This thesis defines the Energy Balance Load ( BL E ) as the difference between the heating requirements plus the electric gains in the building and the cooling coil loads. It then applies a first law energy balance in conjunction with the concepts of analytical redundancy (AR) and trend checking to demonstrate that measured values of BL E can be compared with the simulated characteristic ambient temperature-based BL E to serve as a useful tool to identify bad data. Uncertainty and sensitivity analysis are introduced to analyze the impact of each building or system parameter to the simulated values of BL E . A Visual Basic for Application (VBA) program has been developed through this research work, which applies the methodology illustrated in this thesis to automatically prescreen the measured building energy consumption data with the inputs of several key parameters. Through case studies of six on-campus buildings, the methodology and the program successfully identified monitored consumption data that appears to be erroneous, which may result from incorrect scale factors of the sensors and the operational changes to the building that may enormously affect the key parameters as the simulation inputs. Finally, suggestions are given for the on-line diagnostics of sensor signals.