Based Optimization Drives Mill Profitability, Product Quality
Advanced process control such as model-based analysis, virtual sensors, and model predictive control are now able to impact more aspects of pulping and papermaking
Global competition and escalating raw material and energy costs have forced pulp and paper manufacturers to continuously seek improved efficiencies and lower costs in their operations, while at the same time producing higher quality, more consistent products. Systems and techniques are needed that enable process managers and operators to run their production lines within strategic constraints based on costs, quality, and demand criteria.
One set of emerging applications that delivers on these needs is advanced process control based on model predictive control (MPC) technology. This technology has been utilized for years in many process industries and more recently has moved into pulp mill operations. Now, improved information and control systems are providing the opportunity to extend these applications further into the papermaking process. High quality and robust process models incorporating detailed first principles of pulping and papermaking operations, as well as data driven artificial intelligence techniques, are a key source of systems analysis and online process optimization.
Multivariable process models are now being deployed to improve understanding and control of unit operations as well as entire plants, both in offline and online applications. Solutions currently in use include:
* Model-based multivariable process analysis
* Benchmarking
* Virtual sensors
* MPC
Through innovative product applications, MPC helps paper-makers improve and optimize their processes with low-risk high-return projects. From multivariable analysis and bench-marking to online virtual sensors and controls, process model applications provide the next step in process optimization.
MPC-based applications can be used to optimize specific unit operations or applied across entire operations. With a focus on improved quality and reduced variable costs, these applications will greatly improve paper mill operations in the coming years. Voith Paper Automation provides model-based analysis and optimization through its WebProfit optimization products and services. These applications are delivered as integrated solutions with OnQ quality control system (QCS) installations and OnControl distributed control system (DCS) packages or as stand-alone optimization to legacy systems.
The Growing Meed for Process Optimisation
The business case for new techniques in process optimization has been clearly driven over the past several years by market forces. Increased global competition and higher raw material and energy costs have forced pulp and paper producers to continuously focus on doing more with less. At the same time, converters and end users are demanding better performance and more consistent quality from the products they use.
Under these increasing demands, consistent profitability and even survival of many pulp and paper producers requires improved efficiency and utilization of resources. Beyond basic production procedure and maintenance improvements, sophisticated process optimization approaches hold the key to competitive advantage in today’s global marketplace.
Systems and techniques are needed that enable production lines to run at strategic constraints based on costs, quality, and demand criteria. Typically, the targets of advanced optimization practices are:
* Increased throughput
* Reduced raw material costs
* Reduced energy consumption
* Reduced chemical costs
* Improved quality/reduced variability
Model-Based Optimisation
One approach to process optimization is model-based analysis, prediction, and control. Process models can provide a great deal of information and utility ranging from generalized materials balances through the mill to complex multivariable control of quality properties.
Advanced process control applications have been applied for decades in many continuous and batch process industries. More recently, many pulp mills have applied these same techniques to improve yields and quality based on the ability of these systems to predict results for systems with slow and nonlinear dynamics. Now, improved information technology and measurement systems, including the DCS and robust process historians, are providing the data and connectivity required to extend these advanced process control applications throughout the entire papermaking process.
The preferred process modeling technique is based on first principles knowledge of energy and mass transfer throughout the process. First principles models provide stable and robust descriptions of a process throughout its entire range of operation. However, in a complex system such as a paper machine where all initial conditions are not fully known and where outcomes are affected by unmeasured external influences, equation-based models cannot always accurately describe the process.
Process models can combine process knowledge required to create first principles models with historical data and artificial intelligence, such as neural networks and genetic algorithms. Such models can lead to significant improvements in understanding, control, and profitability. This approach can be used to improve specific unit operations or can be implemented mill-wide.