30th International Conference on Lightning Protection - ICLP 2010

(Cagliari, Italy - September 13th -17th, 2010)

ICLP 2010
AN ANALYSIS OF CONVENTIONAL GROUNDING
IMPEDANCE BASED ON THE IMPULSIVE CURRENT
DISTRIBUTION OF A COUNTERPOISE
Jong-Hyuk Choi1, Jong-Ho Kim1, Bok-Hee Lee1, and Young-Ki Chung2
Inha University1, Uijae Electrical Research Institute,2 Korea
Chjohy80@naver.com

of the length of counterpoise and the current distribution

ABSTRACT
rate according to the front times of the impulse current in
The performance of a grounding system should be evaluated long horizontally buried conductors (i.e. counterpoises)
in terms of the grounding impedance for a lightning surge required to achieve a lower grounding impedance in a
containing high frequency components. Grounding high soil resistivity region or transmission tower. Also
impedance shows resistive, capacitive and inductive the simulated results of the dispersed impulse current at
behaviors. Ground resistance is regarded as a particular
grounding impedance measured in a low frequency range.
each part of the counterpoise and the current distribution
When designing the grounding system, the grounding obtained using the distributed parameter circuit model
impedance and effective length should be considered. In this and an EMTP program are presented.
study, conventional grounding impedance is measured as
functions of the front time of the impulse current and the 2 EXPERIMENTAL METHOD AND
length of the counterpoise used largely as the grounding CONFIGURATION
electrode of the transmission tower. In order to find the
relation between the conventional grounding impedance and
In order to measure the impulsive current distribution
current distribution, the magnitude of the dispersed impulse and conventional grounding impedance, a counterpoise of
current at every 10 m interval along of the counterpoise, 50 m long with a radius of 2.82 mm is installed at a depth
which is 50 m long, is measured and simulated according to of 0.5 m. Before installing the grounding electrode, the
the front time of the injected impulse current. As a result, soil resistivity of the experimental site near the
the conventional grounding impedance of a long counterpoise is surveyed by the dipole-dipole method.
counterpoise is similar to that of a short counterpoise with a Figure 1 shows the soil structure of the test site with two
smaller front time. As the front time of the injected current layers. The lower layer shows a lower resistivity than that
becomes short, the current distribution is increased near the of upper layer. In the case of the soil structure nearby the
current injection point. In addition, the simulated results of
multi-layer soil structure are very similar to the measured
counterpoise, the sections of 10 ~ 20 m and 30 ~ 40 m
results. show a higher resistivity than do the other sections of
counterpoise. The ground resistance of the counterpoise
50 m long is 8.55 and the apparent average soil
1 INTRODUCTION resistivity reversely calculated by equation (1) is 205.34
m [7].
Lightning protection systems require an efficient
grounding system in order to minimize the direct and
induced lightning overvoltages in nearby services and R l
= (1)
structures[1]. According to the standards in Korea and 2l
ln 1
some other countries, the performance of the grounding 2rt
system is evaluated by means of a ground resistance
measured at a low frequency which should be below a
defined security level[2][3]. However, the high frequency 0-10m 10-20m 20-30m 30-40m 40-50m

behavior of grounding systems can be very different from

their behavior at low frequencies[4]. Grounding
impedance presents resistive, capacitive and inductive
characteristics. The ground resistance is restricted by a
particular value of grounding impedance measured in low
Figure 1. Soil structure of the test site nearby the
frequency range [5]. This paper presents the experimental counterpoise of 50 m long
results of conventional grounding impedance as functions

1179-1
 into the soil and the current flowing along the electrode.
Figure 2 shows the configuration of the experimental In order to analyze the relation between the impulsive
setup used to measure the conventional grounding current distribution and the conventional grounding
impedance and current distribution associated with the impedance, current waveforms were observed at the
counterpoise. In order to minimize measurement errors current injection point and at 10 m intervals along the
due to AC mutual coupling, the revised fall-of-potential counterpoise as shown in Figure 2. The current
method, in which the potential probe is extended at an waveforms dispersing into the soil at these of 10 m
angle of 90 with respect to the current injection line as intervals are calculated from the measured current
illustrated in IEEE Std. 81.2-1991, was adapted[6]. In waveforms using the Matlab program. Figure 3 shows
order to reduce the error caused by the earth mutual examples of current waveforms measured by CT installed
resistance, the current and potential auxiliary electrodes at the 10 m intervals, for impulse current with fast and
are set far from counterpoise at the distance of 150 m. slow front times. When an impulse current with a fast
The impulse current is generated by a combination type front time of 4 is injected into the grounding electrode,
impulse generator (1.2/50 in open circuit and 8/20 current waveforms and current peaks appear at times
on closed circuit) and the front time of the impulse which are in the same order as the distance from the
current is controlled by the capacitance of the additional injected point to CT. However the impulse currents
capacitor. The potential difference between the converge to zero at the same time of about 20 . On the
counterpoise and the potential auxiliary electrode at other hand, for an injected current pulse with a relatively
remote earth is measured by the use of a high voltage slow front time of 39 , current waveforms measured at
differential probe with the bandwidth of 70 MHz and the each interval start with injected current waveform and
injected current is measured by with the aid of a current current peaks appear almost simultaneously with each
transformer (Pearson 3525) with a bandwidth of 15 MHz. other and the injected current peak as shown in Figure
Current transformers with same performance are installed 3(b). This phenomenon is caused by the fact that the high
at 10 m intervals along the counterpoise to observe the frequency components of the impulse current associated
current dispersed into each section of the soil. The with the fast front time are dispersed more rapidly near
detected signal is transmitted to a DSO with a 2.5 GS/s the injection point than are the lower frequency
sampling rate and a bandwidth of 500 MHz. The components.
conventional grounding impedance associated with the
front time of the injected current is derived through
calculation of the ratio between the peaks of the current Injection point
At 10m
and the potential of the measured waveforms defined At 20m
according to IEC 62305-1[8]. Dispersed current At 30m
waveforms induced in each 10 m section of the At 40m
counterpoise are considered.

Impulse current generator

Digital storage
oscilloscope

Current auxiliary
electrode(C) 150 m
90 (a) Front time = 4

CT Injection point
0m
15 At 10m
At 20m
50m Counterpoise (E)
At 30m
At 40m
Potential auxiliary
electrode(P)

Figure 2. Experimental setup

3 RESULTS AND DISCUSSION

(b) Front time = 39
3.1 Impulsive current distribution as a function of
the front time of the injected current pulse Figure 3. Examples of impulse current waveforms measured
Ground current is composed of the current dispersing at the injected point and at every 10 m point of counterpoise

1179-2
 The currents dispersed into the soil at 10 m intervals Figure 5 illustrates the average current distribution
along of the counterpoise are calculated from the according to the front time of the injected impulse current
sectional current waveform shown in Figure 3 in order to waveforms. The average current distribution is computed
discover the difference between the fast and slow front from the dispersed current peak in each section of
time current pulses in propagating along the grounding counterpoise as shown in Figure 3 and 4. In case of the
electrode. If there is no current loss in the grounding current with a slow front time, the maximum difference
electrode, the whole current measured at the 40 m point is between the current distributions is about 5 % and almost
dispersed in the 40 to 50 m section of the counterpoise. same with the results obtained for the low frequency
The amount of current dispersed in the 30 to 40 m section current. In the low frequency region or for the slow front
is calculated by subtracting the current waveforms time current, the difference in the current distributions
measured at 40 m from those measured at 30 m. arises from the fact that the soil resistivity is not
Dispersed current waveforms in the 5 sections of the completely spatially uniform. In the results for the slow
counterpoise computed by this method are shown in front time current, the current distributions in of 10 ~ 20
Figure 4. m and 30 ~ 40 m sections are lower than those in the
For the fast front current, a large amount of current is other sections, since the soil resistivity around these
dispersed in the 0 to 10 m section during the front time of sections is relatively higher than that in other regions, as
the current pulse as shown in Figure 4(a). Considering shown in Figure 1. However, as the front time of the
the time difference between current peaks, the current current is reduced, the current distribution for each
peak in the first section is twice as large as those in times section decreases in the order of the distance from the
the other sections. In the section farthest from the current injection point. On the contrary, the current
injection point, most of the current peaks are small. On distribution in the section from 0 to 10 m increases
the other hand, Figure 4(b) shows the front time and sharply for front times shorter than 8.3 . In other words,
magnitude of the current peak dispersing into the soil in fast front time impulse currents containing high
each section of the counterpoise for the current pulse with frequency components are dispersed mostly near the
a slow front time. Because of spatial variations in soil injection point and the part of the grounding electrode
resistivity the current peaks computed for each section of with a small current distribution rate is not a cost
also show some differences. For the fast front impulse effective way to reduce the grounding impedance.
current, most of injected surge current is dispersed near
the injection point as shown in Figure 4(a) 40
0-10m
35 10-20m
20-30m
0~10m
Distribution rate (%)

30 30-40m
10~20m 40-50m
20~30m 25
30~40m 20
40~50m
15

10

0
3.5 4 8.3 10.3 15 20.5 39 63
Front time of impulse current (s)
(a) Front time of 4 Figure 5. Current distribution as a function of front time of
injected impulse current at each section of every 10 m of the 50
0~10m
10~20m
m counterpoise
20~30m
30~40m 3.2 Simulated results of impulsive current
40~50m distribution in multi-layered soil
In order to generalize and simulate the impulsive
current distribution draining through the soil from the
counterpoise, the counterpoise is modeled by a
distributed parameter circuit model. The distributed
parameter circuit model is realized by EMTP and the
results are graphically shown by Matlab program. The
(b) Front time of 39
impulsive current distribution of the counterpoise
simulated in multi-layer soil structures is considered. The
Figure 4. Current waveforms dispersed into the soil of each
section of every 10 m of counterpoise soil layers are divided by into 10 m sections along

1179-3
counterpoise and the soil resistivity of the five sections is the experimental work (as shown in Figure 3) are injected
inversely proportional to the current distribution into the distribution parameter model, the resultant
measured at a frequency of 1 kHz. The results for the AC simulated current waveforms for each section 10 m of the
current distribution at 1 kHz and the calculated grounding counterpoise are shown in Figure 6. In the case of current
resistance and soil resistivity of each part for the waveforms with front times of 4 there are substantial
counterpoise of 50 m long are presented in Table 1. The differences between current waveforms close to the
current distributions measured at 1 kHz show little current injection point. This result arises from the
difference and correspond with the soil structure shown grounding current largely dispersing near the current
in Figure 1. injection point combined with the time lag of the surge
propagation to the end of grounding electrode. On the
Table 1. AC current distribution, grounding resistance other hand, this difference is not shown with the injected
and soil resistivity in each 10 m sections of a
impulse current with a fast front time of 39 since the
counterpoise 50 m long
propagation time is much shorter than the duration of the
0-10m 10-20m 20-30m 30-40m 40-50m impulse current and the dispersed current is not
Current dependent on the current injection point.
distribution at 22.83 17.32 19.84 17.80 22.20
1 kHz (%)
Grounding
Injection point
resistance 27.60 36.38 31.76 35.40 28.38 At 10m
[] At 20m
Soil At 30m
175.83 231.77 202.33 225.52 180.82 At 40m
resistivity

The impulsive current distributions were simulated by

the distributed parameter circuit model simulated using
EMTP. The injected impulse current is modeled by two
Heidler surge functions in the range of the front time
from 3.5 to 70 and the simulated results agree with (a) Front time = 4
the experimental results shown in Figure 3. The five
current probes, as shown in the experimental Injection point
At 10m
configuration, are set at 10 m intervals along the
At 20m
counterpoise. The counterpoise is divided into one At 30m
hundred segments and each parameter is substituted to At 40m
the circuit. In order to simulate multi-layered soil
structures, the parameters of at 10 m intervals are
differently set. The values of G, L, C are calculated from
the equations (2), (3) and (4) respectively expressed in
the references [9, 10] and the relative dielectric constant
is assumed to be 40.
(b) Front time = 39
1
G= (2) Figure 6. Simulated current waveforms dispersed into the
ln l 1
2 soil of each section of every 10 m of counterpoise
2rt

In order to make a close investigation into the

0 r difference between the fast and slow current pulses,
C= (3)
2l dispersing current waveforms in every section of the
ln 1
2rt counterpoise are calculated as in Section 3.2. Figure 7
shows a simulated dispersing current pulse at 10 m
intervals along the counterpoise. In the case of a front
0 2l time of 4 , the dispersing currents in the 10 ~ 20 m
L= (ln 1) (4)
2rt sections is much larger than those in any other sections.
The dispersing current is larger close to the current
When the same impulse current waveforms as used in injection point and at distances up to 30 m from that point.
The dispersing current in the 30 ~ 50 m sections

1179-4
resembles the results measured at a low frequency since decrease gradually. On the other hand, the simulated
the tail part of the injected impulse current has more low results in the 0 ~ 10 m section sharply increased for front
frequency components. On the other hands, the dispersed times less than 8.3 and those in the 10 ~ 20 m section
impulse current peak for the injected impulse current of increase slightly. As the front time of the impulse current
slow front time is inversely proportional to the soil is shortened, the high frequency current component is
resistivity in every section of the counterpoise. For the increased. The high frequency current is largely dispersed
slow front time impulse current containing low frequency around the current injection point and cannot propagate
components, the impulse current begins to disperse near to the end of the counterpoise even though the soil
as the current injection point and the dispersed current resistivity of the section far from the current injection
peak is not much different from that of the impulse point is relatively low. The simulated impulsive current
current. The small difference in dispersed current peaks distributions in the multi-layer soil structure as a function
for each section of the counterpoise comes from the soil of the front time of the impulse current are very similar to
resistivity nearby the counterpoise. the measured results of Figure 5 and these data more
generalize the current distribution in the grounding
system not including the measurement error. Considering
0~10m the measured and simulated results, when the fast surge
10~20m
20~30m
current reaches the grounding electrode, further extension
30~40m of the length of the counterpoise will not improve the
40~50m performance of grounding systems.

35 0-10m
10-20m
30 20-30m
30-40m
Distribution rate (%)

25 40-50m

20
(a) Front time = 4
15
0~10m
10~20m 10
20~30m
5
30~40m
40~50m 0
3.5 4 8.3 10.3 15 20.5 39 63
Front time of impulse current (s)

Figure 8. Simulated current distributions as a function of

front time of impulse current at each section of every 10 m
of counterpoise

(b) Front time = 39 3.3 Conventional grounding impedance as a

function of the front time of impulse current
Figure 7. Simulated current waveforms dispersed into the
soil in 10 m sections of the counterpoise
In order to examine the relationship between the
impulsive current distribution and grounding impedance,
Figure 8 presents simulated current distributions in the the conventional grounding impedance is evaluated for a
5 10 m sections of the counterpoise as a function of the counterpoise of 50 m long as shown in Figure 2.
front time of the impulse current. The current distribution Disconnecting every 10 m of grounding electrode in
dispersing into the soil of in each 10 m section of the opposite side of current injection point, the conventional
counterpoise is calculated from the ratio between each grounding impedance is measured as a function of the
current peak as shown in Figure 7. In the case of slow length of counterpoise.
front times, the current distributions are inversely Figure 9 shows five curves for the conventional
proportional to the soil resistivity and are the same as the grounding impedance of the counterpoise as a function of
measured results at low frequencies. The current front time of the injected current. With a front time of
distribution in the 40 ~ 50 m section having a relatively longer than 50 , the conventional grounding impedance
large distribution decreases sharply for front times shorter shows almost the same values as the ground resistance
than 8.3 . The impulse current distributions in the measured at low frequencies, and the longer the length of
section of 20 m from the current injection point also counterpoise, the lower is the grounding impedance.
However, the conventional grounding impedance

1179-5
decreases slightly at the beginning of the impulse current grounding impedance of the long counterpoise
and begin to increase in the order of increasing length of approaches that of the short counterpoise.
counterpoise at short front time. At front times shorter (3) The impulsive current distributions and current
than 8 , the grounding impedance is the same for 50 m waveforms simulated by EMTP and distribution
and 40 m length counterpoises. The grounding parameter model in multi-layered soil structure are in
impedances for a 30 m long counterpoise is almost the good agreement with the experimental results.
same for front times shorter than 5 and at lengths (4) When designing grounding systems for lightning
greater than 20 m long the impedance is unchanged for 2 protection, the effective length of the grounding
front times. Judging the results at lengths exceeding electrodes should be determined by considering the front
10 m, the grounding impedance at 10 m long are similar time of impulse currents and the soil resistivity.
to those of the other counterpoises for front times shorter
than 1 . As a result, it is meaningless to lengthen the 5 REFERENCES
grounding electrode to decrease the grounding impedance
against the ground current, with fast front time pulses i.e [1] Pedro Llovera, Juan Antonio LLiso, Alfredo Quijano,
Vicente Fuster, "High frequency measurements of
pulses composed of high frequency components. The
grounding impedance on resistive soils", Proc. 28th ICLP,
concept of the effective length of grounding electrode pp.727-729, 2006.
should be considered. In this paper, the effective length [2] Korea Electric Association, Korea Electrical Installation
of the grounding electrode appears when the reduction of Standard, 2006.
the current distribution at the end of grounding electrode [3] Protection of structures against lightning, BS 6651, 1999.
is begins. [4] A. Rousseau, P. Gruet, "Practical high frequency
measurement of a lightning earthing system." Proc. 27th
30 ICLP, pp. 526-530, 2004.
50 m 40 m [5] Silverio Visacro, "A Comprehensive Approach to the
Grounding Response to Lightning Currents", IEEE Trans.
Grounding impedance ()

25 30 m 20 m
10 m
on P.D., Vol. 22 No. 1, 2007.
20 [6] IEEE Std. 81.2-1991, IEEE Guide for Measurement of
Impedance and Safety Characteristics of Large, Extended
15 or Interconnected Grounding Systems, 1991.
[7] Hideki Motoyama, Experimental and Analytical Studies
10
on Lightning Surge Characteristics of a Buried Bare Wire,
Electrical Engineering in Japan, Vol.164, No.3, pp.35-41,
5
2008.
0
[8] IEC 62305-1, Protection against lightning Part 1:
0 10 20 30 40 50 60 70 80
General principles, 2006.
Front time of impulse current (s) [9] M. E. Almeida and M.T. Correia de Barros, Fundamental
considerations on long ground electrodes design, Proc.
13th ISH, pp.241-244, 2003.
Figure 9. Conventional grounding impedance of
[10] C. Mazzetti and D. Mukhedkr, Impulse to impedance of
counterpoise as a function of front time of impulse current
buried ground wires, IEEE Trans. PAS, Vol. PAS-102,
No.9, pp.3148-3156, 1983.

4 CONCLUSIONS
This paper presents the measured and simulated
impulsive current distributions and conventional
grounding impedance of a counterpoise as a function of
the front time of the injected current and the length of
grounding electrode. The results can be summarized as
follows:

(1) The current distribution draining through the soil is

dependant on the soil resistivity for slow front time
current pulses and increases near the injection point for
fast front time pulses.
(2) The conventional grounding impedance for slow
front time currents is similar to the ground resistance. As
the front time of the impulse current is shortened, the

1179-6