Recently, the latest version of the B&R APROL process control system was used by The Lenzing Group to carry out a dynamic process simulation in order to optimize production of their man-made cellulose-based fibers.
The Lenzing Group is a global leader in the manufacturing and marketing of man-made cellulose-based fibers, primarily used in the textile industry.
The fibers are manufactured in several steps, beginning with the renewable resource wood. The wood is converted to natural fibers in large systems, which operate with a high degree of automation, the lowest possible environmental impact and the best possible utilization and recycling of energy and residual materials. Optimization of production is the main motive in the manufacture of the fibers.
While aiming for the highest level of optimization, in-house chemists developed and tested a new manufacturing process in the Lenzing laboratory, which proved to be effective. The goal from there was to implement the results in a full-scale production system.
Because the production processes in this sort of system are complex and are not entirely hazard-free due to the chemical processes used, it is common to build a pilot system after lab testing and before construction of the full-scale system. The pilot system is used to test and optimize processes scaled down to kilograms, before they are applied to the full-scale system. This intermediate step is costly, time consuming, and requires a lot of space. "This drove us to search for a solution,” says Dr. Bernhard Voglauer, managing control technician in the Lenzing Fibers technical planning department.
Voglauer found the answer to this task to be dynamic process simulation, in which the whole production system is simulated as a computer model. In this way, optimization can be performed on a virtual system. As a tool, MATLAB/Simulink from the American manufacturer The MathWorks was their choice. This decision was made because it is a well-known process simulation system used worldwide that could handle the required complexity, and because code that is understood by B&R systems such as the APROL process control system can be generated directly from the simulation models.
Over the course of three months and nearly 400 hours of work, a physical model and then the Simulink process simulation model were created based on construction data, the process graphic, expert knowledge and the results of lab testing.
The "hardware-in-the-loop" concept was used: The process simulation model was connected with the PCS via a LAG simulator, which was supplied with an I/O list by a CAE system.
"This process has brought Lenzing AG enormous savings before and during commissioning," states Voglauer. "180 errors were able to be corrected in the application software before commissioning was started." Actual commissioning then went very quickly because the process control system could be ruled out as the source of errors. The programmers were also able to get through the commissioning phase and carry out their tasks, such as setting the parameters for control loops, limit values and program automats in the office during the process simulation phase as planned without having to rush.
"The time saved and the problems avoided through virtual commissioning convinced even the toughest critics," says Bernhard Voglauer. MATLAB/Simulink is now almost always used to develop new control concepts, and the plan is to use the direct code generation options provided by the integration of Automation Studio and the latest version of the APROL process control system from B&R.
DAANET is the Australian distributor of Bernecker & Rainer's (B&R) products.