C4-301 EXPERIMENTAL EVALUATION OF TRANSFERRED SURGES IN MV TRANSFORMERS FROM HV/LV

Transcription

1 21, rue d'artois, F Paris C4-301 Session 2004 CIGRÉ EXPERIMENTAL EVALUATION OF TRANSFERRED SURGES IN MV TRANSFORMERS FROM HV/LV HERMOSO B.*, AGUADO M., SENOSIAIN V., MARTÍNEZ CID P.M. UNIVERSIDAD PÚBLICA DE NAVARRA IBERDROLA (Spain) In MV lines lightning produces overvoltages by direct action on line conductors or by induced overvoltages. In both cases, if the surge voltage resultant is lower than the line BIL, a surge wave is produced travels by the line arriving, in many cases, to the MV/LV transformers. These transformers allow the path of the waves from the HV to the LV side by different coupling mechanisms (capacitive, oscillatory, electromagnetic), finally the LV installations are the victims for this kind of overvoltages. According with the stard EN , Annexe E is easier to obtain the value of the transferred overvoltage in transformers, using a recurrent surge generator. From the results obtained in the tests describe above considering the CIGRE model for coupling mechanism transformer simulation it has been formulated a simple transformer model that it allows knowing the transferred surge at low side. Keywords: Lightning - Insulation Coordination-Overvoltages-Transformers 1. HV-LV TRANSFERRED OVERVOLTAGES The overvoltages transmission in a distribution transformer from the HV to LV side, due to the lightning, is a main point for the transformer design, for the customer protection. The value of the result transmitted overvoltages is given for the shape of the incident wave (fast front or low front) by the transformer structure (winding shapes, connections...) whose is resumed in capacitive inductive values. The different transmission modalities are indicated in the EN , Annex E [1]: - Electrostatic or capacitive transmission. This component appears the first his frequency is into the range of MHz - Oscillatory transmission. This component is due to the naturals oscillations generated by the capacitance to ground the inductance of transformers - Electromagnetic transmission. This component, also called "normal transmission" is related with the transformation ratio, the inductance the burden of transformer * hermoso@unavarra.es

2 The capacitive component appears the first; his frequency is in the range of MHz. its value is related with the windings capacities the capacity between the windings ground. After the capacitive component, appears the component transmitted by inductive way, whose value depend of voltage distribution in the winding. The transformers work as in normal condition. The oscillatory component is dimmed overlapped to the electromagnetic component. Normally is weak is not very important except when it appears resonance phenomena. Although Stard EN Appendix E gives the expressions to evaluate the different transmission components, this one recognizes the difficulty to applied them, due to the several factors that may appear must been take into account. It recommends, as a more practical method to measure the response with a recurrent generator. 2. TESTS WITH REGURRENT GENERATOR The tests have been achieved with a recurrent surge generator, output 100 volts wave 1,2/50µs (Fig. 2), 35 Hz, applied on different distribution transformers (power, ratio, group connection, insulation). See data in table 1. Table 1 Transformers characteristics KVA HV/LV (kv) Group Insulation 25 0,945/0,4-0,230 Yy0 oil 25 13,8/0,38 Dyn11 oil 25 13,2/0,38 Dyn11 oil 25 13,8/0,4/0,132 Dz0-Dyn11 oil 50 13,6/0,4 Yzn11 oil 50 13,2/0,23 Dyn11 oil ,2-21/0,42-0,24 Yzn11 oil ,8/0,42-0,24 Dyn11 silicone ,86/0,42-0,24 Dyn11 oil ,2/0,42 Dyn11 dry ,86/0,42-0,24 Dyn11 oil ,86/0,42-0,24 Dyn11 silicone /0,4 Ynd11 oil ,86/0,42 Dyn11 silicone /3,19 Dyn11 oil The voltage has been applied in the HV side between two phases, the oscillatory responses obtained in the LV side has been registered as seen in figure 1 figure 2 (100 kva-630 kva) The values (frequency amplitude as p.u. of voltage applied) are showed in table 2. 2

3 Table 2 Test result KVA input/ouput MHz p.u. KVA input/ouput MHz p.u. 25 AB-ca 0,164 0, AB-ca 0,476 0,106 BC-ca 0,172 0,0696 BC-ca 0,49 0,186 CA-ca 0,175 0,0777 CA-ca 0,463 0, AB-ca 0,108 0, AB-ca 0,678 0,0367 BC-ca 0,103 0,0411 BC-ca 0,69 0,025 CA-ca 0,164 0,0345 CA-ca 0,667 0, AB-ca 0,125 0, AB-ca 0,54 0,0847 BC-ca 0,109 0,0411 BC-ca 0,521 0,0851 CA-ca 0,189 0,0345 CA-ca 0,563 0, AB-ca 0,179 0, AB-ca 0,548 0,0596 BC-ca 0,172 0,0505 BC-ca 0,667 0,0176 CA-ca 0,179 0,0777 CA-ca 0,536 0, AB-ca 0,152 0, AB-ca 0,58 0,0827 BC-ca 0,144 0,0077 BC-ca 0,556 0,0303 CA-ca 0,135 0,03 CA-ca 0,606 0, AB-ca 0,458 0, AB-ca 0,667 0,0784 BC-ca 0,465 0,181 BC-ca 0,6 0,157 CA-ca 0,444 0,184 CA-ca 0,595 0, AB-ca 0,5 0,045 BC-ca 0,5 0,035 CA-ca 0,5 0,045 Input Wave 1,2/50 µs V c = 100 V Oscillatory wave f=0,179 MHz V c =783mV=0,783V 0,783/10=0,0783 f=0,172mhz V c =505mV=0,505V 0,505/10=0,0505 f=0,179 MHz V c =777mV=0,777V 0,777/10=0,0777 Fig. 1. Oscillatory response 100kVA; 14,2-21/0,42-0,242 kv; Yzn11; Oil 3

4 f=0,667 MHz V c =784mV=0,784V 0,784/10=0,0784 f=0,6 MHz V c =157mV=0,157V 0,157/10=0,157 f=0,595 MHz V c =849mV=0,849V 0,849/10=0,0849 Fig. 2. Oscillatory response 630 kva; 13,86/0,42 kv; Dyn11; Silicone 3. CORRELATIONS In figure 3 is plotted the transmission coefficient in figure 4 the frequency of the reference transformers sample versus the rated transformers power. correlations. As it is observed in figure 3 it does not exist correlations for all the sample, but it is possible to define a transmission coefficient (average value) around 0,1 p.u., except for one small transformer (25 kva). In figure 4 it may observe two main response frequencies, one around the 0,6 MHz for a wide range of transformers, other one around 0,2 MHz for a small group of transformers with rated power between kva p.u. 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0 transmission coefficient y = 0,0992x -0,0199 R 2 = 0, kva Fig. 3. Transmission coefficient- kva MHz frecuencia y = 0,0762x 0,3122 R 2 = 0,5523 0,8 0,6 0,4 0, kva Fig. 4. Frequency response 4

5 4. SIMPLE TRANSFORMER MODEL The model developed is based on Group III of CIGRE (figure 5) but with some modifications [2]. Fig. 5. CIGRE model, Group III The reason of these modifications is connected with the objective of getting a very simple model, understing simple as feasible, that is without taking so many measures. According above indicated the model proposed is: kr1 R1 L1 C MB k L1 k 2 L2 R2 C W1 W2 MT C BT L Rt Lt Fig. 6. Transformer model proposed being R1 L1 resistance inductance of primary winding R2 L2 resistance inductance of secondary winding C MT, CBT CMB primary secondary capacitance to ground capacitance between primary secondary winding Rt Lt resistance inductance of transformer grounding k R1, L1 k, 2 k L coefficients of primary secondary windings The model in ATP is showing in the figure 7: 5

6 Fig.7. Model in ATP 5. CONCLUSIONS This paper shows that the use of a recurrent surge generator to analyse the transferred overvoltages in distribution transformers is a very useful tool. The transferred voltage coefficient for different distribution transformers is in the range of the 10 % of applied voltage the oscillatory frequency is in the range of 200 khz for transformers with rated power between kva around 600 khz. for a wide range of transformers. This paper presents too a simplified model to analyse the transferred voltages in three-phase distribution transformers. The simplicity lies on the fact that we need only the constructive data (given by manufacturer) the capacitance to ground of each winding between them (feasible measurements). 6. ACKNOWLEDGEMENTS The authors gracefully extent their thanks to IBERDROLA for provide us the necessary technical economical support. 7. REFERENCES [1] EN Insulation Coordination. Part 2: Application guide [2] CIGRE WG.33.02, Guidelines for representation of network elements when calculating transients, October