Anyone here has experience with measuring ground resistance of tower/ structure footing? What testing tools are typically used by utilities to determine the footing resistance? We recently had few back flashovers on our power lines, so we are doing the root cause analysis and starting with the structure footing resistance.
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
Testing equipment for measuring powerline ground impedance
- Thread starter KillBill7
- Start date
If you have wood poles with the shield wire connecting to a ground rod or butt wrap with a ground wire, you can use a clamp type ground resistance tester. See This would also work if you have direct embedded steel poles with insulating coatings on the butts below grade and a separate ground.
If you have a steel pole or tower grounded through with the pole or to a foundation, things are complicated. You can't fit the clamp of a simple clamp type ground tester around the connections to ground (the whole pole or tower base). You normally can't disconnect the tower or pole from the shield, so a fall of potential test won't work because you will be including the multigrounded shield wire in the measurement. There are specialized test sets to make the measurements. See for example.
If you have a steel pole or tower grounded through with the pole or to a foundation, things are complicated. You can't fit the clamp of a simple clamp type ground tester around the connections to ground (the whole pole or tower base). You normally can't disconnect the tower or pole from the shield, so a fall of potential test won't work because you will be including the multigrounded shield wire in the measurement. There are specialized test sets to make the measurements. See for example.
The resistance to ground of a ground rod, ground grid, or any buried metal item is typically measured using a three-point, Fall-Of-Potential test, where two test rods are placed in a straight line out from the equipment under test. The typical distance is 100', which usually suffices for a single ground rod, or maybe three within a few feet. The distance is based on the size of the grid under measure, and may have to be much more if the overall grid size is large. Once the furthest rod (C2) is placed, the other rod (P2) is placed at 61.8% of that distance, in a direct line between the C2 and the rod under test. The P2 rod is then moved to 2-3 locations closer and 2-3 location further than the 61.8% distance. The result from each point is then tabulated and the graph drawn will show if the result at 61.8% is correct, or if you need to move the C2 rod out further to get out of the 'Sphere of Influence' of the rod under test. These tests are best conducted by a NETA trained technician/engineer, who is familiar with the procedure.
Possibly the Ohm Check Concentric Neutral Tester can be used. Its 0-48Vac @ 30Aac output is enough to inject strong currents into the ground cables and the machine will measure and report the resistance. It is not suitable for ground grid resistance testing but great for testing the grounding cables.
- Thread starter
- #6
Thank you everyone for the suggestions. We are currently evaluating this product by EPRI which seems to be specifically designed to measure the grounding performance of the electrode by simulating the lightning surge.
I can't imagine a better tool for flashover impedance measurements using ground references approx 100 m in either direction to measure voltage and current from a safely induced spike to the tower. (unless it also measured PD (partial Discharge) the precursor to corona, hint hint)
I wonder what Quebec Hydro is using (?) but I did see among the evaluators, my current and former suppliers listed for their contributions during evaluation. Hydro One and Manitoba Hydro) Duke also uses it.
Fundamentally it measures V(t)= I(t)R + L dI(t)/dt then the ratio of V(t)/I(t) is impedance Z(t) which may be converted to frequency domain. The time constant for inductive wire is Tau=L/R (63% of target) then there is an air ionization time that depends on path length and capacitance. Lightning tracers are a chain reaction of PD to ionize the air with low resistance then the follow-on return arc goes at the speed of light and ultra-high current limited only by the tower resistance which might be around 5 ohms.
I recall long wire is about 1 uH/m.
This test measures the impedance before the next tower ground can react due to the rise time of the pulse is shorter than the T=L/R.
The standard method relies on the characteristic impedance of ~ 100 m cable over soil which has some distributed coupling capacitance to ground such that the impedance is measured for at least the 300 ns flat portion of pulse before the reflected wave occurs from the cable end. (although it could have been terminated to prevent that) 200m/2e8m/s = 1e-6 s or 1000 nano-nano seconds.
Tony Stewart
EE since 1975
retired but consults on request.
I wonder what Quebec Hydro is using (?) but I did see among the evaluators, my current and former suppliers listed for their contributions during evaluation. Hydro One and Manitoba Hydro) Duke also uses it.
Fundamentally it measures V(t)= I(t)R + L dI(t)/dt then the ratio of V(t)/I(t) is impedance Z(t) which may be converted to frequency domain. The time constant for inductive wire is Tau=L/R (63% of target) then there is an air ionization time that depends on path length and capacitance. Lightning tracers are a chain reaction of PD to ionize the air with low resistance then the follow-on return arc goes at the speed of light and ultra-high current limited only by the tower resistance which might be around 5 ohms.
I recall long wire is about 1 uH/m.
This test measures the impedance before the next tower ground can react due to the rise time of the pulse is shorter than the T=L/R.
The standard method relies on the characteristic impedance of ~ 100 m cable over soil which has some distributed coupling capacitance to ground such that the impedance is measured for at least the 300 ns flat portion of pulse before the reflected wave occurs from the cable end. (although it could have been terminated to prevent that) 200m/2e8m/s = 1e-6 s or 1000 nano-nano seconds.
Tony Stewart
EE since 1975
retired but consults on request.
- Status
Similar threads
- Question
- Locked
- Question
- Question
- Locked
- Question