(712b) A Scheduling Perspective on the Monetary Value of Improving Process Control

Advanced process control (APC) strategies such as Model Predictive Control (MPC) have become the standard control technique in the process industries due to their unique capability of dealing with complex interactions, process nonlinearities, and operational constraints [1]. The use of APC offers many advantages, such as an increase in throughput, process stability, yield, or a reduction in energy consumption, waste, raw material, and costs. However, the implementation of a novel APC technology can be expensive due to the costs of manpower, hardware, software, and production loss due to installation downtime. While there are well-established metrics for the evaluation of the closed-loop performance of a control technology (either in-silico or once implemented on a physical system), a strong case for its implementation can only be established by quantitatively defining the associated monetary benefits. The question of how to assess the economic impact of any control system has therefore been a concern of the process systems community since the inception of the field [2,3].

During steady-state operation, an effective control system minimizes the variability of the controlled variables, such that the process can be operated close to the constraints on those states, which normally translates to higher profits. The most common approach for quantifying the monetary value of any proposed control system has therefore been through performance functions that define profit as a function of the distance between a process value and its operational limit [3]. In transient operation, an effective control system is one that minimizes the transition time between operating points, and limited work has been done on quantifying the monetary value of control systems in the context of transient operation.

In today’s fast-changing markets, transient operation has become ever-more relevant, and thus there have been many efforts towards integrating the dynamic process control calculations within process scheduling frameworks. This integration has been shown to yield significant economic benefits [4]. However, the problem of how to quantify the monetary value of a particular control technology across the horizon of a production schedule remains open.

In this work, we posit that the monetary value of a particular control system can be evaluated by defining control performance metrics and quantifying the change in a production scheduling profit objective value as the performance metric varies. We arrive at an explicit expression that is applicable to any production scheduling problem, and illustrate our methodology for several case studies.

References

[1] S. J. Qin, T. A. Badgwell. A survey of industrial model predictive control technology. Control Engineering Practice 11 (2003) 733-764.

[2] M. Bauer, I. K. Craig. Economic assessment of advanced process control – A survey and framework. Journal of Process Control 18 (2008) 2-18.

[3] T. F. Edgar. Control and operations: when does controllability equal profitability? Computers and Chemical Engineering 29 (2004) 41-49.

[4] M. Baldea, I. Harjunkoski. Integrated production scheduling and process control: A systematic review. Computers and Chemical Engineering 71 (2014) 377-390.