Replacement of faulty sensor signals in a cement kiln is a challenging problem in fault diagnosis due to the increasing demand of fault-free feedback signal to the controller in order to produce good quality cement and estimation technique have been widely practiced for this. It is found from the literature that sensor faults in a highly energyintensive process will have a huge impact on the end product quality and profits. Cement industries, particularly the rotary kiln processes are plagued by temperature sensor faults, oxygen analyzer faults and it is a significant engineering research problem to diagnose the fault accurately amidst both random sensor and system disturbances. Fault in these measurements will severely affect the controller performance resulting in poor quality cement. In this work, Kalman filter based residual signal generation and replacement signal generation technique is employed for detecting bias fault and sensor measurement failure in pyrometer, thermocouple, and preheater oxygen analyzer of the kiln process because of the significance of these measurements in determining the quality of the cement. This work is divided into threephase1) Modelling 2) Residual signal generation 3) Replacement signal generation. The subspace method based state space model has been identified for three sensors and the residuals are found by finding the difference between the estimated states from the Kalman filter and the actual states and the deviation in the residuals will be used as a metric to indicate the nature and occurrence of faults, finally replacement signal has been generated for faulty sensors using the prior estimated states with irrespective of the sensor measurement.