José Eduardo Alves Graciano

Improvements in surrogate models for process synthesis. Application to water network system design

Abstract High accuracy models can be obtained by using different types of surrogate models that accurately approximate equipment phenomenological models and can be used in synthesis problems, leading to faster and more precise solutions. Two types of surrogate models are used to approximate equipment phenomenological models: polynomial and neural network-based. In some cases, these surrogate models are not able to represent more complex equipment. An original methodology to reformulate these models using equations from shortcut equipment design is proposed. A medium-size case study involving fifteen units is presented. The synthesis problem is solved in a short computational time, leading many local solutions. Since several local optima objective function values are very close to each other, the choice of the best configuration among those found should be done qualitatively, because the differences among the objective function values are not significant if compared to the accuracy of equipment cost correlations in the literature.

Adsorption of Zinc Ions from Water Using Zeolite/Iron Oxide Composites

The adsorption characteristics of zeolites synthesized from fly ash were combined in a composite with the magnetic properties of iron oxides to produce adsorbents which were magnetic materials. Such zeolite/iron oxide magnetic composites were prepared with weight ratios of 3:1, 2:1 and 1:1. The experimental data for the equilibrium adsorption isotherms of Zn2+ ions onto the composites were modelled using the Freundlich and Langmuir equations. The presence of iron oxide had no significant effect on the adsorption capacities of the magnetic composites. The experimental data were also employed to determine the kinetic characteristics of the adsorption process. The adsorption of Zn2+ ions was found to follow pseudo-second-order type kinetics. Although intra-particle diffusion occurred in the adsorption processes, it could not be accepted as the primary rate-determining step. The evaluated thermodynamic parameters indicated that the adsorption of Zn2+ ions onto zeolite/iron composites was spontaneous and endothermic.

Performance comparison of parameter estimation techniques for unidentifiable models

Abstract Four different estimation approaches exploiting sensitivities, eigenvalue analysis (rotational discrimination and automatic parameter selection and estimation), reparameterization via differential geometry and the classical nonlinear least squares are assessed in terms of predictivity, robustness and speed. A Monte Carlo methodology is adopted to evaluate the statistical information required to quantify the inherent uncertainty of each approach. The results show that the rotational discrimination method presents the best characteristics among the evaluated methods, since it requires less a priori information than the reparameterization via differential geometry, uses simpler stop criteria than the automatic selection, reduces the overfitting caused by the nonlinear least squares solution and because it estimates parameters with the best predictivity among the methods tested. Additionally, results suggest that assessing the goodness of the estimated parameters solely in the calibration set can be misleading, and that the statistical information obtained from a validation set is more valuable.

Decomposition Techniques for the Real-time Optimization of a Propylene Production Unit

Abstract In this work an application of decomposition techniques on a propylene production unit from REPLAN refinery by Petrobras S.A. is presented. It is shown that the classic Lagrangian Relaxation (LR) technique and the alternative technique called “Pricing Interprocess Streams Using Slack Auctions” (PISUSA) do not converge for the case study presented. The issues involved in each decomposition approach are identified and discussed; then a modification of the Lagrangean Relaxation algorithm is proposed using a new constraint-updating rule. This modified algorithm is able to overcome the issues and solve the decomposed problem properly.

Practical Implementation of a RTO in an Industrial-Scale Vapor Recompression Distillation Process

Abstract In this work a practical implementation of a real time optimization (RTO) in an industrial scale propylene-propane distillation process (Refinaria de Paulinia, SP, Brazil) owned by Petrobras S.A. is carried out. The main steps of a classical RTO cycle are considered: steady-state identification, parameter estimation, data reconciliation and economical optimization. The non-identifiability problems, observed in the parameter estimation module, are mitigated by using the Rotational Discrimination (RD) method (Graciano et al., 2014). The outcomes of this study suggest that the model parameters are successfully estimated by the RD method and the RTO is able to improve the economical gains between 3-26%, depending on the initial condition, which represents saves up to 19 million dollars per year.

Removal of Zn2+, Cd2+ e Pb2+ ions from aqueous solutions by magnetic composite of zeolite from coal ashes

For this study, magnetic composite of zeolite-magnetite was prepared by mixing magnetite nanoparticles suspension with synthetic zeolite. The nanoparticles in suspension were synthesized by precipitating iron ions in a NaOH solution. The zeolite was synthesized from coal fly ash by alkaline hydrothermal treatment. The magnetic composite was characterized by XDR, SEM, magnetization measurements, IR, and BET surface area. Batch tests were carried out to investigate the adsorption of metal ions of Zn2+, Cd2+ and Pb2+ from aqueous solution onto magnetic composite. Adsorption isotherms were analyzed using Freundlich and Langmuir equations. The adsorption equilibrium data fitted well to the Langmuir equation with maximum adsorption capacities in the range of 28.5-127 mg g-1.