Nof S.Y. Springer Handbook of Automation

Подождите немного. Документ загружается.

1015

Automation i

58. Automation in the Wood and Paper Industry

Birgit Vogel-Heuser

Plant automation in the timber (wood) industry

and paper industry has many analogies due to

the similar process characteristics, i. e., hybrid

process. From a plant automation point of view

these industries are challenging because of their

technical requirements. The USA is the largest

producer of paper and paperboard. In a census

by the America Census Bureau, the paper industry

is placed 7 among 21 different industry groups,

with 6% of the total value of product shipments

for the years 2005 and 2006 [58.1]. China, as the

second largest paper producer after the USA, is

expecting a growth rate of 12.4% per annum (for

the forecast from 1990 to 2010) [58.2]. The German

woodworking machinery sector, with more than

26% of the world market share, had a € 3.4 billion

turnover in 2006 and 72% export quota [58.3].

In 2006, the German print and paper industry

grew by more than 7%, to € 8.5 billion and more

than 84% export quota ([58.4], more details

in [58.5]). This chapter will not only highlight the

speciﬁc requirements from a technical point of

view but also from the marketing point of view.

Observations from these two points of view will

lead to a heterogeneous automation system with

58.1 Background Development

and Theory ..........................................1015

58.2 Application Example, Guidelines,

and Techniques....................................1018

58.2.1 Timber Industry ...........................1018

58.2.2 Paper-Making Industry .................1023

58.3 Emerging Trends, Open Challenges.........1024

References ..................................................1025

proprietary devices for real-time and machinery-

safety-related tasks, and standard devices for

the rest. Both industries belong to plant man-

ufacturing industries with their typical business

characteristics. Automation in this ﬁeld is technol-

ogy driven and its importance is growing because

more functionalities are being implemented using

automation software to increase systems ﬂex-

ibility. The interface from automation level to

enterprise resource planning (ERP)systemsisbe-

ing standardized in international manufacturing

companies. Engineering is the key factor for im-

provement that needs to be considered in the

coming years, and therefore also modularity and

reusability where applicable.

58.1 Background Development and Theory

Both processes considered in this chapter are character-

ized by a large number of interrelated but independent

processing steps as well as complex control parameters.

Despite narrow limits and sophisticated process control,

inhomogeneous properties of raw materials will cause

variations in product quality. These variations cannot be

measured directlyat thetime ofproduction and can only

be determined subsequently by destructive testing.

An overview of the requirements of process au-

tomation in plant manufacturing industry is provided in

Table 58.1.

The criteria can be structured according to pro-

cess requirements, automation system architecture, and

project. A process automation system controls a type

of process, e.g., batch, continuous,ordiscrete.Some-

times processes are composed of different process types

and are then known as hybrid process. Both processes

discussed in this chapter are hybrid process.Hybrid pro-

cesses require different control strategies and therefore

also different modeling notations, e.g., block diagrams

(continuous) or state charts (batch). Both processes de-

mand a hybrid automation system architecture.

Part F 58

1016 Part F Industrial Automation

Table 58.1 Overview of the requirements of process automation

Categories/criteria Functionality/notation aspects

Process (hybrid) Batch (continuous) Transfer functions, block diagrams, differential equation

Discrete Status model, ﬂow chart, continuous function chart, Petri net

Real time Hard real-time and event-controlled system, synchronized drive

Reliability Mapping of reliability aspects (failure mode and effects analysis:

FMEA, fault tree analysis: FTA)

Automation system Implementation Programming language (C, IEC 61131-3, PEARL)

(heterogeneous) Operation system [proprietary real-time operating system (RTOS

proprietary)], Windows CE or XT

Hardware platform Programmable logic controller (PLC), distributed control system

(DCS), personal computer (PC), microcontroller

System architecture

heterogeneous

Central, decentralized, distributed

Human–machine

interface (HMI)

PC-based standard HMI tools and proprietary tools for fast trending

of process data

Speciﬁc requirements for both considered processes

in terms of process automation are:

1. Wood, as the input material, ﬂuctuates strongly due

to the environment.

2. The required mechanical construction relies on

heavy machinery.

3. They have hard and fast real-time requirements for

the control of central machinery with a large num-

ber of control loops, which does not allow the use of

one standard automation device, for example, a pro-

grammable logic controller (PLC) or a decentralized

control system (DCS). As an example, during par-

ticle board manufacturing in a hydraulic press, up

to 350 closed-loop controllers need to run within

10–20 ms cycle time.

4. Both timber and paper plants can consist of

3000–6000 analogue or digital inputs and outputs

that are connected via a ﬁeldbus. A paper plant

needs to control up to 3500 control loops using

10–20 central processing units (CPUs) (PLCsor

DCSs) that communicate via Ethernet-based bus

systems.

5. The huge number of analogue control loopsin a ma-

chine, in combination with requirements for fast

real-time operation, results in the use of speciﬁc au-

tomation solutions. In the timber industry a versa

module eurobus (VMEbus)-based control system is

used, whereas in the paper(making) industry, de-

centralized control systems (DCS) with proprietary

automation devices are used.

6. Worldwide environmental conditions lead to differ-

ent automation architectures, i.e., decentralized and

centralized design.

7. Consideration of both cross section and longitudinal

section in control is required. The width ofa pressed

board is up to 2.9 m and in a paper machine up to

11m, therefore the control and operation task has to

deal with these two dimensions.

8. The need for precise synchronization between sev-

eral drives in the machine, else the paper may

be ripped, the timber mat may be ripped or com-

pressed, or the steel belt of the hydraulic press may

be damaged, thus causing damage to the surface of

the panels. Speed of production depends strongly

on product thickness, and ranges from 300 mm/s

to 1.5m/s in the timber industry and up to 25 m/s

in the paper industry. Older solutions used vertical

shafts to synchronize the drives accurately.

9. Continuous production around the clock, 365days

a year, demands high availability of plant automa-

tion system, i.e., about 90% in the timber industry

and 99.9% in the paper industry.

10. Plant automation can automate almost 100% of

tasks, but some important measurements such as

moisture can still not be measured precisely, while

other values such as internal bonding and bending

strength are only available through laboratory ex-

periments such as destructive testing.

11. Implementation of closed-loop controllers for mat

weight or square weight, moisture, temperature, and

thickness is required. Quality criteria for the sur-

Part F 58.1

Automation in the Wood and Paper Industry 58.1 Background Development and Theory 1017

Table 58.2 Languages of the IEC 61131-3 standard

Name/Functionality Application

Ladder logic/circuit diagram On/off, lamps

Instruction list/assembler type Time-critical modules

Function block language/Boolean operations

and functions

Interlocking, controller, reusable functions,

communication

Sequential function chart/state diagram Sequences

Structured text/higher programming language Controller, technological functions

face structure include color and gloss in the paper

industry and varnishing in the timber industry.

12. From the control-theoretic point of view, today both

processes still lack a method that can model the

whole process and the central machinery, i. e., the

continuous hydraulic press in the timber industry

and the paper machine in the paper industry. A ﬁrst

attempt to develop a model for the central ma-

chine has been made for the timber industry [58.6].

A method to model the process has not yet been de-

veloped due to lack of information concerning the

process. Support from the operator and technolo-

gist is needed to bridge this gap, and to do this they

would need, for example, a fast trending tool (sam-

ple time 10 ms for 450 analogue inputs) to analyze

and optimize control loops in the entire machine or

process.

After the discussion of technical requirements the

typical market requirements should be mentioned.

The company- and/or product-speciﬁc technological

knowhow for both considered domains are of great

value. The market is very competitive, which is why

technological knowhow is kept within different pro-

duction companies. Technological feedback to the

machinery supplier is weak and therefore it is difﬁcult

to achieve technical control improvement. Suppliers are

mostly situated in high-wage countries but the mar-

ket is global, with growing competition from low-wage

countries, therefore engineering efﬁciency and reduced

engineering times are important for suppliers. Addi-

tionally, reduced start-up times and improved plant

reliability are growing challenges in every part of the

world. This means that engineering processes need to be

optimized through modularity and reuse, as well as data

integration along the engineering lifecycle. Modularity

and reusability in modeling and simulation provide the

opportunity to increase software quality while reducing

start-up time.

During the design of each plant location, various

requirements need to be taken into account, regard-

ing: standards, type, and stability of power supply from

power companies; ground connection; and the different

qualiﬁcations of operator and service personnel.

Due to the availability requirement (see item 9

above) and the protection of technological knowhow

while optimizing production, there is a need to provide

maintainable and adaptable systems for the worldwide

market. As far as possible standardized automation sys-

tems should be used, i.e., PLCs in the timber industry

and DCSs in the paper industry.

The end customer usually speciﬁes the brand of

PLC, human–machine interface (HMI), or DCS,and

sometimes also the drives used (if technologically pos-

sible). PLC and DCS systems are programmed accord-

ing to the International Electrotechnical Commission

(IEC) 61131-3 standards (Table 58.2). Customers in the

USA prefer ladder logic as the programming language

whenever possible. Technologically complex functions

will be encapsulated in function blocks. However, stan-

dard PLCsandDCSs are not appropriate for fast

real-time requirements and they can be manipulated by

end customers; this should not be allowed as they are

complex devices and there is the risk of machine haz-

ards. This is why heterogeneous automation systems

are standard in both industries. These heterogeneous

systems consist of standard devices, standard HMIs,

and ﬁeldbus solutions, enriched with domain- or even

company-speciﬁc solutions for fast real-time control.

The important difference between the two industries is

that the requirement for intrinsic safety is more relevant

to the paper industry compared with the timber indus-

try. The intrinsic safety requirement has a strong impact

on the selection of appropriate equipment, especially

sensors and actuators as well as ﬁeldbus solutions.

Some green-ﬁeld plants for the timber industry

include glue production and/or paper machinesthat pro-

duce paper necessary for lamination.

Part F 58.1

1018 Part F Industrial Automation

Most machinery will be shipped in parts and is only

commissioned on site, therefore machines or machine

components will only be tested at the supplier’s site.

The ﬁnal test is necessary after commissioning on site.

Training of maintenance personnel and operators

takes place during commissioning and start-up. Oper-

ator training systems (OTS), which include simulation

of the process, will be developed in the near future.

58.2 Application Example, Guidelines, and Techniques

As discussed, the requirements of the timber and paper

industries are similar. This Chapter will ﬁrst explain the

continuous hydraulic press as an application example

to show these requirements in more detail and appro-

priate automation solutions. After that, one solution for

a paper machine will be introduced to highlight the sim-

ilarities in terms of requirements and solutions.

58.2.1 Timber Industry

The production of ﬁber boards in the timber industry is

a hybrid production process which generates a product

(wooden panels) that has to meet many quality require-

ments. The thickness and density of the board or the

consistency of the material are target values that need

to be controlled within the tight intervals. The process

Storage,

debarking

Chip

washing

...

Squeeze out water

Steam supply to digester

Filling level of digester

Digester temperature

Digester steam pressure

Cooking time

Chip quantity

Paraffin supply

Refiner power consumption

Refiner temperature

Refiner steam pressure

Refiner grinding gap

Refiner disc age

Blowout valve outlet

pH value of fibers

Glue quantity ...

...

Fibre discharge quantity

Spreading height

Height of scalping roll

Forming belt speed

Mat weight per unit area

Mat moisture

Spreading width

Prepress pressures

Prepress gap

Mat density

Spray water quantity

Height of mat

Mat temperature

Sprayer height

Mat rejection ...

Defibration

and gluing

Glue

preparation

Fibre preparation Matforming Hot-

pressing

Paraffin

preparation

Fibre

drying

Hogging,

Sifting,

Bunker

Fig. 58.1 Overview of the ﬁber board production process (after [58.7])

consists of multiple steps where wood as the input ma-

terial is prepared (Fig.58.1). In the subsequent steps,

the ﬁber mat is formedthrough continuouspressing pro-

cess, heated, hardened, and sawn into discrete boards.

Wood is anatural material thatis stronglyinﬂuenced

by temperature and humidity conditions. The input ma-

terial’s characteristics for the process varies widely.

The most costly piece of equipment to acquire and

operate is the hydraulic press. It usually determines

the maximum capacity of the production line [58.8].

The simpliﬁed press principle is described as follows.

For the ﬁnished board, the material (already mixed

with glue) has to be pressed with a speciﬁc pressure

to obtain a certain thickness that is related to a set

value.Wood is a natural materialwith underlying strong

disturbing inﬂuences such as temperature or humidity

Part F 58.2

Automation in the Wood and Paper Industry 58.2 Application Example, Guidelines, and Techniques 1019

conditions. Thereby the characteristic of the input ma-

terial to the press varies widely and cannot be measured

with sufﬁcient accuracy for automatic control. Hence

the operator should be able to inﬂuence the automatic

control strongly in order to compensate for such inﬂu-

ences, based on his experience. The operator must be

able to recognize that the situation is no longer in the

range of the selected recipe and that he has to react by

Sys_01

Pressure profile Distance profile

–

Inclination

250

GIC

Frame 5

Frame 4

Frame 3

Frame 1

Frame 2

200

150

100

Sys_02

Sys_03

Sys_04

Sys_05

Sys_06

Sys_07

Sys_08

Sys_09

Sys_010

Sys_01

Sys_03 middle: 220.361

Sys_02

Sys_03

Sys_04

Sys_05

Sys_06

Sys_07

Sys_08

Sys_09

Sys_010

Right

Middle

GIC

Y4HR

PIC

Y4HR

Y4R

PIC

Y4HR

Y4R

Y4L

PIC

Y4L

Y4HL

PIC

Y4HL

Y4M

PIC

Y4M

GIC

Fig. 58.2 (a) Process and instrumentation diagram of a continuous thermohydraulic press (only distance and pressure

control, working direction from left to right).

(b) Pressure and distance proﬁle of a speciﬁc hydraulic press in 3-D; x-axis

shows working direction from left to right, y-axis displays pressure proﬁle (left) and distance proﬁle (right) (after [58.9])

changing the distance or pressure parameters in a spe-

ciﬁc frame or group of frames (Fig. 58.2a). Therefore

the operator needs a good overview of the real values of

pressure and distance in cross-section and longitudinal

directions (Fig.58.2b).

Inclination of the press, which would produce an

improper board, can be monitored on the distance

proﬁle (Fig.58.2b right, section 6 slight inclination).

Part F 58.2

1020 Part F Industrial Automation

Torsion of the press table would be identiﬁed easily

using this presentation. It is possible to identify if the

pressure is too high along and/or across the press us-

ing the pressure proﬁle (Fig. 58.2b left). The shape

allows evaluation of whether the mat will be pressed

regularly and symmetrically (across the cross section).

Due to technical requirements a maximum pressure

is set in order to achieve the thickness of the mater-

ial. Distance control is realized by several hydraulic

systems that consist of pressure transmitter and pro-

portional valve with position sensor. One frame may

consist of ﬁve hydraulic systems. The distance is meas-

ured on the left and right edge of the frame. A press

may have up to 70 or more frames. Therefore, up to

350 pressure values and roughly 140 distance values,

all with a spatial relation to each other, need to be

controlled and displayed. The process and instrumen-

tation diagram for ﬁve frames (Fig.58.2a) shows the

hydraulic cylinders, the pressure measurement, and the

distance measurement. Real-time requirements for the

entire loop need to be taken into account (Fig.58.2a).

Bus

coupler

BusBus

Automation device

PLC

Trans-

ducer

Trans-

mitter

Trans-

mitter

Bus

coupler

5 ms

Pressure transmitterI/O-module (bus)

I/O-module (bus)

Actuator

(proportional valve)

up to 75 ms

2 ms

≈20 ms

≈1–5 ms≈1–5 ms

Servo-

motor

Bus

adapter

Process variable

Fig. 58.3 System components of the control loop that needs to be

considered for (hydraulic press) cycle time calculation. I/O –int-

put/output

The controllers are more or less simple proportional–

integral–differential (PID) controllers with ﬁlters or

nested PID controllers. Complexity depends on their

interdependency. Hydraulic systems are mechanically

coupled to a thick steel plate to heat the mat so that

the glue will harden. Besides the pressure control and

distance control, temperature control and controllers to

synchronize different drives at the inlet and the outlet

of the press are also included. These drives need to be

synchronized with the different forming linedrives (mat

forming and prepress) to avoid material problems in the

press inlet.

Additionally the upper and lower steel belts need to

be controlled depending on the position of the material

in the press during ﬁrst inlet or product changes. At the

outlet of the press, synchronization between the cut-to-

size saw and the cooling and stacking line is required.

The endless board needs to be cut into pieces at high

speed, e.g., 1.5m/s. In the timber industry a cut-to-size

saw is included for this task. Additional controllers in

the forming line or press may be added depending on

product requirements and the speed of the production

process.

Figure 58.3 shows the usual time delays in the en-

tire control cycle from data measurement to its effective

inﬂuence on the process. It is necessary to reduce the

cycle time for each distance controller in the controller

itself to under 20ms as hydraulic systems are highly

dynamic, especially when the press is operated at top

speed. In addition, synchronization between different

frames is required if switch to another control mode is

necessary.

Currently there is one supplier who provides

a method to calculate the tension in the heating plate.

The calculation is done by a machine-safety-related

controller based on a ﬁnite-element calculation in the

automation system.

Speciﬁc process control systems, which depends

on the supplier’s chosen solution, e.g., Motorola and

VMEbus-based or personal computer (PC)-based sys-

tems with real-time operating system (RTOS), are

needed to fulﬁll the strict time requirement. Only one

supplier delivers a PLC solution, which needs less cal-

culation power for a continuous press because it uses

a different mechanical concept.

All distance control needs a data delay less than

10–20ms in order to optimize the control loops for thin

board. In orderto optimizea controller, itis necessaryto

use a data sampling rate that is 5–10 times faster. This

results in an optimized data gathering strategy and the

development of proprietary trending systems [58.10].

Part F 58.2

Automation in the Wood and Paper Industry 58.2 Application Example, Guidelines, and Techniques 1021

Classically, HMI trending functionality is limited to the

fastest sample time of 3 s or even slower because they

focus on long-term trending. PLCs are connected with

an Ethernet-based network (Fig.58.4). Sensors and ac-

tuators are connected with a ﬁeldbus system to ensure

deterministic behavior. The cut-to-size saw is realized

on PLCs with controllers for servo drives, or on more

or less computer numerical control (CNC)-type devices

to enable the required high speed.

Due to customer requirements in plant operation,

such as long mean time between failures (MTBF), plant

maintenance should be carried out by their plant service

personnel. They should have the possibility to manip-

ulate the PLC program within certain limits and to

continue the optimization of processes. Therefore PLC

is the preferred automation approach. The customer can

specify the PLC supplier and bus systems supplier ac-

cording to the market shares in his company and/or

country, the service structure of the PLC supplier, and

the skills of his maintenance personnel.

It is a marketing advantage to implement all control

functionality on PLC. Due to the growth in system ma-

nipulation, a new challenge for PLC is emerging: the

ability to track operator input and manipulations made

on the PLC so that improper manipulations that may

Statistical process (quality)

control (entire plant)

ERP-

interface

Fast trending

press control

Forming line/

press

Forming line

control

Material

preparation

HMIHMI HMI

Forming line/

press

HMI

Forming line/

press

HMI

HMI for

recipe

input

Tren-

ding

MES

Press

forming line

process net

(TCP/IP)

Press control

distance/pressure

PLC preparation

PLC PLC press

(S7) (S7)(S7)

Modem

remote

maintenance

Modem

remote

maintenance

Press control

VMEbus

based

or PC based

FieldbusFieldbusFieldbusFieldbus

Fig. 58.4 Automation architecture (example from one supplier). TCP/IP – transmission control protocol/Internet proto-

col (MES – manufacturing execution system)

cause hazards can be detected and tracked. Password

protection alone is not sufﬁcient.

Regarding safety requirements, a strategy to over-

come loss of power or press stoppage with material

in the press is required because the material may start

burning after a long time. Uninterruptible power sup-

ply (UPS) and emergency power supply are standard

equipment.

The cooling and stacking line (sanding and trim-

ming of particle boards) is needed to cool down

the panel and prepare it for the ﬁnishing line with

lamination or paneldivision,sorting, and packing.Inter-

mediate and delivery storages in timber industry often

follow a simple but chaotic storage strategy. However,

tracking and tracing of boards is becoming increasingly

important to the customer. For this reason, ERP and

production management are becoming more important

and a more precise storage system is needed. Board

handling is often realized by using forklift trucks. The

boards are chaotically stacked in a building until the

truck from thecustomer arrives. They are identiﬁedonly

by a piece of paper with a printed barcode. Due to the

manual handling strategy, tracking of faulty material

is nearly impossible. Two strategies will be discussed

and evaluated later (Sect. 58.2): radiofrequency identi-

Part F 58.2

1022 Part F Industrial Automation

ﬁcation (RFID) of every board and global positioning

system (GPS) detection of every stack.

There are also automated intermediate storage sys-

tems that allow handling of different board sizes on

a mobile electrically powered rail-bound device using

steel pallets [58.13]. The material ﬂow is tracked and

the product data are available. In production of thin

medium-density ﬁber (MDF), board pallets need to be

equipped with top and bottom protection panels to keep

the product ﬂat and protect its surface.

The quality of the board is not only inﬂuenced

by the press but in all sections of the plant, particu-

larly by the properties of the wood mixture and glue

used [58.14]. The plant sections involved in the entire

manufacturing process – starting with the hogged chip

manufacturing, ﬂaking or deﬁbrating, drying, blending,

straining, mat forming, up to prepressing and pressing

sections – are all interdependent and, like the properties

of the raw materials used, are subject to ﬂuctuations.

There are more than 100 parameters that have vari-

ous intensities of effects on a plant’s productivity and

the quality of the product (Fig. 58.5). These parameters

are input values for a process model based on a sta-

tistical algorithm [the three-stage least-squares (3SLS)

algorithm [58.15,16]].

Optimization

cost (Y) → min

nom

(t+1)< Q

pred

(X(t))+Y

opt

(t+1)) +S

pred

(x(t),Y

opt

(t+1))

nom

(t+1)

< Q

pred

(t)+S

pred

(t)

pred

(t): prediction of quality parameter

pred

(t): statistical safety reserve

X(t): (quality relevant) process parameter

Y(t): control parameter

Predictor

(process model)

Controller

Process optimization

Model-based (quality) predictive process control

Online quality control

Nom. value storage

nom

(0)

nom

Data storage

X(t–1), X(t–2), ...

Y(t–1), Y(t–2), ...

Q(t–1), Q(t–2), ...

nom

(t+1) = Y

opt

(t+1)

pred

(X(t), Y

opt

(t+1))

pred

(X(t), Y

opt

(t+1))

opt

(t+1)

Fig. 58.5 Global control loop using model-based quality-predictive and cost-optimized process control (after [58.11,12])

Prerequisites for the analysis of data detailing the

production history of the product, so-called material

tracking and correspondingly time-correlated process

data, need to be calculated. First implementations of

a model-based quality-prediction and cost-optimization

process control are running successfully [58.11,12].

The quality of relevant process data and control

parameters are taken into account to predict the result-

ing quality. Optimized process settings Y

opt

for control

parameters Y are calculated to ensure that nominal re-

quired quality in terms of costs is met

cost(Y) →Minimum

under the condition

nom

(t +1)

< Q

pred



X(t), Y

opt

(t +1)



pred



X(t), Y

opt

(t +1)



and

nom

(t +1)

> Q

pred



X(t), Y

opt

(t +1)



−S

pred



X(t), Y

opt

(t +1)



To close the loop, these optimized process settings are

entered as new nominal values for the control param-

eters.

Part F 58.2

Automation in the Wood and Paper Industry 58.2 Application Example, Guidelines, and Techniques 1023

58.2.2 Paper-Making Industry

Heterogeneous automation systems are implemented in

the paper industry to control force, torque, temperature,

moisture, and vibration. Required reliability is 0.9999

(24h/365 days per year). An overview of process sec-

tions, controllers, and most importantly sensors is given

in Fig. 58.6 (working direction: left to right, with the

process divided into two ﬁgures). The paper machine is

a device for continuously forming, dewatering, press-

ing, and drying a web of paper ﬁbers [58.18].

The automation system is mainly realized using

a DCS system. In the given example, the automation

system was being realized using PCS 7 Siemens based

on the 400 CPUs. The connection to the sensors and

actuators is realized using the PROFIBUS DP ﬁeld-

bus. Speciﬁc automation devices are implemented for

cross-section control and transmission control. Most

measuring devices, e.g., gloss, moisture, caliper and

basis weight, and color, are equipped with their own au-

Wet end control

Stock consistency; extender; retention;

workload; gas (machine longitudinal)

Scan

Moisture temperature

Scanner

Grammage , strip weight

moisture, gloss, caliper

One side

scanner

Moisture

Scanner

Moisture

grammage

Longitudinal control

Closed-loop control walls; thick-matter flow regulator

grammage, separation of steam pressure; coordinated

change of speed, moisture; production output maximization

Grammage

Coat weight Selective elimination of wet strips

Moisture

Track inspection

system

Track inspection

system

Fig. 58.6 Example of a paper machine: process overview with sensors and actuators (after [58.17])

tomation device and will be connected with a controller

area network bus (CANbus) to the cross-section con-

trol and over a measurement server (PC)totheHMI.

Voith, for example, developed its own weight proﬁle

control system software named Proﬁlmatic. Proﬁlmatic

cross-direction control software continuously and au-

tomatically aligns each actuator against its respective

measurement position from the downstream scanner.

An automapping algorithm monitors the movement of

a normal proﬁle control and aligns the cross-direction

measurement data boxes against the actuator control

zones. Each control output array compares the ac-

tual proﬁle change against the expected proﬁle change.

A software model continuously updates the process-

mapping model using the difference between measured

and expected response.

Model-based soft sensors deliver high-quality data

for superior longitudinal control of the paper machine.

Similar to in the wood industry, a multivariate statistical

approach is implemented to forecast, for example, the

Part F 58.2

1024 Part F Industrial Automation

weight proﬁle at the end of the paper machine. Some

basic constraints for the implementation of statistical

process (quality) control in paper industry are given

in [58.12], e.g.,

this assumes that all of the important variables are

measured in a timely manner ... Another point to

consider is the sampling frequency. Generally, the

more variability in a material, the more often it

should be sampled.

Blatzheim gave an example of the beneﬁt of such

system: 15–16km of waste material reduction was

achieved after implementing this system, corresponding

to an approximate increase of 200 000 of sellable paper

per year.

Laukkanen mentioned OTS and appropriate work-

ﬂow guidelines as being one of the keys to reduce

commissioning and start-up, especially abroad [58.3].

Through the engineering life cycle, depending

on the project phase, different tools and methods

are used. In the worst case, the corresponding data

have to be re-entered during the transition from one

phase to the next because there are no appropri-

ate interfaces between the individual tools. The ideal

is to strive for a higher-level tool that consistently

provides all system information in a model and en-

ables the design to be realized both at an abstract

level and in a conventional environment (e.g., electri-

cal engineering computer aided engineering (E-CAE),

IEC 61131-3).

58.3 Emerging Trends, Open Challenges

There are many open challenges in the timber and paper

industries. Some will be discussed in this section, i.e.,

engineering lifecycle and data integration, reduction

of complexity for operators, improvement of operator

training, as well as evaluation of new technologies.

Open challenges are to increase engineering ef-

ﬁciency while reducing costs and start-up time by

applying modularity andreuse. Communicationis a ma-

jor factor to consider when improving engineering

quality because different disciplines such as sales, tech-

nology, mechanical engineering (including hydraulic),

electrical engineering, and computer science are in-

volved through different phases in the engineering

process. Communication could be supported by a com-

prehensive model. Unfortunately modeling of hybrid

systems in process automation is still a ﬁeld for

research and development, especially for reuse and

modularity, and when the target group is engineers

or technicians coming from different disciplines. One

promising approach is to apply a comprehensive mod-

eling notation such as the systems modeling language

(SysML) [58.19] that is based on the uniﬁed modeling

language (UML) and developed for systems engineer-

ing. This basically solves the deﬁciencies of UML for

automation, i.e., modeling of hardware aspects, inte-

gration, and tracking of requirements. The advantage

of UML in terms of supporting modularity through an

object-oriented mechanism is fully available. The task

is to evaluate SysML in terms of ease of application by

engineers and technicians in automation depending on

the availability of strong tools. One of the ﬁrst steps

to integrate modeling into IEC 61131-3 is already in

progress in a research and development (R&D) project

in Germany. Various UML diagrams are implemented

in one of the market leading IEC 61131-3 tools [58.20].

Presenting the modules throughout the engineering life

cycle from customers’ requirements to operation is still

not solved and will not only be a task for research

but also for the development department of leading au-

tomation companies. Plant manufacturing companies

need an easy way to evaluate whether a new plant con-

cept could be realized by reusing existing modules or

to determine the number of new modules that need

to be developed. Data integration throughout the en-

gineering life cycle is theoretically solved but still far

away from realization in the applied tools. Coupling of

computer-aided engineering (CAE) system on the basis

of a tool-to-tool interface is not sufﬁcient for the two

considered industries.

There is also huge potential for cost saving during

start-up by integrating engineering data into an ERP

system, and offshoring in a global market. Additional

elaboration on logistics, e.g., just-in-time machine or

component to site, is needed. Today mainly mechan-

ical construction has been subject to offshoring. This

will change during the next few years. A lot of chal-

lenges for the engineering workﬂow will consequently

be affected.

Operation and maintenance, which is challenging

to improve, is also a very important phase in the en-

gineering life cycle. Application of three-dimensional

(3-D) visualization can successfully reduce complexity

Part F 58.3