Unit 9: Describe the Operation of Programmable Logic Control

(PLC) and Distributed Control System (DCS).

Module 9.2: Describe the Operation of Distributed Control System

(DCS).

Terminal Objective
Terminal Objective
9.2.1 Given access to manufacturer's
Given access to manufacturer's documentation, the trainee will
documentation, the trainee will correctly describe the operation of
correctly describe the operation of Distributed Control System
Distributed Control System

DISTRIBUTED CONTROL SYSTEMS

Generally, the concept of automatic control includes accomplishing

two major operations; the transmission of signals (information flow) back
and forth and the calculation of control actions (decision making).
Carrying out these operations in real plant requires a set of hardware and
instrumentation that serve as the platform for these tasks. Distributed
control system (DCS) is the most modern control platform. It stands as
the infrastructure not only for all advanced control strategies but also for
the lowliest control system. The idea of control infrastructure is old. The
next section discusses how the control platform progressed through time
to follow the advancement in control algorithms and instrumentation
technologies.

1. Historical Review
To fully appreciate and select the current status of affairs in
industrial practice it is of interest to understand the historical perspective
on the evolution of control systems implementation philosophy and
hardware elements. The evolution concerns the heart of any control
system which is how information flow and decision making advanced.

1. Pneumatic Implementation: In the early implementation of

automatic control systems, information flow was
accomplished by pneumatic transmission, and computation
was done by mechanical devices using bellows, spring etc.
The pneumatic controller has high margin for safety since
they are explosion proof. However, There are two
fundamental problems associated with pneumatic
implementation:

1 PCSTP
 • Transmission: the signals transmitted pneumatically
(via air pressure) are slow responding and susceptible
to interference.
• Calculation: Mechanical computation devices must be
relatively simple and tend to wear out quickly.

2. Electron analog implementation: Electrons are used as the

medium of transmission in his type of implementation mode.
Computation devices are still the same as before. Electrical
signals to pressure signals converter (E/P transducers) and
vice-versa (P/E transducers) are used to communicate
between the mechanical devices and electron flow. The
primary problems associated with electronic analog
implementation are:

• Transmission: analog signals are susceptible to

contamination from stray fields, and signal quality
tends to degrade over long transmission line.
Calculation: the type of computations possible with electronic analog
devices is still limited.
3. Digital Implementation: the transmission medium is still
electron, but the signals are transmitted as binary numbers.
Such digital signals are far less sensitive to noise. The
computational devices are digital computers. Digital
computers are more flexible because they are programmable.
They are more versatile because there is virtually no
limitation to the complexity of the computations it can carry
out. Moreover, it is possible to carry out computation with a
single computing device, or with a network of such devices.

Many field sensors naturally produce analog voltage or current signals.

For this reason transducers that convert analog signals to digital signals
(A/D) and vice versa (D/A) are used as interface between the analog and
digital elements of the modern control system. With the development of
digital implementation systems, which DCS are based on, it is possible to
implement many sophisticated control strategies on a very fast timescale.

2 PCSTP
2. Modes of Computer control

Computer control is usually carried out in two modes: supervisory

control or direct digital control. Both are shown in Figure 1. Supervisory
control involves resetting the set point for a local controller according to
some computer calculation. Direct digital control, by contrast, requires
that all control actions be carried out by the digital computer. Both modes
are in wide use in industrial applications, and both allow incorporating
modern control technologies. Measurements are transmitted to computer
and control signals are sent from computer to control valves at specific
time interval known as sampling time. The latter should be chosen with
care.

signals from
digital
computer

Local PID
FC
controller

Supervisory Control mode

Flow measurement to computer

valve
setting
from
compute
r

Direct digital Control mode

Figure 1: Computer control modes.

3 PCSTP
Computer Control Networks
The computer control network performs a wide variety of tasks:
data acquisition, servicing of video display units in various laboratories
and control rooms, data logging from analytical laboratories, control of
plant processes or pilot plant, etc. The computer network can be as simple
as an array of inexpensive PC's or it could be a large commercial
distributed control system (DCS).

Small Computer Network

In small processes such as laboratory prototype or pilot plants, the
number of control loops is relatively small. An inexpensive and
straightforward way to deal with the systems is to configure a network of
personal computers for data acquisition and control. An example
configuration of a PC network control system is depicted in Figure 2. The
network consists of a main computer linked directly to the process in two-
way channels. Other local computers are linked to the main computer and
are also connected to the process through one-way or two-way links.
Some of these local computers can be interconnected. Each of the local
computers has a video display and a specific function. For example, some
local computers are dedicated for data acquisition only, some for local
control only and some other for both data acquisition and local control.
The main computer could have a multiple displays.

4 PCSTP
 All computers operate with a multitasking operating system. They
would be normally configured with local memory, local disk storage, and
often have shared disk storage with a server.

Multiple Display

Main Computer

Local
Data control
acquisitio Display
n Display l
PROCESS

Data Data
acquisition acquisition

Local
control
Display Display

Figure 2: PC network

Programmable Logic Controllers

Programmable logic controller (PLC) is another type of digital
technology used in process control. It is exclusively specialized for non-
continuous systems such as batch processes or that contains equipment
or control elements that operate discontinuously. It can also be used for
many instants where interlocks are required; for example, a flow control
loop cannot be actuated unless a pump has been turned on. Similarly,
during startup or shutdown of continuous processes many elements must
be correctly sequenced; that is, upstream flows and levels must be
established before downstream pumps can be turned on.

The PLC concept is based on designing a sequence of logical

decisions to implement the control for the above mentioned cases. Such a
system uses a special purpose computer called programmable logic
controllers because the computer is programmed to execute the desired
Boolean logic and to implement the desired sequencing. In this case, the
inputs to the computer are a set of relay contacts representing the state
of various process elements. Various operator inputs are also provided.

5 PCSTP
 The outputs from the computer are a set of relays energized
(activated) by the computer that can turn a pump on or off, activate lights
on a display panel, operate solenoid valve, and so on.

PLCs can handle thousands of digital I/O and hundreds of analog

I/O and continuous PID control. PLC has many features besides the digital
system capabilities. However, PLC lacks the flexibility for expansion and
reconfiguration. The operator interface in PLC systems is also limited.
Moreover, programming PLC by a higher-level languages and/or capability
of implementing advanced control algorithms is also limited.

PLCs are not typical in a traditional process plant, but there some
operations, such as sequencing, and interlock operations, that can use the
powerful capabilities of a PLC. They are also quite frequently a cost-
effective alternative to DCSs (discussed next) where sophisticated process
control strategies are not needed. Nevertheless, PLCs and DCSs can be
combined in a hybrid system where PLC connected through link to a
controller, or connected directly to network.

Commercial Distributed Control Systems

In more complex pilot plants and full-scale plants, the control loops
are of the order of hundreds. For such large processes, the commercial
distributed control system is more appropriate. There are many vendors
who provide these DCS systems such as Baily, Foxboro, Honeywell,
Rosemont, Yokogawa, etc. In the following only an overview of the role of
DCS is outlined.

Conceptually, the DCS is similar to the simple PC network.

However, there are some differences. First, the hardware and software of
the DCS is made more flexible, i.e. easy to modify and configure, and to
be able to handle a large number of loops. Secondly, the modern DCS are
equipped with optimization, high-performance model-building and control
software as options. Therefore, an imaginative engineer who has
theoretical background on modern control systems can quickly configure
the DCS network to implement high performance controllers.

A schematic of the DCS network is shown in figure 3. Basically,

various parts of the plant processes and several parts of the DCS network
elements are connected to each other's via the data highway (fieldbus).
Although figure 3 shows one data highway, in practice there could be
several levels of data highways. A large number of local data acquisition,
video display and computers can be found distributed around the plant.
They all communicate to each other's through the data highway. These
distributed elements may vary in their responsibilities. For example, those
closest to the process handle high raw data traffic to the local computers
while those farther away from the process deal only with processed data
but for a wider audience.

6 PCSTP
 The data highway is thus the backbone for the DCS system. It
provides information to the multi-displays on various operator control
panels sends new data and retrieve historical data from archival storage,
and serves as a data link between the main control computer and other
parts of the network.

On the top of the hierarchy, a supervisory (host) computer is set.

The host computer is responsible for performing many higher level
functions. These could include optimization of the process operation over
varying time horizons (days, weeks, or months), carrying out special
control procedure such as plant startup or product grade transition, and
providing feedback on economic performance.

Supervisory
(host)
Computer

Archival Operator Main Operator

Data Control Control Control
Storage Panel Computer Panel

Data
To other highway To other
Processes Processes
Local data acquisition
Local Computer Local and Control
Local Display Computer computers
Local
Display

PROCESS

Figure 3: The elements of a commercial distributed control system

network

A DCS is then a powerful tool for any large commercial plant. The
engineer or operator can immediately utilize such a system to:
• Access a large amount of current information from
the data highway.

• See trends of past process conditions by calling

archival data storage.

7 PCSTP
 • Readily install new on-line measurements together
with local computers for data acquisition and then
use the new data immediately for controlling all
loops of the process.

• Alternate quickly among standard control strategies

and readjust controller parameters in software.

• A sight full engineer can use the flexibility of the

framework to implement his latest controller design
ideas on the host computer or on the main control
computer.

In the common DCS architecture, the microcomputer attached to

the process are known as front-end computers and are usually less
sophisticated equipment employed for low level functions. Typically such
equipment would acquire process data from the measuring devices and
convert them to standard engineering units. The results at this level are
passed upward to the larger computers that are responsible for more
complex operations. These upper-level computers can be programmed to
perform more advanced calculations.

Description of the DCS elements

The typical DCS system shown in Figure 3 can consists of one or more of
the following elements:

• Local Control Unit (LCU). This is denoted as local

computer in Figure 3. This unit can handle 8 to 16
individual PID loops, with 16 to 32 analog input lines, 8 to
16 analog output signals and some a limited number of
digital inputs and outputs.
• Data Acquisition Unit. This unit may contain 2 to 16 times
as many analog input/output channels as the LCU. Digital
(discrete) and analog I/O can be handled. Typically, no
control functions are available.
• Batch Sequencing Unit. Typically, this unit contains a
number of external events, timing counters, arbitrary
function generators, and internal logic.
• Local Display. This device usually provides analog display
stations, analog trend recorder, and sometime video
display for readout.
• Bulk Memory Unit. This unit is used to store and recall
process data. Usually mass storage disks or magnetic
tape are used.

8 PCSTP
 • General Purpose Computer. This unit is programmed by a
customer or third party to perform sophisticated functions
such as optimization, advance control, expert system, etc.
• Central Operator Display. This unit typically will contain
one or more consoles for operator communication with
the system, and multiple video color graphics display
units.
• Data Highway. A serial digital data transmission link
connecting all other components in the system may
consist of coaxial cable. Most commercial DCS allow for
redundant data highway to reduce the risk of data loss.
• Local area Network (LAN). Many manufacturers supply a
port device to allow connection to remote devices through
a standard local area network.

The advantages of DCS systems

The major advantages of functional hardware distribution are

flexibility in system design, ease of expansion, reliability, and ease of
maintenance. A big advantage compared to a single-computer system is
that the user can start out at a low level of investment. Another obvious
advantage of this type of distributed architecture is that complete loss of
the data highway will not cause complete loss of system capability. Often
local units can continue operation with no significant loss of function over
moderate or extended periods of time.

Moreover, the DCS network allows different modes of control

implementation such as manual/auto/supervisory/computer operation for
each local control loop. In the manual mode, the operator manipulates the
final control element directly. In the auto mode, the final control element
is manipulated automatically through a low-level controller usually a PID.
The set point for this control loop is entered by the operator. In the
supervisory mode, an advanced digital controller is placed on the top of
the low-level controller (Figure 1). The advanced controller sets the set
point for the low-level controller. The set point for the advanced controller
can be set either by the operator or a steady state optimization. In the
computer mode, the control system operates in the direct digital mode
shown in Figure 1.

One of the main goals of using DCS system is allowing the

implementation of digital control algorithms. The benefit of digital control
application can include:

9 PCSTP
 • Digital systems are more precise.
• Digital systems are more flexible. This means that
control algorithms can be changed and control
configuration can be modified without having rewiring
the system.
• Digital system cost less to install and maintain.
Digital data in electronic files are easier to deal with. Operating
results can be printed out, displayed on color terminals, stored in
highly compressed form.

Important consideration regarding DCS systems.

The control loop
The control loop remains the same as the conventional feedback
control loop, but with the addition of some digital components. Figure 4
shows a typical single direct digital control-loop. Digital computer is used
to take care of all control calculations. Since the computer is a digital
(binary) machine and the information coming out of the process in an
analog for, they had to be digitized before entering the computer.
Similarly the commands issued by the computer are in binary, they should
be converted to analog (continuous) signals before implemented on the
final control element. This is the philosophy behind installing the A/D and
D/A converter on the control loop. Signal conditioning is used to remove
noise and smooth transmitted data. Amplifier can also be used to scale
the transmitted data if the signals gain is small. Signal generators
(transducer) are used to convert the process measurements into analog
signals. The most common analog signals used are 0-5 Volts and 4 -
20mA. Some of the process variables are represented in millivolts such as
those form thermocouples, strain gauges, pH meters, etc. Multiplexers are
often used to switch selectively a number of analog signals.

All instrumentation hardware (1-9) is designed, selected, installed and

maintained by an instrumentation engineer. The computer is responsible
for making decisions (control actions) . It can host a simple control
algorithm or a more advanced one. The latter can either purchased from a
commercial vendor or developed in-house by a process/control engineer
(See section 7.3) . The terminal is the main operator interface with the
control system. The operator can use the terminal to monitor the control
performance, adjust the set points and tune the controller parameters.

10 PCSTP
The basic units of a digital computer
The digital computer used in DCS systems is a regular microcomputer
with the simplified components shown in Figure 5. It includes the
arithmetic unit, which carry out arithmetic and logic commands. The
control unit is the part of the computer responsible for reading program
statements from memory, interpreting them, and causing the appropriate
action to take place. The memory unit is used for storing data and
programs. Typical computers have Random-Access-Memory (RAM) and
Read-Only-Memory (ROM). The final unit is the input/output interface. The
I/O interface is necessary for the computer to communicate with the
external world. This interface is the most important in the control
implementation. The process information is fed to the computer through
the I/O interface and the commands made by the computer are sent to
the final control element through the I/O interface.

computer

Arithmetic
Unit

Control Memory
unit Unit

Input/output
Interface

I/O devices

I/O devices

Figure 5: A general purpose digital computer

In control application, the design of the I/O devices and interface is
an important part of the overall digital control philosophy. The following
subsections discuss some of these issues.

11 PCSTP
Information presentation and accuracy.
The modern digital computer is a binary machine. This means that
internal data and arithmetic and logic must be represented in binary
format. Therefore all process information flowing into and out of the
computer must also be converted to that form. Traditionally, the computer
memory location is made up of a collection of bits called a word (register).
A typical computer word consists of 16 bits (new computers carry 32-bits
word). Consider, for example, the following machine number:

16-bit computer word: 1011001100010100

The base for this word is 2. Therefore, each bit has the following decimal
equivalent:

Bit
16 … 4 Bit 3 Bit 2 Bit 1 Bit
0 … 0 0 0 0 Machine number
Decimal
216 … 23 22 21 20 equivalent
Each single bit consists of binary elements, i.e. 0 or 1. Therefore, any
integer number from 0 to 7 can be represented by a three-bit word as
follows:

Contents of a 3-bit
Digital Equivalent word
0(20)+0(21)+0(22) = 0 000
1(20)+0(21)+0(22) = 1 001
0(20)+1(21)+0(22) = 2 010
1(20)+1(21)+0(22) = 3 011
0(20)+0(21)+1(22) = 4 100
1(20)+0(21)+1(22) = 5 101
0(20)+1(21)+1(22) = 6 110
1(20)+1(21)+1(22) = 7 111

In this case, analog process information should be first changed to

voltage or current as mentioned earlier. Then it is converted to digital
form by an electronic device called analog to digital converter (A/D).
Similarly, digital information is converted to analog form (Voltage or
current) by a digital to analog converter (D/A). The accuracy (resolution)
of such digitization process depends on the number of bits used to for
representation. The degree of resolution is given by:

1
resolution = [full scale range] × /2m−1

12 PCSTP
where m is the number of bits in the representation. Obviously, higher
resolution can be obtained at higher number of bits. For example,
consider a sensor sends an analog signal between 0 and 1 volt and
assume only a three- bit computer word is available, and then the full
range of the signal can be recognized as follows:

This means that eight specific values for the analog signal can be exactly
recognized. Any values interim values will be approximated according to
the covered analog range shown in the fourth column of Table 1. In this
way, the error in resolution is said to be in the order of 1/14. Assume now
a 4-bit word is available for the same analog signal. Then the full range
will be divided over 15 points, i.e. sixteen equally spaced values between
0 and 1 can be recognized, and the error in resolution will be in the order
of 1/30. Most current control-oriented ADC and DAC utilize a 10 to 12 bit
representation (resolution better than 0.1%). Since most micro- and
minicomputers utilize at least a 16-bit word, the value of an analog
variable can be stored in one memory word. New computers are capable
of using 32-bit word. Therefore, new generation of ADC and DAC with
higher resolution (up to 16 to 20 bit) are emerging.

Table 1: Representation of a 0 to 1 volt analog variable using a 3-

bit word
Analog
range Analog Digital Binary
representatio
covered equivalent Equivalent n
0 to 1/14 0 0 000
1/14 to 3/14 1/7 1 001
3/14 to 5/14 2/7 2 010
5/14 to 7/14 3/7 3 011
7/14 to 9/14 4/7 4 100
9/14 to
11/14 5/7 5 101
11/14 to
13/14 6/7 6 110
13/14 to
14/14 1 7 111

13 PCSTP
Process interface
A typical plant with large number of variables contains abundance
of process information (data). Therefore, process information can be
classified under several classes (groups). Then a specialized device can be
used to transfer all information of a specific class into and out of the
computer. This way designing different I/O interface for each I/O device
to be connected to the computer is avoided. In fact, most process data
can be grouped into four major categories as listed in Table 2.

Table 2: Categories of process information

Example Type
1
Relay Digital .
Switch
Solenoid valve
Motor drive
2
Laboratory instrument output Generalized digital .
Alphanumerical displays
3
Turbine flow meter Pulse or pulse train .
Stepping motor
Thermocouple or strain gauge 4
(millivolt) Analog .
Process instrumentation (4 – 20 mA)
Other sensors (0 -5 Volt)

The digital input/output signals can be easily handled because the

match the computer representation format. The digital interface can be
designed to have multiple registers, each with the same number of bits as
the basic computer word. In this way a full word of 16-bit can represent
16 separate process binary variables and can be transmitted to the
computer at one time and stored. Each bit will determine the state of a
specific process input lie. For example, a state of 1 means the input is on
and 0 means off or vice versa.

The generalized digital information usually uses binary coded decimal and
ranges from 0000 to 9999. Hence, a 16-bit register can be used as
interface device to transmit 4 digits of result because four-bits are
necessary to represent one digit (0-9) of binary coded decimal.

14 PCSTP
 In the input pulse information case, a single register (interface
device) is designed for each input line. The register ordinarily consists of
pulse counter. The accumulated pulses over a specified length of time are
transferred to the computer in binary or BCD count. The output pulse
interface consists of a device to generate a continuous train of pulses
followed by a gate. The gate is turned on and off by the computer.

The analog input information must be digitized by ADC before fed to

the computer. Since the process has a large number of analog sensing
devices, a multiplexer is used to switch selectively among various analog
signals. The main purpose of a multiplexer is to avoid the necessity of
using a single ADC for each input line. The DAC devise performs the
reverse operation. Each analog output line from the computer has its own
dedicated DAC. The DAC is designed such that it holds (freeze) a previous
output signal until another command is issued by the computer.

Timing
The control computer must be able to keep track of time (real time)
in order to be able to initiate data acquisition operations and calculate
control outputs or to initiate supervisory optimization on a desired
schedule. Hence, all control computers will contain at least one hardware
timing device. The so-called real-time clock represents one technique.
This device is nothing more than a pulse generator that interrupts the
computer on a periodic basis and identifies itself as interrupting device.

Operator interface.
The operator interface is generally a terminal upon which the
operator can communicate with the system. Such terminals usually permit
displaying graphical information. Often these display consoles are color
terminals for better visibility and recognition of key variables. The
operator will use the keyboard portion of the terminal to perform specific
tasks. For example, the operator can type in requests for information or
displaying trends, changing controller parameters or set points, adding
new control loop, and so on.

15 PCSTP
Digital control software
To make the best use of a DCS system, an advance control strategy
or supervisory optimization can be incorporated in the main host
computer. In the past, computer control projects are written in assembly
language, an extremely tedious procedure. Nowadays most user software
is written in higher-level languages such as BASIC, FORTRAN, C etc. In
many cases, the user is able to utilize the template routines supplied by
the vendor, and is required only to duplicate these routines and
interconnect them to fit his own application purposes. Another way is to
write his own complete control program and implement it.

Other software in the form of control-oriented programming languages is

supplied by the vendor of process control computers. A simpler approach
for the user is to utilize vendor-supplied firmware or software to avoid
writing programs. Currently, most DCS manufacturers develop their own
advance control and optimization software, which can included in the
package as options. Similarly, many control algorithm developers; (DMC,
ASPEN, etc) design a special interface to allow incorporating their own
control programs into most of the commercial DCS network.

Conclusion

Digitally-based control instrumentation represents a revolutionary

change in the process control paradigm. With digital systems the control
engineer has the opportunity to go beyond the narrow limitation of
standard analog control components to construct a system that is
optimum for the information processing and control requirements of large
processes or even of entire plants. This is why many industrial plants are
updating their hardware and instrumentation systems bearing in mind
that the payout times for installation and commissioning costs is as a low
as three to four months.

16 PCSTP