Bonding of Isolated Grounding system to plant electrode system, i.e. ground loop

[roger]

I just have to convince the client his idea is not a good one.

Tell him that the way his system is set up is a ground loop, non code compliant, and when the link is lifted there will be nothing to clear a fault putting people in harms way.

Roger

 

[Dale001289]

I think I have enough ammo here to compose a 'tactful' email to the client telling him his proposed design is horrible.
Thanks to all.


Sent from my iPhone using Tapatalk

 

[petersonra]

It has occurred to me that an EGC like conductor might not be an EGC if it is not intended to be used as part of a fault clearing path. For instance, if it was used solely to bond shields to earth, it would not be part of any fault clearing path.

 

[roger]

It has occurred to me that an EGC like conductor might not be an EGC if it is not intended to be used as part of a fault clearing path. For instance, if it was used solely to bond shields to earth, it would not be part of any fault clearing path.

Even if that were the case you would still be providing a loop for circulating currents in Dale's description of whats present.

Roger

 

[petersonra]

Even if that were the case you would still be providing a loop for circulating currents in Dale's description of whats present.

Roger

Where would the current path be if it is not connected to a power source?

 

[roger]

Where would the current path be if it is not connected to a power source?

In Dale's original post the conductor in question is connected to the GES, meaning it is an EGC in a loop back to the source.
If we are going to shift gears from an "isolated ground system" to an "auxiliary GE", it is another conversation altogether.

Roger

 

[Smart $]

Where would the current path be if it is not connected to a power source?

What wiring system is not connected to a power source?

 

[petersonra]

What wiring system is not connected to a power source?

If it is solely for instrument wiring shield grounding it is quite possible the shields never have any voltage on them, while the common side of the power supply running through the shielded cable might well be grounded.

 

[Smart $]

If it is solely for instrument wiring shield grounding it is quite possible the shields never have any voltage on them, while the common side of the power supply running through the shielded cable might well be grounded.

It is never the intention of any grounding conductor to have a voltage on it, be it shield or wire. But a voltage needs two points of reference and a potential between them. If you truly want the shielding isolated, just bond all the shields together and do not connect to ground (earth). But how do you propose to handle a 'line' to shield fault?

 

[Dale001289]

It is never the intention of any grounding conductor to have a voltage on it, be it shield or wire. But a voltage needs two points of reference and a potential between them. If you truly want the shielding isolated, just bond all the shields together and do not connect to ground (earth). But how do you propose to handle a 'line' to shield fault?

The main concern here isn't exactly amps/voltage via the shielding, but rather several relative factors (electromagnetic induction, electrostatic coupling, or conduction), collectively known as EMI - Electromagnetic interference, i.e. "noise" feedback - into the DCS (or PLC) I/O, which controls plant operation. Specifically, regarding linear 4-20mA circuitry from critical, process-related transmitters, such as Flow, Temp, Pressure and Level.

 

[Smart $]

The main concern here isn't exactly amps/voltage via the shielding, but rather several relative factors (electromagnetic induction, electrostatic coupling, or conduction), collectively known as EMI - Electromagnetic interference, i.e. "noise" feedback - into the DCS (or PLC) I/O, which controls plant operation. Specifically, regarding linear 4-20mA circuitry from critical, process-related transmitters, such as Flow, Temp, Pressure and Level.

The concern is unwarranted as long as you ground only one end of shields. Disregard instructions which indicate otherwise. There are probably hundreds of thousands of plants out there operating this way with no problem.

 

[Dale001289]

The concern is unwarranted as long as you ground only one end of shields. Disregard instructions which indicate otherwise. There are probably hundreds of thousands of plants out there operating this way with no problem.

Typically, in process related applications, the shield is grounded at the source end only (e.g. DCS cabinet) and left floating at the device (e.g. transmitter). But this is done to prevent circulating current and does not address the broader spectrum of EMI.
The shield should be kept as 'clean' as possible, (as in free of outside interference, capacitance etc) and at the same ohmic level as the isolated equipotential plane, i.e. Triad. This is why bonding the two systems, as required by NEC 250.50, is regarded a negative factor by some Clients.

There's nothing more subjective than grounding.

 

[petersonra]

A few years ago there was an interesting paper presented somewhere that looked at the effect of grounding shields, either at one end or the other, both, or neither. While it was by no means a paper that "proved" much of anything, the results suggested that for the most part it just did not matter much if the shield was grounded or not, but that the best EMI performance was by grounding both ends. The real problem with grounding both ends of a shield is that in most plants there is a significant difference in the voltage between two ground points within the plant and this causes a current flow on the shield. In some cases this has been enough to melt the shield due to heat when a shield was inadvertently grounded at both ends.

 

[Dale001289]

A few years ago there was an interesting paper presented somewhere that looked at the effect of grounding shields, either at one end or the other, both, or neither. While it was by no means a paper that "proved" much of anything, the results suggested that for the most part it just did not matter much if the shield was grounded or not, but that the best EMI performance was by grounding both ends. The real problem with grounding both ends of a shield is that in most plants there is a significant difference in the voltage between two ground points within the plant and this causes a current flow on the shield. In some cases this has been enough to melt the shield due to heat when a shield was inadvertently grounded at both ends.

Its sounds 'legit' to me.

 

[Smart $]

Typically, in process related applications, the shield is grounded at the source end only (e.g. DCS cabinet) and left floating at the device (e.g. transmitter). But this is done to prevent circulating current and does not address the broader spectrum of EMI.
The shield should be kept as 'clean' as possible, (as in free of outside interference, capacitance etc) and at the same ohmic level as the isolated equipotential plane, i.e. Triad. This is why bonding the two systems, as required by NEC 250.50, is regarded a negative factor by some Clients.

There's nothing more subjective than grounding.

Actually having and connecting the shield is addressing rfi/emi rejection.

Isolated equipment grounding systems under the NEC can originate at the main bonding jumper. This is as close to 'earth' as you can get with a compliant EGC. The only point(s) closer is the along GEC(s) or the electrode(s) itself... but non-compliant unless for an SDS (separately-derived system). Achieving the lowest resistance to 'earth' as you possibly can is a better tactic than trying to skirt compliance. This is why many communications companies have a maximum R2E for their hubs. The only way you can buck compliance is to submit a public input to the cycle revision process.

 

[brian john]

The sole connection to plant grounding system as required by 250.50 can result both common mode noise and objectionable currents. The removable link allows the field to 'trouble-shoot' the isolated grounding system and theoretically eliminate both.

I run into this hoodoo voodoo thinking all the time, a properly installed grounding electrode system, MBJ and properly installed wiring with no downstream grounds on the neutrals are your best bet for a "CLEAN" grounding system.

AND Not sure on your building type, but having a truly isolated ground system is difficult due to all the metallic components of a building, building steel, rebar, ducts, water pipes, every large scale IG system we have investigated has been compromised, somewhere in the distribution system, be it misunderstanding by installers, pinched IG conductors or unintentional connections.

A lot of money is spent on additional copper with no real benefit IMO.

As for Triads, no better or worse than a variety of driven electrodes.

 

[ActionDave]

As for Triads, no better or worse than a variety of driven electrodes.

It's all about the triangle and the power that it holds in the universe.

 

[brian john]

It's all about the triangle and the power that it holds in the universe.

For some, this statement is not far from the truth.