1008734 the EPRI Zed-Meter

The EPRI Zed-Meter A New Technique to Evaluate Transmission Line Grounds 1008734

The EPRI Zed-Meter A New Technique to Evaluate Transmission Line Grounds 1008734 Technical Update, October2004

EPRI Project Manager Andrew Phillips

EPRI• 3412 Hillview Avenue, Palo Alto, California 94304• PO Box 10412, Palo Alto, California 94303• USA 800.313.3774• 650.855.2121• [email protected]• www.epri.com

DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM: (A) MAKES ANY WARRANTY OR REPRESENTATION WHATSOEVER, EXPRESS OR IMPLIED, (I) WITH RESPECT TO THE USE OF ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT, INCLUDING MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, OR (II) THAT SUCH USE DOES NOT INFRINGE ON OR INTERFERE WITH PRIVATELY OWNED RIGHTS, INCLUDING ANY PARTY'S INTELLECTUAL PROPERTY, OR (III) THAT THIS DOCUMENT IS SUITABLE TO ANY PARTICULAR USER'S CIRCUMSTANCE; OR (B) ASSUMES RESPONSIBILITY FOR ANY DAMAGES OR OTHER LIABILITY WHATSOEVER (INCLUDING ANY CONSEQUENTIAL DAMAGES, EVEN IF EPRI OR ANY EPRI REPRESENTATIVE HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES) RESULTING FROM YOUR SELECTION OR USE OF THIS DOCUMENT OR ANY INFORMATION, APPARATUS, METHOD, PROCESS, OR SIMILAR ITEM DISCLOSED IN THIS DOCUMENT. ORGANIZATION(S) THAT PREPARED THIS DOCUMENT EPRI Kinectrics, the former Ontario Hydro Research Division

This is an EPRI Technical Update report. A Technical Update report is intended as an informal report of continuing research, a meeting, or a topical study. It is not a final EPRI technical report.

ORDERING INFORMATION Requests for copies of this report should be directed to EPRI Orders and Conferences, 1355 Willow Way, Suite 278, Concord, CA 94520. Toll-free number: 800.313.3774, press 2, or internally x5379; voice: 925.609.9169; fax: 925.609.1310. Electric Power Research Institute and EPRI are registered service marks of the Electric Power Research Institute, Inc. EPRI. ELECTRIFY THE WORLD is a service mark of the Electric Power Research Institute, Inc. Copyright © 2004 Electric Power Research Institute, Inc. All rights reserved.

CITATIONS This document was prepared by Kinectrics 800 Kipling Avenue Toronto, Ontario M8Z 6C4 Canada Principal Investigator W. A. Chisholm Co-author J. G. Anderson 525 Old Windsor Road Dalton, MA 01226 This document describes research sponsored by EPRI. The publication is a corporate document that should be cited in the literature in the following manner: The EPRI Zed-Meter: A New Technique to Evaluate Transmission Line Grounds, EPRI, Palo Alto, CA,: 2004. 1008734

iii

iv

ABSTRACT Traditional methods for measuring the resistance or impedance of transmission tower ground electrodes lose accuracy when overhead groundwires are connected to the towers. Existing methods for accurate tests call for temporary isolation of these parallel connections to remote earth. With mature pulse generation technology and new portable instrumentation, it is now possible to measure the low-current, time domain response of the transmission tower footings without lifting the overhead groundwires. The measured transient impedance is ideally suited for evaluating lightning performance, unlike results from low-frequency measurement methods. A co-operative field trial with AEP staff in the Appalachian Mountains showed that:

• The new Zed-Meter approach could be used on power lines from 69 kV to 765 kV • Nine towers could be tested from the ground, without outages or line work, in a working day by a crew of two. Typically, this crew could only test four towers per day using conventional live-line work methods for temporary isolation of the overhead groundwires.

• The Zed-Meter results ranked footings in the same order as the measured construction footing resistance, but the readings all differed by a factor.

• The Zed-Meter approach can provide measurements of the distributed resistivity, averaged over the length of the current injection system. The existing Zed-Meter configuration could be packaged to improve ease of use and reduce equipment cost. Cross-checks with isolated electrodes in a variety of soil conditions would be useful to further refine the relation between lightning-transient impedance and 60-Hz resistance reported here. The Zed-Meter approach can be adapted to provide direct measurements of touch, step and transferred potential to nearby objects. With further refinement, it may be possible to improve the technique: •

To keep all test conductors within the right of way

•

To use the same test conductors and process to measure earth resistivity near the tower

•

To give a direct indication of ground electrode condition based on the ratio of measured impedance to measured local resistivity.

•

To reconfigure the equipment to excite and measure the lightning surge response of some transmission towers.

v

vi

CONTENTS 1 TRANSMISSION LINE GROUND ELECTRODES ................................ 1-1 Description.................................................................................................................................1-1 Electrical Functions .................................................................................................................1-1 Conventional Test Methods...................................................................................................1-2 During Construction...........................................................................................................1-2 After Installation of Overhead Groundwires.................................................................1-3 Recently-Described Test Methods for Footing Impedance ............................................1-3 High Frequency Resistance Measuring Instrument – BBC HW2W .......................1-3 High Frequency (2-kHz) Directional Impedance.........................................................1-5 Pseudo-Random Noise Injection: The Smart Ground Meter ...................................1-6 Excitation with Existing AC Current ...............................................................................1-8 Test Methods for Earth Resistivity .......................................................................................1-8 Electromagnetic Induction ...............................................................................................1-8 Ground-Based Two-Coil Multi-frequency Electromagnetic Surveys................... 1-10 Ground Penetrating Radar ........................................................................................... 1-12 Comparison of Resistivity Test Methods................................................................... 1-13

2 TRANSIENT INJECTION AT TOWER BASE ........................................ 2-1 Description.................................................................................................................................2-1 Timing Diagrams ......................................................................................................................2-1 Test Lead Design.....................................................................................................................2-4 Interpretation of Measurements............................................................................................2-5 Surge Impedance of the Current Lead .........................................................................2-5 Current Split between Test Lead and Foundation......................................................2-5

3 EQUIPMENT SELECTION ..................................................................... 3-1 Selection of Pulse Generator ................................................................................................3-1 Selection of Impulse and Measurement System Response Time ................................3-2 Selection of Digitizer .........................................................................................................3-3

4 SUMMARY OF PREVIOUS FIELD TRIALS .......................................... 4-1 Initial Field Trials at New Brunswick Power .......................................................................4-1 Equipment Used.......................................................................................................................4-1 Summary of Findings ..............................................................................................................4-2

5 EPRI ZED-METER FIELD TRIALS ........................................................ 5-1 EPRI Zed-Meter Configuration..............................................................................................5-1 Overall Test Conditions ..........................................................................................................5-8

6 EPRI ZED-METER RESULTS ................................................................ 6-1 vii

765 kV Lines .............................................................................................................................6-1 What Was Expected..........................................................................................................6-1 What Was Found ...............................................................................................................6-1 Tower 29-40........................................................................................................................6-1 Tower 29-22........................................................................................................................6-3 Tower 29-19........................................................................................................................6-4 Comparison with AEP Historic Data..............................................................................6-6 138 kV Lines .............................................................................................................................6-7 What Was Expected..........................................................................................................6-7 What Was Found ...............................................................................................................6-7 Line 40-53 Tower 18 .........................................................................................................6-8 Line 40-53 Tower 14 ...................................................................................................... 6-10 Tower 18-133................................................................................................................... 6-11 Axton-Danville No.1 Line 26 Structure 6 ................................................................... 6-12 69-kV Lines............................................................................................................................. 6-15 What Was Expected....................................................................................................... 6-15 What Was Found ............................................................................................................ 6-15 Line 290 Pole 38 ............................................................................................................. 6-16 Line 290 Poles 51 and 52 ............................................................................................. 6-18

7 CONCLUSIONS...................................................................................... 7-1 8 RECOMMENDATIONS........................................................................... 8-1 9 REFERENCES........................................................................................ 9-1

viii

1

TRANSMISSION LINE GROUND ELECTRODES Description Transmission tower foundations are designed to provide mechanical support in a wide range of soil and rock conditions. Depending on application, freestanding steel lattice towers, wood poles and guyed structures are all common. In each case, the tower and foundation also serve some electrical functions by providing a conducting path for lightning and ac system fault current. Even in cases where insulating materials are placed between tower guy wires and guy grips, the high electric stresses associated with lightning will puncture or flash over the gaps. In some areas and for some designs, the foundations are sufficiently large to provide low impedance to remote earth. Intentional bonds should link tower and foundation components to maximize the “natural” geometric resistance. In other cases, such as for wood poles, supplementary ground electrodes such as driven rods will be needed to provide this size and surface area. This chapter will describe the basics of grounding test methods and identify what existing grounding test methods can, and cannot, accomplish. Electrical Functions The electrical impedance of a foundation to remote earth is dominated by the “geometric resistance”, which is a function of the local soil resistivity, the overall size of the combined foundations and the surface area. The size and area include tower anchor rods that are electrically connected to the tower through the guy wires. If we consider the ground electrode of a transmission line structure as a box surrounded by soil of resistivity ρe, then we can calculate a geometric radius g and a surface area A using a length Dx, width Dy and depth Dz with open top:

g=

Dx2 + D y2 + Dz2

2 A = Dx D y + 2( D y Dz + Dx Dz )

The geometric resistance of this box is given by: Rgeometric =

ρ e Ê 11.8 g 2 ˆ ˜ lnÁ 2π g ÁË A ˜¯

The local effective resistivity of the earth ρe is not normally known, and this presently limits the use of Equation 1-3 to overall line design. One desirable function of an improved Zed Meter would be to provide an efficient, fast way to measure effective resistivity during the test.

1-1

Conventional Test Methods During Construction In the time period after foundations are prepared, prior to stringing the overhead groundwires, there is no electrical connection between the ground electrodes of adjacent towers. This provides a good opportunity to measure the resistance of each tower to remote earth and to install any supplementary ground electrodes such as rods or buried radial wires to meet a design specification. Records of construction footing resistance are often retained within utilities long after other construction details are discarded. These values sometimes give excellent insight into the variability in soil resistivity from structure to structure. The records of low-frequency lowcurrent resistance are used directly to evaluate the lightning performance of the line, using such programs as IEEE FLASH and EPRI TFLASH. Since the electrical resistivity is often correlated with other soil properties, a database of construction footing resistance (especially before installation of supplementary electrodes) can also be a useful component for refining grounding and foundation inspection and maintenance programs. While accurate construction footing resistances are of value, the methods under which these measurements were taken tend to vary widely. Normally, the equipment used is common: a four-terminal impedance meter, used in the three-terminal configuration shown in figure1-1 . A four-terminal earth resistance tester has separate terminals for injecting current (C1 and C2) and for measuring the resulting potentials (P1 and P2). For the three-terminal measurement of a structure resistance, both the C1 and P1 terminals would be tied to the tower under test. The C2 terminal would be connected to a vertical probe, driven into the ground at a considerable distance D1 from the foundation. The P2 terminal would also be connected to a remote driven rod, often located often at a distance D2=0.62D1 between the tower under test and the remote C2 probe. D1 = 10 Tower Diameters D2 = 0.62 D1

C1 P1

P2 C2

Earth Resistance Tester

1-2

Figure 1-1: Lead Layout for Fall-of-Potential Test for Isolated Structure Resistance

The variation in practice is two-fold. First, there is a wide range of opinion about how large the “considerable distance” D1 from the foundation should be. For example, when the tower ground electrode is a 3-m (10’) driven rod, adequate results can be obtained for a remote C2 probe that is (D1=30 m) away. However, for a tower with four foundations on 12-m spacing, each contributing about the same low resistance, then a C2 probe location of D1=120m away would be more appropriate. The second variation in practice is in the choice of P2 probe location as a fraction of the C2 probe distance away from the source. For a uniform soil and a concentrated electrode, a distance of D2= 60 to 70% of D1 gives nearly direct readings of the structure resistance without having to correct for the C2 probe influence. However, for non-uniform soil, this is no longer the case. The advice to record measurements at P2 probe locations of (50%, 55% … 80% ) of D1 is commonly given but seldom taken. After Installation of Overhead Groundwires Once the non-insulated overhead groundwires (also called shield or static wires) have been installed, the towers of a transmission line are bonded together electrically. Utilizing the measurement technique described in the previous section would not provide valid results for lightning, since one would be measuring not only the local ground electrode but all the adjacent towers connected via the ground wires. This makes the “normal” three-terminal resistance value at each tower meaningless unless the ground wires are insulated from the transmission structure. Insulating the groundwires is an expensive and labor intensive task that usually requires live-line work and hold-off procedures that can be difficult to schedule. Recently-Described Test Methods for Footing Impedance High Frequency Resistance Measuring Instrument – BBC HW2W In 1976, the Brown Boveri Corporation (now ABB) introduced a three-terminal impedance meter, designed to allow the measurement of a local tower footing impedance without undue influence of adjacent towers connected via the ground wires. This instrument operates at 26 kHz and is designed to operate accurately in the resistance range of 3 to 15 Ω. By using a higher frequency, the effect of the nearby towers is reduced since the impedance of the ground wires is mainly inductive reactance. A power amplifier with 50-V output is provided along with instrumentation with suitable sensitivity to measure 50 mA currents accurately. Measurement probe length is 75 m (250 feet) for the remote current injection (C2 probe length) and 40m (133 feet) or 53% of C2 for the P2 probe. According to the product literature (3.6.84): “The tower diffusion resistances measured with the HW2W roughly correspond to the surge impedance caused by lightning stroke, i.e. the earthing is measured only to such an extent as it becomes effective during shock wave load. This applies for all elongated earthing systems such as ground wires, radial earthings, cable sheathings etc.” In the early 1980s, at least four utilities evaluated the BBC HW2W against conventional measurements [Leppla 1976, Hunt 1981, Carberry 1983, Dick 1986]. Generally, the measurements of low footing impedance values in the range of 5 to 20 Ω were satisfactory, with

1-3

[Carberry 1983] noting large errors between instruments, with a discrepancy of 86 Ω versus 48 Ω for low versus high frequency measurements. Dick [Dick 1986] carried out the most comprehensive calibration studies on the BBC HW2U unit, including bench testing with simulated circuit components, and cross-comparison with field measurements using traditional four-terminal devices, including DET-2 and ET-5 units. A network analyzer was used to obtain complex impedance readings at four frequencies. The footing impedance was also estimated using measured Wenner resistivity profiles along with two-layer soil interpretation. The main disadvantage of the BBC/ABB 26-kHz meter is its poor accuracy when measuring high footing impedance values, greater than about 20 Ω. Table 1-1 shows the results for towers located in areas of low soil resistivity where the BBC meter had adequate accuracy.

1-4

Table 1-1 Cross-Calibration of Tower Footing Resistance Measurements, presented as Magnitude (earth resistivity testers) and Magnitude ∠Phase Angle (network analyzer) [Dick 1986] Q2/6A

Q2/6A

A36/37N

A36/37N

Q3L/4N

Tower 54

Tower 44

Tower 36

Tower 21

Tower 10

Direct Method (between Tower and Isolated Overhead Groundwires) As measured

-

As corrected (OHGW R π0)

7.31 Ω

8.97 Ω

3.3 Ω

8.0 Ω

-

3.45 Ω 2.9 Ω

Fall of Potential (Isolated Overhead Groundwires) DET-2 130 Hz, 75 x 40 m probes

-

2.93 Ω

7.83 Ω

2.35 Ω

2.36 Ω

DET-2, 130 Hz, 100 x 62 m probes

1.70 Ω

3.35 Ω

8.30 Ω

2.63 Ω

2.59 Ω

ET-5, 130 Hz, 75 x 40 m probes

-

2.48 Ω

8.27 Ω

2.93 Ω

-

ET-5, 130 Hz, 100 x 62 m probes

2.03 Ω

2.97 Ω

8.54 Ω

3.24 Ω

2.67 Ω

HW2U, 26 kHz, 75 x 40 m probes

-

3.7 Ω

14 Ω

2.9 Ω

3.0 Ω

Impedance, 67 Hz, 100 x 62 m probes

1.95∠-1°

3.19∠-2°

10.5∠0°

2.75∠0°

2.62∠-1°

Impedance, 600 Hz, 100 x 62 m probes

1.94∠4°

3.17∠1°

13.8∠0°

2.74∠3°

2.05∠-165°

Impedance, 6 kHz, 100 x 62 m probes

2.38∠29°

3.22∠18°

10.9∠10°

3.5∠17°

3.01∠-80°

Impedance, 10 kHz, 100 x 62 m probes

2.96∠39°

3.33∠28°

11.5∠20°

3.2∠19°

6.20∠-172°

Fall of Potential (Overhead Groundwires Connected) HW2U, 26 kHz, 75 x 40 m probes Impedance, 10 kHz, 100 x 62 m probes DET-2, 130 Hz, 100 x 62 m probes

2.4 Ω

3.7 Ω

13 Ω

2.9 Ω

2.45 Ω

-

-

10.8∠0°

2.48∠37°

7∠-172°

1.03 Ω

1.90 Ω

1.12 Ω

0.79 Ω

0.86 Ω

Soil Resistivity Calculation ρ1 , depth, ρ2 Apparent Resistivity ρe Calculated Footing Resistance

30 Ω-m, 45 m , 300 Ω-m

26 Ω-m, 5.4 m , 1560 Ω-m

500 Ω-m, 2.3 m , 115 Ω-m

37 Ω-m, 11 m , 330 Ω-m

33 Ω-m, 13 m , 330 Ω-m

45 Ω-m

57 Ω-m

115 Ω-m

54 Ω-m

46 Ω-m

2.6 Ω

3.2 Ω

8.3 Ω

2.7 Ω

2.6 Ω

Table 1-1 shows that the BBC HW2U readings, with connected or isolated overhead groundwires, did not differ much from the readings obtained from three-terminal resistance when overhead groundwires were isolated. All tested footing resistances were less than the 15-Ω upper limit of the BBC instrument. In four of the five cases, the network analyzer results were also in close agreement, with only one tower giving a rising impedance as a function of frequency. This is a characteristic of a distributed (radial wire) rather than a concentrated (grillage, foundation) ground electrode. High Frequency (2-kHz) Directional Impedance

1-5

For testing transmission and distribution pole bonds, consisting of single conductors with maximum diameter of less than 1”, a number of clamp-on ground resistance testers are offered in the market. Table 1-2 provides detailed specifications of five typical units. Clamp-on ground resistance testers find their main use in multi-grounded systems where it is time-consuming to disconnect the ground under test. These instruments generally run at an excitation frequency of about 2 kHz. The iron-core transformer designs couple test energy in two directions: upwards into the overhead groundwire or neutral system and downwards into the local grounding system. The currents in each direction are monitored to give the driving point impedance, and the fraction of current that flows down into the ground electrode is used to calculate the local electrode resistance. Measurements performed on intact ground systems prove the initial quality of the grounding connections and bonds. Resistance and continuity of grounding loops around pads and buildings may also be measured with these instruments. Table 1-2 Specifications of Clamp-On Ground Resistance Testers (2004) AEMC

Extech

HEME

Meco

Meco

3711

382356

GEO 15

09

4680B

Minimum Resolution

0.01 Ω

0.02 Ω

0.025 Ω

0.1 Ω

.002 Ω

Maximum Reading

2000 Ω

1500 Ω

1500 Ω

200 Ω

1500 Ω

Jaw Size

32 mm

23 mm

23 mm

2x34 mm

35 mm

1.25”

0.9”

0.9”

1.34”

1.38”

8h from 9V

n/a

40 mA from 9V

380 mA

40 mA from 9V

1 mA

0.001 mA

0.001 mA

0.1 mA

0.001 mA

1689 Hz

n/a

1667 Hz

n/a

1667 Hz

Accuracy at 10 Ω

0.25 Ω

0.5 Ω

0.5 Ω

0.5 Ω

0.5 Ω

Accuracy at 30 Ω

0.55 Ω

0.9 Ω

0.9 Ω

2Ω

0.9 Ω

Accuracy at 100 Ω

4Ω

4Ω

4Ω

5Ω

4Ω

Accuracy at 300 Ω

23 Ω

20 Ω

20 Ω

n/a

20 Ω

Approximate Retail Cost, US$

$2000

$1330

$1350

Battery Life / Operating Current

Current Resolution Frequency of Operation

5 Nicad

The main disadvantage of this class of equipment is that the jaw size is not large enough to clamp around a lattice tower leg. The results for multiple-leg towers and for long spans are misleading. Also, at some utilities, the connection between the neutral and overhead static wire is strictly mechanical and presents an open circuit at low voltage. Pseudo-Random Noise Injection: The Smart Ground Meter In the early 1990s, EPRI developed a technology for measuring substation resistance that was later commercialized as the Smart Ground Meter report [Meliopoulos 1993, 1993a, 1994, 2004]. The present Smart Ground Meter configuration uses one current injection probe and six surface potential probes to carry out the following range of measurements:

1-6

1. Ground Impedance of a substation ground system, given the approximate shape and the relative positions of the meter, voltage and current probes as x-y coordinates 2. Soil resistivity measurements, given the separation between a line of nine equally-spaced voltage and current probes 3. Tower ground resistance, using a current return electrode 100 to 150 m (300 to 500 feet) from the tower and six voltage potential probes in two directions, with all probe leads running at right angles to the overhead groundwires. The resistive part of the measured resistance is reported at a measurement bandwidth of 500 Hz. 4. Touch voltage measurements at the earth surface beneath a ground grid. 5. Step voltage measurements using approximate shape of the ground system under test and the relative positions of the meter, voltage and current probes as x-y coordinates. 6. Ground Mat Impedance. 7. Low-Impedance / Continuity measurements to ensure point-to-point integrity of ground grids. 8. Transfer Voltage, using a current return electrode, three surface voltage probes and a connection to the other ground system under test. 9. Fall-of-Potential Method for Ground Impedance, using a current return electrode at distance D and three voltage probes, all located at 0.62D. 10. Oscilloscope function to monitor 60-Hz and harmonic voltages on the test leads or other systems. The equipment injects a pseudo-random sequence of currents using an internal 500-V generator. The frequency spectrum of the sequence extends to 1 kHz. Roughly 70,000 voltage and current samples at 4092 samples per channel are processed to develop statistics on signal reliability and confidence level. Tower ground impedance is plotted as a function of frequency with a horizontal scale of 0 to 250 Hz. Phase relations among signals are used to remove the effects of ground-wire currents that make three-terminal measurements invalid. Utility experience with the Smart-Ground meter has been mixed, with some utilities obtaining good and consistent results and others reporting a wide variation in readings, depending on trivial changes in probe locations.

1-7

Excitation with Existing AC Current The following process can be used to take advantage of stray induced currents present on most transmission towers. • Establish or monitor the current flowing into the tower base from ac system unbalances. • Take measurement of the potential rise to remote earth. • Ratio gives tower footing resistance. • Valid at 60 Hz and possibly harmonics. This approach was tested at one AEP tower but the rapid changes in tower-base current on a five-second time scale did not give meaningful results. However, this is a research idea that could be developed further if its practical advantages (use of existing neutral voltage for excitation) are found to have merit. Test Methods for Earth Resistivity Table 1-1 reported relatively good success in calculating tower footing impedance from local soil resistivity measurements. A Wenner survey, involving placement of four equally-spaced probes at distances ranging from 1 to 100 m, can take about an hour to collect and analyze the data at a single location. However, there are several commercial sensors that can be adapted to provide nearly instant indications of resistivity. These are described generally in the following sections. Electromagnetic Induction

Figure 1-2: Electromagnetic Induction Sensor In Operation [www.dot.ca.gov/hq/esc/geotech]

Soil resistivity measurements using induced electromagnetic fields, rather than direct injection of electrical current into ground probes, are fast and inexpensive because direct contact with the earth is not required. Figure 1-2 shows a typical sensor in use. Electromagnetic induction observations can be used to detect voids, to delineate areas of fill and to make qualitative assessments of soil and rock distribution. In addition, EM induction is used in many areas to locate buried metallic services, well casings, archaeological artifacts and valuables. 1-8

Figure 1-3: Plot of Eddy Current Density in High-Conductivity Soil under Coil at High Frequency

Eddy currents are closed loops of induced current circulating in planes perpendicular to the magnetic flux of a coil. They normally travel parallel to the coil winding and the eddy current flow is close to the area of the inducing magnetic field. Figure 1-3 shows that the eddy currents from a vertical coil concentrate in the ground below, with the strongest density near the coil surface. The eddy current strength decreases exponentially with depth, a phenomenon known as the skin effect. The skin effect is observed in all resistive materials. As eddy currents flow in the ground at any depth, they produce magnetic fields that oppose the primary field. These fields reduce the net magnetic flux and reduce the current flow as depth increases. The depth that eddy currents penetrate into a material is affected by the frequency of the excitation current, the electrical conductivity and the magnetic permeability of the material. The depth of penetration decreases with increasing frequency and increasing conductivity and magnetic permeability. The skin depth δ is the depth at which eddy current density has decreased to 1/e, or about 37% of the surface density.

δ =

1 π fµ oσ

δ is the skin depth in meters; µo is 4π x 10-7 Henry per meter; σ is the conductivity in Siemens per meter; f is the frequency in Hz. Classical metal detectors have operating frequencies of 5 to 30 kHz, with δ=70 to 30m, have been optimized for finding metal objects located less than 1 m beneath the earth surface using the difference in frequency between a reference oscillator (unloaded coil) and an identical oscillator, loaded with a search coil near the ground.

1-9

Ground-Based Two-Coil Multi-frequency Electromagnetic Surveys

Figure 1-4: GEM-2 Multi-frequency Electromagnetic Sensor for Earth Resistivity

Figure 1-4 shows a GEM-2 device for carrying out electromagnetic resistivity surveys [Huang 2003]. This device uses two coils, one to transmit signals and a second, located a fixed distance away, tuned to receive the electromagnetic energy. When the antennas are held near a conductive disturbance such as lossy earth, there are small changes in the in-phase and quadrature responses. Typical returns for 500 Ω-m resistivity are less than 1000 parts per million [Huang 2003]. These changes can be interpreted using the mutual coupling ratio Q, defined as [Keller and Frischknecht 1966]: Equation 1-1: Mutual Coupling Ratio for Horizontal Coplanar Coils • λ2 + j 2π fµσ H sec 3 2 λ − Q= − 1 = −r Ú λ J 0 ( λr )e − 2 λh dλ H pri λ + λ2 + j 2π fµσ 0

Equation 1-2: Mutual Coupling Ratio for Vertical Coplanar Coils • λ − λ2 + j 2π fµσ H sec 2 Q= − 1 = −r Ú λ J 1 ( λr )e − 2 λhdλ 2 H pri λ + λ + j 2π fµσ 0

h is the sensor height r is the coil separation Jo(λ), J1(λ) are Bessel Functions f is the transmitter frequency µ is the magnetic susceptibility σ is the earth conductivity, which is the inverse of resistivity ρ With some considerable mathematical skill, Equation 1-1 and Equation 1-2 can be inverted to obtain estimates of earth conductivity σ if all other variables, including measurement frequency, are known or fixed. By walking around, or by towing a large array under a helicopter at a known height, it is possible to build up a two-dimensional map of conductivity by taking many readings. The mutual coupling of the coils is more strongly affected by conductivity of deep 1-10

layers at low frequency and by surface-layer conductivity at high frequency. This means that different frequencies have different depths of penetration, and a three-dimensional map of the earth conductivity can be built up by taking mutual-impedance measurements at several frequencies at each test point on the test area. The inversion process of the collected data to obtain a good estimate of two-layer soil resistivity on a walking survey is similar to those described for airborne electromagnetic surveys using an array towed under a helicopter. Methods for doing this tend to be proprietary to each provider of resistivity measurement technologies. Surveys have been effectively performed by walking, as shown in Figure 1-5.

Figure 1-5: Raw GEM-2 data (in ppm) at three frequencies from a 4-acre utility plant in New

York. Survey line spacing was 5 feet at a horizontal coplanar mode. [www.geophex.com]

1-11

Ground Penetrating Radar Ground penetrating radar uses high-frequency electromagnetic pulses to create profiles of the shallow subsurface. Both high-resolution profiles and plan maps may be produced with this method. Depths of investigation can range to tens of meters, but very shallow investigation depths, 5 meters or less, are more common. Center frequencies of 1-2 GHz give the best propagation characteristics in most soils. The most beneficial uses of Ground Penetrating Radar are for investigation of very shallow features, such as land mines, underground utilities and pipelines, pavements, rebar and void detection in concrete. Under the proper conditions, Ground Penetrating Radar can also be used for deeper geologic investigations, such as bedrock mapping and water table delineation, and may also be used in fresh water to map scour holes and water bottom depth.

Figure 1-6: Typical Ground Penetrating Radar Equipment and Results [www.dot.ca.gov/hq/esc/geotech/gg/gpr.htm]

Because Ground Penetrating Radar attenuates rapidly with depth in the presence of water and conductive materials, such as clay and saline pore fluids, its use is best suited for dry, sandy soils and rock. The application in Figure 1- 6 shows resolution of pipes in 2-m depth of concrete, which would be similar to 100 Ω-m soil in most respects. The depth of penetration achieved with the radar signal gives a direct reading of the soil resistivity, and in dry soils penetration of up to 50 m can been achieved.

1-12

Comparison of Resistivity Test Methods There is no perfect method for measuring the resistivity quickly at every transmission tower. The traditional Wenner method calls for pounding in at least sixteen probes. The ground-based induction coil has promise but is optimized for detecting anomalies and reporting them in units of parts per million, rather than for reading out resistivity in ohm-meters. Ground-penetrating radar has the promise of being able to find buried counterpoise and nearby services (pipeline, phone) on the right-of-way but its present cost does not deliver sufficient value compared to traditional call-before-digging procedures. Table 1-3: Comparison of Alternative Resistivity Measurement Methods

Equipment Cost

Traditional Wenner Probe

Ground-Based Induction Coil

Ground-Based Two-Coil EM Method

Ground Penetrating Radar

Low

Low

Medium to High

High

Time Per Point

6 probe spacings: 5 s per point at least 600 s

Detail

Lowest probe spacing

Depth of Penetration Ease of Interpretation

5 s per point to Towed array, 10m/s

Towed array, 10m/s

0.1-5 m

Coil Separation

0.02 m

Largest probe spacing

1-10 m

0.5 to 50 m varies with frequency

3 to 10 m

Minimize error sum based on six infinite series

Ad-hoc rules

Inversion of integrals of Bessel Functions

Proprietary highspeed image processing

The Zed-Meter concept was not designed at the outset to measure resistivity. However, as described in the next section, the input impedance of the current reaction lead is affected by the soil resistivity. In contrast to the tiny 1000-ppm (0.1%) change that needs to be measured with two-coil EM methods, the Zed-Meter lead impedance changes by about ±20% near ground, depending on the resistivity. This high sensitivity simplifies equipment design and improves signal to noise ratio for equipment that is being operated directly under power lines.

1-13

1-14

2 TRANSIENT INJECTION AT TOWER BASE Description The overall scope of the Zed-Meter transient injection concept is to: • • • •

Inject a transient current into the tower base. Measure the potential rise at the tower base relative to a remote ground. Compute the ratio of the potential rise to the input current as a function of time. Compute the footing impedance by compensating for the fraction of the surge current flowing into the overhead groundwires. The interpretation of the measured data takes place in the time domain. This allows users to reject noise and early oscillations related to wiring and tower structure, and to retain only the features of the response that are relevant to establishing the peak voltage across insulator strings under lightning surge conditions. Also, it is easier to diagnose instrumentation problems with triggered, averaged time-domain measurements in the field than to identify problems with spectrum-analyzer equipment. The Zed-Meter processing strategy includes: • Taking measurements after the effects of the tower surge response have rung down. • Taking measurements of the potential rise before the effects of adjacent towers have time to affect the reading • Relying on propagation at the speed of light, at a rate of 3x108 m/s (1 foot = 1 ns =10-9 s), in overhead wires • Relying on propagation at a half to a third the speed of light, at a rate of 1-2x108 m/s (1 foot = 2-3 ns), in wire systems near the ground. Generally, the valid results are measured between 200 ns and 2 µs. Timing Diagrams Electromagnetic waves propagate at the speed of light, 3x108 m/s, in free space and wire systems such as transmission lines. If a surge is applied to a tower, it will take at least 1 µs for this surge to reach an adjacent tower, 300 m away, and 1 µs for any influence exerted by the adjacent tower to cause voltage or current changes back at the source. The two-way propagation time of the wave will be two times the span length, divided by the speed of light, or 2 µs.

2-1

Figure 2-1 : Time Sequence of Current Wave, Injected into Tower Base

(0)

(1)

(2)

(3)

(4)

(5)

(6)

(7)

0. Surge is injected between tower leg and horizontal current lead. 1. Surge flows into horizontal current lead whose impedance is constant; surge also flows towards tower into tower leg footing. The measured impedance is that of the current lead in parallel with the single footing. The measured impedance at this stage is not valid since only the impedance of one leg is accounted for. 2. The surge continues to flow into the constant impedance of the current lead. The surge now flows into all four footings and continues up the tower. The measured impedance is somewhat valid, but difficult to extract because the tower impedance is increasing with height. 3. Same as 2, but the surge continues to travel upwards into the tower. The measured impedance is changing as the tower shape changes. 4. Same as 3. The measured impedance changes as the current surge splits into the two overhead groundwires. 5. The measured impedance is the horizontal current lead in parallel with the overhead groundwires (4) and the four footing resistances. The footing impedance is calculating by un-paralleling the impedance of the overhead groundwires using the equivalent circuit shown in Figure 2-2. The overhead groundwire surge impedance is well defined by the height at the tower.

2-2

6. Same as 5. Minor decrease in overhead groundwire height due to sag has negligible effect on result. 7. Same as 6. … 43. On this time scale, current wave arrives at typical adjacent towers. Results still completely valid. 86. On this time scale, reflections from grounds of adjacent towers arrive back at tower base. Results for this individual tower footing impedance no longer valid beyond this time. Results for predicting overall lightning impulse voltage on insulator remain valid. Generally, the impedance results for a transmission tower will no longer be valid when pulses return from adjacent spans.

Figure 2-2: Equivalent Circuit at Steps (5-7) of Time Sequence in Figure 2.1

Overhead Ground Wires: Four in Parallel = 163 Ω

Footing 30 Ω

Current Lead 450 Ω

In the 1980s, a two-way propagation time of 2 µs would allow 4 cycles for the then-typical 2MHz microprocessor CPU – enough computing power to fetch and store away a single reading from an analog-to-digital converter. In 2004, processors running at 3 GHz would execute 6000 cycles in the same propagation time. Digital instrumentation has improved at the same rate as computers, because it uses the same memory and semiconductor technology. This means, approximately, that an engineer in the 1980s would have been able to take a single digital reading for the entire time evolution shown in Figure 2.1, where we can presently take 1500 separate snapshots. The ability to measure a time-evolved history of the injected current and the resulting voltage is what makes the Zed-Meter concept practical now.

2-3

Test Lead Design The Zed-Meter concept in Figure 2.1 shows a horizontal wire, laid on the surface of the ground at right angles to the tower and overhead groundwires. The electromagnetic coupling between these wires will be negligible, and this simplifies the field interpretation. While the wires are conductive and grounded at the remote ends, it is their surge impedance over ground that is important in the measurement process. Figure 2.2 suggests that this wire has a constant surge impedance of about 450 Ω, and this is estimated as follows. The surge impedance Z , in ohms (Ω), of a wire over perfectly conducting ground is a function of the wire height h and radius r, is given by Equation 2.1. Equation 2-1: Surge Impedance of Wire over Ground Plane

Ê 2h ˆ Z wire = 60 lnÁ ˜ Ë r ¯ When the ground is imperfect, the surge impedance is a complex number, corresponding to the substitution of a complex height into Equation 2.1 to give Equation 2.2 [Deri – 1981]: Equation 2-2: Surge Impedance of Wire over Conductive Earth

Ê 2h + p ˆ Z wire = 60 ◊ lnÁ ˜ Ë r ¯ 1 p= jωµ oσ

Where: p is a complex depth in meters (real + imaginary depth) ω is the frequency in radians/s, µo = 4π 10-7 H/m and σ is the soil conductivity in S/m. Soil conductivity is the inverse of soil resistivity ρ, so that σ=1/ρ. Equation 2-2 is an accurate, simple substitute to Carson’s integral expressions. Figure 2.3 illustrates that the computed impedance of an infinitely long 10-mm wire, insulated from but laid directly on the earth’s surface, is a slowly varying function of both earth resistivity ρ and frequency. The significance of this variation lies in our ability to measure the surge impedance of a wire accurately. With the existing Zed-Meter equipment, it was possible to resolve currents and voltages with a precision of better than 2%, even though the leads were located directly under EHV power lines. Using Figure 2-3, this 2% uncertainty in impedance would give a resistivity estimate with about 23% uncertainty. While on face value this seems unimpressive, it is worth noting that processing variations, or a simple repetition of a Wenner survey profile on a parallel survey line, routinely give values that differ by more than 10%.

2-4

Figure 2-3: Surge Impedance of Wire on Ground Surface (Ω) versus Soil Resistivity for Various Frequencies

Interpretation of Measurements Surge Impedance of the Current Lead For a practical range of frequencies of interest to lightning performance, a wire laid on the ground has a surge impedance of about 450Ω. The slow change in surge impedance with a factor of 100 change in soil resistivity makes this a parameter that can possibly be used for resistivity surveys. The empirical expressions fitted to the calculated impedances in Figure 2-3 lead to the following expression for estimating soil resistivity: Equation 2-3

Resistivity of Soil as Function of Input Impedance of 10-mm Wire On Ground

ρ =e

Z − 260Ω 30

The choice of the 260-Ω offset is appropriate for 1-MHz test frequencies, and the variation in Figure 2-3 suggests that the input impedance will rise slowly with increasing pulse duration in areas of constant soil resistivity. Further refinement of this relationship is a recommended task for developing a rapid, relatively accurate estimate of local soil resistivity at the same that the Zed-Meter is used to measure the tower footing impedance. Current Split between Test Lead and Foundation Consider a surge generator that can produce an open-circuit pulse voltage VOC=200V with an internal impedance Zgenerator=50 Ω. If the output of this generator is shorted, a pulse current of Imeas=4A will be delivered. If the 200V is applied between the horizontal wire with surge

2-5

impedance Zwire=450 Ω and a ground electrode with impedance Zx , a current of somewhat less than 0.4A, given by (200 V/ (500 Ω + Zx)), will propagate in both directions. This current can be measured using a wideband current transformer to isolate the measuring system from the pulse generator. If the surge impedance of the wire Zwire over ground was well known, then it would be possible to work out the ground electrode impedance using the following expression: Equation 2-4 Calculation of Foundation Impedance from Observations

Zx =

VOC − Z generator − Z wire I meas

This approach may become feasible in the long run to determine the footing resistance with sophisticated measurement methods. However, the accuracy of measuring Zx will always be improved by using a second wire on the ground. Potential Rise from Foundation to Second Test Lead A second wire, oriented along the line direction, is shown in Figure 2-4 to provide a zero potential reference at constant impedance. The ratio Zx(t) between the transient voltage Vmeas(t) measured between tower base and potential reference wire, and the injected current Imeas(t), will give more accurate estimates of footing response. The current injected into the reference potential wire will be extremely small when modern oscilloscope probes are used, and orientation at right angles to the current injection lead will minimize interference from coupling.

Imeas

Current Lead(s) RG58 with Remote Ground Rod

Vmeas

Potential Lead RG58 with Remote Ground Rod

Figure 2-4: Orientation of Current and Reference Potential Leads for Impulse Testing of Tower Grounds

Figure 2.4 shows considerable separation between the potential lead and the tower leg for clarity, but this separation should be less than 0.1 m in practice to preserve good system risetime. Accounting for the Presence of Overhead Groundwires Figure 2.2 and 2.4 show that the tower and its footing electrode under test, Rx, are connected in parallel with two overhead groundwires. For low frequencies, the overhead groundwires and adjacent towers will form a network of series inductance and parallel resistance. This diversity 2-6

will generally translate into a local impedance of about 2 Ω, relatively independent of the local footing impedance. A good estimate of the overhead groundwire inductance L over perfect earth can be obtained from its surge impedance (Equation 2.1) and its propagation time (given by the span length lspan divided by speed of light c): Equation 2-5: Inductance of Overhead Groundwire

L = 60 ◊

lspan Ê 2h ˆ ◊ lnÁ ˜ c Ë r ¯

For a constant wire height of h=30 m (98’), radius of r=0.007 m (9/16” diameter) and span length 300 m (984’) the self inductance of each groundwire is 543 µH, corresponding to an inductive reactance of 0.2 Ω at 60 Hz and 341 Ω at 100 kHz. At higher frequencies and for timedomain analysis, the self surge impedance of 543 Ω from Equation 2-2 will be seen for 2 µs, at which time the disturbance caused by an impedance mismatch at the remote tower will appear. The wire height is not constant, but takes the shape of a catenary. The sag of typical steel overhead groundwire can be calculated from the approximate expression in Equation 2.5: Equation 2-6: Sag of Steel Overhead Groundwire

Sag steel OHGW =

0.0009 ◊ Span2 % RTS

Span and sag are expressed in meters and a usual stringing tension is about 20% RTS. For a 300-m span at 20% RTS, the sag is calculated to be 4.05 m. The average height of the wire will be the height at the tower, minus 2/3 of the sag. For the model line above, the average height would be 27.3 m and average surge impedance, including the conductor sag, would be 538 Ω, rather than 543 Ω calculated for horizontal wire at tower top. Generally, changes in impedance caused by sag effects can be ignored. Equation 2.2 provided a method to evaluate the effect of soil resistivity on the overhead groundwire impedance. This is used to produce Figure 2.5 for the single wire over ground.

2-7

900

Surge Imepdance, 7-mm Wire at 30 m

800 700 600 500 400 300

60 Hz 10 kHz

200

100 kHz

100

1 MHz

0 10

100

1000

Soil Resistivity, Ω -m

Figure 2-5: Impedance of 7-mm radius Wire, Located 30 m above Lossy Ground

The impedance calculated at frequencies of interest (10 kHz- 1MHz) for the transient injection test is seen to be relatively constant for a wide range of practical soil resistivity values. While the effects of both sag and soil resistivity shown to be small, the mutual coupling of the two overhead groundwires cannot be ignored. The mutual impedance between the two wires is given by: Equation 2-7: Combined Impedance of Four Overhead Groundwires, Two in Each Direction

Ê 2h ˆ Z11 = 60 lnÁ ˜ Ë r ¯ ÊD ˆ Z12 = 60 lnÁÁ 12 ˜˜ Ë d12 ¯ Z + Z12 Z g = 11 4

D12 ª d122 + 4h2

In this case, d12 is the (usually horizontal) distance between overhead groundwires and D12 is the distance from one OHGW to the image of the other OHGW, located at an image height h below the ground. For the case in Figure 2.4, with a separation d12 of 10 m and h=30 m, D12 is 60.8 m, Z11 remains 543 Ω, the mutual impedance Z12 is 108 Ω and the combined impedance of the two groundwires in parallel in each direction is Zg =163 Ω.

Having established a value for Zg, simple circuit theory allows us to un-parallel the unknown footing impedance Zx from the measured impedance of Vmeas/Imeas:

2-8

Equation 2-8: Calculation of Footing Impedance from OHGW Impedance and Measured Values

ZX =

1 I meas 1 − Vmeas Zg

This expression will fail if the measured foundation impedance approaches the calculated impedance of Zg but generally tower impedance Zx << Zg. The mutual coupling from the overhead groundwires to the potential lead on the surface of the earth is relatively small because D12 and d12 from OHGW to potential lead re nearly the same. However, the same is not true in cases where the potential lead is routed near continuous counterpoise or other distributed electrodes. While this is a present weakness with the Zed-Meter concept, it can be overcome with future development by measuring the coupling directly as an extra measurement step. In contrast, measurements at lower frequencies using conventional fallof-potential or Smart Ground Meter approaches are based on fundamental assumptions about the surface gradients that no longer hold true when there is a mix of lumped vertical and distributed horizontal wires in the ground. With existing methods, there is no reasonable way to measure continuous counterpoise resistance. Even if the counterpoise is isolated from both towers, the return electrode for an accurate measurement needs to be at least five span lengths (1500 m or a mile) away, and the low-frequency value obtained, while relevant for system fault current management, does not describe its lightning performance accurately.

2-9

2-10

3 EQUIPMENT SELECTION A wide range of choices, ranging from plug-in cards for laptop computers to specialized fourchannel digital oscilloscopes, can be selected for transient studies at the tower base. In general, test equipment is needed to provide:

• The ability to generate pulse waveforms of suitably large yet safe amplitude • The ability to observe the pulse currents and voltages with suitable fidelity • The ability to store and analyze the results in digital format Selection of Pulse Generator The selection of the pulse generator has the following constraints: • Safe to use • Lightweight and robust • Low power consumption • Capable of providing self triggering • Capable of triggering observation equipment with a digital output that has low jitter Generally, the fast-rising edges and flat top of a rectangular pulse are better for time-domain measurements than excitation using a double-exponential or other high-voltage impulse shape because the current is never constant with the latter waveforms. Many people are familiar with equations for electrocution current, but these equations do not apply for relatively short transients. Figure 3-1 from [Reilly 1998] gives an overall indication of the peak current that causes ventricular fibrillation. The IEC recommend a 500-mA limit for exposure duration of 30 to 200 ms, indicated by the circled (2). The corresponding risk for a 100-µs unipolar (monophasic) pulse would be 30 times higher or roughly 15 A. While this may increase for shorter pulse duration, there is no supporting evidence to say that it does, so a 15-A limit will be respected. Most surge generators with fast rise and narrow pulse width use either a 50-Ω or 75-Ω source impedance. The 50-Ω generators can be obtained with surge outputs of severalkV but, to respect a 15-A pulse current, an upper voltage limit of 750 V is appropriate. Since there is a reasonable chance that a worker could come in contact with the injected surge, it is not sufficient to limit the current to a safe level. The pulse train energy should also be controlled to the same level as an electric fence – annoying but safe. The standards for electric fences establish the maximum pulse energy that would be acceptable in a practical Zed-Meter. In Canada, the CSA [CSA 1992] regulates electric fence apparatus to deliver a charge of less than 1000 µC (= 10 A into 500-Ω animal phantom for 100 µs). In comparison, a “Strong” shock to the fingertip would have a charge of about 2 µC (=2 A times 1 µs).

3-1

Figure 3-1: Electrocution Curve for Various Exposure Times [Reilly 1998]

With the adjacent towers being at most about 500 m away, there is little benefit to providing pulse duration greater than 3.3 µs. A 2-µC limit would allow currents of up to 0.6A into a short circuit, corresponding to an open-circuit voltage of 30V for a pulse generator with an internal impedance of 50 Ω. A 1000-µC limit would allow the use of (15 A) x (3.3 µS) at a repetition rate of about 20 Hz. Most field trials were carried out with a 200-V pulse generator with 0.44-µs to 1-µs duration, using a repetition rate of 20-30 Hz. The 4-A short circuit current provided by this source was needed to provide adequate signal strength and the relatively high voltage tends to punch through thin oxide layers, making test procedures less sensitive to surface preparation. A higher voltage would improve measurement results by further reducing sensitivity to contact resistance. The pulse voltage used in the Zed-Meter is similar to that of the EPRI Smart Ground Meter (SGM) [Meliopoulos 2004] but, because it supplies a continuous signal rather than a series of shortduration pulses, accidental contact between the remote current lead of the SGM and its ground probe would be lethal. Selection of Impulse and Measurement System Response Time It is desirable to excite traveling-wave systems with a step that has a single, fast-rising edge and a flat top. Specialized instruments, called Time-Domain Reflectometers (TDR), have been developed for this purpose. The resolution of a TDR measurement system is directly related to the rise/fall time of the incident pulse. The edge speed of the incident pulse will set the minimum decipherable feature size that can be extracted from the measurement. Also, as the resolution of the measurement system increases, the model of various interconnections can be more detailed, leading to more accurate simulation of the lightning surge response.

3-2

The overall risetime of the measurement system is directly related to the system's TDR resolution. The TDR resolution is a function of the risetime of the incident TDR pulse and the dielectric constant, in cable or dielectric systems. The relation between time-domain step response and feature resolution (such as the location of a cable fault) is: Equation 3-1: Calculation of Minimum Time-Domain Reflectometry Spatial Resolution

tr c 2 εr

λ=

λ is the length of the minimum feature that can be resolved c is the speed of light in air (3x108 m/s) tr is the rise time of the incident TDR pulse εr is the relative dielectric constant The overall system risetime can be evaluated using the root sum of squares of the risetimes of each element in the system, as follows: Equation 3-2: Root Sum of Squares Calculation for System Risetime

tr , system = tr2, pulse generator + tr2,digitizer + tr2, probe + ( 2.57 ◊ jitterrms ) 2

Therefore, providing the fastest-rising pulse possible, using the fastest sampler and probes, and minimizing time base jitter, will all contribute to a system with low risetime and high resolution along the system. The use of a rectangular pulse generator with a trigger output, and a digitizer with a trigger input, will tend to minimize the rms jitter in a noisy environment and this is an important factor in the success of prototype Zed-Meter configurations. Selection of Digitizer The rise time of a digitizer is related to its sine-wave -3dB bandwidth as follows: Equation 3-3: Calculation of Digitizer Bandwidth

f − 3 dB =

0.35 t r ,digitizer

The 0.35 factor changes to about 0.45 for oscilloscopes with more than 1 GHz bandwidth, but these are not needed for accurate characterization of lightning effects on power systems. Given that the minimum feature resolution λ would correspond to a 3-m ground rod, Equation 3-1 gives a system rise time requirement of about 20 ns. If the digitizer rise time is the largest term in Equation 3-2, then Equation 3-3 suggests a 3-dB bandwidth of 18 MHz is sufficient. Normally, 20-MHz bandwidth digitizers will need an external trigger input in order to minimize the rms jitter term in Equation 3-2. The sensitivity of the digitizer is the next selection criterion. Surge currents into the tower were typically in the range of 1 A peak. Typical wideband current sensors with 30-40 MHz bandwidth provide sensitivity of 0.1 to 1 V/A if they have adequate specifications for droop rate (0.1% per µs) and I-t capability (1 mC). This means that the digitizer should have a setting for 100 mV full scale.

3-3

A resolution of eight bits (1/256) is sufficient to give good results for tower footing impedance, but the improved drift and accuracy specifications associated with twelve-bit digitizers will improve measurement precision, which will be needed for successful evaluation of wire surge impedance to obtain local resistivity using Equation 2-3. The memory needed for adequate processing is relatively modest. With a sampling rate of 40MS/s and signal duration of at most 3.3 µs, a record length of 256 samples is adequate for capturing pre-trigger levels. With the high ac fields and wide range of local electromagnetic interference on test leads, the ability to average sixteen or more triggered records is much more useful than a high number of bits in the A/D converter. Adequate records of stimulus (injected current) and response (remote potential rise) have been obtained with two-channel digitizers, as described in Chapter 4, but the Zed-Meter field trials showed that the use of three or four channels had additional merit.

3-4

4 SUMMARY OF PREVIOUS FIELD TRIALS Initial Field Trials at New Brunswick Power In conjunction with a program to measure the resistivity near 345-kV power lines using airborne electromagnetic methods [Fugro 2003, Chisholm 2003], a preliminary study was carried out to measure the surge impedance of transmission tower electrodes, before and after improvements such as installation of six radial counterpoise, embedded in electrically conductive concrete. Equipment Used • • • • • • •

Fluke 123 oscilloscope Pearson Model 2100 current transformer, sensitivity 1 V/A Wideband clamp-on current transformer 10:1 Voltage Probes Avtech AVL-S pulse generator with 50-Ω series matching resistance and 100-nF capacitor Laptop computer, Flukeview software and interface cable Two, 60-m reels of 50-Ω coaxial cable and ground rods for voltage and current reference leads • 12-V to 117-V inverter and 12-V 17 A-h battery Figure 4.1 shows that this equipment configuration is highly portable and robust.

12-V Battery/ 117-V Inverter

3-Hz Trigger Source

Laptop

Pearson Current Transducer (CT)

200-V 50-Ω Pulser

Figure 4-1: Impulse Injection Equipment for NB Power Field Trials

4-1

Fluke 123

Summary of Findings As reported in [Chisholm 2003], the test program led to the following conclusions about impulse injection methods: • • •

It was possible for young engineers and experienced line technicians to collect good raw data for later analysis. There was adequate signal-to-noise on 345-kV lines. Daily AC power requirements met by 17 A-h inverter and the internal laptop computer battery.

Automatic extraction of impedance at end of wave was feasible using weighting of: • Half-sine pulse with 440-ns duration (1.1 MHz) • Quarter-sine pulse weight (570 kHz) • Median from 240 ns to 440 ns. For the sine pulses, a sine wave starting at zero trigger time was multiplied by the experimental values of Vmeas(t)/Imeas(t) . This vector of weighted impedance values was integrated over the 440ns measurement time and then divided by the integrated value of the sine pulse over this same time period.

Figure 4-2: Comparison of Observed Footing Impedance and Observed Soil Resistivity from Airborne Electromagnetic Survey, using 7-m Electrode Radius and Two-Layer Soil Model [Chisholm - 2003]

The recommendations for future development of the method were that:

• Longer pulse width (>440 ns) would be desirable. • Faster (and more) averaging would be desirable, but 8-bit / 20 MHz was adequate. • Booting up the laptop at each tower is undesirable (rate-limiting) step.

4-2

5 EPRI ZED-METER FIELD TRIALS EPRI Zed-Meter Configuration The overall layout of the Zed-Meter lead configuration was shown previously and is repeated here, along with a photo of the equipment configuration.

Imeas

Current Lead(s) RG58 sheath attached towith Remote Ground Rod

Vmeas

Potential Lead RG58 sheath with Remote Ground Rod

Figure 5-1: Zed-Meter Lead Configuration Used in September 2003 Tests

Figure 5-2: Zed-Meter Equipment Configuration Used in September 2003 Tests. Left to Right, Pulse Generator, Tektronix TDS3054VB Oscilloscope, 117-V Battery and Inverter

The input pulse was injected between the tower leg and the current lead. Currents in each direction were measured with wideband current transformers, the green and yellow toroids shown in Figure 5-3. 5-1

Figure 5-3: Detail of Current Injection between Tower and Reaction Lead

The connection sequence was as follows. 1. Connect Red Pulse Lead to tower under test using an alligator clip, not shown. 2. Record standing voltage to the RG58 Current Reaction Lead sheath, running away from tower at a right angle to the line direction. 3. Record standing voltage to the RG58 Potential Reference Lead sheath, running along right-of-way. 4. Insert (Green) Pearson Model 2110, sensitivity of 0.5 V/A into 50 Ω, arrow for electron flow pointing towards tower, into Red Pulse Lead. 5. Connect output of Pearson CT to Channel 1 of oscilloscope. 6. Connect Ground of Avtech AVL-S pulse generator output cable to Red Pulse Lead. 7. Insert (Yellow) Ion Physics CT, sensitivity 1 V/A into 50-Ω termination, around Center of Avtech AVL-S pulse generator output cable, with arrow pointing away from tower. 8. Connect output of Ion Physics CT to Channel 3 of oscilloscope.

5-2

9. Connect Center of pulse generator output cable to RG58 Current Reaction Lead sheath. 10. Connect Ground of 10:1 oscilloscope voltage probe to Red Pulse Lead 11. Connect Center of 10:1 oscilloscope voltage probe to RG58 Potential Reference Lead sheath 12. Connect oscilloscope voltage probe output to Channel 2 of the oscilloscope. The use of a common 117-V supply for pulse generator and oscilloscope calls for the use of an inverted configuration, using the local tower as the ground reference, to avoid shorting out the pulse output with the oscilloscope ground connection if the channels are not dielectrically isolated. An Avtech AVL-S pulse generator with 50-Ω impedance and 200-V nominal open circuit voltage was used along with a variable trigger pulse generator, running at 1 kHz. A pulse width of 440 ns was used initially, and was later extended to 1 µs. Both CTs had rise times of better than 10 ns and fall time constants of more than 10 µs. Figure 5-4 shows the applied voltage between tower and ground leads and Figure 5-5 shows the measured currents for the most typical case, injection to the sheath of a single coaxial cable.

Figure 5-4: Typical Applied Pulse Voltage between Tower Leg and Remote Current Lead

5-3

Figure 5-5: Measured Currents into Tower Leg (Blue) and Remote Current Lead (Violet)

The currents are being injected into two arms of an antenna system. The tower leg has a larger surface area, compared to the sheath of the RG-58 cable, so its impedance is initially lower, leading to a higher current. However, as the surge travels further up the tower, and along the wire, the currents balance each other and reach a steady value, determined by the distributed impedances of the overhead groundwire and the horizontal current reaction wire systems. Figure 5-6 to Figure 5-9 show the effects of four different configurations on the injection impedance at the source. These were tested partially to establish whether it would be possible to store the surge energy into a coaxial cable without unreeling it, and partly to see whether the tower injection current could be controlled through the impedance of the reaction wire. Configurations with a good match between tower and current electrode impedance will have the fastest settling times, giving more time for valid measurements with short pulse widths. This would be especially important in measurements on lattice or guyed towers with short span lengths. Four different lead configurations were tested.

• A single coaxial lead, running away from the tower at right angles to the line, settled to 361 Ω when the wire core was energized

• A single coaxial lead, running away from the tower at right angles to the line, settled to (361 – 50) or 311 Ω when the sheath was energized.

• A pair of coaxial cables, placed about 1 m apart, led to 200 Ω input impedance when the signal was split between the two center conductors

• A pair of coaxial cables, placed 1 m apart, gave a reaction surge impedance of (200 – 50/2)= 175 Ω when the sheaths were both energized.

For the large, 765-kV lattice tower, the two-sheath electrode gave the best results but all were considered acceptable, with constant impedance after about 200 ns. Practical considerations of minimizing the number of wires to be run, and maintaining good electrical contact between the leads and the grounding clamps, led to the selection of the singlesheath configuration used in Figure 5-7. A straight lead of 90 m was selected. The two-way transit time of the 90-m wire would be more than 600 ns, so that any reflection from the remote 5-4

end would arrive too late to contaminate the measurements. Future development of the resistivity measurement process should exploit both the transit time and the magnitude of the reflection measured from the grounded or open end of the lead. Since it is not always possible to run the current lead straight off the right-of-way for a distance of 100 ns , a series of zig-zag configurations were tested, using the full extent of the wire. Generally, these arrangements led to more ringing on the currents and the currents did not settle to the same value as quickly. A figure of merit of the agreement between the two measured currents was adopted in the evaluation of these results. While the main focus of the measurement method is to inject a known current into the tower base and to measure the remote potential rise of the footing, there is some additional information that can be extracted from the fast-front details of the records. To help with these future programs, parameters of the current and voltage leads (length, diameter, geometry) should continue to be recorded and the pulse generator output voltage should be recorded along with the tower base rise.

5-5

200 ns – 357 Ω 300 ns – 366 Ω 400 ns – 361 Ω Pulse Voltage – 180.6 V

Figure 5-6: Input Impedance between 765-kV Tower 29-22 and Core of RG-58 coaxial cable


Figure 5-7: Input Impedance between 765-kV Tower 29-22 and Sheath of RG-58 coaxial cable

5-6


Figure 5-8: Input Impedance between 765-kV Tower 29-22 and Cores of Two Parallel RG-58 cables 200 ns – 162 Ω 300 ns – 173 Ω 400 ns – 167 Ω Pulse Voltage – 158.1 V

Figure 5-9: Input Impedance between 765-kV Tower 29-22 and Sheaths of Two Parallel RG-58 cables

5-7

Overall Test Conditions The tests were carried out in western Virginia at the locations indicated in Tables 5-1 and 5-2. Table 5-1: Locations and Descriptions of AEP Transmission Towers in 2003 Field Tests

Line

No.

Type

Std

Config

Grounding

Longitude

Latitude

138kV Self Supporting Steel Funk Loop

40-53/2

2-Cct lattice

X2A

Dead End

Grillage

-80.16246

37.22891

Funk Loop

40-53/3

2-Cct lattice

S2A

Angle

Grillage

-80.16099

37.22600

Funk Loop

40-53/4

2-Cct lattice

S2A

Tangent

Grillage

-80.15882

37.22216

Funk Loop

40-53/5

2-Cct lattice

S2A

Tangent

Grillage

-80.15683

37.21864

Funk Loop

40-53/13

2-Cct lattice

T4V CA

Dead End

Grillage

-80.13248

37.18832

Funk Loop

40-53/14

2-Cct lattice

S2A

Tangent

Grillage

-80.13123

37.18606

Funk Loop

40-53/15

2-Cct lattice

S2

Tangent

Grillage

-80.13061

37.18438

Funk Loop

40-53/16

2-Cct lattice

S2A

Tangent

Grillage

-80.12949

37.18140

Funk Loop

40-53/17

2-Cct lattice

S2

Tangent

Grillage

-80.12794

37.17719

Axton Martinsville

27-4

2-Cct lattice

T3S 2

Tangent

Counterpoise + 8 rods

-79.70215

36.64309

Roanoke - Carolina

19-194

2-Cct lattice

M

Tangent

Grillage

-79.82044

36.69680

Roanoke - Carolina

19-212

2-Cct lattice

M

Tangent

Counterpoise

-79.74724

36.68387

Roanoke - Carolina

19-291

2-Cct lattice

M

Tangent

Counterpoise

-79.42535

36.62020

Roanoke - Carolina

19-221

2-Cct lattice

BA

Tangent

Grillage

-79.71396

36.67795

Roanoke - Carolina

19-274

2-Cct lattice

M

Tangent

Grillage

-79.49097

36.63639

Glen Lyn Hancock

42-154/2

2-Cct lattice

X2A

Dead End

Grillage

-80.17113

37.23173

Glen Lyn Hancock

42-154/3

2-Cct lattice

Dead End

Grillage

-80.16884

37.23146

Roanoke - Claytor

40-53/18

2-Cct lattice

X2A

Dead End

Grillage

-80.12760

37.17618

Roanoke - Claytor

45-52

2-Cct lattice

B

Tangent

Grillage

-80.12058

37.17798

Roanoke - Claytor

45-53

2-Cct lattice

B

Tangent

Grillage

-80.12644

37.17657

Roanoke - Claytor

45-54

2-Cct lattice

A

Tangent

Grillage

-80.13158

37.17537

5-8

Table 5-2: Locations and Descriptions of AEP Transmission Towers in 2003 Field Tests Line

Number

Type

Std

Config

Grounding

Longitude

Latitude

-79.70115

36.64294

138kV Wood Pole Structures Axton - Danville No. 1

26-6

H-frame

H1S3

Tangent

Butt wrap

Axton - Danville No. 1

26-13

H-frame

H1S3

Tangent

Butt wrap

18-133

Guyed

138kV Guyed Structure

765 kV Lattice Structures Axton - Jackson's Ferry

29-19

1-Cct lattice

T1TBE

Tangent

Grillage

Axton - Jackson's Ferry

29-22

1-Cct lattice

T1TBE

Tangent

Grillage

-79.71035

36.70442

Axton - Jackson's Ferry

29-40

1-Cct lattice

T1VBE

Angle

Grillage

-79.76376

36.75945

69-kV Wood Pole Structures Ballou - Danville

290-4

1 pole

P0HB

Tangent

Ballou - Danville

290-5

1 pole

P0HB

Tangent

-79.40001

36.61436

Ballou - Danville

290-38

1 pole

P0HB

Tangent

-79.42352

36.59109

5-9

Overall, the FCC map of resistivity indicates relatively difficult conditions through most of Virginia, with an average conductivity of 2 mS/m as shown in Figure 5-10. Towers near the Danville area, roughly 36.6° latitude, had higher average conductivity of 4 mS/m.

Figure 5-10: FCC Measurements of Conductivity for Virginia in mS/m. Note: 2 mS/m = 500 Ω-m, 4 mS/m = 250 Ω-m). Survey areas (Danville, Roanoke) indicated by circles.

5-10

Figure 5-11: Extra Precipitation in September 2003, Prior to Tests

There was an extra 150 mm of precipitation on top of a long-term average value of 100 mm in the western part of Virginia in September 2003. Since there is a strong relation between soil moisture content and resistivity, it was reasonable to expect that local readings during the measurement campaign would be lower than long-term average values.

5-11

5-12

6 EPRI ZED-METER RESULTS 765 kV Lines What Was Expected Previously, the surge injection method provided adequate signal-to-noise levels under 345-kV horizontal configuration lines, using 200-V pulse magnitude from 50 Ω, giving 4 A of shortcircuit current. Oscilloscopes have a limited ability (dynamic range) to extract small signals from large, uncorrelated ac noise. With the higher 765-kV system voltage, the ground-level electric fields and induced potentials on measurement wires were expected to be higher. With the large structure height and regular shape, it was expected that the initial transients (up the tower leg before reaching the body panel and up the tower to the overhead groundwires) would be distinct features. The heights from tower base to the tower body was measured to be 8-12m (two way travel time of 50-80 ns). Overhead groundwire heights were between 39 and 47 m, with two-way travel times of 260-310 ns. What Was Found There was a wide range of standing voltages found on the safety tests prior to connecting instrumentation. The largest standing voltages and currents were at Tower 29-40, with 42 V peak-to-peak and 60 mA p-p with a triangular 60-Hz waveform into a remote ground probe, giving a ratio of 700 Ω. The probe resistance was measured to be 570 Ω with a DET-2 megger, which indicated “High Noise” conditions. However, voltages less than 1V were measured on remote leads on Towers 29-19 and 29-22. In spite of the high noise levels, Figure 6-2 from Tower 29-40 shows clean waveforms, leading to a pair of acceptable measurements of the ratio of tower base potential to injected current in Figure 6-3Error! Reference source not found.. Figure 6-2 alsoError! Reference source not found. shows that the initial split of current between the tower and ground electrode was about 400 mA into the ground and 250 mA into the tower, for a total of 650 mA. This persisted for only 50 ns, at which time a reflection from the lower impedance of the three other tower legs appeared. The footing and reference-lead currents start to match at about 220 ns. The currents rise continuously from 570 mA at 430 ns to 640 mA at 900 ns. Figure 6-3 Error! Reference source not found. shows a series of oscillations in the ratio of tower base potential to injected current that settle to a steady-state value of 10 Ω after about 300 ns (3.0E-07 s). Generally, the tower leg response was stronger than the tower-top response, and this is partially because the tower surge impedance above the leg-body junction is fairly well matched to the overhead groundwire impedance (two in each direction) of about 130Ω.

Tower 29-40 This lattice structure had double overhead groundwires at a height of 47.1 m above fourconductor bundles at a height of 35.5 m. A grillage foundation was given in AEP records.

6-1

Figure 6-1: AEP 765-kV Transmission Tower Leg (left) and Structure (right)

Figure 6-2: Measured Voltage and Current in Tower 765-29-40

An impedance profile is obtained by dividing the voltage at each time increment in Figure 6-2 by the corresponding current value. Only the meaningful values up to the end of the pulse (440 ns) are shown in the resulting Figure 6-3.

6-2

Figure 6-3: Calculated Footing Impedance for Tower 765-29-40

This initial trial used the minimum pulse width of 440 ns, which was later extended to 1000 ns by adding coaxial charging cable and reducing repetition rate. The measured footing impedance reached about 27 Ω at 440 ns, just before the falling edge of the injected pulse. Both of the injected currents agree closely after about 80 ns, corresponding to the 12-m distance from ground to tower body panel. Tower 29-22 With additional experience, the test pulse length was increased to nearly 1 µs and the pulse repetition rate was reduced to keep the pulse train energy level within safe limits. The additional pulse length allowed the current and voltage to stabilize better. Note that there are no strong disturbances returning from the ground rods terminating the potential and current reaction wires, which are within 700 ns of the injection point at speed-of-light propagation. This reduced propagation speed was initially noted by Bewley on buried counterpoise.

Figure 6-4 Measured Voltage and Current at 765-kV Tower 29-22

6-3

Figure 6-5 Calculated Footing Impedance versus Time for 765-kV Tower 29-22

The initial overshoot associated with tower leg impedance in Figure 6-5Error! Reference source not found. is smaller in magnitude and narrower in width than in Figure 6-3Error! Reference source not found.. This results from the difference in leg length, giving an overhead groundwire height of 39 m at Tower 29-22 an 47 m at 29-40. Tower 29-19 The injected current waveform for Tower 29-19 is shown in Figure 6-6. Unfortunately, the computer file for the voltage record was corrupted, but the field notes show a measured tower potential rise of 10 V at 400 ns. With 820-mA current, this gives an impedance value of 12 Ω.

Figure 6-6 Observed Injection Current versus Time for 765-kV Tower 29-19

6-4

A calibration check was carried out at Tower 29-19 to make sure that the low impedance values noted on the tail of wave were characteristics of the tower, rather than the measurement circuit. The output of the surge generator was shorted from tower to reference lead. The tower base potential was monitored against the horizontal reference wire. Figure 6-7 shows a 5A pulse and a tower-base transient of nearly 30 V, giving an early peak impedance “noise” level of about 8 Ω. This noise level decayed within about 100 ns to 3 Ω and was negligible at 440 ns.

Figure 6-7 Noise Level (Measured Impedance into Short Circuit) for Impulse Injection Method

6-5

Comparison with AEP Historic Data Table 6-1: Comparison of Zed-Meter Measurements with AEP Records Parameter

765-kV Tower 29-22

765-kV Tower 29-40

26 Ω

7Ω

820 mA

650 mA

570 mA

244 Ω

308 Ω

351 Ω

Resistivity Estimate from Equation 2-3 (1 MHz)

0.6 Ω-m

5 Ω-m

21 Ω-m

Standing Voltage relative to Remote Ground

0.8 Vrms

0.2 Vrms

42 V p-p

Meggered Resistance of Current Probe at End of Wire

525 Ω

n/a

570 Ω

Ratio, Tower Base Rise / Injected Current at 440 ns

12 Ω

10 Ω

27 Ω

Construction Footing Resistance (AEP Records)

765-kV Tower 29-19 n/a

Peak Injected Current (440 ns) Surge Impedance of Current Reaction Wire

The variation in steady-state injection current – 820 mA for Tower 29-19, 650mA for Tower 2922 and 570 mA for Tower 29-40 - suggests some local variation in soil resistivity. Equation 2-3 can be used to invert the impedance values but the results for all towers are extremely low. The Zed-Meter results did not agree with the two values in the AEP records, either, and this was not resolved as the test program refocused on lower-voltage lines with shorter spans.

6-6

138 kV Lines What Was Expected In the western Virginia area, the AEP 138-kV system crosses some rugged and difficult terrain. With steep changes in elevation, it was anticipated that there would be a wide range of grounding conditions, with low resistivity in valleys and high resistivity in rocky areas. Preliminary data from AEP suggested footing resistance values that would vary from 10 to over 200 Ω. What Was Found The impulse injection method gave results that also varied over a wide range, but tended to be a factor of three lower than the AEP historical records. Generally, the two sets of measurements agreed about which towers were good or bad. With right-of-way restrictions, some measurements were taken using less-than-ideal current injection wire orientation. The compromise was to orient the current lead sideways to the edge of the right-of-way, and then turning 90° away from the auxiliary potential lead. In these cases, the difference between the injected current in the reaction wire and the injected current into the tower took a longer time to decay. The relation between the soil resistivity, estimated from the impedance of the current reaction lead, and the Zed-Meter footing impedance, was not perfect. This is as it should be, since the grounding electrodes (driven rods, drilled well) at some towers were different from others. A “footing dimension”, given by the measured resistivity divided by the measured impedance, gives a very rough indication of the type and condition of the footings, with high dimension indicating a large (good) extent and surface area. Four representative records are given. The first two show the lowest and highest impedance measured on steel lattice towers. The third set of data show results for a guyed tower, and the fourth set give footing impedance for an H-frame design. Table 6-2 then shows the measured values of footing impedance on all the 138-kV lines, along with the available AEP values measured with the overhead groundwires isolated from the tower.

6-7

Line 40-53 Tower 18 Results from this tower are shown because it had the lowest measured impedance in the 138-kV test series.

Figure 6-8 shows the injected current stabilized at 800 mA in about 300 ns, and the measured potential difference reached a 4.3-V plateau after this time. The calculated impedances at 400 ns, 600 ns and 800 ns were 6.3 Ω, 5.8 Ω and 5.1 Ω respectively.

Figure 6-8: Measured Voltage and Current for 138-kV Line 40-53 Tower 18

6-8

Figure 6-9: 138-kV Line 40-53 Tower 18 and its Calculated Impedance

6-9

Line 40-53 Tower 14 Only four towers uphill from Tower 18, this structure had the second-highest ground resistance (100 Ω) of any tested 138-kV self-supporting steel tower in the AEP program and the highest footing surge impedance. The resistivity estimate was surprisingly low, with a wire surge impedance of 271 Ω to be interpreted. However, Figure 6-10 shows the tower base voltage settled to about 25 V at the tail, giving an overall (footing + groundwire) impedance of 40 Ω.

Figure 6-10: Measured Voltage and Current on 138-kV Line 40-53 Tower 14

Figure 6-11: 138-kV Line 40-53 Tower 14 and its Calculated Impedance

After accounting for the parallel impedance of the overhead groundwire, the footing impedance rises from 40 to 47 Ω. As a point of historical interest, this tower was one of those treated by with the very first GE ceramic-housed transmission line surge arresters, appearing in horizontal orientation in Figure 6-11 and apparently still intact.

6-10

Tower 18-133 Previous research on guyed 345-kV structures showed that the decay time was long, compared to lattice structures. The improved instrumentation and methods in the AEP test program were used on a single guyed structure to evaluate effectiveness and to provide some data on whether the initial waveform structure could be eventually used to evaluate tower surge impedance effects.

Figure 6-12: Measured Voltage and Current on 138-kV Guyed Structure 18-133

Figure 6-13: 138-kV Structure 18-133 and its Calculated Impedance

The injected current reached a stable value at about 250 ns, which was similar to the settling time for the taller double-circuit lattice towers. The potential peaked early at 19V and decayed to 15V by 1000 ns. The impedance calculated from the ratio of voltage to current also fell from 22 Ω at 400 ns to 18 Ω at 800 ns.

6-11

Axton-Danville No.1 Line 26 Structure 6 While Structure 13 had a higher footing impedance, Structure 6 is presented because it had a recorded ground resistance measurement of 345 Ω in the AEP records.

Figure 6-14: Measured Voltage and Current on 138-kV H-Frame Structure 126-6

Figure 6-15: 138-kV H-Frame Structure near 126-6 and its Calculated Impedance

The injection current at this site was low, with an impedance of (200 – 40)V / 452 mA = 360 Ω at 1000 ns that gives a resistivity estimate among the highest found. The current took a long time to settle, considering the overhead groundwire height is only 16.3 m, giving a two-way travel time of 110 ns. The loop structure of the H-frame pole bonds and the pole separation of 4.4 m contributed to these oscillations. The impedance at the tail of wave was 80 Ω, and this represents the parallel impedance of footing and overhead groundwires. Equation 2-7 and Equation 2-8 were used to compute a parallel impedance of 150 Ω and a footing impedance of 172 Ω.

6-12

Table 6-2: Comparison of Construction Footing Resistances, Zed-Meter Impedance and Zed-Meter Resistivity Estimates for 138-kV Lines

Tower Number

Tower Resistance in AEP Records

Tail-ofwave Impedance from ZedMeter

Footing Impedance (Removing OHGW Z withEquati on 2-8)

Tail-ofWave Current, mA

Impedance of Wire Over Ground

Zed-Meter Calculated Resistivity

Footing “Size”: Resistivity/ Impedance

40_53_18

11 Ω

5.12 Ω

5.2 Ω

814 mA

241 Ω

0.5 Ω-m

0.1 m

45_53

32 Ω

6Ω

6.1 Ω

795 mA

246 Ω

0.6 Ω-m

0.1 m

19_212

4Ω

6.2 Ω

6.4 Ω

804 mA

243 Ω

0.6 Ω-m

0.1 m

45_54

28 Ω

6.4 Ω

6.6 Ω

788 mA

247 Ω

0.7 Ω-m

0.1 m

27_4

19 Ω

7Ω

7.2 Ω

546 mA

359 Ω

27 Ω-m

3.8 m

575 mA

340 Ω

14.4 Ω-m

1.7 m

40_53_05

25 Ω

8Ω

8.3 Ω

40_53_15

49 Ω

8.3 Ω

8.6 Ω

528 mA

371 Ω

40 Ω-m

4.6 m

19_194

22 Ω

10 Ω

10.4 Ω

583 mA

333 Ω

11.4 Ω-m

1.1 m

19_221

n/a

11 Ω

11.5 Ω

671 mA

287 Ω

2.5 Ω-m

0.2 m

40_53_02

34 Ω

13 Ω

13.7 Ω

540 mA

357 Ω

26 Ω-m

1.9 m

602 mA

318 Ω

7.0 Ω-m

0.5 m

42_154_02

17 Ω

14 Ω

14.8 Ω

40_53_04

49 Ω

17 Ω

18.2 Ω

668 mA

283 Ω

2.1 Ω-m

0.1 m

19_291

26 Ω

17.9 Ω

19.2 Ω

362 mA

535 Ω

9400 Ω-m

491 m

42_152_03

44 Ω

20 Ω

21.7 Ω

550 mA

344 Ω

16 Ω-m

0.8 m

18_133

n/a

21.1 Ω

24.2 Ω

877 mA

207 Ω

0.2 Ω-m

0.0 m

639 mA

291 Ω

2.8 Ω-m

0.1 m

40_53_03

56 Ω

22 Ω

24.0 Ω

40_53_13

119 Ω

23 Ω

25.2 Ω

645 mA

287 Ω

2.5 Ω-m

0.1 m

40_53_17

74 Ω

25 Ω

27.6 Ω

588 mA

315 Ω

6.3 Ω-m

0.2 m

45_52

188 Ω

31 Ω

35 Ω

480 mA

386 Ω

67 Ω-m

1.9 m

40_53_16

99 Ω

37.1 Ω

43 Ω

539 mA

334 Ω

11.9 Ω-m

0.3 m

644 mA

271 Ω

1.4 Ω-m

0.0 m

40_53_14

100 Ω

40 Ω

47 Ω

126_6

345 Ω

80 Ω

172 Ω

452 mA

361 Ω

29 Ω-m

0.2 m

126_13

n/a

83 Ω

187 Ω

436 mA

375 Ω

46 Ω-m

0.2 m

* Large dimension = More Effective Footing

6-13

Figure 6-16: Comparison of Zed-Meter Impedance and AEP Records for 138-kV Lines

Figure 6-16 shows that the relation between the Zed-Meter impedance and previous AEP test values is statistically strong, with a linear regression coefficient R2=0.86, but there is a scale factor of about 0.4 that relates the two readings. As mentioned in Chapter 5, there had been anomalously high rainfall in the month of the tests, and some of the discrepancy can be attributed to soil moisture. However, some of the difference may be related to the change in effective resistivity (combining resistivity, dielectric constant and multilayer effects) at high frequency compared to 60 Hz.

6-14

69-kV Lines What Was Expected With a relatively simple grounding system, consisting of a single bonding conductor and overhead groundwire, the 69-kV subtransmission system is easier to analyze. The only possible complication is the short spacing of the spans, meaning that reflections from adjacent towers can arrive before the measurements have fully stabilized. Complicating this aspect is the fact that the lines are located in urban or suburban areas, where there were sometimes less-than-ideal places to run the test leads. What Was Found Satisfactory current injection was obtained in all five test cases, even when the current injection lead was run over concrete or asphalt parking lots. A zig-zag array was tested and found to be reasonably effective as well. A configuration was tested where current was injected between the bases of two poles, and the potentials from each tower measured against the reference lead along the right of way. This would reduce setup time somewhat and is especially suitable in urban settings.

Figure 6-17 Urban Clutter around 69-kV Wood Pole Circuit, with Test Instruments at Pole Base.

6-15

Line 290 Pole 38

This pole, shown above in Figure 6-17, was an exciting challenge for the Zed-Meter concept. With poor access in the lateral direction, it was decided to keep the current injection “pure” by running this lead along the right-of-way. The potential lead ran 180’ to the side, with the remainder of the 300’ coaxial cable left on the spool. The current stabilized quickly at about 600mA but strong reflections from the adjacent poles (90 m and 126 m away) arrived at about 750 ns, forcing the current negative. Overhead shield wire height was measured to be 21.4 m.

Figure 6-18: Measured Voltage and Current at 69-kV Wood Pole 38 in Parking Lot

< Discuss voltage, current going negative quite soon >

Figure 6-19: Calculated Impedance at 69-kV Wood Pole 38

The low 7.65-Ω impedance measured at about 400 ns was double-checked by inserting a 55-Ω resistor in series with the footing. This caused the injected current to decrease to 580mA, the potential rise to increase to 44V and the calculated impedance to be76 Ω. In other cases where this calibration check was performed, the increases in measured impedance were much closer to the expected 55 Ω.

6-16

Line 290 Pole 4 This 69-kV three-phase wood-pole structure was 34 m and 54 m from adjacent poles. Driven rods at the ends of potential and current leads gave 174 Ω and 207 Ω respectively, indicating reasonably low soil conductivity. The current lead was run on the ground in a zig-zag pattern, roughly 15 m wide and 20 m long.

Figure 6-20: Measured Voltage and Current at 69-kV Tower 290-4

Figure 6-21: Calculated Footing Impedance at 69-kV Tower 290-4

Any initial transients associated with the pole bond should settle down after two travel times. The overhead groundwire height was measured to be 17.0 m, giving a two-way travel time of 113 ns. The oscillations in current and voltage waves continue until about 400 ns, and this is the difficulty in using a non-ideal current lead geometry. However, after this time, the readings are relatively clean and stable. With the relatively short spans, it would be expected that the impedance would start to fall when returns from adjacent towers arrive back at the base of 290-4, similar to the results above for Tower 290-38. The overall time needed for this propagation would be 450 ns to the 34-m tower and 580 ns for the 54-m span. While Figure 6-21 shows the highest impedance at 400 ns, the rate of reduction is surprisingly small at Tower 290-4, giving about 10% reduction at 1000 ns.

6-17

Figure 6-22: Tests on 69-kV Wood Pole Line with Short Spans (J.G.Anderson)

Line 290 Poles 51 and 52 These poles were situated in the middle of a suburban parking lot and again provided an opportunity to stretch the Zed-Meter concept to more difficult situations. In this case, the surge generator was located half-way between the two poles, a distance of 45 m in each direction. The loop current, ground-bond current and ground-bond potential were measured at the base of each pole. The overhead groundwire was 20.5 m above ground and a distribution underbuild shared the pole with the three-phase 69-kV circuit.

Figure 6-23: Measured Voltage and Current at 69-kV Line 290 Pole 51

6-18

Figure 6-24: Measured Voltage and Current at 69-kV Line 290 Pole 52

Figure 6-25: Calculated Impedances for 69-kV Line 290 Poles 51 and 52

In this case, the majority of the current injected into the loop (CH1 in Figure 6-23) initially found its way into the ground connection (CH3), with an increasing fraction being diverted into the overhead groundwire to adjacent structures as time increased. The voltage trace on this pole also shows some high-frequency noise at 1.3 MHz (AM broadcast). A sine wave was fitted to the pre-trigger information, and continued analytically through the impulse record. This procedure was effective in this case, and demonstrated that a potential reference lead, laid on the surface of an asphalt parking lot, without a driven a ground rod at the end, could be considered. While the injected currents were the same in each case, Pole 52 had nearly three times as much potential rise and impedance. Evaluation was carried out at 400 ns to avoid the disturbance of reflections from adjacent poles. AEP’s records show a value of 86 Ω for pole 38 and 464 Ω for pole 5. These are much greater than the impedances found with the current injection method.

6-19

6-20

7 CONCLUSIONS 1. The Zed-Meter approach offers several advantages over conventional testing using isolated overhead groundwires or the other technologies shown in Table 7-1. These advantages include speed (nine towers in a day), signal to noise ratio, insensitivity to minor corrosion on tower steel, safety and data quality, with impedance values taken at time scales that are appropriate for lightning protection analysis. 2. The correlation coefficient between Zed-Meter and AEP records of footing resistance was excellent (R2= 0.86) but the Zed-Meter values were, on average, 40 % of the AEP values for their 138-kV lines. 3. The Zed-Meter approach can probably be used to estimate local soil resistivity but the inversion model using wire impedance and Carson’s equations does not give realistic results. Modeling of “Beverage” horizontal-wire antenna systems with SommerfeldNorton ground model is needed to improve the results. 4. The Zed-Meter approach can be used to obtain estimates of “Footing Dimension”, the local resistivity divided by the local footing impedance. More experience is needed to establish whether this quality measure can be related to the footing condition. 5. It was possible to achieve good signal-to-noise levels in the grounding measurements with a 200-V pulse generator having a source impedance of 50 Ω, even under 765-kV lines. 6. With a 90-m horizontal wire (the sheath of a coaxial cable) as a reaction electrode, the surge current into the tower base was approximately 500 to 800 mA. 7. The surge impedance of the 90-m horizontal wire is exponentially related to the local soil resistivity, with a 10000:1 variation computed using Deri’s formulation of Carson’s equations at 1 MHz. 8. The currents into the tower and into the reaction wire tended to converge after about 250 ns on the 765-kV towers, with faster settling times on shorter, lower-voltage lines. 9. The measured potential rise to remote earth, again using a wire laid on the ground as a constant surge impedance reference, was initially high (corresponding to tower surgeresponse) but settled to a relatively constant value of between 10 and 40 V, depending on local soil resistivity and ground electrode size. 10.The surge current into the reaction electrode can probably be used to establish the resistivity of the earth near the tower. 11.In the 138-kV line measurements, the ratio of tower base voltage to tower current settled to values in the range of 4 to 35 Ω. The measured values were roughly one-third of the recorded construction footing resistance values. 12.With the standard test equipment and laying out two reaction wires, it was possible to carry out a surge injection measurement in less than an hour. 13.The orientation of the current lead was not critical to the success of the injection. For tests carried out when the right-of-way width was restricted, adequate signals were obtained with a zig-zag pattern. Tests should be repeated with a current reaction lead oriented away from the potential reference, using a differential method to quantify and subtract off any mutual coupling effect. Table 7-1: Comparison Table of Methods to Measure Footing Impedance

7-1

Principle of operation Range Shield Wire Removal Required Measures Resistance of Multi-footed towers Measures Steel Pole Resistance Needs Remote Current Probe Measures Overhead Groundwire Surge Impedance Provides Measurement of Resistivity Source Voltage and Safety Comment Frequency of Operation Power Supply

26 kHz injection into tower and ground

1Ω to 1200 Ω No

Smart Ground Meter SGMD2001 and -3001 Broad Spectrum bipolar injection into Tower and Ground 0.01Ω to 1kΩ No

2 to 25 Ω No

Step Transient injection between tower and wire over ground 3Ω to 3kΩ No

Yes

No

Yes

Yes

Yes

Yes

No

Yes

Yes

Yes

Yes

No

Yes

Yes

Yes

No

Assumes impedance is low at 2 kHz

Assumes wire path is inductive, only reports resistive component

Assumes impedance is high at 26 kHz

Yes With 6-10 extra probe setups and measurements

Yes with known length of driven wet iron rod and one measurement

Yes With one extra probe setup and measurement

No

10 to 100 V

Induces current, no open terminals, safe 2 kHz for most meters

250 or 500 V, at least 5 A, lethal continuous output 0-250 Hz, 0-500 Hz or 0-1 kHz, up to 4k samples 1 kVA sine-wave generator / inverter

50 V, Safe body withstand is higher at 26 kHz 26 kHz

Measures parallel impedance of footing and OHGW paths, the value needed for lightning analysis Yes simultaneously using surge impedance of current lead 200 V impulse, safe as an electric fence Time domain to 2 µs; 20-50 MS/s, 4-16k samples Portable ac inverter and 18 Ah 12-V gel-cell battery

Fall of potential meter Waveform generator and four-terminal ohmmeter 0.01Ω to 20kΩ Yes

Square or sine pulse, 1 kHz typical Internal D-cells

Clamp on meter Impedance to 2kHz Induced Current

Internal 9-V

7-2

ABB Meter HW2A

Internal NiCad Batteries (4 A-h)

Zed Meter 2003

Fall of potential meter

Clamp on meter

Smart Ground Meter SGMD2001 and -3001 August 2004

ABB Meter HW2A

Zed Meter 2003

Advantages Well Trusted and understood Cheap

Does not require shield wire removal Cheap

Does not require shield wire removal Reduces noise by averaging and statistics

5 lb, 4 cable reels

Lightweight

20 lb, generator, 3 cable reels

Other Uses

Measures ac current, 1 mA to 30 A

Reduces noise by averaging 16-64 waves 20 lb + 2 cable reels

20 lb + 2 cable reels

Measures step, touch and transferred potentials on lines and at stations

Measures step, touch and transfer impedance under surge conditions for EMC

Has oscilloscope and spectrum analyzer function, 4k records at 4kS/s

Is a wideband oscilloscope with built-in FFT; Pulse generator can test low-voltage surge arresters

Disadvantages Physical Issues

Need to remove shield wires

Probe Routing Issues

Uncertain choice in probe spacing depends on assumptions about size of tower electrode

Issues with Continuous Counterpoise or Buried Radial Crowfoot

Can’t be used - far enough away for accuracy starts to include parallel resistance of adjacent towers

Only works on poles with single down conductor that can be clamped around

Resistances of all probes must be less than 1000 Ω difficult in highresistivity soil Current return is 300-500’ away; six potential probes in both directions sideways at least 100’ away from tower base, need 200’ right of way Can’t be used – long buried wires influence pointsource models for fall-of-potential; far enough away for accuracy includes adjacent towers too

Works fine

7-3

Fixed probe spacing of 75 and 40 m from tower may not be feasible in all cases

300’ Current and potential leads along right-of-way, best results with straight paths away from buried wires

Probe spacing too short for correct results

Minor mutual coupling errors can be resolved by extra steps

7-4

8 RECOMMENDATIONS Cross-calibration experience suggests that the use of construction footing resistance as a substitute for lightning impulse impedance is warranted on the AEP 138-kV lines. However, there was a wide discrepancy in the two data sets, both for 69-kV poles and for 765-kV steel lattice structures. Thus, the Zed-Meter approach should be refined and applied to a wider selection of towers. Uncertainty about the local soil resistivity on the day of the test should and can be resolved by developing better analysis of the input surge impedance of the remote current lead. Measurements in two directions should take advantage of the presence of the remote potential lead. Patent protection of this finding may have a business case. A measurement lead configuration that keeps both current and potential leads along the right-ofway is nearly essential for efficient testing, even if it calls for an extra set of measurement records to characterize the mutual coupling. The alternative of zig-zag layout for the current lead will not give good values of resistivity, so straight leads should be used in the near-term work. Calibration factors between the Zed-Meter transient impedance and values obtained with fall-ofpotential, EPRI Smart-Ground Meter and ABB 26-kHz meters should be established. The values with and without overhead groundwires should be compared to validate the treatment of this factor using a constant parallel surge impedance in the Zed-Meter approach. Calibration factors between the Zed-Meter resistivity measurement and results from two-layer interpretation of conventional Wenner surveys should be established. It would be particularly desirable to establish whether simple waveform features such as late-time creep-up of impedance correspond to soil properties (such as an increase in Wenner resistivity with probe spacing).

8-1

8-2

9 REFERENCES 1. [Deri – 1981] Deri, A., G. Tevan, A. Semleyn and A. Castanheira. 1981. “The Complex Ground Return Plane: A Simplified Model for Homogeneous and Multi-Layer Earth Resistivity”. IEEE Transactions on Power Apparatus and Systems. Vol. PAS-100. No. 8. August. 2. [CSA – 1992] Canadian Standards Associatoin. 1992. “Electric Fence Controllers”. Canadian National Standard CAN/SAP-=C22.2 No.103-M92. 3. [Hunt – 1981] D.A. Hunt and J.F. Pinto, “Measuring Ground Resistance with a High Frequency Instrument”, Consolidated Edison Report, 8 May 1981. 4. [Carberry 1 – 1983] R.E. Carberry, Northeast Utilities memo “High-Frequency EarthResistance Testing”, April14, 1983 5. [Carberry 2 – 1983] R.E. Carberry, Northeaset Utilities letter to Richard Leppla, “HF Earth-Resistance Measurements”, October 3, 1983 6. [Leppla 1976] R.R. Lepla, “High Frequency Resistance Measuring Instrument”, memo to A. Taddeo of Commonwealth Edison Co, October 15, 1976 7. [Dick 1983] E.P. Dick, “Tower Footing Resistance Measurements”, Ontario Hydro Research Division Report 85-315-H, January 10, 1986. 8. [Keller and Frischknecht 1966] G.V. Keller and F.C. Frischknecht, Electrical Methods in Geophysical Prospecting (Oxford: Pergamon Press), 1966. 9. [Reilly 1998] J.P. Reilly, Applied bioelectricity :From Electrical Stimulation to Electropathology (New York : Springer), 1998 10. [Huang 2003] H. Huang and I.J. Won, “Real-time resistivity sounding using a hand-held broadband electromagnetic sensor”, Geophysics, Vol. 68 No.4, pp 1224-1231, July/August 2003. 11. [Meliopoulos 1993] A.P.S. Meliopoulos, F. Xia, E.B.Joy, G.J. Cokkinides, “An Advanced Computer Model for Grounding System Analysis”, IEEE Transactions on Power Delivery, Vol. 8 No. 1, January 1993, pp 13-23

9-1

12. [Meliopoulos 1993a] A.P.S. Meliopoulos, G.J. Cokkinides, H. Abdallah, S. Duong, S.Patel, “A PC Based Ground Impedance Measurement Instrument”, IEEE Transactions on Power Delivery, Vol. 8 No.3, July 1993, pp 1095-1106 13. [Meliopoulos 1994] A.P.S. Meliopoulos, S. Patel, G.J. Cokkinides, “A New Method and Instrument for Touch and Step Voltage Measuremetns”, IEEE Transactions on Power Delivery, Vol. 9 No.4, October 1994, pp 1850-1860. 14. [Meliopoulos 2004] Smart Ground Multimeter SGM Models SGMD-2001 and SGMD-3001 Operating Manual, Software Version WINSGM-5.6, August 2004

9-2

9-3

About EPRI EPRI creates science and technology solutions for the global energy and energy services industry. U.S. electric utilities established the Electric Power Research Institute in 1973 as a nonprofit research consortium for the benefit of utility members, their customers, and society. Now known simply as EPRI, the company provides a wide range of innovative products and services to more than 1000 energyrelated organizations in 40 countries. EPRI’s multidisciplinary team of scientists and engineers draws on a worldwide network of technical and business expertise to help solve today’s toughest energy and environmental problems. EPRI. Electrify the World

Export Control Restrictions Access to and use of EPRI Intellectual Property is granted with the specific understanding and requirement that responsibility for ensuring full compliance with all applicable U.S. and foreign export laws and regulations is being undertaken by you and your company. This includes an obligation to ensure that any individual receiving access hereunder who is not a U.S. citizen or permanent U.S. resident is permitted access under applicable U.S. and foreign export laws and regulations. In the event you are uncertain whether you or your company may lawfully obtain access to this EPRI Intellectual Property, you acknowledge that it is your obligation to consult with your company’s legal counsel to determine whether this access is lawful. Although EPRI may make available on a case by case basis an informal assessment of the applicable U.S. export classification for specific EPRI Intellectual Property, you and your company acknowledge that this assessment is solely for informational purposes and not for reliance purposes. You and your company acknowledge that it is still the obligation of you and your company to make your own assessment of the applicable U.S. export classification and ensure compliance accordingly. You and your company understand and acknowledge your obligations to make a prompt report to EPRI and the appropriate authorities regarding any access to or use of EPRI Intellectual Property hereunder that may be in violation of applicable U.S. or foreign export laws or regulations.

© 2004 Electric Power Research Institute (EPRI), Inc. All rights reserved. Electric Power Research Institute and EPRI are registered service marks of the Electric Power Research Institute, Inc. EPRI. ELECTRIFY THE WORLD is a service mark of the Electric Power Research Institute, Inc. 00000000000001008734 Printed on recycled paper in the United States of America

EPRI • 3412 Hillview Avenue, Palo Alto, California 94304 • PO Box 10412, Palo Alto, California 94303 • USA 800.313.3774 • 650.855.2121 • [email protected] • www.epri.com

1008734 the EPRI Zed-Meter

Recommend Documents