Plantwide control: the search for the self-optimizing control structure

The limitation of fossil resources and the global efforts to reduce anthropogenic carbon dioxide emissions demand innovative strategies for the sustainable production of fuels and chemicals from renewable raw materials. Carbohydrates derived from lignocellulosic biomass comprise the largest fraction of terrestrial biomass feedstocks. Their conversion into valuable products can be envisaged by subsequent selective transformations of a set of platform molecules (new building blocks) that constitute the key intermediates for molecular diversity. The industrial implementation of such advanced biorefinery concepts faces a number of ecological, economic, and logistical challenges. On the most fundamental level, it is essential to develop novel synthetic methods and catalytic processes for the selective deand refunctionalization of biogenic substrates to yield a broad range of versatile chemical structures. Ultimately, this could allow the design of the optimum reaction pathway starting from the desired product structure, in analogy to a retrosynthetic analysis in traditional organic synthesis. Whereas much effort currently focuses on production processes for platform molecules, their true potential for further conversion remains largely unexplored. Scheme 1 exemplifies how a diverse range of highly attractive products and building blocks could be derived from levulinic acid (LA), which is accessible from wood-based feedstocks, and itaconic acid (IA), which can be obtained from green biomass. In particular, the cyclic ester g-valerolactone (GVL) has been proposed as one of the key components in a biomass-derived economy. Biogenic diols are interesting building blocks for biodegradable polyesters and other polymeric materials, as demonstrated by the commercialization of Sorona, a polyester which is made from glycerolderived 1,3-propanediol. Among the cyclic ethers, 2-methyltetrahydrofuran (2-MTHF) is advocated as an alternative solvent in the pharmaceutical industry and is also considered a fuel component. 3, 9] Hydrogenation routes involving heterogeneous catalysts to produce GVL, 2-MTHF, or 3-MTHF have been proposed. Organometallic catalysts based on ruthenium phosphine systems were reported to allow the conversion of succinic acid into g-butyrolactone and LA into GVL. 14] In Scheme 1. Analysis of reaction pathways for a variety of chemical structures formed from levulinic acid (LA) and itaconic acid (IA) as platform chemicals in a cellulose-based supply chain.

Selective and flexible transformation of biomass-derived platform chemicals by a multifunctional catalytic system

Decentralized control of the Tennessee Eastman Challenge Process

Feedback control for optimal process operation

Computer-Aided Chemical Engineering is reviewed from its beginnings in the 1950s to the present state in which virtually all chemical engineering is computer-aided. Over 200 computer-based computations routinely undertaken by chemical engineers are listed. Computer-aids are used at every stage from deciding what chemical species to make, through the conceptual design of the processes, the detailed design, the on-line control, optimization and retrofit design, up to the decommissioning. Computer-aids are important for assessing and minimizing environmental impacts and hazards. This chapter does not discuss any of these topics in detail. It concentrates on the design of reliable software and the correct use of software supplied by third parties. Guidance is given for designing software that properly meets the technical requirements of the end users. The program structure and test procedures necessary to validate the software are described. Methods for ensuring that the program properly incorporates the physical models on which it is based are outlined including emphasis on dimensionally consistent programming. The importance of data validation is emphasized. It is further emphasized that these design guidelines apply equally to models written in special-purpose modeling systems as they do to models written in general-purpose computer-programming languages. The elements of numerical analysis are introduced and illustrated with common examples from chemical engineering. Even simple computations, such as log–mean temperature difference, can be grossly in error if the numerical limitations of computers are ignored. Reliable, easily solved models should be near-linear and explicit, and steps to achieve such models are included. The treatment is designed to be adequate for engineers writing small programs, or small parts of larger programs. References are introduced to enable professional engineering programmers to pursue the topics in greater depth.



End-users of chemical engineering software have responsibility for the decisions taken based on computed results. They must ensure that the software is adequate for its purpose, that all data is properly validated, and that the computed results are properly interpreted. They are also responsible for scoping the uncertainties in the computation and assessing their impact on the decisions taken. This chapter describes the steps that should be taken to ensure that computer aids are properly used and decisions are properly reached. It gives further reading for engineers likely to manage projects in which extensive use is made of computer tools.



A brief introduction is given to some of the rapidly advancing areas of computer-aided chemical engineering.


Keywords:

CAPE (Computer-Aided Process Enginerring);
program design;
software validation;
Numerical Analysis;
Software use;
CAMD (Computer-Aided Molecule Design);
process synthesis;
flexible process design

Computer‐Aided Chemical Engineering

A new procedure for synthesizing process flow sheets and base-case designs has been developed. The procedure is evolutionary in nature and proceeds through a hierarchy of decision levels, where more fine structure is added to the flow sheet at each decision level. Heuristics are used to obtain some of the structural elements of the flow sheet, and other heuristics are used to make some of the decisions required at the various decision levels. In many cases, no heuristics are available, so that process alternatives are generated. The analysis protion of the procedure has strong focus on the economic tradeoffs that are associated with the significant design variables.

A hierarchical decision procedure for process synthesis

This paper reports on the hierarchical decision procedure described by Douglas which provides a simple way of identifying potential pollution problems early in the development stages of a design. The procedure focuses on the decisions required to complete a design. If these decisions are changed, then process alternatives are generated. Some of the decisions affect the exit streams from (and the feeds to) the process, and in some cases these exit streams have an adverse environmental impact. Hence, if we can make decisions, i.e., find alternatives, that do not lead to pollution problems, we can develop cleaner processes. The hierarchical decision procedure also can be used to decompose existing designs, and then process alternatives that eliminate existing pollution sources can be identified.

Process synthesis for waste minimization

A prototype expert system for synthesizing chemical process flowsheets

A hierarchical procedure for synthesizing an optimal plantwide control system is proposed for an existing continuous chemical process. Alternative plantwide control systems are synthesized and are compared based on economics. The cost associated with dynamic controllability is quantified by the controllability index ν introduced by Zheng and Mahajanam (1999). The procedure is illustrated on a simple reactor-separator-recycle system. Although the procedure is discussed for an existing plant, a simple modification of the approach can be used to determine the optimum surge capacities of a process, including reflux drums and column sumps—a problem of increasing interest.

Hierarchical procedure for plantwide control system synthesis

La gouvernabilite d'un processus peut etre atteinte: en modifiant le flowsheet pour obtenir de plus nombreuses variables manipulees, en surdimensionnant certains appareils pour que les contraintes ne deviennent pas actives sur tout le procede, en ignorant l'optimisation des variables de moindre importance. On cherche a determiner le moyen le moins couteux parmi les trois proposes

James M. Douglas

Papers

A hierarchical decision procedure for process synthesis

Process synthesis for waste minimization

A prototype expert system for synthesizing chemical process flowsheets

Hierarchical procedure for plantwide control system synthesis

The interface between design and control. 1. Process controllability