Process Instrumentation and Control System
Process Instrumentation and Control System
Instrument line
Alarms
PV PAH PIC
Actuator PT
PAL
Controller
Process or
utility stream
Final control Sensing
element element
Error Controller
Set + signal Function
point - generator
Input Output
Process
Controller computes error between input and set point and adjusts
output based on a control algorithm
P: Proportional SP
Controller output is proportional to error
The proportionality constant is called the controller gain
High gain gives fast response, but can lead to instability
Low gain can cause offsets time
I: Integral
Output is proportional to integral of error
Eliminates offsets from P control, but makes response more sluggish
D: Derivative
Output is proportional to rate of change of error
Damps out instability and allows higher gain to be used
Other functions such as summation, multipliers, and advanced models can
be easily coded in digital controllers
de
Controller Output = G e + (1 / T i) e dt - T d
o dt
where: G = Controller gain
e = Error (set point minus measured variable)
Ti = Integral time constant
Td = Derivative time constant
stoichiometry FFV
of distillation column
control to set reflux or
boil-up ratios
• Example: reactor
temperature (primary
controller) cascades
onto coolant flow
controller (secondary)
to control reactor
temperature
Easy Difficult
Temperature
L/L level
Pressure
pH
Flow rate
Certain components
oxygen, sulfur, hydrogen,
V/L Level CO
Pressure difference Composition
Conductivity Density
Voidage
THERMOCOUPLE HEAD
LEAD WIRE
A A + Cu
MILLIVOLT
TRANSMITTER
B B - Cu
HOT JUNCTION (T1 )
COLD JUNCTION (T2 )
Place a restriction in the flow path and measure the resulting pressure
drop using a differential pressure (PD) cell
If fluid properties are known, results can be calibrated to flow rates
PD PD
Sensor element
PD
Control Valves
Actuators
S D M
*AH
*AL Shared display device with software alarms (* is measured variable)
PIC
PV
If vapor has a high
loading of
condensable
material, then
pressure control is on
the vent gas stream
from the condenser PIC
PT
PT PV
Process
vapor
Process PIC
PT
PV
Coolant
FIC FIC
FV
PI M PI FT
FT
FIC
FIC
Vapor
Vaporizer flow control needs
LIC to prevent liquid accumulation
FT
FV Hence use level controller to
actuate heat input to the
LV
vaporizer and maintain a
LT constant inventory
Feed Control of liquid flow in is
easier than control of vapor
Steam flow out
Trap
Condensate
TV
TV
M M M M
TV
TO VARIABLE
TIC SPEED MOTOR TIC
TE TT TE TT
CONTROL CIRCUIT
Direct control of
LC distillate composition
by using tray
temperature to infer
composition and
TC control distillate flow
rate
Flow control on
(constant) boil-up
LC
rate could be set in
FC ratio to feed if feed
flow rate was highly
variable
Indirect control of
distillate composition
LC
by using tray
temperature to infer
TC
composition and
control reflux rate
Flow control on
LC
(constant) boil-up
FC
rate could be set in
ratio to feed if feed
flow rate varies
Direct control of
LC
bottoms composition by
FC
using tray temperature
to infer composition
and control bottoms
flow rate
TC
Flow control on
LC
(constant) reflux rate
could be set in ratio to
feed if feed flow rate
varies
Indirect control of
PC
LC
bottoms composition
FC by using tray
temperature to infer
composition and
control boil-up rate
Flow control on
TC
FV
FT
FIC
TE
FIC TT FY
FT FV
Intermittent
Steam drain
Intermittent Trap
charge
To vent
system
M
FIC
FV FT
Feed A
FIC TT
TIC FIC
FV FT
LAH LIC
LT
TE LA
FV L
FT
Feed B
Coolant
Temperature cascade
Product
control of coolant flow
Independent flow control
of feeds
Feed
Reactor
Product
Remember the simple flowsheet introduced in the lecture on
simulation?
Liquid feed is mixed with recycle gas, heat exchanged against reactor
effluent, heated to reactor temperature then passed over fixed bed of
catalyst. Product is cooled and liquid product is recovered.
Unconverted gas is recycled with purge to prevent accumulation of
inerts
Feed
Reactor
Product
What controllers would you use?
Where would you place them?
If the measured variable exceeds a preset limit an alarm should alert the
operator to take appropriate action
Alarm limits should be set far enough from normal process variation to avoid
nuisance alarms
If the measured variable exceeds a safe operating limit then an
automatic plant shutdown may be necessary
Shutdown limit should be set far enough from alarm limit that the operator has a
chance to respond to the alarm
But not so far that no time is left to safely shut the plant down
LIC LIC
LT LT
LAL LAL
TRIP
LSL LAL
UC
A
UC
A
All new plants and most older plants have the instruments and
controllers connected to a plant-wide digital control system (DCS).
Using the DCS:can update individual controller set points via shared
Operators
displays
Plant information systems can log data
Feedforward and multivariable predictive control can be
implemented more easily
Plant real-time optimization can be implemented through the
control system
Measured variable 2
Measured variable 3
Control Valve 1
Control Valve 2
Control Valve 3
FV
FT PT TT
Manual Control
Pneumatic Analog
Electronic Analog
Digital
?!?!?!?
Source: UOP
Source: UOP
The controls companies are currently putting a lot of effort into developing the devices and
standards for implementing wireless control
Optimizer
User User
Plant DCS
Off-line Optimization
Users take plant data and run the optimizer then send instructions to the
plant operators to update the DCS settings
Labor intensive and difficult to update more than daily
Optimizer Optimizer
Optimizer Optimizer