A virtual in-situ sensor calibration method has been recently studied in order to solve the practical problems of sensor calibration in building energy systems: (1) time and monetary cost; (2) disruption to a normal operation; (3) difficulty in accessing various embedded sensors in equipment; and (4) large quantity of sensors. The proposed in-situ calibration method is able to approximate the measure and establish benchmark values for a calibration statistically or by using system models, without removing the working sensor or adding reference sensors as in a conventional calibration.