Process Systems Analysis

Module 4:
Computer Aided Solution of the Process Flow-sheet

Course Learning Outcomes

 Solve material and energy balances of a process flowsheet using
Process Simulators

Lecturer Name: Dr Youssef Touhami

Academic Year: 2022-2023
Course Coordinator Name: Dr Youssef Touhami

1
 Structure of a
Chemical Process Simulator
Role of Process Simulation
What is Simulation?

Simulation is the act of

representing some aspects of
the real world by numbers or
symbols which may be
manipulated to facilitate their
study
What Is A Process Simulator?
An Engineering Tool which performs,
* Automated calculations
* Material and/or energy balances
* Physical property estimations
* Design/rating calculations
* Process optimization
It is not a Process Engineer !
Solving Material and Energy Balances
using Flowsheeting Codes
Flowsheeting:
steady state process material and energy balances

Flowsheeting Package or Code:

the computer code for solving the material and
energy balance Equations in time domain or in
space domain
Material and Energy Balances
• The general balance equation
input + generation – output – consumption = accumulation
• Valid for Batch, Continuous and Semi batch
• The procedure for a single unit
– Define the basis
– Write the flowchart .. write all the known variables, label
unknowns
– Convert all the data in consistent units
– Perform the degree of freedom analysis
– Write the equations in an efficient solver and solve the system
– Calculate the quantities requested in the problem statement
Material and Energy Balances
• The general balance equation
input + generation – output – consumption = accumulation

• Balances on multiple-unit processes (recycle – bypass)

• Balances on reactive processes

– Molecular or atomic species
– Extent of reaction

• Single phase and multiple phase systems

Simple Example: Mixing Block routine

• Two streams are mixed adiabatically

– Each stream may contain any of five components (A, B, C, D, E)
– No phase changes take place
– The heat capacity of all components may be approximated as constants and the
heat of mixing can be neglected
– Calculate the component molar flow rates and T of the product stream from
specified values of these quantities for the feed streams
• Solution
– 5 + 1 (Temp) unknowns
– 5 material balance equations
– 1 energy balance DH = 0 equation
• MS Excel solution Mixer
• Fortran Code solution
• Process Simulator solution
Example: Cyclic Process
Na2 (mol A/s)
Nb2 (mol B/s)
Na1 (mol A/s) Na3 (mol A/s)
Nc2 (mol C/s)
P-1 Nb1 (mol B/s) Nb3 (mol B/s)
100 mol A/s
Nc3 (mol C/s)
20 °C 90 °C
T1 °C
30 °C

E-2 Separator
E-3 Pump
E-1 Reactor

N4 (mol/s)
0.9 (mol A /mol)
0.10 (mol B/mol)

85°C

• Gas-phase dehydrogenation of isobutane to isobutene

– Reaction: C4H10  C4H8 + H2
– Pure isobutane (A) is mixed adiabatically with a recycle containing 90% mole isobutane and the
balance isobutene (B)
– Single pass isobutane conversion is 35%
– All hydrogen (C) and 10% of the isobutane (as well as some isobutene) is sent to another part
of the plant
– Qr and Qs are the net rate of heat transfer in the reactor and separator
Cyclic Process – Degree of freedom
analysis Na2 (mol A/s)
Nb2 (mol B/s)
Na1 (mol A/s) Na3 (mol A/s)
Nc2 (mol C/s)
P-1 Nb1 (mol B/s) Nb3 (mol B/s)
100 mol A/s
Nc3 (mol C/s)
20 °C 90 °C
T1 °C
30 °C

E-2 Separator
E-3 Pump
E-1 Reactor

N4 (mol/s)
0.9 (mol A /mol)
0.10 (mol B/mol)

85°C

• Unit operations
– Mixing point:
• 4 unknowns (Na1, Nb1, N4, T1); 3 eq. (2 mat. Ba., 1 en. Bal.)  1 local d.o.f.
– Reactor
• 7 unknowns (Na1, Nb1, Na2, Nb2, Nc2, T1, Qr); 4 eq. (2 atomic species balances, 1 en.bal, 1 conv. spec.)  3
local d.g.f.
– Separator
• 8 unknowns (Na2, Nb2, Nc2, Na3, Nb3, Nc3, N4, Qs); 5 eq. (3 mat. bal., 1 en. Bal., 1 split spec.)  3 local
d.g.f.
• Overall process
– 7 local d.g.f. (1+3+3); 7 ties (Na1, Nb1, Na2, Nb2, Nc2, N4, T1 counted twice)  0 d.g.f.
Cyclic Process –
Degree of freedom analysis
Na2 (mol A/s)
Nb2 (mol B/s)
Na1 (mol A/s) Na3 (mol A/s)
Nc2 (mol C/s)
P-1 Nb1 (mol B/s) Nb3 (mol B/s)
100 mol A/s
Nc3 (mol C/s)
20 °C 90 °C
T1 °C
30 °C

E-2 Separator
E-3 Pump
E-1 Reactor

N4 (mol/s)
0.9 (mol A /mol)
0.10 (mol B/mol)

85°C

• MS EXCEL solution
– Initial estimate of the recycle set to 100 mol/s and assumed value of mixing point
temperature of 50°C.
– The value of N4a is varied until the calculated recycle flow rate (N4c) equals the
assumed value of N4a that drives the value of (N4a-N4c) to zero.
– Once the flow rates are correct, the mixing point temperature will be varied to
determine the values that drives DH =0 for the adiabatic mixer
• Process simulator solution
 Spreadsheets: very tempting
but be careful!
• Hidden equations?
• Range checking?
• Hidden/implicit defaults?
• Thermo physical properties?
• Do you REALLY know what it does?
• Has it really been validated on YOUR kind of
problem?
• Has it been properly documented?
Design & analysis through process
simulation/optimization

Divider

Mixer Reactor Separator

A chemical plant is our real world

Main steps in process simulation

Develop
process
model

Define
Problem Divider
Collect
additional
Separator data
Mixer Reactor
Analyze
results Chemical Plant

Solve model
equations
Steady state simulation - solve
algebraic equations
Develop
process
model
h(x, y) = 0

Define
Problem
Divider
Collect
additional
Mixer Reactor Separator data
Analyze
results
Chemical Plant

Solve model
equations -
AE solver
Process optimization - minimize
function s.t. constraints
Develop
process
model
h(x, y) = 0

Define
Problem Divider
Collect
additional
Mixer Reactor Separator data
Analyze
results Chemical Plant

Solve model
Min  (x, y)
equations -AE
solver + NLP s.t g(x, y)  0
method h1 (x,y) = 0
Dynamic simulation - solve ordinary
differential equations
dx/dt = f(x,y)
Develop
process
model
h(x, y) = 0

Define
Problem Divider
Collect
additional
Mixer Reactor Separator data
Analyze
results Chemical Plant

Solve model
equations -
DAE solver
Problem Definition -
What information do we need?

Divider

Mixer Reactor Separator

Chemical Plant Stream

data

Equipment
Mixture Unit Process Known data
compounds operations flowsheet data
Operating
data
 Problem Definition -
What do we need to select (from a
simulator)?

Divider

Mixer Reactor Separator

Chemical Plant Unit

models

Thermo-
Mixture Unit Process Library models
compounds operations flowsheet
Method of
solution
 Some Definitions

Unit 1 Unit 2 Unit 3 Unit 4 Unit 5

Module; Stream; Flowsheet decomposition; Partitions; Recycle

loop; Tear stream; Calculation sequence; Simulation approach;
Convergence technique
Process Simulation
• Process: a group of operations that transform
input streams into product streams by means of
chemical-physical transformations

• Simulation: the mathematical representation of

the reality by using a computer

• Dynamic process: a process which is studied in

the time domain rather than in steady state
Process Simulation
• Thermo-physical properties: the crucial point

• Data Banks: the basic value

• Unit operations: mathematical modelling

• Other modules such as optimization, numerical

procedures,...

• Cost estimation methods

• .......
The fundamentals

• Different possibilities for process simulation

– Steady state simulation

– Dynamic simulation

– Integrated steady state - dynamic simulation

The fundamentals

• Different stages in process design

– Process synthesis: identification of the optimal choice
of units and the connection between them, selection of
solvents,…

– Process design and simulation: establishment of the

optimal operating conditions and definition of each unit
operation

– Process analysis: study the effectiveness of design in

the light of operational consideration (flexibility,
controllability, safety, environmental effects, …)
The fundamentals

• Process simulation impact on industry

– The way engineering knowledge is used in processes

– The design procedure of the process (plant)

A typical flow-sheeting code:
steady state

Numerical Energy and Material

System Editor
Subroutines Balances
Manager
Graphics
Utilities
Reports
Equipment sizing
Forms

Cost Estimation Data Base

(physical
properties,
Economic and costs, etc.)
Environmental
Evaluation
Mathematically speaking
• n non linear material balances equations
• 1 energy balance non linear equation
• set of differential - algebraic equations (dynamic
simulators)
• In presence of:
– Very many components;
– Complex thermo-physical models for phase equilibrium
calculations
– A high number of subsystems (equipment)
– Rather complex equipment (distillation column,…)
– Recycle streams
– Control loops
Types of Solution Algorithms

• Sequential Modular
• Equation Solving (Simultaneous nonmodular)
• Simultaneous Modular

The Sequential Modular Algorithm

is the Most Widely Used.

29
Numerical strategies
• Equation oriented strategy - simultaneous solution
– Write down the entire set of equation
– Identify the constraints
– Solve the non linear system

• Sequential Modular approach

– Each subsystem is solved independently, starting from the first one
– Output streams for the solved subsystems are input streams for the next
subsystem
– Problems for the recycle streams (of material, energy and information)

• Combination of the two extreme approach

– Equation can be lumped into modules
– Modules can be represented by polynomials that fit input-output information
 Equation oriented flow-sheeting ...

• Solution of a set of non liner equations with constraints

• Definition of the matrix of the stream connection (process matrix)

• Definition of the inequality constraints

– Linearization of non-linear equations
– Process limits for Temperature, Pressure, concentration
– Requirements that variable be in a certain order
– Requirements that variables be positive or integer

• Define the procedure for determining the order in solving the

equations

• The treatment of feedback (recycles)

Equation oriented flow-sheeting

• Method of solution
– Newton Raphson
– Secant
• Tearing = selecting certain output variables from a set of equations
as known values so that the remaining variables can be solved by
serial substitution

• Partitioning = partition of equations into blocks containing common

variables

• Definition of initial guess

• Scaling the variables (the same order of magnitude)

• Scaling the equations (the same deviation from zero)

Sequential Modular Algorithm

• Most widely used algorithm

• Without recycle streams, only one flowsheet iteration produces a
converged solution.
 Recycle Streams in Sequential Modular
• Solving processes containing recycle streams is more
complicated.

Iterate flowsheet
repeatedly until
streams r1 and
r2 are within a
prescribed
tolerance.
Sequential modular approach
• Most common approach

• Each unit operation is described by a subroutine (or DLL or object)

• The output of a module is the input of the next module

• Other subroutines take care of

– equipment sizing and cost estimation
– numerical calculations
– handle recycle calculations
– optimize and serve as controllers for the whole set of modules

• Tearing is the process of solving the recycles by deciding which stream

should be interrupted and guessed.

• Partitioning: grouping of equations foe fast solving

• Fortran or C++ codes, VB and VBA

SMA: Advantages and disadvantages
• Advantages of sequential modular approach
– The flow-sheet architecture is easily understood because it closely
follow the process
– Individual modules can easily be added and removed
– Modules of different levels of accuracy can be substituted
• Drawbacks of sequential modular approach
– The input of a module is the output of a module: you cannot arbitrarily
introduce an output or input
– The modules need extra time to generate derivatives (perturbation of
the input)
– The modules may require a fixed procedure for the order of solution:
slow convergence
– Parameter specification is done with control loops: possibility of
introducing nested loops
– Phase equilibrium instability during the convergence of the process
may lead to inconsistency in the process specifications
• SMA is the most popular approach, but EO is the
most applied in modern Process Simulators
 Process simulation
and the Engineering Work Process
Process R&D

Plant Retrofits
Process
Synthesis

Process simulation

Process & Process

Equipment Design – Improvements –
Simulation Analysis
 The ‘life cycle’ of a process
Control
Process
System
Definition
Design

Control system and

Process Optimization Process integration

Dynamic simulation
Economic evaluation for training
Design estimation

Process Data reconciliation and

Synthesis model maintenance Start up
Plant
Operation
 Integration of
Process Synthesis & Process Simulation
Quantify “real” impact

Process Simulation
• Evaluate Ideas Operating Plant
Steady state, • Existing Performance
Dynamics

How far from the “best”?

Does it work?

Process Synthesis
Propose process • Generate Ideas Set
configuration Pinch technology objective
& design Train of columns targets

How do I change my system?

From a traditional way of using process
simulation …

Flow sheet design

Equipment Critical parameters definition (such as
distillation column stages, column diameter,…)
 … to the comprehensive use of Process
simulation in the entire ‘life’ of the plant
Control strategies design
Process parameters optimization ( --> ‘better’ processes)
Time evolution of the process (start up and shut down) ( --> risk analysis)
Operator training
Definition of procedure to reduce the unsteady state operations
Process synthesis and design
Maintenance optimization
Data Acquisition and Interface to ERP systems and supply chain interface
 Benefits of process simulation
• Partial or total replacement of Pilot Plant operations
– Reduction of the number of runs
– Runs planning
• Reduction of Time to market for the development of new
processes
– New processes
– Modification of existing processes (different solvent,...)
– Production of new materials
• Fast screening of process alternatives to select the best
solution
– economic aspects
– environmental aspects – environmental evaluations
– energy consumption aspects
– flexibility of the proposed process
Steady state simulators:
the core product
Directly used in
process design - evaluating process changes
analyzing what-if scenarios - Sensitivity analysis -
Optimization
And is the basis for
dynamic simulation - process synthesis - detailed
equipment design - off-line and on-line equation-
based optimization - application technologies for
vertical markets, e.g. polymers
Steady State Process simulation: the
procedure…
• Identify the problem: perform a detailed analysis
• Obtain all the relevant information
– Get process data: flow rates, operative conditions,
concentrations
– Get thermodynamic data:
• In house data
• Data Banks (Dechema, …) or literature
• Test run on laboratory / pilot plant
• Via estimation methods (be suspicious on Group Contribution
methods)
• Via molecular modeling
– Get kinetic data
• Directly from pilot plant
• from excess Gibbs energy calculations (if possible)
• directly from plant data

• TIP: avoid a rigorous definition of kinetic model and use concept of

yield and conversion wherever possible and reasonable, at least in
the first stage of the development
… Process simulation: the procedure ...
• Select the software
– Stady state simulators
• Aspen Plus (Aspentech)
• PRO II (Sim Sci)
• Hysys.Process (Hyprotec – AEA Technology)
• Chem CAD
• Process
• .....
– Dynamic simulators
• Speedup  Aspen Dynamics (Aspentech)
• Batch model and DynSim (Sim Sci)
• Hysys.Plant (Hyprotec – AEA Technology)
• gPROMS (PSE)
• ABACUS (MIT)
• .....
– Integrated solution
• Aspen Dynamic, Hysis, Pro II, ...
• Select the Hardware
… Process simulation: the procedure…

• Training
– Material and energy balances
– Basic course on process simulation
– Thermodynamic, phase equlibria and model selection
– Specific topics in thermophysical property calculation
• Electrolytes
• Polymer systems
• Kinetic data and kinetic modeling
– Specific topics in unit operation modeling
• heat exchangers design
• Batch distillation and reaction
• heat integration
– Process dynamics and control and dynamic process simulation
– Economic factors, cost analysis and energy consumption
– Batch process modelling
– Environmental impact and environmental evaluation
– Control system
… Process Simulation: the procedure…
• Run the base case
– The base case is a complete representation of the plant (or
part of the plant) in which the goals are
• the closure of the degree of freedom analysis
• the convergence of all the variables
– The base case may not correspond to the goals to be
achieved
– The base case is the starting point of ‘what if analysis’
– The base case is the validation procedure for
thermodynamic models and for data used
– The base case is the validation for the unit operation models
used in terms of:
• Reliability
• Convergence and numerical methods
• Completeness of the results
– The base case is the validation for the reaction modeling
and for the kinetic assumption
… Process Simulation: the procedure…
• Fine tune the base case by adjusting
– Thermodynamic models and parameters (thermodynamic
analysis)
– Kinetic models
– Lumping of components
• Assign Design Specification to each block
– Assign internal design specification which involves variables and
equation of specific blocks, such as solvent recovery, outlet
temperature, number of stages, reflux ratio,…
– The simulator will close again the degree of freedom analysis by
considering the new design specification
– The simulator will solve the unit operation by varying an
indicated variable so that the design specification is met
– Different way of performing such task in different process
simulators
… Process Simulation: the procedure…

• Perform Sensitivity analysis

– A sensitivity analysis is the observation of a process condition
consequent to the variation of a single process variable (partial
derivative)
– One can perform one sensitivity per each variable
– One can observe the effect of the variation in several process
conditions
– Sensitivity analysis is of paramount importance for determining
the most ‘sensitive’ variables for an optimization of the process
– Plot of the results are normally prepared
Sensitivity Analysis
• Your most useful ally in identifying problems and
assessing risk is the Sensitivity Analysis (Tony Perris)
– “what if …?”
– vary the input data ( and even the models)
– explore the defaults
– vary the assumptions
– bound your region of confidence
– explore combinations of uncertainties

• Sensitivity Analysis ….
– identify key sensitivities & their engineering consequences
(efficiency vs. structure)
– focus on what really matters (ie. what has a serious effect)
– allocate design margins accordingly
… Process simulation; the procedure…
• Design specification for the entire process
– Is the natural result of a correct sensitivity analysis
– Allows to define a given process condition by modifying
a variable in any part of the process
– IS a one – to – one relationship between specification
and variable

• Case Studies
– Run the simulation with different operating condition
– Different from the sensitivity analysis
– The simulation converges completely for each process
condition
… Process Simulation: the procedure.
• Optimization
– Is the procedure that allow to find the best conditions in
terms of a well defined objective function
– The objective function may be:
• Energy consumption
• Operating cost
• Equipment cost
• Environmental impact
• ….
– The number of variables is more than one
– Points of attention:
• Multiple minimum
• Feasible and non-feasible path
• Numerical convergence
Process simulation: the recipe.
• Design the flow sheet connectivity
• Components definition
• Physical - Chemical properties definition
• Feed conditions definition
• Unit operation internal definitions for each unit
operation involved in the process
• Process specification definition (design specifications)
• Analysis tools
– Sensitivity analysis
– Case studies
– Optimization
• Cost analysis and evaluation
Structure of a Process Simulator
Results obtainable
 Verification of the process operating conditions
 Information on intermediate streams (not measured)
 Enthalpy balances information
 Verification of the plant specifications
 Validation of phase equilibrium models for the real system to
be used in similar conditions
 Influence of the operative parameters on the process
specifications
 Process De-bottlenecking for each individual section
 A priori Identification of process control strategies and tuning
of instrumentation
 Possibility to verify security systems behavior for variation of
process condition
Problems involved
• Availability of all the required thermodynamic properties of the
pure components involved
• Definition of an accurate thermodynamic model (Equations of
state or Excess Gibbs energy model) for binary and multi-
component mixtures
• Availability of all the necessary interaction parameters
• Availability of all the necessary unit operations modules
• Necessity of defining dummy operations, non always easy to
identify
• Tear streams identification to achieve rapid convergence if in
presence of recycles
• Necessity of defining user models and user thermo (In C++,
FTN or VB - VBA)
Conclusions
• Process simulation is a powerful methodology for
– Material and energy balances in steady state conditions
– Material and energy balances in dynamic condition
– Investigation of process dynamics and batch process
– Implementation of a control strategy

• Process simulation is applicable in different field of the

process engineering
– Analysis of existing processes (optimization, de-bottlenecking,...)
– Synthesis of new processes (solvent selection, environmental impact,...)
– Operator training, process dynamics start up and shut down ...

• Process simulation is applicable in the framework of

environmental impact study and sustainable development
– What if analysis
– Safety analysis
– Environmental evaluation
– New and cleaner processes investigation
Simulation Trends
• Move to increase simulation usage in operating
environment
• Increased trend to integrate and link applications
• Desire for common interfaces to reduce learning curve
• Linking of simulation technology to control systems
and plant data
• Integration of steady state, dynamic and real time
optimization technologies
• Support of enhanced engineering workflow and
productivity
• Enables economic analysis and optimization
• Enhances design and operational decisions
• Use web infrastructure to improve/monitor process
Some VERY IMPORTANT Principles:

 the program is an AID in making calculations and

decisions: YOU must ensure that it is “fit for
purpose”
 YOU are responsible for any results you generate
and for any use which you make of them
 it is your PROFESSIONAL, ETHICAL & LEGAL
RESPONSIBILITY to take care and to exercise good
judgement
 in essence a program is no different to a pencil &
paper!

From ‘Good Practice Guidelines’ – Tony Perris

User’s Guidelines: Some Very Basic Advice!

• Think very hard about your OBJECTIVES

– Your objective is NOT to develop a model, it is to solve an
engineering problem? Never lose sight of this!
• Keep it SIMPLE
– Start EXTREMELY simple: this needs less data, is easier to get
going, provides a sound basis for anything more complex and
may even satisfy your objectives
• Explore SENSITIVITIES
• Add detail SELECTIVELY
– ONLY where you NEED it and only a bit at a time: adding detail
means adding problems! Detailed models need more data,
which you often don’t have and have to guess anyway

From ‘Good Practice Guidelines’ – Tony Perris

User’s Guidelines:Solving the equations
• has the calculation converged
– and check what that actually means: converged
doesn’t necessarily mean right!?
– Any sign of asymptotic behaviour?
• make sure you understand any error/warning
messages
• problems?
– check the input data again
– is the combination of specifications physically
realistic?
– could you use a simpler model to “get things moving
in the right direction”?
User’s Guidelines: Checking the results

• Assess the results against your defined

objectives:
– do they make engineering sense?
– are they within expected bounds?
– are they within valid ranges of (eg) thermo?
– can all “surprises” be explained in engineering
terms?
• Check sensitivity of your results to input data,
assumptions, etc
• Add detail where NECESSARY
Q: Why PS is important?
• Because it represent with high accuracy the real
world
• Because it allows to focus on the interpretation of
the results rather than on the methods for
obtaining the results
• Because it allows a global vision on the process
engineering by assembling theories and models
• Because it is essential in the design of new and
existing processes
• Because it is a simple tool for treating real cases
• .....
User’s Guidelines: Input data: KIS
• get the units of dimension right!
• make sure the specifications are consistent and
feasible!
• does the program correctly understand your
problem definition? (check any data-
reprint/summary very carefully)
• does it use any default values & are they suitable
for your problem?
• ESPECIALLY check the thermophysical property
data (and then check it again!)
• old models & data for a new problem? Be careful!
Q: Which are the limitations?
• No equipment design nor momentum balances
• Models of some important units missing
• Multiple solutions for flowsheeting and poor recycle convergence
• PS is a tool: it cannot interpret results
• Thermodynamic modeling is of paramount importance (accuracy
and reliability is still questionable…)
• Kinetic models
• Availability of parameters must be checked
• Calibration of the results versus existing data is necessary
• Time scale of some phenomena are still crucial (reaction, fluid
dynamic,..)
• Changing results depending on simulators release data

• Avoid the GIGO….. Approach = garbage in gospel out

Q: Common Sources of Problems?

• model/program does not apply to your problem

• errors in input data, inappropriate defaults
• over-complication: Keep It SIMPLE!
• units of measurement
• convergence problems: ignoring error & warning
messages
• applying models or data outside their range of validity
• failure to consider transients
Q: Common Sources of Problems?
• Thermophysical Properties
– Basic validity of model?
– Data from different sources (especially mixture data)?
• Is the data valid for YOUR mixture under YOUR conditions?
• Potential Azeotropes (real or imaginary!)?
– Extrapolation properties:
• Ranges of validity ( composition too)?
• Anywhere near any of the critical points?
• Transients?
• Convergence paths?
– Trace components can be important

• Sensitivity analysis
– “Pre-flight drill”!
 Industrial Processes are complex

To get benefits from Process simulations one

must CRITICALLY SIMPLIFY THE PROCESS

 The need of engineering knowledge and

experience
Process simulation is a simple and helpful
tool…

... to be used by chemical engineers that

fully understand the process

When using a mathematical model, careful

attention must be given to the uncertainties in
the model – Richard Feynman
References
1. Gil Chaves, I.D., López, J.R.G., García Zapata, J.L., Leguizamón Robayo A.
& Rodríguez Niño, G. (2016). Process analysis and simulation in chemical
engineering. Springer International Publishing Switzerland.

2. Chemmangattuvalappil, N. G., Chon, C. H., Sum, D. N. K., Elyas, R., Chen,

C-L., Chien, L., Lee, H-Y, Elms, R. D. (2017). Chemical engineering process
simulation. Elsevier.