Michael Nikolaou

Chemical Engineering Department

Texas A&M University

College Station, TX 77843-3122

Presented at AIChE World Congress, San Diego, CA (1996)

Computer-aided process engineering (CAPE) refers to the use of computers to design and operate chemical processes effectively and efficiently. In technical jargon, CAPE entails software development for process optimization, planning, scheduling, simulation, and control. While recent CAPE advances span all facets of CAPE, the emphasis of this article will be on computer-aided process operations and control, in the spirit of two recent conferences (ISPE 95, CPC V). The interested reader can find a plethora of reports on technical developments related to CAPE in the above two references. This article presents our viewpoint of the CAPE research climate.

CAPE represents a unique research opportunity for chemical engineers. There are several factors in support of the above claim:

• Market forces: While the construction of new large chemical process plants in the US is almost at a standstill, improvements in existing plants are common. This trend is bound to continue, given increasingly strict safety and environmental regulations, as well as stiff domestic and foreign competition. Improved process automation and process design modifications are important elements in meeting these challenges. In particular, the need to quickly respond to rapidly changing market conditions and government regulations is forcing companies to seek thorough computer integration of chemical processes, from the lowest level of process control to the highest layers of management and business optimization.
• Installed base: Computers are already a commodity that is widely available in most chemical plants. However, the computing potential such computers afford is far from being fully exploited. To take advantage of the computing power available in chemical plants, techniques are needed that convert the myriad or numbers continually collected and stored by computers into useful knowledge and understanding, which, in turn, could be used in human-aided or automated decision making.
• Technology drive: Availability of increased computer CPU power at predictable rate (doubling every 18 months) and well-publicized recent advances in computer networking and communications are creating enormous potential for the development and application of new CAPE techniques. In addition, CAPE techniques that used to be practically infeasible due to lack of sufficient computing power, are gradually becoming practical.
• Scientific advances: There have been important theoretical and practical advances in the general science of how engineering problems are solved with computers. For example, artificial intelligence (AI) methods are maturing and evolving from marginal academic curiosities to mainstream industrial realities, particularly in the area of "intelligent" process automation and decision making. New optimization paradigms are also emerging. The full potential of these areas has yet to be explored.
• New applications: In addition to the oil and petrochemical industries, on whose research problems a large number of chemical engineers are currently working, diverse industries such as the microelectronics, food, and biochemical industries are increasingly relying on chemical engineering talent for the development and application of CAPE solutions. Moreover, new technologies for environmentally benign and safe manufacturing have to consider the use of computer tools as an integral part of process development and operation.
• Industry-university partnership: Because several chemical companies are downsizing, they are examining new ways of partnering with universities, to meet their research needs. The traditional role of the academic side in such partnerships has been to focus on long-term research. Industry, however, desires quick solutions for its current pressing problems. CAPE defines a research area where the seemingly conflicting goals of long- and short-term research are, in fact, highly complementary. This has created an excellent climate for a new kind of industry-university research collaboration, with long-term fundamental orientation (based on the abstraction and projection of applied ideas) but also industrial relevance and careful monitoring of milestone achievements. This collaboration can capitalize on

CAPE problems are engineering problems. For their solution, they require thorough understanding of both the process involved and current computing technology. Chemical engineers are uniquely qualified to combine both skills.

The need for CAPE researchers to combine expertise from several disciplines presents a challenge and an opportunity. It is, for example, becoming increasingly evident that no single paradigm will dominate process control research in the near future. While existing paradigms (e.g. Model Predictive Control) have proven useful both in academic and industrial research, it appears that ideas from diverse areas are needed for the successful handling of the complexity level associated with today’s processes. As an example, while control theory has long focused on algorithms, "intelligent" control systems can now include elements such as logic, sequencing, reasoning, and heuristics. As another example, the use of statistics becomes increasingly more widespread in the automatic interpretation of huge volumes of data accumulated by data acquisition systems in chemical plants. The diversity of tools used by CAPE researchers today is gradually reflected in the titles of sessions organized by the CAST division of AIChE in national meetings.

Computer-aided process engineering is a research field with predictably high potential. Future advances are likely to come from an interdisciplinary approach to CAPE problems, and from good interaction between industry and academia.

Fifth Conference on Chemical Process Control (CPC V), Tahoe City, CA, January 1996.

Intelligent Systems in Process Engineering (ISPE 95), Snowmass, CO, July 1995.