Corona Points Discharge Current Measurement on Atmospheric Electric Field

Transcription

1 Corona Points Discharge Current Measurement on Atmospheric Electric Field Luis Forero, Juan Chavarro, Rafael Valenzuela, Francisco Román * Universidad Nacional de Colombia, Ciudad Universitaria, Edificio 411, Bogotá D.C., Colombia * fjromanc@unal.edu.co Abstract: Under natural fair weather conditions, the electric field amplitude is positive -the earth surface is negative-, with circa 100 V/m. This natural electric field polarity changes with the presence of thunderstorm clouds. 5 to 10 kv could be the background electric field magnitude at ground level, under bad weather - thunderstorm- conditions. Its polarity is usually negative -the earth is positive-, depending on the cloud electrical charge polarity. This natural electric field can be amplified by low curvature radius objects such as needles, cactus like-electrodes or Franklin rods. The electric field amplification at such low curvature objects produces the so called point discharge currents. In the present research work we investigate the point discharge currents produced by cactus-like objects. These current measurements are important to understand the operation and protection principles given by such devices under bad weather natural atmospherical conditions. Measurements were performed at 2600 m above sea level, in Bogotá, Colombia (4 36' N, 74 5' W). In this research work we also investigate the polarity, magnitude and shape of the Corona discharge currents produced in two different electrode types. The influence of some atmospherical conditions on the point discharge current variation, such as atmospheric pressure, relative humidity and temperature was also investigated. Keywords: Natural electric field, Franklin rods, current measurements, natural atmospheric conditions. INTRODUCTION In the present research work two prototypes of electrodes were developed to study the possibility to extract usable energy from the electric field, as it was suggested by Román [1]. The studied electrodes were a cactus-like electrode called C-Electrode and a plate electrode with needles called P-Electrode. Designs were theoretically supported by the works [2-5]. Additionally a small weather station was designed and constructed. The registered variables were: environmental temperature, atmospheric pressure, relative humidity and C and P- Electrode currents. The sampling rate was 3 s, enough to characterize the point discharge currents DC current component. Initial measurements show a close relationship between some of the measured atmospherical variables and point discharge currents. ELECTRODE SYSTEM C-Electrode The geometry of C-Electrode is based on previous laboratory test performed by [4] and [5]. Some characteristics of C-Electrode are shown in Fig.1. Fig.1 Design planes and assembly picture of C- Electrode, (a) Top view, notice that needles are turned π/4 rad, (b) needles detail (c), Front view, notice the needles inclination on the z axis, (d) and (e) assembly picture. Notice the rain water protection. C-Electrode consists on a bronze rod of 14,3 mm diameter and 2,3 m length, finished on a needle tip, with 100 steel needles placed around as can be seen in Fig.1. 4 needles were placed around the electrode each 0,04 m, turned π/4 rad on the horizontal plane and inclined π/4 rad in z axis, as it is shown in Fig.1 (a) and (c). Additionally the C-Electrode assembly has a water rain protection (see Fig.1 (e)), in order to avoiding possible unexpected electrode-support contacts (the support is electrically connected to earth). P-Electrode The P-Electrode is a plate with 82 needles, with the same dimensions that those on C-Electrode. The minimum distance between needles is 100 mm to avoid shielding effects between them [5]. Design and assembly are shown in Fig.2.

2 1,0 I corona [ma] 0,8 0,6 0,4 100 needles 50 needles 25 needles (a) (b) Fig.2 (a) P-Electrode design. Notice 4 orthogonal perforations for the supports and another central perforation for the measurement signal, (b) assembly picture, P-Electrode is placed 0,80 m over the building ceiling level (10,5 m). LABORATORY TESTS C- and P- Electrodes laboratory tests were performed in order to characterize their corona current. C-Electrode In this test C-Electrode was placed (without support, as it is shown in Fig.3) within a cylinder, with dimensions φ = 0,50 m and l = 1 m, with purpose to simulate quasihomogenous electric field conditions in a coaxial arrangement. The number of needles were 100, 50 and 25, to evaluate the shielding effect between them. Test was performance for a background electric field varying within 13,2 kv/m and 140 kv/m. Background electric field is defined here as applied voltage divided by cylinder interelectrode distance. Fig.3 shows the schematic circuit and the arrangement test. Fig.3 C-Electrode characterization. Test circuit: (1) High Voltage source, from 3,2 kv to 34 kv, (2) C- Electrode in coaxial cylinder arrangement, (3) measuring system composed of a measuring resistance (10 kω), a coaxial cable RG 58 and a oscilloscope. As it is indicated in Fig.4, there is no a linear relation between the number of needles and the measured corona current, for the same background electric field: The current measured with 100 needles is neither 4 times the current measured with 25 needles nor twice as large as the current measured with 50 needles. This could indicate a shielding effect between needles. However, in all cases the largest current measured was with 100 needles. For this reason the assembly with 100 needles was used in the here described point discharge current experiment. 0,2 0, Backgraund Electric Field [kv/m] Fig.4 C-Electrode s corona current as a function of the background electric field and the number of needles. Notice that in all cases the electrode with 100 needles produced the largest corona current. P-Electrode Two goals were follow in present experiment: The first one was to measure the current produced by the complete assembly with 82 needles and the second was to test if the variation of the series connected resistor between 100 kω and 1 MΩ would affect or not the current measuring system. Fig.5 P-Electrode characterization. Test circuit: (1) High Voltage source, from -15 kv to -45 kv, (2) P- Electrode assembly, (3) Measuring system, composed by: A variable resistance (between 100 kω and 1 MΩ), a coaxial cable RG 58 and a multimeter. The distance between the high voltage plate electrode and the P-Electrode was 1,35 m. I corona [ua] Background Electric Field [kv/m] Rs = 1M Rs = 100 k Fig.6 P-Electrode s corona currents as a function of the background electric field. Notice the independence of the measured corona current of the series resistance. As it can be seen in Fig.6, test results have shown that the measured corona current was independent of the measuring series resistance. 100 kω and 1 MΩ were tested.

3 SPACE CONSIDERATIONS The electrodes were located at 10,5 m over floor level on the building 411 of the National University of Colombia in Bogotá D.C. (4 36' N, 74 5' W) at 2600 meters over sea level. Fig.9 Electric filed simulation for P-Electrode, on the top of the university building, frontal view. Notice the background electric field intensification over P- Electrode and its edges. MEASURING SYSTEM Fig.7 Schematic diagram of corona electrodes and the weather station. All variables are stored on a PC. Before the corona electrodes were installed, a finite element simulation with the 3D software ANSYS was performed. The purpose was to estimate the atmospherical electric field amplification due the building and the electrodes geometry. Results have shown that the background electric field was 4 times amplified due to the building itself. A simulation with the electrodes placed on their final position (without considering the needles), has show that the background electric field was circa 130 times amplified around C- Electrode (see Fig.8), while for P-Electrode, the amplification was just 15 times (see Fig.9). All simulations were performed with a background electric field. The measuring system registered following variables: Environmental temperature, atmospherical pressure and relative humidity, the corona currents from C- and P- electrodes (Ic = Corona current from C-Electrode, Ip = Corona current from P-Electrode). These variables were measured continuously every 3 s. The measuring system was divided in two parts: The first one considers the sensors and the signal conditioning and the second one the signal register. The data was registered on the PIC16F877-4 card (8 bits, 5 channels), through the interface port MAX232 (RS232 port driver). The software support was made in Visual Basic and the data was recorded on Microsoft Excel. The card input data was filtered to reject noise with a 470 pf capacitor. Environmental temperature measurement Environmental temperature was measured with the sensor LM335, which has a 3 μs response time (the device spends 15 s to arrive to the true value). Under atmospherical conditions (airflows and humidity variations) together with the register system, the sensor typical error is 1,3 %. Equation (1) is the calibration result of the environmental temperature T measuring system: V s = 10 mv T -1 [K] [V] (1) Where: Vs is the voltage measured as a function of environmental temperature. The schematic circuit implemented to measure environmental temperature measurements is shown in Fig.10. Fig.8 Numerical simulation of the university building with C-Electrode, placed on its corner. Notice the background electric field intensification on the upper part of the rod. Fig.10 Environmental temperature measuring system schematic circuit. The sensor is exposed to environmental conditions and the measured data is recorded on register channel 1.

4 Atmospherical Pressure Measurement Atmospherical pressure was measured with the sensor MPX115A, which has a 1 ms response time, and including the measuring system its typical error is 1,5%. Equation (2) is the calibration result of the atmospherical pressure P measuring system: V s = 0,045 P[kPa]-0,484 [V] (2) Where: Vs is the voltage measured as a function of atmospherical pressure P. The schematic circuit implemented to measure atmospheric pressure is shown in Fig.11. Fig.11 Atmospherical pressure measuring system schematic circuit. The sensor is exposed to environmental conditions and the measured data is recorded on register channel 2. Relative Humidity measurement Relative humidity was measured with the sensor HIH3610, which spends 15 s to arrive to the true value, and including the measuring system its typical error is 3%. Equation (3) is the relative humidity HR calibration result: V s = 0,032 HR [%] +0,814 [V] (3) Where: Vs is the voltage measured as a function of HR. The schematic circuit implemented to measure relative humidity is shown in Fig.12. implemented for the corona currents measurement is shown in Fig.13. Fig.13 Corona currents measuring system schematic circuit. In the figure: (1) Resistance arrangement: 100, 20, 10 and 1 MΩ, 1 μf capacitor as high frequency filter and a gas discharge -300 V-, (2) Amplifier G = 10 (Z in = 1TΩ 12 pf), (3) 1 V adder circuit, (4) register system PIC16F877-port RS232-PC-user. OBSERVATIONS AND MEASUREMENTS From the performed measurements it was evident that there exist a close correlation between the atmospheric variables and the corona currents behavior. C-Electrode is noticeable, as it can be seen in Figs The results have shown that C-Electrode corona current precedes 20 and 30 minutes important environmental temperature (see Fig.14) and atmospherical pressurechanges (see Fig.15). The recorded values were observed during important relative humidity variations (20 % - 100%). It is possible that the mentioned changes are due to the proximity of a electrically charged cloud, which is a risk for the some human activities. In other words, C-Electrode is susceptible to be calibrate and to work as early warning system. Fig.12 Relative humidity measuring systems schematic circuit. The sensor is exposed to the environmental conditions and its response is recorded by register system channel 3. Current measurement Current was measured as a voltage drop on a high value resistance. The circuit implemented has a commonmode impedance and rejection equal to 1 TΩ 12 pf and 100 db, respectively, and gain (G) equal to 10. In addition a resistance arrangement was constructed in order to change the measuring scale. The current measuring circuit is shown in Fig.13 (same for both electrodes). Additionally the circuit has an adder between amplifier and register system, in order to compensate the own DC voltage of the registry system. This voltage level was compensated during the performed measuring system calibration procedure. The schematic circuit Fig.14 Measured corona currents for C-Electrode (Ic curve) and P-Electrode (Ip curve), environmental temperature (T curve) measured on Notice that almost 30 minutes before of the first important temperature increase, Ic increased to important values. The same behaviour was observed two times later.

5 CONCLUSIONS (1) Two different corona electrodes were developed: A cactus-like electrode and a plate with needles. Both electrode types were tested and corona currents were measured. Higher corona current were obtained with C- Electrode than with P-Electrode. Fig.15 Measured corona currents for C-Electrode (Ic curve) and P-Electrode (Ip curve), atmospherical pressure (Patm curve) measured on As it is shown in Fig.14 current Ic precedes the first increase of atmospherical pressure almost 30 minutes. (2). In an initial approach, there are a close relation between corona currents and some atmospheric parameters, such as pressure and temperature. The correlation between these parameters will be studied in future studies. ACKNOWLEDGEMENTS Authors would like to thank Universidad Nacional de Colombia and COLCIENCIAS for supporting present investigation. REFERENCES Fig.16 Measured corona currents for C-Electrode (Ic curve) and P-Electrode (Ip curve), relative humidity (RH curve) measured on Notice that there is no a clear correlation between corona currents and relative humidity. [1] Román, F. Effects of Electric Field Impulses Produced by Electrically Floating Electrodes on the Corona Space Charge Generation and the Breakdown Voltage of complex Gaps. Acta Universitatis Upsaliensis. Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology pp. Uppsala. ISBN Sweden, [2] Arévalo, L., Díaz, O. Estudio de la física del efecto corona para diferentes configuraciones de electrodos en un arreglo de cilindros coaxiales. Department of Electrical and Electronic Engineering, Universidad Nacional de Colombia. Bogotá D.C., Agosto [3] Gómez, C., López, H. Caracterización de la corriente Corona acumulativa producida por Electrodos hiperboloides en una Configuración de cilindros coaxiales. Department of Electrical and Electronic Engineering. Universidad Nacional de Colombia. Bogotá D.C., Fig.17 Measured corona currents for C-Electrode (Ic curve) and P-Electrode (Ip curve), environmental temperature (T curve), atmospherical pressure (Patm curve) and relative humidity (RH curve), measured on Notice a possible correlation between the three atmospherical variables: Temperature, Pressure point discharge current in the C-Electrode. The collected data in both dry- and rainy-days have shown that by strong rain, corona currents decrease. Additionally when cloud-to-ground lightning appear, fast polarity changes have been observed. [4] Aubrecht, L., Pekarek, S., Koller, J., Stanek, Z., Multineedle-to-plane Corona Discharge, Department of Physics Faculty of Electrical Engineering, Czech Technical University. Prague, [5] Osorio, G., Rosero, R. Disminución del campo eléctrico de inicio del efecto corona mediante la utilización de distintas configuraciones de electrodos basados en las formas de la naturaleza, Department of Electrical and Electronic Engineering, Universidad Nacional de Colombia, Bogotá D.C., 2003.