Traditional approaches in process performance evaluation rely on characteristics and time trends of critical process variables such as controlled variables and manipulated variables. Ranges of variation of these variables, their frequency of reaching hard constraints, or any abnormal trends in their behavior have been used by many experienced plant personnel to track process performance. Variances of these variables and their histograms have also been used. More formal techniques for process performance evaluation rely on the extension of statistical process control (SPC) to continuous processes.