During the past decades, reactive distillation has received intensive attention due to the well known benefits of integrating distillation with reaction in a single unit. Significant capital savings, improved conversion and selectivity, avoidance of azeotropes, together with heat integration are some of the main advantages of using reactive distillation. Many applications have proven to be economically advantageous by using reactive distillation (e.g. MTBE and TAME synthesis, production of methyl-acetate, manufacture of di-isopropyl-ether, oligomerisation of linear butenes and others). Whereas there are many procedures available for the synthesis of non-reactive columns or reactive-separation systems, the synthesis of reactive distillation columns is still a challenge, due to the complexity and the high number of design parameters involved. Available conceptual design methods generally address three (or four) components and fully reactive columns, but there is still a lack of systematic conceptual design methods for more general column configurations and for multi-component systems. The aim of this work is to develop a methodology to identify promising column configurations and to obtain column design parameters (number of reactive and non-reactive stages, reflux and reboil ratios, feed condition) for a given feed mixture and a set of desired products. A new systematic design method for reactive systems reaching equilibrium allows the analysis of the impact of different configurations (fully reactive or hybrid columns) and feed policies (single- or double-feed columns) on column performance. The methodology is extended to account for kinetically-controlled reactions in synthesis and design of reactive distillation columns. Systems with two degrees of freedom (according to the Gibbs phase rule) were considered for equilibrium reactions, and ternary and quaternary systems for kinetically-controlled reactions. Reactive distillation column designs generated by the methodology are presented as illustrative examples. Their predicted performances are shown to be in good agreement with those predicted by rigorous simulation using HYSYS. The approach can easily be automated and typically generates multiple designs, allowing a design engineer to efficiently compare various design options including hybrid and fully reactive columns, single- and double-feed configurations, and different sets of operating parameters for a given column configuration. The new methodology developed in this work facilitates a stepchange in conceptual design practice, offering a systematic and easy to use tool for the synthesis and design of reactive distillation columns.

Reactive distillation is a combination of a traditional multi-stage distillation column with a chemical reaction. The primary benefits of a reactive distillation process are reduced capital costs for equipment and energy in addition to enhanced conversion for equilibrium-limited reactions. One such equilibrium-limited reaction is an olefin metathesis. Olefin metathesis is a catalyzed reaction that breaks the double bond in olefins and rearranges the alkene fragments into new olefinic products. A comprehensive investigation of a reactive distillation based olefin metathesis and supporting experimentation is documented here. A small pilot plant study was performed for pilot scale performance comparison. Bench reactor experimentation was conducted for the purposes of learning detailed information on specific metathesis reactions. Lastly, a process simulation study was completed for comparison in performance with the small pilot plant process. The small pilot plant study involved the design, construction, testing, operation, and optimization of a reactive distillation column. Continuous operation campaigns at two different hydraulic capacities within the reactive zone were performed and their performances were compared. A higher hydraulic capacity proved to be more efficient and more selective for the conversion of medium molecular weight olefins into both lighter and heavier olefinic products. Bench reactor experiments were designed with the intent of investigating specific alpha olefin metathesis reactions and obtaining conversions, selectivities, and yield structures for future simulation work. However, under conditions similar to that within the small pilot plant process, there existed a high frequency of secondary double bond isomerization (possibly due to an isomerization activity for alumina). There was also an observed dependence on temperature for both the primary metathesis and secondary isomerization reactions. A process simulation representative of the small pilot plant process was constructed in AspenPlus. Using a simplified reaction network based on assumptions and analysis of the reactive zone, its performance was compared with that of the small pilot plant process. The simulation performance tended to underpredict overhead compositions, but accurately simulated the bottoms product composition. Because reactive distillation has not been used with a heavy olefin metathesis reaction, this dissertation demonstrates the uniqueness and effectiveness of a reactive distillation based heavy olefin metathesis. / text

The production of 7-tetradecene was examined. Properties for this compound were estimated using group contribution methods and compared to experimental data. Process simulation was used as a tool to identify competitive processing strategies. For reactive distillation, three different models were compared to determine the model complexity needed to describe the process: Model A, with the assumption of physical and chemical equilibrium; Model B, with kinetics described by a second order reaction and physical equilibrium; and Model C, a non-equilibrium stage model that accounts for mass transfer. A conceptual design was obtained with Model B and was checked with Model C, which described the process more accurately but was more difficult to converge. Since, Model A was easier to converge, it was used to predict process conversions at different pressures. Predictions favor working at 1 bar, due to the lower heat duty and the minimum stages required.

Reactive distillation has enormous potential for the economical synthesis of tertiary ethers. Methyl tert-butyl ether (MTBE) has been commercially produced with this technology since the early 1980s and it appears that the process also has application for Ethyl tert-butyl ether (ETBE) and other ethers. However, the combination of reaction and distillation in a single unit operation produces a process complexity that inhibits expeditious design and tight control, and presents a technology risk for potential developers. This particularly applies to hybrid reactive distillation where both reactive and non-reactive column sections are employed.The steady state simulation of a series of reactive distillation columns and processes for the production of ETBE and MTBE illuminated a number of important issues related to the optimal design techniques. Many of these issues are peculiar to reactive distillation and would not reasonably be anticipated without a priori knowledge of the phenomena involved. For example, the addition of theoretical equilibrium stages and an increase in the reflux ratio do not always have a directionally equivalent effect. The trade-off between energy consumption and capital cost which is the basis for most distillation designs cannot always be applied to reactive distillation. Importantly, the use of standard modelling techniques for equilibrium processes was also validated for reactive distillation design.The use of residue curve diagrams and reactive residue curve diagrams for the design of reactive distillation processes was investigated and shown to provide useful information regarding the feasibility of reaction-separations. Combined with simulation tools (e.g. Pro/II and SpeedUp), these techniques form the basis of a proposed design strategy for hybrid reactive distillation. It is important to apply these design tools appropriately and to ++ / select the correct process for a given application. The optimal design must also consider economics and the relative values of products, reactants and energy. These issues were studied with respect to ETBE production for gasoline oxygenation.The complexity of hybrid reactive distillation not only presents design challenges but potentially makes the process more difficult to control. Dynamic simulations of ETBE and MTBE reactive distillation processes were used to explore some unusual dynamic phenomena and to elucidate the process non-linearity and bidirectionality of reactive distillation. The presence of multiple steady states for some reactive distillation columns has been documented previously but the analysis of this behaviour has been incomplete and somewhat flawed. It was shown that the distinction between molar inputs and physically realisable mass or volumetric inputs is crucial and that multiplicity could be present in one case and not in the other. Multiplicity that is only observed with molar inputs (relatively common) was termed pseudo-multiplicity. Pseudo-multiplicity has few implications for the operation and control of practical reactive distillation processes although most literature examples of multiple steady states fall into this category. Four distinct causes of output multiplicity were identified including one new cause, reaction hysteresis, which is only applicable to hybrid reactive distillation. It was shown, using dynamic simulations, that transitions between parallel steady states are possible for a range of physically realisable and practical disturbances. This contrasts with other work in the area, which examines only unrealisable events and control schemes.An extensive analysis of reactive distillation control was also undertaken with respect to ETBE and MTBE hybrid columns. Manual (open-loop) control was shown to be impractical due to ++ / the need to sustain the operating conditions at close to the optimal values in order to produce acceptable process performance. One-point composition control was found to be relatively easy to implement and effective with either an energy-balance or a material-balance control scheme provided only one steady state was present. Where multiple steady states exist, there are restrictions on the feasible control structures due to unavoidable instability in the inventory controllers. For example, if multiple steady states exist for the one value of the reboiler duty, only the bottoms product draw rate can be used to control the reboiler sump level. Thus, a material-balance control structure that uses the reboiler duty to control the sump level cannot be implemented in practice. Two-point control was also investigated and found to effectively prevent transitions between parallel steady states. Although more complex and difficult to implement than one-point control, a two-point scheme could be used successfully to control both the product composition and the reactant conversion and this could be desirable in some cases.A reactive distillation pilot plant was designed and operated for ETBE synthesis from ethanol and a locally available refinery hydrocarbon stream. The design of the pilot plant was based on simulation studies and the objective of operating in the industrially significant ranges of product purity and isobutene conversion. A fully automatic control system was designed and installed on the pilot plant to permit precise control of the manipulated variables and the framework to implement a range of control structures and schemes.Keywords: reactive distillation; process simulation; process design; process control; dynamic simulation; multiplicity; bidirectionality; distillation control; inferential control; pilot plant design and operation.

No description available.

Yes / Batch reactive distillation is well known for improved conversion and separation of desired reaction products. However, for a number of reactions, the distillation can separate the reactants depending on their boiling points of them and thus not only reduces the benefit of the reactive distillation but also offers operational challenges for keeping the reactants together. Methyl lactate (ML) synthesis via the esterification of lactic acid (LA) with methanol in a reactive distillation falls into this category and perhaps that is why this process has not been explored in the past. The boiling points of the reactants (LA, methanol) are about 490 K and 337 K while those of the products (ML, water) are 417 K and 373 K respectively. Clearly in a conventional reactive distillation (batch or continuous) methanol will be separated from the LA and will reduce the conversion of LA to ML significantly. Here, first the limitations of the use of conventional batch distillation column (CBD) for the synthesis of ML is investigated in detail and a semi-batch reactive distillation (SBD) configuration is studied in detail where LA is the limiting reactant and methanol is continuously fed in excess in the reboiler allowing the reactants to be together for a longer period. However, this poses an operational challenge that the column has to be carefully controlled to avoid overflow of the reboiler at any time of the operation. In this work, the performance of SBD for the synthesis of ML is evaluated using model based optimization in which operational constraints are embedded. The results clearly demonstrate the viability of the system for the synthesis of ML.

Dynamic models of a reactive distillation column have been developed and implemented in this work. A model describing the steady state behavior of the system has been built in a first step. The results from this steady state model have been compared to data provided from an industrial collaborator and the reconciled model formed the basis for the development of a dynamic model. Four controlled and four manipulated variables have been determined in a subsequent step and step tests for the manipulated variables were simulated. The data generated by the step responses was used for fitting transfer functions between the manipulated and the controlled variables. RGA analysis was performed to find the optimal pairing for controller design. Feedback controllers of PID type were designed between the paired variables found from RGA and the controllers were implemented on the column model. Both servo and regulatory problems have been considered and tested.

This work presents a modeling and control study of a reactive distillation column used for hydrogenation of benzene. A steady state and a dynamic model have been developed to investigate control structures for the column. The most important aspects of this control problem are that the purity of the product streams regarding benzene need to be met. At the same time as little toluene as possible should be converted. The former is a constraint imposed by EPA regulations while the latter is tied to process economics due to the high octane number of toluene. It is required to satisfy both of these objectives even under the influence of disturbances, as the feed composition changes on a regular basis. The dynamic model is used for developing transfer function models of two potential control structures. Pairing of inputs and outputs is performed based upon the Relative Gain Array (RGA) and PI controllers were designed for each control structure. The controller performance was then compared in simulation studies. From our results, control structure 2 performed better than control structure 1. The main advantage of CS2 over CS1 is noticed in the simulation of feed composition disturbance rejection, where CS2 returns all variables back to steady state within 3 hrs while it take CS1 more than 20 hrs to return the temperature variables back to steady state.

Dynamic models of a reactive distillation column have been developed and implemented in this work. A model describing the steady state behavior of the system has been built in a first step. The results from this steady state model have been compared to data provided from an industrial collaborator and the reconciled model formed the basis for the development of a dynamic model. Four controlled and four manipulated variables have been determined in a subsequent step and step tests for the manipulated variables were simulated. The data generated by the step responses was used for fitting transfer functions between the manipulated and the controlled variables. RGA analysis was performed to find the optimal pairing for controller design. Feedback controllers of PID type were designed between the paired variables found from RGA and the controllers were implemented on the column model. Both servo and regulatory problems have been considered and tested.

A reactive distillation column for the alkylation of benzene with long chain linear olefin was studied. The study involved design, construction, experimentation, and simulation of the column. Establishing the design required study of reaction rates, thermodynamic relationships, and packing structures. A heuristic was developed for the design of such columns. This heuristic involved estimating an amount of catalyst loading and subsequently determining the operating parameters for a column. This method is particularly applicable to systems with high concentrations of inert feeds. A column was constructed following the design. Data was collected from the column and compared to simulations. The simulations were performed with Aspen Plus RADFRAC. In this manner, the data was used to validate the commercial steady state models for reactive distillation. In addition, dynamic simulations of the system were performed. These dynamic simulations provided insight into more design considerations. For example, steady state simulations indicated an optimal feed stage based on steady state conversion of the olefin. However, the dynamic simulations showed a potential disadvantage to the utilization of the optimal feed stage. With some disturbances, a column configured with the feed stage with the highest steady state conversion also deviated from the steady state faster and with greater amplitude than other configurations. These considerations were further explored in developing a control scheme for reactive distillation columns. Control of reactive distillation differs from traditional distillation in that one control variable is conversion. Traditional distillation generally focuses on production rates and product purity. To this end, control schemes were analyzed and dynamic simulations were performed. These simulations showed an advantage to a variable pairing in which duty is paired with conversion. The conversion was inferred from a stage temperature in the reactive zone. In addition, distillate rate may be paired with product composition. In conclusion, the reactive distillation column design for long chain olefin alkylation of benzene requires careful estimation of catalyst requirements and valid simulation tools. In addition, dynamic response should be considered in the design. Finally, a simple inferential control scheme may be adequate. / text