In this chapter, a review of earth impedance calculation techniques and soil measurements methods are given. Analytical methods of calculating earth impedance of some earth electrode configurations are described. The performance of earth electrode under transient conditions is then presented. A review of testing and modelling the wind turbine earthing system under both power frequency and lightning conditions are carried out.
The early work on earthing system was concentrated on the performance of a simple earth electrodes such as vertical electrodes and horizontal electrodes, using field tests, laboratory experiments and simple circuit models. However, the results of the previous studies applied on the earthing system and improvements have been achieved an improvement in the design of the earthing system subjected to power frequency faults and fast transient currents. The most recent work for analysing the performance of earthing system subjected to impulse currents by using computer simulations techniques, which is enables to investigate complex earthing system configurations. In this section a brief review of the work related to the behaviour of the earth electrode subjected to a surge conditions.
2.2 Calculation of earth impedance
A considerable amount of work has been carried out to formulate accurate expressions of earth impedance [2.1 - 2.11] for a wide range of earth electrodes. Earthing systems consist of various components such as earth electrodes, metallic cable sheaths and tower line earthing systems associated with tower footings.
Earth electrode resistance
The longitudinal impedance of earth electrodes is significant only if the electrode is very long (more than several hundreds meters). For typical concentrated earth electrodes, the inductive component is negligible, and therefore, such electrodes can be considered to be predominantly resistive at low frequency.
Dwight [2.1] proposed different formulae to calculate the earth resistance of several configurations of rods, a short horizontal wire, a buried horizontal plate, a horizontal strip and a ring of wire. The expressions proposed are based on the analogous relationship between capacitance and resistance
(2.1)
where R is electrode resistance and r is the soil resistivity
C is the capacitance between the electrode and its image above the surface of the earth and given by
(2.2)
where
l= electrode length and
a= electrode radius
Applying this method leads to a general equation applicable to any form of electrode [2.4]. However, most of the configurations proposed by Dwight are not encountered in practice, such as the three-point star, burial horizontal round plate, and four-point star.
In 1954, Schwarz [2.2] carried out an analytical investigation into the calculation of earth resistance for various electrode configurations. An average potential method was used with modifications to allow for the influence of one part of the electrode on another which is known as the proximity effect. This modification included the consideration of an additional quantity, which he described as the ''density'' of number and length of conductors over the area covered. In other words, to make it a dimensionless quantity, the density is taken as the ratio of conductor quantity per linear extension of the area. He compared the results obtained by the expression with the results obtained by measurements made on a scale model, and found a good agreement. The expressions proposed by Schwarz are more likely to be found in practice rather than those developed by Dwight [2.1].
Laurent [2.3] derived expressions for the calculation of earth electrode resistance for different electrode configurations, some of these expressions are based on expressions derived by Dwight. The expressions he proposed are based on the assumption that the electrode can be divided into small segments, and the current flowing to earth is distributed between these segments. The potential at any point can be calculated as the sum of the potentials resulting from each segment.
Sunde [2.5] provided a formula for the calculation of the earth resistance of earth electrodes based on the voltage rise at the electrode midpoint. In order to find the potential on the surface of the conductor, an assumption of neglecting the longitudinal voltage drop along the conductor is made. With this assumption, the change in potential along the conductor surface can be considered to be zero.
Tagg [2.4] performed most of his work on the measurement of the earth electrode resistance. However, he suggested a formula to determine the resistance to earth of an earth electrode based on expressions developed by Dwight [2.1].
Electricity industry earthing standards such as ER/S34 [2.6] provide formulae to determine the resistance of practical earth electrodes based on expressions proposed by Schwarz [2.2], Sunde [2.5] and Tagg [2.4].
The following table (Table 1.1) shows the various resistances expressions of vertical earth electrode obtained for a uniform soil and given by Sunde [2.5], Laurent [2.3], Tagg [2.4] and ER/S34 [2.6].
Table 1.1 different expressions for calculating the earth resistance of vertical earth electrodes
Where
r the soil resistivity
l the electrode length
a radius of the electrode
d diameter of the elect rode
Earth electrode under impulse condition
Many authors [4-25] have investigated the behaviour of earth electrode subjected to high impulse currents. However, it was assumed that the earthing system subjected to high impulse current will behave differently from expected behaviour at 50Hz faults.
As early as 1928 Towne [4] carried out tests on galvanised - iron pipes of different lengths up to 6m and 10.65mm radius, buried in loose gravel soil. Impulse current used up to 880 A with rise time between 20 ms to 30 ms. The resistance of 6m pipe was 24W, at 60 -cycle, when the impulse current injected the resistance drops to 17W, which is 71% from the 60 - cycle measured value. The same kind of tests carried out on different lengths, obtained the same results, the resistance were less than the 60 - cycle values. Towne was the first who noticed that the impedance of the earth electrode presented to an impulse current could be lower than that seen by power frequency. This was attributed to the arc sparks in which, expand the contact area between the electrode and the soil. Since then the researches began to investigate this subject in order to explain the behaviour of the earth electrode subjected to high magnitude current.
Then Bewley [5] carried out impulse tests on counterpoises of different lengths. Impulse current of 6ms to 12ms rise time with currents between 2kA to 8kA were injected. It was found that the transient impedance of the counterpoises less than the 60Hz resistance. Moreover, the transient impedance defined as the ratio of voltage to current as a function of time.
Bellaschi [6] conducted a set of tests on vertical earth electrode with current between 2kA and 8kA with rise time between 6ms.and 12ms. The impulse resistance was measured as a ratio between the peak voltage and the peak current. However, the same findings were reported, that the resistance decreased with the current magnitude to a value lower than power frequency resistance. This is attributed to the soil ionisation effect, due to the sharp decrease in the voltage value after the peak value immediately.
In subsequent paper Bellaschi [7] conducted a series of tests using a current between 400A and 15.5kA with 20/50 ms, 8/125ms and 25/65 ms rise time. He found that the degree of reduction of impulse resistance, which is can be determined by the ratio of the 60Hz resistance to the impulse resistance. This reduction depends on type of soil and earth electrode arrangement, but independent of impulse rise time. It was also observed that the impulse resistance of electrode buried in low soil resistivity has the maximum reduction degree.
The findings of Towne [4] Bewley [5] Bellaschi [6-7] highlighted that the non linearity of the transient resistance of an electrode and the transient resistance of an earth electrode resulted form high impulse current could be less than the power frequency resistance of the same earth electrode.
Berger [8] carried out experiments on spherical electrode, 1.25 cm radius, half buried in 2.5 m diameter hemispherical pit filled with different soils. Impulse current applied ranged from 3.8 kA to 11.4 kA with rise time between 3ms and 30 ms, initially water was used surrounding the electrode, and the results obtained showed that a constant resistance for water, equal to the power frequency value. However, he obtained different values for different soils resistivities values 300Wm and 57Wm. The results obtained for different soils used showed that when the current reached a certain value the v - i characteristic curves show liner correspondence to the power frequency resistance, but when the current exceeded that value the characteristic curves show that the resistance below that obtained at 60Hz. Also he carried out tests on 110 m earth wire, 6 mm diameter buried at 20 to 30 cm depth. From the results obtained, the resistance decreases to a value below the power frequency resistance. The increase of the resistance initially was explained by the inductance of the earth electrode and the decrease due to the soil breakdown.
Petropoulos [] conducted set of test using a shaft electrode and electrodes with spikes. He found that the spiked electrodes give lower impedance when subjected to an impulse current. This reduction in impedance value was attributed to soil break down effects. The results show that as the length of the spikes increased the impulse resistance decreases.
Liew and Darveniza [9] carried out series of tests on vertical electrode and hemispherical electrode buried in different types of soils with resistivity ranging from 50 Wm to 310 Wm using impulse currents between 1 to 20 kA were applied with rise time 6 to 16 and 18 to 54ms. The results showed that the, non- - linear V/I characteristic and the minimum resistance occurs after the peak current occurs, the peak voltage occurs before the peak current. They propose a dynamic model to explain the behaviour of earth electrode subjected to a high impulse current. Figure 2. shows the behaviour of the earth electrode is divided into three stages, stage (a) represent a constant resistivity of homogeneous and isotropic soil, and the stage (c) where the current exceeds the critical current density, soil breakdown occurs in an exponential manner with decrease in the soil resistivity by a value known as ionisation time constant. In the third stage as the current decreases the soil recovers towards the steady state value in an exponential manner by a value known as de - ionisation time constant. They applied the model on 100 kA current which is expected in practice, assuming 10000 ohm-m soil resistivity and electric field intensity 3 kV/c, the results showed that the resistance of above 15 kA were equal to 15 kA resistance and in case of 100 kA a significant reduction in the resistance value occurred. The amount of reduction will depend on the soil resistivity value, would be greater in case of high resistivity and lower breakdown gradient.
Figure 2 dynamic model for soil ionisation process Reproduced from reference [9]
Gupta et al [10, 11] investigated the effect of impulse currents on square earth grids and rectangular grids by conducting experimental work. The impulse impedance found to be higher than the power frequency resistance. However, the impulse impedance increases as the soil resistivity increases. They defined the impulse impedance as the ratio of the voltage to the current at the injection point. Moreover, the impulse coefficient defined as the ratio of the impulse impedance to the dc resistance. It was found that the grids give higher impedance when injected at the corner rather than the centre. Their definition of the impulse impedance, it seems to be not correct due to the peak current and peak voltage occurs at different times. They conducted a laboratory work by using scale model of the square and rectangular grids in different soil resistivities. The same type of results was obtained the impulse impedance always higher than the dc resistance in all soil resistivities cases. They conclude their results that the soil ionisation effect for earth grids very small and can be ignored. However, the impulse impedance decreases with the area of the grid increases until a certain area which no decrease beyond this area and referred to as the effective length. The same finding was reported by Ramamoorty et al [12]
Velazquez and Mukhedkar [13] developed a dynamic model taking into account the soil ionisation process. However, they varied the radius of the ionised zone assuming zero resistivity. Moreover, the electrode divided into a number of segments each having a different current density and as a result different radii of ionised zones. They reported the capacitance becomes dominant when soil resistivity more than 1000Wm. furthermore, the study extended to include different electrode lengths between 30m to 150m in soil resistivities ranging form 1000Wm to 5000Wm. They reported that the transient behaviour of earth electrode depends on the length of electrode, soil resistivity, permittivity, and the shape of the impulse wave. According to their results, the impulse impedance of the earth electrodes increases to a maximum value equal to the surge impedance then decreases, eventually reaching the dc resistance of the earth electrode.
Kosztaluk et. al [14] carried out tests on 4 electrodes encased in concrete representing a tower footings and buried in soil two months before the tests were conducted, due to the concrete thoroughly saturated with the moisture containing in the soil surrounded the earth electrode. However, a peak current up to 26kA with rise time of 3ms/35ms were applied. The results show at initial rise current (at low currents), Ri (impulse resistance) is equal to 60Hz or dc resistance. Moreover, as the current increased gradually until it reached 2kA a decrease in the transient resistance observed. This reduction were attributed to two conditions (electrolytic and channel mechanism).
Geri et al [15], conducted high-voltage tests on a 1m steel earth rod and a 5m steel horizontal wire installed in the field. Impulse currents up to 30kA magnitude and 2.5ms rise-times were applied. The 'equivalent impulse resistance' was defined as the ratio of the peak current and peak voltage. The resistance decreases as the peak current increases, for example from 18W to 6W for the rod and from 10W to 4W for the wire. It was observed that the voltage peak preceded the current peak for the vertical rod but followed it for the horizontal electrode. The longer horizontal wire would have had a greater inductance compared with the vertical rod and therefore the voltage peak, in contrast to the results, would be expected to lead the current by a greater amount.
Almeida and Darveniz [18] carried out tests on earth electrode buried in sand and gravel mix soil, the electrode length is 0.61 and radius 0.075 m. current of 3.5 kA double exponential with 5 to 16 ms rise time is applied. Typically results reproduced in figure 3. As shown in the curve the resistance is constant until the critical current has been exceeded, soil breakdown is modelled decrease in soil resistivity, therefore the electrode resistance decrease by a time determined as `ionisation time constant'. As the current decreases then the soil recovers towards the initial resistivity value, as a result the electrode resistance recovers towards the normal resistance as can be seen in the curve, this is determined by the de - ionisation time constant.
Figure 3 dynamic model for soil ionisation process Reproduced from reference [16]
Sekioka et al [19] carried out field tests on three types of earth electrode (concrete pole, buried conductor and grounding net), used current up to 40kA and front waves with few microseconds rise time. A non-linear behaviour observed for the concrete pole and the buried conductor. However, the resistance decreases as the current increases for both. Moreover, the resistance of the grounding net was crest current independent, this is due to the large surface area of the net which is exhibited a small steady state resistance as shown in figure No 4.
Grounding net 8.1m buried conductor 17m buried conductor
Figure 4 Earthing resistance of the different earth electrode vs. crest current reproduced form reference [18]
The same finding to Ramamoorty et al [12] were reported by Stojkovic et al [20] no ionisation process occurring in soil when large area earthing system subjected to high impulse currents.
Cotton [] used hemispherical model in order to describe the process of soil ionisation. Two hemispherical electrodes 2.4cm and 5cm in diameter were placed in a75cm diameter inverted hemispherical constructed out of concrete. An impulse generator was used to inject current up to 3.25kA with rise times 4ms and 10ms. The results showed that a decrease in the impulse resistance of the earthing system as the applied voltage increases except in case the lowest applied voltage which did not appear to cause any ionisation. He applied the same voltages on the larger hemispherical electrode where gave no evidence of significant soil ionisation since the electric field levels at the surface of the hemisphere were lower than the critical soil ionisation gradient, which in the order of 4.5kV/cm. the results shown that the impulse resistance independent from rise time.
Wang et. al [23] proposed a model that the model is an extension to a dynamic model developed by Liew [9]. They claimed that this model is accurately describe the impulse behaviour of the concentrated earthing system at high currents which result in discrete breakdown. They stated that the previous model did not attempt to describe the surge behaviour of the earth at high currents which result in discrete breakdown paths. However, they introduce a fourth region to the proposed three regions called the sparking region, in this region the tracking puncture the earth and earth resistivity drops to very low value in which reduces the concentrated earthing resistance to the minimum.
The work of characterisation of earth electrode [4 -25] under transient condition, including experimental work, laboratory and computer simulation have highlighted significant observations such as:
The impulse resistance is lower than the power frequency resistance.
The reduction is dependent on soil type and electrode arrangements but independent of the current rise time.
Slightly reduction in impedance in high soil resistivity and small reduction for low power frequency resistance.
Inductive effects can be seen in case of high impulse currents.
No ionisation phenomena occur for large earthing system
However, computer simulation techniques helped to represent complicated earthing system and to verify the experimental and laboratory test. Also contribute to obtain better understanding of earthing system behaviour under transient conditions.
Further studies required in order to quantify the effects such as current rise time, sand grain sizes and electrode dimensions of the earthing system.
2.3 Standards guidelines for high frequency earthing requirements
In this section, a brief review is provided of recommendations for earthing systems subjected to transient and lightning surges
EA TS 41-24 [25] (Guidelines for design, testing and maintenance of main earthing system in substations) requested a low impedance value in order to dispraise the high frequency currents safely to the earth, also recommends for impulse earthing the connections from the equipment to the earth should be as short and as free form changes in directions as is practical. The standard suggests that for effectiveness and improving the operation of the arresters by connecting it to high frequency electrode in the immediate vicinity for example an earth electrode.
IEEE 80 [26] (Guide for safety in substation grounding) no guidelines were given in the standard for designing earthing systems subjected to lightning surges but considers that the earthing systems designed for power frequency faults will provide protection against high magnitude transient currents.
BS 6651 [3] Protection of structures against lightning the standard recommends that the earthing system designed for lightning protection should have an earth resistance of less than 10W. The same requirement appears in BS 61400-24:2002 Wind turbine generator system __ Lightning protection with some details of earthing system arrangements of individual wind turbine.
2.4 Windfarm Earthing System
The earthing system of a single wind turbine normally achieved by placing a ring electrode around the foundation and bounding to the tower through the foundation structure, in accordance with the regulation in the standards, the minimum dimension of earth electrodes embedded in concrete is a diameter of 10 mm solid round steel. Vertical rods or strips electrodes are often used in conjunction with the ring electrode to obtain a certain value of a resistance. A resistance of 10 W or less before connected to any other system is stated in standards [3]. The interconnection between the wind turbines by the cable sheath formed an extended earth electrode occupying a large area may result in low resistance. However, this resultant large earthing system requires further investigation in terms of low frequency and high frequency currents, in order to obtain better understanding, hence effective design can be developed.
Different investigators [27-38] studied the earthing system of the wind turbine and wind farm extended earth electrode subjected to power frequency and transient currents.
N, Jenkins and A Vaudin [27] studied the earthing system of windfarm consisting of turbines ratings 300kW to 400kw by conducting site measurements for soil resistivity and resistance of the extended earth electrode of the wind farms. However, they calculated the resistance value of the earthing system using lumped parameter and a compression between the calculated and measured results using slope method showed that the calculated values higher than measured which may be attributed to backfill of the trenches were not compacted, resulting in higher measured values. In addition their results of the soil resistivity show a considerable variation over the site. In their calculation and measurement only resistance mentioned. Earthing systems occupying a large area and long horizontal electrode requires a quite large distance for placing the current electrode in order to obtain reliable results also the impedance should be measured not the resistance due to the inductive component of the long electrode and any calculation of potential rise based on resistance value will be underestimate the hazards.
Nikos Hatziagyrious, Maria Lorentzou, Ian Cotton and Nick Jenkins [28] they reported that the same earthing system is normally used for protection against both power system fault and lightning strikes, the response to either energization source is dramatically different, due to the high frequency components being injected into the earthing system. They investigated the earthing system for individual wind turbine and interconnected wind turbines forming an extended electrode using simulation methods by using computer modelling, EMTP and CDEGS soft wares for modelling the wind farm earthing system subjected to lightning strike. The injection current was 5.5/75 ms rise time and half time with peak current 30kA surge current. They found that the grounding impedance contain significant amount of inductive and became higher compared to the resistance value and also, when using one turbine the difference between testing with low frequency and high is very small in terms of earthing resistance. They conclude that the lower the earthing resistance the lower the voltage rise by the lightning strike, also the benefit of splitting the 50m horizontal electrode into 4 (12.5m) strips in reducing the resistance and voltage rise in both cases as can be seen in the figure number 5.
Figure 5 Effect of short electrode in reducing lightning voltage rise reproduced from reference [28]
I Cotton [29] reported that, following of determining the earth impedance determines the potential rise of the earthing system when current (DC, 50/60 Hz or surge) is injected to it. As a high rise in potential poses a risk to living beings and equipment it desirable to keep the impedance of an earthing system to a low value. However, resistivity measurement has to be taken at each proposed wind turbine location. The lightning protection system earth termination can then be individually designed at each location. He developed a model similar to transmission line model, by splitting the earthing system into small parts and representing each part as a series and shunt impedance. The two series elements represent the impedance of the conductor itself in terms of a resistance and inductance. The two shunt elements meanwhile represent the connection between the conductor and the soil in terms of conductance and a capacitance (the impedance of the shunt connected would be very high for a power cable sheath insulated with a material such as PVC. The site described is based on a windfarm equipped with ten wind turbines, 1.5MW each turbine is being separated by a distance of 300m the substation is 500m from the nearest wind turbine. The radius of the wind turbine foundation is 6.5m the thickness of the metallic sheath is 2.5mm.
The windfarm sited has soil resistivity of 500 Wm (average value). Injection of 1000A at 50Hz into turbine 10 for two different designs the second one an extra earthing electrodes run between the wind turbines and between the final wind turbine and substation. The result show that the maximum potential rise in the first arrangement 826 V at the faulted turbine and 822 V for the second arrangement. The addition of extra earthing electrodes caused a slightly reduction. The impedance of the earthing system uses power cable only is 0.93@29 and with additional earth electrodes are added, the impedance becomes 0.88@33. The same conclusion is drawn the addition of electrodes has not reduced the earth impedance substantially. He reported that another study on different wind farm indicated that a reduction of 20% of the potential when an extra electrodes employed. Therefore, 'studies must be made on an individual basis and what may be true for one windfarm may not true for another' [29].
The relatively large inductive component of the windfarm earthing system impedance show that DC resistance measurement techniques should not be used on a site such as a windfarm. If the resistance value is used in determining the maximum potential rise during a fault a hazard to human safety may occur as the earth potential rise would be underestimated.
A lightning current injected to the turbine 10 base with 30kA peak current and 2.5ms rise time and time to half value20 ms. the results show that higher potential rise in the case of the only turbine 10 when connected with cable sheath considerable decrease. He reported that only the earthing system in the immediate vicinity of the lightning strike is important when determining the maximum rise in earth potential.
N. Hatziargyrious, M. Lorentzou, I. Cotton, N. Jenkins, [30] conducted computer simulations on wind turbine earthing system for both power frequency current and fast transient injection. Moreover, they recommend the use of package software's techniques to analyse complicated earthing systems such as wind farms earthing systems. They developed a model representing the WT earthing electrode as an octagonal structure of 6.5m radius consisting of 4mm radius of copper conductors, buried at 1.5m depth. This model gives a DC ground resistance of 24.8W when buried in 500 Wm soils; Figure 6 illustrates the model. 9kA 1.4/1.7ms impulse current and 9kA at 50Hz injected to the octagonal model attached with 300 m horizontal electrode. The results show that in the case of 50Hz the EPR decreases as the length increased. On the other hand in the case of impulse the decrease for shorter line compared with 50Hz case. The effective length of high frequency is shorter than at 50Hz. They extended the study to include 5 wind turbines and substation earth grid when interconnected. The distance between two successive wind turbines is equal to 400m the same distance between the nearest wind turbine to the substation. The soil is considered homogenous with 500 ohm meter. The interconnected earth electrode has a radius of 4mm and buried at 1.5m depth. The substation earth grid dimension is 15mX30m. Its earth resistance is 9.57W. The investigation carried out includes the following situations:
(i) The wind turbines earth electrodes are interconnected by the armour of the power cable.
(ii) 25m horizontal earth electrodes are connected to either side of each wind turbine in parallel to the power cables.
(iii) Horizontal earth electrode connected in parallel to the whole length of the power cable.
A lightning current of 30kA with 5.5ms and 75ms peak and half peak time is injected to the middle wind turbine. The maximum EPR always appears on the turbine injected as can be seen in figure No. 7. The potential rise reduced when connected the individual wind turbine together no reduction noticed when an extra electrode in parallel with power cable is connected.
a) Simplified model of WT base earthing system
1 2 3 4 5
S/S
b) Connected 5 Wind turbines 400m spaced from each other
Figure 6 illustrates the turbine model and the wind farm arrangement
Figure 6 maximum EPR in % shows that the max at injected turbine (reproduced form reference [31]
Prousalidis et al [31] investigated the soil ionisation of the wind turbine earth electrode using EMTP programme. Different electrode shape were studied the simple vertical electrode, the horizontal and the square arrangement electrode which represent the wind turbine foundation. However, impulse current of 1kA to 100kA with rise time 1.2ms and tail 50 ms is considered. The results show that the ionisation phenomenon reduces the total grounding resistance and the potential rise. The study considered only for one turbine as has a small earthing system, but in fact wind turbine base has a bigger earthing system in which the ionisation may not occur there for a study on a complete model required in order to include the ionisation in the case of wind turbine or not. More simulation and site tests required to investigate the response of the interconnected earth electrode of the windfarm.
Lorentzou et al [32] studied the wind turbine earthing system using computer simulation when, turbine earthing system subjected to power frequency faults and transient conditions. The simulations were conducted on different earthing arrangements as shown in Figure 8 in order to select the effective earthing system design in terms of dispersion of imposed currents and minimisation of raised potentials. The results show that arrangement B has the lowest resistance compared to the other arrangements. They calculated the resistance of one turbine and the impedance of the interconnected turbine, their results show that the increase of the burial depth has no important reduction in resistance but, reduces the touch and step potentials. Moreover, a lightning hit a single turbine show that the reactive component can be neglected, due to small earthing system can be represented by pure resistance. However, when turbines interconnected the highest potentials found on the injection point. The interconnection of the wind turbines reduces the overall impedance with conductor length on both cases due to the effective length. Effective length is the length at which no important reduction in impedance value and depends on soil resistivity and frequency. The difference between the effective length in both cases are shown bellow in Figure 9.
A B C
Figure (8) Wind turbine earthing arrangement used by [ ]
Figure 9 the maximum potential rise vs. conductor length reproduced from reference [32]
In a review paper by Ukar et. al [33] reported that when connecting each wind turbine earthing system to other windfarm earthing system makes between (1-2) W resistance in steady state condition. Moreover, the resistance has to be replaced by impedance, due to the high frequencies the reactive component become dominant. Consequences of this effect are not quantified accurately. Also they highlighted the difficulties of designing the WT earthing system such as the high resistivity site and high cost installation. The authors mentioned that only few works published in reference to the behaviour of lightning discharges energised wind turbines earthing network using commercial software packages. Also the standard methods of determining the earth impedance are not suitable in the case of windfarm. In the review they conclude that in order to minimise the earth potential rise (EPR) the use of horizontal electrode to join different rings that surrounds the wind turbines is recommended forming an extended earth electrode of the windfarm, but taking in account the importance of exceeding the effective length.
Yasuda and Funabashi [34] conducted sets of computer simulation on windfarm earthing system hit by lightning using digital transient simulator ARENE. However, their model consists of two turbines of 1MW each and local transformer spaced by 1km intervals arranged in three different ways as shown in figure (10). The analysis conducted by varying parameters such as the resistance is from 1W to 10W of wind turbine foundation resistance and inductance from 0 to 10mH in lumped constant and the impulse current 30kA with 2ms front wave and 70ms wave tail. The results show that the surge propagating form the thunderstruck wind turbine to the next one may become large depending on the condition of the earthing system. The lower wind turbine earth resistance provides safer protection. Moreover, the inductance has a large effect in terms of high earth impedance and high potentials around the struck turbine and to the next turbine due to the surge propagation. It was found that the arrangement B gives safer results than the other arrangements in terms of surge back-flow. They concluded that it's necessary to conduct more detailed analysis considering wind farm arrangement and wind turbine earth impedance in various soil resistivities.
Figure 10 Turbines arrangement and lightning struck reproduced from reference [34]
In subsequent paper Yasuda and Funabashi [35] studied the effect of the surge propagation resulted form wind turbine subjected to direct lightning to the next wind turbine. By using EMTP software and extended the investigation to include 10 interconnected wind turbines model with the influence of changing earth resistance value of individual wind turbine on the surge propagation. They reported that when the earth resistance high a great amount of transient current will flow into the thunder-struck turbine and quite high amount will propagate to the next turbines. This surge propagation decreases along the turbines until the 5th turbine the decrease became not noticeable, this can be explained by the effective length, the length at which no decrease beyond this point. When the earth resistance very small they noticed that the surge power at the thunder- struck turbine is weaker, but the surge propagation almost still the same to the distant turbine. The results show that the magnitude of the EPR higher when the resistance is high and weaker when the resistance is low but the propagation is independent from the resistance value.
Kontargyri et al [] studied an interconnected wind turbines earthing system and the influence of the of the two layer soil model with various soil resistivity values for top and lower layer on the wind farm earthing system using CDEGS software for frequencies up to 1Mhz. The model adopted in their study is shown in figure 1 which containers 5 interconnected wind turbines. The burial depth of the outer ring 2.5m with radius of 6m, inside ring 0.9m with radius of 1.5m and the outer rectangular is 35X25m buried at 0.9m. The radius of the conductor and the length of the interconnected conductor between each turbine were not given. However, their results show that the impedance of the earthing system is lower as the top layer has low resistivity with high thickness also for low resistivity of the top layer with low depth the influence of the lower layer will be significant. Furthermore, the when the current injected to the central wind turbine the earth impedance will have a value compared to the injection to the terminal wind turbine. This simple mode may be can be used for prediction of the performance of the earthing system for low frequency. In order to obtain accurate results when the earthing system of the turbines subjected to impulse currents the above ground structure has an influence which is effect on the performance of the earthing system. Therefore, the steel tower for the wind farm should be included.
Interconnected wind turbine earthing system (Reproduced from reference 1)
Yasuda [36] proposed installation of two rings attached to the nacelle and the tower as can be seen in figure number 11. These rings make the spark over between the two rings resulted form the lightning current to flow safely to the earth without damaging any part of the wind turbine. However, the author developed 1/100 scale model of 2MW wind turbine. The injection was made by using an impulse generator, 664 kV and 1.3micro second front time and 49 micro second tail were applied to the scale model a spark over was occurred between the two rings. The results show that the space between the two rings should be less than the space between the ring and the nacelle, in order to lead the current safely to the earth through the outer down conductor. Moreover, the author conducted a computer simulation to the same model using FDTD method to clarify the results. When compared the simulation results to the results obtained by the scale model were in good agreement. However, in order to recommend this method a field test using this proposal should be conducted.
Figure 11 proposal of the two ring electrode for lightning protection for wind turbine reproduced from reference [36]
Yamamoto et. al [37] investigated the performance of the Wind turbine earthing system by conducting an experimental work on a scale model represents 3/100 size of the actual wind turbine size. All parameters are 3/100 from the actual quantities in order to obtain actual results. A fast front current generated by the pulse generator is injected to the turbine foundation. The results show that the current has ramp wave and its rise time is approximately 5ns. The voltage is inductive at the wave front this can be seen from the wave shape oscillation in the figure no 12. The resistance obtained form the ratio of the maximum voltage at wave front and current at the same time is 250W while the steady state resistance is 75W. They changed the point of injection the results obtained show that the waveform is changed depend on the injection point. When the lightning hits the nacelle the wave front more steeper and large than those in case of a lightning strike to one of the blades. Their conclusion can be summarised as following the earthing system of the foundation become inductive at the wave front and became capacitive after that. Moreover, the voltage rise on the foundation and the surrounding soil could cause a break down in the outermost insulating layer of the cable.
Figure 12 earthing response due to the earth foundation injection reproduced from reference [37]
Yasuda and Ueda [37] conducted a simulation study on different wind turbine earth electrode arrangements, in order to clarify the recommendation of placing a ring earth electrode around the wind turbine foundation by IEC, TR61400-24 [2] using FDTD method. IEC 61400-24 recommends arrangement B for wind turbine earthing system, this arrangement was originally defined as an earthing method for ordinary houses or buildings in IEC 61024-1 2003 [3] and IEC 61024-1, 2 1998 [3]. The concept of the earth electrode is to create equipotential bounding surrounding a house or a building to avoid values of step and touch potentials that conventionally are considered dangerous. The authors reported that, a discussion concerning earth ring electrode especially for wind turbine has not been exhaustive. Furthermore, the impedance of the ring may have both inductive and capacitive characteristics in high frequency region such as lightning surges.
The study was carried out for different wind turbine arrangement as follows:
foundation with 4 vertical electrode attached to the corners
foundation with earth electrode and outer ring earth electrode
comparison of 15m radius of earth electrode to vertical electrode
In the case with the vertical rods the impedance curves show that evident of vertical rods in reduction of impedance compared with the case of foundation only as can be seen in figure no 13. Also, the impedance decreases with the depth of the vertical electrode. Moreover, the potential rise curves have moderate peaks and show slight inductivity. In the case of outer ring electrode the result curves show the suppressed impedance value and its value depends on soil resistivity as can be seen in figure 14. However, they concluded their results by conducting a comparison between 15m square placed around the wind turbine foundation and vertical rods, they found that the effect expected from 15m square have an equivalent effect from vertical rods driven to 25m depth.
Figure 13 Impedance of WT foundation Figure 14 Impedance of the with Vertical rods foundation with ring electrode
Reproduced form reference [37]
In subsequence paper Yasuda et. al [39] investigated the recommendation of the technical report 61400-24 of the additional electrodes is required if the radius of the foundation of the wind turbine is less than the minimum radius stated by IEC 62305-3 [3]. The additional electrodes given by 62305-3 are originally considered for larger earthing systems. Moreover, is it suitable for wind turbine with smaller foundation area? And what is the best combination of ring earth electrode and vertical rods to obtain effective lightning protection?
The authors reported that a few work and reports have been published as far as wind turbine earthing system and wind farm earthing system concerned. They stated that 'the technical report 61400-24 describes an earthing system for wind turbine but, the description is slightly short and does not seem to be most-to-least' [39]. However, in previous work they investigated the benefits of the ring earth electrode in terms of the lightning protection. They conducted a computer simulation on the following earth electrode combination using FDTD method:
4 vertical electrodes are installed in the bottom of the WT foundations' corners.
4 vertical electrodes are installed in the bottom of the ring electrode.
Combination of i) and ii)
The results obtained show that in the case one the recommendation can not be met, but in case two the results satisfy the standard and in the last case the results stay within the recommended limit as can be seen in figure (15). However, they recommend that improvement to the standard should be taken in account in order to obtain effective design for the WT earthing system.
Figure 15 resistance vs. vertical electrode length for various wind turbine electrode configuration Reproduced form reference [39]
Conclusion
A brief review of literature relating to the high frequency and transient performance of earth electrodes and windfarm earthing system has been carried out. Published work dates from early of the last century and the subject continues to attract various investigators. However, the few work published recently on characterising the behaviour of the earthing system of the wind turbine subjected to both power frequency and to high impulse currents.
Investigators have attempted to characterise the behaviour of earth electrodes using a number of different approaches. Work has involved high-voltage testing both in the laboratory and on earth electrodes installed in the field. A brief review of simulation methods is presented. Computer software packages for analysing complex systems are described. The simulation methods can be carried out on a model based on the detailed geometry of the earthing system buried in either uniform or layered soil structures.
The open literature contains comprehensive studies on simple electrode subjected to transient currents. Moreover, the behaviour of the earth electrode subjected to power frequency faults have been studied and well understood.
The literature clearly identified the difficulties of the measurement and the test procedures of the earthing system of the wind turbines and the extended earth electrode of the windfarm. Different earthing arrangement for wind turbine presented
The standards dealing with the recommendation for earthing systems contain guidelines for design of the earthing system and the earthing system is designed primarily for power frequency earth fault conditions. However, the earthing system subjected to high impulse will behave differently form expected power frequency faults.
Wind turbines earthing systems have been reviewed and the limitations of the measurement procedures have been highlighted. The wind farm occupying an extensive area with very high soil resistivity, in order to measure the earth impedance using fall of potential method a large distance for the current electrode required. This distance is not attainable for such sites. Also, the conventional methods of calculating the earth impedance of such earth electrode were unsuitable. Large earthing systems, the extended earth electrode of the windfarm exhibit a significant reactive component and any estimation of earth potential rise (EPR) based on d.c. measurements is likely to contain a significant error. Therefore, a.c. tests should always be employed on such systems.
Published work has identified the length of an earth electrode as being an important parameter affecting its performance. An 'effective length' has been defined for a wind turbine attached to the interconnection conductors. A comprehensive study required in order to justify the effective design for the wind turbine and interconnected windfarm due to the high soil resistivity sites and the considerable variations in soil resistivity values within one site. Moreover, field tests and site measurements required to compare results with simulation results obtained with detailed earthing arrangements from an operating wind turbine. However, the published literature did not contain experimental work of characterising the performance of the wind farm earthing system for dc, power frequency and transient faults. The field tests and developed models have to be carried out, in order to obtain better understanding of the behaviour of the earth electrode of the wind turbine subjected to both power frequency and high transient currents. The models that presented in the available literature were simulated as a simple electrode and the tower of the wind turbine was not taken in account as the tower may contain inductive element which has to be taken in account when modelling the wind turbine.
References
[1] Global Wind Energy Council: Press Release, Global Wind Energy Market Continue to Boom-2006 another record year, (2007, 2).
http//www.gwec.net/uploads/media/07-02_PR_Global_Statistics_2006.pdf
[2] IEC Technical Report, Wind Turbine Generation System -24: Lightning Protection, TR61400-24, 2002.
[3] IEC International Stand Protection of Strictures against Lightning IEC 61024 - 1998.
[4] H.M.Towne, Impulse Characteristics of Driven Grounds, General Electric Review, Vol.31, No.11, p.605, November, 1928.
[5] L.V.Bewley, Theory and Tests of the Counterpoise, Electrical Engineering, No.53, p 1163, August 1934.
[6] P.L.Bellaschi, Impulse and 60-Cycle Characteristics of Driven Grounds, AIEE Transactions, Vol.60, p.123, March 1941.
[7] P.L.Bellaschi, R.E.Armington, A.E.Snowden, Impulse and 60-Cycle Characteristics of Driven Grounds -II, AIEE Transactions, Vol.61, p.349, 1942.
[8] K.Berger, 'The Behaviour of Earth Connections Under High Intensity Impulse Currents', CIGRE Paper No.215, 1946.
[9] A.C.Liew, M.Darveniza, Dynamic Model of Impulse Characteristics of Concentrated Earths, Proc IEE, Vol.121, No.2, p.122, February 1973.
[10] B.R.Gupta, B.Thapar, Impulse Impedance of Grounding Grids, IEEE Transactions on Power Apparatus and Systems, Vol.PAS-99, No.6, p.2357 Nov/Dec 1980.
[11] B.R.Gupta, V.K.Singh, Impulse Impedance of Rectangular Grounding Grid, IEEE Transactions on Power Delivery, Vol.7, No.1, p.214, January 1991.
[12] M.Ramamoorty, M.M.Babu Narayanan, S.Parameswaran, D.Mukhedkar, Transient Performance of Grounding Grids, IEEE Transactions on Power Delivery, Vol.4, No.4, p.2053, October 1989.
[13] R.Kosztaluk, M.Loboda, D.Mukhedkar, Experimental Study of Transient Ground Impedances, IEEE Transactions on Power Apparatus and Systems, Vol.PAS-100, No.11, p.4653, November 1981.
[14] R.Velazquez, D.Mukhedkar, Analytical Modelling of Grounding Electrodes Transient Behaviour, IEEE Transactions on Power Apparatus and Systems, Vol.PAS-103, No.6, p.1314, June 1984.
[15] A.Geri, G.M.Veca, E.Garbagnati, G.Sartorio, Non-linear Behaviour of Ground Electrodes Under Lightning Surge Currents: Computer Modelling and Comparison with Experimental Results, IEEE Transactions on Magnetics, Vol.28, No.2, March, 1992.
[16] A.M.Mousa, The Soil Ionisation Gradient Associated With Discharge of High Currents Into Concentrated Earths, IEEE Transactions on Power Delivery, Vol.9, No.3, p.1669, July 1994.
[17] L.D.Grcev, F.E.Menter, Transient Electromagnetic Fields Near Large Earthing Systems, IEEE Transactions on Magnetics, Vol.32, No.3, p.1525, May 1996.
[18] M.E. Almeida, M.T Correia de Barros accurate modelling of Rod Driven Tower Footing Resistance, IEEE Transaction on Power Delivery, Vol.11, No.3, pp.1606-1609., July 1996
[19] S. Skioka, H. Hayashida, T. Hara, A. Ametani: 'Measurement of Grounding Resistance for High Impulse Currents', IEE proceedings Generation, Transmission and Distribution, Vol. 145, pp. 693-699, 1998.
[20] Stojkovic, M.S. Savic, J.M. Naham, D. Salamon, B. Bukorovic, 'Sensitivity Analysis of Experimental Determined Grounding Grid Impulse Characteristics', IEEE Transaction on Power Apparatus and Systems, Vol. PAS 13, No. 4, pp 1136-1141, October 1998.
[21] L.D.Grcev, V. Arnautovski, Toseva, 'Grounding and Transient: Some Fundamental Considerations', IEEE Bologna PowerTech conference, Italy, June 23-26, 2003
[22] Y. Chen and P. Chowdhuri, Correlation between laboratory and field tests on the impulse impedance of rod-type ground electrodes, Proc. Inst.Elect. Eng. Generation Transm. Distrib., vol. 150, no. 4, pp. 420-426, Jul. 2003.
[23] J. Wang, A.C.Liew, M.Darveniza, 'Extension of Dynamic Model of Impulse Behaviour of Concentrated Ground at High Currents', IEEE Transaction on Power Delivery, Vol.20, No.3, pp.2160-2165., July 2005.
[24] N.M. Nor, 'Review: Soil Electrical Characteristics Under High Impulse Currents', IEEE Transaction on Electromagnetic Compatibility, Vol.48, No.4, pp.826-829., November 2006.
[25] Electricity Association Technical Specification 41-24: 'Guidelines for the design, installation, testing and maintenance of main earthing systems in substations.' Electricity Association, London, 1992.
[26] IEEE guide for safety in AC substation grounding, ANSI/IEEE standard 80, 1986
[27] N, Jenkins and A Vaudin, Earthing of Wind farms Renewable Energy, 17-19 November 1993, Conference Publication No. 385 IEE 1993.
[28] N. Hatziargyrious, M. Lorentzou, I. Cotton, N. Jenkins Windfarm Earthing The Institution of Electrical Engineers, IEE, Savoy Place, London, 1997
[29] I. Cotton, Windfarm Earthing in Proc International Symposium on High Voltage Engineering, Vol.2, pp. 288-291, London, UK, 1999.
[30] N. Hatziargyrious, M. Lorentzou, I. Cotton, N. Jenkins, Transferred Overvoltages by Windfarm Grounding Systems Proceedings of ICHQP conference, 14-16 October 1998, Athens, Greece.
[31] J. M. Prousalidis, M. P. Philippakou, N. Hatziargyrious, B. Papadias, The Effect of Ionization Wind Turbine Grounding Modelling, Proc 10th Mediterranean Eletrotechnical Conference (MELECON 2000, pp940-943, Vol.3), 2000.
[32] M. Lorentzou, N. Hatziargyrious, B. Papadias, Analysis of Wind Turbine Grounding System, Proc 10th Mediterranean Eletrotechnical Conference (MELECON 2000, pp. 936-939, Vol.3), 2000.
[33] O. Ukra, I. Zamoua, R. Idiondo, A. Mugica Analysis of high frequency of Grounding Systems, For Wind Turbines, Proceedings of the 26th international Conference on Lightning Protection (ICLP), Vol.2, pp.1053-1057, Avignon, France, September, 2004.
[34] O. Yasuda, T. Fnabashi:, Transient Analysis on Wind Farm Suffered from Lightning Proceeding of 39th International Universities Power Engineering Conference, pp.202-206, 2004.
[35] Y. Yasuda, T. Fnabashi: Lightning Analysis on Windfarm-Sensitivity Analysis on Earthing Proceedings of the 27th International Conference on
[] V.T. Kontargyri, I.F. Gonos and I.A. Stathopulos, 'Frequency Response of Grounding System of Wind Turbine Generators', Proceedings of the XIVth International Symposium on High Voltage Engineering, Tsinghua University, Beijing, China, August 25-29, 2005
Lightning Protection (ICLP), pp.1041-1046, September 2004.
[36] Y. Yasuda: Proposal of Ring-shaped Lightning Conductor for Wind Turbine,
Proc. of European Wind Energy Conference & Exhibition 2005 (EWEC2006),
2005, Athens
[37] K. Yamamoto, T. Noda, S. Yokoyama, A. Amentani: Ground Characteristics
of a Wind Turbine Generation System and Voltage Rise Around It
Proceedings of International Conference on Grounding and Earthing
(GROUND'2006), November 2006.
[38] Y. Yasuda, U. Toshiaki, FDTD Transient Analysis of Ring Earth
Electrode Proceeding of 41stInternational Universities Power Engineering
Conference, pp.133-136, 2006.
[39] Y. Yasuda, F. Toshiaki, U. Toshiaki, How Does Earth Electrode Effect on
Wind Turbine Proceeding of 42ndInternational Universities Power
Engineering Conference, pp.796-799, 2007.
2.10 References:
[2.1] Dwight H.B.: 'Calculation of resistance to ground.' Electrical Engineering, December 1936, pp 1319 - 1328.
[2.2] Schwarz S.J.: 'Analytical expressions for the resistance of grounding system.' July 1951, pp368-402 (Translated from the original French in IEEE guide for safety in AC substation grounding., ANSI/IEEE standard 80,1986).
[2.3] Laurent P.: 'Les bases générales de la technique des mises la terre dans les
installations électriques.' Le bulletin de la Société Franaise des Electriciens,
1967(Translated from the original French in IEEE guide for safety in AC
substation grounding, ANSI/IEEE standard 80,1986.
[2.4] Tagg G.F.: 'Earth Resistance.' (George Newnes Ltd, England, 1964)
[2.5] Sunde E. D., Earth conduction Effects in Transmission line Systems, Dover Publications Inc., 1968.
[2.6] Electricity Association: `Engineering Recommendation S.34: A guide for assessing the rise of earth potential at substation sites.' Engineering Management Conference, May 1986.
[2.7] Edison Electric Institute.: 'EHV Transmission Line Reference book.' USA, 1968.
[2.8] Endrenyi J.: 'Analysis of Transmission Tower Potential during Ground Fault' IEEE Transaction on PAS, Vol.86, No.10, pp 1274 - 1283 IEEE, Vol. 73, Part III-B, 1954, pp 1011 - 1016.
[2.9] IEEE guide for safety in AC substation grounding., ANSI/IEEE standard 80,1986.
[2.10] BS 7354:1990: 'Code of practice for the design of high-voltage open terminal stations.' British Standards Institution, 1990.
[2.11] Electricity Association Technical Specification 41-24: 'Guidelines for the design, installation, testing and maintenance of main earthing systems in substations.' Electricity Association, London, 1992.
