Grounding issue has, I guess, always been a confusing matter for people like me. I have two grounds, Safety and Instrument. All cabinets, chasis are terminated on safety ground for the protection of humans. Whereas instrument shields are cut and taped in the field and terminated on instrument ground on marshalling cabinets for the protection of instruments and equipment.

So far so good, but what about -24V DC, where should it be connected? IN MY OPINION IT SHOULD BE CONNECTED to Instrument ground. Is this right?

Furthermore, in the soil safety and instrument grounds are more than 4m apart, but on the mesh, these are connected on one point only. Is it OK?

Need advice.

 

kkk,

Most of the larger plants I have been in have a central point in the switchroom which I call the “Main ground bar” where all the MCCs, switchgear, cable tray grounds and the main neutrals are connected together. From here also is a connection to the main perimeter buried ground. I believe the “Main ground bar” is where your instrument system should connect also. 24V power supply negatives should as you say connect to the instrument ground bar inside each panel. In theory if you were to lift the instrument ground connection from the “Main ground bar” you would find it and all the instrument shields isolated from ground. Not that I recommend trying. Large pieces of equipment on ground tanks and building columns are usually connected to the buried perimeter ground.
I have sometimes seen the instrument system connected to a separate buried ground, I believe this is wrong. I visualize the grounding system like a Faraday shield, as long as you stay inside you are safe, connecting to a separate buried ground is akin to sticking your arm out.
No doubt I will get shot down in flames.
Oh well.

Roy

P.S. Does anyone else find these pages slooooooow to load?

 

First of all you are right, this page was too slow to download.

Thanks, i am relieved, For lifting the instrument ground from the main ground, I have checked that, before energising my system and it was isolated from the main ground.

kkk

 

One point to consider is the earthing of any Intrinsic Safety barriers. In the IEC-based codes, the barrier earth must be connected to the neutral of the power supply with no more than 1 ohm resistance. Since the barrier earth is often also the instrument earth, you generally need to connect the two directly rather than rely on the uncertain resistance between 2 separate ground electrodes.

Bruce.

 

Regarding Roy's comments about his desire to have all grounds connected together, I would like to add a bit of field experience. There is no need to picture a grounding system as a Faraday shield, Ohms Law will suffice. The reality an most industrial plants is that there is some ground leakage from the power system. With the popular use of VFD's and other solid state drives, a fair bit of that leakage has high frequency components. From the main ground terminal this leakage current flowing through the resistance to 'perfect earth ground' gives rise to a noisy voltage drop. This is avoided by going to a separate instrument ground. But this ground will have to be out of the sphere of influence of the power ground, at least several tens of meters away.

Hugo

 

Roy Matson,

You are not alone in your beliefs--though we are in the extremely small minority of believers. There is this huge and unstoppable movement to separate control system grounds from "main neutrals" as you call them. Why? Because on the face of it, it *appears* to make sense. And appearance is just about everything to those people selling and purchasing power plants and industrial plants. They *refuse* to look at the control systems being provided and how they are designed to operate, and *insist* on *forcing* control system vendors to "adapt" their control systems to fit a plant design scheme which is not in keeping with the control system design.

And the control system vendors make some, sometimes expensive, modifications to accommodate these requirements which results in unexplained and nuisance control system problems. I have been to many sites where I measure a difference in potential between the "main neutral" (sometimes referred to as "protective earth") and the instrument ground "system" (a term I am using *very* loosely) using my multimeter. When I use a simple alligator jumper to connect the control system instrument ground to the main neutral" (quickly, before anyone realizes what is happening because they will physically try to restrain me from doing so) all the instability and problems suddenly stop. After the plant "engineers" (another term used loosely in many parts of the world) and instrument tech's calm down and see that nothing untoward has happened, they insist the jumper be removed. And in many cases, almost immediately upon doing so the instability and nuisance problems start. When I try to say it's because of the potential differences between the two "alleged" grounding/earthing systems (isn't ground ground? shouldn't any ground system be at zero potential with respect to any other ground system?) I'm soundly thrashed for my heretical contention and informed that the problem is the control system, not the ground system.

Yet at other similar plants where there is only one ground system, there are few, if any, similar problems, which I directly atrtibute to a single ground system and the absence of different potentials between "alleged" ground systems.

I agree whole-heartedly that the concept makes inherent good sense--however, until control system manufacturers *and* plant designers start isolating instruments and devices from the "main neutral" and connecting them to the "instrument neutral", *or* designing entire control systems (the control panels/hardware *and* the instruments and devices they are connected to) to be unaffected by potential differences, *or* the designers and constructors of plants are forced to design, build, and test ground systems to conform to what would appear to be a simple concept of all ground systems at one single potential.

We'll go down in flames together, in the knowledge that change begins with a few people who challenge conventional "wisdom."

 

Hugo,

It's an interesting subject grounding. Picture a magnetic flow-meter installed in a pipe, it's enclosure is solidly connected to the electrical ground by the pump and equipment grounds and its power supply 120 VAC also but the analog output is connected to the instrument ground through the control system therefore if you have the instrument ground separate from the electrical ground any difference (noise) shows up across the meters electronic components. I believe that using the same common ground avoids that.

I'm not sure what you meant by the Ohms Law reference?

By some ground leakage, do you mean a short to the frame, that should make it's way back to the supply transformer via conduit and equipment ground because it is trying to get back to the power source (transformer), or are you referring to the potential difference between different areas of the plant? To clarify my case, I am referring to a typical installation where the supply transformer is just outside the switch-room sharing the same ground mat and any on-ground tanks are also grounded to the same mat.

Why do we put up with the noise from VFDs, I like to specify Line and Load reactors which I'm sure help. Lately I see a move toward special cable for VFD-motor with 3 grounds but that's to benefit the motor and VFD.

Roy

 

In the farming community, the problems of multiple grounds have been long understood. The possibility of setting up stray currents and the associated voltage drops is very real and has all sorts of nefarious effects on cattle and other farm animals who are much more sensitive to this sort of stuff than we humans are. The same principles I am sure apply to the instrumentation world where small unwanted and unpredictable differences of potential can wreak havoc with control systems. The utilities companies also specify... one ground ONLY per service entrance!

Gerald Beaudoin

 

CSA,

Right on, I had a similar experience with trying to get a PLC to communicate RS232 with a DCS in a cabinet 6 feet away, wouldn't talk until I connected the DCS ground to the electrical ground, I never measured it but concluded that the potential difference must have been enough to bias the signal out of range.

Bruce Durdle's response re: IS barriers was also confirmation that we are on the right track.
BTW the Wikipedia article on Faraday cage is worth reading.

Roy

 

On one site I worked on, there was a PLC and a DCS - each connected to a single instrument earth point in the respective cabinets, and these were connected to a sub-floor instrument earth. A fairly standard tree configuration. But there were some interconnections for control purposes between the two. One was designed to open a steam bypass valve in the event of a turbine trip, and was implemented with a resistor network switched by the PLC and connected to a lead network in the DCS. During a shutdown some guys were working on the bypass valve while others were checking PLC alarm circuits. The valve (14" so no toy) decided to move through about 1/3 of a stroke, narrowly avoiding the fingers of the guy who was making some adjustments at the time. We traced this back to the effect of the alarm lamp current surge (120 plus lamps at 110 V) when the Lamp Test button was pressed, flowing through the very small series inductance of the PLC earth connection. This gave about a 0.2 V step at the DCS input, amplified by the effect of the lead network. Solution was to use reed relay outputs in the PLC and ensure that the power supply and earth connections to the resistor network were taken back to the DCS pane.

Aaah earthing - guaranteed to be good for some very heated debates...

Cheers,

Bruce

 

Well, now, isn't this interesting? The farming community has figured out that grounding is extremely important, and, yet, the power- and industrial plant design and construction communities can't make it work!

Proper design and construction of the use of multiple grounding systems is critical--and yet it's one of the most over-looked aspects. I've only ever been on one site where proper ground system testing was implemented during the construction of the plant, and that plant used only one ground system for the entire plant. One would presume that ground system testing would be critical, especially for plants which use multiple ground systems.

Ideas are wonderful things. Implementation is sometimes quite another story, and can destroy a good idea (and has destroyed more than one).