Statistical measurement calibration based on state estimator results

An approach to a statistical calibration of active power, reactive power and voltage measurements is shown. Measurement errors are studied according to their nature: systematic, installation and random errors. Effect of installation and systematic measurement errors on an energy management system (EMS) is analysed, paying special attention to its effect in state estimator behaviour. State estimator accuracy is discussed in order to analyse its use as a measurement standard. Statistical calibration is deeply explained: series of measured and estimated data are statistically processed getting valuable information regarding each measurement. This information is further analysed in order to remotely calibrate those measurements with systematic and installation errors, along with the determination of unknown state estimator measurement weights. The paper shows that once measurements have been statistically calibrated, immediate replacement of outdated equipment can be avoided, and furthermore, measurement errors can be dealt with as pure random errors (normally distributed)