Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
a r t i c l e i n f o a b s t r a c t
Article history: Numerical techniques enable realistic analysis of lightning currents effects on tall structures includ-
Received 29 April 2016 ing actual installation conditions. Characteristics such as the frequency dependence of soil parameters,
Received in revised form 24 January 2017 inductances and capacitances due to nearby metallic elements, and the actual geometry of the structure
Accepted 16 February 2017
and its grounding system are commonly neglected or simplified in lightning grounding system design.
Available online 6 March 2017
In this paper, the impact of neglecting these characteristics on a tall-mast lightning grounding system
design is discussed. The effects are determined using three risk assessment approaches based on energy
Keywords:
calculation, voltage thresholds, and current integration, which allows comparing the performance of the
Lightning
Lightning earthing system
grounding system under different conditions. Results obtained from numerical simulations and then
Frequency dependence of soil verified by measurements show that a metallic protecting fence not connected to the earthing mesh
Grounding produces lower human safety risks to lightning currents than when it is connected. This result is in oppo-
Human safety sition to common design practices and shows the need to assess each specific grounding design under
Lightning protection true conditions to guarantee human safety requirements.
© 2017 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.epsr.2017.02.013
0378-7796/© 2017 Elsevier B.V. All rights reserved.
120 J.J. Pantoja et al. / Electric Power Systems Research 153 (2017) 119–127
ZTh
+ ~0 Ω RB
VTh
-
1 kΩ
Fig. 3. Thevenin equivalent circuit used to calculate the current and energy deliv-
ered to a human being.
Using this model, the frequency response of the soil permittivity Fig. 4. Setup for a direct lightning strike simulation. (a) Lateral and (b) perspective
and conductivity can be calculated as [20] views.
ε∞ 0
εr (ω) = 1+ (1) Fig. 2 is based on alternating currents below 100 Hz, extrapolations
ε0 fε∞
to transient currents are used. In Ref. [22], for example, a current
impulse of 199 A and 340 ns of time to half value, which corresponds
4fε∞ to 13.5 J, is used as a reference value to assess grounding systems
(ω) = 0 1+ [S/m], (2)
0 against lightning.
To calculate the energy given by the lightning induced step and
where ε0 is the free-space permittivity, 0 is the DC or low fre- touch voltages, the methodology of the Thevenin equivalent circuit
quency conductivity, ε∞ = 8ε0 , and f is the frequency. The complex proposed in Ref. [21] and shown in Fig. 3 was used. In the calcula-
permittivity in a material is given by ε = ε0 (εr − jεr ), where εr = tions, the Thevenin voltage, VTh , corresponded to the step or touch
/ (ωε0 ). Therefore, a soil can be characterized using just 0 , which voltage, the Thevenin impedance, ZTh , was neglected, assuming that
can be calculated as the inverse of the low-frequency soil resistivity the feet contact resistance was zero, and the body resistance, Rb ,
LF . was assumed to be 1000 . Finally, the energy was calculated as
2.2. Assessment parameters V 2 (t)
W= P (t) dt = dt, (4)
Rb
2.2.1. Energy calculation
The first estimated parameter is the amount of energy delivered where V (t) is the step or touch voltage calculated as a function of
to a person by a lightning induced overvoltage. The limit of the time and P (t) is the dissipated power as a function of time.
amount of energy W in Joules, which a person could withstand
without fibrillation risk, can be calculated using both the Dalziel 2.2.2. Step voltage threshold
curve shown in Fig. 4 and Eq. (3): The second method used to assess the lightning protection
system was to compare the induced step voltage with a voltage
W = I 2 · R · t [J] , (3)
threshold calculated as [23]:
where I and t are the pair of current and time values along the
(165∼250) + g
Dalziel curve in Fig. 4 and R is the human body resistance estimated Ust = √ , (5)
T
in 1000 , as specified in the IEEE Std. 80 [21]. When the energy
that a person could withstand is calculated using Dalziel’s charac- where g is the soil resistivity and T is the lightning duration. Using
teristic curve, a constant energy value of 10 J is obtained. Although g = 70 m and T = 1 ms, a 10.12 kV threshold voltage is obtained.
122 J.J. Pantoja et al. / Electric Power Systems Research 153 (2017) 119–127
Fig. 5. Calculated step voltage induced at 2.83 m from a direct lightning strike on Fig. 6. Step voltage induced due to an impulsive current at 0.23 m from the current
a soil with = 70 m. Simulation results are obtained using a resistive soil with injection and with a step of 0.7 m. Measurement results are taken from Ref. [1].
constant parameters = 70 m and εr = 1 and using the Messier model for a soil Simulation results are obtained using a 2nd order fit of the soil parameters reported
with LF = 70 m. in Ref. [1] and using the Messier model for a soil with LF = 2700 m.
This expression is restricted to long duration lightning strikes, such Fig. 5 shows the comparison between step voltages obtained
as the first positive impulse. from theoretical and simulation results. A good agreement is
observed between both approaches, validating the numerical simu-
lation. In Fig. 5 is also shown the calculated step voltage considering
2.2.3. Severity parameter Si1.4
the frequency dependency of the soil parameters. To include the
The third method is to use a parameter for safety condition
soil Messier model in the electromagnetic simulator, a dielectric
for ventricular fibrillation when short duration impulsive currents
dispersion fit with an error of 3.7% with respect to the parameters
flow in the human body. The criterion applies for healthy adults
produced by the model was used. As previously reported [1,3,8],
and currents due to touch, transfer, and step voltages. The severity
the effect of this frequency dependency is to reduce the peak value
parameter Si1.4 is given by [24,25]
of the induced voltages, as it is confirmed in Fig. 5.
1.4
Si1.4 = |ib (t) | dt (6)
3.2. Experimental results comparison
where ib (t) is the current through the body. The limit of the severity In Ref. [1], impulsive currents are impressed from a hemispheric
parameter is 1.52 A1.4 ms for negligible probability of ventricular electrode to a distant auxiliary grid while the induced step volt-
fibrillation and 2.96 A1.4 ms for a probability of 5% [24,25]. age at the soil surface is measured. Using the simulation setup
presented in previous paragraph, the developed step voltage was
3. Validation of the simulation calculated. In this case, the soil parameters with the frequency
dependence between 100 Hz and 4 MHz are used, since they were
First, a validation of the numerical simulation was performed. calculated in Ref. [1] from impulsive current measurements. In
The validation consisted in the comparison of numerical results the simulation, an extrapolation of these data up to 100 MHz was
with theoretical and experimental data for simplified cases. The used. For comparison, another simulation using the soil parameters
step voltage induced due to a direct lightning strike to earth was obtained with the Messier model was performed.
considered for this purpose. Fig. 6 shows the induced voltage obtained from measurements
and simulations. This figure shows that the voltage simulated using
3.1. Theoretical comparison both sets of soil parameters, predicts well the waveform experi-
mentally obtained. Although the voltage obtained using the Messier
The step voltage given on the ground surface at a distance r0 model presents a higher peak value, this result describes the gen-
from the lightning strike, with a step length l, soil conductivity , eral behavior of the induced voltage only with the lowest frequency
and a lightning current I (t) as a function of time, can be calculated resistivity reported in Ref. [1]. This can be considered as an adequate
as [11] response due to the high dependence of the voltage waveform on
I (t)
1 1
the response of the soil parameters.
Up (t) = − . (7)
2 r0 r0 + l
4. Lightning earthing system assessment
For the numerical simulation, the lightning channel was repre-
sented using a 15 m long cylindrical perfect conductor connected 4.1. Earthing system design
to a current source buried 0.5 m, as shown in Fig. 4. A soil with a
low frequency resistivity of 70 m was used. The step voltage was The lightning protection design consisted of two concentric
measured by integrating the electric field calculated in the simu- squared earth loops around the tower, as shown in Fig. 7. To avoid
lations using the integration line shown in Fig. 4a. The step length, unauthorized personnel access, the mast base is protected with a
represented as the arrow number 2 in this figure, was 1.4 m and the 2 m high square metallic enclosure. This metallic enclosure is not
distance to the lightning channel was 2.83 m. bonded to the earthing mesh and it is separated 2 m from the mast
J.J. Pantoja et al. / Electric Power Systems Research 153 (2017) 119–127 123
Fig. 7. Simulation Setup. (a) Perspective, (b) lateral, and (c) top views.
center. Due to the metallic enclosure geometry included in the second case, the corners of the metallic enclosure are bounded to
design, high electric field intensities are presented at its corners [9]. the earthing mesh, while in the third case the metallic enclosure is
For this reason, the protection assessment was based on calculating not connected to the earthing mesh.
the step- and contact- induced voltages at the metallic enclosure
corners. These voltages were calculated by integrating the electric 4.2. Effect of the metallic enclosure
field in the paths shown in Fig. 7a; where the step length is 1 m and
the touch path is formed by two straight lines, one horizontal and Figs. 8 and 9 show the respective induced voltages due to first
one vertical with respective lengths of 1 m and 1.5 m. positive and first negative lightning impulses. These figures show
Three possible situations of the metallic enclosure were con- that the earthing condition of the metallic enclosure has an impor-
sidered. In the first one, the metallic enclosure is removed, in the tant effect on the waveform and amplitude of the induced voltages.
Table 1
Risk assessment parameters calculated from simulated voltages.
Excitation Metallic enclosure Delivered energy (J) Step voltage Severity parameter Si1.4 (A1.4 ms)
(kV)
Fig. 9. Magnitude of the step and touch voltages produced by the 100 kA, 1/50 s
lightning strike at the top of the tower considering the earthing system (a) without
Fig. 8. Magnitude of the step and touch voltages produced by a 100 kA, 10/350 s metallic enclosure, (b) with a metallic enclosure bonded to the earthing mesh, and
lightning strike at the tower top, considering the earthing system: (a) without metal- (c) with a metallic enclosure not bonded to the earthing mesh.
lic enclosure, (b) with a metallic enclosure bonded to the earthing mesh, and (c) with
a metallic enclosure not bonded to the earthing mesh.
J.J. Pantoja et al. / Electric Power Systems Research 153 (2017) 119–127 125
Fig. 10. Experimental setup used to measure the touch voltage produced by an
impulsive current injected into the mast base.
energy levels for each case were calculated applying Eq. (4) to the
voltage waveforms presented in Figs. 8 and 9. Table 1 shows that ter was obtained when the enclosure is not bonded to the earthing
removing the metallic enclosure from the simulations yields to very mesh since it does not surpass the step-voltage limit and slightly
high energy values delivered by the step path. For the other con- surpass the touch-voltage limit.
sidered cases, with the metallic enclosure bonded and not bonded It is important to note that the three calculated approaches indi-
to the earthing mesh, the delivered energy levels are below the 10 J cate that the lowest risk to persons in the vicinity of the mast is
limit for the touching integration line. However, for the step path, obtained for the case when the metallic enclosure is not bonded
the not bonded condition is the only one that is below the limit. to the earthing mesh. Although the 1/50 s waveform is shorter
For both impulses, the design with the enclosure not bonded to the than the maximum values of lightning parameters defined by the
earthing mesh delivers less energy than the earthed one. standard IEC 62305-1 for the first negative impulse (i.e. 100 kA,
Peak step voltages are also presented in Table 1, which shows 1/100 s), the obtained energy levels are considerable below the
that the design with the enclosure not bonded to the earthing mesh threshold of 10 J for the not bonded design.
has lower voltage levels than the step voltage threshold of 10 kV The calculated results indicate that there is an influence of the
determined by using Eq. (5). impulse waveform on the risk assessment parameters.
The severity parameter Si1.4 was calculated applying Eq. (6) to
the current produced in a body resistance of 1 k by the voltage 4.4. Experimental test
waveforms presented in Figs. 8 and 9. Table 1 shows that the limit of
2.96 A1.4 ms is not surpassed in any case for the 1/50 s waveform. An experimental test was performed to verify that the design
For the 10/350 s waveform, both simulations: with and without with the enclosure not bonded to the earthing mesh provides a
the metallic enclosure bonded to the earthing mesh surpass the lower risk when the mast carries impulsive currents. An impulse
2.96 A1.4 ms limit. The best performance according to this parame- current of −1.1 kA, 17/62.5 s was injected into the base of the mast
Fig. 11. Schematic of the setup used to generate and measure the impulsive current injected in the mast base.
126 J.J. Pantoja et al. / Electric Power Systems Research 153 (2017) 119–127
5. Conclusions
[6] R. Alipio, S. Visacro, Modeling the frequency dependence of electrical [16] H. Karami, F. Rachidi, M. Rubinstein, On practical implementation of
parameters of soil, IEEE Trans. Electromagn. Compat. 56 (5) (2014) electromagnetic models of lightning return-strokes, Atmosphere 7 (10)
1163–1171. (2016).
[7] R. Alipio, S. Visacro, Frequency dependence of soil parameters: effect on the [17] The Modelling of Lightning Strikes, CST AG, Computer Simulation Technology.
lightning response of grounding electrodes, IEEE Trans. Electromagn. Compat. Article ID: 342, 2012.
55 (1) (2013) 132–139. [18] D. Johns, Designing building structures for protection against EMP and
[8] F.H. Silveira, S. Visacro, R. Alipio, A. De Conti, Lightning-induced voltages over lightning, IEEE Electromagn. Compat. Mag. 5 (1) (2016) 50–58.
lossy ground: the effect of frequency dependence of electrical parameters of [19] F.M. Tesche, On the modeling and representation of a lossy earth for transient
soil, IEEE Trans. Electromagn. Compat. 56 (5) (2014) 1129–1136. electromagnetic field calculations, in: Theoretical Notes, 2002, no. Note 367.
[9] A. Sowa, J. Waiter, Ground potential rise, step and touch voltages during [20] M. Messier, Another Soil Conductivity Model, Internal Report, JAYCOR, Santa
lightning strokes to GSM base station, in: International Conference on Barbara, CA, 1985.
Lightning Protection, Avignon, France, 2004. [21] IEEE Std 80-2000: IEEE Guide for Safety in AC Substation Grounding, 2000.
[10] B. Markovski, L. Grcev, V. Arnautovski-Toseva, Step and touch voltages near [22] D.S. Gazzana, A.S. Bretas, G.A.D. Dias, M. Telló, D.W.P. Thomas, C.
wind turbine grounding during lightning strokes, International Symposium Christopoulos, A study of human safety against lightning considering the
on Electromagnetic Compatibility (EMC EUROPE), 2012 (2012) 1–6. grounding system and the evaluation of the associated parameters, Electr.
[11] J.J. Pantoja, F. Roman, Analysis of a lightning earthing system using Power Syst. Res. 113 (2014) 88–94.
electromagnetic simulations, International Symposium on Lightning [23] Y. Zen, Earthing Technique, Water Conservancy and Electric Power Press,
Protection (XIII SIPDA), 2015 (2015) 104–108. Beijng, 1979.
[12] A. Sowa, J. Wiater, Reduction of the step voltages around building during [24] E. Amiri, S.H.H. Sadeghi, R. Moini, A probabilistic approach for human safety
direct lightning strike, presented at the IX International Symposium on evaluation of grounding grids in the transient regime, IEEE Trans. Power
Lightning Protection, Foz do Iguaçu, Brazil, 2007. Deliv. 27 (2) (2012) 945–952.
[13] C.F. Dalziel, Dangerous electric currents, Trans. Am. Inst. Electr. Eng. 65 (8) [25] C. Portela, Frequency and transient behavior of grounding systems II –
(1946) 579–585. practical application examples, presented at the IEEE 1997 International
[14] L.B. Gordon, The physiological effects of electric shock in the pulsed power Symposium on Electromagnetic Compatibility, Austin, Texas, United States,
laboratory, Pulsed Power Conference, 1991. Digest of Technical Papers. Eighth August 1997, 1997.
IEEE International (1991) 377–380.
[15] D. Li, et al., On lightning electromagnetic field propagation along an irregular
terrain, IEEE Trans. Electromagn. Compat. 58 (1) (2016) 161–171.