Dividing wall columns are state of the art distillation arrangements performing three separation tasks within one unit. Compared to using two conventional columns in series this saves both capital costs and energy but on the other hand it brings a higher risk of malfunction. This simulation study analyses what can go wrong during the operation of dividing wall columns. The emphasis is on the operation of the prefractionator section, that is, on the choice on the liquid and vapor splits, which is crucial for the overall performance. The resulting two-way flows between the prefractionator and main column gives a broader feasible operating range than in a conventional column arrangement. This can lead to peculiar behavior, including circulation of components around the dividing wall. This paper identifies 15 non-optimal operating regions for the prefractionator with specific internal flow patterns, characteristic temperature and composition profiles for the separation of a fairly ideal mixture of benzene, toluene and p-xylene. From these results it is possible to identify and hopefully rectify wrong choices for the liquid and vapor splits in a dividing wall column.

This article examines the effect of individual and combined uncertainties in thermodynamic models on the performance of simulated, steady-state Pareto-optimized Dividing Wall Columns. It is a follow-up of the previous work analogously treating deviations in process variables. Such deviations and uncertainties that may even be unknown during the design process can significantly influence the separation result. However, other than process variables, uncertainties in thermodynamics are usually not systematically considered during design. For the first time, the effects of uncertain thermodynamic properties on Pareto-optimized DWCs with different numbers of stages and for different mixtures are presented and compared qualitatively and quantitatively. Depending on the number of stages and mixture characteristics, particularly critical properties are identified. On the one hand, this provides information on aspects requiring special attention prior to design, and on the other hand, it also indicates in which section of the DWC a stage supplement might be most beneficial.

Introductory studies on the relationship between design and robustness of optimized Dividing Wall Columns are presented. These columns are optimized multicriterially for different mixtures and numbers of stages. The internal distribution of stages depending on the total number is analyzed. Supporting Material analyzes the stagedependent vapor demand of the binary sub-systems and shows that the relationships can change fundamentally, especially for low stages. The deviations in process variables considered are defined and screening results are presented. Supporting Material provides an analysis of the nominal and disturbed concentration profiles for deviating internal splits, and shows that at high numbers of stages, the formation of pinch zones leads to significantly higher purity losses. Finally, results for combined deviations are presented. For all result parts, patterns are worked out that depend only on the number of stages, specific mixture properties like characteristics of the Vmin diagram or the individual product.

Multiple dividing wall columns are a promising intensified process to reduce energy consumption of conventional distillation sequences by up to 55 %. Their operation is of course more complex, in particular with regard to the internal split ratios of the liquid and vapor flows at the dividing walls. Theoretical studies have shown that the split ratios may not need to be at a specific value but in a certain range, which would simplify operation. However, until recently no experimental setup existed to prove this observation in reality. This work presents the first comprehensive experimental study in this regard obtained from an approximately 10 m high multiple dividing wall column pilot plant. The plant has two dividing walls, so the column consists of a prefractionator, a middle and a main column. The results validate the theoretical observations. With the chosen mixture in particular the middle column is found to be operable in a certain range. Temperature profiles prove that internally different separations are performed, while the resulting product compositions are unaffected.

Free-form shape optimization techniques are investigated to improve the separation efficiency of structured packings in laboratory-scale distillation columns. A simplified simulation model based on computational fluid dynamics (CFD) for the mass transfer in the distillation column is used and a corresponding shape optimization problem is formulated. The goal of the optimization is to increase the mass transfer in the column by changing the packing's shape, which has been previously used as criterion for increasing the separation efficiency of the column. The computational shape optimization yields promising results, with an increased mass transfer of nearly 20%. For validation, the resulting optimized shape is additively manufactured using 3D-printing and investigated experimentally. The experimental results are in good agreement with the performance improvement predicted by the computational model, yielding an increase in separation efficiency of around 20%.

Although annular centrifugal contactors (ACCs) have been widely applied, their complex operation and its dependence on operational variables are still not fully quantified. That hinders their acceptance and precludes revealing their potential. Intuitive techniques to informedly operate ACCs while maximizing their gain are thus demanded. This work introduces a procedure to define the operation window of an extraction task on ACCs, using CINC V 02 variant, that demonstrates the feasible ranges of the operational variables and the maximum possible throughput. The overlapping impacts of the variables are also visualized, thus understood and, to enhance practicality, the procedure is converted into user-interface software. The concept is intended to adapt to any task on ACCs and it is based on merging a hydraulic part focusing on the rotor and a statistical one modeling the annulus mass transfer. Additional insights on ACC's hydraulics and mass transfer dynamics are presented, including the time needed to establish the steady state.

Although the system water + n-propanol + toluene is important for research and industry, its liquid–liquid equilibria LLE are scarcely covered in literature, with no thermodynamic modeling reported using typical equations of state or activity coefficient models. The current work addresses this gap by presenting the equilibria of the system at 20°C and atmospheric pressure. That includes the binodal curve and twelve tie lines covering the entire miscibility gap, whereby the compositions of the conjugate phases are determined using high-precision refractive index measurements. The quality of the data is verified by applying the techniques of conjugate lines and mass balances, whereas it is shown that empirical correlations cannot be used for that purpose. Three different methods to extrapolate the critical point are then employed and, using the more reliable one, that point is located and added to the dataset. Further thermodynamic correlation of the data is conducted using the activity coefficient models, UNIQUAC and NRTL. While the former did not perform well, the latter was able to model the LLE and adapt to temperature changes.

In early project stages often no simulation results are available for dividing wall columns. Hence, shortcut methods are helpful providing estimates for vapor and liquid splits which can be used for initializing rigorous process models. In a previous paper it was shown that diagrams are a suited tool to satisfy this need. However, it has turned out that the approach shows weaknesses for columns with finite and non-optimally allocated stage numbers. This contribution closes this gap and presents a new approach to derive suited initial guesses. For this purpose, the original diagram is combined with a heuristic approach to calculate Pareto-optimal column designs. The resulting stage-adapted diagram can then be used to graphically determine suited ranges of the vapor and liquid splits. A comparative study shows that the new approach is a powerful tool to enable reliable guesses for multiple dividing wall column simulations with finite stage numbers.

[This corrects the article DOI: 10.1021/acs.iecr.3c04102.].

Multiple dividing wall columns are a promising extension of the concept of dividing wall columns offering energy savings of up to 55 % compared to conventional distillation sequences for four products. This paper presents the first systematic study on the start-up of a multiple dividing wall column. The study aims to develop a start-up strategy for the first pilot multiple dividing wall column worldwide commissioned at Ulm University. The start-up process is investigated using a dynamic simulation model that facilitates the investigation of different start-up strategies and the influence of different factors on the start-up process. Based on these dynamic simulations a viable start-up procedure for the column is found. Through improvements to the procedure a significantly shortened start-up process is achieved. The dynamic response of the column shows that the cold and dry plant is able to reach the desired operating point within few hours. Simulation results are confirmed with experimental data from the pilot column.

The minimum (S/F)min and maximum (S/F)max solvent-to-feed ratios with their respective number of stages are essential in designing countercurrent solvent extraction processes, as they define the operating limits of the separation task. The (S/F)min is typically associated with an infinite number of stages, and the common state-of-the-art method to graphically determine it for ternary systems depends on finding the limiting tie line that passes through a known feed composition. However, the method applicability ceases when solute concentration in the feed is high enough so that no limiting tie lines exist anymore. This work comes to extend for the first time the said method beyond its validity region to cover feed compositions for which it fails. To support the presented theory, rigorous simulations are performed that clearly show the typical asymptotic behavior at infinite stages when the graphical (S/F)min is approached. For (S/F)max, although it is associated in literature with one required stage, it is not clear how this applies and whether the term “stage” is even applicable. By discussing it from different perspectives, this work shows how the corresponding minimum stages can be obtained and that they can be different from one. The latter is also supported by simulations.

Large-eddy simulations of mircrostructured pipes are carried out in this study. The pipes are helically ribbed with several starts from which shavings are cut out along the rib. A Reynolds number of 16,000 and a Prandtl number of 9 are applied. Cyclic boundary conditions are used to simulate an infinitely long pipe and a LES turbulence model is applied. The results show that both the number of starts and the distance of the shavings along the rib, have an impact on turbulence and heat transfer. An optimum was found for the number of starts regarding heat transfer. The heat transfer area increases for a higher number of starts, but the flow-related heat transfer decreases. Thus, an optimal arrangement of starts was identified for this pipe configuration. The variation of the arrangement showed that the pressure loss can be reduced by an inline configuration, whereby the heat transfer is only slightly decreased. It can be shown that the simulations can contribute to optimising complex structures in the future and thus accelerate the development process of new geometries. It can be shown that the development of new heat exchanger pipes is also possible with complex structures, this can be demonstrated here for the first time for structures that can actually be fabricated.

This publication presents a general approach for the enhancement of packings using 3D printing. Within a joint research project of the Ulm University, the Technical University of Munich and BASF SE, the presented methodology is used to develop miniaturized, scalable distillation columns for process development and scale-up applications. Therefore, a combination of design, computational fluid dynamics, 3D printing and experiment is used to overcome current limitations in the design of structured packings. The packing to be developed should have a high, constant separation efficiency independent of the F-factor at the target diameter of 20 mm. Based on a 3D printable version of the Rombopak 9M, an improved structure is introduced using the proposed methodology. This packing is an intermediate step, but already exhibits a higher, more constant separation efficiency and an improved reproducibility. This publication acts as proof of concept for this methodology.

The now industrially established concept of a dividing wall column can be evolved into that of a multiple dividing wall column, and thus theoretically achieve any number of pure fractions with a single condenser and reboiler in one column shell. The possible investment and energy savings compared to conventional sequences are up to 55 % for four products. These benefits, however, are associated with increased complexity regarding the plant design and operation, which has so far prevented its realization. At Ulm University, the first pilot-scale multiple dividing wall column was designed, built, and commissioned. This paper provides a summary of the theoretical background and previous work, which served as the foundation for the design of the column. introduces the pilot plant, and presents the first operating data of its kind. In the course of this work, a start-up procedure for the pilot column was developed, that allows for a reliable start-up process in a time frame of only few hours. The experimental data obtained during this research serve as proof that the concept of a multiple dividing wall column can be practically implemented.

This article introduces a technique to simulate the liquid flow inside different additively manufactured structured packings with enclosing column wall on laboratory scale using the software OpenFOAM. The physical problem is handled by means of a periodic setup and makes use of common numerical discretization techniques of the software. The periodic setup reduces the required computational resources while permitting simulations of the liquid distribution in a large part of the column even for complex packing structures. A novel post-processing tool is used for a systematic evaluation of the liquid distribution. When optimizing packing structures, these computational fluid dynamics simulations eliminate the necessity to manufacture every developed geometry. As a proof of concept, this work simulates the liquid flow inside two additively manufactured packing structures. The results are validated with experimental data and show, that an improved liquid distribution can be obtained inside laboratory-scale packing structures.

Additive manufacturing (3D printing) is a promising approach to creating packings for laboratory-scale distillation columns. Current research focuses on experiments and simulations to tailor packing geometry regarding performance (e.g., pressure drop, fluid distribution). These performance benchmarks are, in large part, dependent on the wettability of the manufactured surface. Research shows that the 3D-printing process settings affect wetting significantly. This effect must be quantified to accurately assess the effectiveness of printed packing geometry. Due to the interdependence of wetting, surface roughness, and involved substances, the required experimental effort is not feasible. Sessile drop experiments show that analytical models underpredict the resulting wettability. In this study, a novel method to address this issue is introduced. The rough surface of a printed sample is reverse-engineered, and CFD simulations are performed to predict the static contact angle. The results show agreement between the computational model and experimental investigations.

Although annular centrifugal extractors (ACCs) have been increasingly applied in the process industry recently, their complex fluid dynamics is still not fully explained in the literature. This could be one of the major reasons why their current industrial implementation is not up to their potential. This work aims to enhance the understanding of ACC hydrodynamics by focusing on liquid holdup and distribution in the variant CINC V 02. First, several sources causing holdup variation are investigated with the help of a model simulating the liquid surface inside the device. Then, the factors determining liquid content in both chambers are presented. The standard shut-down procedure to measure holdup is eventually elaborated on, and an improved procedure is proposed.

For an in-depth investigation of the separation process in small-scale distillation columns, knowledge about the exact vapor load inside the column is highly important. However, since columns with small diameters have a comparatively high surface-to-volume ratio, heat losses have a significant impact on fluid dynamics, as they lead to unwanted condensation, and thus, to changes in the internal flows. This work presents a procedure used to measure heat losses in a 9.6 m high distillation column with three partially parallel segments (multiple dividing wall column). The evaporator is made of stainless steel, and the column walls are made of double-walled, evacuated, mirrored glass, and additionally, these can be heated. It is found that significant amounts of heat are lost in the evaporator. Throughout the column height, around 0.8 kW are additionally lost, even with external wall heating. To determine the main reason for this significant loss, thermal images are taken, indicating that the problem mainly arises because of the flanges. Based on this, it can be concluded that proper insulation and additional heating jackets for the column walls are highly recommended for small-scale distillation columns in order to increase their thermal efficiency.

Lab-scale distillation columns with a diameter of 50 mm (DN50) are widely established for experimental investigations during process development and scale-up. In recent years, the diameter was successfully reduced to 20 mm (DN20) by utilizing additive manufacturing to attain new miniaturized packings that attend to the specific miniaturization obstacles. However, since heat losses have a higher influence in a DN20 column, this publication investigates their impact on the operation and data obtained from an additively-manufactured laboratory distillation column. A comparison is provided between the state-of-the-art DN50 size and a miniaturized DN20 size. The heat loss rate is estimated by simulations and experimentally measured for both column diameters with and without thermal insulation. The impact on the internal flow variation and the gas load (F-factor) uncertainty is presented. The findings provided motives for further improvement to minimize the heat loss rate in the DN20 column. This led to the development of a novel active insulation system where custom-dimensioned flexible silicon heaters are utilized for external heat compensation. The system reduces the heat losses significantly and led to a decrease in the F-factor uncertainty from ± 0.09 Pa0.5 to 0.04 Pa0.5 compared to the passively insulated case. Finally, the effect of the heat loss reduction on the mass transfer efficiency is discussed using two additively-manufactured packings, a 3D-printable version of the Rombopak 9M structure (RP9M-3D) and a new miniaturized packing (XW-Pak). The RP9M-3D showed a decrease in separation performance with heat loss reduction while the performance of the XW-Pak remained unaffected.

In this work we present a new approach that we use to simulate and optimize multiple dividing wall columns at the same time. Instead of considering all model equations as constraints and all process variables as optimization variables in a large and highly nonlinear optimization problem we only incorporate a subset of the model equations as constraints and a subset of the process variables as optimization variables. The remaining process variables are calculated from this subset by a robust and fast calculation procedure. This calculation procedure also ensures that the remaining model equations are satisfied. A comparison with the commercial process simulator Aspen Plus shows that with the new approach multiple dividing wall columns can be optimized more stable and better solutions are found. Moreover the time needed to find an optimal design decreases significantly.