Isolating Static and Neutral Wires for Grounding Grid Impedance Testing
	or
	"How to Get Your Boss's Attention"
	by R. D. Southey, W. Ruan, Y. Li

"Uh, disconnect the static and neutral wires from the substation grounding grid," you mumble, "To measure its ground resistance."

"Are you out of your mind?" is the curt retort.

So much for the direct approach. Yet you dearly wish to measure the substation grounding grid impedance, isolated from the rest of the system, as this provides a valuable validation of your grounding analysis and design predictions (or those of others) and allows you to check this value over time. Unfortunately, once a station is in service, it is unlikely that anybody will agree to disconnect the grounding grid for you.

Does this mean that you must now resign yourself to measuring not only the resistance of the grounding grid resistance, but also that of the whole system of interconnected neutral wires, water pipes, static wires and pole/tower grounds of which the substation grounding grid is an appendage? While such a measurement may provide useful results, it may simply not be what you are after. Furthermore, measuring the ground impedance of the whole town may represent an effort that you (or your boss!) are not prepared to undertake just yet. There is a way, however, to measure the grid resistance with the judicious use of variable frequency instrumentation.

What you need to carry out this field work is a variable frequency power supply (e.g., 45 Hz to 5000 Hz or more, 1500 VA), a sensitive dynamic signal analyzer, a clamp-on ammeter with minimal phase angle distortion, and long test leads. Plan a fall-of-potential impedance test such as to minimize coupling between your current and voltage test leads and between these leads and long grounded conductors: this is achieved by maximizing separation distance and angle. Be particularly careful where you route your potential lead.

The key to success, now, is to select the right injection frequency. You want the frequency to be high enough to increase the reactance of the static and neutral wires such that they are large compared to the expected grounding grid resistance, thus minimizing the loss of test current along these conductors; on the other hand, you do not want the frequency to be so high that you are introducing a large reactive component into the grounding grid impedance that is not there at power frequency. The reactive impedance of neutral and static wires is typically over 1 ohm/mile at 60 Hz; a proportionate increase with frequency can be expected. Thus, at 5 kHz, the impedance of such a conductor is on the order of 100 ohms/mile.

SES recently tested this method with the HIFREQ software module and then applied it in the field. In one case, for example, a substation ground resistance of 6 ohms was measured, using a frequency of 5 kHz; HIFREQ modeling demonstrated that the required correction for inductive effects was approximately 1.7 ohm. On the other hand, at 70 Hz, a ground impedance of 0.2 ohm was measured - a result of the municipal water system connected to the distribution neutral wires. No useful information about the substation grid would therefore have been available at this frequency. For a 0.04-ohm power plant ground which was also tested, however, a frequency of 70 Hz was preferable: because of its low impedance, the plant grounding system attracted almost all of the test current, which did not seek to flow in the 9 ohm/mile 3/8" static wires. Use of a higher frequency would have been undesirable, as the reactance of the plant grounding system, already on the same order as the resistive component, would have been unduly exaggerated and would have made it more difficult to estimate the true impedance of the grounding grid. Intermediate ground impedances would require intermediate frequencies.

"You did WHAT?" sputters your boss.

Now you nimbly explain how able you have been at not doing what he thinks you´ve done and doing what he thinks he´s forbidden you to do. Then you deftly request his authorization to attend the next CDEGS Users Group Meeting in Los Angeles this June. See you there!