Omega-3 Fish Oil,Fish Oil 10/50 Tg,Omega 3 Fish Oil,10/50 Tg Fish Oil ZHOUSHAN SINOMEGA BIOTECH ENGINEERING CO.,LTD , https://www.sinomegabiotech.com
Application of Predictive Control (PFC) in Vinyl Chloride Distillation Process
In response to the control challenges in the VCM (Vinyl Chloride Monomer) rectification process, our company implemented the Zhejiang University Central Control Predictive Function Control Software (APC-PFC) to establish an advanced control system for the rectification process. This system was successfully launched in March 2005 and has since been effectively used for VCM rectification. The implementation significantly reduced the workload of operators by using a forecasting model to predict the trends of controlled variables. It also enabled rolling optimization and deviation correction of control loop setpoints, which improved the accuracy and stability of key process variables. As a result, product quality was ensured, operational conditions were maintained smoothly, and the overall control performance was greatly enhanced.
The VCM rectification process typically involves two main stages: a low-boiling column system and a high-boiling column system, forming a classic two-stage cascade distillation setup. The crude vinyl chloride from the compressor is first sent to a full condenser, where most of the gas is condensed. After water separation, it enters the low-boiling tower. Uncondensed gas goes to the exhaust condenser, and the condensate is returned to the water separator. The low-boiling tower reboiler is heated with hot water, and the material separated at the top is sent to the tail gas condenser. The uncondensed gas is recovered by a tail gas adsorber, while the remaining inert gas is discharged. The bottom liquid flows into the high-boiling tower via an intermediate tank, where it is heated again to produce refined VCM, which is then condensed and stored in a finished tank. To maintain monomer purity, the bottom of the high-boiling tower contains some high-boiling components that must be periodically removed.
Due to the continuous nature of production, the VCM rectification process is closely connected with the previous synthesis stage and the following polymerization process. The output from the synthesis section serves as the feed for the VCM rectification section. Therefore, both the flow rate and composition of the feed to the low-boiling tower frequently change, which is a major source of disturbance in the VCM rectification process. These changes affect both the material and energy balances. Due to the long time lag of the crude VCM in the distillation column, traditional feedback control struggles to respond quickly enough. Additionally, the complex and often unmeasurable feed composition poses significant challenges to maintaining stable product quality.
For the high-boiling tower, the feed changes are frequent and large, making it difficult to control the liquid level. Frequent fluctuations in the liquid level can lead to variations in the column temperature, which in turn influence the top temperature, pressure, and reflux of the tower. The interdependent relationships between these parameters make the control of the high-boiling tower particularly complex. Many factors collectively determine the operating conditions of the unit, and each parameter has specific allowable limits. Coordinating the control variables is essential to ensure the overall control performance and final product quality. Thus, the control challenge in the high-boiling tower is especially prominent in the VCM rectification plant.
The APC-PFC control software is based on Predictive Function Control (PFC), a member of the model predictive control family. Introduced by J. Richalet et al. in 1986, PFC focuses on the structural form of control inputs, treating them as weighted combinations of basis functions. It incorporates error feedback correction and performs online optimization. Key features of APC-PFC include a simple algorithm with minimal computational load, ease of integration into existing DCS systems, and the ability to handle measurable disturbances through feedforward control. The system also allows quick parameter adjustments, significantly reducing setup time.
Before implementing the predictive control scheme, the VCM rectification process was characterized by strong variable coupling, multiple disturbances, and varying feed conditions. The presence of an intermediate buffer tank between the low- and high-boiling towers allowed for relative independence, enabling the design of separate controllers for each. The advanced control system uses a predictive function controller as the supervisory layer and conventional PID as the control layer, combining the reliability of PID with the dynamic performance of PFC. Feedforward control was added to enhance disturbance rejection, ensuring smoother operation and better control performance.
Since the deployment of the APC-PFC predictive control software, the stability of production and product quality have seen significant improvements. Before the system was implemented, the top temperature of the low-boiling tower and the liquid level of the high-boiling tower fluctuated widely, resulting in unsatisfactory control. After the system was put into use, process smoothness improved dramatically, with notable reductions in variance and range. Other parameters also showed improvement. Despite minor disturbances caused by intermittent discharge of high-boiling materials, the system demonstrated strong anti-interference capability. The purity of the monomer now consistently reaches 99.999%, fully meeting the requirements of PVC production.
The APC-PFC predictive control software was installed and debugged without disrupting the production process. It supports bumpless switching between conventional PID and predictive control with just one switch. Operators require minimal training to use the system. The few adjustment parameters are easy to modify, and maintenance engineers can perform effective system upkeep with basic training.
Author: Xue Jian, Qinqiao Liang, Chen Dongmeng, Fan Minghua, Nantong Jiangshan Agrochemical & Chemical Co., Ltd.
Source: Control Engineering China