Polypropylene (PP) is widely used in packaging, labeling, textiles, stationery, plastic parts and containers worlldwide. In 2013, the global market for polypropylene was about 55 million tons, and today, Polypropylene is the world's second-most widely produced synthetic plastic after polyethylene.

Polypropylene is made from the polymerization of propylene gas in the presence of a catalyst system. Polymerization conditions such as temperature, pressure and reactant concentrations are set by the polymer grade to be produced.

There are three traditional manufacturing processes used to produce polypropylene* taking place either in a gas-phase or a liquid-phase process (slurry or solution). An example of Spheripol process is illustrated bellow:

The polypropylene process is a semi-batch process, switching from one grade of product to another. When a certain grade of product is under production, it is a typical continuous process consisting of reactors, recovery units, etc.

The implementation of Advanced Process Control (APC) on PP is not a new topic because it is a semi-batch process. It has been studied and practiced for years by different vendors and the main focuses include how to handle the nonlinearity of the process and how to automate the grade change. In this article, a typical APC application which combined Yokogawa APC as well as procedure automation technology is presented based on a real project.

The project KPI was set at the beginning:

1. Controller Uptime >=90%

2. Quality control: controlling a better reactor operation profile to reduce the “give-away” products

3. Transition control: Automating and standardizing the transition procedure with the integration of nonlinear APC control to shorten the grade transition time

4. Throughput maximization: pushing the process to its limitation to increase production yield and get better industrial assets use

Automate & Standardize the Transition Procedure

The automation of grade switch is configured in Yokogawa procedure automation package – Exapilot. Exapilot is an operation efficiency improvement support package aimed at automating the tasks performed by operators such as irregular operations or error-response operations in a plant. In Exapilot, the typical steps in operation procedures, such as stopping and starting pumps, MV/SV ramping and issuing field operation instructions, are provided in the form of “unit procedures”. 

In the project, a number of standard procedures were created including fouling calculation, SQC and grade change. Grade change introduced shortly hereafter.

There are 3 types and more than 90 grades of products that are produced by this customer. Exapilot runs grade change procedures based on the detection of current grade name and the selection of the target grade name. Together with the customer, Yokogawa specialists review the operation procedures and classified them into 15 standard STEPs. These 15 STEPS can be applied to any grade change with different combinations. Some of the STEPS can be skipped based on the current grade and target grade. An example of this can be seen below:

During the grade change, operators will select the target grade and simply press the “star” button. Based on current grade and target grade, a certain combination of the STEPs is automatically selected by Exapilot from thedatabase. Then Exapilot will run the procedures automatically and sequentially, including catalyst preparation, H2 feeding, C2 feeding, process setting, valve switching, temperature ramping, etc.

Controlling the Process with APC

In this process, Yokogawa APC was implemented in the loop reactor, gas reactor, and a few other units.

One of the key control targets of loop reactor was the density of slurry, which is directly related with final product quality – melt-index. A stable density control can bring customer not only a stable production rate but also a better product quality. At this site, the effluent of the loop reator-1 (RX1) is fed to loop reactor-2 (RX2), therefore the density control of the two reactors are correlated.

Yokogawa APC technology has a unique feature called “intermediate variable.” Gray box modeling allows this technology to impose process knowledge to solve complex control problems. By employing the intermediate variable, an input can be connected to one more intermediate variables and then to the final dependent variable. The series of interconnected low order models enables us to build a complex higher order model, which is also smoother than the FIR (Finite Impulse Response) model. In this case, the prediction of RX1 density can propagate to RX2, therefore the density of RX2 is predicted by the main controller and passed to the sub controller for dynamic optimization. This avoids the “sub optimal” control moves.

After the implementation of the APC controller, the standard deviation of density is reduced by 20% ~ 80% depending of the product grade. In addition, the production rate is also stabilized as one of the results. Below trend shows the production rate before and after APC control.

Integration of Exapilot and APC

During grade change, Exapilot downloads the targets of new product grade to the APC controller. The APC controller manipulated the related variables to reach the targets in a smooth and quick manner.

For example, the APC controller takes over the control of H2 concentration in the loop reactors.  Exapilot automatically gives the target of H2 concentration of the new grade. The APC controller moves the H2 flow to meet to new targets.

In the trend below, we present the differences of H2 control between APC and the traditional control scheme. It can be seen that the APC controller can reach the new target in a much faster speed. In fact, with APC controller, it takes about 22mins to increase the H2% to new level while without APC, it takes 70mins.

Achieved Benefits

The main benefits of the APC project are from Exapilot and Shell Multivariable Optimizing Control (SMOC) technology. Exapilot is configured to realize logic control which automates the grade change procedure. Yokogawa APC controllers are designed to stabilize the key control variables and drive the process towards economic optimization.

Stabilization of Key Variables:

Production Rate Increasing: In general, APC can increase the production from 2.7% or more, depending on the production grade, catalyst and other operation factor.

Grade Transition Time Reduction: Based on the calculation, the combined solution of APC and Exapilot can reduce the transition time by approximately 6%

Citation:

1. "Polypropylene Production via Gas Phase Process, Technology Economics Program". by Intratec, ISBN 978-0-615-66694-5, Q3 2012