Page 371 - Design of Simple and Robust Process Plants
P. 371

358  Chapter 9 Operation Optimization
                  It was said earlier that the accuracy of a performance meter is essential for its
                function. Such accuracy is determined by the mass balance reconciliation in which
                three factors play a role:
                  1.  Measurement accuracy
                  2.  Degree of redundancy of the flowsheet
                  3.  Process variability

                9.4.1.1  Measurement accuracy
                This is initially determined by the selected instrument. It is common practice to
                measure the volumetric flow and to calculate the mass flow based on an assumed
                density, but this already introduces a considerable error. Most streams are subject to
                density variations caused by temperature or concentration, while the density is often
                not measured together with the volumetric measurement. In addition, there is also
                a volumetric measurement, such as an orifice, that is already density dependent. In
                this application a mass flow meter is recommended, not only to avoid these implicit
                errors, but also to create a high accuracy.

                9.4.1.2  Degree ofredundancy
                This is achieved by installation of additional measurements for reconciliation of the
                flowsheet. An example of adding redundancy is shown in Figure 9.9. The addition
                of more redundant nodes is illustrated, in the bottom part also the storage area is
                included in the flowsheet. The accuracy of the individual streams can be calculated
                for the selected flowsheet based on the installed instrument accuracy. and described
                in the Appendix 9.1 (Heyen, 1994). This has been extended by Heyen et al. (1996) by
                the introduction of the contribution of the variance of measurement k in the estima-
                tion of the variance of reconciled state variable i. These contributions might be listed
                in order of impact. Conclusions drawn in this leading article, about these sensitivity
                analysis are:
                  .   Which are the measurements that contribute significantly to the variance of
                      the validated results for a set of state variables?
                  .   Which are the state variables whose variance is influenced significantly by
                      the accuracy of a given measurement?
                  .   How is the value of a state variable influenced by the value and the standard
                      deviation of all measurements?
                The sensitivity analysis has been implemented in Vali II software from Belsim.

                9.4.1.3  Process variability
                Depending on the application of the performance meter, the process dynamics play
                a role in the performance measurement. There are different ways to limit the effect
                of process variability:
                  ±   Integrate the measurement over a larger time interval.
                  ±   Include process hold-up variations in the model.
                  ±   Improve process stability.
   366   367   368   369   370   371   372   373   374   375   376