First step is cost benefit analysis. User needs to define what process units or areas are appropriate candidates advanced process control (APC). It needs to also quantify benefits and improvement opportunities of advanced process control (APC) of defined process unit. Mistakes lead to incorrect cost estimation, which will have negative consequences on the advanced process control (APC) project.
The second step is system identification and APC controller design. To obtain appropriate advanced process control (APC) models plant tests needs to be conducted and manipulated, controlled variables and process or/and equipment constraints needs to be defined. Many production companies were faced with lost margins during the lengthy design phase and disruptions to the process of gathering data to build advanced process control (APC) models.
The third step is APC controller commissioning. First, the communication with the existing plant control system is tested. Next, each multivariable advanced process controller module must be tested in an open-loop model. Each control loop needs to be tuned and the controller actions needs to be tested in an advisory mode. The optimization algorithm can then be commissioned after the multivariable controllers have exhibited the desired performance
The last step is APC models and controller maintenance. This is required to ensure continued benefits from the implemented advanced process controls. Advanced process control algorithms need regular analysis. When product specifications and process conditions change or new product specifications are added, re-tuning of the advanced process controller is required. Also, any changes to the process or equipment after deployment would require re-identifying the advanced process control (APC) models. This again require costly step testing.
Some companies today have moved beyond the traditional methods to a more advanced technology by integrating adaptive process control with advanced process control (APC) technology. Adaptive process control allows to experience faster deployments and sustained benefits through continuous advanced process control (APC) models updates in the background with no disruptions to the process, reduction of time and money within the organization.
Authors: Tushar Singh & Kate Kulik, Aspen Technology Inc., & Ali Awais Amin, Intech Process Automation
PiControl Solutions COLUMBO is the latest breakthrough advanced process control (APC) technology for identifying open-loop advanced process control (APC) dynamic models using completely past closed-loop data without any step-tests in the plant required. It helps to identify bad (wrong) model predictive control (MPC) / advanced process control (APC) models and helps to generate new and correct models. It allows fixing all known parameters like dead time, time constant or even process gain based on operator experience/knowledge, engineering calculations, vessel dimensions, and vendor data. COLUMBO uses data from any model predictive control (MPC) / advanced process control (APC) controller, even data of slave PIDs in auto, manual or even cascade modes. It can identify simultaneously as many as ten model predictive control (MPC) / advanced process control (APC) dynamic models.
Recent Comments