded system to exist and be safe the supply can not be grounded at any point. Some seem to think that just because they don't ground a 3 wire delta service at the building disconnect that they think they have an ungrounded system. That would only be true and safe if the POCO does not ground it at their end.

 

[kwired]

Thanks guys. I was implicitly taught that if electricity cannot return to its source, it can flow freely to ground. Cannot believe I've been so wrong for 17 years. So if the source is not grounded, then power will not flow to ground? Correct?

Unfortunately even some electricians do not understand this. Electricity is trying to return to another terminal of it's source, for a simple two wire source that means whatever leaves line A is trying to get back to line B. For a three phase system, whatever leaves line A is going to seek B, C or N if there is a N, of the same source. It only will seek earth as a path if there is another connection to earth to complete the circuit. An "ungrounded" system has no such other connection under normal conditions, and therefore when one system conductor does become connected to earth nothing happens for the most part other than the system does become temporarily grounded. Should another ground fault happen while this first temporary ground is in place is when you will get some objectionable current flow.

Ungrounded systems are quite safe, but do need a way of monitoring for ground faults or they can become pretty unsafe. They are typically used in processes where instant shut down because of a ground fault may be more harmful than than continuing to run - but indicating a fault - and the idea here is to shut down the process in an orderly fashion as soon as possible. If you don't remedy the problem as soon as possible, as second ground fault may instantly shut the process down anyway, which is trying to be avoided with the ungrounded setup in the first place.

Usually the ungrounded system is isolated to the process that needs to shut down in an orderly fashion and is not supplying the entire facility, things like HVAC wouldn't be typical to have on the ungrounded system, unless maybe heating or cooling is very critical to the process involved.

This is just kind of a start of the basics of an ungrounded system feel free to ask more questions if you don't understand something.

I spoke to an electrician who had been involved in searching for the corner fault, and asked him about the continuity. He said well obviously there's continuity, this is an ungrounded system.

I didn't press, he wasn't giving me that opening.

That remark was a little more of a smartass remark from him, he needed more information to give you a good answer.

 

[pintofstout]

Guys I can't thank you enough. I have learned a tremendous amount in a very short period here. It feels clear to me now.

Texie, I did not intend to suggest that the substation feeding the circuit in question was grounded. Their are 5 feeding the plant, 3 are grounded. This one is not.

HVAC is critical to the process, however I believe that if phase c goes to ground and there is no longer potential to the chassis, there will be control transformers loosing power on several at the same time. I need to have another look.

Now I need to consider how I can remove my hvac machines from the equation when they do ground a phase. I suppose fuses should open while the motor short phases, so I could give individual loads their own fuse block.

Seems like I'm begging to short another phase that way and invite a potential phase-phase short. Maybe then we just trip the breaker anyway?

The building engineer has a history of trouble with phase monitors, nuisance trips can be very detrimental to the overall process in this instance.

This makes me very concerned about the sustained balance, these big screw and centrifugal chillers need <2%, but he doesn't want them on some critical pieces. Voltages are very tight right now but we just got this plant, summer will see much more electrical load.

So if I can push just a little bit further....what methods are best for keeping my phases contained within the load, following a fault in that load?

Thank you again

 

[Smart $]

Petersonra, I just noticed that you commented within my quote. Thank you, resistance readings were all taken l-g and did not settle or seem to charge like a cap would. Very clear continuity.

Could be degraded insulation. FWIW, current leakage at 460V and 70kohms is only about 6.6mA. If 70kohm reading was taken with just a DMM, the electrician should megger those conductors before putting back in service... if nothing else but just to verify their integrity.

 

[Smart $]

...

So if I can push just a little bit further....what methods are best for keeping my phases contained within the load, following a fault in that load?

Local ground-fault protection.

 

[George Stolz]

Stout, all your transformers should be line-to-line. There is no circuit path line-to-ground.

(While there is a circuit path L-G in a grounded system, it's a code violation to use the ground as a neutral also, incidentally.)

 

[kwired]

Guys I can't thank you enough. I have learned a tremendous amount in a very short period here. It feels clear to me now.

Texie, I did not intend to suggest that the substation feeding the circuit in question was grounded. Their are 5 feeding the plant, 3 are grounded. This one is not.

HVAC is critical to the process, however I believe that if phase c goes to ground and there is no longer potential to the chassis, there will be control transformers loosing power on several at the same time. I need to have another look.

Now I need to consider how I can remove my hvac machines from the equation when they do ground a phase. I suppose fuses should open while the motor short phases, so I could give individual loads their own fuse block.

Seems like I'm begging to short another phase that way and invite a potential phase-phase short. Maybe then we just trip the breaker anyway?

The building engineer has a history of trouble with phase monitors, nuisance trips can be very detrimental to the overall process in this instance.

This makes me very concerned about the sustained balance, these big screw and centrifugal chillers need <2%, but he doesn't want them on some critical pieces. Voltages are very tight right now but we just got this plant, summer will see much more electrical load.

So if I can push just a little bit further....what methods are best for keeping my phases contained within the load, following a fault in that load?

Thank you again

Your control transformers will not care if an incoming lead becomes grounded. If they are 480 volt primary all they will care about is if there is 480 volts between the two input leads, they will not care one bit if either one is grounded or becomes grounded, so long as there is still 480 volts to the other input lead.

Now the results of what happens if there is a ground fault on the secondary of the control transformer is going to depend on if the secondary is a grounded system or not because it is a separately derived system and you kind of start all over again with a new source here.

 

[texie]

Guys I can't thank you enough. I have learned a tremendous amount in a very short period here. It feels clear to me now.

Texie, I did not intend to suggest that the substation feeding the circuit in question was grounded. Their are 5 feeding the plant, 3 are grounded. This one is not.

HVAC is critical to the process, however I believe that if phase c goes to ground and there is no longer potential to the chassis, there will be control transformers loosing power on several at the same time. I need to have another look.

Now I need to consider how I can remove my hvac machines from the equation when they do ground a phase. I suppose fuses should open while the motor short phases, so I could give individual loads their own fuse block.

Seems like I'm begging to short another phase that way and invite a potential phase-phase short. Maybe then we just trip the breaker anyway?

The building engineer has a history of trouble with phase monitors, nuisance trips can be very detrimental to the overall process in this instance.

This makes me very concerned about the sustained balance, these big screw and centrifugal chillers need <2%, but he doesn't want them on some critical pieces. Voltages are very tight right now but we just got this plant, summer will see much more electrical load.

So if I can push just a little bit further....what methods are best for keeping my phases contained within the load, following a fault in that load?

Thank you again

While ungrounded systems have there place and advantages, there are also a number of down sides. Your comments lead me to think this may be a rather large chiller. That would indicate manufacturer and engineering supervision would be advised regarding the power source. Does the manufacturer support this ungrounded power supply? Not all modern equipment is suitable for operating from an ungrounded system or even a corner grounded 3 wire system. Most equipment today of this type is supplied by a grounded Y source for a number of reasons.
Just thoughts from experience, not making any hard and fast assertions.

 

[George Stolz]

Fortunately for me, such selections are usually above my pay grade, and likely Stout's.

 

[pintofstout]

Ungrounded plant

So "corner grounded"...that's not reference to this system with a phase-ground fault, that's a system designed with one leg grounded huh?

I work for the manufacturer directly, they are ok with the system being ungrounded. The 2000 ton machines were designed for paper mills that use similar setups. They have soft starters, that's the key. Already called the factory guys on that one.

I have about 30 smaller tonnage chillers (-500) and two dual centrifugals that are the real workhorses. Most are on the system we are discussing, some are grounded.

So far as the 24v controls...the 24v side on some package a/c units has a terminal that goes right to the chassis, as well as out to the neutral of all the contactors and controllers. I'm not really able to visualize what happens if that chassis becomes part of one of the phases that feeds the xfmr. Seems like it would have an effect.

Is this system, being floating ground, definitely going to be a delta?

 

[pintofstout]

Kwired, to be clear I am by means disagreeing with you.

 

[Smart $]

...

So far as the 24v controls...the 24v side on some package a/c units has a terminal that goes right to the chassis, as well as out to the neutral of all the contactors and controllers. I'm not really able to visualize what happens if that chassis becomes part of one of the phases that feeds the xfmr. Seems like it would have an effect.

Shouldn't have any effect. The system is still required to have equipment grounding conductors. A phase to ground fault on an ungrounded system actually has less effect than one that is grounded. On the latter, because fault current is high until the overcurrent protection trips, there's voltage drop between the fault and system bond, which elevates local voltage on non-current-carrying-metal parts. With an ungrounded system, there's not enough fault current to eleevate the local voltage on non-current-carrying-metal parts.

Is this system, being floating ground, definitely going to be a delta?

Not necessarily... but it don't matter if there's only three line conductors run from the source transformer.

 

[pintofstout]

Not enough CURRENT to elevate VOLTAGE.

And now two steps back...

You know, I'll spend all the time it takes to read about this but I just can't follow some of this stuff.

As I thought about it- even if the chassis has a phase shorted to it from somewhere across the plant, the difference between the two leads of the xmfr is still 24v so it will continue to work.

That I can get on board with. But what you just said really spun me out. I'm still looking at something through a somewhat distorted lens.