M. Abderrazak Latifi

Leak identification in a water distribution network using sparse flow measurements

Abstract A linear programming technique is proposed to balance the flows in a network, matching flow measurements where available. Where necessary, a “leak-out” flow is invoked on a pipe section in order to achieve the balance. Usually, multiple solutions are possible, and these are sounded out by progressively increasing an integrity weight for each pipe section. A feature of the method is that it overlays “snapshots” of the network at a series of points in time, in order to progressively narrow down the part of the network which can commonly account for all observations.

Sensitivity, estimability and correlation of parameters describing equilibrium and nonequilibrium transports of bromide tracer in the field lysimeter

Modelling of solute transport under transient flow conditions requires many input parameters, especially when using the nonequilibrium model. Most often, these parameters are not accurately measured but estimated by means of the inverse methods. Among different factors causing the nonuniqueness issues of estimated parameters, limited information content and correlations among parameters are the important ones. Therefore, it is necessary to evaluate these factors to better define optimisation approaches. In this study, sensitivity and estimability analysis were performed to evaluate sensitivity, estimability and correlations of parameters describing the equilibrium and nonequilibrium tracer transport in the field lysimeter. Estimability analysis was based on a sequential calculation algorithm of a sensitivity coefficient matrix. The comparison of estimability and sensitivity results indicated that estimability analysis is much more useful than sensitivity analysis. The former allows evaluation of the information content of the tracer breakthrough data and possible correlations among parameters. The estimability results indicated that the tracer breakthrough data contained limited information content for estimating the hydraulic and solute parameters in the nonequilibrium model. Furthermore, the high correlation between the first-order rate solute and water transfer coefficients was found based on those data. This study may be considered as warning for the use of tracer breakthrough data to investigate the nonequilibrium solute transport at the field lysimeter scale. This further suggests that additional data are required to increase the ability of estimating the hydraulic and solute parameters at the given experimental conditions.

Multicriteria dynamic optimization of an emulsion copolymerization reactor [conference paper]

A multicriteria optimization procedure based on an evolutionary algorithm has been developed to determine the optimal control policies for a fed-batch emulsion copolymerization reactor, particularly for styrene and butyl acrylate in the presence of n-C12 mercaptan as chain transfer agent (CTA). The process model was elaborated and validated experimentally in order to predict the global monomer conversion, the number and weight average molecular weights, the particle size distribution and the residual monomers mass fraction. The process objectives were to produce core-shell particles (hard core and smooth shell) with specific end use properties and high productivity. This has been achieved by the maximization of the conversion at the end of the process and the minimization of the error between the glass transition temperature and a designed profile subject to a set of operational constraints. The nondominated Pareto solutions obtained were ranked according to a decision aid strategy based on a decision maker preferences and experience using multiple attribute utility theory (MAUT). Finally, the best solution was implemented experimentally.

Optimized Production of Multilayered Monodisperse Polymer Nanoparticles.

Abstract A dynamic optimization frame work is used to produce in a controlled way multilayered latex nanoparticles. The key feature of the method is to track a glass transition temperature profile, which is designed to produce polymer layers with the targeted properties. Several constraints are considered to achieve better control and produce nanoparticles with a specified particle diameter and layers’ thicknesses. To enhance the control of the different layers,two separate monomer feeds are considered under starving conditions throughout the fed-batch stages. The emulsion copolymerization of styrene and butyl acrylate in the presence of n-C12 mercaptan, as chain transfer agent (CTA), is illustrated here as a case study. The optimal feed profiles of the pre-emulsioned monomers are obtained using a genetic algorithm.

Modeling and parameter estimation of coke combustion kinetics in a glycerol catalytic conversion reactor

A two dimensional, heterogeneous model was developed for the simulation of a coked catalyst regeneration fixed-bed reactor in a green acrylics process. A kinetic model which consists of combustion of coke containing carbon, hydrogen and oxygen was derived. The unknown parameters (including the kinetic parameters, initial composition and concentration of coke, heat transfer parameters and catalyst pore diameter) were identified using the results of regeneration experiments with the coke formed in a glycerol dehydration process. Good agreements between the model predictions and experimental results were obtained.