OPTIMISED TOOL FOR THE MEASUREMENT OF WINDING RESISTANCES ON POWER TRANSFORMERS 1Haefely Test AG, Basel, Switzerland *Email: [email protected] Abstract: The topic of this paper is the analysis of the fundamentals of the winding resistance measurement and their application for the development of an optimized tool. The ultimate goal is to create a portable device, which measures a complete transformer the fastest and easiest way possible. The influence of the charging voltage and the measuring current on the measuring time has to be analyzed. Another important issue is the stabilization time on low ohmic delta windings. Further, testing time can be reduced by introducing a demagnetization feature, which eliminates the need of applying high voltage AC after a DC resistance test. Finally, the implementation of the optimized tool is presented with its efficient connection scheme and the multi-channel architecture. Fast, efficient and accurate measurement of winding resistances on large power transformers creates several difficulties. Long charging and discharging times, unstable values on closed delta winding systems due to long stabilization times, inaccurate temperature measurements for resistance correction, residual magnetism and its unwanted effects, inefficient connection and disconnection of the measuring equipment are just some of the difficulties to deal The paper describes an integrated, mobile instrument developed to speed up the stabilisation time when supplying DC to a transformer winding by an intelligent magnetic-flux optimised charging. With example diagrams of various transformers the charging and stabilisation effects are shown. After applying DC to a transformer, the core remains magnetized. This can cause problems for further measurements or reconnecting the transformer to the grid. Thus, an integrated, low voltage demagnetization function to bring a power transformer into a defined, demagnetised state will Typical problems arising when connecting a measuring system to a power transformer have been solved. In particular these problems are time consumption and faulty connections, which have been solved by a fully integrated connection set and a multiplexing circuit. Error possibilities in setting up the measurement equipment have to be faced and eliminated by offering state-of-the-art graphical, self-explanatory user interfaces with online information about all related values and conditions in an easy and well-arranged way. 2 SUPPLY VOLTAGE AND CURRENT 2.1 General measuring principle Figure 1 shows a schematic of a common winding resistance measurement technique. Figure 1: Winding resistance measuring principle using a DC power supply The instrument consists of a programmable power supply, which is normally operated in constant current mode. It is used to supply a user-defined current into the DUT. Additionally, there is a voltage and current measurement unit. Thus, the resistance is calculated according equation (1): Also, Equation 1 shows the main difference between a normal and a winding resistance measurement. There is a large inductance in series with the resistance of interest. This inductance is typically in the range of L0 = 0.1 H to Haefely brands HAEFELYJ&S ILUJUMM HAEFELYJ^W? HIGH VOLTAGE TEST | » & j ft J M E H T J TCCHHQLCHJV WWW Haefely Test AG ■ Birsstrasse 300 ■ 4052 Basel ■ Switzerland ■ Phone +41 61 373 41 11 ■ www.haefely.com ■ www.haefelyemc.com Haefely is a subsidiary High ifoliaga of Hubbell Incorporated, ---Test Business

5000H. In figure 2, the current dependency of the magnetizing inductivity is depicted. Figure 2: Inductance of a transformer core By rule of thumb the saturation current can be expressed in terms of the no load current l0: Because of the large magnetizing inductivity of the transformer, the measuring current can not be applied instantaneously. The current can only change according to equation (3): For a given transformer, the rate of change of the measuring current depends only on the voltage applied to the magnetizing inductivity LM. Current charging and discharging time depends mainly on the...

Delta winding considerations Figure 5 shows a common transformer configuration with a delta connected winding on the low voltage side. A Figure 5: Example transformer YNd11 The common method for measuring the winding resistance of a delta connected winding is illustrated in figure 6. The example shows a measurement at the transformer terminals ‘a’ to ‘b’. But in steady state the current distribution will be according equation 6 (for Ra≈Rb≈Rc): Using the method depicted in figure 6, the measured resistance value will only be correct when reaching steady state. By introducing a virtually circulating...

FLUX OPTIMIZED YN-DELTA METHOD Experimental Results This section describes another method to decrease the stabilization time on a low ohmic delta winding. The goal of this method is to reduce the magnetizing inductivity by saturating the core, too. But this method uses the high voltage winding to saturate the core, because the saturation current on the high voltage side can be significantly lower than on the low voltage side (depending on the turns ratio): Figure 7 shows the relative steady state flux distribution in the transformer core, when applying a current l2 from 'a' to 'b' on the transformer...

Figure 9: Resistance measurement on a large generator transformer (YNd5, 1100MVA) using the flux optimized YN-Delta method with h = 8A (HV) And figure 9 illustrates the stabilization time of the resistance reading, when the flux optimized YN- Delta method is used with 8A on the high voltage side and 8A on the low voltage side. In both cases the measurement current injected on low voltage side is 8A, which is smaller than the saturation current lSat2 ~ 12A. But the stabilization time with the traditional method is much longer (25 to 30 minutes) than with the optimized method (6 By increasing the...