D.W.T. Rippin

Simulation of single- and multiproduct batch chemical plants for optimal design and operation☆

Abstract Reasons for the increasing interest in batch plants and their improved design and operation are discussed. The extensive literature on the optimal operation of individual batch equipment items is reviewed and summarized. Topics covered include simple reactors with various kinetic schemes, polymerization, reactors with decaying catalysts including deactivating enzymes, batch distillation and crystallization. The study of multi-item batch plants is a much more recent development. A personal perspective is given of methods currently available for the quantitative treatment of the design and operation of multiproduct and multipurpose batch plants. Interconnections between existing areas of competence are pointed out and some directions suggested for further developments.

Batch process systems engineering: A retrospective and prospective review

Abstract The development of batch process systems engineering in the last ten years is reviewed with particular reference to a number of representative problem areas. The relation between research activity and industrial practice is examined and some future directions suggested.

Theory and application of the modulating function method—I. Review and theory of the method and theory of the spline-type modulating functions

Abstract The progress of research on modulating function methods and their applications over the period of 1954–1990 is summarized and analyzed herein. After introducing the concept of the modulating function method and defining the key properties of modulating functions the history of the method with its developments and applications is given. Applicable to nonlinear systems, the structure the models may have in terms of nonlinearities is explored in detail, which is followed by a discussion of the method in the framework of process identification. The second section introduces the analytically derived spline-type modulating functions and discusses in detail their specific properties.

Deactivation and regeneration of PdAg membranes for dehydrogenation reactions

Abstract Experiments on commercially available palladium—silver (PdAg) tubular membranes were carried out to: (1) verify the validity of both Sieverts and Ficks laws for H 2 permeation through the membrane; (2) quantify the deactivating effects of methylcyclohexane (MCH), toluene (TOL), sulphur (S), and chlorine (Cl) on membrane permeability; and (3) develop regeneration procedures to restore the original H 2 permeation rates. The results showed that the H 2 permeability data were represented by Sieverts and Ficks laws. Initial exposure of the fresh membrane to MCH or TOL caused inhibiton of H 2 permeation but after air oxidation, the membrane permeability was restored and constant for 32 h at 633 K in the presence of MCH. The H 2 permeation rates through the membrane were 14% lower in the presence of MCH or TOL. S and Cl poisoned the membrane strongly but redox procedures for S and a water vapour treatment for Cl were sufficient to restore the original H 2 permeation rates. The application of a catalytic membrane reactor (CMR) with a sulphided, monometallic, commercial catalyst is illustrated for MCH dehydrogenation where twice the equilibrium value of TOL was obtained.

Exceeding equilibrium conversion with a catalytic membrane reactor for the dehydrogenation of methylcyclohexane

A catalytic membrane reactor containing a tubular palladium—silver (PdAg) membrane sealed in the centre to separate in situ the hydrogen produced, was used to exceed equilibrium limitations in the dehydrogenation of methylcyclohexane. A sulphided, monometallic, noble metal catalyst produced higher than equilibrium yields of toluene and hydrogen from methylcyclohexane at economically viable throughputs. Experiments in the membrane reactor in the temperature range of 573–673 K, pressure range of 0.5–2.0 MPa and liquid hourly space velocity range of 2–12 volume feed/h/reactor volume showed conversions up to 4 times higher than equilibrium values after 300 h onstream and repeated temperature cycling. Exceeding equilibrium was due to the selective permeation of one of the reaction products, i.e. hydrogen, through the membrane. Reactor axial temperature profiles for the reaction with and without a membrane showed that the total amount of catalyst is more efficiently utilized in the membrane reactor.

Production planning and scheduling for multi-purpose batch chemical plants

Abstract A computer program for solving the general production planning problem in the case of multi-purpose batch chemical plants is presented. The procedure starts with an evolutionary enumeration of alternative production strategies. First, alternative ways of producing one batch of each of the various products are sought. Then, alternative production lines are constructed for each product by assembling types of batches of this product sequentially on the time axis. Finally, production lines of the same, or different products are set in parallel to form alternative campaign candidates. After the alternative production strategies (campaign candidates) have been generated, they are screened by identifying and rejecting non-dominant (i.e. inefficient) strategies. The screenign procedure is based on some principles of linear algebra and uses a duality between points and hyperplanes in a multi-dimensional space. Finally an optimal production plan is obtained by solving an optimization problem in which production time is allocated to the various production strategies.

Modelling, simulation and controllability analysis of an industrial heat-integrated distillation process

A controllability analysis of an industrial heat-integrated distillation process is described. The rigorous process model includes the important interactions of the liquid inventories between each other and with the heat transfer rate. Three different control schemes are investigated using steady-state and dynamic controllability measures. These schemes are compared with an unconventional control structure. The application of controllability analyses to real world examples is discussed.

A newflexible and easy-to-use general purpose regression program for handling a variety of single- and multi-response situations

Abstract A general purpose regression program based on Bayes and Maximum Likelihood principles has been developed. It is able to solve most of the practically occurring parameter estimation problems in single- and multi-response algebraic and ordinary diffirential equation models (including stiff systems) when the independent variables are assumed to be error-free. The program is written in a way that it is flexible and easy to use for a non-expert user while being efficient in computation due to the effective numerical methods and implementation structure.

Radial flow reactor optimization for highly exothermic selective oxidation reactions

Abstract Radial flow reactors (RFRs) with catalyst external to cooling tubes are compared with fixed bed, axial flow cases for highly exothermic, selective oxidation reactions. Steady state, one dimensional mass, heat and momentum balances were integrated for the butane to maleic anhydride (ma) reaction at commercially relevant conditions. “Triangular” kinetics for the main, total oxidation and product decomposition reactions were utilised (Wellauer et al., 1986). From this work, diffusion restrictions in the particles and interphase heat transfer were approximated making the reactor model quasi-heterogeneous. From the literature, heat transfer for radial flow through a packed bed with cooling tubes was estimated in addition to pressure drop for radial flow using flow parameters based on “local” values and including the bed voidage occupied by the cooling tubes. Comparison of the optimal cases at the same ma yields showed significant advantages for RFRs with respect to an increase in productivity, a decrease in reactor volume, lower hotspot temperatures and steam production and a significantly lower pressure drop. Stable operation was predicted for butane concentrations (5 mol %) within flammability limits but below ignition temperatures and well within runaway conditions for axial flow.

Knowledge-based tools for batch processing systems

Abstract The very varied way in which batch processes are used makes them particularly suitable for study using knowledge-based systems. The facilities needed for a general representation of batch systems, as presented in an earlier problem classification, are examined and methods for realizing these within a particular knowledge based system (KEE) are presented. Open questions about the further requirements of the system are discussed and preliminary experience of the simulation of a small batch plant is described.