Hierarchical Computer Monitoring And Grinding Control At Pinto Valley

Monitoring and control of conventional industrial wet grinding circuits is conceptually straightforward. However, since production objectives and operating constraints can vary widely from plant to plant, there is probably no single overall grinding circuit control strategy which is "best" in all cases; in-variably some compromise is necessary between performance desired under ideal conditions and performance achievable in practice. The recent advent of relatively low cost, reliable, high speed small computer systems is revolutionizing daily operations in increasing numbers of metallurgical processing plants. Of special power and capability is a hierarchical computer network like that currently in use at the Pinto Valley Operations of the Cities Service Company. Several satellite processors for key plant areas communicate with a master computer operating under a powerful multitasking real-time operating system. Hardware and software structure is such that downtime on one of the processors -including the master -does not have a major impact on overall system operation. Admittedly, the powerful, core efficient table driven satellite software, which handles process/computer communication, calculation of nonmeasured process variables, and basic Direct Digital Control functions, is difficult to modify. However, important high level process monitoring, report preparation, graphic display, and supervisory control functions -including economic optimization -can be implemented by project engineers at the plant site. Thus, for example, new control strategies can be quickly examined and rapidly evaluated. Such computer monitoring capability also permits use of short duration process pulse tests for approximate dynamic modeling and rapid calculation of controller tuning constants for single loop dominant systems. Following a brief back-ground discussion of the Pinto Valley Project, we present some of the more interesting results and conclusions of our work over the past year.