motor frame grounding conductor sizing 7

[electricpete]

The NEC states that equipment grounding conductor should be sized according to the rating of the overcurrent device feeding the equipment.

It seems like the logic is that in event of a fault within the motor causing ground current to flow in equipment ground conductor, we want the cable to be protected over a range of time and currents that is bounded by the protection. i.e. the motor must trip before the cable melts.

From that logic, should we be permitted to use much smaller grounding conductors for motors that are protected by sensitive ground fault relaying?

 

[DanDel]

You may have a point, but don't count on the NEC to change. Also per 430.52(B), 'the motor branch-circuit short-circuit and ground fault device shall be capable of carrying the starting current of the motor.', so you can't really set the GF for a very sensitive value.
Besides, cable melting would occur from high fault current which would be taken out by the instantaneously-operating short-circuit protection before the GF time delay expired.

 

[busbar]

electricpete—that’s an interesting proposal, but I think that it’s not something that consensus standards folks [NFPA…IEEE] would be willing to buy into. Taking the conservative approach, with the typical limitation of not needing to be larger than the phase conductors, one aspect that would not be effected by relaying is the magnitude of potential difference between equipment frames until protective relaying operates to clear the fault. [In the whole scheme of things, equipment grounding/bonding is typically a small fraction of electrical-installation costs.]

Accepted practices vary by voltage level, but I think most conscientious engineers would review limiting I²t as to not exceed the usual 250°C “emergency” copper-conductor rating, which is probably the weakest link in the chain.

 

[electricpete]

Thx for comments, I'll need awhile to digest them.

By the way, I'm not trying to cut corners on a new installation. Just questioning an existing installation.

 

[Shortstub]

I think the essential point of grounding-cable sizing has been missed!

Cable sizing is not related to fault-clearing time, but instead, to personnel safety. The voltage-drop from machine-to-ground must not exceed the touch-potential considered dangerous to personnel!

 

[Shortstub]

A related point. Grounding-conductors should not be placed in magnetic enclosures, like steel conduit. And, if it is, then, the conductor must be bonded where it enters and leaves the conduit!

Thermalstation's comment regarding European practice is as close as one can get to a scientific approach. Think of the touch-voltage problem as a voltage divider consisting of the phase-conductor and the ground-return path. The point of connection is the motor's carcass/frame.

 

[electricpete]

Hi Shortstub - You may be right, but it is still not a clearcut logic to me.

I am trying to discern the basis/logic (and resulting application guidelines) for the NEC requirement which ties ground conductor size to protective device setting.

Article 250.122 requires ground conductors to be sized according to the "rating or setting of the overcurrent protective device in the circuit ahead of the equipment..."

For 15A rating/setting it requires 14AWG copper
For 20A rating/setting it requires 12AWG copper
For 30A rating/setting it requires 10AWG copper
etc

I honestly do not know what the basis of these numbers are.
Clearly they are not to protect against conductor melting as I suggested since they are so low.

It seems to me that they are likely intended to protect the ground cable itself, since they are almost identical to the protection which would be applied to insulated wire in free air (although I agree ground conductor doesn't need as much protection as insulated conductor since it has no insulation to damage).

Is it your belief that these numbers are based instead upon steady state voltage drop? If so then shouldn't there be a limit on the length of ground conductor in addition to the size?

A related question for the group. Let's say I have an industrial motor fed from load center with the following protection:
residual-connected ground protection 51G - 10A
Time overcurrent - 200A
Instantaneous - 1200A.
(All three currents are expressed on primary side of CT)
Which of these three settings would we pick as the basis for the ground conductor?

 

[electricpete]

Correction:
"It seems to me that they are likely intended to protect the ground cable itself from steady state overload..."

 

[Shortstub]

Electricpete,

The presumption is, that because both the phase- and grounding-conductors have to conform to the protective device size criterion, then their per unit length impedances are nearly equal. Therefore, the grounding-conductor voltage-drop will always be smaller than the phase conductor's voltage-drop. Thus, by voltage-divider effect, the touch-voltage will be smaller.

Of course, for the above to be true, then the current flows must be equal in both conductors, i.e., phase and grounding.

 

[electricpete]

I'll tell you one thing is I am thinking about a power plant environment where we have a grounding grid and equipment safety ground is connected to structures attached to grounding grid... does not have to run back toward the power supply. Maybe that is different than most equipment covered by the code?

Even if we did have a grounding conductor running the same distance as the phase conductor, what you are suggesting is that a voltage divider which provides approx equal voltage drop in the phase conductor and ground conductor will ensure the voltage does not rise to dangerous levels? I have a very hard time believing that we are relying on votlage drop in the phase cable to reduce voltage to safe level in event of a fault... it is just not reasonable (would 1/2 of line-to-ground voltage be safe?). I can envision that we might take the fault current level and multiply by the ground cable impedance to determine voltage which a human may be exposed to, but the table does not account for fault current level or resistance (no lenght specified). Without a lenght it seems like it must be looking at thermal effects, not resistance (My opinion)

 

[DanDel]

I believe voltage drop in this discussion (touch potential, or touch voltage) should refer to the voltage between the faulted motor frame and any other grounded equipment in the close vicinity which can be 'touched' simultaneously by personnel, not between the phase and ground return cables at the power source.
The theoretical voltage divider should be between the faulted frame(zero volts) and any nearby grounded metal components(measured as the voltage drop across the grounding connection through which the fault current is traveling). The better the bonding and grounding between these components, the lower the potential between them, and the less risk to personnel near them.
This is also the reasoning behind 'step voltage', and why there is a ground grid to bring the potential of the ground to near-equal values everywhere in a switchyard.

 

[electricpete]

Hi Dan. Are you saying that you also think the basis for the NEC guidelines is associated with providing a sufficiently large cable to minimize step potential or voltage drop across cable due to fault current, and not thermal considerations?

 

[Shortstub]

Electricpete,

Im not speaking for DanDel, but it is what I mean!

 

[DanDel]

'pete, I believe it is probably a combination of several factors, including both of these you mention. Besides, these two factors(voltage drop and thermal damage from fault current) are two sides of the same coin.
A long time ago, the prevailing school of thought was to isolate all electrical equipment and not ground anything. This was supposed to limit exposure to dangerous conditions. Eventually, grounding and bonding came to be the standard, so that if something went to ground, a solid ground connection would assure that the protective device would trip and de-energize the fault.

 

[ielivaz]

For a grounding wire, I use as thumb rule half the cooper in the phase wire and always a naked wire, because there is no rubber or PVC to be melted, it makes possible to support very high temp. (350 C)

Also, grounding must be as close as possible to the electrical device.

 

[GordS]

'pete...In Canada, we have a similar ruling. The rationale that CSA provides ( in a book that supplements the Canadian Electrical Code)is the bonding conductor size is determined by 2 factors - big enough to carry the current and big enough that it does not excessively increase the impedance of the ground return ciruit.

 

[peebee]

ielivaz:

Your half-size rule will put you in violation of NEC for many installations.

Regarding the ground-wire insulation: you'll just end up burning the insulation on your phase conductors instead (you are running that bare ground wire with your phase conductors, aren't you?). Also, your ground wire will corrode much faster without insulation. There's no code issue with using bare wire, just maybe not as much advantage as you think.

 

[electricpete]

Thanks Peebee - you have stated an obvious point that I had overlooked. If the ground conductor is run with phase cables than it's loading/temperature would need to be controlled just the same as an insulated conductor and it would have to be protected at the same level.

From my perspective it now seems very logical that the basis for these ground conductor ratings are thermal protection of cables. The other comments are welcome and appreciated but I just don't see how we could hope to address touch/step potential issues without considering the fault current level and length of the ground cable.

More discussion is welcome if you think I'm missing soemthing.
Thx

 

[busbar]

Any edition of IAEI Soares' Book on Grounding is fifty bucks well spent.

 

[peebee]

Wow, thanks Pete -- A compliment from you, jbartos or busbar is better than ten eng-tips red stars in my book.

BTW, jbartos, congrats on yet another well-deserved re-election as TipMaster.