Grounding and bonding in a refinery

[Skogsgurra]

Dear all

Are there any general rules as to how often one should check the quality of ground lead connections in an oil refinery? In this case, there are Cu bars above ground to which Cu ground leads are connected and then one Cu lead is connected down to a buried grid of Cu cables that are cross-connected and connected to earth plates and earth rods.

We are thinking of doing a continuity check on some of the ground loops in the system and are looking at the Fluke 1630 clamp as well as a few other makes. These clamps work with an elevated frequency (seen 2.5 and 3.33 kHz) and have a rather poor resolution. They are said to measure resistance, but Fluke (seems to be the only guys that mention this) say that the result given is for an inductance-free loop. I cannot think of an inductance-free loop in a grid with meshes measuring at least 100 feet and more often 300 feet square.

Would this make a serious measurement of ground grid mesh/loop resistance impossible?

We are also looking at the use of a transformer which we feed with 50 Hz (Sweden) and which has something like a few volts/turn (say 100 turns connected to 230 V) and measure the 50 Hz current that the voltages can drive through the circuit. We have done that to check other low-impedance circuits but never used the technique for ground grid meshes and similar.

All tips and experience you can share is more than welcome!

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ScottyUK]

The clamp-on type earth electrode testers I have looked at seem to rely on the electrode under test being much higher in resistance to earth than the return path via multiple electrodes in an array. I am dubious of these instruments, especially in poor earthing conditons, although I like the convenience they purport to offer. They do prove that the conductor under test is connected to the buried earthing grid, but I am sceptical of the accuracy of the numbers displayed. They are bordering on useless for a single point electrode.

Have you looked at the fall of potential and Wenner methods for testing earth electrodes? Their downfall is that the test leads can become unreasonably long for a large electrode like a substation earth grid, and impossibly long for a large and heavily interconnected array like a power plant or refinery. I had an almost comical discussion with an auditor regarding the overall resistance to earth of a massively interconnected earth grid measuring roughly ½ mile by ¾ mile at a power station, especially when he insisted that I should prove that the resistance was less than one ohm. I'm not quite sure what instrument he proposed to use but I'm glad I wasn't carrying it.

 

[Skogsgurra]

This is not about resistance from earth grid to ground. That doesn't matter very much in this case. It is more about finding out why there was such a big potential difference between different points in the grid as such. So, I need to verify that quite a few grounding bars (where the different PE conductors are connected) are connected via the grid and that the impedance in each connection is reasonably low.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[waross]

I think that the clamp on testers find their greatest use checking the "Magic triangle" used for instrumentation grounds. (Three 10 foot long rods in a 10 foot triangle or something similar.) I haven't used one but I did see a confused instrument tech misuse such an instrument. His misuse gave me a very good idea of how the instrument was supposed to be used. The instrument is clamped around one ground rod and induces a current in the rod. The instrument indicates the ground resistance from the rod under test to the two other rods.
It may be worthwhile to use such an instrument to verify the quality of connections of the ground leads.
The petro-chemical installations that I am familiar with have multiple connections from from the ground grid to the steel structures supporting the pipe racks. If there are cable trays running with the pipes, at least one tray in a group will have a 2/0 ground conductor running in the tray. This is connected to the steel and to the buried grid every 50 feet. In some cases, the steel support structure is similar to a series of squat towers with twenty foot long gaps bridged by the pipes and cable trays and the continuity of the steel support structures may not be dependable from section to section.
It may be worthwhile to use the Fluke 1630 to verify the quality of connections of the ground leads.
A high resistance reading will indicate an open circuit on the conductor under test and the need for repairs.
Testing the quality of individual conductors/connections and locating any open circuits may be more valuable than trying to measure the resistance to ground of a large grid.
Any lead used as a single point ground of a piece of equipment may test open, but that is not an issue if you are aware of the possibility.
In any event, equipment of any size may be grounded with two or more connections.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[ghostbuster7]

You may want to watch this recent grounding video that displays the different readings obtained using different instumentation.

http://forums.mikeholt.com/showthread.php?t=159914

 

[waross]

Watch out for cathodic protection!

Bill
--------------------
"Why not the best?"
Jimmy Carter