Simulated Moving Bed Unit Research Articles

Raising the Research Octane Number using an optimized Simulated Moving Bed technology towards greater sustainability and economic return

The scale of CO2 emissions from light-duty vehicle (LDV) fleets worldwide has led governments to mandate substantive improvements in vehicle fuel economy, thereby mitigating climate change. Raising the Research Octane Number (RON) of fuel through isomerization, alongside mandates to recalibrate existing LDV engines, promises to contribute substantially to climate action. This study has aimed to develop a highly efficient adsorption separation technology for isomerization refining that can contribute significantly to the sustainability and economics of the overall “well-to-wheel” outcome for LDV fleets globally. This study developed a rigorous dynamic model for a Simulated Moving Bed (SMB), comparing the SMB to conventional distillation as the next best alternative. The SMB was optimized using a genetic algorithm, maximizing RON and Gross Margin as objective functions. This study showed that SMB effectively separates high octane components from low octane components, producing a fuel with a RON of 95 when maximizing the RON, thereby enabling lower emissions associated with recalibrated LDV engines. Compared to 11 MW of steam duty associated with conventional distillation, the SMB unit utilized 3.4 MW and 5.7 MW of electricity when optimizing the RON and the Gross Margin respectively, appreciably reducing comparative greenhouse gas emissions. Finally, compared to conventional distillation as measured by Gross Margin, the optimized SMB unit increased the economic return by 47% when maximizing the RON and by 82 % when maximizing the Gross Margin. In summary, this study motivated for rapid capital investment into retrofitting isomerization facilities with SMB unit operations, replacing outmoded distillation as the primary separation technology. Future work will focus on optimizing the separation technology alongside the overall isomerization process using a rigorous techno-economic analysis as the objective function for optimization.

Strategies to reduce the extent of product dilution in a tandem simulated-moving-bed process for ternary separation without loss of throughput

One of highly-efficient separation processes for recovery of a target product in the pharmaceutical and fine chemical industries is a simulated-moving-bed (SMB) process. Although an SMB can ensure high separation capability and high throughput, it always has a problem of product dilution, which increases the energy consumption for post-processing of SMB product streams. This issue is even more important in a tandem SMB, which uses two subordinate SMB units in series for implementing a ternary-separation task. To address this issue, we sought to develop effective strategies for reducing the extent of product dilution in a tandem SMB without loss of throughput. For this task, a standing-wave-design (SWD) method was used to find a clue to reducing the extent of a tandem-SMB product dilution, which revealed the following results. First, for the considered issue, it could be effective to increase the column length of the SMB unit under the influence of column-efficiency limiting factor while decreasing the column length of the other SMB unit accordingly for the purpose of keeping total bed volume constant. It could also be effective to place more columns in the zones containing both edges of solute band for a target component. The two aforementioned strategies (called “S1” and “S2” respectively) were applied to the tandem SMB for recovery of neoagarotetraose from neoagarooligosaccharides. It was found that when the S1 method was utilized, the extent of product dilution in the tandem SMB was reduced by 56%. Furthermore, if the S1 and S2 methods would be used together, the extent of the tandem-SMB product dilution could be reduced by up to 85%. It is thus expected that the proposed strategies can contribute to a substantial improvement in the energy efficiency for post-processing of product stream from a tandem SMB process for ternary separation.

Optimization studies for improving the throughput and solvent usage levels of a tandem simulated-moving-bed process for recovery of galactotriose from crude galacto-oligosaccharides

The single-objective and multi-objective optimizations based on a multi-component standing-wave-design were carried out for the tandem simulated-moving-bed (SMB) process for continuous-mode separation of galactotriose (G3) from crude galacto-oligosaccharides. This process (named “G3-SMB”) consisted of two subordinate SMB units. The results from the G3-SMB optimization showed that its two subordinate SMB units had a significant difference in column efficiency, which acted as a major factor for restricting the level of attainable throughput and deteriorating solvent usage. This problem could be overcome by allowing the use of different column lengths within the G3-SMB (called “S1″ method), which was applied in a way that increased the column length of the subordinate SMB unit under the influence of column-efficiency limiting factor while decreasing the column length of the subordinate SMB unit with high column-efficiency. This method led to 32% increase in maximum achievable throughput and 37% reduction in solvent usage. It was also found that the adjustment of column configuration in each unit (called ”S2 method“) was as effective in improving throughput as the S1 method, but less effective in reducing solvent usage than the S1 method. Finally, it was confirmed that the simultaneous use of the aforementioned S1 and S2 methods could create a synergy effect, thereby resulting in 74% increase in maximum achievable throughput and 39% reduction in solvent usage.

A long short-term memory based Quasi-Virtual Analyzer for dynamic real-time soft sensing of a Simulated Moving Bed unit

The evaluation of unmeasurable quantities is an issue faced in several fields. One example is in the production of pharmaceuticals, in which, typically, isomer properties related to different effects are found: one enantiomer may kill while the other may cure. Thus, strict quality control is necessary in this field, which requires accurate, reliable and frequent measurements of the system. However, the measurement of the main quality properties associated with the production of some pharmaceuticals has a low frequency. Consequently, in addition to safety-related issues, financial losses are also related to these low frequency of measurements, as the product can easily run towards an out of spec production. In this scenario, the present work proposes a novel Deep Artificial Intelligence structure, which has an intrinsic (Nonlinear Output Error) NOE structure, associated with a (Nonlinear AutoRegressive with Exogenous input) NARX predictor, to be used as an online soft sensor in order to provide information about the main properties of a Simulated Moving Bed chromatographic unit, commonly used in the production of pharmaceuticals, in order to mitigate the low frequency of measurement associated with this unit. The proposed structure is here called Improved Quasi-Virtual Analyzer. The model can adapt itself as the process evolves, having the possibility of online learning through measurements obtained in the laboratory periodically. The proposed structure was tested in a software-in-the-loop scenario and compared with a more traditional alternative. These tests showed a robust capacity of the Improved Quasi-Virtual Analyzer to provide reliable predictions in real-time, as well as to outperform the traditional artificial network structure.

Global Approach for Simulated Moving Bed Model Identification: Design of Experiments, Uncertainty Evaluation, and Optimization Strategy Assessment

Simulated moving bed (SMB) chromatography is a widely used technique for the resolution of compounds difficult to separate. SMB parameter estimation is traditionally carried out following a time and money consuming series of experiments in an SMB unit where deviations may arise. This work aims to present a novel global and straightforward parameter estimation procedure together with uncertainty analysis. Particle swarm optimization (PSO) is employed to search for parameters in an eight-dimensional space, avoid local minima, and enable uncertainty analysis. The proposed methodology is validated in a software-in-the-loop approach. A new parameter estimation is then carried out using the data of one experimental run from the literature, together with uncertainty evaluation based on the PSO-generated population that enables model validation and definition of confidence regions for the model. A robust method for the parameter estimation of an SMB unit is presented in order to produce a more precise and reliable model. In addition, it also significantly reduces the number of necessary steps for parameter estimation, leading to a more efficient procedure. The results show that it is possible to perform parameter estimation from SMB chromatography producing a more trustworthy model.

Optimizing the Yield of a Pure Enantiomer by Integrating Chiral SMB Chromatography and Racemization. Part 2: Theory

We address the purification of the target enantiomer of a chiral compound from its racemic mixture, through simulated moving bed (SMB) chromatography either as a standalone or combined with the recycle and racemization of the undesired enantiomer. We carry out a comparative assessment of the two processes with focus on the role of the racemization kinetics, on the effect of the selectivity of the chiral stationary phase, and of the upstream symmetric synthesis of the racemate. The analysis is general thanks to the methods adopted for the optimal design of the integrated process (with racemization) and for the sizing of the SMB unit, as well as for the use of the four generalized Langmuir adsorption isotherms. In this way, we determine quantitative criteria that show that the standalone SMB process outperforms the integrated process when the performance of the racemization catalyst or the chiral stationary phase is poor and the racemic feed is inexpensive.

Optimal Design of SMB Units: A Novel Strategy Based on Particles Swarm Optimization

Abstract The Simulated Moving Bed (SMB) is a separation device whose use has been increasing, especially in separations that demand high purities like the chiral separation. Process Design is a topic often discussed in the literature and is usually limited to a sensitivity analysis. An optimal design of a SMB unit offers challenges due to its complexity on dynamics and numerical levels. Thus, there is a lack of a consistent methodology to optimize the SMB design. This work is focused on the SMB design which will be made using the Particles Swarm Optimization (PSO) method. For the first time, PSO will be used to choose the configuration, i.e., length of each section, of the True Moving Bed (the theoretical model of the SMB) unit for the case of the separation of the bi-naphthol enantiomers. The operating conditions, in terms of flowrates, will also be optimized. Globally, with the design strategy that was implemented, the system’s productivity is at least 30% higher than previous results reported in the literature for the TMB without optimization of the device configuration. The SMB that was designed from this configuration enables to increase the separation productivity at least 20% in comparison with previous results.

Optimization of a Simulated Moving Bed Unit within an Existing and Revamped Aromatics Complex with Crystallization and Toluene Methylation Units

The aromatics complex is a petrochemical facility where the main products are benzene and p-xylene. The xylene isomer is produced through an energy-intensive cycle loop in which the p-xylene separa...

Assessment of process configurations to combine enantioselective chromatography with enzymatic racemization

Enantioselective chromatography is nowadays a reliable tool for single enantiomer production from a racemate. The recovery of the distomer by racemization and recycling is a promising method to tackle the 50% yield constraint and to increase the productivity. In this paper three process configurations are compared. The production of enantiopure mandelic acid and methionine enantiomers exploiting different enzymes for racemization are evaluated as part of different chromatographic process configurations. First, the benefits of conventional simulated moving bed (SMB) chromatography in contrast to a single column batch separation unit are assessed in integrated configurations. Then, a concept of coupling the racemization with a simpler three-zone SMB unit, where one regeneration zone is removed, is evaluated.

Mathematical modeling and optimization of industrial scale ELUXYL simulated moving bed (SMB)

In this work, a detailed study on the modeling, simulation and optimization of the industrial-scale Eluxyl simulated moving bed (SMB) process are presented. Commercial SPX3000 (Ba-faujasite type zeolite) was used as an adsorbent, and para-diethylbenzene applied as a desorbent for separating para-xylene from other C8 aromatics isomers. In the first step, the xylenes and para-diethylbenzene adsorption equilibrium on the commercial adsorbent were measured at operating conditions (175 °C, 9 bar) experimentally. The density of particle and porosities were determined from commercial plant operation and adsorbent analysis. Furthermore, mass transfer coefficients, including internal (diffusion in the pore) and external (diffusion in the liquid film) resistance, have been investigated in details. A common mathematical model encompassing the SMB strategy, node model, extended Langmuir isotherm, and rate expression were constructed to implement a numerical simulation of the industrial-scale Eluxyl SMB process with and without considering backwash stream. The accuracy of the mathematical model was validated by the industrial data. Besides, a dynamic optimization framework with Sequential Quadratic Programming (SQP) optimization algorithm was used for obtaining optimal conditions. Two separate single objective optimizations by considering minimum desorbent consumption and maximum SMB unit productivity as objectives were firstly designated. By comparing the results, it was concluded that minimizing desorbent consumption was more favorable to be selected as an objective function, which could achieve 6.35% increase in productivity and 7.28% decrease in desorbent consumption. Finally, two-level optimization was planned for reaching maximum feed throughput with minimum desorbent consumption. The results showed that productivity could be increased by 13.75%, and para-diethylbenzene consumption could be reduced by 10.81% at the optimal condition.

Enhanced separation of bioactive triterpenic acids with a triacontylsilyl silica gel adsorbent: From impulse and breakthrough experiments to the design of a simulated moving bed unit

A simulated moving bed (SMB) unit was designed to separate oleanolic and ursolic acids, two naturally occurring triterpenoids with structural isomerism, with remarkable nutraceutical and pharmacological properties. A triacontylsilyl silica gel adsorbent (stationary phase of an Acclaim C30 column) was considered and impulse tests with different solvents were performed to select a mobile phase, from which methanol/water 95/5 (%, v/v) emerged as the most suitable. Equilibrium and global mass transport coefficients were then determined through breakthrough experiments using pure compound solutions and the C30 column. Afterwards, these parameters were applied to the simulation of two model binary mixture separations, whose breakthrough curves were also experimentally measured.Finally, the SMB unit was designed and optimized. It was demonstrated that using the packing of an Acclaim C30 column and methanol/water 95/5 (%, v/v) as mobile phase it is possible to separate both acids with purities of 99.9 wt.%, a productivity of 1.705 kg/(m3adsorbent day), and a configuration of two columns per section (2–2–2–2). The simulated results obtained in this work with the C30 stationary phase represent a significant improvement over literature data.

Chromatographic separation of betulinic and oleanolic acids

In this work, a simulated moving bed unit (SMB) was designed for the separation of betulinic and oleanolic acids, two naturally occurring triterpenic acids that exhibit key pharmacological properties. Preliminary impulse experiments were conducted to select appropriate mobile and stationary phases, from which methanol/acetonitrile 50/50 (%, v/v) and an Apollo C18 column emerged as the most favorable option. Breakthrough experiments were conducted with the pure compounds to determine the equilibrium and mass transfer coefficients necessary for SMB simulations. These parameters were then validated through the successful prediction of breakthrough experiments of two binary mixtures of betulinic and oleanolic acids.Optimum SMB operating conditions were determined via an optimization strategy combining the design of experiments and response surface methodologies (DoE-RSM) with phenomenological computer simulations. It was demonstrated that an SMB unit consisting of two columns per section allows the recovery of betulinic acid in its raffinate stream with a purity of 99.0 % and oleanolic acid in the extract stream with 99.4 % purity. Moreover, it was also demonstrated that the addition of acetonitrile to methanol is translated into a gain of 16.1 % in terms of productivity when compared to other mobile phase such as methanol/water 95/5 (%, v/v).

Modelling of the separation of long-chain normal paraffins from kerosene in a simulated moving bed process: effect of the desorbent

Linear paraffins can be selectively separated from the rest of components of kerosene (branched hydrocarbons, aromatics and naphthenes) by means of liquid phase adsorption on 5A zeolite using the technology of simulated moving bed (SMB). In previous works, the kinetic and equilibrium parameters required for modelling and design of the SMB unit were obtained for pure n-paraffins and n-paraffin mixtures. However, the simulation of the SMB process indicated the presence of n-C5, used as a desorbent, in the separation zone, especially after the feed mixture is introduced. This finding motivated this work, in which n-paraffin mixtures (n-C10, n-C12, n-C14) including n-C5 were studied to address its influence in the process. The kinetic and equilibrium parameters for these mixtures were obtained and included in the model for the simulation of an SMB unit. While mixtures without n-C5 preferentially adsorbed shorter n-paraffins, it was found that including n-C5 in the mixtures reverses the selectivity of the adsorbent. In this case, longer n-paraffins are preferentially adsorbed, matching the trend observed for pure n-paraffins. In addition, n-C5 significantly increases the mobility of n-paraffins, as indicated in their higher mass transfer coefficients. The model was validated by comparing the predicted performance with the reported separation achieved by a commercial SMB unit that separates n-paraffins from hydrotreated kerosene fractions. The predicted separation performance is very similar to that achieved in our previous works, slightly improving the purity (99.6%) of the extract as a trade for a small loss in recovery (95.4%).

Optimization of a True Moving Bed unit and determination of its feasible operating region using a novel Sliding Particle Swarm Optimization

The optimization of Simulated Moving Bed units has become an important topic to be developed in this field. The present work proposes a novel methodology to optimize a True Moving Bed unit and draw the feasible operating region through a novel improved Self-Organizing Hierarchical Particle Swarm Optimization with Time-Varying Acceleration Coefficients and Mutable Searching Region, here called Sliding Particle Swarm Optimization, concomitantly with an adapted method to define the feasible operating region. Two main contributions can be highlighted, first the process optimization with its feasible operational regions, leading to better results when compared with the numerical optimization based on the equilibrium theory, second a new particle swarm method is presented which can deal more efficiently with multi local minima problems. Finally, the concept of feasible operating region concept is presented as a support tool for the process operation. Three different operational scenarios were simulated here in order to verify the consistence and efficiency of the methodology. The main advantage of the methodology here proposed is the possibility to tracking all the possible operating regime of the unit while meeting a given process requirement.

Modelling of the separation of normal paraffins from kerosene fractions by a simulated moving bed process

Linear paraffins are widely used in the manufacturing of industrial and domestic detergents. Some adsorbents selectively separate these linear hydrocarbons by adsorption from petroleum feedstocks. LTA molecular sieves (5A zeolite) adsorb linear paraffins while excluding the rest of the components of kerosene (branched hydrocarbons and aromatics). Equilibrium and kinetic parameters are available in the literature for light paraffins in the vapour phase, however, there is scarce information concerning high molecular weight paraffins in liquid phase, especially at the operating conditions of commercial processes. In a previous work, the equilibrium and kinetics of high molecular weight n-paraffins C5, C10, C14 and C18 were studied for the adsorption in liquid phase on 5A zeolite. The aim of this work is to study the equilibrium and kinetics of n-paraffins C12 and C16, as well as mixtures of n-paraffins C10, C12 and C14 in the same conditions. n-pentane has been included in the study as it is mainly used as desorbent in the cyclic simulated moving bed (SMB) commercial process. Pure component isotherms were obtained, as well as a multicomponent isotherm. By comparing them, it was observed that selectivities are significantly lower in mixtures (for example, selectivity towards C14 with respect to C12 is lowered from 2.84 for pure paraffins to 1.05 for mixtures). A theoretical model has been developed to describe the column adsorption dynamics of the studied systems. The model has been included in an SMB simulation program (SMBSIM), and the model prediction has been validated by comparison with the separation performance data reported for a commercial SMB unit that separates normal paraffins from a hydrotreated kerosene fraction The model predicts the separation of linear paraffins with 99.2% purity and 96.3% recovery (5% error obtained for n-paraffin concentration in the extract and non-adsorptives in the raffinate).

Design and optimization of a simulated moving bed unit for the separation of betulinic, oleanolic and ursolic acids mixtures: Experimental and modeling studies

Abstract Betulinic, oleanolic and ursolic acids are naturally occurring triterpenic acids that have attracted considerable interest due to their nutraceutical and pharmacological properties. These compounds can be extracted from natural sources, however, their simultaneous occurrence and very similar structures make their separation a challenging task. In this work we designed a simulated moving bed (SMB) unit for the separation of a representative natural extract containing betulinic, oleanolic and ursolic acids into high purity compounds using a two-step process: firstly, betulinic acid was isolated from oleanolic and ursolic acids, and secondly oleanolic and ursolic acids were fractionated. Preliminary HPLC experiments were conducted to select appropriate mobile and stationary phases. Equilibrium and mass transport parameters were determined through breakthrough experiments with pure compounds in a single column. Subsequently this information was successfully applied in the simulation of a ternary mixture separation, whose results were validated with ternary breakthrough measurements. Finally, the SMB was designed and optimized using a Design of Experiments approach combined with Response Surface Methodology (DoE-RSM), and simulated using a phenomenological rigorous model. It was concluded that using the designed SMB unit it is possible to produce betulinic, oleanolic and ursolic acids with purities of at least 99.4%, 99.1%, and 99.4% (all mass fraction), from a feed mixture containing 20%, 25%, and 55% of each acid, respectively.

A systematic simulation and optimization of an industrial-scale p-xylene simulated moving bed process

Abstract This work presents a detailed study for the systematic analysis, simulation and optimization of an industrial-scale simulated moving bed process with Ba/K-exchanged faujasite-type zeolite as adsorbent for separating p -xylene from the C 8 aromatic mixture. The adsorption equilibrium of C 8 aromatic isomers and p -diethylbenzene on the commercial adsorbent were determined experimentally, while the lumped mass transfer coefficients were estimated through pulse experimental results. A generic mathematic model incorporating the fixed model, node model as well as dead volume model was developed to implement a numerical simulation of an industrial-scale p -xylene simulated moving bed process. Moreover, the accuracy of mathematic model was validated by industrial data. Furthermore, a dynamic optimization framework, with state of the art Sequential Quadratic programming optimization algorithm, was firstly proposed to obtain the optimal operating conditions for an existing industrial simulated moving bed unit. Optimization results reflect that the desorbent consumption of p -diethylbenzene was favorable to be selected as objective function to be minimized, which could achieve 3.12% increase in SMB unit productivity and 8.66% decrease in desorbent consumption for seven zone p -xylene simulated moving bed process. Two-level optimization procedure was also constructed to get the maximum feed throughput with a desired minimized desorbent consumption. Under the optimal operating conditions, the productivity increases by 16.63% while the PDEB consumption decreases by 18.85% for seven zone p -xylene simulated moving bed process. Eight zone p -xylene simulated moving bed process as a new operating mode, which was also analyzed, emulated and optimized through numerical calculation. Results show that the eight zone operation could achieve a much higher productivity and the desorbent consumption of p -diethylbenzene was almost the same in comparison with seven zone operation.

Dynamics of a True Moving Bed separation process: Linear model identification and advanced process control

The control of Simulated Moving Bed (SMB) units is challenging due to their complex dynamic behaviour and the difficulty of measuring their main properties. Furthermore, for the SMB units, the transfer function identification when the unit is operating at its optimal point is not easy to be done through the usual way. This work presents the development of a novel strategy to identify transfer functions of TMB/SMB and its application on classical linear model predictive controllers (MPC). However, for the process in study, due its unique dynamics, only the identification of the linear model is not enough to solve its control problem. Therefore, it is proposed a modification in the MPC prediction, that consists in a strategy based on a switching system where the most adequate transfer function is employed in the controller to overcome the problems related with the process dynamic behaviour. The results show that the used methodology enables the easy identification of transfer functions at the process optimal operating point and that the MPC can control the process in both the servo and regulator problem cases. It is also showed that the transfer function identified can be applied in the control of a SMB unit with four columns, under its optimal conditions.

Dynamic response to process disturbances—A comparison between TMB/SMB models in transient regime

The modelling and design of Simulated moving bed (SMB) processes is normally done using the True moving bed (TMB) approximation. Several studies show that average values obtained at cyclic steady state for SMB units approach the TMB unit at steady state and that this approach is better as the number of columns in the SMB increases. However, studies that evaluate this equivalence under dynamic conditions are scarce. The objective of this work is to perform an analysis of the transient behaviour of two SMB units, with four and eight columns, and compare the results with the ones obtained for a TMB unit. An analysis of the impact of operating variables on the processes performance parameters is performed. The results show that TMB/SMB equivalence is valid only for conditions that do not violate the regeneration/separation regions and that the transient behaviour of the four columns SMB can resemble more the TMB.

Improvement of the performances of a tandem simulated moving bed chromatography by controlling the yield level of a key product of the first simulated moving bed unit

One of the trustworthy processes for ternary separation is a tandem simulated moving bed (SMB) process, which consists of two subordinate four-zone SMB units (Ring I and Ring II). To take full advantage of a tandem SMB as a means of recovering all three products with high purities and high economical efficiency, it is important to understand how the separation condition in Ring II is affected by that in Ring I, and further to reflect such point in the stage of designing a tandem SMB. In regard to such issue, it was clarified in this study that the Ring I factors affecting the Ring II condition could be represented by the yield level of a key product of Ring I (YkeyRing I). As the YkeyRing I level became higher, the amount of the Ring I key-product that was reloaded into Ring II was reduced, which affected favorably the Ring II separation condition. On the other hand, the higher YkeyRing I level caused a larger dilution for the stream from Ring I to Ring II, which affected adversely the Ring II separation condition. As a result, a minimum in the desorbent usage of a tandem SMB occurred at the YkeyRing I level where the two aforementioned factors could be balanced with each other. If such an optimal YkeyRing I level was adopted, the desorbent usage could be reduced by up to 25%. It was also found that as the throughput of a tandem SMB became higher, the factor related to the migration of the Ring I key-product into Ring II was more influential in the performances of a tandem SMB than the factor related to the dilution of the stream from Ring I to Ring II.