applied process control systems

7/30/2019 Applied Process Control Systems

1/263

Appied Process Control Systems

Fundamentals of Process

Control SystemsLe c t u r e - 1

PEC UET Lahore Dr. Shahid Naveed 9. April 2009 / Dr. Ing Naveed Ramzan 1

. .


2/263

Sequence of Presentation

1.Why Process Control?

.

3.Role of Sensors and Instruments

4.Modes Of Control Systems

6.Examples of Servo Control Systems

7.Conclusions



3/263

1. Why Process Control ?

1.Safety of Equipment and Personnel

.

quantity

.

4.Observe Environmental and Countryaws

5.Economics



4/263

Control Objectives

1. Safety ofEquipment &

Give example

personnel

2. ProductionSpecification of

Vapo

ProductT6 PC

quality &quantity

3. Operational F1

T2

T4 T3 L1

Feed

Constraints

4. EnvironmentalRegulations

F3F2

5. Economics productSteamProcessfluid L. Key



5/263

Safety of Equipment & Personnel


High pressurein drum isdangerous

Va o


Product

T1 T5Feed

quantity

3. OperationalConstraints

F1 T4T3 L1


Liquid

F3F2

A1

. conomcs productSteamrocessfluid L. Key



6/263

Control Objectives


Give example

personnel


Vapo

ProductT6 PC

quality &quantity

3. Operational F1

T2

T4 T3 L1

Feed

Constraints


F3F2




7/263

Safety of Equipment

1. Safety ofEquipment & VaporT6 PC

No flow coulddamage thepump

personne


Product

T1 T2 T5qua yquantity

3. Operational

F1 T4 T3L1

ee

4. EnvironmentalRegulations Liquid

F3F2

A1

5. Economics pro ucfluid L. Key



8/263


9/263

Quality Assurance


Give Example

personnel


Product

quality &quantity

3. Operational F1 T4 T3 L1

Feed

Constraints


F3F2




10/263

Quality Assurance

1. Safety of

Achieve L.Keyby adjusting

personnel2. Production

VaporProductT6 PC

quality &quantity

T1

F1

T2

T4

T5

T3

Feed

.Constraints

4. Environmental F3F2

5. EconomicsLiquidproductSteamProcess

fluid L. Key

A1



11/263

Observe Environmental and Other Regulations

1. Safety ofNever releaseh drocarbons

qupmen

personnel

2. Production

VaporProduct

T6 PCto atmosphere

pec ca on oquality &

quantity

T1

F1

T2

T4 T3 L1

Feed

.Constraints


5. Economicsqu

productSteamProcessfluid L. Key

A1



12/263

Economics

1. Safety ofGive Example

qupmen

personnel

2. Production

VaporProductT6 PC

pec ca on oquality &

quantity

T1

F1

T2

T4

T5Feed

.Constraints


5. EconomicsLiquidproductSteamProcess

fluid L. Key

A1



13/263

Economics

1. Safety ofEquipment &personnel


Give Example

quality &quantity

3. Operationalonstrants


5. Economics



14/263

2. What to Control ?

1.The Influence of External

2.Ensure Stable Operation

(Avoid Unsteady State)

. p m ze pera on

6. Februar 2008 / Dr. Ing Naveed Ramzan #PEC UET Lahore Dr. Shahid Naveed 9. April 2009 / Dr. Ing Naveed Ramzan 14


15/263

Suppress the Influence of External Disturbances

Disturbances?



16/263

Ensure Stable Operation



17/263


18/263

Optimized Operation (online)



19/263

Potential Stochastic Gains through effective Process Control

When we control a process, we reduce the variabil ity.



20/263

Potential Stochastic Gains through effective Process Control (Contd)

Variability is moved from controlled to manipulated variable!



21/263


What statisticscan we cacua e

from this data?

How do we relatevariability to process



22/263


Process erformance

efficiency, yield,production rate, etc. It

for a control objective.

Calculate the process

performance using the

,average value of the key

variable!



23/263


24/263

Comment on Sensors and Instruments

operator that can see & feel the intenseprocess variable inside the vessels

` Accordingly the measurements should be

possible with reasonable costs.

` Process control go hand in hand

If you cannot measure you cannot control!



25/263

4. Modes of Control Systems

` Feedback Control Systems (Regulatory)

` Feedforward Control Systems(Servo, Tracking)

` Sequential Control System

` Distributed

Modes of Implementation

` Pneumatic

` Electronic

` Digital



26/263

Basics of feedback control system



27/263

Examples of feedback system



28/263


29/263




30/263




31/263




32/263

Example of Temperature Control (Heat exchanger)



33/263

Example of Composition Control



34/263

Where is Control Done?



35/263


36/263

Control Systems Implementations In Industry



37/263




38/263




39/263



40/263

PEC Islamabad UET Lahore Dr. Shahid Naveed 9. April 2009 / Dr. Ing Naveed Ramzan 40

Applied Process Control System


41/263


pp y

Sensors &

Instrumentation

Dr. Ing. Naveed Ramzan

Le c t u r e - 2

Sequence of Lecture


42/263

Instruments are our eyes

Performance characteristics of Instruments

Principle measurements desired in industry

(a) Temperature

(b) Pressure(c) Level(d) Flow(e) Others ( Composition, pH etc.)

q



43/263


44/263

Importance of effectivemeasurement in process

industry

http://h/BP%20Explosion.avihttp://h/BP%20Explosion.avihttp://h/BP%20Explosion.avihttp://h/BP%20Explosion.avi


45/263

Failure to measure effectively the level of liquidin bottom of the tower lead to

--- Fire

--- Explosion



46/263

Functional Elements of an Instrument


47/263

Process/Measured medium

Primary

Sensing

Element

Variable

Conversion

Element

Variable

Manipulation

Element

Data

Transmission

Element

Data

Presentation

Element

Observer


Functional Elements of an Instrument (Contd)


48/263

Typical Example:


Functional Elements of an Instrument (Contd)


49/263

Primary

SensingElement

Variable

ConversionElement

Data

TransmissionElement

Variable

Manipulation

Element

Data

Presentation

Element

TemperatureMeasured

Quantity

Pressure

Variable

Conversion

Element

Pressure

MotionMotion

Fluid

Temperature Tube Tubing

Spiral Bourdon

Tube

Linkage GearScale & Pointer

Observer

Process/Measured medium



50/263

Performance Characteristics of Instruments (Contd)


51/263

Static characteristics

Static characteristics of an instrument includes;

Accuracy

Precision

Repeatability

Range

Resolution

Others ( Sensitivity , Dead zone etc.)


Static Characteristics


52/263


1. Accuracy

The ability of a device or asystem to respond to a truevalue of a measured variable

under reference conditions.

In general accuracy

expressed as Limit of Error



53/263

Static Characteristics


54/263


3. Repeatability

Repeatability is the variation in measurements takenby a single person or instruments on the same itemand under the same conditions.


Calibration


55/263

Assigning standardvalues to anequipment iscalibration.


Performance Characteristics


56/263

Dynamic characteristics

Dynamic Characteristics of an instrument includes;

1. Speed of response

2. Fidelity

3. Lag

4. Drift


Principle measurements desired in Industry


57/263

Principle measurements desired in industry

(a) Temperature

(b) Pressure

(c) Level

(d) Flow

(e) Others ( Composition, pH etc.)



58/263

It is time to turn up the heat but

first you must learn how tomeasure it

Temperature Sensor


Methods of Temperature Measurement


59/263

1. Thermocouples

2. Thermistors

3. Electrical resistance change (RTD)4. Expansion of materials

5. Pyrometers



60/263

Thermocouple Types


61/263

TCs are identified by a single letter type andgrouped according to their temperaturerange

` Base Metals up to 1000 C

` Type J, Type E, Type T, Type K

` Noble Metals up to 2000 C

` Type R, Type S, Type B

` Refractory Metals up to 2600 C

` Type C, Type D, Type G


Metal Combinations


62/263

TC

Type

Colours Range C Positive Lead

(Coloured)

Negative Lead

(all Red)J White/Red -210 to 1200 Iron Constantan

E Purple/Red -270 to1000 Chromel Constantan

T Blue/Red 0 to 400 Copper Constantan

K Yellow/Red -270 to1372 Chromel Alumel

R Black/Red -50 to 1768 Platinum-13%rhodium

Platinum

S Black/Red -50 to 1768 Platinum-10%rhodium Platinum

B Grey/Red 0 to 1700 Platinum-30%rhodium

Platinum-6% rhodium

C White-Red/Red 0 to 2320 Tungsten/5%rhenium Tungsten 26% rhenium

Chromel = Nickel-chromiumAlumel = Nickel-aluminum

Constantan = Copper-nickel


Thermocouple Tables


63/263

Type T Thermocouple (Blue & Red) Reference Junction 0 C

C 0 1 2 3 4 5 6 7 8 9

0 0.000 0.039 0.078 0.117 0.156 0.195 0.234 0.273 0.312 0.352

10 0.391 0.431 0.470 0.510 0.549 0.589 0.629 0.669 0.709 0.749

20 0.790 0.830 0.870 0.911 0.951 0.992 1.033 1.074 1.114 1.155

30 1.196 1.238 1.279 1.320 1.362 1.403 1.445 1.486 1.528 1.570

40 1.612 1.654 1.696 1.738 1.780 1.823 1.865 1.908 1.950 1.993



64/263

Thermistors


65/263

` The thermistors can be in the shape of a rod,

bead or disc.


` Manufactured from oxides of nickel, manganese,

iron, cobalt, magnesium, titanium and othermetals.

Thermistors


66/263

Advantages:

` Small sizes and fast response

` Low cost

` Suitability for narrow spans

Disadvantages:

`More susceptible to permanent decalibration athigh temperatures.

` Use is limited to a few hundred degrees Celsius.

` Respond quickly to temperature changes, thus,especially susceptible to self-heating errors.

` Very fragile


Electrical Resistance Change (RTD)


67/263

Resistance Temperature Detector- RTD

` RTD (Resistance Temperature Detector) is a

temperature sensitive resistor.

` It is a positive temperature coefficient device, which

means that the resistance increases with temperature.` The resistive property of the metal is called its

resistivity.

The industry standard is the platinum wire RTD(Pt100) whose base resistance is exactly 100.00

ohms at 0.0 C.




68/263

Platinum Wire RTDs (PRTs)

PRTs have established themselves as the de-facto industrystandard for temperature measurement, and for many reasons:

` linear temperature sensors

` Resistance vs temperaturecharacteristics are stable andreproducible

` linear positive temperature

coefficient (-200 to 800 C)

` very accurate and suitable for use asa secondary standard




69/263

OtherRTDs` 10 ohms Copper RTD - .00427 coefficients

` 100 ohms Platinum RTD - .00385 coefficients

(new IEC)

` 100 ohms Platinum RTD - .00392 coefficients (old)

` 120 ohms Nickel RTD - .00672 coefficient

` 604 ohms Nickel-Iron RTD - .00518 coefficients

All base resistances are specified at a temperature of 0 degrees C

A Pt1000 will have a base resistance of 1000 ohms at 0 deg. C



70/263

RTD Colour Code


71/263

Not standardized but this is commoncolour arrangement. Some (like in thelab) will use BLK-BLK-RED


Expansion Thermometers


72/263

` Bimetallic Thermometer

(Expansion of solids)

Effect of unequal expansion of a bimetallic strip

Bimetallic thermometer

-Different metals have difference coefficient.

-Configured as spiral or helix for compactness

- Can be used with a pointer to make aninexpensive compact rugged thermometer.



73/263

` Filled Thermal Systems

(Filled System Thermometer, Filled Bulb Thermometer)

Similar operation as the liquid in glass

`Bulb

` Capillary tube

` Pressure element

` Scale

Spiral TypeBourdon Tube




74/263

` Filled Thermal System Classes(Filled System Thermometer, Filled Bulb Thermometer)

` Class l A,B Liquid filled

` Class ll A,B,C,D Vapour filled

` Class lll A,B Gas filled

` Class V A,B Mercury Filled

Temperature Range Response

` Class l: -125 F to + 600 F Slowest

` Class ll:

-40 to 32 or 32 to 600 F Fastest

` Class lll: -450 F to +1400 F Fast

` Class V: -40 F to +1200 F Fast


Pyrometer


75/263

Pyrometry is a technique for measuring temperature without

physical contactAn apparatus for measuring high temperatures that uses

the radiation emitted by a hot body as a basis formeasurement.

` Radiation pyrometers ( measurement of radiant energy)

` Optical Pyrometers (comparison of the intensities )



76/263


77/263

Pressure Sensors


In any given plant, the number of

pressure gauges used is probablylarger than all other instruments

put together

Pressure Measurement Methods


78/263

1. Elastic pressure transducers

2. Electric pressure transducers

3. Pressure measurement by measuring vacuum

4. Pressure measurement by balancing forcesproduced on a known area by a measured force

5. Manometer method


Elastic Pressure Transducers


79/263

1. Bourdon tube pressure gauge

2. Diaphragm pressure transducers

3. Bellows


Bourdon Tube Pressure Gauge


80/263

40

Bourdon tubes are generallyare of three types;

1. C-type

2. Helical type3. Spiral type


Diaphragm Pressure Gauge


81/263

They are used to measure gauge pressures over very low

ranges.` Two types of diaphragm pressure guages are:

1. Metallic diaphragms gauge2. Slack diaphragms gauge


Bellows


82/263

` More sensitivethan bourdontype gauge.

` Used to measurelow pressures


Electric Pressure Transducers


83/263

Electrical pressure transducers consists of three

elements

1. Pressure sensing element such as a bellow , adiaphragm or a bourdon tube

2. Primary conversion element e.g. resistance orvoltage

3. Secondary conversion

element



84/263

Strain Gauge Pressure Transducer


85/263

A strain gauge is a passive type resistance pressure transducer

whose electrical resistance changes when it is stretched orcompressed

A pressure transducer contains a diaphragm which is deformedby the pressure which can cause a strain gauge to stretch orcompress. This deformation of the strain gauge causes thevariation in length and cross sectional area due to which its

resistance changes.


Capacitive Pressure Transducer


86/263

The sensing diaphragm and capacitor form a differentialvariable separation capacitor. When the two inputpressures are equal the diaphragm is positioned centrallyand the capacitance are equal. A difference in the two input

pressure causes displacement of the sensing diaphragmand is sensed as a difference between the two capacitances


Differential Pressure Cell (DP Cell)


87/263

It is a device that measures the differential pressure

between two inputs.

Depending on what class theDP-cell is, it will give youfeedback with a current signal.

Normal in Europe is 4-20mA, where 4 is lowest and 20is highest.


Using a d/P Cell Transmitter


88/263

In a closed tank, the Low side of the d/P cell is

connected to the top of the tank and will cancelthe effects of the vapour pressure above thesurface.

Closed Tank Measurement Lo side of the d/P cell measures the vapour pressure above the surface. Hi side measures the hydrostatic head pressure which is proportional to

the height of the liquid and its density + vapour pressure

24 VDC mA

4 20 mA

To PLC orController

H L


89/263

Level Sensors


Level is another common process variable that is

Level Measurement


90/263

Level is another common process variable that ismeasured in many industries. The method used will

vary widely depending on the nature of the industry,the process, and the application.

Inventory:

-- a constant supply or storage of materialControl:-- continuous, batch, blending, and mixing control

-- stabilize flow to the next processAlarming:-- hi/lo limits, safety shut downData Logging:-- material quantities for inventory and billing

purposes and where regulatory requirements arenecessary


Methods ---- Direct or Indirect (inferential)


91/263

` Hydrostatic Head` Float` Load Cells` Magnetic Level

Gauge` Capacitance

Transmitters` Magnetostrictive

` Ultrasonic` Microwave` Laser` Radar

` Guided WaveRadar

` Dip Stick` Vibration


Selection Criteria


92/263

When determining the type of level sensor thatshould be used for a given application, there area series of questions that must be answered:

` Open tank or closed tank?` Can the level sensor be inserted into the tank or

should it be completely external? Contact or non-

contact?` Continuous measurement or point measurement?

` Direct or Indirect measurement?

` What type of material is being measured? Liquidor Solid? Clean or Slurry?


Dip Stick


93/263

` Simple and cheap

` Can be used with any wet

material and not affectedby density.

` Can not be used with

pressurized tanks` Visual indication only

(electronic versions are

available)

RodGauge - similar to a dipstick found in a car, it has weighted

line markings to indicate depth or volumePEC UET Lahore Dr. Shahid Naveed 9. April 2009 / Dr. Ing Naveed Ramzan 53

Th f th fl id i th t k th

Resistance Tape


94/263

The pressure of the fluid in the tank causes the

tape to short-circuit, thus changing the totalresistance of the measuring tape. An electroniccircuit measures the resistance; it's directlyrelated to the liquid level in the tank.



95/263

Magnetic Level Sensor


96/263

Used where the sightglass level gauge cannot be used.

Magneto-resistive typescan provide an

electrical output.

Liquid/liquid interface (such as water and oil) can be measured bychanging the buoyancy of the magnetic float


Floats


97/263

Float rides the surface level to provide the

measurement. Many different styles areavailable.

Liquid density does not affect measurementPEC UET Lahore Dr. Shahid Naveed 9. April 2009 / Dr. Ing Naveed Ramzan 57

Hydrostatic Head Level Sensors


98/263

` These methods infer level by measuring

the hydrostatic head produced by theliquid column.

`A pressure sensing element is installedat the bottom of the tank and pressure

is converted to level.

`Different liquid densities or closed tankapplications must be accounted for.


Practical Considerations when usingHydrostatic Head Level Sensors (Contd)


99/263

g

head type instrumentsThe reference point of the tank vs

instrument input must be considered.

Height

(H)

Pressure PSI

Liquid Density

(D)Height

(H)Pressure

PSI

Liquid Density

(D)

This may not be practical in some applications where the tank elevationis below grade or where a remote visual reading is required.


Bubblers


100/263

Bubblers allow theindicator to belocated

anywhere.

The air pressure in

the tube varieswith the headpressure of the

height of theliquid.

Bottom of tubedetermines reference

point

P

Regulated purgesystem

(air or nitrogen)

Instrument inputdoes not matter

Cant be used in closed tanks or where purging a liquid is not allowed (soap). Very

popular in the paper industry because the air purge keeps the tube from plugging.



101/263

Conductivity Level Measurement


102/263

Po i n t Le v e l M e a su r em e n t Co n t i n u o u s Le v e l M e a su r em e n t

Advantages and disadvantages

Low Cost

Conductive, non-coating liquids onlyInsulating coatings can cause problems


UltraSonic Level Measurement


103/263

` Non-Contact direct levelsensor

` Level is a function of thetime it takes an ultrasonicpulse to hit the surface andreturn

Limitations include:

Surface foam absorbs signal, agitation create reflections

High Pressure & High Temperatures affect the signal speed

Vapour and condensate create false echos



104/263

Load Cells


105/263

Tank level is determined by theweight of the quantity of

materialLoad Cells (strain gauge

transducers) placed at thebottom of the tank measurethe weight and then convert itto an electrical signal.



106/263

Flow Sensors


` Pl t t l f d t lit d f t

Reasons for Flow Metering


107/263

` Plant control, for product quality and safety

reasons.` Custody transfer, both interplant and selling to

outside customers.

` Filling of containers, stock tanks andtransporters.

` Energy, mass balancing for costing purpose

and health monitoring of heat exchangers.` Health monitoring of pipelines and on-line

analysis equipment, Government and

company legislation may dictate the use hereof such equipment.


` Differential Pressure Meters

Types of Flow Meters


108/263

` Differential Pressure Meters.

` Rotary Meters.1. Displacement2. Inferential

` New Flow Meters.1. Electromagnetic2. Vortex Shedding3. Ultrasonic4. Cross Correlation

5. Tracer6. Swirl7. Fluidic

` Point Velocity Meters.` Mass Flow Maters.


` Orifice Plate

Differential Pressure Meters


109/263

` Orifice Plate

` Dall Tube` Venturi Tube

` Pitot Tube

` Rota meter

` Target mater

` Averaging Pitot

` Nozzle

` Spring Loaded

` Intake Meter` Elbow Meter

` Bypass Meter


Differential Pressure Meters ( Contd)


110/263

Elbow Flow meter Rota meter


Di l t M t

Rotary Meters ( Displacement Meters)


111/263

Displacement Meters

` Gear

` Oval wheel

` Vane Meter

` Gear (Roots)

` Diaphragm Meter

`Liquid Sealed Meter

Inferential Meters

` Turbine Meter

` Hoverflo Meter


Turbine Flow Meters


112/263


` Electromagnetic

New Flow Meters


113/263

` Electromagnetic

EM Meter

` Vortex Shedding Meter

Vortex Generation Meter

` Ultrasonic Flow Meters


New Flow Meters (Contd)


114/263

Magnetic Flowmeter


New Flow Meters (Contd)


115/263

Swirl Meter


Coriolis Mass Flowmeter


116/263



117/263


118/263

Discussion & Questions?

Dr. Shahid Naveed 9. April 2009 / Dr. Ing Naveed Ramzan 78



119/263


Design of Control Systems& Standard Practices

Prof. Dr. Shahid Naveed

Le c t u r e - 3



120/263

1.Design Objectives and Variables2.Steps in Design of a Control System

3.Standard Practices

4.Controller Selection & Tuning

5.Conclusion


Variables in a Chemical Process


121/263


Hardware for a process Control System


122/263


Steps in Design of a Control System


123/263

1.Define the control objective(Purpose)

2.Select the measurement variable/s

(Primary / Secondary measurements, Easily, Rapidly, Reliably)

3. Select the manipulated variable/s (Utility/process)

4. Choose the control configuration (ConfigurationStandard/New)

5. Design/choose the controller

(Degree of controllability desired)

6.Design of multiple input multiple outputsystems (MIMO)


General Control Configuration

F d b k t l


124/263

Fe e d - f o r w a r d co n t r o l

Fe e d b a ck co n t r o l



125/263

Recommended ControlSchemes and Standard

Practices


Level Control


126/263



127/263

Flow Control


128/263


Pump Control


129/263


Steam Turbine Driven Pump Control


130/263


Rotary Pump Control


131/263


Temperature Control of HE


132/263


Temperature Control of HE (contd)


133/263



134/263

Temperature Control of HE (contd)


135/263


Pressure Control of Pressure Vessel


136/263


Pressure Control of Pressurized Vessels


137/263



138/263

Cascade Control


139/263


Split Range Control


140/263


Ratio Control


141/263



142/263

Pressure Control (Contd)


143/263



144/263

Controller Selection &Tuning Methods


Available Methods

Controller Selection & Controller Tuning


145/263

Available Methods

Ziegler-Nichols Tuning Technique

Cohen-Coon Setting

Quartering Technique

Minimum Offset

Minimum Integral Square Error ISE Minimum Integral Time Average Error ITAE

Minimum Integral Average Error IAE


Ziegler-Nichols tuning technique

It is for closed response system.


146/263

Kc Ti (min) Td (min)

Proportional - -

Proportional -integral

Proportional

integral derivative

2Ku

2.2Ku

7.1Ku

2.1Pu

2Pu

8Pu

M1

oWc

2cycle

min

Where;Ultimate gain = Ku =

Ultimate period =Pu =

(M is the amplitude ratio for sustained oscillations )


Cohen Coons Method


147/263

Input

Outpu

A

BK == (at steady state)

S

BT = B=ultimate response, td=time elapsed until system started to response

S=slope of curve


Other Methods of Tuning Controllers


148/263

QuarteringMinimum Offset

Minimum Integral-Square Error

( )=

dttISE2

( )=

dtttITAE

( )dttIAE =


The general guidelines for tuning are as follows:

Selecting the Best controller and the values of P, I, D


149/263

` If you want to strongly suppress the Large Errors , ISEis better than IEA because the errors are squared andthat contribute more to the value of the integral.

` If you want to suppress the Small Errors, IAE is betterthan ISE because when square small numbers (smallerthan one) they become even smaller.

` To suppress errors that persist for Long Times, the ITAEcriteria will tune the controllers better because thepresence of large t amplifies the effect of even smallerrors in the value integral.


Selecting the Best controller and the values of P, I, D


150/263


General Criteria to Select Feed Back Controller

` Define approximate performance criteria


151/263

` Define approximate performance criteriaISE, IAE or ITAE.

`

Compute the value of the performancecriteria using P or PI or PID controller withthe best adjusted Kc, Ti, and Td.

` Select the Controller that gives the BESTvalue for the performance criteria.


Rules of Thumb

` If possible use simple proportional (P)


152/263

controllers.

` If simple Proportional action is

unacceptable use a Proportional Integral(PI) controller.

`Use a Proportional Integral Derivative(PID) control to increase the speed ofresponse and retain robustness.


Critical Features of P, PI & PID Controllers.


153/263


Summary & Conclusion


154/263

Design of Control System is a science andan art.

Recommended control practices play an

important role in it.

The role of engineer stretches far beyond

its design in application throughunderstanding



155/263


Applied Process Control System


156/263


Computer ControlSystems

Dr. Ing. Naveed Ramzan

Le c t u r e - 4



157/263

Dr. Shahid Naveed 9. April 2009/ Dr. Ing Naveed Ramzan 2

` Basics of Computerized Control

` Essential components of computer control system

` Advantages of Computer Control

`Types of computer control implementation

Evolution of Process Control


158/263


Feedback Control Loop

Single Loop Control (SLC)


159/263


Traditional Single Loop Controller

Single Loop Control (SLC) (Contd)


160/263


From Traditional Analog Controller to Computer Control

Single Loop Control (SLC) (Contd)


161/263


Multiple loop Control


162/263


Actuators Process

Essential Components of Computer Control


163/263

Hold Amplifier

DA Converter

Multiplexer

Digital

Computer

OperatorLog

Sensors

Instrumentation

Multiplexer

Amplifier/

Attenuator

A D Converter


Essential Components of Computer Control (Contd)


164/263



Types of Signal


165/263


MemorySlowFast



166/263

Input-OutputInterface

Digital Analog

I/0 I/0

Mass Storage

DisksFixed HeadMoving HeadFloppy

Magnetic Tapes

CommunicationPeripherals

PrintersType writersVideo Display unitGraphic TerminalsCard Reader

X-Y Platters

CentralProcessing Unit (CPU)

Basic FeaturesControllersArithmetic UnitLogical UnitRegisters

OptionsFloating Point ProcessorReal time ClockPower fail-safe auto restartWatchdog timer

ROM


Components of digitalcomputer

Plant Inputs : Multiplexing



167/263


` Control System

Process control computer vs data processing computer


168/263


Figure 1b

` Monitoring System

Process control computer vs data processing computer


169/263


Figure 1a

Advantages of Computer Control


170/263


-- all plant data is available at a central point and/or distributedpoints around the plant or works

-- alarm conditions are detected and reported quickly

-- plant is run according to the plan set by management, whichmay not be the case where there is a lot of manual

intervention

-- displays and printouts of information are available readily

-- changes to the operating conditions or recipes are made insmooth, repeatable and closely controlled manner

-- written records of all alarms and operator actions are

Advantages of Computer Control


171/263


available for post-incident or routine analysis

-- alarm conditions are detected and reported quickly

-- control room and control panel size is reduced

-- all batches of material are treated identically

-- batch logs are recorded automatically

-- complex start-up and shut-down procedure can be automated

-- safety interlocks can be programmed into the system

Direct Digital Control loop

Types of Computer Control Implementation


172/263


Figure 4


173/263

Direct Digital Control loop



174/263


Digitally Directed Analog Control (DDAC)

- Supervisory computer control



175/263


Figure 4

Digitally Directed Analog Control (DDAC)

- Supervisory computer control



176/263


Figure 4

Simplified Diagram of supervisory control



177/263



178/263

Supervisory computer control



179/263


Distributed Control System



180/263


Distributed Control System



181/263


Programmable Logic Controller



182/263




183/263





184/263





185/263





186/263





187/263





188/263


Comparing DCS with PLC for ease of configuration


189/263



190/263

Emerging (Ideal) Configuration


191/263


Integrated Computer Control System



192/263


Integrated Computer Control System for a paper and pulp mill



193/263


Traditional SCADA



194/263


Field Instrumentation

Remote Stations

Communications Network

Central Monitoring Station

Water distribution system using SCADA



195/263


Discussion


196/263

Discussion

Dr. Shahid Naveed 9. April 2009 / Dr. Ing Naveed Ramzan 41



197/263


Fundamentals of

Distributed ControlSystems

Prof. Dr. Shahid Naveed

L ec t u r e - 5

Mainframe To Distributed Control


198/263



199/263


DCS is a Computer Network with Control Objectives

1. The usual benefits of computer network on

Ethernet (or like) are availed e.g. resourcesharing (hard/soft) backup/archive database


200/263

Ethernet (or like) are availed e.g. resourcesharing (hard/soft), backup/archive, databaseupdating, control algorithm computation, datalogging etc

2. Each Local Control Unit (LCU) is used tocontrol the process variables. The informationin shared through communication interfaces

for multiple usage and Low/High LevelInterfaces.

3. The High Level Human Interface (HLHI) is

provided in the control room and formanagerial applications throughout the plant


Typical Benefits of DCS

Scalability & Expandability of System


201/263


Scalability & Expandability of System Improved Control Potential

Improved Operator Interface Capability

Integration of System Function

Single Point Failure

Lower Installation costs

Easy to Maintain

Local Control Unit Architecture

1. It is the smallest collection of hardware ina distributed control system that performs

closed loop control.


202/263

closed loop control.

2. LCU malfunctioning can cause a conditionthat is hazardous to both people and

equipment its proper design is critical


Typical Requirements of LCU

` Flexibility of changing control configuration

` Ability to use the controller without being expert


203/263

Ability to use the controller without being expert` Ability to bypass the controller in case it fails so

that the process can be controlled manually

` Ability of LCU to communicate with other LCUsand other elements of the system

Architectural Parameters of LCUSize of ControllerFunctionality of Controller

Performance of ControllerCommunication channels out of ControllerController Output Security


2

Analog inputs

4

Digital Inputs

A LCU


204/263

Capacity:

10 : Continuous

Function Blocks

40: Logic

Function Blocks

LCU A

Analog outputs Digital outputs


A LCU (Contd)


205/263

.. ..

....

LCU B1

LCU B2

16

Analog inputs

32

Digital Inputs

Capacity:

40 Continuous

Functions Blocks

Capacity:

160 Continuous

Functions Blocks

8

Analog outputs

16

Digital outputs



206/263

Architectural parameters of LCU

` Size of Controller

` Functionality of Controller


207/263

y

` Performance of Controller

` Communication channels out of Controller` Controller Output Security

A large number of configuration are possible

` Single loop

`Two LCU types

` Multiple Loops


208/263

Comment

` Typical cost effectiveness of architectures is

estimated with each of the configuration


209/263

g

` Configuration-A 12 controllers are

required

` Configuration-B 3 controllers + 1 backup

required

` Configuration-C 2 controllers + 1 backup

required

Typical cost effectiveness


210/263


Typical cost effectiveness


211/263


Nos of function blocks in control sys

Control Complexity Ratio


212/263

Nos. of function blocks in control sys

` CCR = --------------------------------------------

Nos. of control system output

Nos. of continuous function blocks

` CCR = ---------------------------------------------

Nos. of control analog output

Fo r com p l e x s y s t e m t h e CCR i s 3 0 - 4 0


Other Architectural Issues

How long each cable can be extended?


213/263

How many communication channels the

system can support?

What kind of delay is expected?

What is the internal communication system?

Distributed Communication


214/263

PEC I UET Lahore Dr. Shahid Naveed 9. April 2009 / Dr. Ing Naveed Ramzan 18

Level of Sub networks

Several high level operation interfaces and

communication elements located in the centralcontrol room area must communicate with


215/263

control room area must communicate witheach other at moderate load of messagetraffic.

The message communication has generally three

levels.1. A load bus or subnet

2. A local network in the central control room

3. A plant wide communication

Field-bus Reduces Wirings


216/263


Benefits of Distributed CommunicationA comparison between non-distributed and distributed

Communication facility reveals the following benefits:

Cost of wiring is reduced significantly


217/263


Flexibility of making changes in software &firmware increases

Less time to implement a large control system The control system becomes more reliable due

to significant reduction of physical connection

Typical worries for Distributed CommunicationExisting one-is-to one configuration had notime delay. Sharing the data buses may havesome delay! The answer is NOOverloading of channels!The answer is Overloading may take place.

Requirements of Communication System

1. Minimize time delays and maximized security oftransmission.

2. Communication between various modules and HLHI/


218/263

LLHI should be effective.

3. Communicate set points, operating modes and control

variable from HLHI devices.Additional Requirements

Download control configurations, tuning parametersand user programs for HLHI.

Transmission of information from data input/ outputunits to high level computing devices.

Synchronization of real- time and elements of

machines. Transfer large data for high level processing take trends

and searching applications.

Communication Issues

1. Maximum permitted size of the system

2. Maximum acceptable time3 Maximum allowed delay in systems


219/263

3. Maximum allowed delay in systems

4. Communication rate

5. Rate of undeleted errors occurring in the system ( due tonoise)

6. Sensitivity to traffic loading (traffic of signal load)

7. System scalability (expandable/ large/ small)8. Fault Tolerance (failure of one channel should not effect

other)

9. Interacting requirement (protocol issues RS 232c, RS422, IEEE 488, smart)

10. Ease of application/ maintenance

Smart Hubs In Networks


220/263


Network Topologies

Topology refers to the structure ofphysical connections among the


221/263


physical connections among theelements in a network.

Some of the popular topologies are:Star

BusMeshRing

Network Topologies (Contd)


222/263


Network Topologies (Contd)

Staris simplest but its failure shall cause theentire sub network to stop functioning

l l b h


223/263


Bustopology is similar to star but the sameproblem exist i. e. sharing one bus

Meshtopology tries to overcome the limitationof star topology and is generally adopted in

industry

Ring or looptopology is a special case ofMash topology that provides connectionsbetween active switching devices in a loopsequence

The Way Ahead-- Management Executive Systems


224/263



225/263

6. Februar 2008 / Dr. Ing Naveed Ramzan #



226/263


The Opportunity inStochastic Process

ControlProf. Dr. Shahid Naveed

L ec t u r e - 6


1.Objective of SPC

2.Basic Concepts of SPC


227/263

3.Techniques & Analysis

4.Common & Special Causes of ProcessDeviations

5.Stable Process

6.Trend Analysis

7.Process Capability


What statisticscan e calculate

Recalling From Lecture -1 Stochastic gains


228/263

can we calculatefrom this data?

How do we relatevariability to processperformance?



229/263

In Nutshell Squeeze the Variability is the 1st Objective


230/263


Another Example


231/263


2. Basic Concepts for SPC

`What is a process?` Converting raw materials into products


232/263

`What is control?` Achieving the objectives in the desired manner

` Have we done the job correctly?` Only Historic Data can reveal

`Can we do the job more consistently?` Through online statistical monitoring/trend analysis


SPC Related Queries

Are we doing the job correctly?Detection, prevention, confirmatory to design aspects


233/263

Detection, prevention, confirmatory to design aspects

Can we do the job better?The process and human capability is under question

Can capability/ quality be enhanced?

Through process improvement and training


3. Techniques & Analysis

` What statisticaltools are


234/263

available?

` Process flowCharts

`

Check or tallycharts

` Histograms


3. Techniques & Analysis (contd)

` Graphs` Pareto Analysis


235/263

` Pareto Analysis

` Cause & effect

diagrams

` Scatterdiagram

` Control Charts


What is a Control Chart?


236/263


What is a Control Chart?


237/263


4. Objective of Control Charts

Process monitoring

for extendedperiods


238/263

p

To identify common

causes of processvariation / deviation

To identify special

causes of processvariation/ deviation


Common Causes of Process Variations

` Inherent or natural variation in the inputs andtransformation activities.


239/263

How to identify them?

` Repeatable, predictable, non- symmetric orrandom.

` Global to the system and originate from manysources.

` Lie within the process band within which a

process is expected to vary.


Special Causes of Process Variations

` These are variations due to process inputs andtransformation activities that occur irregularly.They are not part of the system but originatefrom outside


240/263

from outside.

How to identify them?

`They are localized to an operator, shift orpiece of equipment.


Strategy of Process Improvement

` The general policy to improve process

monitoring is to minimize commoncauses and eliminate special causes ofdi t b


241/263

disturbances.

Com m o n Ca u s e s can be minimized by:

` Reduction in variability in inputs andtransformation activities

` Improve the process

` Improve work practices

` Improve operating guidelines


Strategy of Process Improvement

Sp e c i a l Ca u s e s need to be attended specifically

` Whenever a control chart signals a special


242/263

cause, search immediately for what was done

differently on that instance.` The discovery of special cause of variation is

usually the responsibility of someone who isdirectly connected with the process.

` Do not make a fundamental change inprocess, identify and permanently remove thespecial cause to prevent the recurrence.

` If a special cause can not be pin-pointed makeprocess adjustment.


5. What is a Stochastic Stable Process ?A Process is said to be in a state of Stochastic Stable Controlif the distribution of measurement data has the same shapelocation and spread over time.In other words, a process is stable when the effects ofall specialcauses have been removedfrom a process, so that the remaining


243/263

p , gvariability is only due to common causes.



244/263

Trend Analysis

Trend Analysis leads to answer the queries like

1. Are we in control ?


245/263

2. Do we continue to be in control ?

and is instrumental to

1. Fault diagnostics

2. Capability enhancement


Trend Analysis (contd)

After a process is recognized to be out of control

trend analysis is employed to search for thesources of problems.


246/263


The chart is divided into three zones. Zone A isbetween +/- 3. zone B is between +/- 2 andzone C is between +/- 1

The following eight tests can be performed.


Test 1Pattern: One or more points falling outside thecontrol limits on either side of the average.


247/263


Qu

ality

Ch

ara

cte

rist

ics

Problem source:

Equipment breakdownNew OperatorDrastic change in raw materialquality

Change in method, machine,or process settingAction:Go back and look at what

might have been donedifferently before the out ofcontrol point signals.


Test 2Pattern: A run of nine points on one side of the average.

Problem source:This may be due to a small


248/263


This may be due to a smallchange in the level of process

average. This change may bepermanent at the new level.

Action:

Go back to the beginning of therun and determine what wasdone differently at that time orprior to that time

Quality

Characteristics


Test 3Pattern: A trend of six points in a row either increasing or

decreasingProblem Source:


249/263


Qual

ity

Characteristics

Gradual tool wearChange in characteristics suchas gradual deterioration inmixing or concentration of achemical.Deterioration of plating oretching solution in electronicsor chemical industriesAction:Go back to the beginning of therun and search for the sourcein procedure


Test 4:Pattern: Fourteen points in a row alternating up and down within

or outside the control limits.Problem source:


250/263


Quality

Characteristics

Sampling variation fromtwo different sources suchas sampling systematicallyfrom high and low conditionwith two different averages.orAdjustment is being madeall the time (over control).Action: Look for cycles inthe process, such ashumidity or temperaturecycles or operator overcontrol of process


251/263


Test 6Pattern: Four out of five points in a row on one side of the

average in zone B or beyond

Problem source:


252/263


Quality

Characteristics

Problem source:This may be due to a

moderate shift in theprocess

Action: Go back three or

four points in time to findthe root cause.


Test 7Pattern: Fifteen points in a row on either side of the

average in zone CProblem Source:Unnatural small fluctuations or


253/263


Quality

Characteristics

Unnatural small fluctuations orabsence of points near the control

limitsIt may appear to be good controlbut this is not.Incorrect selection of subgroups.

Sampling from various sub-population and combining theminto a single subgroup for chartingIn correct calculation of control

limitsAction: Look very close to thebeginning of the pattern


Test 8Pattern: Eight points in a row on both sides of the center line

with none in zone C.


254/263

Qual

ity

Characteristics

Problem source: No sufficientresolution on the measurementsystem

Action: Look at the range chartand see if it is in control

6. Process Capability

Understanding Capability!


255/263

` Whether the process is capable of meeting the

requirements ?

` Whether the process is meeting the

requirement at any point in time ?

` Can the parameter be corrected or adjusted

when it is not meeting the requirements ?


Process CapabilityThe capability of a stable process is defined in

terms of a distribution. It is the spread of all

values of the process distribution.Capable Process ( Cp)


256/263


Cp measures the effect of the inherent variability

Cp is defined mathematically asUSL - LSL

Cp = ---------------6

Allowable process spreadCp = -----------------------------

Actual process spread

Where:USL = upper specification limitLSS: lower specification limit


Process Capable but not Centered


257/263

The Control implementation is erring towards on lower side.(Or underutilized)



258/263

Centred & Capable Process


259/263


Capability index ( Cpk)The long term objectives of Cpk is to continuously improve

process and reduce variability

Mathematically defined as:


260/263


Cpk = Minimum {USL- X / 3, X-LSL/3}

Where X= overall process averageCpk is applicable for process centering.

If for a two sided specification the capability index (Cpk) isequal to or greater than 1.33, then the process may beadequately centered.

Cpk may be employed even when there is only one sidedspecification. This is used to determine the percentage ofobservations out of specification.

` A process is capable (Cp1) if its naturaltolerance lies within the engineering tolerance

Capability index ( Cpk) (contd)


261/263

or specifications. The measure of process

capability of a stable process is 6 where isthe inherent process variability estimated fromthe process.

` A minimum value of Cp = 1.33 is generallyused for an on going process. This ensures avery low reject rate of 0.007% and thereforeis an effective strategy for prevention ofnonconforming items.


Conclusion

Stochastic Process Control has greatOpportunities leading to Production,


262/263

Opportunities leading to Production,Quality Assurance, Fault Diagnostics,

Capability Enhancement etc.


263/263

applied process control systems

Documents