Fundamentals of

Paper Machine Control

Page 1 Solutions for Superior Results

Purpose

The purpose of this training is to better understand

how a Honeywell Measurex system contributes to the
optimization of the papermaking process and to assist
with troubleshooting the mill-Honeywell Measurex
interface.

Page 2 Solutions for Superior Results

What Will We Be Doing?

We will discuss:
• The theory of feedback control systems
• Typical process behavior
• How Honeywell Measurex models process behavior
• How process behavior characteristics are determined
• How supervisory control loops work
• Cross Direction control
• Tips for maintaining optimum performance

Page 3 Solutions for Superior Results

How Will I Benefit?

• You will have a better understanding of the role of Honeywell -

Measurex measurement and control in papermaking
• You will be better able to differentiate between Honeywell -
Measurex system and mill field device problems.
• You will learn basic operation of the Honeywell - Measurex system
Operator Interface.
• Honeywell - Measurex service personnel and mill personnel will be
able to communicate better

Page 4 Solutions for Superior Results

 The Fourdrinier Paper Machine

PRESS CALENDER
HEADBOX REEL
SECTION STACK

STOCK FOURDRINIER DRYER Honeywell Measurex

PREPARATION TABLE SECTION SCANNER

Page 5 Solutions for Superior Results

Stock Preparation

Ad
dit Pulp er
iv es at
W
The basic objective in
stock preparation is to mix
fibrous raw materials
(pulp), non-fibrous
components (additives),
and water into a uniform
papermaking stock.

To Headbox

Page 6 Solutions for Superior Results

The Headbox

SLICE PROFILE
ADJUSTER
The headbox takes the
incoming pipeline stock flow
and distributes it evenly
across the forming fabric.

RECTIFIER
ROLLS FORMING
FABRIC

FROM STOCK
PREPARATION

Page 7 Solutions for Superior Results

The Fourdrinier Table

By means of an endless,
fourdrinier fabric the
papermaking stock is
formed into a continuous
sheet while the fourdrinier
table drains the water by
suction forces.

Page 8 Solutions for Superior Results

 The Fourdrinier Table

The forming fabric (wire) is an

SUCTION endless, finely woven belt
COUCH
ROLL
(usually a plastic mesh) that
travels between two large
rolls, the breast roll and the
couch roll. The breast roll is
solid and serves only to
BREAST support the fabric. The couch
ROLL roll is a hollow, perforated
FORMING
FABRIC shell containing one or two
stationary high-vacuum
suction boxes for dewatering
the sheet.

Page 9 Solutions for Superior Results

The Fourdrinier Table

The first static element

under the wire is the
forming board. It supports
the wire and is used to
retard initial drainage, so
that fines and fillers are not
washed through the sheet.
FORMING HYDROFOIL The next element, the
BOARD ASSEMBLIES
hydrofoil assemblies,
provide sheet drainage by
inducing suction.

Page 10 Solutions for Superior Results

The Fourdrinier Table

“Wet boxes” use vacuum-

assisted suction to drain the
sheet.
“Flat boxes” or “dry boxes”
also use vacuum-assisted
suction, however, a much
WET DRY
higher pressure is used.
BOXES BOXES

Page 11 Solutions for Superior Results

The Fourdrinier Table

LUMP
DANDY
BREAKER The dandy roll reworks the
ROLL
ROLL
top section of the sheet to
improve formation and
surface characteristics.
The lump breaker roll
improves water-removal at
the couch and consolidates
the sheet.

Page 12 Solutions for Superior Results

The Press Section

PRESS The rather fragile sheet is

ROLLS
transferred from the forming
section and conveyed on
specially-constructed felts
through a series of roll press
nips into the dryer section.
The objective of the press
PRESS section is to remove water
FELT
and force the fibers into
intimate contact.

Page 13 Solutions for Superior Results

The Dryer Section

The wet web from the press

TOP
FELT
section, containing 55 - 60%
moisture, is passed over a
series of rotating steam-
heated cylinders (dryer
cans) where water is
evaporated and carried
away by ventilation air. The
web is held tightly against
DRYER
the cylinders by a synthetic,
CANS BOTTOM permeable fabric called a
FELT
dryer felt.

Page 14 Solutions for Superior Results

The Calender Stack

CALENDER
STACK Calendering is a general
term meaning pressing
with a roll. The sheet is
calendered through a
series of roll nips to
reduce thickness and
smooth the surface.

Page 15 Solutions for Superior Results

The Reel

After drying and

calendering, the paper
REEL REEL FINISHED product must be
DRUM SPOOL REEL
collected in a convenient
form for subsequent
processing. Typically,
this is done by winding
the sheet onto a reel.

Page 16 Solutions for Superior Results

 Measurement and Control

All industrial processes have inherent variability which must

be minimized if the plant operation is to yield a uniform,
high-quality product that makes efficient use of raw
materials, energy, and staff. Minimizing the variation
requires automation of various aspects of the operation. A
typical automatic control loop consists of three basic
components: sensor, control element, and controller.

Page 17 Solutions for Superior Results

Functional Structure of Feedback Control

Sensor - The measuring

SENSOR
component
Controller - Compares the
measurement to the set
point to determine the
CONTROLLER
CONTROL
ELEMENT
amount of “error”
Control Element - Actuates
toward reducing the error

Page 18 Solutions for Superior Results

 Basic Control Concepts

The variables involved:

FT Controlled Variable - a measurement
of the variable you wish to control
STOCK
FLOW Manipulated Variable - some method
to manipulate or change the
MANIPULATED CONTROLLED controlled variable, often done
VARIABLE VARIABLE through the use of a control valve

Page 19 Solutions for Superior Results

Basic Control Concepts

The variables involved:

Setpoint - the desired value of the controlled variable
Disturbances - any variable, other than the manipulated variable, that
drives the controlled variable away from the desired setpoint

DISTURBANCES

MANIPULATED CONTROLLED
PROCESS
VARIABLE VARIABLE

Page 20 Solutions for Superior Results

 Manual Control

Before studying automatic process

control, it’s helpful to spend a moment
reviewing typical manual operation. In
manual, the operator visually inspects
the current value of the controlled
FT
variable (in this example stock flow).
Then, if desired, the operator can
STOCK change the flow by manipulating the
FLOW valve. The operator acts as the
controller.

Page 21 Solutions for Superior Results

 Automatic Control

Automatic control of a process is

CONTROLLER done by sending the sensor output
to the controller. The controller
compares the actual value to the
desired value. Then, the controller
FT calculates the change needed to the
manipulated variable in order to
STOCK bring the controlled variable to the
FLOW
desired value.

Page 22 Solutions for Superior Results

Automatic Control

SETPOINT The feedback control system

monitors the process and
automatically makes
CONTROLLER adjustments as required.

The operator only needs to

provide the setpoint to the
FT
controller.
STOCK
FLOW

Page 23 Solutions for Superior Results

Measurement

To better understand how

feedback control systems work
SENSOR we need to take a closer look
at its three components. We
will begin with sensors.
To minimize product variation
CONTROL
requires automation of various
CONTROLLER ELEMENT aspects of the operation - a
process that begins by getting
accurate and reliable
measurements. These
measurements are the heart of
any control system.

Page 24 Solutions for Superior Results

Measurements

Our review of measurement will focus on the typical

process sensors found in modern paper mills,
including:
• Stock Flow • Steam Pressure
• Head Box Pressure • Machine Speed
• Head Box Level • Temperature
• Consistency

Page 25 Solutions for Superior Results

Measurements

We will also review the Honeywell - Measurex sensors

found on the online scanners.

• Basis Weight
• Caliper
• Moisture - Reel

Page 26 Solutions for Superior Results

Elements and Transmitters

There are two components used

TRANSMITTER in measurement, the element
which is the sensing device and
the transmitter which converts
FT
the sensor’s output into a usable
signal by the measurement and
FE
control system. The element and
transmitter are shown for stock
flow measurement.
ELEMENT

Page 27 Solutions for Superior Results

 Stock Flow Measurement

STOCK
To ensure consistent
PREPARATION
HEADBOX
product quality the stock
FT entering the headbox is
measured as it enters the
headbox.

Page 28 Solutions for Superior Results

Magnetic Flow Meters

Although flow transmitters

UPPER
MAGNETIC come in many different
COIL
types, stock flow is typically
measured using a magnetic
flow meter.
The measurement principle
is based on Faraday’s Law
of Magnetic Induction.
LOWER
MAGNETIC
COIL

Page 29 Solutions for Superior Results

 Magnetic Flow Meters

Faraday’s Law states that a

charged particle passing
through a magnetic field
produces an electromotive
e
force (e) that is
B
perpendicular to both the
magnetic field (B) and the
V e velocity vector (V).

Page 30 Solutions for Superior Results

Magnetic Flow Meters

Stock flowing through the

magnetic field produces a
voltage proportional to the
flow rate (as the flow
increases the voltage
increases).
This voltage is measured by
an electrode pair, amplified,
and sent to a computer for
ELECTRODE processing.
PAIR

Page 31 Solutions for Superior Results

Magnetic Flow Meters

Advantages Disadvantages
• Obstructionless flow path • High cost
• No pressure loss • Fluid must be a conductor of
• Can be used in corrosive electricity
service • Calibration can shift due to
electrode coating

Page 32 Solutions for Superior Results

 Headbox Pressure Measurement

SLICE PROFILE
ADJUSTER The function of the headbox is to
PT deliver stock to the wire at a
uniform velocity. The discharge
velocity of the stock depends
directly on the headbox pressure
(total head) therefore,
measurement and control of total
head is critical.

RECTIFIER
ROLLS FORMING
FABRIC

FROM STOCK
PREPARATION

Page 33 Solutions for Superior Results

 The Strain Gage

One of the most common ways to

P2
measure pressure in process control
DIAPHRAGM
applications is through the use of a
STRAIN GAGES
strain gage sensor. The sensor uses
a material whose electrical
resistance changes as a function of
how much it is bent. This resistive
material is bonded to a metal
diaphragm.

P1

Page 34 Solutions for Superior Results

 The Strain Gage

When pressure is applied to the diaphragm, two of the four strain

gages will be in compression (their resistance decreases) and two will
be in tension (their resistance increases).

P2 P2
STRAIN GAGE

DIAPHRAGM
P1 P1

P1 = P2 P1 > P2

Page 35 Solutions for Superior Results

The Strain Gage

The four strain gages are connected into a Wheatstone bridge circuit
to yield an electrical signal proportional to the strain or pressure on
the diaphragm.

Page 36 Solutions for Superior Results

 Headbox Level Measurement

P
SLICE PROFILE
P2 ADJUSTER Similar to headbox pressure,
headbox level is a critical factor in
P1
producing a uniform sheet. Strain
gage technology is used to measure
the pressure at two points in the
headbox (P1 and P2). The transmitter
converts the difference between P1
and P2 to a level measurement.

RECTIFIER
ROLLS FORMING
FABRIC

FROM STOCK
PREPARATION

Page 37 Solutions for Superior Results

 Consistency Measurement

Consistency is a measure of the

density of the stock. Changes in
STOCK the stock make-up or the refining
PREPARATION process can cause the consistency
HEAD BOX
to change. Since these changes
CT
impact sheet weight, consistency
measurement and control is
necessary to produce a high
quality product.

Page 38 Solutions for Superior Results

Consistency Measurement

Obtaining accurate and reliable

consistency measurement is
difficult. One technique often used
is to have a motor drive a paddle
wheel in the stock, and monitor the
current draw on the motor (as the
density of the stock increases the
motor will draw more current).

Page 39 Solutions for Superior Results

 Steam Pressure

After pressing, the sheet is

conveyed through the dryer section
where residual water is removed by
evaporation. The evaporation rate
is directly related to the steam
pressure inside the dryer cans.
Steam pressure is measured using
strain gage technology, discussed
earlier.

DRYER
CANS PT
STEAM
FROM
BOILER

Page 40 Solutions for Superior Results

 Speed Measurement

Speed measurements are used for machine set-up and production

information. Typically they are taken at the fourdrinier table (wire
speed) and at the dryer section or reel (machine speed).

PRESS CALENDER
HEADBOX REEL
SECTION STACK

STOCK FOURDRINIER DRYER

Honeywell Measurex
PREPARATION TABLE SECTION
SCANNER

Page 41 Solutions for Superior Results

Digital Tachometers

Speed measurement, on modern

paper machines, is typically done
using digital tachometers.

DIGITAL
TACHOMETER

Page 42 Solutions for Superior Results

Digital Tachometers

Similar to a generator, magnets

MAGNETS
are rotated past a coil of wire
creating an electrical current.
This current is in the form of a
square wave whose frequency is
directly related to the speed of
rotation.

Page 43 Solutions for Superior Results

 Temperature Measurement

The two most common sensors for temperature measurement are the
thermocouple and the resistance temperature detector.

CERAMIC PROTECTIVE
POWDER SHEATH
ALLOY A

COLD LEAD
HOT
CURRENT WIRES
JUNCTION JUNCTION

ALLOY B PLATINUM
ELEMENT
THERMOCOUPLE
RESISTANCE TEMPERATURE DETECTOR

Page 44 Solutions for Superior Results

 Thermocouples

Thermocouples use two wires of

unlike metals, joined at one end
called the “hot” end. At the cold
ALLOY A end the open circuit voltage is
measured. This voltage depends
HOT COLD upon the difference in temperature
CURRENT
JUNCTION JUNCTION
between the hot and cold junctions
ALLOY B and the Seebeck coefficient of the
two metals.

Page 45 Solutions for Superior Results

 Resistance Temperature Detectors

Resistance temperature detectors

(RTDs) use high resistance wire
whose resistance changes
CERAMIC PROTECTIVE
POWDER SHEATH relative to temperature. Changes
in resistance are detected by
LEAD
Wheatstone bridge circuits.
WIRES

PLATINUM
ELEMENT

Page 46 Solutions for Superior Results

Signal Flow

Typically, the current

4 - 20 ma produced by these sensors
TO Honeywell Measurex
is amplified and transmitted
FT
as a 4 - 20 ma signal (the
exception is the digital
tachometer).
This signal is sent through a
pair of wires to the
Honeywell Measurex
system for processing.

Page 47 Solutions for Superior Results

The Controller

The objective of the controller is

to:
SENSOR
• process the input signal from the
sensor
• compare the desired process
value to the input signal from the
sensor
CONTROL • act to reduce the error and bring
CONTROLLER ELEMENT
the actual value toward the
desired value

Page 48 Solutions for Superior Results

The Controller

All feedback control systems have a comparator and a controller.

COMPARATOR
TO
SETPOINT CONTROLLER CONTROL
ELEMENT

FROM
SENSOR

Page 49 Solutions for Superior Results

The Comparator

The comparator takes the difference between the setpoint and the
sensor signal. The error signal, which is the output of the
comparator, becomes the input to the feedback controller.

COMPARATOR
TO
SETPOINT CONTROLLER CONTROL
ELEMENT

FROM
SENSOR

Page 50 Solutions for Superior Results

The Controller

Based on the error signal from the comparator, the controller

calculates a signal to the final element. The mathematical function
performed in the controller is determined by the process dynamics.
Today’s modern controllers use proportional, integral, and derivative
(PID) functions.

COMPARATOR
TO
SETPOINT CONTROLLER CONTROL
ELEMENT

FROM
SENSOR

Page 51 Solutions for Superior Results

Control Elements

SENSOR
The output of the controller is
a signal to the final control
element which governs the
control of the manipulated
CONTROL variable.
CONTROLLER ELEMENT

Page 52 Solutions for Superior Results

 Control Valves

SIGNAL FROM
CONTROLLER
In many process control
systems, the final control
element is a valve - which is
ACTUATOR typically driven by an
actuator. The actuator
CONTROL
VALVE moves the stem of the control
valve to open or close the
valve. Control valve
actuators may be pneumatic,
electric, hydraulic, or manual.
A TYPICAL SLIDING STEM CONTROL
VALVE WITH A PNEUMATIC ACTUATOR

Page 53 Solutions for Superior Results

 Control Examples

SETPOINT The amount of stock placed on the

wire determines the basis weight of
the finished sheet. For example; if
CONTROLLER the sheet was too light, the
operator would increase the
setpoint of the stock flow
controller. This action would
cause the stock flow valve to open,
FT
placing more stock on the wire and
STOCK
increasing the sheet weight.
FLOW

STOCK FLOW

Page 54 Solutions for Superior Results

 Control Examples

DILUTION The consistency of the stock is

WATER
also a contributing factor in the
basis weight of the finished sheet.
As the consistency increases the
SETPOINT
sheet will get heavier. Attempts to
maintain the consistency at the
desired level are made by adjusting
DILUTION
CONTROLLER the amount of dilution water added
VALVE
to the system.

CT

STOCK
FLOW
CONSISTENCY

Page 55 Solutions for Superior Results

Control Examples

DRYER
CANS The amount of steam in the dryer
cans determines the moisture of the
finished sheet. For example; if the
sheet was too wet the operator
would increase the setpoint of the
steam pressure controller. This
action would cause the steam valve
to open, putting more steam in the
dryer cans - which would dry the
CONTROLLER PT
sheet.

STEAM
FROM
BOILER

STEAM PRESSURE

Page 56 Solutions for Superior Results

 Control Examples

COUCH
Many paper machines use a steam
turbine to drive the machine
PRESS
through a line shaft and a series of
belts and pulleys. To control the
machine speed, the master speed
DRYERS controller makes outputs to the
turbines’ speed control or
MASTER governor.
SPEED
CALENDER CONTROLLER

REEL

MACHINE SPEED TACHOMETER

Page 57 Solutions for Superior Results

Process Dynamics

All processes exhibit some form

of dynamic behavior. To study
this behavior, it’s helpful to look
at how a process responds to a
MANIPULATED step change in the manipulated
OUTPUT %

VARIABLE variable.

CONTROLLED
VARIABLE

TIME

Page 58 Solutions for Superior Results

Process Dynamics

The most common type of

dynamic behavior is the first-
order lag. It’s called this because
the output lags the input by an
MANIPULATED amount that can be described by
OUTPUT %

VARIABLE a first-order differential equation.

A simplified equation that
describe 1st order lag is
CONTROLLED
VARIABLE æ
ç − t ö÷÷
1− e
ç
è TC ø

TIME

The Response of a First-Order Lag

to a Step Input

Page 59 Solutions for Superior Results

 The First-Order Lag

The shape of the response curve

is described by a time constant.

100
90 The time constant is the amount
OUTPUT %

80 of time it takes the process to

70
experience 63.2% of the
60
50 remaining change.
40
30
20
10

TIME TIME
CONSTANT

Page 60 Solutions for Superior Results

 The First-Order Lag

There are two other

characteristics that are important
factors in processes with first-
100 order lags:
90
• Process Gain
OUTPUT %

80
70
60
• Time Delay
50
40
30
20
10

TIME TIME
CONSTANT

Page 61 Solutions for Superior Results

Process Gain

Process gain is determined by

changing the manipulated
MANIPULATED variable and observing the
VARIABLE
change that occurs in the
∆ MV
controlled variable.
OUTPUT %

CONTROLLED
VARIABLE
G = ∆CV / ∆MV

∆ CV

TIME

Page 62 Solutions for Superior Results

Time Delay

In many processes you will see a

100 response curve that exhibits a time
90
delay before any dynamic response
OUTPUT %

80
70 is observed.
60
50
40 Time delay can be calculated by
30
20
subtracting the time when the
10 process was changed from the time
the change was first observed.
TIME TIME
DELAY

PROCESS
CHANGED

Page 63 Solutions for Superior Results

 PID Controllers

For Da Vinci, a DCS System is used whenever outputting to a final control

element (e.g. valve). These systems have many options and can perform
many functions.
A common controller is the Proportional Integral Derivative (PID)
controller. These are implemented in most DCS systems and as single
loop controllers.

TO
PID OUTPUT
SETPOINT CONTROL
Components Components
ELEMENT

INPUT FROM
Components SENSOR

Page 64 Solutions for Superior Results

Input

The purpose of the PID loop controller input is to convert process

inputs into engineering units.

TO
PID OUTPUT
SETPOINT CONTROL
Components Components
ELEMENT

INPUT FROM
Components SENSOR

Single Loop Controller

Page 65 Solutions for Superior Results

PID Input

The 4 - 20 ma current
signal from the field
transmitter is connected
INPUT
Components to the input terminals.

FT

Page 66 Solutions for Superior Results

 ADC Conversion

Many controllers utilize

digital electronic
components. To process
0 - 20 ma 0 - 5VDC information from a field
transmitter, the signal
from the transmitter must
ADC
CONVERTER
be converted .

Page 67 Solutions for Superior Results

 Filtering

To attenuate noise from the

transmitter the raw signal is
smoothed using an
CONVERSION
VOLTAGE TO FILTER
exponential filter. This type
ENG. UNITS of filtering is similar to RC
circuits and is useful when
the noise frequency is below
1 Hz.

Page 68 Solutions for Superior Results

Filtering

The configured Filter Factor (FF) is defined by the

following equation:
FF = [1 - exp(-t/TC)]

Where: t = sample rate

TC = the desired time constant

The Filtered Value (value after the filter) is calculated

using the filter factor.
FV = New Reading - [(New Reading - Old FV) * (1 - Filter Factor)]

Page 69 Solutions for Superior Results

 Filtering

This graph shows the effects

1.0
FF = .8 of filtering a step response
FF = .4
using three different filter
FF = .2 factors (.8, .4, and .2 with a
0.5
sample rate t = .25). Note:
filter factors will always be
0.0
between 0 and 1, and as they
0.5 1.0 1.5 2.0 2.5 approach 1 less filtering is
Time (SEC) applied.

Page 70 Solutions for Superior Results

Alarm Limits

The controller can perform alarm

checks on each input signal. These
alarms warn operating personnel of
various undesirable conditions
ALARM
depending on the type of controller
FILTER and controller mode.
LIMITS

Page 71 Solutions for Superior Results

 Setpoint Comparison

The processed input signal is then compared to the desired setpoint.

Differences between the actual value and the setpoint are called
“error”.

TO
PID OUTPUT
SETPOINT CONTROL
Components Components
ELEMENT

INPUT FROM
Components SENSOR

Page 72 Solutions for Superior Results

The PID Controller
All controllers function as a special-purpose calculator. Using the
error signal from the comparator they calculate the changes needed
in the manipulated variable. The PID Controller allows you to
configure any combination of proportional, integral, or derivative
control actions.
The PID Controller computes a control move for the Output device,
based on the error between the PID’s setpoint and the process input.

TO
PID OUTPUT
SETPOINT CONTROL
Components Components
ELEMENT

INPUT
FROM
Components
SENSOR

Page 73 Solutions for Superior Results

Closing the Loop

So far we have devoted our

attention to the open-loop
SENSOR
behavior of the individual
components of a process control
system.

Closing the loop refers to how

CONTROL the process behaves when each
CONTROLLER ELEMENT of these components function as
one unit.

Page 74 Solutions for Superior Results

Closing the Loop

Each component of a control

OPEN-LOOP system has it’s own dynamic
behavior.

CLOSED-LOOP
When placed together in a
closed-loop, it’s likely that the
system’s output will oscillate.

TIME

Page 75 Solutions for Superior Results

Effects of Time Constant & Time Delay

In general, the more time constants

and time delays associated with a
control loop, the worse the control
problem.

It doesn’t make any difference if the

time constant or time delay is
associated with the valve, the sensor,
or the process itself, it will have the
TIME same ill effect on the process.

Page 76 Solutions for Superior Results

Stability vs Speed

Therefore, closing the loop

requires a trade off between speed
of response and stability.

For example, if we decreased the

controller gain the loop would
become more stable, however, it
would respond slower.

TIME

Page 77 Solutions for Superior Results

Optimum Control

Optimum control is achieved by

tuning the various parameters
within the controller to insure a
proper trade off between stability
and speed of response.

TIME

Page 78 Solutions for Superior Results

Tuning Parameters

Control tuning refers to the process of determining the optimum

parameters for the controller. Control loops use some or all of the
following tuning parameters:
– Gain (Process or Proportional)
– Integral Gain (or Reset)
– Time Constant
– Time Delay
– Dead Zone
•Valves can exhibit
– Stiction
– Backlash

Page 79 Solutions for Superior Results

Process Gain

Previously we defined process gain as ∆CV / ∆MV. Sometimes it’s

helpful to think of process gain as a units converter.

TO
SETPOINT CONTROLLER STOCK
(STOCK FLOW) VALVE
(SECONDS)

FROM
SENSOR

Page 80 Solutions for Superior Results

Process Gain

MANIPULATED
VARIABLE To calculate the process gain
make a “bump” to the process.
∆ MV Then divide the change in the
OUTPUT %

∆CV) by the
controlled variable (∆
CONTROLLED change in the manipulated
VARIABLE
∆MV) .
variable (∆
∆ CV

TIME

Page 81 Solutions for Superior Results

 Negative Gain Factor

SIGNAL FROM
CONTROLLER
The negative gain factor is used
to modify the process gain on
negative outputs. In effect the
PID controller has two gains one
ACTUATOR for positive moves and one for
negative moves.
CONTROL
VALVE

A TYPICAL SLIDING STEM CONTROL

VALVE WITH A PNEUMATIC ACTUATOR

Page 82 Solutions for Superior Results

Time Constant and Time Delay

To calculate the time constant

100
90
and time delay make a “bump” to
the process. Then determine the
OUTPUT %

80
70 total change in engineering units
∆EU
60 ∆EU). Multiply this amount by
(∆
50
40 .63 and add it to the starting
30 value.
20
10 ∆EU
∆ = 80
0 10 20 30 40 50 TIME 80 * .63 = 54.4
54.4 + 20 = 74.4

Page 83 Solutions for Superior Results

Time Constant and Time Delay

Transfer the 74.4 from the “Y”

100
90
axis to the curve. Transfer the
intersection point down to the
OUTPUT %

80
70 “X” axis.
60
50
40
30
The time constant is the
20 difference in time from when the
10 process began to move until it
0 10 20 30 40 50 TIME reached 63% of its final value.
23 - 10 = 13 seconds
TIME
CONSTANT

Page 84 Solutions for Superior Results

Time Constant and Time Delay

The time delay is the difference

100
90
in time from when the bump was
first made until the process
OUTPUT %

80
70 began to move.
60
50 10 - 0 = 10 seconds
40
30
20
10
0 10 20 30 40 50 TIME

TIME
DELAY

Page 85 Solutions for Superior Results

 Stiction

Stiction is the amount of time in

seconds needed to force the
OUTPUT %

actuator from standstill to a

moving condition. It can be
determined through graphical
analysis of the “bump test” data.

STICTION TIME

Page 86 Solutions for Superior Results

Backlash

Backlash compensates for

mechanical slack in an actuator
whenever the actuator reverses
direction.

It can be calculated using the data

from the “bump test” turn-around
moves .

Page 87 Solutions for Superior Results

Dead Zone

Dead Zone is the amount that a controlled variable is allowed to

vary around its setpoint without taking any control action. The
size of the dead zone is usually based on the amount of natural,
short term variation that remains when the control loop is in
manual. It can also be due to offsets caused by stiction values
that are smaller than the minimum output the controller is
capable of. The implementation of a dead zone can prevent
excessive wear of the control actuators by reducing unnecessary
control moves.

STOCK FLOW IN MANUAL

510

AVERAGE DEADZONE

500
TIME

Page 88 Solutions for Superior Results

Cascade Control

The general concept of cascade

control is to nest one feedback
SENSOR loop inside another feedback
loop.

TO LOWER
LEVEL LOOP
CONTROLLER

Page 89 Solutions for Superior Results

 Cascade Control

An example would be basis weight control.

BASIS WEIGHT STOCK FLOW

SETPOINT SETPOINT TO
BASIS WEIGHT STOCK FLOW
STOCK
CONTROLLER CONTROLLER
VALVE

INPUT FROM INPUT FROM

CONTROLLER SCANNER CONTROLLER STOCK FLOW
SENSOR

Page 90 Solutions for Superior Results

 Cascade Control

In this arrangement the output of the basis weight controller adjusts

the setpoint of the stock flow controller. These higher level cascade
loops are often called “supervisory” loops since their outputs
supervise the lower level loops.

BASIS WEIGHT STOCK FLOW

SETPOINT SETPOINT TO
BASIS WEIGHT STOCK FLOW
STOCK
CONTROLLER CONTROLLER
VALVE

INPUT FROM INPUT FROM

CONTROLLER SCANNER CONTROLLER STOCK FLOW
SENSOR

Page 91 Solutions for Superior Results

 Cascade Control

Cascade control is particularly useful when a very slow process is

involved. For example, a typical process will respond to a stock flow
change in less than five seconds. However, the resulting change to
the basis weight may take one to two minutes.

Cascade control allows you to control this intermediate variable (stock

flow) and to take corrective action on disturbances more promptly.

BASIS WEIGHT STOCK FLOW

SETPOINT SETPOINT TO
BASIS WEIGHT STOCK FLOW
STOCK
CONTROLLER CONTROLLER
VALVE

INPUT FROM INPUT FROM

CONTROLLER SCANNER CONTROLLER STOCK FLOW
SENSOR

Page 92 Solutions for Superior Results

Scanning Sensors

Many of the cascade

controls used in paper mills
rely on inputs from
Honeywell Measurex
scanners.
Measurements provided by
these scanners include
many sheet characteristics
such as: basis weight
(mass), moisture , caliper,
color, formation, ash,
smoothness, and opacity.

Page 93 Solutions for Superior Results

Scanning Sensors

The measurements from

the scanning sensors
provide accurate “real-time”
data that can be used to
adjust machine operation.
The result is improved
machine efficiency and
higher product quality.

Page 94 Solutions for Superior Results

Ash Measurement
• Accurate Ash /
Mineral
measurement
• High signal-to-
noise tuned X-ray
tube source
• Non-contacting
• Utilizes inputs from
Basis Weight and
Moisture gauges
• Compensation for
environmental
sources of error
– Air Temperature
– Z-axis Changes
– Dirt Buildup
Page 95 Solutions for Superior Results
 Basis Weight Measurement

• Fast Measurement Response

• High Resolution CD Profile
Measurement
• High Resolution MD Trend
Measurement
• Fast Profile Measurement
• High Frequency Process &
Quality Analysis

Page 96 Solutions for Superior Results

 Basis Weight Measurement

Air cylinder Air cylinders

for source for internal
shutter standards SOURCE

Air actuated
source shutter 2 internal
standards

Source
capsule

Integrated
“Close RECEIVER
geometry”
air curtain

Ionization
chamber
receiver

Model 4201 - BW Sensor

Precision Basis Weight Sensor

viewed from MD side
Model 4202 - BW Sensor
Source open, one internal standard inserted

Page 97 Solutions for Superior Results

Radiation Safety
• HMX Basis Weight and Ash sensors utilize ionizing
radiation sources.
– Basis Weight uses Krypton-85, an inert gas.
– If accidentally released, Krypton-85 dissipates quickly and
is rendered harmless when thoroughly dispersed.
– Some sensors at other mills contain solid sources, such
as Strontium or Americium, which must be treated with
greater care.
– Ash uses an X-ray tube which has its’ own 4kv power
supply. When the power is off, no X-rays are emitted.
• Radiation Indicators are located on each end of the
scanner frame.
– Red light indicates source shutters OPEN
– Green light indicates source shutters CLOSED
– Yellow light indicates Ash Power ON
Page 98 Solutions for Superior Results
Radiation Safety

• Safety Interlocks
– Hardware interlocks to force shutters closed if heads
accidentally become detached from scanner or are
damaged.
– Software interlocks to force shutters closed if heads are
separated; e.g. while cleaning sensors or if scan drive belt
breaks.
• Radiation Safety Checks are performed quarterly
by trained HMx personnel. Functionality of all
safety interlocks is tested.

Page 99 Solutions for Superior Results

 Caliper Measurement

• Magnetic reluctance
principle

• Accuracy through
contacting
measurement

• Full sheet measurement

and signal processing

Page 100 Solutions for Superior Results

 Caliper Measurement
Contacting Design for Accurate Measurement
Pressurised Bellows:
• Heat and chemical resistant
silicon rubber for long life
• Compresses sheet slightly for
correlation to laboratory caliper
• Follows passline variation and
sheet flutter
• Adjustable air pressure for Gimballed Skis Pressurised
Bellows
optimum performance
Gimballed Skis:
• Rugged stainless steel
construction
• Protects contacts and bellows
• Provides precise contact
alignment

Page 101 Solutions for Superior Results

 Caliper Measurement
Magnetic Reluctance Measurement
Frequency

Oscillator

Ferrite Core
Magnetic Flux
Sapphire Contact
Sheet
Passive Ferrite Contact

• Oscillator frequency is proportional to the

length of the magnetic path,
– i.e. the gap between the sapphire and the ferrite
contacts
Page 102 Solutions for Superior Results
Color Measurement
MODEL 2250 COLOR SENSOR Sensor Function
The Color Sensor measures the color,
brightness, and fluorescence of a moving
paper web by shining a specified illuminant
on the web, then measuring the properties of
the reflected light.

The Color Sensor consists of three functional

modules:

• Source of illumination
• Receiver, light analyzer, and computer
• Standards and sheet backing

The three modules are contained in the two

sensor heads. The illumination source,
receiver, light analyzer, and microcomputer
are in one head (generally the upper head),
and the standards and sheet backing are in
the other.

Page 103 Solutions for Superior Results

Moisture Measurement - Reel

• Fast measurement response

• High resolution CD profile
measurement
• High resolution MD trend
measurement
• Fast profile measurement
• High frequency process &
quality analysis
• Accurate measurement in
the presence of different
fillers, coatings, recycled
fibers and sheet temperature
variation

Page 104 Solutions for Superior Results

Moisture Measurement - Reel
IR Transmission
Reference Measurement
Wavelength Wavelength

Bone Dry
Moisture

Low
Moisture

High
Moisture

1.8 µ 1.9 µ Wavelength

• Moisture-sensitive measurement at 1.9 microns varies with

water content and basis weight
• Reference measurement at 1.8 microns varies with basis
weight

Page 105 Solutions for Superior Results

 Moisture Measurement - Reel

Receiver

Diffusing Reflecting
Surface

Sheet

Diffusing Reflecting
Surface
Infrared Light Source

• Patented INFRAND Optics

• Simple, powerful technique to achieve scattering insensitivity
• Quartz plates INFinitely RANDomly scatter source IR

Page 106 Solutions for Superior Results

Moisture Measurement - Size Press

• Non-contacting, reflective
measurement eliminates
breaks and sheet marking
• Low sensitivity to sheet
flutter
• Insensitive to formation and
non-uniform moisture
distribution
• Carbon correction permits
accurate measurement in the
presence of re-cycled fibre
• Small spot-size, fast
response and unique signal
processing provides
measurement to the edge of
sheet
Page 107 Solutions for Superior Results
Moisture Measurement - Size Press
Moisture
Detector
• Parallel processing 1.9µ
wavelength analysis
provides same-spot
measurement of all channels Carbon Reference
for: Detector Detector
2.1µ 1.8µ
– Formation-insensitive
measurement
– Fast sensor response
• SingleCal™ calibration
provides one calibration
group for most grades
• Carbon correction provides
accurate moisture
measurement on non- Sheet
deinked recycled stock

Page 108 Solutions for Superior Results

Sheet Measurement

Each sensor in the scanning head provides a machine direction and a

cross direction reading of the sheet.

MACHINE
DIRECTION CROSS
DIRECTION

Page 109 Solutions for Superior Results

 Machine Direction Measurement

The scanner continuously collects data as it travels across

the sheet. At the end of each scan, this data is averaged
and displayed as machine direction last average. Several
last average readings are combined to produce other
machine direction readings such as:
• Machine Direction Trend Average
• Machine Direction Reel Average

Page 110 Solutions for Superior Results

Cross Direction Measurement

CROSS DIRECTION PROFILE The data collected by the

scanner is sorted into cross
direction measurement
Basis Weight Last Average 28.5 lanes (slices), averaged,
Basis Weight Roll Average 28.7
and displayed as a cross
direction profile.

Page 111 Solutions for Superior Results

 Basis Weight Control Example

Machine direction basis weight measurements are compared to the basis

weight setpoint. Differences are acted on by the weight control and
outputs are cascaded to the setpoint of the stock flow controller. The stock
flow controller then adjusts the stock valve.

FT

STOCK
WEIGHT WEIGHT
FLOW
CONTROLLER SETPOINT
CONTROLLER
STOCK FLOW SETPOINT

Page 112 Solutions for Superior Results

 Moisture Control Example

Moisture measurements from the scanner are compared to the moisture

setpoint. Differences are acted on by the moisture controller and outputs
are cascaded to the setpoint of the steam pressure controller. The steam
pressure controller then adjusts the steam valve.

STEAM
PRESSURE
TRANSMITTER

PT

STEAM
STEAM
VALVE MOISTURE MOISTURE
PRESSURE
CONTROLLER SETPOINT
STEAM FROM CONTROLLER
BOILER HOUSE STEAM PRESSURE SETPOINT

Page 113 Solutions for Superior Results

 Tuning Cascade Loops

Like the loops they supervise, cascade loops have gain and time constant
tuning parameters. You may also find significant time delay in these loops.
For example, on a basis weight loop the time delay is the amount of time
from when the stock valve is moved until the effects are seen at the
scanner.

PT

Page 114 Solutions for Superior Results

 Decoupler Control

When feedback loops interact with one another, a control system is needed
that will decouple the loops. An example of this interaction can be seen on
a paper machine between weight and moisture. The effects can be
minimized through decoupler control.

STOCK
WEIGHT WEIGHT
FLOW
SETPOINT CONTROLLER
CONTROLLER
D1

D2
STEAM
MOISTURE MOISTURE
PRESSURE
SETPOINT CONTROLLER
CONTROLLER

Page 115 Solutions for Superior Results

Cross Direction Control

Using inputs from the scanner

measurements, the cross
direction properties of the sheet
can be controlled using various
cross direction actuators.

CROSS
DIRECTION

Page 116 Solutions for Superior Results

Cross Direction Control

ThermaTrol or Autoslice
(CD Weight)
Devronizer (CD Moisture)

AquaTrol (CD Moisture)

CalTrol or Calcoil (CD Caliper)

Page 117 Solutions for Superior Results

Weight Profile Optimizer

The basis weight profile control

supervises the slice lip
actuators to provide tight
control of the weight profile.

The control incorporates many

advanced models and
strategies to determine the
correct output for each
actuator.

Page 118 Solutions for Superior Results

Actuators

ThermaTrol Cross direction basis weight

MANUAL control is accomplished using
ADJUSTMENT ThermaTrol or Autoslice Motors
Actuators.

ELECTRIC
CURRENT The actuators open and close
APPLIED TO the slice lip.
HEAT ROD

SLICE ROD
EXPANDS AND
CONTRACTS

TO SLICE LIP

Page 119 Solutions for Superior Results

Moisture Profile Optimizer

MOISTURE PROFILE OPTIMIZER

Moisture profile control is a
flexible package that offers a
wide range of actuators including
Steamboxes and Aquatrol.

Page 120 Solutions for Superior Results

Aquatrol

Aquatrol actuators are used

DRYERS
near the dry end of a paper
machine to re-wet the sheet.

The result is reduced moisture

streaks and increased moisture
levels.

CALENDER
AQUATROL STACK

Page 121 Solutions for Superior Results

Caliper Profiler

The CalTrol II High Powered

Caliper Profiler precisely controls
the caliper profile by adjusting
variable temperature, high
turbulence air jets.

CALENDER
STACK
CALTROL

Page 122 Solutions for Superior Results

CalTrol

HEATING When hot air is applied to the

ELEMENT calender roll the roll will expand
- causing a corresponding
decrease in the caliper of the
CALENDER sheet.
AIR ROLL
FLOW
Numerous control zones
provide optimum caliper
profiling.

Page 123 Solutions for Superior Results

Display Profiles

MINISLICE PROFILE
The Lo-Res Profile is an
average of the Hi-res
readings taken over each
20 40 60 80 100 120 140 of the Lo-Res zones.
PROFILE
TRANSFORMATION
Typically, the display
profile is aligned as
closely as possible to the
actuator spacing on the
10 20 30 40 50 60 headbox.
DISPLAY
PROFILE

Page 124 Solutions for Superior Results

Lo-Res Zone Width

To create the Lo-Res Profile

the size of the display zone is
determined from the headbox
actuator spacing and
MINI
shrinkage factor. All the
SLICE zones in a Lo-Res profile are
the same size.

DISPLAY

Page 125 Solutions for Superior Results

Scanner Alignment

The purpose of scanner

alignment is to position
the sheet with reference
to the machine.

Alignment is necessary to
insure that cross direction
actuator movement
corresponds with the
desired position on the
sheet.

Page 126 Solutions for Superior Results

Control Profiles

MINISLICE PROFILE Control Profiles are an

average of the Hi-Res
readings taken over each
20 40 60 80 100 120 140 of the control zones.

PROFILE
TRANSFORMATION This allows accurate
alignment of the scanner
to the paper machine and
the cross direction
10 20 30 40 50 60 actuators.
CONTROL PROFILE

Page 127 Solutions for Superior Results

Control Profiles

On Da Vinci systems, control

profiles are derived from the
5 10 15 20 25 30 Hi-Res readings. A
DISPLAY
PROFILE
mathematical transformation
is used to calculate control
PROFILE
TRANSFORMATION
zones

10 20 30 40 50 60
CONTROL PROFILE

Page 128 Solutions for Superior Results

Control Zone Alignment

PROFILE CONTROL SUMMARY Since every cross

WEIGHT PROFILE OPTIMIZATION REEL MOI PROFILE OPTIMIZATION
direction controller has
different actuator spacing,
CALIPER PROFILE OPTIMIZATION R/S MOI PROFILE OPTIMIZATION
each controller must have
its own control zone
profile.

Page 129 Solutions for Superior Results

Cross Direction Control

ACTUATORS

Like all other control

loops, cross direction
control compares the
current measured value
to the desired setpoint.
Differences are acted on
by the controller and
CONTROL ZONE PROFILE signals are sent to the
TO REDUCE THE HIGH SPOT
appropriate actuator.
OPEN ACTUATOR # 3

Page 130 Solutions for Superior Results

Cross Direction Control

Most cross direction

controllers use traditional
proportional and integral
control techniques.
However, there are many
other factors that must be
accounted for such as:
• Actuator response time
• Process delay time
• Scan speed

Page 131 Solutions for Superior Results

Evaluating Control Performance

STOCK VALVE SETPOINT CHANGE

100 Short-term evaluation of a

control loop can be done
0
TIME by observing its response
to a setpoint change on a
STOCK FLOW RESPONSE
trend plot display.
1000

0
TIME

Page 132 Solutions for Superior Results

Evaluating Control Performance
STOCK VALVE SETPOINT CHANGE

100
Typical responses to a
0
TIME setpoint change when the
TOO AGGRESSIVE
control loop is not
1000
optimally tuned (could
also be the result of a
0
TIME
process that is in need of
repair).
TOO SLOW
1000

0
TIME

Page 133 Solutions for Superior Results

Assessing Loop Behavior

Oscillation in a control loop is

often believed to be caused by
1000 excessive controller gain.
However, this problem can
come from any of the following
sources:
0
TIME • The control valve or actuator
• Outside disturbances
• The controller tuning

Page 134 Solutions for Superior Results

Assessing Loop Behavior

Oscillations are a common control problem that

deserves our attention. Your first step, should be to
determine whether the oscillation is caused by the
controller, or comes from another source. This can be
done by placing the loop in manual.
DISTURBANCES

MANIPULATED CONTROLLED
PROCESS
VARIABLE VARIABLE

Page 135 Solutions for Superior Results

Assessing Loop Behavior

PUT THE LOOP IN

MANUAL MODE If the oscillations are still
present, after placing the
NO
STILL
YES
loop in manual, then they
OSCILLATING
? are being generated
outside the loop.
SEARCH FOR
CHECK THE VALVE
THE SOURCE

If the oscillations go away,

PROBLEM
NO they are being generated
? inside the loop.
YES
UNDERTAKE VALVE CHECK CONTROLLER
MAINTENANCE TUNING

Page 136 Solutions for Superior Results

Assessing Loop Behavior

BASIS WEIGHT MD SPREAD Control behavior should also

0.5
be assessed on a long term
basis. A simple and
effective way to do this is to
0
TIME
plot machine direction and
cross direction variation over
BASIS WEIGHT CD SPREAD time.
0.5

Tracking this variation over

time gives you a good
0
TIME indication of the controls
effectiveness.

Page 137 Solutions for Superior Results

Troubleshooting

• Common Problems
– System wide problems
M Blank videos.
M Can’t change video frames.

– Control problems
M Loop won’t go on control.
M Loop is unstable.
M Loop is too slow.
M Unable to change manipulated variable at all.
M Loss of process input readings.

Page 138 Solutions for Superior Results

Troubleshooting

• Common Problems (Continued)

– Scanner/Sensor problems
M Won’t scan.
M Scans OK but goes off-sheet by itself.
M Loss of sensor readings.
M Sensor readings unstable.
M Sensor readings don’t match lab checks.

Page 139 Solutions for Superior Results

Troubleshooting

• Approaches To Troubleshooting
– Interpreting video display information, status, etc.
M Control frames
M Trend plots
M Profiles
M Diagnostic frames
– Operating loops in manual.
– Operating loops in DDC control.
– Testing basic control inputs and outputs (increment,
decrement).
– Operating loops in supervisory control.

Page 140 Solutions for Superior Results