Detecting CVT Measurement Errors Using a Voltage Online Monitoring System and Maximum Approximate Coefficient Growth Rate

The capacitive voltage transformer (CVT) plays a critical role in power grid metering; however, over time, CVT measurement errors can become a prevalent issue. This study establishes an online detection method for CVT measurement errors utilizing a voltage online monitoring system and the maximum approximate coefficient growth rate (MACGR) method. First, a voltage online monitoring system integrated TV54A71 voltage sensors and a ZU3EG chip is designed. The TV54A71 voltage sensors and ZU3EG chip are, respectively, used to capture and deal the three-phase voltage on the CVT’s secondary side. Subsequently, the collected CVT secondary voltage is decomposed using the wavelet transform method in the monitoring system, and the MACGR of each phase is calculated to serve as a statistical index for CVT measurement errors. A kernel density statistical method is then applied to establish a measurement error threshold of 0.2%. The efficacy of the proposed MACGR method is validated through tests conducted using the CVT online test system of the China Electric Power Research Institute (CEPRI). The results indicate the ability of the MACGR method to accurately estimate CVT measurement errors. Furthermore, the designed monitoring system is assessed at a Xuzhou substation of the State Grid Corporation of China (SGCC). The system consistently collects CVT secondary voltage accurately over an extended period. Employing the proposed MACGR method on the measured data confirms its effectiveness in estimating the measurement errors of CVT of the SGCC.