The aim of this work package is to quantify the quality of voltage and current measurement data from sensor networks already installed in power grids, and to develop and validate methods to identify unreliable measurement data points in real-time, for example, detection of invalid data caused by sensor malfunction. Methods will include comparing grid data to a trained machine-learning profile based on reliable datasets.

Currently, available reliability and data quality assessment methods are insufficient, because they are not standardised, do not consider the uncertainty of the measurement chain or instrument models and/or the impact of different operating conditions, are typically non-real-time, and do not provide a quantification of data quality.

WP1 will collect and document all available real measurement data concerning five different types of grid phenomena, develop methods to identify unreliable measurement data points in real-time, and quantify the data quality of power grid measurement data, and validate, calibrate and provide traceability to the developed methods.