Investigation on the Performance of Different Lightning Protection System Designs

IX- Investigation on the Performance of Different Lightning Protection System Designs Nicholaos Kokkinos, ELEMKO SA, Ian Cotton, University of Manchester Abstract-- In this paper different lightning protection system (LPS) are modeled so that the scalar potentials on the LPS components and the magnetic fields that are generated due to the flow of the lightning current on the LPS s can be recorded. Furthermore an investigation regarding the induced voltage and currents due to the magnetic fields on cables inside the LPS will also be presented. Index Terms-- Lightning, lightning protection system, magnetic fields, overvoltages. I. INTRODUCTION In the event of a lightning strike on an earthed structure, the external lightning protection system (LPS) should safely conduct the lightning current through the LPS s into the earth. Although a direct contact between the structure and the lightning flash might be avoided (due to the presence of the LPS), the current flow on the LPS s will generate high potentials on the s themselves, which might lead to a flashover [] between the LPS s and other metallic objects (e.g. Air Condition units, Doors, Windows and Fences) that are in the vicinity of the LPS s. But this is not the only problem that might arise due to a poor LPS design. The flow of the lightning current on a will generate strong magnetic fields, which will induce overvoltages and surge currents on cables and electrical installations inside the structure. The magnetic field is proportional to the current, therefore by distributing the current on more LPS s (act as a Faraday cage) the magnetic fields can be reduced. Therefore it is the aim of this paper to analyse different LPS designs and to present the current distribution and Contact Address: Dr. Nicholaos Kokkinos ELEMKO SA, Tatoiou 90 str, 52 GR,Metamorphosis, Attiki,, Hellas e-mail: nkokkinos@elemko.com the potential on the LPS s and also the magnetic fields near and inside the structure. The results can be calculated either by using mathematical formulas and methods that have been invented and analyzed by previous researchers [2], [3], [] or by using one more powerful method, a computer software simulation package. The software that was used for all the following simulation results is called CDEGS [5] (Current Distribution, Electromagnetic Fields, Grounding and Soil Structure Analysis). II. CDEGS SIMULATION MODELS For a simple rectangular building with the following dimensions: Length 0 meters width eters and height eters, four different LPS designs were examined and the resultant current distribution, potential on the s and the magnetic fields were recorded. The four different LPS to be examined are the following: Simple type lightning protection with just a rod to intercept with the direct lightning current and one down connected to an earth ring 5x20m, see figure. LPS with four meshes 20xm for air termination, four down s near each edge of the structure, connected to the same earth ring as in the previous case, see figure 3. LPS2 with four meshes 20xm for air termination, twelve down s ten meter apart from each other all connected to the same earth ring as in the previous two cases, see figure 5. LPS3 (figure 8) with sixteen meshes x5m for air termination, twelve down s, ten meter apart from each other similar to LPS2 and all connected to an earth ring similar to the previous cases. The LPS designs were based on European (EN) and International (IEC) standards, although that the first two -- th International Conference on Lightning Protection

cases did not fulfil standards [6], [] there are real cases where such installations occur. The earthing system (which is a part of the external LPS) was installed at 0.5 meter depth in a uniform soil with a 0 Ωm resistivity, a relative permittivity of and a relative permeability of are also assumed. The lightning surge considered throughout this paper is defined as a double exponential type function αt βt I t = I e e where I m =30 ka, α=.x () ( ) m sec - and β=6x 6 sec -. The waveform is characterized by a rise time of.2 µsec and a tail time of 50 µsec. III. CURRENT DISTRIBUTION AND POTENTIALS ON LPS CONDUCTORS A. Simple type lightning protection The LPS s will be in much higher potential than any metallic object (e.g. Doors, Windows, and Fences) in the vicinity of them (unless effective bonding is provided). Therefore a potential difference between the two parts might lead to a flashover (the generation of an arc between the LPS and the metallic objects). This event is known as a sideflash..50e+06 6.00E+06.50E+06 3.00E+06.50E+06 -.50E+06-3.00E+06.00E-06 2.00E-06 3.00E-06.00E-06 5.00E-06 6.00E-06 Time (usec) Fig. 2: Potential at the top of the down of Simple type lightning protection Even nowadays where national and international regulations clearly specify on how an efficient lightning protection system should be designed there are cases were LPS installations do not fulfill these specifications. It is a fact that LPS installations with just a single rod and only one down are currently used to protect a structure against lightning. Figure represents the schematic configuration of this particular design. Even if such an LPS installation can provide protection against a direct lightning strike there are many disadvantages, which prove that it should not be used. These disadvantages will be discussed in this chapter. The energisation point in figure is at the top of the lightning interception. The lightning current will flow through the only down and discharge through the earthing ring into the soil. The potential at the top of the down can be seen in figure 2. B. Lightning protection system LPS LPS (see figure ) design consists of four 20 x meshes on the air termination system, four down s at each edge of the structure and a buried earth electrode at 0.5 m into the soil, which connect all the down s. The lightning current is equally subdivided in the down s. The equal distance and geometry between the four down s and the energisation point and the fact that all the down s are bonded together with the earth ring in a uniform soil explains the equal subdivision of the lightning current. 0. 0 m 0. 0. Fig. 3: LPS design and current distribution on the down s according to Kirchhoff's voltage law based on the resistive part of the s (DC-Resistance) only 0. Fig. : Simple type lightning protection system design 0 m The voltage was measured approximately.mv with respect to earth. The oscillations and the negative s can be explained by traveling wave theory. (If a surge propagates down a and finds a discontinuity point, i.e. a change in impedance, e.g. where the down joints to the earthing system, some of it will be reflected back to the energisation point. For more information on traveling wave theory see [8] ) Figure shows the voltage variation between the top and the bottom of the down s. The voltage difference between the two points can be easily observed especially at the very initial stages (i.e.0-0.5 µsec) due to the voltage drop across the surge impedance of the LPS s. The important information that was extracted from these results was that due to the high potential at the top of the LPS s, a flashover is more likely to happen at the top than at the bottom of a structure. Therefore care must be taken when designing the air termination and the down part of the LPS in order to avoid flashover. Unnecessary loops within the LPS s -- th International Conference on Lightning Protection

must be avoided and bonding requirements must be considered..20e+06.00e+06 8.00E+05 6.00E+05.00E+05 2.00E+05 the actual lightning current that will flow though and not the real current, which also includes the electromagnetically induced current. Figure 6 and show the variation of current and voltage for each set the down s respectively. Table contains the values from the voltage and current graphs. -2.00E+05 -.00E+05-6.00E+05.00E-06 2.00E-06 3.00E-06.00E-06 5.00E-06 6.00E-06.00E-06 8.00E-06 Bottom Top Fig. : Voltage variation between the top and the bottom of LPS down s C. Lightning protection system 2 LPS2 Figure 6 shows the second lightning protection design LPS2, which will be under consideration. In the previous case of LPS, although the principles of the design were based on IEC 62305-3 and EN 62305-3 it did not fulfil all of the necessary requirements exactly as they are mentioned in the standards. Regardless of the lightning protection level, distances between the down s should be a maximum of 20m or m according to both standards. Therefore in this case the same design as before will be examined but with twelve down s placed m apart from each other. The energisation point will remain in the middle of the air termination system. It should be noted that sets of the s, whose distance and symmetry are equal and identical respectively from the energisation point, will have an equal current flow through them. This is also due to the fact that they are all bonded together in a uniform soil. 0.06 0.099 0.0 0. 0.0 Half face middle down Face middle down 0.06 0.099 0.06 0.0 0. 0.0 0.06 Side down Edge down Fig 5: LPS2 design and current distribution on the down s according to Kirchhoff's voltage law based on the resistive part of the s (DC-Resistance) only Conductors near the energisation point, carry more current. Comparing the side s with half face middle s, shows that the distance from the energisation point is the same however, the current that flow through them is not the same. That is because the side s are directly connected to the energisation point (without any intermediate node) while the half face middle s are connected through one node. In other words the current will flow through the easiest path, which will drive it to earth. Note that the current values on the down s in figure 6 is the percentage of Current (A) 500 000 3500 3000 00 2000 500 00 500 0 0.0E+00.0E-05 2.0E-05 3.0E-05.0E-05 5.0E-05 6.0E-05.0E-05 8.0E-05 Edge s Side s Half face middle s Face middle s Fig. 6: Current distribution on the down s of LPS2 9.00E+05 8.00E+05.00E+05 6.00E+05 5.00E+05.00E+05 3.00E+05 2.00E+05.00E+05 0.0E+00.0E-0 2.0E-0 3.0E-0.0E-0 5.0E-0 6.0E-0.0E-0 8.0E-0 9.0E-0.0E-06 -.00E+05-2.00E+05 Edge Conductors Side s Half face middle s Face middle s Fig. : Voltage variation on LPS2 down conduct TABLE : PEAK CURRENT AND VOLTAGE VALUES FROM FIGURE AND FIGURE 8 Conductor Edge Side Current Voltage Half face middle Face middle,9 A 2,803 A 2,3 A,55 A 69 kv 80 kv 593 kv 653 kv D. Lightning protection system design 3 LPS3 Figure 8 represents LPS3, which is similar to LPS2 but with an improved air termination system. The air termination system was designed according to the mesh method. The mesh size is x 5 meters. -2- th International Conference on Lightning Protection

5 m IV. ELECTROMAGNETIC FIELD ANALYSIS WITHIN A STRUCTURE DUE TO THE FLOW OF LIGHTNING CURRENT ON THE EXTERNAL LIGHTNING PROTECTION SYSTEM 0.0 0.08 Side down 0.085 0. 0.085 Edge down Half face down Face middle down 0.08 0.0 0.085 0. 0.085 0.0 Fig. 8: LPS3 design and current distribution on the down s according to Kirchhoff's voltage law based on the resistive part of the s DC-Resistance) only Figure 9 and show the current and voltage variation on LPS3 down s respectively. Table 2 contains the values from the voltage and current graphs. The lightning current has been more evenly distributed between the down s. The edge, side and half face middle s carry approximately the same amount of current. The potential difference at the top of the LPS s is not very high. Therefore by using more meshes on the air termination system, flashover probability can be reduced. TABLE 2: PEAK CURRENT AND VOLTAGE VALUES FROM FIGURE AND FIGURE Conductor Edge Side Current Voltage Current (A) 3500 3000 00 2000 500 00 500 Half face middle 0.0 Face middle 2,026 A 2,29 A 2,560 A 3,300 A 556 kv 62 kv 560 kv 53 kv 0 0.0E+00.0E-05 2.0E-05 3.0E-05.0E-05 5.0E-05 6.0E-05.0E-05 8.0E-05 Edge s Side s Half face middle s Face middle s Fig. 9: Current distribution on the down s of LPS3.00E+05 6.00E+05 5.00E+05.00E+05 3.00E+05 2.00E+05.00E+05 -.00E+05.00E-0 2.00E-0 3.00E-0.00E-0 5.00E-0 6.00E-0.00E-0 8.00E-0 9.00E-0.00E-06 Edge s Side s Half face middle s Face middle s Fig. : Voltage variation on LPS3 down s CDEGS can compute the magnetic fields at single point and at a surface, which consists of many observation points. For the purpose of the following simulations three observation surfaces were selected, one at the top one at the middle and one at the bottom of the structure. Each surface consists of a set of profiles; each individual profile consists of a number of observation points. Figure 2 describes the schematic configuration of the computation procedure. Top surface Middle surface Bottom surface Number of points per profile 0 m Number of profiles per surface Fig. : Three-dimension view of the simulation structure with the three observation surfaces The top observation surface was 0.5 meters from the top of the air termination system. The magnetic field is measured in amperes per meter. Every color represents a range of magnetic field magnitude. All the magnetic field results are extracted at.2 microseconds, which is written on the bottom of the graph. The 3D graph presentation was used in order to represent the magnitude of the magnetic fields at this particular time instant (.2 µsec). The maximum field was recorded at the time of the current, which is represented in table 3. The magnetic field is proportional to the current and therefore LPS designs whose s carry more lightning current are expected to generate stronger magnetic fields. The benefit of having a more detailed air termination system can be understood by comparing LPS3 (figure 5), which has sixteen meshes, with the simple LPS design (figure 2), LPS (figure ) and LPS2 (figure ) whose air termination system consists of one and four meshes only. The lightning current has to flow from the energisation point to the earth. Therefore by creating more conductive paths a better distribution can be succeeded and a better distribution means lower currents through the LPS s and therefore lower magnetic fields near the LPS s. By using more LPS s the structure acts as a faraday cage, which results to the reduction of the electromagnetic fields inside the structure. For the purpose of this paper only the magnetic fields that were recorded on the top of the structure will be graphically presented. The results on tope middle and bottom are discussed further in the paper and are also numerically presented in table 3. -- th International Conference on Lightning Protection

A. Discussion of magnetic field results S2 S S S5 S S S9 S S5 S3 S Fig. 2: Resultant magnetic field at the top of simple LPS at.2 usec (Peak value at that time was 500 A/m) S S 0-200 200-00 00-600 600-800 800-00 00-200 200-0 0-00 00-800 800-2000 2000-00 00-200 200-2600 2600-00 S S S S S Fig. : Resultant magnetic field at the top of LPS at.2 usec (Peak value at that time was 3000 A/m) S2 S S S5 S S S9 S S5 S3 S 0-200 200-00 00-600 600-800 800-00 00-200 200-0 0-00 00-800 800-2000 2000-00 00-200 Fig. : Resultant magnetic field at the top of LPS2 at.2 usec (Peak value at that time was 200 A/m) 0-200 200-00 00-600 600-800 800-00 00-200 200-0 0-00 3 3 3 3 3 3 0 0 0 From the simulation results some basic factors that influence the efficiency of each LPS design in terms of minimizing magnetic field can be stated. It is now evident that a more detailed air termination system can minimize the electromagnetic fields at the top of the structure. Comparing the observation results for LPS2 and LPS3, a significant reduction regarding both the values and the area that is covered by the high magnetic field can be achieved by using more meshes on the air termination system. The number of the down s is another very important parameter, which influences the efficiency of the external LPS. From the simple LPS type result it is obvious that just one down can not provide effective protection against electromagnetic pulses. The use of four down s (LPS) provides a significant reduction in both the values and the area that is covered by strong magnetic fields. LPS2 and LPS3 prove that by using even more down s a further reduction to the magnetic fields can be achieved. The results at the bottom of the structure prove that buried s can also influence the generation of magnetic fields. Regarding the middle and bottom results, it is possible to see that in the center of the structure the magnetic fields are much lower than close to the down s. Another important factor that should be mentioned is the fact that there is no other element apart from air between the inside of the structure and the LPS s. So in a real situation all the magnetic fields should be lower because concrete and especially reinforced concrete is more conductive than air. Reinforced concrete structures act as a Faraday cage, which has as a result the reduction in the magnetic fields inside it. As a total overview of the results it is possible to state that the most efficient design was LPS3. The generated magnetic fields due to the lightning current on the LPS3 s present lower values than in all of the other simulations and also the total area inside the structure that is under a strong magnetic field can be reduced to a minimum by using the LPS3 design. Table 3 summarizes the maximum-recorded magnetic field values inside each lightning protection design for the top, middle and bottom observation surfaces. TABLE 3: SUMMARY OF MAXIMUM MAGNETIC FIELD VALUES FOR DIFFERENT LPS DESIGNS S S S S S S S S S Fig. 5: Resultant magnetic field at the top of LPS3 at.2 usec (Peak value at that time was 0 A/m) 3 3 0 3 6 9 Maximum magnetic field A/m Simple LPS LPS2 LPS3 Top 55,000 8,00,500 6,050 Middle 9,000,000 6,500 5,200 Bottom 2,000 9,50 3,200 2,500 -- th International Conference on Lightning Protection

V. INVESTIGATION OF INDUCED VOLTAGE AND CURRENT SURGES ON CABLES INSIDE THE LPS DESIGNS A metallic loop was simulated near the down s inside of each LPS. The induced voltage and current surges on the metallic loop were depending on the position of the loop inside the LPS design. Therefore two different positions of the metallic loop inside each LPS were examined and the maximum-recorded surges will be presented. The loop dimensions and positions can be seen in figure. Inside LPS, LPS2 and LPS3, the position of the metallic loops are the same but for the simple type the loops were installed near the only down. The energzation point for all the simulations was the center of the LPS. The cable type that was used in the loop was a co-axial type cable with an external 0.0 mm foil screen and 0. mm of PVC insulation. Each loop was individually simulated as an open circuit and as a short circuit. The voltage across the open circuit point is the value of the inductive and capacitive induced voltage across the cable. The current that will flow through the short circuit loop is the generated current due to the induced voltages. Table summarizes the recorded overvoltages and surge currents values for the previously described metallic loops inside the four different LPS designs. Loops inside LPS2 and LPS3 present the lower overvoltage and surge current values, which proves that a higher efficiency LPS design, not only provides higher protection against direct flashes on the structure, but can also minimize the induced voltages and currents on the cables inside the structure. TABLE : INDUCED VOLTAGE AND CURRENT ON A METALLIC LOOP INSIDE DIFFERENT LPS DESIGNS Efficiency and Level Voltage surge (kv) Current surge (A) Simple LPS - (*),550,5 LPS - (*) 380 530 LPS2 80%, IV 5 0 LPS3 95%, II (*) Simple type LPS does not satisfy any of the international standards therefore its efficiency could not be evaluated. (**) LPSdoes not fulfil all of the necessary requirements for a level IV LPS, it only has four instead of six down s. Area = 2 m 2 8 m 5 m 2 m m m m Short Circuit Horizontal arrangement Open Circuit Fig.: Dimensions and position of the metallic loop inside the LPS designs The recorded voltage and current values on the cables are the inductive and capacitive coupled surges and are not due to resistive coupling. Although resistive coupling surges may be more severe than inductive and capacitive coupling, the protection against them is easier that the last two. By installing surge protection devices [-2] at the main distribution boards the resistive coupled surges can be minimized. But the inductive and capacitive coupling surges may appear across installations even if they are not connected to any part of the structure, which carry lightning current. The magnitude of the induced voltage surges varies between some tens of kv up to some hundreds of kv and the current magnitudes may reach values of some hundreds of amperes. Electrical installations connected at these cables will experience these overvoltages and surge currents. The damage that may be caused to sensitive electronic equipment, which operate with only a few tens of volts and few ma could be severe. VI. CONCLUSIONS Designing an efficient external lightning protection system (LPS) the risk of having high induced overvoltages on cables inside the structure can be reduced. International standards provide guidance on how an efficient LPS should be designed where necessary. Simulations investigating the performance of different LPS designs have proved that just a single down can not provide efficient protection regarding the generation of induced overvoltages on metallic loops in the vicinity of the down s. By increasing the number of down s the potential across the LPS can be reduced and the probability of sideflashes can also be minimised. The air termination system is also very important and it has been proved that by using a meshed air termination system a better distribution of the lightning current on the down s can result, which then reduces the potential differences across the LPS components. By using more down s and a meshed air termination system, the generated magnetic fields can also be reduced and therefore, induced overvoltages and overcurrents on cables inside the LPS can be partially controlled. VII. REFERENCES [] Allen, Fundamental aspects of air breakdown, High voltage engineering and testing, IEE Power, 96 [2] Flisowski, Stanczak, Kuca, Mazzeti, Orlandi, Yarmarkin, Induced currents and voltages inside LPS models due to lightning currenr, 23 rd ICLP, Firenze, 96 [3] Noda, Takami, Nagaoka, Amantani, Basic investigation of lightning induced voltages to an electronic appliance, 23 rd ICLP, Firence, 96 [] Sowar, Gosling, Lightning radial electric fields and their contribution to induced voltages, IEEE Symposium on EMC, 99 [5] CDEGS, Safe Engineering Services (SES) and Technologies limited, Montreal, Canada. [6] EN 62305-3, Protection against lightning Part 3: Physical damage to structures and life hazard [] IEC 62305-3, Protection against lightning Part 3: Physical damage to structures and life hazard [8] Pritindra Chowdhuri, Electromagnetic transients on power systems, John Wiley & Sons, 96 [9] IEC 63 2, Low voltage surge protective devices Part 2: Surge protective devices connected to low voltage power distribution systems Selection and application principles -5- th International Conference on Lightning Protection