O Li E Data Reco Ciliatio for Process Co Trol

O LI E DATA RECO CILIATIO FOR PROCESS CO TROL Combined data reconciliation with estimation of slowly changing parameters has been implemented for closed-loop control in a Chemical Plant. Goals include streamlining use of redundant measurements for backing up failed instruments, filtering noise, and, in some cases, reducing steady state estimation errors. Special considerations include bumpless transfer from failed instruments and automatic equipment up/down classification. Parameters are calculated and filtered, then held fixed during each data reconciliation. I TRODUCTIO The purposes of this paper are to clarify the nature of the online estimation problem, and provide a “tool kit” for practical online data reconciliation applications. In this paper, “online” data reconciliation implies use in closed-loop control or optimization. It will be seen that there are different considerations in using data reconciliation in a process control environment than when doing typical offline applications like adjusting pilot plant or lab data. The necessary tools presented here include new techniques, extensions of previous work, ideas adapted from process control, and different philosophies for applying existing algorithms. It is also hoped that this paper will demonstrate some common ground for practitioners of process control and data reconciliation. It is written from the standpoint of the practicing process control engineer. As a result, there is an emphasis on small applications of data reconciliation, easy to use techniques, practical problems, and a heavy use of ideas from process control. This paper was written following the implementation of online data reconciliation in a process plant. It was recognized that the existing literature did not address all the concerns adequately, and that some new tools had been developed. The new tools include equipment up/down indicators, dynamic reconciliation, cascade estimation, and a new approach to bias estimation. Also, it was difficult to find a summary of other existing tools that might be relevant. This paper is an attempt to fill the gap by summarizing both new and existing tools. A brief summary of the contents of this paper follows. Following a brief review of data reconciliation, the paper starts with a discussion of the special needs for process control and online optimization. The needs discussed include forcing flows to zero for equipment that is shut down, handling instrument failure and out-of-range problems, recognition that noise at different frequencies should be handled differently (including estimation of slowly-changing process parameters and handling instrument bias), dealing with process dynamics, and recognition that in some cases steady state estimation errors are acceptable. One of the first needs addressed in this paper is forcing flow estimates to zero when a part of a unit is shut down. The new technique of using up/down status indicators, and classifying this status automatically, is introduced to solve this problem. Next, some practical details of implementing online data reconciliation are discussed. After some general comments on the selection of model equations, specification of measurement variances, propagation of variance, and processing interval for data reconciliation, the section discusses preprocessing of data, the actual data reconciliation, and some comments on how the reconciled data is used. Preprocessing of the