It doesn’t take an electrician or engineer to know that lightning is a major cause of electrical interference and damage. Protecting the electrical system against such a danger is paramount. A lightning stroke can be hotter than the surface of the sun, with voltage estimated as high as 120 million volts, including 400 kA and 500 megajoules of energy. Utility transmission towers are particularly inviting targets due to their large metallic mass and their height above nearly all surrounding structures.
Therefore, a grounding wire runs above the conductors as a convenient point of initial contact and is continuous down the tower legs and into a grounding electrode at each foot. Such a structure provides excellent protection, but it must be maintained. Transferring massive lightning strokes into the earth in microseconds is a demanding task, and the electrode can become damaged or deteriorated in the process, even while preventing any notice of its presence on the electrical system. As part of installation and maintenance, it’s important to test and evaluate transmission line grounding impedance.
Prevailing methods to test tower grounds are performed at lower test frequencies, and their thoroughness and accuracy has been challenged. The connection to adjacent towers and the entire string further complicates the testing. Wouldn’t it be better to simulate a lightning-like impulse?
A pioneering test of an alternate method uses a pulse generator and digital oscilloscope to inject a 440-nanosecond transient pulse into the tower base and measures the transient response at a remote ground — the effect of the simulated lightning pulse on the immediate soil environment. The remote ground was the sheath of a 50Ω coaxial cable, oriented at a right angle to the tower line. A 60-m reference potential lead to prevent reflections from distorting the measurements ran along the same line. The test equipment was centered at the tower base to counter problems associated with other proposed test methods, such as high-frequency attenuation and lethal test signals (Figure 1).
Results were evaluated against airborne electromagnetic data and traditional fall of potential test results and found to give a good indication of the effectiveness of the grounding under lightning impulse conditions.
A 200 V pulse applied between the tower base and the remote current terminal yielded a pulse width of 1.4 milliseconds (Figure 2). For confidence in the result, the electromagnetic value was calculated using NEC-4, Numerical Electromagnetics Code, which is a widely used computer code for analyzing the electromagnetic response of antennas. For this exercise, NEC-4 computations were carried out in the range of 195.3 kHz–100 MHz to simulate lightning conditions.
The method’s success was modeled after the behavior of a Beverage antenna, which is a horizontal wire antenna placed near lossy ground. Studies indicate that phase velocity will be slow for a surge injected into antennas near the earth, and losses will reduce the magnitude of the surge. This models the excitation of a large transmission tower with overhead ground wires using a signal that gives practical ground impedance results using low-voltage pulses. Using the Beverage antenna as a model, a step wave can be applied to the middle of the wire. The observed ratio of the step voltage to the currents in each leg of the dipole establishes the transient surge impedance.
This model and its test results are shown in Figure 3a, Figure 3b, and Figure 3c. Figure 3a shows a 200-meter dipole over lossy ground with a resistivity of 50 Ω-meters and a relative permittivity of 10, with a wire 5.4 mm above earth surface and with a radius of 1.8 mm. The insulation is pure dielectric, with relative permittivity of 3 and radius of 2.5 mm. Both ends of the dipole are terminated with ground rods buried 0.1 m in earth. The injected pulse meets the description shown in Figure 2. The slow propagation speed of the current along the wire at different distances from the source is shown in Figure 3b. The transient impedance, obtained by dividing the source voltage by the NEC-4 computed current (Figure 3c), settles fairly rapidly to a value of 370 Ω and stays fairly constant until the effects of reflections from the ground rods arrive at 1.3 μsec. These reflections do not influence the result, as it is read out before their arrival. Translating this into robust transient impedance measurement, this model demonstrates that careful selection of pulse width and dipole length can enable a time window large enough for the measurement to be made. Tower ground measurement can be modeled on the example of the Beverage antenna.
Note that the result obtained (Figure 3c: 370 Ω) is high only because this was done on an experimental model of a Beverage antenna, not on an actual tower ground. It serves the purpose of illustrating that the method can be transferred to an actual tower serving in the place of the Beverage antenna with appropriate test parameters. Transient impedance varies with wire height and earth resistivity. The strong influence of dipole height (Figure 4) is shown for heights of 1 m and 0.01 m.
The results of an experiment performed with 200 m coaxial cable and a pulse injected into the sheath and current measured in both directions with CTs (Figure 5) compare favorably with the results shown in Figure 3c.
This experimental model can be adapted to actual tower grounds as depicted in Figure 1. Current injection and remote potential leads are 100 m and positioned parallel with the tower line, terminated with ground rods. The tower is 36 m tall with a foundation at a depth of 4.6 m. The adjacent towers are modeled as simple vertical stakes, spaced at 200 m, with an aerial ground wire bonded to structure. In the corresponding numerical model, the leads are at a constant height of 10 cm. The wave shape of injected current is shown in Figure 6 and tower base voltage shown in Figure 7. Tower transient impedance then settles to 5 Ω, a reasonably acceptable value, and holds until the reflections from the lead ends arrive.
The experiment was conducted in a relatively agreeable soil type with a resistivity of 50 Ωm (Figure 8). As an illustration of the profound effect of soil on grounding protection, a similar test conducted in 1,000 Ωm soil (Figure 9) revealed 46 Ω of ground impedance, not a desirable value. This further illustrates the importance of testing and maintenance, as a good ground cannot be presumed merely on structure.
It should also be noted that one objection to more traditional methods of testing is the effect of adjacent towers along the line that cannot be electrically separated. In both the Beverage antenna experiment and the tower ground actual test, the effects of the termination ground rods (Beverage antenna) and the adjacent towers are nullified, as the median current is measured before any reflections arrive. To cope with the variability of the electrical footprints of different towers, all that must be done is to increase the lead lengths and pulse duration. Note how this parallels the solution to issues with traditional testing, like Fall of Potential, where increasing the lead distance is often the most effective remedy.
Numerous trials have shown good agreement between results calculated by NEC-4 and those observed using the transient injection method. NEC-4 results tend to be more pronounced in terms of faster return time and stronger magnitude of reflection, but this effect can be reproduced by adjusting the height of the leads above ground. The transient injection method has the advantage of indicating potential rise at frequencies commensurate with those imposed by a lightning stroke, and its results match up well against those obtained by established methods.
E. Petrache, W. A. Chisholm, A. Phillips, “Evaluating the Transient Impedance of Transmission Line Towers,” IX International Symposium on Lightning Protection.
Jeffrey R. Jowett is a Senior Applications Engineer for Megger in Valley Forge, Pennsylvania, serving the manufacturing lines of Biddle, Megger, and Multi-Amp for electrical test and measurement instrumentation. He holds a BS in biology and chemistry from Ursinus College. He was employed for 22 years with James G. Biddle Co., which became Biddle Instruments and is now Megger.