Digital Grounding?

[Nicolas2]

Good afternoon.
I participate in the design of a plant whose electrical buildings are sharing space electrical equipment and I&C equipment (sensitive loads).
For sensitive loads we have included surge protection and isolation transformer, and follow the NEC Exhibit 250.13 connection schematic. Sensitive loads have a separate grounding which is connected to general grounding grid, so it is not an insulated grounding. Now, some people on site tell they want a "digital grounding" which is an insulate green conductor, going to an insulate grounding electrode, and they say this is a normal practice here in USA. But, cheking the Erico grounding practical guide, says:

"An unnecessary conflict can be created however, such as when someone attempts to create a "separate","dedicated" or "clean" grounding connection that is
not permitted by the NEC!"

So, my question is, am I misinterpreting the NEC and this is a general practice, or the design is correct and I can't install a "clean grounding" in USA? I know in IEC countries is mandatory, but USA is IEEE, and I am less familiar with IEEE.

Thank you

 

[winnie]

You are _never_ permitted to completely isolate a ground, in the sense that you cannot have a completely isolated connection back to soil.

In fact, doing so is dangerous and likely will result in worse problems with equipment. Consider, for example, what happens in the event of any current flowing through the soil because of power distribution systems or because of nearby lightning strikes; current could flow 'into' one ground electrode and 'out' of another, _through_ the sensitive equipment which you are trying to protect.

There is something called 'isolated ground' which is permitted. This involves the use of insulated 'ground' wires which are kept separate from 'common' ground conductors such as conduit and the like, all the way back to the main neutral-ground bond back at the service location or at the transformer supplying the circuit. At the main service point location this 'isolated' ground is connected with the rest of the ground system.

-Jon

 

[Nicolas2]

So, at the end, is almost the same, my separate grounding conductors are joined to general grounding systme out of the building.

Thank you

 

[winnie]

Yes. However the whole concept of a 'clean' ground is flawed at its core. It stems from the FALSE idea that the soil is some sort of infinite sink for electricity, and that connecting to the soil has magical properties of protection.

If your 'sensitive equipment' actually requires a quiet ground, then this must be achieved not by complete isolation from other grounded metal, but rather by careful selection of the grounding and bonding circuit topology to make sure that problematic currents and voltages are not imposed upon sensitive segments of the 'grounded' metal.

For example, some communications devices use the voltage on a signal conductor relative to a 'ground' conductor as the method of passing information between two devices. If each of these devices shares that 'signal ground' with its power supply 'safety ground', then you have a situation where the electrical wiring ground is connected at two locations (the two computer devices) via a very sensitive communications conductor.

While you can apply lots of band-aids in the form of trying to reduce the noise on the safety ground conductors, what you should really do is use a proper differential pair communications circuit, and not depend upon a quiet ground for reliable signal communications.

Please also see: https://electrical-engineering-port...lated-common-and-integrated-grounding-systems

All the best,
Jon

 

[kwired]

Yes. However the whole concept of a 'clean' ground is flawed at its core. It stems from the FALSE idea that the soil is some sort of infinite sink for electricity, and that connecting to the soil has magical properties of protection.

If your 'sensitive equipment' actually requires a quiet ground, then this must be achieved not by complete isolation from other grounded metal, but rather by careful selection of the grounding and bonding circuit topology to make sure that problematic currents and voltages are not imposed upon sensitive segments of the 'grounded' metal.

For example, some communications devices use the voltage on a signal conductor relative to a 'ground' conductor as the method of passing information between two devices. If each of these devices shares that 'signal ground' with its power supply 'safety ground', then you have a situation where the electrical wiring ground is connected at two locations (the two computer devices) via a very sensitive communications conductor.

While you can apply lots of band-aids in the form of trying to reduce the noise on the safety ground conductors, what you should really do is use a proper differential pair communications circuit, and not depend upon a quiet ground for reliable signal communications.

Please also see: https://electrical-engineering-port...lated-common-and-integrated-grounding-systems

All the best,
Jon

And many modern protocols use proper differential pairs compared to many years ago. Might still be some specifications at times for "isolated ground circuits" but the need for them isn't as great as it once was.

 

[paulengr]

Truly isolated grounds are used in the US and are legal and in fact required in mining. The situation is that the mine substation is subject to lightning strikes and utility transients and similar disturbances compared to equipment in the pit or underground. So the station and load grounds are separated by some distance usually 25-50 feet. Equipment on the primary side of the transformer are grounded to the station ground. Equipment on the secondary side is grounded to the equipment ground. Since the transformer is usually delta-wye it blocks zero sequence currents like transients which are coupled to the station ground.

Not only that but utilities typically do not supply bonding (share grounds) and separately derived systems can’t be separate if they share grounds. Never mind fairly complicated grounding issues with multiple generators.

Have to disagree with the statement that there is no such thing as an isolated ground. Over very short distances (under twice the length of a ground rod) you get coupling. So agreed they are not isolated. Away from this ground resistance is inversely proportional to distance. So at distances of a mile or two effectively all ground rods are shorted together. But at intermediate distances even low resistivity soils are pretty effective insulators.

“Sensitive electronics” grounding rules mostly deal with shielding and the idea of maintaining the bonding like a Faraday cage condition.

Modern electronics has gone from the days of unipolar RS-232 to RS-485 and Ethernet derivatives neither of which are affected by anything but the most severe noise. But we still have plenty of sensors out there to watch out for. For instance most load cells are easily confused by a truck driver keying up a CB radio. And some vendors still try to get by with nonisolated voltage and even current input cards (Allen Bradley). But that’s more of a case by case basis.

 

[winnie]

I have to say both fascinating and quite different than the requirements specified by the NEC for structure wiring.

I think that the situation describe is the application of the same physics to a very different domain of electrical installation, and thus with different rules, and of course this means that there has to be some sort of middling situation where both sets of rules are equally problematic. Curiously, this weaves back to a discussion I had on the 'ECN' board which led to Mike Holt inviting me to participate here. (At the time I was arguing that for large enough distance between buildings, that it might be safer to have a combined neutral/egc with bonding at the outbuilding, rather than the separate neutral and egc required for NEC installations.)

The basic physics remains that when you have separate ground electrodes, that current in the soil can enter one electrode, flow through connected wiring, and exit another electrode. If there is not solid bonding between the various electrodes, then these currents will flow via unintended paths such as signal wiring shields, or the associated voltages will be impressed across systems designed for much lower voltages (eg. telephone wire insulation).

For wiring of structures under the NEC, all ground electrode systems are required to be tied together. The grounded conductor of each 'SDS' is tied back to the same grounding electrode system as everything else. If a truly _separate_ system is permitted, then it is an _ungrounded_ system. In this case the goal is to make all of the grounded/bonded metal that a person might come in contact with 'float' to the _same_ potential at all times.

I think that this makes good sense for buildings which are small compared to the size of a cloud.

I have no experience with mining, but I can see how you might want to isolate one system from the rest of the world, so that currents flowing on say the utility multi-earth-neutral don't get carried into the mine. I presume that the NEC doesn't apply, but a different electrical code specific to mining.

My bet is that the mine hardware represents a rather good grounding electrode, and that if you did have such connection to the 'outside world' that you would have significant current flowing into the mine bonded metal. I am also willing to bet that when you have a substation with two 'isolated' by ground electrode systems (one on the utility side, one on the mine side) that significant effort has to be made to design the substation to tolerate the chance of voltage difference between these two electrode system.

I'd love to hear more about this.

-Jon

 

[brian john]

Good afternoon.
I participate in the design of a plant whose electrical buildings are sharing space electrical equipment and I&C equipment (sensitive loads).
For sensitive loads we have included surge protection and isolation transformer, and follow the NEC Exhibit 250.13 connection schematic. Sensitive loads have a separate grounding which is connected to general grounding grid, so it is not an insulated grounding. Now, some people on site tell they want a "digital grounding" which is an insulate green conductor, going to an insulate grounding electrode, and they say this is a normal practice here in USA. But, cheking the Erico grounding practical guide, says:

"An unnecessary conflict can be created however, such as when someone attempts to create a "separate","dedicated" or "clean" grounding connection that is
not permitted by the NEC!"

So, my question is, am I misinterpreting the NEC and this is a general practice, or the design is correct and I can't install a "clean grounding" in USA? I know in IEC countries is mandatory, but USA is IEEE, and I am less familiar with IEEE.

Thank you

Install a proper distribution system and save all the magic copper comprised of 0's and 1's (I have never seen digital copper I assume it has magic powers?). Make sure all your neutrals are ISOLATED, SPDs/TVSS will not help one iota with system noise. If you real want protection install a separately derived system for your critical loads.