technische universität dortmund Fakultät Bio- und Chemieingenieurwesen Lehrstuhl für Fluidverfahrenstechnik

Transkript

1 Experimental Investigation, Analysis and Optimisation of Hybrid Separation Processes EFCE Excellence Award in Process Intensification 2009 Carsten Buchaly

2 Motivation Reduced volumina of apparatuses / capital costs Less energy consumption No auxiliary components required Strong interactions of both unit operations Detailed process know-how necessary Less experience existent 1

3 Process Description 1-Propanol (H 2 O) Propionic acid 1-Propanol H 2 O Membrane separation 1-Propanol H 2 O H 2 O Reactive Distillation Integration of reaction and separation Increased conversion and selectivity 1-Propanol Membrane Separation 1-Propanol recovery High selectivity Independent on VLE n-propyl propionate 2

4 Methodology Hybrid Process (reactive distillation + membrane separation) Theory Experimental design Process analysis Optimisation Experiments Pilot-scale Start-up behaviour Process know-how 3

5 Methodology Hybrid Process (reactive distillation + membrane separation) Theory Experimental design Process analysis Optimisation Experiments Pilot-scale Start-up behaviour Process know-how Unit operation 1 (reactive distillation) Unit operation 2 (membrane separation) Experiments Pilot-scale Operational parameters Model validation Theory Experimental design Process analysis Experiments Lab-scale Separation characteristics Model parameters Experiments Pilot-scale Scale-up 6

6 Reactive distillation: pilot-scale plant Distillation column (DN 50) ProAc POH Sulzer BX 2.3 m 5.5 m packing height Investigated operating parameters Pressure: atmospheric Katapak SP 11 Distillate-to-feed ratio: m Molar feed ratio (c): Sulzer BX 0.5 m Reflux ratio (RR): m Katapak SP 11 Database of 15 successfully realised RD experiments 4

7 Vapour permeation: pilot-scale plant Steady state vapour permeation experiments Connection with reactive distillation column to hybrid separation process Sulzer Pervap 2201(D) with A Memb = 0.5m² Plate-and-frame module Buchaly, Kreis, Górak (2007): Chemical Engineering and Processing, 46, p

8 Vapour permeation: pilot-scale plant Steady state vapour permeation experiments Connection with reactive distillation column to hybrid separation process Sulzer Pervap 2201(D) with A Memb = 0.5m² Binary system 1-propanol/water Water concentration: wt.-% Feed temperature: C Feed pressure: atmospheric Permeate pressure: mbar Buchaly, Kreis, Górak (2007): Chemical Engineering and Processing, 46, p

9 Hybrid separation process: experimental investigations RD and VP separated sequential configuration closed recycle 6

10 packing height / m technische Hybrid separation process: experimental investigations 5,5 5,0 4,5 4,0 3,5 POH (exp) ProPro (exp) ProAc (exp) Water (exp) ProAc-Feed Successful experiment with coupled unit operations 3,0 2,5 2,0 POH-Feed 1,5 1,0 0,5 reactive section 0,0 evaporator -0,5 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 molar fraction in the liquid phase / mol mol -1 Buchaly, Kreis, Górak (2008): Chemie Ingenieur Technik, 80, p

11 packing height / m technische Hybrid separation process: experimental investigations 5,5 5,0 4,5 4,0 3,5 3,0 2,5 POH (exp) ProPro (exp) ProAc (exp) Water (exp) ProAc-Feed POH-Feed Successful experiment with coupled unit operations Increase of ester concentration in the bottom Change of distillate composition 2,0 1,5 1,0 0,5 reactive section 0,0 evaporator -0,5 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 molar fraction in the liquid phase / mol mol -1 Buchaly, Kreis, Górak (2008): Chemie Ingenieur Technik, 80, p

12 packing height / m technische Hybrid separation process: experimental investigations 5,5 5,0 4,5 4,0 3,5 3,0 2,5 POH (exp) ProPro (exp) ProAc (exp) Water (exp) ProAc-Feed POH-Feed Successful experiment with coupled unit operations Increase of ester concentration in the bottom Change of distillate composition 2,0 1,5 1,0 reactive section 0,5 0,0 evaporator -0,5 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 molar fraction in the liquid phase / mol mol -1 Strong interactions between both unit operations Buchaly, Kreis, Górak (2008): Chemie Ingenieur Technik, 80, p

13 Modelling approach Reactive distillation based on non-equilibrium stage model Multicomponent mass transfer (Stefan-Maxwell) Packing specific correlations (hold-up, Dp, k g a, k l a) Steady state and dynamic process simulation Vapour permeation based on Solution-Diffusion-Model Polarisation effects (c,t) Hydrodynamics (co-, counter-current; Dp) Membrane materials Implemented in Aspen Custom Modeler TM (ACM) 8

14 Modelling approach Reactive distillation based on non-equilibrium stage model Multicomponent mass transfer (Stefan-Maxwell) Packing specific correlations (hold-up, Dp, k g a, k l a) Steady state and dynamic process simulation Vapour permeation based on Solution-Diffusion-Model Polarisation effects (c,t) Hydrodynamics (co-, counter-current; Dp) Membrane materials Implemented in Aspen Custom Modeler TM (ACM) 8

15 packing height [m] technische Reactive distillation: model validation 5,5 5,0 4,5 4,0 3,5 3,0 POH (exp) ProPro (exp) ProAc (exp) Water (exp) ProAc-Feed Feed: c POH/ProAc : RR: Distillate: 2,0 kg/h 2,068 2,49 0,85 kg/h 2,5 2,0 POH-Feed 1,5 1,0 reactive section 0,5 0,0 reboiler -0,5 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 molar fraction in liquid phase [mol/mol] 9

16 packing height [m] technische Reactive distillation: model validation 5,5 5,0 4,5 4,0 3,5 3,0 POH (exp) ProPro (exp) ProAc (exp) Water (exp) ProAc-Feed Feed: c POH/ProAc : RR: Distillate: 2,0 kg/h 2,068 2,49 0,85 kg/h 2,5 2,0 POH-Feed Correlations for column internals: 1,5 1,0 0,5 reactive section Sulzer BX -> Bravo et al.*; Rocha et al.** Katapak-SP -> Brunazzi*** 0,0 reboiler -0,5 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 molar fraction in liquid phase [mol/mol] * Bravo et al. (1985): Hydr. Proc., 1, p ** Rocha et al. (1993): Ind. Eng. Chem. Res., 32, p *** Brunazzi, Viva (2006): Distillation & Absorption

17 liquid phase mole fraction experimental / mol mol -1 technische Reactive distillation: model validation 1,0 0,8 +15% +5% -15% -5% Feed: c POH/ProAc : RR: Distillate: 2,0 kg/h 2,068 2,49 0,85 kg/h 0,6 0,4 0,2 Correlations for column internals: Sulzer BX -> Bravo et al.*; Rocha et al.** Katapak-SP -> Brunazzi*** 0,0 0,0 0,2 0,4 0,6 0,8 1,0 liquid phase mole fraction simulated / mol mol -1 * Bravo et al. (1985): Hydr. Proc., 1, p ** Rocha et al. (1993): Ind. Eng. Chem. Res., 32, p *** Brunazzi, Viva (2006): Distillation & Absorption

18 liquid phase mole fraction experimental / mol mol -1 technische Reactive distillation: model validation 1,0 0,8 +15% +5% -15% -5% Feed: c POH/ProAc : RR: Distillate: 2,0 kg/h 2,068 2,49 0,85 kg/h 0,6 0,4 0,2 Correlations for column internals: Sulzer BX -> Bravo et al.*; Rocha et al.** Katapak-SP -> Brunazzi*** 0,0 0,0 0,2 0,4 0,6 0,8 1,0 liquid phase mole fraction simulated / mol mol -1 Model validation successful * Bravo et al. (1985): Hydr. Proc., 1, p ** Rocha et al. (1993): Ind. Eng. Chem. Res., 32, p *** Brunazzi, Viva (2006): Distillation & Absorption

19 packing height / m technische Hybrid separation process: model validation 5,5 5,0 4,5 4,0 3,5 POH (exp) ProPro (exp) ProAc (exp) Water (exp) ProAc-Feed 3,0 2,5 2,0 POH-Feed 1,5 1,0 0,5 reactive section 0,0 evaporator -0,5 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 molar fraction in the liquid phase / mol mol -1 10

20 packing height / m technische Hybrid separation process: model validation 5,5 5,0 4,5 4,0 3,5 POH (exp) ProPro (exp) ProAc (exp) Water (exp) ProAc-Feed Sequential configuration X ProAc,exp = 74.4 % X POH,exp = 29.8 % X ProAc,sim = 73.6 % X POH,sim = 29.4 % 3,0 2,5 2,0 POH-Feed 1,5 1,0 0,5 reactive section 0,0 evaporator -0,5 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 molar fraction in the liquid phase / mol mol -1 10

21 packing height / m technische Hybrid separation process: model validation 5,5 5,0 4,5 4,0 3,5 POH (exp) ProPro (exp) ProAc (exp) Water (exp) ProAc-Feed Sequential configuration X ProAc,exp = 74.4 % X POH,exp = 29.8 % X ProAc,sim = 73.6 % X POH,sim = 29.4 % 3,0 2,5 2,0 1,5 1,0 0,5 reactive section POH-Feed X ProAc,exp X POH,exp Closed recycle = 73.3 % = 31.8 % X ProAc,sim = 73.5 % X POH,sim = 31.9 % 0,0 evaporator -0,5 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 molar fraction in the liquid phase / mol mol -1 Excellent agreement between experiments and simulation 10

22 Hybrid separation process: process analysis D = const RR = const A memb =? w H2O varied Q H =? 11

23 Hybrid separation process: process analysis 1,0 Strong influence on A Memb High degree of dewatering requires large A Memb 0,8 membrane area / m 2 reboiler heat duty / W A memb , , ,2 0,0 Column: D/F mass = RR = 2.5 Membrane: p perm = 30 mbar D T sup = 4.0 C 0,0% 2,5% 5,0% 7,5% 10,0% 12,5% 15,0% mass fraction of water in the recycle

24 Hybrid separation process: process analysis 1,0 0,8 Strong influence on A Memb High degree of dewatering requires large A Memb Lower Q H at higher recycle purity required 0,6 membrane area / m 2 reboiler heat duty / W A memb Q H , ,2 0,0 Column: D/F mass = RR = 2.5 Membrane: p perm = 30 mbar D T sup = 4.0 C 0,0% 2,5% 5,0% 7,5% 10,0% 12,5% 15,0% mass fraction of water in the recycle

25 Hybrid separation process: process analysis 1,0 0,8 Strong influence on A Memb High degree of dewatering requires large A Memb Lower Q H at higher recycle purity required 0,6 membrane area / m 2 reboiler heat duty / W A memb Q H , Optimisation problem 0,2 0,0 Column: D/F mass = RR = 2.5 Membrane: p perm = 30 mbar D T sup = 4.0 C 0,0% 2,5% 5,0% 7,5% 10,0% 12,5% 15,0% mass fraction of water in the recycle

26 Hybrid separation process: process analysis 1,0 0,8 Strong influence on A Memb High degree of dewatering requires large A Memb Lower Q H at higher recycle purity required 0,6 membrane area / m 2 reboiler heat duty / W A memb Q H , Optimisation problem 0,2 0,0 Column: D/F mass = RR = 2.5 Membrane: p perm = 30 mbar D T sup = 4.0 C 0,0% 2,5% 5,0% 7,5% 10,0% 12,5% 15,0% mass fraction of water in the recycle Modified stochastic optimisation algorithm based on a differential evolution* approach (MDE) * Babu, Angira (2008): Comp. Chem. Eng., 30, p Angira, Babu (2006): Chem. Eng. Sci., 61, p Frerick et al. (2008): Chem. Ing. Tech., 1, p

27 Hybrid separation process: rigorous optimisation using MDE Objective function: min Q H = f (RR, D/F mass, w H2O,recycle ) Solution space: 0.8 RR 5; 0.25 D/F mass 0.55; 0.02 w H2O,recycle 0.15 Constrain: x ProPro,bottom 0.75 w H2O =? no LLE A memb =? RR =? D/F mass =? Q H =? 13

28 objective function technische Hybrid separation process: rigorous optimisation using MDE Objective function: min Q H = f (RR, D/F mass, w H2O,recycle ) Solution space: 0.8 RR 5; 0.25 D/F mass 0.55; 0.02 w H2O,recycle 0.15 Constrain: x ProPro,bottom w H2O =? RR =? D/F mass =? no LLE parent 1 parent 2 parent 3 parent 4 A memb =? parent 5 parent 6 parent 7 parent 8 65 generations with 8 parents CPU-time ~ 12 hour Q H,min = 1059 W RR = 0.98; D/F mass = 0.52; w H2O,recycle = A memb = 3.04 m² Q H =? generation feasibility and applicability of the now available optimisation package based on detailed process models proven 13

29 Conclusion Choice of an appropriate chemical test system Development of a generic methodology for the design of hybrid processes Provision of reliable experimental data for the applied unit operations Experimental investigation of the fully coupled unit operations Reliable experimental database for both unit operations and hybrid process Model validation for the stand-alone unit operations and the hybrid process Analysis of the influence of decisive operational parameters on the performance of the hybrid separation process Optimisation of the hybrid process using an evolutionary algorithm 14

30 Special thanks to: Stefan Kluckhenn; Ildefonso Campos-Velarde; Christian Schroeder; Marcel Kotora; Martin Cruz-Dias; Carlos Duarte; Katrin Kissing; Siggi Weiss; Piotr Mitkowski; Paulo Perez; Gencay Cengiz; Rocio Castillo-Tornay; Dominik Plaßmann; Markus Hengsbach; Bharat Karanth; Mayur Dalwani; Mayuratheepan Puthirasigamany & the Laboratory of Fluid Separations EU-Project INSERT Integrating Separation and Reaction Technologies (Contract No. NMP2-CT ) Marie Curie Training Site Reactive Separations (Contract Nr. HPMT-CT ) Fonds der Chemischen Industrie

31 Experimental Investigation, Analysis and Optimisation of Hybrid Separation Processes EFCE Excellence Award in Process Intensification 2009 Carsten Buchaly