Chapter 3.

Brief Introduction to Distillation Control

All distillation columns have to be carefully operated in order to achieve the required production rates and
product quality. The 3 main objectives of column control can be stated as:

· To set stable conditions for column operation

· To regulate conditions in the column so that the product(s) always meet the required specifications
· To achieve the above objective most efficiently, e.g. by maximising product yield, minimising energy
consumption, etc.

Process variables like temperatures, pressures, flow rates, levels and compositions must be monitored and
controlled in all distillation processes. These process variables within a distillation system affect one
another, whereby a change in one process variable will result in changes in other process variables. Thus,
in column control one should be looking at the whole column and not focusing on any particular sections
only.

Each column has a control system that consists of several control loops. The loops adjust process
variables as needed to compensate for changes due to disturbances during plant operation.

Each of the process variables has its own control loop, which typically consists of a sensor and transmitter,
controller and control valve. See the Figure below. Each control loop keeps track of the associated process
variable. An adjustment is made to a process variable by varying the opening of its control valve. The
stream flow rate is therefore adjusted and a desirable variable is being controlled.

The sensor measures the process variable from the plant (i.e. plant data) and the transmitter sends the
information to the DCS (Distributed Control System) controller located in the control room.
The controller checks if the process variable agrees with the set point. If not, it will send corrective signal to
the control valve that will make adjustment in the plant so as to match the process variable to the set point.
This goes on continuously, essentially in a loop - hence the term "control loop".

[ Back on Top ]

Distillation Control Philosophy

Some of the general guidelines are noted below:

column pressure normally controlled at a constant value

feed flow rate often set by the level controller on a preceding column
feed flow rate is independently controlled if fed from storage tank or surge tank
feed temperature controlled by a feed preheater. Prior to preheater, feed may be heated by bottom product via
feed/bottom exchanger
top temperature usually controlled by varying the reflux
bottom temperature controlled by varying the steam to reboiler
differential pressure control used in packed columns to monitor packing condition, also used in tray columns to
indicate foaming
the compositions controlled by regulating the reflux flow and boiled-up (reboiler vapour)

Pressure is often considered the prime distillation control variable, as it affects temperature, condensation,
vapourisation, compositions, volatilities and almost any process that takes place inside the column. Column
pressure control is frequentlyintegrated with the condenser control system.

Reboilers and condensers are integral part of a distillation system. They regulate the energy inflow and
outflow in a distillation column.

[ Back on Top ]

A column is controlled by regulating its material and energy balances

Reboiler Control

This is required to provide good response to column disturbances, and to protect the column from
disturbances occurring in the heating medium. The reboiler boil-up is regulated either: (1) to
achieve desired product purity, or (2) to maintain a constant boil-up rate.

In a typical reboiler control (see Figure below), the control valve is located in the reboiler steam inlet line.
For inlet steam controlled reboiler, the heat transfer rate is regulated by varying the steam control valve
opening, thereby changing the steam condensing pressure and temperature. 

When an additional boil-up is required, the valve opens and raises the reboiler pressure, which increase the
temperature, and in turn increases the boil-up rate. This scheme has the disadvantage of non-linear
relationship between pressure and boil-up, and is affected by fouling in the reboiler.

[ Back on Top ]

An alternative is to control the condensate flow, i.e. by putting the control valve on the condensate line
( see Figure below). The main disadvantage is that this scheme has poorer dynamic response than the
previous scheme. Manipulating the inlet valve immediately changes the vapour flow, giving faster dynamic
response. On the other hand, the condensate outlet valve has no direct effect on vapour flow. The
response time varies with the condensate level in the exchanger.

The other main disadvantage is the sizing of the condensate valve. If condensate cannot be drained in
time, vapour flow may be restricted as much of the reboiler remains flooded. On the other hand, too fast of
condensate draining (faster than vapour condensation in the reboiler) as result in loss of liquid seal in the
reboiler and steam will pass into the plant's condensate recovery system.

[ Back on Top ]

Some reboiler control features the use of condensate pot. This is particularly important in fouling or
corrosive services (where leakage is a serious problem). An example is shown in the Figure below.
In the system shown, by varying the level control set point, the tube surface area in the reboiler that
is exposed for vapour condensation can be adjusted, thus changing the available heat transfer area. The
heat transfer rate can therefore be adjusted.

This arrangement also automatically minimise the condensation (and therefore, tube wall) temperature.
A pressure-balancing lineis provided to maintain a steady pressure and level in the condensate pot.

Condenser and Pressure Control

The 3 main methods of pressure and condensation control are:

(1) vapour flow variation, 

(2) flooded condenser, and
(3) cooling medium flow variation.

Vapour Flow Variation

The simplest and direct method for column producing a vapour product. The pressure controller regulates
the vapour inventory and therefore the column pressure. See the Figure below.

An important consideration here is the proper piping of the vapour line to avoid liquid pockets.

Flooded Condenser
This method is used with total condensers generating liquid product. Part of the condenser surface
is flooded with liquid at all times. The flow of condensate from the condenser is controlled by varying the
flooded area. Increasing the flooded area (by reducing flow) increases the column pressure (less surface
area for condensation).

[ Back on Top ]

Cooling Medium Flow Variation

Pressure can also be controlled by adjusting the flow of coolant to the condenser ( see Figure below).
Operation using cooling water can cause fouling problems at low flow condition, when cooling water
velocity is low and outlet temperature is high.

For air-fin condensers, the controller varies the fan speed or fan pitch to control pressure
(see Figure below). This arrangement is energy-efficient as it minimises fan power consumption, but
requires the use of variable-pitch fan or variable speed motor.
[ Back on Top ]

Other method: pressure control using inerts (see Figure below).

When column pressure falls, an inert gas is admitted to raise the column pressure. 

Or: split-range pressure control venting excess gas to flare (see Figure below)

[ Back on Top ]

In most instances, both vapour and liquid phase are present in the column overhead. The vapour contains
components that cancondense out but are undesirable in the liquid, i.e. excessive condensation may lead
to off-specification liquid product. In addition, it is also undesirable to lose liquid product (through insufficient
condensation) to the vapour. It is therefore important to control the rate of condensation to obtain the
desired vapour-liquid split.

This is usually done by controlling the temperature of the liquid product just downstream of the condenser.
One common scheme used is shown the Figure below.
Temperature Control

Column temperature control is perhaps the most popular way of controlling product compositions. In this
case, the control temperature is used as a substitute to product composition analysis.

Ideally, both top and bottom compositions should be controlled to maintain each within its specifications.
See the Figure below.

In practice, simultaneous composition control of both products suffer from serious "coupling"

(interaction) between the 2 controllers, resulting in column instability. In the system shown, suppose that
there are concentration changes in the feed conditions that result in lower column temperature. The top and
bottom temperature controllers will respond by decreasing reflux and increasing boil-up respectively.

If the actions of the 2 controllers are perfectly matched, and response is instantaneous, both control
temperatures will return to their set points without interaction.

However, the 2 actions are rarely perfectly matched, and their dynamics are dissimilar - usually the boil-up
response is faster. The reflux and boil-up will "cycle" as shown in the Figure above.

The interaction can be avoided by controlling only 1 of the 2 product compositions.

On-line analyser can be used together with temperature control to control product composition. The
principal control action is rapidly performed by the temperature controller, while the analyser slowly adjusts
the temperature set point to prevent off-specification product purity. A set up is shown in the Figure below.

In the above set-up, delayed analyser response is acceptable, as its time lags become a secondary
consideration. The fast temperature controller action renders this control method less sensitive to upsets
and step changes in an analyser-only control system.

Another advantage is that, should the analyser become inoperative, the temperature controller will maintain
automatic control of the process.

Feed Preheat Control

Feed preheat is usually practised for heat recovery or to attain the desired vapour and liquid traffic above
and below the feed tray. The objective of the preheat control system is to supply the column with a feed of
consistent specific enthalpy. With a single-phase feed, this becomes a constant feed temperature control;
with a partially vapourised feed, a constant fractional vapourisation is required.

As an example, consider case (a) as shown in the Figure below whereby the feed is a cold liquid. In this
case, all the liquid feed will go to the stripping section. In addition, because the feed is cold, it will also
condense some of the rising vapour.
As a result, the amount of liquid flow in the stripping section is much larger than the liquid flow in
the rectifying section. Thevapour flow in the rectifying section is lower than the vapour flow in
the stripping section because of the condensation into the liquid. 

[ Back on Top ]

The following Figures showed 2 other feed conditions: case (b) for saturated liquid (left) and case (c)
for vapour-liquid mixture(right):

And the following Figures showed 2 other feed conditions: case (d) for saturated vapour (left) and case (e)
for superheated vapour (right):

[ Back on Top ]

Subcooled feed or superheated feed can be controlled (see Figure below) by preheating (left)

or desuperheating (right) the feed prior to column entry:
A superheated bottom feed can be cooled by injecting a quench stream as shown in the Figure below.