Data reconciliation on multicomponent network process

Abstract A method for data reconciliation applied to steady-state chemical processes, with or without chemical reactions, is presented in this paper. In the first part, the use of available process information for the formulation of the minimization problem is discussed. An analysis of the process units and topology leads to the use of general and automatic algorithms that point out redundant measurements, variables to be calculated and redundancy equations. In the second part, the emphasis is laid on the numerical strategy used to estimate measurements for large multicomponent systems. An original method for the deduction of unmeasured variables is also discussed. Illustrative examples based on published applications and industrial test runs allowed to evaluate the efficiency of the proposed approach.