Collaborators:  University of Maine, Argonne National Laboratory, and S.D. Warren Company
Funded By:      Department of Energy

Contact:            Mohamad Musavi

Project Summary:

The technical objective of this project is to develop process models based on a combination of expert systems, simple process-physics based models, artificial neural networks and fuzzy systems for characterization, analysis, control, and understanding of pulp and paper processes. The overall objective is to create a government-university-industry partnership for development and transfer of emerging intelligent technologies to the pulp and paper industry.

The proposed project will consider several different processes in pulping, bleaching, refining, and paper making operations. These operations include important issues such as on-line chip characterization, digester level, paper machine draw control, Titanium usage, modeling of pulp property, and net energy usage in the refining operation. Every one of these processes have multi-variable input space that is related to an objective output variable. Since the functional relationship between the input-out variables are not well understood, human operators are currently controlling these processes in an unknown environment and are relying on their experiences. As a result, adjustments are made very cautiously so as not to overreact to transient or erroneous information. This leads to significantly less than optimal operation that contributes to waste in energy and raw material consumption, and shorter plant runtime and life span. These factors have reduced returns on capital investment. The overall result has been a setback for the US pulp and paper industry. Once the leader in the world, the industry is now behind Canada and Scandinavian countries.

In this proposal, we will design and develop novel intelligent models to characterize process variables, to understand unknown functional relationships between process input-output variables, to predict future values of control variables, and to provide control strategies for optimal process operation. The research will provide a clear operational environment for the operators to base their decision on known boundaries and act quickly and efficiently than conservatively. We will build the success of this project on the success of our current projects with our industrial partners. The energy and waste saving, as the result of this project, has been projected to be in the order of $10-15M/year for one company alone. The net benefit to the pulp and paper industry and other related industries could be even more substantial.

In addition to its technical merit, this project fosters a triangle of government-university-industry partnership that will enable an efficient flow of information in all directions. It will provide the opportunity for the Argonne National Laboratory and the University of Maine Scientists to become more familiar with real-time industrial processes, apply their knowledge to solve some of the existing problems and work closely with their industrial partners. It will also provide a great opportunity for students to learn about an industrial problem and work with engineers from the industry. On the other hand industrial partners will economically and educationally benefit from this project. Some of the bottlenecks in their manufacturing operations will be removed and they get training and education in emerging intelligent technologies.

Return to Project Index

[Overview] - [Goals] - [Faculty] - [Students] - [Industrial Partners]
 [Publications] - [Projects] - [Axon] - [Downloads] - [Links] - [Contact INTSYS]