Dynamic characteristics and cascaded coordination of limiting‐type SPDs under subsequent negative strokes
2016; Institution of Engineering and Technology; Volume: 10; Issue: 16 Linguagem: Inglês
10.1049/iet-gtd.2016.0803
ISSN1751-8695
AutoresYuwei He, Zhengcai Fu, Bengang Wei, Jian Chen, Anfeng Jiang,
Tópico(s)Polyoxometalates: Synthesis and Applications
ResumoIET Generation, Transmission & DistributionVolume 10, Issue 16 p. 4197-4204 Research ArticleFree Access Dynamic characteristics and cascaded coordination of limiting-type SPDs under subsequent negative strokes Yuwei He, Yuwei He Department of Electrical Engineering, Shanghai Jiao Tong University, No. 1954 Huashan Road, Shanghai, 200030 People's Republic of ChinaSearch for more papers by this authorZhengcai Fu, Corresponding Author Zhengcai Fu zcfu@sjtu.edu.cn Department of Electrical Engineering, Shanghai Jiao Tong University, No. 1954 Huashan Road, Shanghai, 200030 People's Republic of ChinaSearch for more papers by this authorBengang Wei, Bengang Wei State Grid Shanghai Municipal Electric Power Company, No. 1122 Yuanshen Road, Shanghai, 200122 People's Republic of ChinaSearch for more papers by this authorJian Chen, Jian Chen Department of Electrical Engineering, Shanghai Jiao Tong University, No. 1954 Huashan Road, Shanghai, 200030 People's Republic of ChinaSearch for more papers by this authorAnfeng Jiang, Anfeng Jiang State Grid Shanghai Municipal Electric Power Company, No. 1122 Yuanshen Road, Shanghai, 200122 People's Republic of ChinaSearch for more papers by this author Yuwei He, Yuwei He Department of Electrical Engineering, Shanghai Jiao Tong University, No. 1954 Huashan Road, Shanghai, 200030 People's Republic of ChinaSearch for more papers by this authorZhengcai Fu, Corresponding Author Zhengcai Fu zcfu@sjtu.edu.cn Department of Electrical Engineering, Shanghai Jiao Tong University, No. 1954 Huashan Road, Shanghai, 200030 People's Republic of ChinaSearch for more papers by this authorBengang Wei, Bengang Wei State Grid Shanghai Municipal Electric Power Company, No. 1122 Yuanshen Road, Shanghai, 200122 People's Republic of ChinaSearch for more papers by this authorJian Chen, Jian Chen Department of Electrical Engineering, Shanghai Jiao Tong University, No. 1954 Huashan Road, Shanghai, 200030 People's Republic of ChinaSearch for more papers by this authorAnfeng Jiang, Anfeng Jiang State Grid Shanghai Municipal Electric Power Company, No. 1122 Yuanshen Road, Shanghai, 200122 People's Republic of ChinaSearch for more papers by this author First published: 01 December 2016 https://doi.org/10.1049/iet-gtd.2016.0803Citations: 4AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract Subsequent negative impulses are steep and common in lightning flashes. Under such currents, the dynamic behaviour of metal-oxide varistors (MOVs) is expected to be significant and would affect the coordination of cascaded surge protective devices (SPDs). This study experimentally investigates the dynamic behaviour of MOVs under steep currents. Based on the dynamic characteristics, simulations are conducted to investigate the coordination of two-cascaded limiting-type SPDs with different cascaded combinations, different separation distances, and different loads under subsequent strokes with 0.25/100 μs current waveform. Meanwhile, the results are compared with that under 8/20 and 10/350 μs currents. Experiments are performed to investigate SPD coordination performance under subsequent strokes. The results show that voltages on MOVs of low-voltage SPDs increase about 6.8-18.7% under steep currents, compared with that under 8/20 μs current. Thanks to the significantly changed dynamic characteristics, the cascaded SPDs could more easily realise effective coordination under subsequent strokes, compared with that under 8/20 and 10/350 μs currents. Moreover, effective coordination can be achieved in large ranges of separation distances and loads under subsequent strokes. The SPD combinations with clamping voltage of upstream SPD a little higher than downstream SPD have the best behaviour for lightning protection in low-voltage systems. 1 Introduction Two-stage cascaded surge protective device (SPD) composed of metal-oxide varistors (MOVs) is a common scheme to protect sensitive equipment against surge currents in electrical and electronic systems [1–4]. In the two-cascaded SPDs, the SPD installed at the service entrance is denoted as class I, which is the upstream SPD employed to divert large amount of the transient energy to ground. The SPD installed near the electrical device is denoted as class II, which is the downstream SPD used to clamp the overvoltage to the level that the protected device can sustain [5]. In such a protection scheme, effective coordination between the cascaded SPDs is of extreme importance. The coordination performance of two-cascaded SPDs could be affected by the waveforms of the impulse currents applied on the SPD circuits, the decoupling elements between the SPDs, the dynamic characteristics of MOVs, and so on [6–9]. Coordination issues on two-cascaded limiting-type SPDs have been addressed under impulse currents normally using 8/20 μs standard waveform, 0.5 μs–100 kHz ring wave, 10/1000 μs long wave, and 10/350 μs wave simulated first positive stroke [8–13]. The 'high–low' cascaded condition, where the clamping voltage of the upstream SPD is higher than the downstream SPD, is suggested for the impulse currents with short duration and low energy, such as 8/20 μs impulse and 0.5 μs–100 kHz ring wave [10]. However, under impulse currents with slow wave and longer duration, such as 10/1000 μs long wave and 10/350 μs impulse, the two-cascaded SPD coordination can be achieved by 'low–high' or equally rated SPDs [8, 10]. Subsequent negative impulses, often existing in natural lightning flashes, have steep wave-front and relative long duration with standard 0.25/100 μs waveform recommended by the IEC [5, 14–16]. The observations on SPDs installed at the service entrance show that subsequent negative impulse is a common current component encountered by the SPDs [17]. Under such steep currents, as the IEEE working group 3.4.11 has pointed out, the dynamic characteristics of MOVs would change significantly, especially when the current front time decreases to near 1 μs [6, 18]. Due to the difference of the waveforms between subsequent negative impulse currents and the normally used impulse currents, as well as the significantly changed dynamic characteristics of MOVs, the cascaded SPDs that have been verified as effectively coordinated under normally used impulses need to be checked under subsequent negative impulses. Due to the difficulty of theoretical analysis, experiments are carried out to investigate the dynamic behaviour of MOVs used for low-voltage (LV) SPDs under subsequent strokes. Based on the dynamic characteristics of MOVs, the coordination performance of two-cascaded limiting-type SPDs under subsequent negative impulses with different SPD cascaded combinations, different distances between the SPDs, and different loads being protected are simulated employing the transient software EMTP. Meanwhile, simulations considering the above-mentioned issues are also conducted under normally used 8/20 μs impulse current and 10/350 μs wave simulated first positive impulse current for comparison. Experiments with steep current simulated subsequent strokes are also carried out to investigate the coordination performance on the actual SPD circuits. The results of these works could provide reference for the improvement of the selection and configuration design for cascaded SPD protection circuits in LV systems. 2 Case study under consideration 2.1 Simulation model and parameters The analysed system and corresponding equivalent circuit is shown in Fig. 1. In this paper, the surges due to flashes to the structure (source S1), which is protected by a lightning protection system (LPS), are taken into account. LV supply TN system with two-cascaded limiting-type SPDs is considered as basic arrangement [7, 19, 20]. Fig. 1Open in figure viewerPowerPoint Schematic diagram of the analysed system and corresponding equivalent circuit Fig. 1 represents a simple example of flash to a LPS conductor near an electrical circuit loop, where SPD1 is the SPD installed in the distribution panel at the entrance of the building and bonding the phase conductor to the equipotential bonding bar; SPD2 is the SPD added inside the equipment or in the socket; l is the cable length between SPD1 and SPD2; L is the load to be protected with equivalent impedance ZL ; Z is the grounding system conventional impedance [7, 19, 20]. The simulation analyses are carried out by means of the transient software EMTP. Three shapes of lightning currents, namely representative of subsequent negative stroke (0.25/100 μs), first positive stroke (10/350 μs), and 8/20 μs standard waveform have been considered. The lightning impulse currents are simulated by an ideal bi-exponential current generator. Three types of ZnO MOVs used for actual SPDs, indicated as V275, V320, and V660, are selected to represent the limiting-type SPDs. Their main parameters are shown in Table 1 [8]. The MOVs are simulated in order to match the U–I characteristics derived by actual voltage–current measurements [7]. Table 1. Main parameters of the MOVs subject to analysis MOV type Diameter, mm Max. permitted AC voltage, V Capacity, kA)(8/20 μs) Max. clamping voltage, V V275 20 275 6.5 710 V320 20 320 6.5 810 V660 20 660 6.5 1650 The Polyvinyl chloride insulated cable with single core and no shielding sheath is selected as the connecting cable for the internal circuit [3]. The cables are simulated by means of transmission line model with electric parameters: R = 0.018 Ω/m, L = 1.51 μH/m, C = 0.606 nF/m, G = 0 s/m. Simulations with different cable lengths of 1, 5, 10, 20, and 40 m, respectively, between SPD1 and SPD2 are carried out for the purpose of evaluation of effective coordination distance [10]. Different loads with resistive loads of 1, 10, 100, and 1000 Ω, inductive loads of 35, 70, 100, and 1000 μH, and capacitive loads of 10, 200, 800, and 80,000 pF [2–4, 8] are taken into account to investigate the influence of different protected loads. For the convenience of describing the simulation conditions and considering the impact of different factors on the coordination effects, a 'reference simulation condition' with some assumptions are defined as: the subsequent negative impulse current is simulated by 0.25/100 μs waveform; the current amplitude is 3 kA; the cable length l between two SPDs is 10 m [10]; the protected load is represented by a resistance with 10 Ω [3, 4, 8]; the conventional grounding impedance of the structure is assumed Z = 10 Ω [7, 19, 20]. Only parameters which differ from that defined in 'reference simulation condition' will be indicated in specific cases of simulation. 2.2 Experiment conditions As for the experimental investigation with subsequent negative strokes, the generator for simulating subsequent negative impulses has been developed by High Voltage Laboratory of Shanghai Jiao Tong University. The generator could produce the steep currents with front time <1 μs and tail time around 3 μs (abbreviated as '1 μs steep current') [16, 21], with arrangement shown in Fig. 2. Fig. 2Open in figure viewerPowerPoint Impulse current generator for simulating subsequent negative impulses Two Pearson coils with output ratio of 0.01 and 0.1 are adopted to measure the currents flowing through MOVs. Two voltage dividers with ratio of 13.4 and 14.3 are used to measure the voltages on MOVs. A four-channel digital oscilloscope of TEK DPO3014 is adopted to record the test waveforms. 3 SPD dynamic characteristics under subsequent negative strokes The dynamic behaviour is described as the voltage across a MOV increases as the front time of the MOV current decreases, and the voltage reaches a peak before the current reaches its peak [18]. IEEE working group 3.4.11 has pointed out that the MOV residual voltage for a given impulse current magnitude is increased by ∼6–12% as the front time of the surge current is reduced from 8 to 1.3 μs [18]. This conclusion is based on the MOVs used for medium- or high-voltage surge arresters. Although it is confirmed that the residual voltages on MOVs used for LV SPDs would increase significantly under steep currents [6], the exact voltage increments on such MOVs have not been investigated thoroughly. Experiments on the MOVs used for LV SPDs are carried out to investigate the dynamic characteristics. The impulse currents of wave-front time 8, 4, and 2 V660–V320 V320–V275 k = 1.14 equally rated V660–V660 k = 1 V320–V320 V275–V275 low–high V275–V320 k = 0.88 V275–V660 k < 0.5 V320–V660 Considering the distance between cascaded SPDs is 5 m (i.e. the cascaded distance is changed to 5 m from the 'reference simulation condition'), the energy absorption in each SPD in the nine SPD combinations under 0.25/100 μs wave simulated subsequent negative impulse are calculated and shown in Fig. 4a. For comparison, the energy distributions under 3 kA, 8/20 μs standard lightning current and 3 kA, 10/350 μs waveform simulated first positive impulse current are simulated and shown in Figs. 4b and c, where E1 and E2 represent the energy absorbed in SPD1 and SPD2 in each combination, respectively. Fig. 4Open in figure viewerPowerPoint Energy absorption in each SPD in the nine SPD combinations under the three types of impulse currents a Under 0.25/100 μs impulse current b Under 8/20 μs impulse current c Under 10/350 μs impulse current It can be seen from Fig. 4a that for combinations V660–V275 and V660–V320, the energy absorbed in SPD2 is higher than that in SPD1 under 0.25/100 μs impulse currents. This indicates that 'high–low with k > 2' SPD combinations could not reach effective energy coordination under 'reference simulation condition'. The other SPD combinations, including 'high–low with k = 1.14' (V320–V275), 'equally rated', and 'low–high' SPD combinations, seem could realise effective energy coordination because most of the surge energy are absorbed by SPD1. Comparing the results in Figs. 4a and b, the ratio of energy absorption in SPD2 under 0.25/100 μs impulse current increase significantly compared with that under 8/20 μs impulse current. This may be due to the relative long tail of the 0.25/100 μs current, during which the SPD2 would divert the surge current. On the other hand, although the longer tail also exists in 10/350 μs current, the SPD2 in combinations V275–V660 and V320–V660 absorb zero energy under 10/350 μs current, as shown in Fig. 4c, while they could conduct and divert the surge currents under 0.25/100 μs current. Fig. 5 shows the current and voltage waveforms on the SPDs in combination V275–V660 under 10/350 and 0.25/100 μs impulse currents, respectively, where the 0.25/100 μs current adopts the positive impulse in order to compare with the results under positive impulse of 10/350 μs current. Fig. 5Open in figure viewerPowerPoint Current and voltage waveforms on the SPDs in combination V275–V660 under 10/350 and 0.25/100 μs impulse currents a Under 10/350 μs impulse current b Under 0.25/100 μs impulse current It can be seen from Fig. 5a that under 10/350 μs impulse current, the current I2 on SPD2 is 0, while the SPD2 could divert the surge current under 0.25/100 μs impulse current as shown in Fig. 5b. This may be due to the significantly changed dynamic characteristic of the cascaded SPDs and the significantly increased voltage drop on the connection cable between the SPDs under 0.25/100 μs current, which result in the greatly increase of the voltage on the SPD2 and makes the SPD2 conductive for protecting the load. Therefore, it can be concluded that under subsequent negative strokes, the significantly changed dynamic characteristic with increased voltages on the SPDs would help the 'low–high with k < 0.5' SPD combinations to realise effective energy coordination, while the SPD2 in such combinations do not conduct under relative slow-wave currents, such as 8/20 and 10/350 μs impulse currents. However, effective coordination for two-cascaded SPDs not only demands the proper energy distribution, but also needs the SPD2 to limit the overvoltage on the protected loads within its maximum withstand value. It should be noted that the overvoltage limit value for indoor LV equipment is 1.5 kV [22]. The subsequent negative impulse would lead to higher voltages on the MOVs, thus more attention should be paid on the residual voltages on SPD2. Considering the most serious conditions, where the magnitude of the impulse current is the maximum flow capacity (6.5 kA) of the SPDs (i.e. the current magnitude is changed to 6.5 kA from the 'reference simulation condition'), the residual voltages on SPD2 in the nine SPD combinations under 0.25/100, 8/20, and 10/350 μs current, respectively, are calculated and shown in Fig. 6. Fig. 6Open in figure viewerPowerPoint Residual voltages on SPD2 in the nine SPD combinations under the three types of impulse currents As shown in Fig. 6, the residual voltages on SPD2 in combination V660–V660 are higher than 1.5 kV under the three types of impulse currents, which means that this kind of combination is not acceptable for indoor LV systems. The residual voltages on SPD2 in combinations V275–V660 and V320–V660 are below the limit value under 8/20 μs current and 10/350 μs current, while they increase significantly and exceed 1.5 kV under 0.25/100 μs current. It is noticed that the SPD2 in these combinations is the varistor V660 with the highest limiting voltage among the three types of MOVs. This indicates that for SPD coordination under subsequent negative impulses, it is not reasonable to adopt the downstream MOVs with their maximum clamping voltage higher than 1.5 kV. Therefore, combinations V660–V660, V275–V660, and V320–V660 are not considered in the later analysis. In addition, it can be noticed that the simulation results between the V660–V275 and V660–V320 are similar under 0.25/100 μs current, and common characteristics also exist in the energy distribution between the three 'equally rated' SPD combinations, as shown in Fig. 4a. Therefore, for the simplification of the analysis, only four representative combinations are considered in later analysis, namely V660–V275, V320–V320, V320–V275, and V275–V320, representing the 'high–low with k > 2' combination, the 'equally rated' combination, the 'high–low with k = 1.14' combination, and the 'low–high with k = 0.88' combination, respectively. 4.2 SPD coordination under different separation distances The influences of the separation distance on the SPD coordination under subsequent negative impulses are investigated by changing the length of the connection cable l between the cascaded SPDs from the 'reference simulation condition' (1, 5, 10, 20 and 40 m, respectively). Meanwhile, simulations under 8/20 and 10/350 μs currents are also carried out under corresponding simulation conditions for comparison. The effective coordination demands that more energy of the surge currents should be absorbed by the SPD1 than by SPD2. Therefore, the ratio between the energy absorbed in SPD1 and SPD2, which is represented as E1/E2, may intuitively describe the energy distribution between the SPDs. When the ratio E1/E2 is larger than 1, the cascaded SPDs are regarded to realise effective coordination. However, the ratio E1/E2 should not be too large since it indicates that the SPD2 receives almost zero energy. In other words, the SPD2 is nearly redundant. Therefore, the ratio E1/E2 of effective coordination for cascaded SPDs should be within a certain range, which is set as 1–10 in this paper. Of course, for different MOV types, the effective ratio range may be different and the range can be adjusted. Fig. 7 shows the calculated ratio E1/E2 under different separation distances considering the four SPD combinations. Fig. 7Open in figure viewerPowerPoint Energy absorption ratio between the cascaded SPDs under different separation distances a V660–V275 b V320–V275 c V320–V320 d V275–V320 Some results could be obtained from Fig. 7 : Fig. 7a shows that when the separation distance between V660 and V275 is shorter than 8 m, the ratio E1/E2 under 0.25/100 μs current is lower than 1. This indicates that under subsequent negative strokes, the 'high–low SPD with k > 2' combination could not realise effective energy coordination when the cascaded distance is short. Similar result has also been obtained in Fig. 4a. Therefore, for the 'high–low with k > 2' SPD combination, the distance between cascaded SPDs should be relatively long or some extra decoupling elements should be added to realise effective coordination. As shown in Figs. 7b–d, the ratios E1/E2 calculated in combinations V320–V320, V275–V320, and V320–V275 are always within the defined criteria range under 0.25/100 μs current as the separation distance changes. That is to say, the SPD combinations with k between 0.88 and 1.14 could realise effective energy coordination in large range of separation distance with 1–40 m under subsequent negative strokes. As shown in all the subfigures in Fig. 7, the energy absorption ratio E1/E2 increases with the increase of the separation distance. In other words, with the increase of the separation distance, the cascaded SPDs are more easy to reach effective energy coordination no matter under what kind of impulse currents. However, the effective coordination distance should be limited since too long distance may result in the excess of ratio E1/E2 over the criteria range, especially for the SPDs under 8/20 μs current. With the increase of the separation distance, the ratio E1/E2 for SPD combinations under 8/20 μs current is the largest and increase most significantly, compared with that under 0.25/100 and 10/350 μs currents. In addition, the effective coordination distance for V660–V275 is longer than that for V320–V275 under 8/20 μs current, as shown in Figs. 7a and b. It means that better coordination could be achieved when the difference of the clamping voltages between cascaded SPDs is relatively large for 'high–low' SPD combinations under 8/20 μs impulse current. The combination V320–V275 seems to have the best behaviour among all the SPD combinations since its distance range for effective coordination under the three impulse currents is the largest. The effective coordination distance for the combination is 1 m to more than 40 m under 0.25/100 μs current, 5 m to more than 40 m under 10/350 μs current, and 1–16 m under 8/20 μs current. 4.3 SPD coordination with different protected loads To investigate the influence of the protected loads on SPD coordination under subsequent negative strokes, simulations considering different load types with various values are taken into account during simulations, namely: resistive loads (1, 10, 100, and 1000 Ω), inductive loads (35, 70, 100, and 1000 μH), and capacitive loads (10, 200, 800, and 80,000 pF) [2–4, 8]. The simulations under impulse currents of 8/20 and 10/350 μs are also carried out for comparison. Except for the varying protected loads, the other parameters are kept unchanged in the 'reference simulation condition'. Fig. 8 shows the energy absorption ratio E1/E2 between the cascaded SPDs with different protected loads. Some values for E1/E2 are not presented, such as the combination V275–V320 with inductive loads 35, 70, and 100 μH under 8/20 and 10/350 μs currents in Fig. 8d. These points are omitted because the ratio E1/E2 is quite large (larger than 25), indicating that the SPD2 absorbs little energy and the cascaded SPDs are not effective coordination. Fig. 8Open in figure viewerPowerPoint Energy absorption ratio between the cascaded SPDs under different protected loads a V660–V275 b V320–V275 c V320–V320 d V275–V320 Some results could be obtained from Fig. 8 : As shown in all the subfigures in Fig. 8, the capacitive loads nearly have no influence on the energy distribution in the cascaded SPDs. For resistive loads, when the resistance is larger than 10 Ω, the influence on the energy distribution is quite small. With the increase of the inductive loads, the ratio E1/E2 decreases gradually under 0.25/100 μs impulse current, while it decreases sharply under 8/20 and 10/350 μs impulse currents for the SPD combinations with k between 0.88 and 1.14, as shown in Figs. 8b–d. For these four SPD combinations, the E1/E2 is always within the effective range (1–10) under 0.25/100 μs impulse current though the protected load changes. Similar conclusion could also be made for the impact of the separation distance, as shown in Figs. 7b–d. Therefore, it can be concluded that effective SPD coordination under subsequent negative strokes can be achieved in large ranges of separation distances and protected loads. In addition, considering the various separation distances and protected loads, the SPD combinations with k between 0.88 and 1.14 are suggested for SPD coordination under subsequent negative strokes. As shown in Fig. 8b, the V320–V275 still has the best behaviour under varying loads, because the combination could achieve effective coordination with most of the considered load conditions. In addition, this combination also has wide distance range for effective coordination under varying separation distances as mentioned in Section 4.2. Therefore, among the nine SPD combinations, the 'high–low with k = 1.14' combination are believed to have the best coordination performance considering the common lightning currents in LV systems. 5 Experimental investigation To validate the simulation results, experimental investigation on the SPD coordination under '1 μs steep current' simulated subsequent lightning stroke is carried out. The actual SPD circuits in the aforementioned analyses are constructed during experiment. The measured current and voltage waveforms on the SPD1 and SPD2 in combination V320–V275 are shown in Fig. 9a, where the subsequent lightning stroke is simulated by a 3 kA steep current with front time of 0.89 μs, the separation distance between the cascaded SPDs is 10 m, and the protected load is represented by a resistance of 10 Ω. Fig. 9Open in figure viewerPowerPoint Experimental measured results a Measured current and voltage waveforms on the SPDs in combination V320–V275 b Measured energy absorption in each SPD under different separation distances As shown in Fig. 9a, the voltage on SPD1 reaches peak value before the current on SPD1, which shows the dynamic behaviour on SPD1 under steep current. The experimentally measured peak current and voltage on SPD1 are 2.81 kA and 1.38 kV, respectively, which are close to the 2.89 kA and 1.34 kV calculated value in simulation, though the waveform of '1 μs steep current' is somehow different from the waveform of standard subsequent negative impulse. The experimentally measured energy absorbed in SPD1 and SPD2 is 32.4 and 1.2 J, respectively. Due to the short tail of '1 μs steep current' used in the experiment, the measured energy absorbed in cascaded SPDs, especially in SPD2, is much less than that calculated in the simulation. To experimentally investigate the influences of separation distance on SPD coordination under '1 μs steep current', experiments are carried out on the combinations V660–V275, V320–V275, and V275–V320. The amplitude of the '1 μs steep current' is 3 kA and the protected load is 10 Ω. The energy absorption in cascaded SPDs under different separation distances is shown in Fig. 9b, where E1 (320–275) and E2 (320–275), respectively, represent the energy absorbed in SPD1 and SPD2 of combination V320–V275, and the other marks have similar meaning. Fig. 9b shows that under '1 μs steep current', the energies absorbed in SPD1 increase with the increase of the separation distance, while the energies absorbed in SPD2 decrease. This verifies the simulation conclusion that the energy absorption ratio E1/E2 increase with the increase of the distance between the cascaded SPDs. In addition, when the separation distance is 2' SPD combination, the separation distance for effective coordination should be relatively long. To investigate the influences of different protected loads on SPD coordination under '1 μs steep current', experiments are carried out on the combination V320–V275. The peak current and energy absorbed in the SPD1 and SPD2 are measured, respectively, as shown in Table 3. Table 3. Measured peak current and energy absorption in the SPDs for combination V320–V275 under different protected loads SPD cascaded combination Resistive load, Ω Inductive load, µH Capacitive load, pF 10 1000 35 70 200 800 V320–V275 I1, A 2810.0 2812.6 2811.4 2811.8 2811.2 2811.0 I2, A 134.6 173.3 100.2 114.6 172.9 173.0 E1, J 32.4 33.7 20.1 23.2 32.8 32.9 E2, J 1.2 1.2 0.2 0.5 1.3 1.3 It can be seen from Table 3 that the impact of inductive loads on the energy distribution for cascaded SPDs is the most significant among the three types of loads. When the inductance is increased from 35 to 70 µH, the ratio E1/E2 almost decreases by half, which decreases from 100.5 to 46.4. Similar conclusion could also be found in the simulation that the ratio E1/E2 decreases from 4.8 to 2.5, as shown with the line marked with triangle in Fig. 8b. The influence of resistive loads on the energy distribution is quite small when the resistance is increased from 10 to 1000 Ω. Moreover, the change of the capacitive loads has no impact on the SPD coordination. These conclusions all correspond with that obtained in the simulation. It should be noted that due to the difference of the simulation and experimental waveforms for simulating the subsequent negative strokes, the experimentally measured values may have some divergence with the simulation results. However, the variation regularities of the measured absorbed energy in cascaded SPDs conform with the simulation results, which validate the simulation conclusions to some extent. 6 Conclusions In this paper, experiments are performed to investigate the dynamic characteristics of MOVs under '1 μs steep current' simulated subsequent strokes. Based on the dynamic characteristics of MOVs, simulation analysis is carried out to investigate the coordination of two-cascaded limiting-type SPDs under 0.25/100 μs current simulated subsequent negative strokes. Simulations under normally used 8/20 μs standard lightning current and 10/350 μs wave simulated first positive stroke are also conducted for comparison. Moreover, the influence factors on the SPD coordination are investigated in the simulation and some of the conclusions are validated by experimentally measured results. The following main conclusions can be drawn: The residual voltages on MOVs used for LV SPDs increase about 6.8–18.7% under '1 μs steep current' simulated subsequent strokes, compared with that under 8/20 μs impulse current. With the impulse currents peaks increase, the increments of the residual voltages under '1 μs steep current' become larger. For MOVs with lower clamping voltage, the increase of the residual voltages is more significant. However, when the wave-front time of the impulse currents decrease from 8 to 4 μs, the increment of the residual voltages is small. Thanks to the significantly changed dynamic characteristics of MOVs under subsequent negative impulses, the 'low–high with k < 0.5' SPD combinations could realise effective energy coordination, while the SPD2 in such combinations do not conduct under relative slow-wave currents, such as 8/20 and 10/350 μs impulse currents. However, the increased voltages on SPDs may also result in some of the SPD combinations unacceptable under subsequent negative impulses, because the overvoltage limit value on the protected load is exceeded. The SPD coordination under 8/20 and 10/350 μs impulse currents is strongly influenced by the cascaded distance and the protected loads, while the effective SPD coordination under 0.25/100 μs impulse current can be achieved in large ranges of the separation distances and protected loads. The SPD combinations with clamping voltage of upstream SPD approximately equal to the downstream SPD (such as k between 0.88 and 1.14) are suggested for SPD coordination under subsequent negative strokes, considering the varying separation distances and protected loads. Among the different SPD combinations, the V320–V275 could realise effective coordination with most of the considered simulation conditions as well as common lightning currents including first positive impulse, 8/20 μs current, and subsequent negative impulse. Therefore, the SPD combinations with clamping voltage of upstream SPD a little higher than the downstream SPD are suggested for lightning protection in LV systems. 7 References 1IEC Std. 62305-4: 'Protection against lightning – part 4: electrical and electronic systems within structures', 2010 2Metwally, I.A., Heidler, F.H.: 'Enhancement of the SPD residual voltage at apparatus terminals in low-voltage power systems', IEEE Trans. Power Deliv., 2007, 22, (4), pp. 2207– 2213 (doi: 10.1109/TPWRD.2007.905563) 3He, J., Yuan, Z., Wang, S. et al.: 'Effective protection distances of low-voltage SPD with different voltage protection levels', IEEE Trans. Power Deliv., 2010, 25, (1), pp. 187– 195 (doi: 10.1109/TPWRD.2009.2035297) 4He, J., Yuan, Z., Xu, J. et al.: 'Evaluation of the effective protection distance of low-voltage SPD to equipment', IEEE Trans. Power Deliv., 2005, 20, (1), pp. 123– 130 (doi: 10.1109/TPWRD.2004.835254) 5IEC Std. 62305-1: 'Protection against lightning – part 1: general principles', 2010 6Schmidt, W., Meppelink, J., Richter, B. et al.: 'Behaviour of MO-surge-arrester blocks to fast transients', IEEE Trans. Power Deliv., 1989, 4, (1), pp. 292– 300 (doi: 10.1109/61.19216) 7Kisielewicz, T., Piparo, G.B.L., Fiamingo, F. et al.: 'Factors affecting selection, installation and coordination of surge protective devices for low voltage systems', Electr. Power Syst. Res., 2014, 113, pp. 220– 227 (doi: 10.1016/j.epsr.2014.02.038) 8He, Z., Du, Y.: 'Influence of different factors on coordination of two cascaded SPDs', Electr. Power Syst. Res., 2016, 139, pp. 139– 145 (doi: 10.1016/j.epsr.2015.11.029) 9Zhang, D., Fu, Z., Zhao, G. et al.: 'Protection distance of surge protective devices in low voltage distribution systems', J. Shanghai Jiaotong Univ.(Sci.), 2008, 13, pp. 477– 481 (doi: 10.1007/s12204-008-0477-x) 10Lai, J.S., Martzloff, F.G.D.: 'Coordinating cascaded surge protection devices: high–low versus low–high', IEEE Trans. Ind. Appl., 1993, 29, (4), pp. 680– 687 (doi: 10.1109/28.231979) 11Martzloff, F.D.: 'Coordination of surge protectors in low-voltage AC power circuits', IEEE Trans. Power Appar. Syst., 1980, 99, (1), pp. 129– 133 (doi: 10.1109/TPAS.1980.319618) 12Zhang, D., Fu, Z., Zhao, G. et al.: 'The coordination mechanism of cascaded SPDs in low voltage distribution systems', Trans. China Electrotech. Soc., 2009, 24, (7), pp. 145– 149 (in Chinese) 13Shin, H.K., Kim, D.S., Chung, Y.K. et al.: 'Energy coordination of ZnO varistor based SPDs in surge current due to direct lightning flashes'. Proc. 32th Int. Conf. on Lightning Protection (ICLP), Shanghai, China, 13–17 October 2014, pp. 136– 140 14Clifford, D.W., Crouch, K.E., Schulte, E.H.: 'Lightning simulation and testing', IEEE Trans. Electromagn. Compat., 1982, 24, (2), pp. 209– 224 (doi: 10.1109/TEMC.1982.304032) 15Modrusan, M., Gockenbach, E., Lutz, M.: 'Simulation of lightning stroke parameters in a laboratory'. Proc. 20th Int. Conf. on Lightning Protection (ICLP), Interlaken, Switzerland, September 1990, pp. 56– 61 16He, Y., Fu, Z., Jiang, A.: 'On the simulation of lightning application in lightning arrester and material ablation tests'. Proc. Int. Conf. on 2015 Asia-Pacific Symp. on Electromagnetic Compatibility, Taiwan, China, 25–29 May 2015, pp. 428– 431 17Li, C., Jianguo, W., Mi, Z. et al.: 'Observation of induced voltage on power lines under a nearby natural lightning', High Volt. Eng., 2013, 39, (4), pp. 1000– 1008 (in Chinese) 18 IEEE Working Group 3.4.11: 'Modeling of metal oxide surge arresters', IEEE Trans. Power Deliv., 1992, 7, (1), pp. 302– 309 (doi: 10.1109/61.108922) 19Kisielewicz, T., Mazzetti, C., Piparo, G.B.L. et al.: 'Stress to surge protective devices system due to direct flashes to low voltage lines', Electr. Power Syst. Res., 2015, 129, pp. 44– 50 (doi: 10.1016/j.epsr.2015.07.010) 20Kisielewicz, T., Piparo, G.B.L., Mazzetti, C.: 'Surge protective devices efficiency for apparatus protection in front of direct flashes to overhead low voltage lines', Electr. Power Syst. Res., 2016, 134, pp. 88– 96 (doi: 10.1016/j.epsr.2016.01.008) 21Jiang, A., Fu, Z., Sun, W. et al.: 'Experimental and analytical investigation of the effects of single and multiple simulated lightning impulse currents on metal-oxide arrester blocks', Electr. Power Syst. Res., 2016, 139, pp. 146– 152 (doi: 10.1016/j.epsr.2015.11.017) 22IEC Std. 60664-1: 'Insulation coordination for equipment within low-voltage systems – part 1: principles, requirements and tests', 2007 Citing Literature Volume10, Issue16December 2016Pages 4197-4204 FiguresReferencesRelatedInformation
Referência(s)