IMPROVING CARBON PLANT OPERATIONS THROUGH THE BETTER USE OF DATA

摘要

In most Carbon Plants an enormous amount of data is collected and used for short and long term control, improvement of plant operations, reporting, and to monitor anode quality for the customer. In the experience of the Authors, however, the way that data are used can be improved considerably - full value is rarely extracted from the information and several "data traps" are common: 1. Control decisions and actions are based only on comparisons to arbitrary specifications 2. Averages are used inappropriately, e.g. with a lack of attention to variability in the data and data is over-aggregated 3. Insufficient attention is given to the impact of sampling methods on the usefulness of data This paper will demonstrate these data traps, describe their implications, and suggest ways of avoiding these pitfalls using plant examples including information from the sampling and testing of anode cores.