Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Integrated Scheduling and Dynamic Optimization ofBatch Processes
Yisu Nie1 Lorenz T. Biegler1 John M. Wassick2 Carlos M. Villa2
1Center for Advanced Process Decision-makingDepartment of Chemical Engineering
Carnegie Mellon University
2The Dow Chemical Company
March 14, 2012
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 1 / 11
IntroductionBackground and motivation
Current industrial practice for batch production
Development of a batch process
Product recipescale-up−−−−→ Process recipe
scheduling−−−−−→ Production process
− Lost productivity and lowered performance in unit operations− Schedules sensitive to uncertainties in manufacturingI Opportunities for process optimization
Optimizing the performance of a batch process in real-time
Approach An optimization-based framework with integrated
Scheduling: assignment of batch operationsOptimal control: execution of batch operations
Benefit + Adding value to existing assets+ Improving plant profitability and reliability
Steps Off-line ⇒ On-line
Integrated optimizer
Batch plant
Schedule
Control strategy
⇓Integrated optimizer
Batch plant
Schedule
Control strategy
Feedback
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 2 / 11
IntroductionBackground and motivation
Current industrial practice for batch production
Development of a batch process
Product recipescale-up−−−−→ Process recipe
scheduling−−−−−→ Production process
− Lost productivity and lowered performance in unit operations− Schedules sensitive to uncertainties in manufacturingI Opportunities for process optimization
Optimizing the performance of a batch process in real-time
Approach An optimization-based framework with integrated
Scheduling: assignment of batch operationsOptimal control: execution of batch operations
Benefit + Adding value to existing assets+ Improving plant profitability and reliability
Steps Off-line ⇒ On-line
Integrated optimizer
Batch plant
Schedule
Control strategy
⇓Integrated optimizer
Batch plant
Schedule
Control strategy
Feedback
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 2 / 11
Problem DescriptionIntegrated scheduling and dynamic optimization of batch plants
GivenI A batch plant with existing equipmentI A time horizon to make productsI Knowledge of process dynamics
DetermineI The optimal production scheduleI The optimal equipment control strategy
Via1 Process representation using the state
equipment network (SEN)2 Mathematical optimization formulation
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 3 / 11
Problem DescriptionIntegrated scheduling and dynamic optimization of batch plants
GivenI A batch plant with existing equipmentI A time horizon to make productsI Knowledge of process dynamics
DetermineI The optimal production scheduleI The optimal equipment control strategy
Via1 Process representation using the state
equipment network (SEN)2 Mathematical optimization formulation
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 3 / 11
Problem DescriptionIntegrated scheduling and dynamic optimization of batch plants
GivenI A batch plant with existing equipmentI A time horizon to make productsI Knowledge of process dynamics
DetermineI The optimal production scheduleI The optimal equipment control strategy
Via1 Process representation using the state
equipment network (SEN)2 Mathematical optimization formulation
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 3 / 11
Problem Solving MethodologyIntegrated optimization based on the SEN
The SEN represents the process system as a directed graphconnecting two kinds of nodes
Material Feed, intermediate and final products
Equipment Process units carrying out operations
The integrated formulation
Objective function Max Profit = Product sales - Material cost - Operating cost
Constraints I Scheduling considerationsAssignment constraints, material balance, capacity constraints, timing
constraintsI Unit operation
Dynamic first-principle models, limits on controls and statesI Material quality measurement
Material blending, quality requirementsI Auxiliary tightening constraints
Tightening timing constraints, mass balance of process units
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 4 / 11
Problem Solving MethodologyIntegrated optimization based on the SEN
The SEN represents the process system as a directed graphconnecting two kinds of nodes
Material Feed, intermediate and final products
Equipment Process units carrying out operations
The integrated formulation
Objective function Max Profit = Product sales - Material cost - Operating cost
Constraints I Scheduling considerationsAssignment constraints, material balance, capacity constraints, timing
constraintsI Unit operation
Dynamic first-principle models, limits on controls and statesI Material quality measurement
Material blending, quality requirementsI Auxiliary tightening constraints
Tightening timing constraints, mass balance of process units
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 4 / 11
Case StudyA jobshop batch plant
Equipment units and products manufactured in the plant
Heater×1 Reactor×2 Filter×2 Distillation Column×1
Packaging Line×2
FeedA
Prod1
Prod3
Prod2
30(MU/ton)
50(MU/ton)
20(MU/ton)
5(MU/ton)
Maximizing the net profit within a 10-hour time horizon
MILP solved by CPLEX, MINLP solved by SBB(CONOPT)
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 5 / 11
Case StudyA jobshop batch plant
Equipment units and products manufactured in the plantHeater×1 Reactor×2 Filter×2 Distillation
Column×1Packaging
Line×2
FeedA
Prod1
Prod3
Prod2
30(MU/ton)
50(MU/ton)
20(MU/ton)
5(MU/ton)
Maximizing the net profit within a 10-hour time horizon
ModelStatistics
Type Var.(Discrete) # Nonlinear Var. # Cons. #
Recipe-based MILP 676(90) 0 1079Integrated MINLP 4978(90) 2292 12507
ModelSolution
Profit(MU) CPU time (s) Node(Best) # Gap(%)
Recipe-based 1374 0.366 288(199) 0.0Integrated 1935 9564 5000(1602) 67.9
MILP solved by CPLEX, MINLP solved by SBB(CONOPT)
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 5 / 11
Case StudyA jobshop batch plant
Equipment units and products manufactured in the plantHeater×1 Reactor×2 Filter×2 Distillation
Column×1Packaging
Line×2
FeedA
Prod1
Prod3
Prod2
30(MU/ton)
50(MU/ton)
20(MU/ton)
5(MU/ton)
Maximizing the net profit within a 10-hour time horizon
ModelStatistics
Type Var.(Discrete) # Nonlinear Var. # Cons. #
Recipe-based MILP 676(90) 0 1079Integrated MINLP 4978(90) 2292 12507
ModelSolution
Profit(MU) CPU time (s) Node(Best) # Gap(%)
Recipe-based 1374 0.366 288(199) 0.0Integrated 1935 9564 5000(1602) 67.9
MILP solved by CPLEX, MINLP solved by SBB(CONOPT)
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 5 / 11
Case StudyOptimal production schedules in Gantt charts
Recipe-based approach
Line2
Line1
Column
Filter2
Filter1
Reactor2
Reactor1
Heater
0 2 4 6 8 10
Time(hr)
<1> 80.0 <2> 80.0
<1> 35.0 <2> 35.0 <3> 35.0
<1> 34.4 <2> 45.0 <3> 45.0
<2> 69.4
<3> 80.0 <4> 80.0
<3> 60.0 <4> 60.0
<5> 37.0 <6> 25.0
<4> 39.5 <6> 50.8
HeatingReaction1Reaction2Filtration1Filtration2
Distillation1Distillation2Packaging1Packaging2Packaging3
Integrated approach
Line2
Line1
Column
Filter2
Filter1
Reactor2
Reactor1
Heater
0 2 4 6 8 10
Time(hr)
<1> 73.6 <2> 62.8 <3> 40.0
<1> 34.9 <2> 28.7 <3> 17.8 <4> 35.0
<1> 34.0 <2> 44.9 <3> 45.0
<2> 68.9
<3> 73.6 <4> 57.8 <5> 40.0
<3> 60.0 <4> 58.5 <5> 45.4
<5> 40.3 <6> 33.2
<4> 38.0 <6> 30.0
HeatingReaction1Reaction2Filtration1Filtration2
Distillation1Distillation2Packaging1Packaging2Packaging3
Numbers in a slot: <event pt.> batch size
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 6 / 11
Case StudyOptimal operating profiles for batch units
Optimal temperature profiles of Reactor 1
4.0 5.0 6.0 7.0 8.0 9.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8Sca
led
tem
p.
Time(hr)
Event slot 1: Reaction 1Recipe
Integrated
0.4 0.6 0.8 1.0 1.2
2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6Sca
led
tem
p.
Time(hr)
Event slot 2: Reaction 2
0.4 0.6 0.8 1.0 1.2
3.8 4 4.2 4.4 4.6 4.8 5 5.2 5.4Sca
led
tem
p.
Time(hr)
Event slot 3: Reaction 2
0.6 0.9 1.2 1.5 1.8
5.6 5.8 6 6.2 6.4 6.6 6.8 7 7.2Sca
led
tem
p.
Time(hr)
Event slot 4: Reaction 2
Dynamic profiles also obtained for Reactor 2 and Distillation Column
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 7 / 11
Industrial Application at DowOptimizing a real-world process with the integrated method
Methodology development
Main resultsI An integrated optimization formulation for batch scheduling and dynamic
optimization based on the state equipment networkI A proof-of-concept case study: verified benefits of the integration
Application at Dow: an alkoxylation process
Process characteristicsI Polymerization reactions in semi-batch stirred tank reactorsI Finishing train as a continuous processI Multiple products differ in composite monomers, molecular weights,
functionality, etc.
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 8 / 11
Industrial Application at DowOptimizing a real-world process with the integrated method
Methodology development
Main resultsI An integrated optimization formulation for batch scheduling and dynamic
optimization based on the state equipment networkI A proof-of-concept case study: verified benefits of the integration
Application at Dow: an alkoxylation process
Process characteristicsI Polymerization reactions in semi-batch stirred tank reactorsI Finishing train as a continuous processI Multiple products differ in composite monomers, molecular weights,
functionality, etc.
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 8 / 11
Industrial Application at DowAlkoxylation process
Process chemistry*
Key ingredientsI Epoxides (ethylene oxide (EO), propylene oxide (PO))
OO OOI Molecules containing active hydrogen atoms (alcohols, amines)
OH N
H
HI A basic catalyst (potassium hydroxide (KOH))
Basic proceduresI Starters are first mixed with catalyst in the liquid phaseI Alkylene oxides are fed in controlled rates
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 9 / 11
* Ionescu, M. Chemistry and Technology of Polyols for Polyurethanes; Rapra Technology: Shawbury, U.K., 2005.
Industrial Application at DowDevelopment of process dynamic models
Modeling reaction kinetics
Open literature example: Ethoxylation reaction*I Initiation
HOCH2CH2O−K+ + EO → HOCH2CH2OEO−K+
I Propagation
HOCH2CH2O(EO)−i K+ + EO → HOCH2CH2O(EO)−i+1K+
I Exchange
PiO−K+
growing chain+ PjOH
dormant chain
↔ PiOHdormant chain
+ PjO−K+
growing chain
P = HOCH2CH2(EO)
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 10 / 11
* Di Serio, M. Tesser, R. Dimiccoli, A. Santacesaria, E. Kinetics of Ethoxylation and Propoxylation of Ethylene Glycol Catalyzed
by KOH Ind. Eng. Chem. Res. 2002
Industrial Application at DowProposed future developments
Planned work for the project
Dynamic model developmentI Alternative monomers, copolymersI Vapor-liquid equilibriumI Molecular weight distributions
Scheduling method developmentI Discrete-time resource task network
Integration and on-line implementation
Conclusions and acknowledgments
Integration of scheduling and dynamic optimization using SENrevisited
Application at Dow
Thank you and any questions?
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 11 / 11
Industrial Application at DowProposed future developments
Planned work for the project
Dynamic model developmentI Alternative monomers, copolymersI Vapor-liquid equilibriumI Molecular weight distributions
Scheduling method developmentI Discrete-time resource task network
Integration and on-line implementation
Conclusions and acknowledgments
Integration of scheduling and dynamic optimization using SENrevisited
Application at Dow
Thank you and any questions?
Yisu Nie (Carnegie Mellon University) EWO meeting March 14, 2012 11 / 11