Introduction

Process Automation Systems (PAS) are widely known to be a crucial element in processing plants worldwide. Chemicals, petrochemical, oil refining and power generation industries are vast and complex, and said to be controlled by just a handful of major players worldwide. In an environment that is ever-changing, industry leaders employ state of the art PAS's when possible to mitigate complexities -- with Advanced Process Control (APC) systems specifically leading to higher efficiency, less operator interaction and increased profits

Advanced Process Control (APC) is a proven control and optimization technology delivering measurable and sustainable improvements in production yield, coupled with the added value of energy savings. Most engineers will agree that stabilizing control loops, with big process dead time and strong process interactions is very difficult. APC has become a standard solution for realizing stable control processes – and to be put in layman’s terms – APC is the added value upgrade to a Process Automation Systems.

"MPC (Model Predictive Controllers) rely on dynamic models of the process, most often linear imperical models obtained by system identification. The main advantages of MPC is the fact that it allows the current timeslot to be optimized, while keeping future timeslots in account." - Wikipedia

"Soft Sensor or (inferential analyzer) is a common name for software where dozens of measurements are processed together. The interaction of the signals can be used for estimating properties of material being processed." - Wikipedia

APC is typically comprised of, Model Predictive Control (MPC) or Soft Sensors (Inferential property estimation) enabling more advanced stability and control by utilizing the multivariate analysis and estimated product properties in real-time (such as RON, RVP, Density, etc.)

The information that is gathered is then used to stabilize the process control without user interaction, and then adjust the set point closer to the limit - as shown below (Exhibit 1). Changing the specification point of the operation from high energy consumption to a point with less energy consumption creates the opportunity to use energy much more efficiently and improve profit margins.

Implementation

Step 1: A stable base layer control is the first step towards achieving stability in your overall operation. Since MPC systems generally manipulate the set points of the targeted PID controller, it is important to tune PID parameters and introduce enhanced regulatory control prior to implementing MPC system [enhanced regulatory control improves the controllability of PID loops by putting advanced functions between related control loops in the Distributed Control System (DCS) level].

Step 2: Implementing soft sensor and MPC is the next step. Improving potential benefits, with reduced cost and efforts comparing with the final step of optimization. Using prioritized constraints, MPC realizes optimization of control variables and economic formulas for both steady-state and static optimization objectives.

Step 3: The last step – but certainly not the least – is implementing the first principle model-based real-time process optimization. Great efforts are required not only for defining rigorous static process models and setting data reconciliation, but for maintaining these models as well.

A typical APC implementation schedule can be seen in Exhibit 3. Estimating production quality in real-time is essential in controlling product quality in MPC. Therefore, soft sensor (inferred property) modeling by historical process data and commissioning are conducted prior to, or in parallel, with the MPC design.

Step response testing is required to get dynamic response data for MPC modeling. After MPC modeling based on the dynamic response data and configuration based on the control strategy MPC simulation with static parameters are completed, tuning is then executed to confirm MPC behaviors. With tuning MPC dynamic parameters and soft sensor calibration parameters, commissioning is executed, and implementation is complete.

Note: Post audit is important. Measuring the control performance and controllability of key variables, with prior result being compared to results after APC introduction, will give a clearer understanding of the improvements and benefits of APC (Exhibit 5)

Key Benefits

Key quantitative benefits of APC implementation are both quantitative and qualitative. The quantitative benefits are obvious, coming mostly from improvements in yield, capacity increase, and energy savings. The pie chart below displays this quite clearly.

In the special case when product prices are stable and conversion rates are improving, feed stock throughput will generate even greater benefits. Qualitative benefits attributed to APC equal less wear and tear, maintenance, and operator dependence. The automated optimizing control will stabilize operations on a complete scale

The typical benefits of implementing APC in refining process units are shown in Exhibit 4. The benefits of APC range based on the operating conditions and the equipment capacities of a particular operation, but the benefits of APC introduction site-wide are estimated US$ 20 cents per bbl crude, in usual cases.

For example, if a medium-capacity refinery were to implement an APC application in a crude distillation unit (e.g. 140 crude kbbl/day) the benefit is estimated to be as much as 2M$ per year (0.04$*140,000*365) - small investment, relatively, in APC equals big benefits – quantitative and qualitative.

Full returns on APC investment after implementation in major processes are typically within 12 months – i.e. Investment in APC can be fully recovered by the benefits realized by the APC system within one year i.e. If you spend 1M$ to implement an APC application in a crude distillation unit at a medium-capacity refinery the ROI will be 6 months.

The new APC system is the best in class suite to Rapidly Deploy and sustain Advanced Control applications, generating substantial benefits and improvements throughout the plant life-cycle. Please check the "Latest in APC" link for more information: