Docstoc

Hybrid Distillation

Document Sample
Hybrid Distillation Powered By Docstoc
					Optimal synthesis of batch
separation processes

Taj Barakat and Eva Sørensen
University College London


iCPSE Consortium Meeting, Atlanta, 30-31 March 2006
Motivations

  Many valuable mixtures are difficult to
   separate
  Need to optimise efficiency of current
   processes
  Select most economical separation process

  Explore novel techniques and alternatives




                      2
Objectives

    Development of models/superstructure to
     determine the best design configuration,
     operating policy and control strategy for
     hybrid separation (distillation/membrane)
     processes.

    Develop general guidelines for design,
     operation and control of such processes

                         3
Project Features

  Economics objective function
  Rigorous dynamic models
  Encompassing (most of) the available
   decision variables
  Considering novel configurations




                      4
Outline

 1.   Optimal synthesis of batch separation
      processes
 2.   Multi-objective optimisation of batch
      distillation processes
 3.   Concluding remarks




                        5
   Optimal synthesis of
batch separation processes




             6
Configuration Decisions
                       Separation problem

                                ?
                           Process
                        Superstructure

  Batch Distillation    Batch Pervaporation   Batch Hybrid




                                    7
Design and Operation Decisions
   Design Alternatives             Operational Alternatives




                          Min
                         capital
                          cost
                                    • Vapour loading rate
  • Trays
                                    • Reflux/reboil ratios
  • Membrane stages
                                    • Recovery/No. batches
  • Membrane modules       Min
                         running    • Withdrawal rate
                           cost     • Task durations




                             8
Process Superstructure


                Rc                   Ns , Nm,s
           Nt                                         Rr
                                 Qs
      Lr
                       Fs            P

                                                 Retentate


                                     Rp          Permeate


                Feed            Qr               Offcut


                            9
Batch Distillation


              Rc             Product 1


                             Offcut
      Nt      Rp
                             Product 2




                        Qr
             Reboiler

                             10
Batch Pervaporation


            Ns   Separation Stage


                                    Nm,s
                                            Rr


                                           Retentate


     Feed
                          P
                                    Rp     Permeate
            Qf
                                           Offcut



                                     11
Hybrid Distillation I


          Rc
                          Product

                                   Feed
    Nt
                     Ns Nm,s

                     P             Rp

                     Qr   Offcut      Permeate
          Reboiler



                                     12
Hybrid Distillation II



                     Ns Nm,s
         Rc


        Nt       P

                               Retentate
                 Rp
                               Offcut

         Feed   Qr             Permeate


                                  13
Hybrid Distillation III



                Rc            Ns , Nm,s
           Nt                                      Rr


                        Fs     P
      Lr
                                          Retentate
                               Rp         Permeate
                 Feed
                             Qr Rpr       Offcut


                                    14
Problem Formulation – Objective Function
  Maximise
  Annual Profit =       Revenues – Operating Costs Av. Time – Capital Costs
                         Batch Processing Time

                    Nonlinear, (OC/CC, Guthrie’s correlations)
Subject to :
        Model equations                                DAE/PDAE, nonlinear
        Design variable bounds                         discrete and continuous
        Operational variable bounds                    continuous

To determine :
        Design variables
        Operation variables (time dependent)




                                         15
Problem Formulation - Solution
• Mixed integer dynamic optimisation (MIDO) problem
• Complex search space topography (local optima, nonconvex)
• Need robust, stable and global solution method


       DAE
           gPROMS (Process Systems Enterprise Ltd., 2005)


       MIDO
           Genetic Algorithm (GA)



                                   16
Optimisation Implementation
                        Genetic
                       Algorithm    GAlib
                        Module
      Genome Set                    Simulation Output

                           Batch
                    Distillation/Pervap gPROMS
                           Model

      Model State                   Physical Properties


                    Thermodynamics
                                   Multiflash
                        Model

                             17
Case Study




    18
Case Study ( Acetone – Water )
    Separation of a binary tangent-pinch mixture
    Acetone dehydration system ( 70 mol % acetone feed )
    20,000 mole feed

    Subject to:
        Purity            ≥ 97%
        Recovery          ≥ 70%

    Maximise:
        Annual profit

    Assuming:
        Single membrane stage
        Single retentate recycle location
                                      19
Case Study Superstructure

               Rc               NsNm,s
          Nt                                      Rr



     Lr               Fs        P


                                         Retentate
                                Rp
                                         Permeate
               Feed        Qr
                                         Offcut

                            20
Optimal Process - Hybrid
                                                                         Rr
                                                                         1.00 – 1.8%
                                                    Nm = 2
                                                                         0.83 – 96.3%
                                 To = 330 K
                                                                         0.24 – 1.9%
               Lr =3
                                  Fside = 2.5 mole/s
                       Nt = 30
 tf = 5119 s                                    P = 300 Pa               Rp
                           Fs = 9                                        0.79 – 1.8%
Profit 18.07 M£/yr                                           Retentate
                                  VReb = 5 mole/s                        1.00 – 96.3%
                                                             Permeate
                                                                         0.88 – 1.9%

                                  Feed                       Offcut



                                               21
Fixed Configuration – Distillation only
                                                         Rr
                                                         1.00 – 0.10%
                                             Product 1
                                                         0.68 – 99.7%

                                             Offcut      0.70 – 0.20%

                     Nt = 30                 Product 2

    tf = 8964 s                                          Rp
                           VReb = 5 mole/s               1.00 – 0.10%
Profit 14.30 M£/yr
                                                         1.00 – 99.7%
        -26%
                                                         0.00 – 0.20%

                               Reboiler




                                    22
Case Study Summary
    Approach for process selection based on overall
     economics
    Allows determination of best process alternative
     for maximum overall profitability
    Company specific costing can easily be included




                           23
Multi-objective optimisation of
 batch distillation processes




               24
Batch Distillation



                Rc             Product 1


                               Offcut
       Nt       Rp
                               Product 2




                          Qr
               Reboiler

                               25
Problem Formulation – Objective Function

 Minimise                        Minimise
 Investment Costs         &      Operating Costs


Subject to :
       Model equations                        DAE/PDAE, nonlinear
       Design variable bounds                 discrete and continuous
       Operational variable bounds            continuous
To determine :
       Design variables
       Operation variables (time dependent)



                                  26
                                       f(x)
Optimisation
Single-objective optimisation:
To find a single optimal solution x*
of a single objective function f(x)


                                       0                x
                                                 x*
Multi-objective optimisation:
To find array of “Pareto optimal” solutions with respect to
multiple objective functions



                               27
Multiobjective Optimization Problem
    Maximize f (x)  ( f1(x), f 2 (x), ..., f k (x))
    subject to x  X
                                      f 2 ( x)
                                                  Pareto Optimal
Several Pareto-optimal sets                             Solutions




                                   Minimise
                                                             f1 (x)
                                                 Minimise


                              28
Ranking
           k
          nc
          i 1 i   if solution is infeasible
          
f ( g )  2       if solution is feasible but dominated

          3
          
                   if solution is feasible and non-
                   dominated




                        29
Ranking

                                      Max = 1
   F2
                    2
        3

        3                2
                                 2
            3                          2

                3   3        3
                        F1                      better

                                 30
Problem Formulation - Solution
 • Multi-objective Mixed integer dynamic optimisation
   (MO-MIDO) problem
 • Need robust, stable and global solution method


     DAE
         gPROMS (Process Systems Enterprise Ltd., 2005)


     MO-MIDO
         Multi-Criteria Genetic Algorithm (MOGA)



                                 31
Case Study




    32
Case Study ( Acetone – Water )
    Separation of a binary tangent-pinch mixture
    Acetone dehydration system ( 70 mol % acetone feed )
    20,000 mole feed

    Subject to:
        Purity           ≥ 97%
        Recovery         ≥ 70%

    Minimise:
        Investment costs
        Annual operating costs




                                  33
Case Study Summary




                34
Case Study Summary
    Approach for multi-criteria process optimisation
     using Genetic Algorithm
    Allows determination of process alternatives
     through Pareto optimality
    Company specific costing can easily be included




                           35
Concluding Remarks
 For hybrid batch separation processes:
  Optimum synthesis and design procedure
  Multi-criteria optimisation


    Simple extension to continuous hybrid
     processes




                          36

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:6
posted:3/23/2012
language:
pages:36