Packed Distillation Column Research Articles

Design, conception, and assessment of an innovative liquid distributor for separation column

Liquid distributor is an important internal in packed distillation columns. One of the key features of this equipment is the drip point density, which corresponds to the number of irrigation points per area of the cross-section. For gravity liquid distributors, increasing the number of irrigation holes generally requires small hole diameters, which makes the equipment more sensitive to plugging. In this work, we disclose a new gravity liquid distributor that achieves large drip point densities without requiring smaller irrigation holes. This distributor is an adaptation of the classic orifice-pan type, in which we add a tree-like structure made of wires to spread the liquid dripping from the orifices. The liquid flows outside the wires and splits several times over the different branching levels before falling on the packed bed. From an orifice-pan configuration with drip point density of 736 pts/m2, we designed and constructed enhanced distributors with a theoretical irrigation density of 14,800 pts/m2. Performance comparison using a recently disclosed structured packing shows that these enhanced distributors can decrease HETP from 0.45 m to 0.27 m.

Fluid dynamic and separation performance of foam loaded packed distillation columns

Foam formation in distillation columns leads to various negative effects such as reduced distillation throughput and reduced separation performance. In order to determine the influence of foam on distillation throughput and separation performance, measurements with chlorobenzene/ethylbenzene (CB/EB) and water/monoethyleneglycol (water/MEG) mixtures in a DN50 laboratory distillation setup were carried out. As internals, structured packings were compared to randomized packings with different materials. Results show that structured packings were more susceptible to foam and had greater losses in separation performance under the influence of foam as compared to randomized packings. However, in absolute terms, foam-loaded structured packings showed the highest separation efficiencies compared to randomized packings. An increase in column wall temperature above adiabatic conditions resulted in an increased foam loading and a decreased separation performance. Pressure drops were similar between packings, but foam loaded randomized packings showed a slightly lower pressure drop.

On Fractioning the Tire Pyrolysis Oil in a Pilot-Scale Distillation Plant under Industrially Relevant Conditions.

Tire pyrolysis oil (TPO) is one of the most interesting products derived from the pyrolysis of end-of-life tires. Among others, it contains valuable chemicals, such as benzene, toluene, ethylbenzene, and xylene (BTEX), as well as limonene. In order to recover these chemicals, a pilot-scale distillation plant has been designed, erected, and operated using TPO derived from an industrial-scale pyrolysis plant. The distillation facility consists of a packed column (20 kg/h) and is within the fifth technological readiness level. This work describes for the first time the fractioning of the TPO in a continuous operational mode under industrially relevant conditions. For this purpose, different reboiler temperatures (250-290 °C) and reflux ratios (up to 2.4) were preliminarily assessed on the yields and properties of the resulting products: light fraction (LF) and heavy fraction (HF). Thus, the distillation plant is capable of producing 27.0-36.7 and 63.3-73.0 wt % of LF and HF, respectively. The highest BTEX concentration in the LF (55.2 wt %) was found using a reboiler temperature of 250 °C and a reflux ratio of 2.4. Contrarily, the highest limonene concentration (4.9 wt %) in the LF was obtained at 290 °C in the reboiler without reflux. In this sense, the lower the reboiler temperature, the higher the BTEX, and the lower the limonene concentration in the LF. The main results herein obtained serve to gain key insights to operate packed distillation columns using complex and promising hydrocarbons as TPO in order to recover valuable products. In addition, this work provides significant information for optimizing the recovery efficiencies of both BTEX and limonene, as well as their potential applications including that for the resulting HF.

Dynamic Optimization Study for Cryogenic Distillation in Hydrogen Isotope Separation System

This paper presents a dynamic optimization study on a cryogenic distillation unit in hydrogen isotope separation system. A packed distillation column with an equilibrator is considered. The packed column was modeled as a theoretical equilibrium-based stage column. The equilibrator was placed to enhance separation performance. An optimal control problem was formulated to minimize the tritium holdup under two product quality constraints, in a situation where feed flowrate is periodically fluctuated. An optimal control policy was obtained for two manipulating variables, namely distillate flowrate and reboiler heat duty. The optimal control policy was analyzed by examining the operating variable profiles of the column.

Program for calculating the technological and structural characteristics of plate and packed distillation columns for separating two, multicomponent mixtures and isotopes

The article describes the developed and software implemented methods for calculating the technological and structural characteristics of plated and packed distillation columns for separating two-component, multicomponent and isotopic mixtures. The results of the verification of the methods are presented. A good convergence of the calculation results with the known literature data is shown.

Modelling of a continuous distillation process with finite reflux ratio using the hydrodynamic analogy approach

The hydrodynamic analogy (HA) approach offers an alternative to stage models describing packed distillation columns. So far, HA models have been successfully applied to separation processes, in which the properties of fluid streams entering the packing bed are pre-set as input parameters. However, in a continuous distillation process, the internal fluid streams are coupled via a reboiler and a condenser, thus making the inlet compositions a part of the solution.In the present work, an extension of the HA approach for the description of a continuous distillation process with finite reflux ratio is presented. Equations describing the relationships between the inlet and outlet streams at the column top and bottom are incorporated in the model. Furthermore, changes in compositions and flow rates induced by the feed stream are considered. The model is verified by comparison with experimental data on distillation of a methanol/ethanol/water mixture in a column of 2.3m height and 210mm diameter. The simulated composition profiles are in good agreement with the experiments.

CFD study of oscillatory motion effect on liquid flow distribution into structured packed divided wall column

Packed distillation columns are key equipment in natural gas pre-treatment process within LNG production process train operated at offshore. In this work, the impact of offshore operating conditions and that of divided wall column shape were explored via developing transient three-dimensional CFD model for a divided wall distillation column filled with structured packing. Inter-phase drag force, porous resistance force, mass transfer, and liquid dispersion were incorporated into volume averaged equations to map the distortion in liquid distribution in response to column oscillatory motion. The results show that introducing a dividing-wall into distillation column affects fluids partition and distribution. Column motions induce complex fluid flows in radial direction that result in oscillatory patterns around the steady-state solution of vertical column. Furthermore, liquid distributions in a divided wall column under influence of column motion differ with respect to column system positions throughout rotational axes, mass transfer also fluctuates according to column oscillations. A position with a dividing wall making an angle of 45˚ with rotational axis gives a scenario in which rolling motion resembles to pitching motion. The thereof position leads to moderate liquid maldistribution in comparison to liquid maldistribution from scenario with dividing wall parallel to rotational axis.

Dynamic Modeling of a Sub‐zero, Fed‐Batch, Packed Distillation Column to Produce Tetrafluoroethylene

AbstractEquilibrium and rate‐based dynamic models are proposed for a laboratory‐scale, fed‐batch, packed distillation column that operates at sub‐zero temperatures. The purpose of the column is to separate tetrafluoroethylene from a ternary mixture that also contains hexafluoropropylene and octafluorocyclobutane. The simulation results for both models indicate second‐order type responses, with either increases or decreases from the expected steady‐state values. It is recommended to build a start‐up routine into the simulations. Experimental validation should also form part of future work.

Моделирование структуры потоков в насадочной ректификационной колонне по профилю скорости паровой фазы

An algorithm has been developed for calculating the main parameters of the flow structure by the velocity profile of the boiling liquid phase in a packed distillation column: differential and integral response and dispersion functions without determining the response curves by the indicator method. An example of the calculation of the above parameters and modeling according to standard and combined models is given: cell, one-parameter diffusion and with a series connection of zones of displacement and mixing.

Experimental Study on Mass Transfer in a Packed Distillation Column

Experimental Study on Mass Transfer in a Packed Distillation Column

Investigating azeotropic separation of hydrochloric acid for optimizing the copper-chlorine thermochemical cycle

In this paper, an atmospheric-pressure distillation system is designed and constructed for partial to separation of hydrochloric acid and water. The system concentrates HCl(aq) between the electrolyzer and hydrolysis processes of the Copper–Chlorine (Cu–Cl) cycle for hydrogen production. The motivation behind this study is to investigate azeotropic separation of HCl(aq), as needed for integration of unit operations in the Cu–Cl cycle. The separation is only partial, as the mixture is unable to cross the azeotrope with only a single pressure. The distillation system consists primarily of one packed distillation column, which employs heating tapes and thermocouples to achieve a desired axial temperature profile. The column can be operated in batch or continuous mode. The distillate is H 2O(l) and the bottoms is HCl(aq) near the azeotropic concentration; feed concentrations are less than azeotrope. Thus, the degree of separation is determined to be independent of the feed concentration. The bottoms concentration varies from experiment to experiment, but does so independently of feed concentration, likely the result of corrosion impurities affecting the calculation of its concentration. It is found that HCl(aq) can be concentrated up to approximately 0.1068 mol/mol from an initial concentration of 0.0191 mol/mol. A simulation of pressure-swing distillation (PSD) is also performed, but due to safety constraints (a column operating at 10 atm must be certified to CSA B51), a single-pressure (single-column) distillation is physically performed. A single-pressure column is beneficial to the Cu–Cl cycle because it partially recycles HCl, which reduces the cost of the cycle, and still provides valuable results for analysis. The maximum HCl concentration achieved experimentally is 0.1068 mol/mol and the maximum HCl concentration determined from simulation is 0.11 mol/mol (the azeotropic concentration). The novelty of this research is that the experimental column built to study HCl partial separation is designed to be simple yet safe to integrate within the Cu–Cl cycle for hydrogen production.

Model Discrimination for Multicomponent Distillation – A Geometrical Approach for Total Reflux

AbstractWhile rate‐based models are available in commercial flowsheet simulation tools, packed distillation columns are still mostly designed based on the equilibrium stage model in combination with HETP values. In order to discriminate between both types of models in a simple way, this article proposes an algorithmic test, based on a geometric criterion for total reflux operation. Substantial differences are illustrated especially for wide‐boiling mixtures, while component‐specific mass transfer rates either increase or reduce the deviation. The derived results are validated by dedicated experiments.

Fractionation of D/H and 18O/16O Water Isotopes in a Packed Distillation Column

An experimental packed fractionating column was designed to study the process of separation of water hydrogen and oxygen isotopes, including the separation in the salt-effect distillation process. The vapor rate upon flooding and the dependence of the pressure drop in the packed column on the vapor load when operating under full reflux conditions are determined experimentally. A comparative analysis of the column operation in the processes of fractionation of pure water and fractionation of water in the presence of salt separating agents is performed. It is shown that the use of urea for separating hydrogen isotopes and MgCl2 for separating oxygen isotopes can give rise to an increase in the column performance with respect to the target component by an order of magnitude.

Minimum entropy production in a distillation column for air separation described by a continuous non-equilibrium model

In this work, we apply the rate-based model of Taylor and Krishna to describe the separation of air in a low-pressure, packed distillation column. By use of numerical optimization, we identify the temperature profile for heat exchange with the column and its surroundings that minimizes the total entropy production. Optimal operation of the column reduces the entropy production by nearly 50%, and the total heating- and cooling duties by 30% and 50%, respectively. We find that the local entropy production is more uniform for the optimal solution than in the adiabatic column, a property that may be helpful for new designs. Using the equilibrium stage model, the state of minimum entropy production has higher cooling/heating duties than in the rate-based model case. This shows that more sophisticated models can be beneficial for the development of reliable strategies to improve the energy efficiency of distillation columns.

Simulation of the process of rectification with the account of longitudinal diffusion

Based on the differential equations of material balance and mass transfer, taking into account the longitudinal mixing of the boiling solution, the formulas of the working lines of the reinforcing and exhaustive parts of the packed distillation column, the concentration profiles along the nozzle height, and the algorithm for calculating the height of the nozzle are proposed.The equations of the working lines of the reinforcing and exhaustive parts of the distillation column are analytically derived, taking into account the longitudinal diffusion, which, when the longitudinal diffusion number or the longitudinal diffusion coefficient, is transformed into the known equations of the working lines for the flow patterns of ideal displacement in both phases. Analytically, on the basis of the differential equations of material balance, the equations of working lines and mass transfer in a rectifying column for binary mixtures with allowance for longitudinal diffusion are derived. It is shown that in its absence, when the flow structure for both phases corresponds to an ideal displacement, that is, the Peclet number of longitudinal diffusion, and the longitudinal diffusion coefficient itself, the resulting equations of working lines and mass transfer become known equations for the reinforcing and exhaustive parts of the column. Comparative examples of calculations for typical structures of the ideal displacement flow and for derived equations that take into account longitudinal diffusion for the binary mixture “ethyl alcohol-ethyl alcohol” are given.

Моделирование процесса ректификации с учетом продольной диффузии на уравнения рабочих линий

Analytically, on the basis of the differential equations of material balance, the equations of working lines and mass transfer in the distillation column for binary mixtures for the strengthening and exhaustive parts of the packed distillation column with a diffusion structure of the flow in a continuous liquid phase are derived. Comparative examples of the calculation for typical structures of the flow of ideal displacement are given using the derived equations that take into account the longitudinal diffusion for a “ethanol – ethyl alcohol” binary mixture. A calculation algorithm is proposed and technological and geometrical parameters are compared with similar parameters of the same column, calculated according to a typical algorithm, when the dispersed gas phase has the structure of an ideal displacement flow, and the continuous liquid phase is an ideal mixing. The scheme of material flows and their concentrations is presented in the exhaustive part of the rectification column taking into account the longitudinal diffusion.

Simulation and optimization of reactive packed distillation column for biodiesel production using heterogeneous catalyst

It is important to simulate a process to see how its production would be in real time. However, there is a few simulation study in literature on continuous flow-biodiesel production in reactive distillation (RD) column. In this study, simulation and optimization of continuous flow-biodiesel (FAME) production in RD column packed with cheaper heterogeneous basic CaO catalyst by using Aspen HYSYS 3.2 software was investigated to contribute to the literature. In study, low-priced waste cooking oil (WCO) and methanol were fed separately into first stage at top of RD column. In the literature, transesterification was considered as pseudo-first order forward reaction, and activation energy (Ea) and frequency factor (A0) for WCO of 79 kj/mol and 2.98 x 1010 min-1 were determined, respectively. After discovering composition of WCO by GC, a set pseudo-first order forward reaction based on triglyceride (as tripalmitin, tristearin, triolein and trilinolein) were written to the simulator. The developed model using General NRTL fluid package was simulated to converge by Sparse Continuation Solver. After simulation, optimum conditions were determined by Optimizer tool and Box algorithm. In the optimization, objective function was selected so as to maximize the sum of conversion and mole fraction of m-oleate in bottom product. As a result, optimum values were determined as reflux ratio of 0.1, reboiler duty of 17.9 W, total feed flow rate of 11.2x10-4 kgmol/hour and methanol/WCO molar ratio of 6.42 for maximum conversion of 99.97% and mole fraction of 70.69%. Consequently, good results were very compatible with literature, thus showing suitability of suggested model, economically feasible biodiesel production and Aspen HYSYS 3.2 capability of handling this process successfully.

The Effects of Physical Property Estimation on Packed Column Distillation Simulation with a Rate-based Mode

The Effects of Physical Property Estimation on Packed Column Distillation Simulation with a Rate-based Mode

Optimization-Based Design of a Reactive Distillation Column for the Purification Process of Cyclohexanone Using Rigorous Simulation Model and Validated Using an Experimental Packed Column

This paper presents a model for a reactive distillation column where 2-cyclohexen-1-one is eliminated from a mixture of cyclohexanone and cyclohexanol. In the purification of cyclohexanone, 2-cyclohexen-1-one must be removed to produce high quality caprolactam. A common practice in industry is to promote the condensation of 2-cyclohexen-1-one with cyclohexanone using NaOH as catalyst. However, it promotes the cyclohexanone self-condensation with the drawback of incurring raw material losses. The column model was developed using gPROMS and validated using a packed distillation column on laboratory scale at steady state. The kinetic model of involved reactions and the binary interaction parameters of NRTL equation were implemented. The model can predict adequately the concentration profile for the involved compounds. The model was used to optimize the operating industrial conditions to minimize the unitary total cost to produce high purity cyclohexanone. gPROMS algorithms were used to solve the resulting mixed-integer nonlinear problem.

High-purity liquefied gases: Aspen Dynamics simulation of a purification process in a middle-vessel batch distillation column

The paper explores the use of Aspen Plus and Aspen Dynamics for simulating and optimising a high-precision process of liquefied gases high purification in a middle-vessel packed distillation column. The example studied involves the removal of lower- and higher- boiling impurities from the upper- and lower rectifying sections and maintaining the purified product in the middle vessel. Total reflux operation is studied in the pressure range between the normal boiling point of the purified product and the critical point. Among the considered aspects are the column design and sizing, hydrodynamics study (flooding calculations), the effect of operating pressure and temperature on the thermodynamics of vapour-liquid equilibrium and mass-transfer kinetics in a wide range of operating conditions. The paper demonstrates that the effects and dependencies observed previously through the experiments and by using user-developed custom models are adequately reproduced in the commercial simulator (Aspen Plus) by setting up a fictitious steady-state continuous process with rigorous built-in models. The optimised flowsheet can be converted to Aspen Dynamics as a pressure-driven process where a rigorous dynamic simulation gives a realistic dynamic response, and alternative control structures can be explored.