Design, Simulation, and Pilot Verification of a Coupled Azeotropic and Extractive Distillation Process for the Production of Propylene Oxide with High Purity

Abstract

To obtain the production of propylene oxide (PO) with high purity, a coupled azeotropic and extractive separation process is proposed to remove the impurities such as aldehyde, methanol, methyl formate, 2-methylpentane, water, propylene glycol (PG), and so forth from crude PO. n-Butane is screened as an azeotropic entrainer for the disposal of the light impurities, while n-octane is employed for PO purification. For the PG removal, the process of withdrawing the PG and n-octane azeotropic mixture from the side of the solvent recovery column and then separating them through liquid–liquid phase split is designed. The simulation and the total annual cost (TAC) analysis are performed by Aspen Plus with the non-random-two-liquid thermodynamic model. The key design variables such as the ratio of solvents to crude PO, the theoretical plate number, the feed stage, and so forth are analyzed and determined based on the minimal TAC method, in which the capital, energy, and raw material costs are comprehensively considered. The result that is verified through the pilot experiment shows that the purity of the PO product exceeds 99.99 % in mass and the impurity contents of formaldehyde and water are lower than the national standard value of China (GB/T 14491-2015).

Biocatalytic Furfuryl Alcohol Production with Ethanol as the Terminal Reductant Using a Single Enzyme

“…RCSTR module was chosen as the equilibrium reactor block, and the reaction equilibrium constant was estimated by Aspen Plus using the Gibbs free energy change of the reaction at 25 °C. The NRTL model was chosen for property estimation as it is an activity coefficient-based model, which accounts for the non-ideality in aqueous-organic systems and has been adopted by other researchers in non-ideal systems …”

“…The NRTL model was chosen for property estimation as it is an activity coefficient-based model, which accounts for the non-ideality in aqueous-organic systems and has been adopted by other researchers in non-ideal systems. 46 Batch Reactions. All batch reactions in this study were performed using 100 mM sodium phosphate buffer (pH 6−8) and 100 mM sodium pyrophosphate buffer (pH 8−9.5).…”

Mixed Ionic Liquids as Entrainers for Aromatic Extraction Processes: Energy, Economic, and Environmental Evaluations

“…The studied cases (1–8) are listed in Table , and they correspond to different extractants: taking pure sulfolane, pure [EMIM]­[NTf 2 ] and [EMIM]­[EtSO 4 ], and the binary IL mixtures ([EMIM]­[NTf 2 ]/[EMIM]­[EtSO 4 ] = 9:1, 7:5, 5:5, 3:7, and 1:9). The separation processes with sulfolane and IL extractants were optimized respectively according to the separation sequences and the importance of the operation parameters. − In the industrial sulfolane process, sulfolane in the raffinate phase needs to be removed by a water scrubber. Although this method can reduce the temperature of the bottom of the column and prevent sulfolane from thermal decomposition, it will inevitably bring water as an impurity.…”

Conceptual Process Design and Techno-Economic Analysis of Biocatalytic Furfural Hydrogenation Using Ethanol as the Terminal Reductant

“…However, the annual operating cost would be crucial for economic feasibility. The annual operating cost includes raw material cost, utility cost used in various process units, labor cost, maintenance cost, overhead cost, supervision and management cost, and solvent cost, and can be calculated using the following equation: , normalAnnual 0.25em normaloperating 0.25em cos normalt = C RM + C utilities + C OL + C normalS normal& normalM + C overhead + C maintenance + C solvent where, C RM = raw materials (EtOH and FAL) cost, C utilities = total utilities cost, C S&M = supervision and management cost, C overhead = overhead cost, C maintenance = maintenance cost, and C solvent = makeup solvent cost, all calculated in $/y.…”

“…Comprehensive process simulations were carried out using Aspen Plus v12.1 and its built-in unit operation modules. The nonrandom two-liquid (NRTL) model was used for property estimation to account for the nonideality of the vapor–liquid and vapor–liquid–liquid equilibria for the investigated system. , Binary component pairs for which experimental data were not available, the Aspen Plus built-in group contribution method R-PCES (Property Constant Estimation System) was used for NRTL binary interaction parameter estimation. The accuracy of the NRTL-predicted binary VLE was compared with available experimental data for FAL-FOL, , AcH-H 2 O, , and AcH-EtOH, as representational binary systems and reasonable accuracy was observed in each case (Figures s1–s3).…”