Xavier Joulia

Prediction of water adsorption curves for heterogeneous biocatalysis in organic and supercritical solvents

Abstract In the framework of heterogeneous biocatalysis in non-aqueous media, knowledge of water adsorption curves is necessary for the choice of operating conditions. Indeed, water content plays a crucial role in the performance of such enzymatic reactors. A method to estimate water adsorption curves in biphasic solid-fluid systems is proposed here. It is grounded on the hypothesis that the water adsorption of the solid is not dependent on the solvent and on the theoretical evaluation of water activity in the solvent. It is tested in the case of 3-methyl-1-butanol, 5-methyl-2-hexanone and supercritical carbon dioxide as solvents and an immobilized enzyme, Lipozyme ® , as the solid phase. The method proved to be in good agreement with experiments. The method uses only a single experimental curve but necessitates the development of numerical modeling of thermodynamics. However, a simplified method, restricted to the range—hydrophobic to weakly hydrophilic solvent—is also proposed and tested.

A general strategy for parameter estimation in differential-algebraic systems

Abstract In this paper, a general method is presented for parameter estimation in differential and algebraic equation systems (DAE systems). It is more specifically developed for the kinetic parameter estimation of an isothermal batch reactor kinetic model. How to improve the Gauss-Newton method used for parameter estimation in the context of dynamic simulation involving DAE systems is shown. Emphasis is put on the calculation and conditioning of the Gauss-Newton operator. At the minimum, the joint linearized confidence region is evaluated and used to discriminate among rival models. The given methodology has been applied to determine a suitable Langmuir—Hinshelwood model for multiple catalytic hydrogenations.

Improving steady-state identification

Abstract The use of online data together with steady-state models, as in Real Time Optimization applications, requires the identification of steady-state regimes in a process and the detection of the presence of gross erros. In this paper a method is proposed which makes use of polynomial interpolation on time windows. The method is simple because the parameters in which it is based are easy to tune as they are rather intuitive. In order to assess the performance of the method, a comparison based on Monte-Carlo simulations was performed, comparing the proposed method to three methods extracted from literature, for different noise to signal ratios and autocorrelations. The comparison was extended to real data corresponding to 197 variables of the atmospheric distillation unit of an important Brazilian refinery. A hierarchical approach was applied in order to manage the dimension of the problem.

Maxwell-Stefan approach coupled drop population model for the dynamic simulation of liquid-liquid extraction pulsed column

Abstract Zimmermann and al. (1995) have proposed a model for the simulation of a multi-component extraction process based on Maxwell-Stefan approach and incorporating a drop population model. In the continuation of this work, this paper presents some recent developments, notably the extension to the dynamic and some improvements at the hydrodynamic level. The model is characterised by the coupling of two main aspects in separation processes: hydrodynamics and multi-component mass transfer. Hydrodynamics is described by a drop population model: fundamental mechanisms like transport, axial back-mixing and forward-mixing, drop breakage and interdrop coalescence, including Marangoni effect, are described in a detailed way. Multi-component mass transfer is computed using the Maxwell-Stefan approach. The proposed model is able to predict the component concentration profiles in the dispersed and continuous phases, hold-up and drop size distribution throughout the column and the interactions between mass transfer and hydrodynamics under operating conditions up to flooding.

Dynamic simulation of Brazed Plate-Fin Heat Exchangers

Abstract In this paper a dynamic simulator of Brazed Plate-Fin Heat Exchangers (PFHEs) ProSectm is presented. A rigorous model is used which allows one to represent the very broad range of configurations for this kind of equipment. A DAE solver (DISCo) is used with a sparse direct solver (MA38) to handle the large number of equations. Applications for control and design are highlighted through an industrial case study involving shutdown.

Simultaneous modular dynamic simulation: Application to interconnected distillation columns

Abstract Key issues related to the use of the Simultaneous Modular Approach as a basis for an efficient dynamic simulator are presented. The proposed strategy is articulated roughly at two calculation levels. First, the module level, where the model and local control loop equations are solved all together via a global approach using an adapted Gears integration code for DAE systems. Features and problems (consistent initialization, index problem, discontinuities, …) encountered at this level are discussed in the proposed context. Second, at the flowsheet level recycle and control equations are solved simultaneously using a DAE integrator along with Dynamic Sensitivity Matrix techniques used to promote the convergence. Testing of this strategy - with comparison to other approaches - was previously performed on flowsheets with interconnected flash units. It is extended here to flowsheets with highly interconnected distillation columns. Future perspectives and evolutions are then discussed.

New configuration for hetero-azeotropic batch distillation: I. Feasibility studies

Abstract For heterogeneous batch distillation a new double column configuration operating in closed system is suggested. This configuration is investigated by feasibility studies based on the assumption of maximal separation and is compared with the traditional batch rectifier. The calculations are performed for a binary (n-butanol—water) and for a ternary heteroazeotropic mixture (isopropanol—water+benzene as entrainer).

Moving finite difference method for tracking a shock or steep moving front

For tracking a shock or steep moving front in the convective Partial Differential Algebraic Equations (PDAEs), some spatial discretization methods such as central, upwinding, Essentially Non-Oscillatory (ENO) and Weighted Essentially Non-Oscillatory (WENO) schemes, and the moving grid techniques are examined respectively and simultaneously with regards to accuracy, temporal performance and stability. The uniform fixed-grid method is efficient owing to relatively short CPU time. However, the solution of the Fixed Stencil (FS) approach such as the central and upwinding spatial discretization methods is unstable. Many mesh points are needed for all of the fixed grid methods to obtain enough solution accuracy. In the moving grid techniques, stability of the solution is enhanced even at small mesh numbers but it is prohibitive because of the addition of a complex and nonlinear mesh equation to solved PDAEs. The combination of the ENO/WENO schemes (based on an adaptive stencil idea) with moving grid techniques gives a stable and accurate solution event at small mesh numbers and is efficient to track a moving shock front.

Development of a computational tool for the transient kinetics of complex chemical heterogeneous reaction systems

We present a program for the modelling of the transient of complex heterogeneous chemical reaction systems. The philosophy of the program is based on the formulation of a generic model for heterogeneous reactions derived from a phenomenological description of the system, where all the reaction steps should be detailed., the challenge of dealing with the complexity of the model thus generated is addressed by the robustness of the numerical tools employed for its simulation and parameter estimation. The package is able to obtain a fit even for complex models. This fact justifies a meaningful discussion on the usual philosophy of model analysis which consists of steps of reduction and discrimination. The possible issues are illustrated with a study on the hydrogenation of acetophenone over Rhodium catalyst.

Tools for computer aided analysis and interpretation of process simulation results

A process modelisation for results analysis is presented. This modelisation has recovered concepts already used for design (aggregation of units, pinch technology, etc) and come up with specific ones of its own (aggregation of components, process objectives, etc). The understanding of the process has been enhanced by means of objective functions and path visualisation. General or user-specified automatic checking constraints allow the engineer to concentrate on important features of the process. These techniques may be extended to other procedures (test of alternatives, etc). A prototype containing the ideas presented in this paper is processed. The next step of the concepts development is a closer study of subfunctions, and an extension of constraints checking using automatic consistency checking.