Technological
Papers
Know-How in Process Control
Danieli Automation
Pickling line modeling
for advanced process monitoring and automation
1 Different designs of
pickling line tanks:
a) Deep-type tanks
b) Shallow-type tanks
c) Turbulence tanks
d) Turboflo tanks.
A virtual platform for sensorless monitoring
and consumption optimization of the pickling process.
downcoil temperature (i.e. the strip
temperature when leaving the rolling mill),
and cooling process duration. This oxide
layer must be removed before further strip
processing in cold rolling plants.
Pickling of carbon steel
The local current reduces the trivalent ions
in the wustite, transforming them into acidsoluble bivalent iron ions. The scale is
quickly dissolved. The chemical reaction is
generally accelerated by heating the acid
solution to 65–85 °C.
Different design concepts of strip processing
tanks characterize different pickling systems
as shown in Figure 1.
They are briefly described in the following.
The descaling of hot strips of carbon steel is
normally achieved in pickling lines consisting
of many processing tanks where the steel
strip comes into contact with a corrosive
solution, normally hydrochloric acid. The
acid reacts with the oxide layer. Iron
chlorides and iron ions are dissolved into the
acid baths.
Maintaining the baths with fresh acid and
discharging the exhaust solution maintains
the processing capability of the line. In most
cases, the pickling line is coupled with an
Acid Regeneration Plant (ARP), which
regenerates the exhaust solution by removing
the iron content, thus reducing the
consumption of fresh acid.
Usually, the scale thickness is 10 to 20 μm.
Its so-called scale structure is not uniform,
but mainly consists of a layer of wustite
(FeO). According to the descaling model
described by Frisch and Thiele (see Gruchot
et al., 2001), the acid penetrates the scale
structure, thus reaching the free metal
surface. The reaction occurring between acid
and free metal generates a local current
flowing from the scale-free metal portion,
acting as an anode, to the conductive wustite
layer, playing the role of a cathode.
384
Edited by
C. Aurora
D. Sclauzero
F. A. Cuzzola
Deep tank pickling
b)
Shallow tank pickling
c)
Turbulence pickling
d)
TurboFlo pickling
Flat products
After leaving the rolling mill, hot steel strip
reacts with both air and cooling water,
which leads to scaling, the formation of a
surface oxide layer. The scale’s properties
depend on many factors, including the
steel’s chemical composition, the so-called
a)
[ Danieli Automation ]
Deep-type pickling lines consist of deep
working tanks where the acid solution presents
an extremely slow flow motion.
Shallow-type pickling lines rely on a different
tank design, which guarantee a higher pickling
effect on the bottom surface of the strip with
respect to conventional deep-type lines, even
with slow motion of acid flow.
Turboflo™ pickling lines, developed by Danieli
Wean United, consist of pickling tanks divided
into a number of cells of 2 m length with
special tank covers (Gruchot et al., 2001).
The high kinetic energy of the directly injected
acid accelerates the descaling process and
increases the strip/acid heat transfer
coefficient. High strip speeds are achieved: up
to 400 m/min for light gauge (0.7–0.8 mm) strip.
Pickling of stainless steel
Stainless steel exists in many different alloys
and is available in the form of cold rolled and
hot rolled strip. In both cases, a re-crystallizing
treatment in annealing furnaces is required
before the strip can be used or cold rolled
again. Thus, repeated descaling of stainless
steel strip needs to be performed, to remove
mill scale as well as scale resulting from the
annealing process.
With respect to carbon steel, the scale from
stainless steel is more difficult to remove.
Moreover, the hot rolling and annealing
processes lead to the diffusion of chromium
from the upper layer of the base material into
the scale. The resulting chromium depleted
layer on the strip surface needs to be removed
by the pickling process.
The annealing and descaling of stainless steel
are often combined into a single processing
line (Kani et al., 1996). Since the annealing
section governs the strip speed, the pickling
process must guarantee suitable operational
flexibility to avoid unwanted underpickling (i.e.,
incomplete scale removal) or overpickling
phenomena. This requirement led to the
development of specific descaling processes
for stainless steel.
Neutral-electrolytic pickling lines, in particular,
offer some unquestionable advantages in
terms of process controllability and acid
consumption reduction as compared to the old
stainless steel descaling technologies of acid
or fused-salt baths. The self-regenerating
neutral electrolyte (Na2SO4) is inexpensive
and operates over a wide range of pH values
(4.5–7.0). The pH value can be properly
adjusted by the addition of caustic soda, if the
pH is too low, or by the addition of sulphuric
acid, if the pH is too high. Electric current
controls the descaling process: in case of a
plant stoppage, the neutral electrolyte does not
attack the base material. Moreover,
overpickling never occurs, since the descaling
reaction practically stops by itself after the
complete removal of the scale.
385
Danieli Automation
Pickling line modeling
for advanced process monitoring and automation
1 Different designs of
pickling line tanks:
a) Deep-type tanks
b) Shallow-type tanks
c) Turbulence tanks
d) Turboflo tanks.
A virtual platform for sensorless monitoring
and consumption optimization of the pickling process.
downcoil temperature (i.e. the strip
temperature when leaving the rolling mill),
and cooling process duration. This oxide
layer must be removed before further strip
processing in cold rolling plants.
Pickling of carbon steel
The local current reduces the trivalent ions
in the wustite, transforming them into acidsoluble bivalent iron ions. The scale is
quickly dissolved. The chemical reaction is
generally accelerated by heating the acid
solution to 65–85 °C.
Different design concepts of strip processing
tanks characterize different pickling systems
as shown in Figure 1.
They are briefly described in the following.
The descaling of hot strips of carbon steel is
normally achieved in pickling lines consisting
of many processing tanks where the steel
strip comes into contact with a corrosive
solution, normally hydrochloric acid. The
acid reacts with the oxide layer. Iron
chlorides and iron ions are dissolved into the
acid baths.
Maintaining the baths with fresh acid and
discharging the exhaust solution maintains
the processing capability of the line. In most
cases, the pickling line is coupled with an
Acid Regeneration Plant (ARP), which
regenerates the exhaust solution by removing
the iron content, thus reducing the
consumption of fresh acid.
Usually, the scale thickness is 10 to 20 μm.
Its so-called scale structure is not uniform,
but mainly consists of a layer of wustite
(FeO). According to the descaling model
described by Frisch and Thiele (see Gruchot
et al., 2001), the acid penetrates the scale
structure, thus reaching the free metal
surface. The reaction occurring between acid
and free metal generates a local current
flowing from the scale-free metal portion,
acting as an anode, to the conductive wustite
layer, playing the role of a cathode.
384
Edited by
C. Aurora
D. Sclauzero
F. A. Cuzzola
Deep tank pickling
b)
Shallow tank pickling
c)
Turbulence pickling
d)
TurboFlo pickling
Flat products
After leaving the rolling mill, hot steel strip
reacts with both air and cooling water,
which leads to scaling, the formation of a
surface oxide layer. The scale’s properties
depend on many factors, including the
steel’s chemical composition, the so-called
a)
[ Danieli Automation ]
Deep-type pickling lines consist of deep
working tanks where the acid solution presents
an extremely slow flow motion.
Shallow-type pickling lines rely on a different
tank design, which guarantee a higher pickling
effect on the bottom surface of the strip with
respect to conventional deep-type lines, even
with slow motion of acid flow.
Turboflo™ pickling lines, developed by Danieli
Wean United, consist of pickling tanks divided
into a number of cells of 2 m length with
special tank covers (Gruchot et al., 2001).
The high kinetic energy of the directly injected
acid accelerates the descaling process and
increases the strip/acid heat transfer
coefficient. High strip speeds are achieved: up
to 400 m/min for light gauge (0.7–0.8 mm) strip.
Pickling of stainless steel
Stainless steel exists in many different alloys
and is available in the form of cold rolled and
hot rolled strip. In both cases, a re-crystallizing
treatment in annealing furnaces is required
before the strip can be used or cold rolled
again. Thus, repeated descaling of stainless
steel strip needs to be performed, to remove
mill scale as well as scale resulting from the
annealing process.
With respect to carbon steel, the scale from
stainless steel is more difficult to remove.
Moreover, the hot rolling and annealing
processes lead to the diffusion of chromium
from the upper layer of the base material into
the scale. The resulting chromium depleted
layer on the strip surface needs to be removed
by the pickling process.
The annealing and descaling of stainless steel
are often combined into a single processing
line (Kani et al., 1996). Since the annealing
section governs the strip speed, the pickling
process must guarantee suitable operational
flexibility to avoid unwanted underpickling (i.e.,
incomplete scale removal) or overpickling
phenomena. This requirement led to the
development of specific descaling processes
for stainless steel.
Neutral-electrolytic pickling lines, in particular,
offer some unquestionable advantages in
terms of process controllability and acid
consumption reduction as compared to the old
stainless steel descaling technologies of acid
or fused-salt baths. The self-regenerating
neutral electrolyte (Na2SO4) is inexpensive
and operates over a wide range of pH values
(4.5–7.0). The pH value can be properly
adjusted by the addition of caustic soda, if the
pH is too low, or by the addition of sulphuric
acid, if the pH is too high. Electric current
controls the descaling process: in case of a
plant stoppage, the neutral electrolyte does not
attack the base material. Moreover,
overpickling never occurs, since the descaling
reaction practically stops by itself after the
complete removal of the scale.
385
Danieli Automation
Pickling line modeling for advanced process monitoring and automation
Advances in pickling line automation
Processing lines (pickling and annealing) are
considered simple processes, which do not
present critical control problems. Pickling
lines, in particular, are often managed with
semi-manual operating practices, defined on
the basis of the operators’ experience and
plant-specific knowledge. Nevertheless,
in recent years the field has become quite
competitive, and modern control and
automation solutions are required to achieve
significant improvements in quality and
production, and reductions in consumption
of acid and steam. The increasing interest in
model-based applications is motivated by the
need to provide additional process monitoring
capabilities (see Dornemann, 1997 and Frank
et al., 1999) and to offer valid support to the
operators’ decisions. Some of the main
motivations for improving automation systems
on pickling lines are listed below.
For all the above-mentioned reasons, Danieli
Automation is devoting great efforts to the
development of top-quality automation
software for pickling line management and
control, including model-based applications.
> Modern Turboflo pickling lines (Gruchot et
al., 2001) are characterized by fast dynamics,
which require higher control performance with
respect to traditional push-pull or even lowspeed continuous lines (Frank et al., 1999
and Dornemann and Teichert, 1994).
> Real-time chemical analysis of pickling
baths is usually unavailable. The process
monitoring capabilities of plant automation
systems often reveal extremely useful
information for plant operators (Gohring et al.,
1972).
> Acid regeneration plants usually guarantee
the best performance and efficiency if the
process control systems are able to maintain
the metal concentration in the exhaust acid
solution from the pickling line to near-constant
levels.
> The increasing demand for high-quality steel
products requires that the correct degree of
descaling is maintained by process
automation. Underpickling is responsible for
corrosion damage to steel strip, while
overpickling reduces strip quality and
increases roughness, which leads to different
friction values during cold rolling.
> Only through the support of valid process
control systems is it possible to improve the
plant’s efficiency by reducing the consumption
of steam and fresh acid.
386
3 Pickling stage
schematic
representation.
Architecture of control software
The hierarchical structure of the control
software usually adopted for processing lines,
rolling mills and technological processes in
general is depicted in Figure 2.
Level 1 automation directly interacts with lowlevel actuators and transducers. Real-time
control loops and logic sequences are
implemented here. Fast sampling (1 ms) and
high computing power are achieved through
VME (Versa Module Europa) architecture
technology. Standard PLCs, by comparison,
provide a sample time of about 10 ms. The
Human-Machine Interface (HMI) offers the
operators a real-time look of the process.
Level 2 automation provides higher-level
control functions and utilities, like optimal
plant setup calculation, generation of
production reports and statistical analysis of
product quality. In particular, mathematical
models of technological processes are used to
generate proper plant setups. The reliability of
physical models, in different and even timevarying working conditions, is guaranteed by
the self-adaptation of model parameters,
based on plant feedback. Technological
information and a historical archive of
production are stored into the database (DB),
while the Process Workstation (PWS) offers a
graphic interface to all the Level 2 utilities.
In many cases, Level 3 automation is supplied
too, implementing additional utilities for toplevel production supervision, storage yard
Level 3
Production
planning &
scheduling
Data
Base
Level 2
Setup calc.
Reports gen.
Consumption
Sales dept.
PWS
Process
workstation
management, and coordination of Level 2
automation of different processes belonging to
the same plant. For pickling lines, Level 1
provides real-time control of strip speed and
tensions, acid solution level inside treatment
tanks, temperature of pickling baths and
electric current for electrolytic processes. A
model-based approach to real-time process
control can offer valuable advantages for fast
Turboflo processes (Frank et al., 1999 and
Dornemann and Teichert, 1994). Level 1
automation receives the proper setpoint values
of the above-mentioned variables from Level 2,
depending on the features of the coil to be
processed and on the status of the line.
The support of a mathematical model can
guarantee optimal setup calculation and coilto-coil estimation of bath degradation, even if
real-time chemical analysis of pickling baths
is unavailable. Moreover, optimal tank refilling,
with the consequent saving of fresh or
regenerated acid, is supported.
Pickling lines components and configuration
In this section, a detailed description of basic
pickling line components and structure is
provided. The same structure is a reference
for the process modeling work.
Main components of pickling lines
Pickling lines present a modular structure:
the descaling process takes place in many
equally sized consecutive stages. The typical
pickling stage is shown in Figure 3.
The list of its main components follows:
> In the Working Tank the strip comes into
contact with the acid reactant. The working
tank is continuously refilled with heated acid
solution (sprayed at high pressure in Turboflo
2 Hierarchical structure of
control software.
4 Typical pickling line
configuration.
Working thank
Steel strip
Heat exchanger
(hot steam)
Recirculation
Pump
to the previous
stage
to the ARP
3
from the following
stage
Recirculation
Tank
from the ARP
pickling lines), while the exhaust solution is
drained, and the fluid level is kept constant.
> The Recirculation Tank provides acid
solution to the working tank, and receives from
it the pickling liquor, enriched with metal
content. The recirculation tank can exchange
acid solution with the ARP and/or with the
adjacent stages.
> The Recirculation Pump guarantees a
continuous flow of pickling solution between
the working and recirculation tanks. Turboflo
systems are usually equipped with variablespeed recirculation pumps. This introduces a
fast control action on the acid concentration
inside the working tank, especially useful in
case of fast transients or plant stops.
> Through the Heat Exchanger, the pickling
liquor is heated by steam. Temperature is
regulated through a closed-loop controller
of steam flow.
Pickling lines configuration
The typical configuration of a pickling line is
shown in Figure 4. Assuming the strip
Working thanks
Exit side
Entry side
Recirculation thanks
PLC
Level 1
Real-time
control loops,
Logic sequences
HMI
Human
Machine
Interface
Exaust
acid solution
Acid regeneration
plant (ARP)
Regenerated
acid solution
387
Flat products
Management and control of pickling processes
Danieli Automation
Pickling line modeling for advanced process monitoring and automation
Advances in pickling line automation
Processing lines (pickling and annealing) are
considered simple processes, which do not
present critical control problems. Pickling
lines, in particular, are often managed with
semi-manual operating practices, defined on
the basis of the operators’ experience and
plant-specific knowledge. Nevertheless,
in recent years the field has become quite
competitive, and modern control and
automation solutions are required to achieve
significant improvements in quality and
production, and reductions in consumption
of acid and steam. The increasing interest in
model-based applications is motivated by the
need to provide additional process monitoring
capabilities (see Dornemann, 1997 and Frank
et al., 1999) and to offer valid support to the
operators’ decisions. Some of the main
motivations for improving automation systems
on pickling lines are listed below.
For all the above-mentioned reasons, Danieli
Automation is devoting great efforts to the
development of top-quality automation
software for pickling line management and
control, including model-based applications.
> Modern Turboflo pickling lines (Gruchot et
al., 2001) are characterized by fast dynamics,
which require higher control performance with
respect to traditional push-pull or even lowspeed continuous lines (Frank et al., 1999
and Dornemann and Teichert, 1994).
> Real-time chemical analysis of pickling
baths is usually unavailable. The process
monitoring capabilities of plant automation
systems often reveal extremely useful
information for plant operators (Gohring et al.,
1972).
> Acid regeneration plants usually guarantee
the best performance and efficiency if the
process control systems are able to maintain
the metal concentration in the exhaust acid
solution from the pickling line to near-constant
levels.
> The increasing demand for high-quality steel
products requires that the correct degree of
descaling is maintained by process
automation. Underpickling is responsible for
corrosion damage to steel strip, while
overpickling reduces strip quality and
increases roughness, which leads to different
friction values during cold rolling.
> Only through the support of valid process
control systems is it possible to improve the
plant’s efficiency by reducing the consumption
of steam and fresh acid.
386
3 Pickling stage
schematic
representation.
Architecture of control software
The hierarchical structure of the control
software usually adopted for processing lines,
rolling mills and technological processes in
general is depicted in Figure 2.
Level 1 automation directly interacts with lowlevel actuators and transducers. Real-time
control loops and logic sequences are
implemented here. Fast sampling (1 ms) and
high computing power are achieved through
VME (Versa Module Europa) architecture
technology. Standard PLCs, by comparison,
provide a sample time of about 10 ms. The
Human-Machine Interface (HMI) offers the
operators a real-time look of the process.
Level 2 automation provides higher-level
control functions and utilities, like optimal
plant setup calculation, generation of
production reports and statistical analysis of
product quality. In particular, mathematical
models of technological processes are used to
generate proper plant setups. The reliability of
physical models, in different and even timevarying working conditions, is guaranteed by
the self-adaptation of model parameters,
based on plant feedback. Technological
information and a historical archive of
production are stored into the database (DB),
while the Process Workstation (PWS) offers a
graphic interface to all the Level 2 utilities.
In many cases, Level 3 automation is supplied
too, implementing additional utilities for toplevel production supervision, storage yard
Level 3
Production
planning &
scheduling
Data
Base
Level 2
Setup calc.
Reports gen.
Consumption
Sales dept.
PWS
Process
workstation
management, and coordination of Level 2
automation of different processes belonging to
the same plant. For pickling lines, Level 1
provides real-time control of strip speed and
tensions, acid solution level inside treatment
tanks, temperature of pickling baths and
electric current for electrolytic processes. A
model-based approach to real-time process
control can offer valuable advantages for fast
Turboflo processes (Frank et al., 1999 and
Dornemann and Teichert, 1994). Level 1
automation receives the proper setpoint values
of the above-mentioned variables from Level 2,
depending on the features of the coil to be
processed and on the status of the line.
The support of a mathematical model can
guarantee optimal setup calculation and coilto-coil estimation of bath degradation, even if
real-time chemical analysis of pickling baths
is unavailable. Moreover, optimal tank refilling,
with the consequent saving of fresh or
regenerated acid, is supported.
Pickling lines components and configuration
In this section, a detailed description of basic
pickling line components and structure is
provided. The same structure is a reference
for the process modeling work.
Main components of pickling lines
Pickling lines present a modular structure:
the descaling process takes place in many
equally sized consecutive stages. The typical
pickling stage is shown in Figure 3.
The list of its main components follows:
> In the Working Tank the strip comes into
contact with the acid reactant. The working
tank is continuously refilled with heated acid
solution (sprayed at high pressure in Turboflo
2 Hierarchical structure of
control software.
4 Typical pickling line
configuration.
Working thank
Steel strip
Heat exchanger
(hot steam)
Recirculation
Pump
to the previous
stage
to the ARP
3
from the following
stage
Recirculation
Tank
from the ARP
pickling lines), while the exhaust solution is
drained, and the fluid level is kept constant.
> The Recirculation Tank provides acid
solution to the working tank, and receives from
it the pickling liquor, enriched with metal
content. The recirculation tank can exchange
acid solution with the ARP and/or with the
adjacent stages.
> The Recirculation Pump guarantees a
continuous flow of pickling solution between
the working and recirculation tanks. Turboflo
systems are usually equipped with variablespeed recirculation pumps. This introduces a
fast control action on the acid concentration
inside the working tank, especially useful in
case of fast transients or plant stops.
> Through the Heat Exchanger, the pickling
liquor is heated by steam. Temperature is
regulated through a closed-loop controller
of steam flow.
Pickling lines configuration
The typical configuration of a pickling line is
shown in Figure 4. Assuming the strip
Working thanks
Exit side
Entry side
Recirculation thanks
PLC
Level 1
Real-time
control loops,
Logic sequences
HMI
Human
Machine
Interface
Exaust
acid solution
Acid regeneration
plant (ARP)
Regenerated
acid solution
387
Flat products
Management and control of pickling processes
Danieli Automation
Implementation of a pickling line model in
plant automation systems allows status
monitoring of pickling baths, calculation of
optimal setup and tank refilling. and, finally,
true real-time control of faster Turboflo lines.
The model is based on flow-rate balances of
pickling stages and properly simplified
representations of electro-chemical reactions,
describing both acid and electrolytic pickling
processes (Aurora et al., 2006).
The recirculation tank model represents the
dynamics of both the pickling solution volume
and the concentration of each chemical agent,
as a function of all the input-output solution
flows. Similarly, the working tank dynamics
depends on a flow-rate balance, where mass
flow rates dealing with the descaling chemical
reaction are expressed as a proportional
function of the mass flow of removed scale,
computed by the reaction model. For example,
for pickling lines using hydrochloric acid, the
chemical reaction can be summarized as
The ratio between the consumed hydrochloric
acid HCl and the quantity of removed iron
oxide FeO is given by the ratio of the
corresponding stoichiometric coefficients and
molecular weights. As proposed in Frank et al.
(1999), a proper mathematical description of
the descaling process can be achieved by
introducing a quantity called Average Reaction
Speed (ARS). It is defined as the rate of
removed oxide quantity per surface and time
388
3
2
1
0
30
35
40
45
50
55
60
65
HF = 25 g/l
HF = 35 g/l
unit. ARS can be determined by means of
laboratory tests, and it is also widely used to
express the effectiveness of a line operated at
specified concentrations and temperatures.
An example of model tuning, based on
experimental data, is shown in Figure 5.
Flat products
Pickling line model
5 Example of model
tuning, Average
Reaction Speed (ARS)
vs pickling temperature.
4
Kg/m2/h
direction as a reference, the last stage, directly
refilled by the ARP, is characterized by the
highest concentration of fresh acid, which
guarantees complete scale removal from the
strip. The content of metal is low, because
the strip has been processed already by the
previous stages. In the intermediate stages,
the acid concentration is proportionally lower,
and the metal content is proportionally higher.
The first stage has the largest concentration of
connected acid and metal ions. The exhaust
pickling solution is drained and sent to the ARP.
In electrolytic pickling lines the rectifiers are
positioned inside the working tank. The
descaling process is driven by current,
essentially, so the acid reactant is usually
replaced with a neutral one, with the same
concentration for all the pickling stages.
Pickling line modeling for advanced process monitoring and automation
Model implementation
The pickling line mathematical model has
been integrated into Level 2 applications of
different plants. In particular, for Turboflo lines
using hydrochloric acid, the model is used to
compute the optimal line setup for strip
processing: flow-rate of fresh acid solution
required to the ARP, exhaust pickling liquor to
be drained from the line, bath temperatures
and strip speed.
Conclusions
The introduction of new fast-dynamics
descaling processes like Turboflo asks for
adequate performance from the control
systems. The increasing demand for highquality steel requires that the correct grade
of pickling be guaranteed for all kinds of
products. Finally, the consumption of chemical
reactants must be optimized. All these
requirements suggest the development of
model-based control applications.
The mathematical model of a pickling line
presented in this chapter can be integrated
into high-level control applications, giving
support to the plant operators in process
monitoring, the definition of optimal pickling
practices for each kind of material,
and tank refilling.
389
Danieli Automation
Implementation of a pickling line model in
plant automation systems allows status
monitoring of pickling baths, calculation of
optimal setup and tank refilling. and, finally,
true real-time control of faster Turboflo lines.
The model is based on flow-rate balances of
pickling stages and properly simplified
representations of electro-chemical reactions,
describing both acid and electrolytic pickling
processes (Aurora et al., 2006).
The recirculation tank model represents the
dynamics of both the pickling solution volume
and the concentration of each chemical agent,
as a function of all the input-output solution
flows. Similarly, the working tank dynamics
depends on a flow-rate balance, where mass
flow rates dealing with the descaling chemical
reaction are expressed as a proportional
function of the mass flow of removed scale,
computed by the reaction model. For example,
for pickling lines using hydrochloric acid, the
chemical reaction can be summarized as
The ratio between the consumed hydrochloric
acid HCl and the quantity of removed iron
oxide FeO is given by the ratio of the
corresponding stoichiometric coefficients and
molecular weights. As proposed in Frank et al.
(1999), a proper mathematical description of
the descaling process can be achieved by
introducing a quantity called Average Reaction
Speed (ARS). It is defined as the rate of
removed oxide quantity per surface and time
388
3
2
1
0
30
35
40
45
50
55
60
65
HF = 25 g/l
HF = 35 g/l
unit. ARS can be determined by means of
laboratory tests, and it is also widely used to
express the effectiveness of a line operated at
specified concentrations and temperatures.
An example of model tuning, based on
experimental data, is shown in Figure 5.
Flat products
Pickling line model
5 Example of model
tuning, Average
Reaction Speed (ARS)
vs pickling temperature.
4
Kg/m2/h
direction as a reference, the last stage, directly
refilled by the ARP, is characterized by the
highest concentration of fresh acid, which
guarantees complete scale removal from the
strip. The content of metal is low, because
the strip has been processed already by the
previous stages. In the intermediate stages,
the acid concentration is proportionally lower,
and the metal content is proportionally higher.
The first stage has the largest concentration of
connected acid and metal ions. The exhaust
pickling solution is drained and sent to the ARP.
In electrolytic pickling lines the rectifiers are
positioned inside the working tank. The
descaling process is driven by current,
essentially, so the acid reactant is usually
replaced with a neutral one, with the same
concentration for all the pickling stages.
Pickling line modeling for advanced process monitoring and automation
Model implementation
The pickling line mathematical model has
been integrated into Level 2 applications of
different plants. In particular, for Turboflo lines
using hydrochloric acid, the model is used to
compute the optimal line setup for strip
processing: flow-rate of fresh acid solution
required to the ARP, exhaust pickling liquor to
be drained from the line, bath temperatures
and strip speed.
Conclusions
The introduction of new fast-dynamics
descaling processes like Turboflo asks for
adequate performance from the control
systems. The increasing demand for highquality steel requires that the correct grade
of pickling be guaranteed for all kinds of
products. Finally, the consumption of chemical
reactants must be optimized. All these
requirements suggest the development of
model-based control applications.
The mathematical model of a pickling line
presented in this chapter can be integrated
into high-level control applications, giving
support to the plant operators in process
monitoring, the definition of optimal pickling
practices for each kind of material,
and tank refilling.
389
Worldwide Danieli Companies
GERMANY
FRANCE
USA
CHINA
600 Cranberry Woods Drive
Suite 200
Cranberry Township, PA 16066
Phone +1 724 7785 400
Fax +1 724 7785 401
[email protected]
Jingyuan Street, 8
BDA
Beijing, 100176
Phone (86) 10.58082828
Fax (86) 10.58082929
[email protected]
Poligono Sondikalde
Calle Portu Bidea, 2
48150 Sondica - Vizcaya
Phone +34 944 804 343
Fax +34 944 631 183
[email protected]
USA
CHINA
AUSTRIA
INDIA
Brunshofstrasse, 12
D-45470 Mülheim - Ruhr
Phone +49 208 3780 000
Fax +49 208 3780 200
[email protected]
Les Mercuriales
F-93176 Bagnolet Cedex
Phone +33 149 722 269
Fax +33 149 722 538
[email protected]
UNITED KINGDOM
SPAIN
722 Prince of Wales Road
Sheffield S9 4EU
Phone +44 (0) 114 2800 300
Fax +44 (0) 114 2800 319
[email protected]
SWEDEN
Nya Ågatan, 23
S-77782 Smedjebacken
Phone +46 240 668 500
Fax +46 240 668 501
[email protected]
Max Planck Strasse, 5
A - 9100 Völkermarkt
Phone +43 4232 51440 6101
Fax +43 4232 51440 6105
[email protected]
9800 Southern Pine Boulevard
Suite L
Charlotte, NC 28273-5522
Phone +1 704 5528 024
Fax +1 704 5220 264
Technopolis Building
Plot 4, Block - BP, 5th Floor
Wing - B, Sector - V, Salt Lake
700 091 Kolkata - West Bengal
Phone +91 33 3984 7777
Fax +91 33 3984 7501
[email protected]
No.19, Xing Gang Road
CEDZ
Changshu, Jiangsu 215513
Phone (86) 512.52267088
Fax (86) 512.52267223
[email protected]
THAILAND
Land Plot N. K11
The Eastern Seaboard Ind. Estate
Tambol Puakdaeng, Amphur
Pluakdaeng, 21140 Rayong
Phone (66) 38.955888
Fax (66) 38.959096
[email protected]
Worldwide Offices and Service Centers
CANADA
headquarters
DANIELI
Via Nazionale, 41
33042 Buttrio (UD) Italy
TEL +39 0432 1958 111
FAX+39 0432 1958 289
www.danieli.com
[email protected]
DANIELI AUTOMATION
Via B. Stringher, 4
33042 Buttrio (UD) Italy
TEL +39 0432 518 111
FAX+39 0432 673 177
www.dca.it
[email protected]
4210 South Service Road
Burlington
Ontario L7L 4X5
Phone (1) 905.3049337
Fax (1) 905.3047975
[email protected]
BRAZIL
Av. Ibirapuera, 2332
Cj 121 – 12th floor
CEP 04029 200 São Paolo
Phone (55) 11.35085900
Fax (55) 11.35085924
[email protected]
MEXICO
Edificio Sierra Madre
Ave. Vasconcelos Oriente 310
Colonia del Valle
66250 Garza Garcia, N.L.
Phone (52) 81.83781055
Fax (52) 81.83781058
[email protected]
EGYPT
KSA
Silver Tower 6th floor
P.O. BOX 4867
Al-Khobar 31952
Phone (966) 3.8993145
Fax (966) 3.8993198
[email protected]
RUSSIA
Kutuzoysky Prospekt 12
Building 2, Floor 5
121248 Moscow
Phone (7) 495.6612636
Fax (7) 495.6612637
[email protected]
RUSSIA
Ul. Kominterna, 16
Office 406
620078 Ekaterinburg
Tel (7) 343.3101334
Fax (7) 343.3101335
[email protected]
UKRAINE
Mohamed Farid Street, 37
11351 Heliopolis West-Cairo
Phone (20) 2.26379229
Fax (20) 2.26379525
[email protected]
Most Citi Business Center
Glinky Street, 2, 3rd Floor
49000 Dnepropetrovsk
Tel (38) 056.7904301
Fax (38) 056.7904304
[email protected]
UAE
INDIA
Prestige Car Bldg.
Flat 1604 – Khalidiya
P.O. Box 127231
Abu Dhabi
Phone (971) 2.6812268
Fax (971) 2.6813368
[email protected]
271 Business Park, 5th floor
Model Industrial Colony
Off Aarey road
Goregaon (E) Mumbai 400 063
Phone (91) 22.399117100
Fax (91) 22.39917191
[email protected]
THAILAND
Muang Thai-Phatra Office
16th Floor, Tower II
252-91 Rachadaphisek Road
Huaykwang – Bangkok 10310
Phone (66)2.6933520
Fax (66) 2.6933529
[email protected]
TAIWAN
26F-1, No. 31
Hai-Bian Road
Kaohsiung City, Taiwan 802
Phone (886) 7.3358655
Fax (886) 7.3358755
[email protected]
KOREA
Room 201, Daecheon Building
157-4 Samsung-dong
Gangnam-gu 135-090, Seoul
Tel (82) 2.5626622
Fax (82) 2.5623639
[email protected]
JAPAN
42F, Yokohama Landmark Tower
2-2-1, Minatomirai, Nishi-ku,
Yokohama-City
220-8142 Japan
Tel (81) 45.651.7077
Fax (81) 45.651.7033
[email protected]