Process Simulation
Process Simulation
Process Simulation
Answer: In future as the energy consumption expands and resources of material and energy become
more expensive and less certain, substantial changes are needed in many conventional chemical
processing plants. These have to incorporate a high degree of energy integration and achieve greater
efficiencies through process modifications, thus conserve material and energy resources. At the
same time protection standards are tightening, therefore, lower temperatures and pressures are
desired. Higher reaction rates, fewer stages and smaller plants will reduce burden of the material in
the process, leading to safer plants, With the growing consciousness about pollution, it is essential
to produce less effluents as well as to treat them properly.
For facing all these challenges, larger effort is needed on the part of process engineers in evaluating
alternative flowsheets. Fortunately, recent developments in computing procedures combined with
increasing capabilities of successive generations of computers enable chemical engineers to tackle
the process design problems on a much broader basis. Figure 1.1 shows a strategy of development
from conception to commercialization of a chemical process, in which thick blocks indicate the
steps totally computerized today and constitute elements of modern computer-aided-design (CAD).
The broken blocks will eventually be computerized in the course of time.
Answer: It is clear from Fig. 1.1 that the total process design effort can be broadly divided into
three stages:
1. Process synthesis
2. Process analysis
3. Optimization
1. PROCESS SYNTHESIS
In the first stage (shown by broken blocks in figure 1.1), a process concept is generated and a
flowsheet structure is primarily chosen with various types of equipment and their interconnections.
For this purpose, information from several sources is required, most important being the experience
piled up over several years in solving both simple and complex design problems in a certain area of
process industry. All this information, along with the knowledge of local conditions, and the facts
and relationships that govern behaviour of the processing systems should be put up to the maximum
use. The latter usually involve disciplines of chemistry, physics, mathematics and economics.
Process synthesis depends on design heuristics (trial and error methods) and on intuition of the
process engineers. This, however, can lead only to that upperbound of efficiency which is inherent
to an already synthesised system. On the other hand, there may exist other plausible system
structures (i.e. flowsheets) possessing higher upperbounds which should be synthesised in order to
achieve maximum benefit. For this reason, during the current years attention is being focused to
develop a general and systematic approach to process synthesis. Though still in a state of infancy,
process synthesis will be carried out in future on a more rational basis with the help of computer
aids.
2. PROCESS ANALYSIS
Once a flowsheet is synthesised, an analysis is required to solve material and energy balances for a
steady state process, sizing and costing the equipment and evaluating worth of the flowsheet. The
area termed as “steady-state simulation” in Fig. 1.1 has seen a phenomenal growth during the last
two decades. Also known as “flowsheeting”, chemical process simulation is represented by a
mathematical model with an aim to obtain information about response of a plant to various inputs.
The model is often applied in the form of a computer program for a continuous steady-state
deterministic system simulation. There are various features and requirements of general purpose
simulation. Important requirement of any successful simulation is availability of accurate properties
of pure chemical species and their mixtures. Dynamic simulation is needed in the analysis of control
aspects as well as for studying transient response during startup or shutdown of a plant.
3. OPTIMIZATION
Finally, optimization is performed leading to the final flowsheet with its optimal decision variable
values. This field has also witnessed tremendous development during the last three decades with
numerous techniques perfected for unconstrained and constrained optimization. The functional
optimization is required for calculating optimal control policies for implementation of feedforward
control through microprocessors and real time computers.
Question: What calculations are required for process design?
Answer: Among the most tedious and repetitive problems of process design is that of calculating
heat and mass balances. During the initial stages of a flowsheet study, whether of a new process or
an existing one, simple material balances may be sufficient. At this level, specifications may be set
from the plant data if an existing plant is being considered, or from engineering experience or pilot
plant data if a new process flowsheet is under development.
In the later stages of design, however, heat and mass balances must be calculated along with the
equilibrium and rate equations, P-V-T relationships and those governing counter current operations.
All these equations are generally strongly nonlinear. Moreover, in the simplest case if a process is of
a sequential configuration, as shown in Fig. 2.1-I, it is easy to proceed from the feed streams until
the products are obtained calculating sequentially for one process unit after the other. Unfortunately,
majority of the chemical plants are of complex configuration involving recycle of streams, mass
(Fig. 2.1-II) or energy (Fig. 2.1-III) or both; they represent interlinked networks of units. In such
situations for making calculation procedure sequential, it is essential to decompose the network.
Figure 2.1
Design mode is numerically less stable than that in the analysis or performance mode. Being
numerically more stable, the analysis mode is more often used to perform design calculations
iteratively for making case studies, Hence, majority of the application packages are written in this
mode of simulation.
Question: What is sequential modular approach for simulation?
Answer: The whole process is divided into units. These units are called modules. Calculations are
performed in sequence for each module starting from the feed stream until the product streams are
obtained.
The modular viewpoint involves collecting equations and constraints for each process unit into a
separate computational subroutine or module as well as the solution procedures for these equations.
This concept is parallel to that of a unit-operation, hence termed as a “unit-computation”. Thus each
module calculates values pertaining to the output streams for the given input conditions and
parameters for that process unit or equipment, irrespective of the source of input information or the
sink of output information. A typical module embedded in be a flowsheet network simulation
program is shown in Fig. 2.3. Using such a concept, it is thus possible to evolve a library of
modules and use them to simulate variety of flowsheets in a building block approach.
Figure 2.3: Unit Computation Module
The sequential part of the approach involves carrying out calculations from module to module,
starting with feed streams until products are obtained.
Dealing with recycle streams: Calculations cannot be performed in sequence from module to
module if recycle streams are present. In order to make it totally sequential, it is then necessary to
identity recycle loops of units in a given flowsheet and to "tear" certain streams, in other words,
assume values of the variables associated with these streams and adopt an iterative procedure which
should converge on the assumed values. For instance, tearing streams 2 and 7 in the network of
modules shown in Fig 2.4, the whole procedure becomes sequential in the order of modules (1), (4),
(3), (5) and (2).
Figure 2.4: Network of Modules with Tear Streams
A general purpose sequential modular program has the following structural components:
(a) A store of physical and thermodynamic properties for pure chemical species and their
mixtures.
(b) A system working data base derived from the data obtained from the store as well as the
problem input.
(c) Individual module data base derived from the working data base.
(d) Modules or subroutines for individual process units, computational procedures including
input/output for each module and internally iterated variables, if any.
(e) Routines for recycle calculations and convergence analysis.
(f) Other mathematical service routines.
Question: List some simulation programs using the sequential modular approach.
Answer: Some typical steady-state simulation programs using the sequential modular approach are
given as follows:
1. PACER 245
2. CHESS
3. FLOWTRAN
4. PROSIM
5. CONCEPT
6. ASPEN
Input Phase
In this phase, the steps executed are data input, data checking and storage. Data are organized up to
a maximum number of 100 streams. Maximum number of 50 equipment modules or nodes can be
simulated; the parameters of each node are given in a vector of 25 elements.
Preprocessing Phase
This phase involves retrieval from the data base of physical property estimation, coefficients for
various pure species present in the streams. Data are organised for a maximum of eight input or
output streams per node and a maximum of 20 components per stream.
Calculation Phase
This phase is executed under the supervision of the executive program by the various unit module
subprograms, which are aided by the thermodynamic and physical property package, TPACK. The
basic set of modules which allows any process to be modelled and simulated at an elementary level
consists of mixer, divider, splitter, phase separation/determination, heat exchange, pressure change
and reactor.
Output Phase
In this phase, tables with stream data, operational data of different equipment modules, error
messages and warnings about certain difficulties are printed.
Question: What is TPACK?
Answer: The thermodynamic and physical properties package includes 11 basic parameters for
each of the 100 pure components. Based on these, it computes for a designated stream vaporization
equilibrium ratio, single phase enthalpy, compressibility factor (liquid or vapor), dew point or
bubble point temperatures, and temperature at the given enthalpy.
Question: Compare sequential modular and simultaneous modular approaches for simulation.
Answer: The first comparison is with respect to the executive system required for each of the two
modular approaches. The sequential modular executive is easier to write. It has to work out a
computational order for the unit modules by tearing certain streams for loop calculations.
In the simultaneous modular approach no decomposition of streams is required. But its executive
has to arrange solving a set of linear balance equations in each iteration in addition to calculating
mathematical model the process units.
Secondly, the time needed by the two approaches is compared in simulating a particular problem. If
the system has no recycle streams, then sequential modular approach takes less time for calculations
than simultaneous modular approach. Based on the time consumed for various other processes the
recommended approaches are given the Table 2.2.
*****