Introduction to API Process

Simulation
Pharmaceutical API Process
Development and Design
 Module Structure
• Process modeling basics
 Model applications
 Model types
 Modeling procedure
• Simulation packages
 DynoChem
• Examples
 Heat transfer
 Batch reactor with accumulation effects
 Model Applications
• Effects of process parameter changes
• Optimal operating policies for batch
operations
 Compare different reactant or solvent feed
strategies
 Maximization of yield in crystallization
 Minimize side-product formation in batch reaction
• Safety
 Loss of cooling
 Model Types
• Mechanistic (white box)
• Empirical (black box)
• Combined models (grey box)
• Lumped parameter
• Distributed parameter
• Continuous
• Discrete
• Hybrid discrete/continuous
 Modeling Procedure
1. Problem definition
a. Level of detail
b. Inputs and outputs
2. Identify controlling mechanisms
3. Evaluate problem data
a. Measured data
b. Parameter values
4. Construct model
5. Solve model
Controlling Mechanisms
1. Chemical reaction
2. Mass transfer
a. Diffusion
b. Boundary layer
3. Heat transfer
a. Conduction
b. Convective
c. Radiation
4. Fluid flow
5. Mixing
6. Evaporation
 Model Construction
1. System boundary and balance volumes
2. Characterizing variables
3. Balance equations
4. Transfer rate specifications
5. Property relations
 Model Components
1. Model equations and variables
a. Overall and component mass balances
b. Energy balance
c. Momentum balance
d. Transfer rates
e. Physical properties
2. Initial conditions
3. Parameters
 Software Packages
• Examples
 gPROMS, DynoChem, Daesim Studio, MATLAB
• Desired features
 Solution of differential algebraic equation systems
 Parameter estimation
 Optimization
 Model templates, physical properties estimation
• Software used for examples in this module
 DynoChem
 DynoChem Features
• Tools for simulation, optimization and
fitting
• Excel spreadsheets for data entry and
utility calculations
• Model library
 Templates for common API Unit Operations
• Utilities for physical properties, vessel
characterization
 DynoChem Model Structure
• Component Definitions
 Name, molecular weight, functional groups for
property calculations
• Process Definition
 Statements
• Scenarios
 Initial values, parameters
• Data sheets
 Profiles for measured variables
 Statements
• Phase
 Represents vessel (e.g. header tank,
condenser, receiving vessel) or compartment
(e.g. headspace)
 Solid, liquid, gas
• Flow
 Transfer, feed, remove
• Reactions
 Take place in phases or flows
 Statements (contd.)
• Heat transfer
 Heat or cool a phase with a jacket (flow)
 Heat exchange between phases
 Heat duty
• Mass transfer
 Liquid-liquid (transfer between immiscible
phases)
 Gas-liquid (e.g. hydrogen into solvent)
 Solid-liquid (e.g. dissolution)
 Statements (contd.)
• Condense
 V-L phase equilibrium (Antoine eqn)
• Calculate
 Set up user defined equations
• Integrate
 Integrate variables during a simulation
• Solver
 Solution method, accuracy
Example 1: Heat Transfer
Through Jacket
(see handout for detailed process
description)
 Balance Volumes

bulk
jacket

1. Bulk liquid bulk

2. Heating fluid jacket

 Assumptions and Controlling
Mechanisms
• Assumptions
 Neglect agitator work
 Neglect heat losses to environment
 Neglect evaporation
 Constant properties
• Controlling Mechanisms
 Flow of heating liquid
 Heat transfer between jacket and tank
 Perfect mixing
 Model Variables
M bulk Bulk mass
c p ,bulk Bulk specific heat
Tbulk Bulk temperature
F jacket Jacket mass flow rate
c p , jacket Jacket specific heat
T jacket ,in Jacket inlet temperature
T jacket ,out Jacket outlet temperature
 Heat Transfer Equations
M bulk  specified
c p ,bulk  specified
Fjacket  specified
c p , jacket  specified

F jacket c p , jacket T jacket,out  T jacket,in   (UA)Tlm  q


Tlm  T jacket,out  T jacket,in  ln Tbulk  T jacket,in  Tbulk  T jacket,out  
d
M bulk c p ,bulk  Tbulk   q
dt
 Model Objectives
1. Determine UA by fitting experimental data
2. Estimate time to heat bulk liquid to boiling
point for different jacket temperatures
 DynoChem Model Summary
• Components
 solvent (methanol), htfluid
• Process definition (statements)
 Phase bulk liquid
 Heat bulk liquid with jacket
• Scenarios (initial values and parameters)
 Bulk liquid: Initial temperature, solvent mass, specific
heat
 Jacket: Inlet temperature, flow, specific heat
 UA (to be determined by fitting data)
 Data Sheets
Jacket and Bulk temperature profiles

120

100

80 Jacket
see legend

Temperature C
60
Bulk liquid
Temperature C
40

20

0
12/8/06 6:57 AM 12/8/06 7:04 AM 12/8/06 7:12 AM 12/8/06 7:19 AM 12/8/06 7:26 AM 12/8/06 7:33 AM
Time, min
 Simulation Tool
• Requires UA value
• Obtain by fitting simulated temperature
profile to plant data
120

100 Jacket Temperature (Imp)

80 Bulk liquid Temperature (Exp)

60
Bulk liquid Temperature
40 (UA=400)
Bulk liquid Temperature
20 (UA=100)
0
0 5 10 15 20 25 30 35
 Fitting Tool
• Least squares fitting (Levenberg-Marquardt)
 Scenarios
• Compare heating time with different jacket
parameters
Heating time

70
Temperature (C)

60
Jacket Temperature=104
50
Jacket Temperature=120
40
Jacket Temperature=88
30
20
0 10 20 30 40 50 60
Time (minutes)
 Example 2: Fed-batch
reaction with safety constraint
(see handout for detailed process
description)
 Balance Volumes
feed1
header

bulk
jacket

1. Bulk liquid bulk

2. Heating fluid jacket

3. Header tank header

 Process Description
• Exothermic reaction
 substrate + reagent → product
• Isothermal operation, fed-batch
• Objective
 Minimize time to produce given amount of
product
• Manipulated variable
 Feed rate of reagent
 Model Variables
c X ,bulk concentration of species X in reactor;
Vbulk volume of material in reactor;
Vmax maximum volume;
qin feed rate;
c X ,header concentration of X in header tank;
k kinetic rate constant;
Tbulk reactor temperature (normal process operation);
MTSR Maximum temperature of synthetic reaction
(temperature attained after cooling failure);
Tmax maximum allowable temperature;
H r heat of reaction;
qr Reaction heat generation;
 density;
c p ,bulk heat capacity of material in reactor
 Safety Constraint
• MTSR (maximum temperature of synthetic reaction)
MTSR  Tbulk  Tad  Tmax

Tbulk
 Safety Constraint
• Cooling failure → Stop feed→ Reaction
continues till unreacted components are
exhausted
• Maximum attainable temperature
MTSR(t )  Tbulk  min(c substrate,bulk (t ), c reagent ,bulk (t )) (H r ) c p

extent of reaction after feed is stopped

• Without safety constraint, batch operation

(add all B at t=0) is optimal
Srinivasan et al., (2003), Computers and Chemical Engineering, 27(2003) 1-26
 Feed Profile
1 3
qinMax qinMax

2
qincon
qin

4
qinMin

time
• Max flow (1, 3): Volume and safety constraints are
inactive
• Controlled flow (2): Safety constraint is active
• No flow (4): Volume at maximum value

Srinivasan et al., (2003), Computers and Chemical Engineering, 27(2003) 1-26

 Reaction Equations
rate  Vbulk kc AcB

qr  rate  H r

MTSR(t )  Tbulk  min(c substrate,bulk (t ), c reagent ,bulk (t )) (H r ) c p

Heat transfer equations as in Example 1

 DynoChem Model Summary
• Components
 solvent, coolant, reagent, substrate, product
• Process definition (statements)
 Phase bulk liquid
 Heat bulk liquid with jacket
 Phase header tank
 Transfer to bulk liquid from header tank
 Reactions in bulk liquid
 Calculate MTSR
 DynoChem Model Summary
• Scenarios (initial values and parameters)
 Bulk liquid: Initial temperature, solvent mass,
specific heat, substrate moles, reagent moles
 Header tank: Temperature, solvent mass,
reagent moles
 Jacket: Inlet temperature, flow, specific heat,
UA
 Data Sheet for Simulation
• Adjust feed profile to satisfy MTSR and volume
constraints
• Isothermal temperature profile is imposed through data
sheet (DynoChem calculates required jacket temperature internally)
Feed and Temperature Profiles for Fed Batch Reactor

80

70

60

50
see legend

Qin L/hr
40
Temperature C
30

20

10

0
0 200 400 600 800 1000 1200 1400
Time, min
 Simulation Results
Controlled flow

Volume (l)

110
100
90
Volume (l)
80
70
60
0 200 400 600 800 1000 1200 1400
Time (min)

Maximum flow No flow

 Simulation Results
Safety constraint active Volume constraint active

MTSR

80.5
80
Temperature (C)

79.5
79
78.5 MTSR
78
77.5
77
76.5
0 200 400 600 800 1000 1200 1400
Time (min)

Safety and volume constraints inactive

 Scenarios
• Increase reactor volume, reduce cycle time
Volume constraint no longer active
120
110
100
Run1
Vol (l)

90
Run2
80
70
60
0 200 400 600 800 1000 1200 1400
Time (min)

70
60
50
Product (mol)

40 Run1
30 Run2
20
10
0
0 200 400 600 800 1000 1200 1400
Time (min)
 References
• Katalin Hangos and Ian Cameron, Process
Modeling and Model Analysis, Academic Press,
2001, London.
• P.E. Burke, Experiences in Heat-Flow
Calorimetry and Thermal Analysis, in W. Hoyle
(ed), Pilot Plants and Scale-Up of Chemical
Processes, Royal Society of Chemistry, 1997,
Cambridge.