THE COPPERBELT UNIVERSITY

School of mines and minerals sciences

CE 560 / MT 580
Process control

Lecture 07
Control strategies

C. Botha (Mr.)
Contacts: Chemical Engineering Department
Email: [email protected]
Alt. Email: [email protected]

January, 2024
 Ratio control

o Type of feedforward control used to maintain proportion of two streams in a variety of

industrial applications.

 Objective is to maintain ratio of two process variables at or near a specific value.

 Variables are usually flow-rates, one is manipulated and the other is a disturbance
variable.

 The ration in Eq. 7.1. is thus controlled rather than individual variables.
𝑢
𝑅= 7.1
𝑑
o Where 𝑢 and 𝑑 are physical variables and not deviation variables.

o Application examples of ratio control includes:

 Ratio control

 Specifying relative amounts of components

Disturbance (wild) stream, d
in blending.

 Keeping a specific reflux ratio in distillation. FT-1

dm
 Holding air-fuel ratio to a furnace at the Divider
Rm
RC
Ratio set point, Rd

optimum level. um

FT-2
 Keeping reactants to a reactor in a
stoichiometric ratio.
Manipulated stream, u

o In Fig.7.1, flow-rates for the disturbance Fig.7.1 Ratio control

and manipulated variables are measured.
 Ratio control

o Flow rates are then used to calculate the ratio

𝑢𝑚
𝑅= 7.2
𝑑𝑚

 Calculated ratio from the divider element is sent to the ratio controller.

 Controller compares this value to the desired ratio (set point) 𝑅𝑑 .

 Manipulated variable, 𝑢, is then adjusted accordingly.

 I.e. until the calculated ratio is as close as possible to set point ratio.

o When 𝑅𝑚 > 𝑅𝑑 , controller reduces flow rate 𝑢 until 𝑅𝑚 ≈ 𝑅𝑑 .

o And when 𝑅𝑚 < 𝑅𝑑 , controller increases flow rate 𝑢 until 𝑅𝑚 ≈ 𝑅𝑑 .

 Ratio control

o Ratio control typically uses a PI controller with the desired ratio as it set point.

o Since a measured ratio 𝑹𝒎 is calculated, a divider is included in the loop.

o Divider element results in process gain 𝑲𝒑 to vary in a nonlinear manner, i.e.,

∆𝑜𝑢𝑡𝑝𝑢𝑡 ∆𝑅𝑑 1
𝐾𝑝 = = = 7.2
∆𝑖𝑛𝑝𝑢𝑡 ∆𝑢𝑑 𝑑

o Process gain is inversely related to the disturbance, which is non-linear.

o Makes it difficult to properly tune the PI controller.

o Since controller will have to be retuned every time the disturbance changes.
 Ratio control

o This is the main disadvantage of the

configuration in Fig. 7.1. Disturbance (wild) stream, d

Ration station
o Hence, alternative configuration is
typically used, Fig. 7.2. FT-1 RS
Set point, usp
o Disturbance flow is measured and
transmitted to a ratio station. um
FC

o Measurement multiplied by FT-2

adjustable gain 𝑲𝑹 .

o Value of 𝑲𝑹 is desired ratio. Manipulated stream, u

Fig.7.2 Alternative ratio control strategy

 Ratio control

o Output from ratio station provides set point for the flow controller.

o Flow-rate of 𝑢 is adjusted so that it is as close as possible to 𝒖𝒔𝒑 .

o At this point, the two streams are in a ratio which is equal or close to the desired
ratio.

o During transient conditions, it may not be possible for the desired ratio to be
attained.

o A setback for both strategies depicted in Figs. 7.1. and 7.2.

o This is due to the dynamics associated with the flow control loop for 𝑢.
 Ratio control

o Thus, after a change in the disturbance, manipulated variable requires some time to
reach its desired value.

o The time required depend on the dynamics of the control loop.

o If the flow control loop has slow dynamics, a considerable time is required and vice
versa.

o Fortunately, flow control loops tend to have fast dynamics so that the settling time is
short.

o Although some transient mismatch between 𝑑 and 𝑢 will always be there but is
acceptable.
 SISO and MIMO control systems

o The control systems fall into one of the two categories:

 Single-input, single-output (SISO), and

 multiple-inputs, multiple-outputs (MIMO).

o E.g., blending process for which control objective is to keep the composition of the exit
stream at or within a desired range.

 Achieved by manipulating the flow-rate of pure A.

 Control strategy (whether feedback or feedforward is used) is a SISO system.

 SISO and MIMO control systems

o Since only one input and one output is involved, i.e., one manipulated variable (input)
and one controlled variable (output).

o If on the other hand the control objectives (controlled outputs) are many, then then
control system is a MIMO.

 E.g., keeping the outlet composition and liquid level within a desired range.

 Achieved by manipulating both flow-rates of feed stream and pure A.

 In which case, two controlled variables (outputs) and two manipulated variables
(inputs) are involved.
 SISO and MIMO control systems

o Figs. 7.3 and 7.4 illustrates MIMO control systems for different control strategies.

Disturbances Disturbances

Measured outputs
Manipulated variables (controlled variables) Controller

Process

Measured outputs
(controlled variables)

Process
Unmeasured outputs

Manipulated variables

Controller

Unmeasured outputs
Set points

(a) (b)
Fig.7.3 General structures for control configurations (a) MIMO feedback and (b) MIMO Feedforward
 SISO and MIMO control systems

Disturbances

Manipulated variables Measured outputs

Process

Unmeasured outputs
(controlled variables)

Estimator: computes
the values of the
Controller unmeasured
controlled variables

Fig. 7.4 General structure for a MIMO inferential control configuration

 More examples of MIMO control systems – Distillation column

o MIMO control systems typically require multiloop control.

o Since each controlled variable has to be paired with an appropriate manipulated
variable.
o Thus, several control loops are required for several pairs of controlled and
manipulated variables.
o E.g. MIMO control system is required for the distillation column in Fig. 7.5.
o Since it has five controlled variables and five manipulated variables.
o Controlled variables are as follows:
 product compositions, 𝑥𝐷 and 𝑥𝐵 ,
More examples of MIMO control systems – Distillation column

PT

Coolant
QD

LT hD AT

Reflux Distillate
R D
Feed xD

Heating fluid hB LT
Hexane
QB AT

Bottoms
B
xB

Fig. 7.5 Distillation column with multiple controlled and manipulated variables
 More examples of MIMO control systems – Distillation column

 liquid levels in the reflux drum and column base ℎ𝐷 and ℎ𝐵 , and
 the column pressure 𝑃.
o The manipulated variables are:
 Product flow rates, 𝐷 and 𝐵,
 Heat duties for condenser and reboiler, 𝑄𝐷 and 𝑄𝐵 , and
 The reflux ratio.
o Feed stream may come from an upstream unit and is a disturbance, thus cannot be
manipulated.
o Though, it can be measured and used in feedforward control.
 More examples of MIMO control systems – Distillation column

o Thus, five feedback loops would be required for control of the distillation column.
o Each loop using a single manipulated variable for each controlled variable.
o The controlled and manipulated variables can be paired in different ways.
o I.e., the total number of different multiloops could be 5! Or 120.
o But most of these control configurations are infeasible and can be eliminated.
o E.g., control configuration that attempts to:
 Control base liquid level by manipulating condenser heat duty.
 Control condenser liquid level by manipulating heat duty to the reboiler.
o There are criteria that can help to select the most promising loops.
 More examples of MIMO control systems - milling circuit

o Thus, two control strategies are possible:

1) Control product size by controlling ore feed rate, and circulating load amount.

 Variables associated with circulating load can also be controlled instead.

2) Control product size by simply controlling ore feed rate.

o Choice depends on which loop responds faster to disturbances.

 I.e., particle size loop or circulating load loop.

More examples of MIMO control systems - milling circuit

Water SP1

PC
FT

PS
FC
WC
Product
T

Rod mill Ball mill  Flow-rate

Feed  Density

FC
 Flow-rate
 level
 Density
F  Density
Water  Pressure

Pump speed
D=Density, F = Flow-rate and P = Pressure

Fig.7.5 Control strategy for mill product size