Вы находитесь на странице: 1из 15



This chapter includes a fully-referenced review and discussion of previous studies which are
relevant to the research.


Distillation Column

Distillation is a physical process of separating the mixture into two or more products that
have different boiling points, by preferentially boiling more volatile components of the
mixture. When the liquid mixture of two volatile material is heated, the vapour that comes off
will have higher concentrations of volatile (i.e., lower boiling point) of liquid material from
where it has grown. On the other hand, if the vapour is cooled, the less volatile (i.e., the
higher the boiling point) materials have a tendency to condense on a larger share of the more
volatile substances.


Model Design

Feed is introduced approximately centrally into the global vertical cascade. Steam rising in
section at the top of the feed was washed with a liquid to remove or absorb the less volatile
component. Since no external substances are added, as in the case of absorption, washing

liquid in this case is provided by the condensation of steam removing from the top, which is
rich in the more volatile component. The liquid on the top tower is called reflux, and
materials are removed permanently is distillate, which may be vapour or liquid, rich in the
more volatile component. In the bottom of the liquid feed was stripped of volatile
components with steam generated at the bottom by partial evaporation of liquid in the bottom
reboiler. Rich liquid produced in less volatile component, or the bottom of the tray. Inside the
tower, liquid and vapour are always in bubble point and dew point, respectively, so that the
highest temperature is at the bottom and the lowest at the top. The figure below shows the
example of a typical distillation column model:

Figure 1: Distillation Column Model


Model Assumptions

1. Constant relative volatility. In this case the vapour-liquid equilibrium between any
two components, where is independent of composition. This assumption holds well
for the separation of similar components, for example, for alcohols or for

2. Constant molar flows. In this case the molar flows of liquid and vapour along the
column do not change from one stage to the next, that is, if there is no feed or product
removal between stages i and i+1.
2.1.3 Design Parameters

Column proposal in this work is designed to produce a distillate product with D = 0.5 mol /
min and the composition of the product yds = 0.99 and B = bottom with 0.5 mol / min and
composition xB = 0.01, from feed equimolal F = 1 mol / min of methanol / ethanol at 55C. =
So we have the following nominal conditions:

Feed rate F = 1 (mol / min)

Food composition ZF = 0.5 (unit mole fraction)
Feed temperature TF = 55C
YD = 0.99, P = 0.55 mol / min
xB = 0.01, B = 0.55 mol / min
With ZF = 0.5 and TF = 55C we calculate the pet food q: q = 1.0405
Reflux stream is L = 2.6889 mol / min and flow boilup is V = 3.2294 mol / min.
Detainees nominal liquid is Mi = 0.5 mol for all levels, including reboiler and

The relative volatility of methanol-ethanol system is considered to be 1.5.


Operating Points

Industrial process usually works with different quality in their products. In this case, we will
considers that the distillation column should get three different quality, that is, three different
operating point:












Simulation studies are often used to examine the operational behaviour of distillation
columns. A rigorous model for the simulation of the dynamic behaviour of distillation column
has been developed. This model allows the analysis of the influence of different disturbances
(failures) on the dynamic behaviour of a methanol/water column.



Simulink is an environment for multi domain simulation and Model-Based Design for
dynamic and embedded systems. It provides an interactive graphical environment and a
customizable set block libraries that let you design, simulate, implement, and test a variety of
time varying systems, including communications, controls, signal processing, video
processing, and image processing.
Simulink is integrated with MATLAB , providing immediate access to various tools that lets
you develop algorithms , analyse and visualize simulations, create batch processing scripts,
customize the modelling environment, and define signal, parameter, and test data.
Simulink provides a graphical user interface (GUI) for building models as block diagrams,
allows you to draw a model as you would with pencil and paper. Simulink also includes
Comprehensive block libraries sinks, sources, linear and nonlinear component, and connector.
The interactive graphical environment simplifies the process of modelling, eliminate the need
to devise a different, and the differences or similarities in the language program.
It is also hierarchical, so you can build them using both top-down and bottom-up approach.
You can see the system at a high level and then double-click the block to see the level of
improvement detailed model. This approach provides insight into how the model organized
and how its parts interact.


Distillation Column Model

The original model uses an S-Function block that calls a MATLAB function that implements
the system of differential equations. This nonlinear model has four manipulated inputs (L, V,
D and B), three disturbances (F, zF and qF) and 2N states returned by S-Function colas:

State 1: liquid composition in reboiler x1 = xB,

Then follow the stage compositions xi up the column, i = 2, , N-1
State N: composition stage N (condenser), xN = yD
State N+1: holdup reboiler, M1 = MB
Then follow the stage holdups up the column Mi, i = N+2, , 2N-1
State 2N: condenser holdup, M2N = MD

Figure 2: Distillation Column Simulink original model.



Several types of disturbances have been implemented in the simulation model. We can
specify the time at which the disturbance will be activated and the size, type (spike, ramp and
pulse) and duration of the disturbance.



Monitoring of chemical processes is becoming increasingly difficult as a result of the more

complex and larger scale operations. In this experiment, LabVIEW is used to monitor the
operation of a laboratory scale distillation column and to identify process state. Occurrence of
a fault will result in the deviation from the normal operating trajectory. Root cause
identification can also be performed through simple visualization of component planes.



LabVIEW software is suitable for any size or control system, and the heart of NI design
platforms. Integrates all the tools that engineers and scientists need to develop a variety of
applications in less dramatic, LabVIEW is a development environment for problem solving,
accelerated productivity and continuous innovation. NI LabVIEW system design software is
at the heart of the National Instruments platform. It provides a comprehensive tool that you
need to build any size or control applications in less time, LabVIEW is an ideal development
environment for innovation, discovery, and accelerated results. Combining the power of
LabVIEW and modular software, hardware reshaped to address the growing complexity
involved in providing measurement and control system on time and under budget.


Process Control

A process is the conversion of feed materials for products using chemical and physical
operations. In practice, the term tends to be used for both processing operations and
processing equipment. The main objective is to maintain control of the process in the
operating process required conditions, safely and efficiently, while satisfying environmental
quality and product needs. The subject of process control is concerned with how to achieve
this goal. In large-scale, integrated processing plants such as oil refineries or ethylene plants,
thousands process variables such as composition, temperature, and pressure is measured and
must controlled. Fortunately, a large number of process variables can usually be manipulated
for this purpose. Feedback control system compares the size with their then adjust the desired
value and the manipulated variable accordingly.


Proportional control

The key concepts behind proportional control are the following:

The controller gain can be adjusted to make the controller output changes as sensitive

as desired to deviations between se point and controlled variable

The sign of Kc can be chosen to make the controller output increase (or decrease) as
the error signal increases.

An inherent weakness of proportional only control is a steady-state error (or offset) applies
after the change set - points or disorder suffered. Basically, the offset can be eliminated
manually reset the set point ySP. However, this approach is not appropriate because operator
intervention is required and the new ySP usually have to be sought by trial and error. For
control applications where offsets can be tolerated, proportional-only control is attractive
because of its simplicity.


Integral control

Integral control measures are widely used because it provides an important practical
advantages, offset losses. When important measures were used, p automatically changes until
it reaches the required to make zero steady-state error. Reasonable condition always applies
unless the controller output or final control elements saturates and therefore cannot be
brought under control changed back to set point. Controller saturated or harassment occurs
when set- point the change is so great that it is beyond the range of the manipulated variable.
Therefore, an important control measures are commonly used in in conjunction with
proportional control as proportional - integral (PI) controller.

One disadvantage of using integral control is that it tends to produce oscillations of responses
controlled turns and reduce the stability of the feedback control system. A limited number of
swing usually acceptable because it is often associated with a more rapid response. That
undesirable effects were too many important actions can be avoided with proper tuning
controller or by acts including derivatives that tend to counteract the effects of instability.


Derivative control

Derivative control action function is to anticipate the future behaviour of the error signal
considering the rate of change. Note that a proportional controller reacts to temperature
deviations only, without any difference to the time period in which the deviation is growing.
Important control measures are not effective for sudden deviations in temperature due
corrective action depends on the duration of the deviation. Therefore, the derivative control is
never used alone; it is always used in conjunction with proportional or proportional control integral. To provide control measures expectations, modes tend issue to stabilize the
controlled process. Therefore, it is often used to overcome the tendency of instability
important mode.


Proportional-Integral-Derivative Control

The combination of proportional, integral, and derivative control mode as PID controllers can
be done in many variations. The three most common form is the form of parallel and series.
PID controllers are used in a loop where the signal is not noisy and where tight dynamic
response is important. Derivative action helps to offset the lag in the loop. Controller PID
temperature in the reactor normally. The controller senses the rate of movement of set points
and start moving the control valve earlier than the PI only.


Multivariate Statistical Process Control and Engineering Process Control

Statistical Process Control (SPC) and Engineering Process Control (EPC) are two techniques
that are used for improving process productivity and product quality by reducing the
variability of process from target while keeping it stable and under control.


Statistical Process Control (SPC)

Statistical process control, a technique widely used, accomplishes the above tasks to monitor
and track key changes in the behaviour of the system. It is an effective technique for
monitoring process variables as far as possible otherwise advised by an independent
statistical variable whose value falls around the deterministic.


Engineering Process Control (EPC)

Engineering process control attempts to reduce variability by controlling process variables to

ensure output of the target process. It is an ongoing procedure that adjusts the manipulatable
variables to ensure output at set point or target.


Integration of SPC and EPC

Integration of Statistical Process Control and Engineering Process Control got first attention
in 1988 when (MacGregor 1988, Box and Kramer 1992) proposed this concept of integration
and convinced the SPC research community that control charts can be used to monitor a
controlled system. He reviewed the two schemes, their similarities, overlap, contradictions,
reasons behind their isolation and the need to integrate them.

One of the popular schemes of SPC/EPC integration involves triggering of EPC controller
only when the SPC signals the presence of faults or assignable cause as termed by
statisticians; (Nembhard and Mastrangelo) were the earliest researchers who have advocated
that EPC-based process adjustments should only be triggered if SPC detects the out-ofcontrol state of the system which reflects the presence of fault(s).

The most powerful approach of SPC and EPC integration involves continuous adjustments
using EPC and detection of assignable cause using SPC monitoring. Several researchers have
explored different EPC techniques along with different control charts for this purpose.



The type and magnitude of disorders affecting the distillation column has a direct effect on
diversity of product. This is the main type of harassment faced in distillation column:

Feed composition upsets

Changes in the composition of the feeds that represent the most significant disruption to the
distillation column systems have to be deal on a continuous basis. Changes in feed
composition switch composition profile through the column which resulted in a huge upset in
the composition of the product. Most of the industrial column has no feed composition
analyser; therefore, the composition of the feed disorders usually appear as unmeasured


Feed flow rate upsets

The flow rate in the steady state model column with a continuous scale tray efficiency
continues with a feed rate. Therefore, the ratio control is an effective means to address the
upset flow rate of feed. Dynamic compensation is usually necessary to take into account
when the dynamics of the response does not match the composition of the product to feed and
flow rate changes in response to changes in the MV.


Feed enthalpy upsets

For column that use a low reflux ratio, enthalpy changes can significantly alter the feed rate
of vapour / liquid in the column, which led to major changes in the internal composition and
profile, therefore, a significant upset in the composition of the product. This upset can be
difficult to identify because most industrial column has no temperature measurements of feed

and though the measuring the temperature of feed available, it does not detect a change in
enthalpy of feed for two-phase feed.


Subcooled reflux changes

When a thundershower passes over a plant, the reflux temperatures for the columns can drop
sharply. Columns that use finned-fan coolers as overhead condensers are particularly
susceptible to rapid changes in ambient conditions. If internal reflux control is not applied,
severe upsets in the operation of the columns result because of major shifts in the
composition profiles of the columns.


Loss of reboiler steam pressure

When a steep drop in steam header pressure occurs, certain columns experience a sharp drop
in reboiler duty. This results in a sharp increase in the impurity levels in the products. When
the steam header pressure returns to its normal level, the composition control system for the
column attempts to return to the normal product purities. Because of the severity of this
upset, if the composition controllers are not properly tuned, the upset can be amplified by the
composition controllers, requiring the operators to take these controllers off-line to stabilize
the column, greatly extending the duration of the period of production of off-specification


Column pressure upsets

Column pressure has a direct effect on the relative volatility of the key components in the
column. Thus, changes in the column pressure can significantly affect product compositions.
A properly implemented pressure control scheme maintains column pressure close to its set
point, with only short-term and low-amplitude departures. A large class of columns is

operated at maximum condenser duty to maximize column separation, which minimizes

steam usage.