Feasibility of Instrument Transformer Calibration using PMU Data based upon Innovation Approach

Tracking state estimator is a well established method for state estimation in transmission network. The method of using innovation vector for bad data detection is available by default in tracking state estimators. In this paper, we investigate the possibility of extending the usage of the innovation vector available with tracking state estimator to estimate the ratio correction factor (RCF) of instrument transformers (ITs) using synchrophasor measurements. For this purpose, tracking and normal state estimators are considered on a three-bus, two-line system. One voltage measurement and one current measurement was biased, once with a magnitude correction factor (MCF) of 1% and then with a phase angle correction factor (PACF) of 1 degree. Normalized innovation is used for detecting bias errors. It is shown that the tracking state estimator will not identify a systematic error in IT. If we treat moderate or small RCFs as bad data in tracking state estimators, we find that the innovation vector is no longer zero mean. However, it is not large enough to pass statistical thresholds of bad data detection. Also, due to smearing, pinpointing the location of bad data is difficult. Hence, it appears challenging to develop IT calibration methods along these lines.