Loesche is launching its new LM-Master for advanced vertical roller mill (VRM) control. This automation solution stabilises the VRM to allow operation with lower vibration. It also optimises mill performance by making accurate control actions several times per minute and offers continuous monitoring of the milling circuit for improved operational visibility.

1. FOR VERTICAL
ROLLER MILLS by Matthias Authenrieth, Thomas Hyttrek and
Andreas Reintke, Loesche GmbH, Germany,
and Steven McGarel, Senior Consultant, USA
LMmaster_forVRMs.indd 1 19.06.12 11:46

2. PROCESS CONTROL
I LM-Master for VRMs
by Matthias Authenrieth, Loesche is launching its new LM-Master for advanced vertical roller mill
Thomas Hyttrek and (VRM) control. This automation solution stabilises the VRM to allow
Andreas Reintke, operation with lower vibration. It also optimises mill performance by
Loesche GmbH, Germany, making accurate control actions several times per minute and offers
and Steven McGarel, continuous monitoring of the milling circuit for improved operational
Senior Consultant, USA visibility.
V
RMs have become the grinding
circuit of choice in cement plants
for raw meal, clinker and coal
grinding due to their lower capital cost
and energy consumption compared to
ball mills. Loesche GmbH, Germany, has
been a leading supplier of roller mills for
over 100 years and in more recent times,
has expanded by adding a subsidiary
automation company for both hardware
and software delivery and project
management.
The latest addition to the automation
software portfolio, which already includes
mill simulator and SCADA solutions,
Figure 1: reducing process
is an advanced mill controller called variability improves performance
LM-Master. This provides fully-automatic
control and real-time optimisation of
the mill process through precise control allow for some margin of error due to requirements for production rate, product
of mill behaviour according to defined mill process variations, eg, from feed consistency and energy consumption, the
objectives of throughput, energy materials. The development of the optimal combination of set points can
efficiency and mill availability. The LM-Master solution reverses this situation be calculated on a high-frequency cycle
software platform is installed on a PC and by monitoring and controlling more to drive the process to the economic
interfaces with common PLC systems. frequently than any operator and allows optimum continuously.
With high controller availability the mill for operation at higher capacity and levels
is continuously operated close to physical of stability than can be achieved by an Reduction of vibration levels
limits yet also with improved stability and operator alone (see Figure 1). Roller mills are constructed of major
reduced vibration levels, ensuring high mill mechanical components that can be
efficiencies for low capital expenditure. easily damaged by high mill vibrations. It
New-generation optimisers such
is very important that a mill is operated
LM-Master mill optimiser as LM-Master use a dynamic in a stable manner and that corrective
Loesche has unique knowledge of all model of the actual mill, resulting actions are taken to prevent damage
aspects of roller mill design, operation and in improved regulatory control to mechanical equipment and mill
automation, allowing the development of and set points closer to the outages that result in lost production.
detailed and accurate mill control models. Unfortunately, high-vibration scenarios
physical limits of the process.
These have been incorporated into an arise very rapidly and a mill may trip
advanced control software platform which before an operator even has time to
allows the mill control functionality to be The control solution monitors multiple assess the situation and take corrective
configured and held in-house to protect process variables and calculates new set action. LM-Master is valuable in these
proprietary knowledge and intellectual points several times a minute to maintain situations in that, not only does a mill
property. mill performance at very high levels. operate at generally lower vibration levels
The kiln is the equipment that demands Operating at an economic optimum under LM-Master control – reducing
the most attention from operators in a is a challenging control problem since wear on gearboxes and other major
cement plant, hence mills are generally this constantly varies. However, by components – but the continuous
operated below optimum capacity to specifying the overall – and competing – monitoring of the mill process allows for
INTERNATIONAL CEMENT REVIEW MAY 2012
LMmaster_forVRMs.indd 2 19.06.12 11:46

3. PROCESS CONTROL
timely, accurate and rapid
control actions to address
issues with high vibration
and prevent mill trips.
Requirements of
VRM operation
Safe, reliable and productive
use of a mill plant requires
consideration of the
following:
• high production rates in
the presence of potential
limitations in:
– mill table power
– mill differential pressure via
the mill fan
– external recirculation
system via bucket elevator or
other conveyors
– drying capacity via mill exit
temperature control Figure 2: control structure diagram
• low product size variability:
– balance throughput with stability of mill regulatory and operational safety as Control (MPC) which calculates the
– adjust for changes in feed hardness to primary objective. Without coupling of optimum combination of set points to
promote consistent grinding the dynamic relationship between loops, meet multiple objectives of throughput,
• low specific power consumption: holistic optimisation across the process is product quality and process limits
– reduce variability in the load parameters basically impossible. In practice, loops are simultaneously. The structure of the
(mill table power, mill differential pressure, de-tuned to maintain process stability and control scheme is shown in Figure 2.
mill fan power) do not control the process very effectively. In addition to a cross-coupled
– reduce specific power (mill fan and mill In VRMs in particular, highly-interactive multivariable control structure, the
table). variables and rapid process dynamics technology has an inherent capability for
make PID control difficult while variations disturbance rejection which is desirable in
Control technology in the hardness of mill feed represent a grinding circuit control, due to the varying
and structure major disturbance to the circuit. nature of feed streams.
Traditional PID single input/single output New-generation optimisers such as The main features of MPC technology
control has some inherent disadvantages LM-Master use a dynamic model of are:
in that each loop operates where a single the actual mill, resulting in improved • a dynamic model as opposed to a
variable set point is compared to the regulatory control and set points closer to steady-state model
controlled variable. Multiple PID loops the physical limits of the process. • true closed-loop control with built-in
are used on a mill circuit, yet these are The technology selected for the control optimisation
connected as stand-alone loops with solution is based on Model Predictive • mathematical basis allows optimal
solving for control actions in real time
Figure 3: system architecture: LM-Master connected to the control network • computationally efficient, allowing
controllers to run at high frequency
• predictive capability as opposed to
reacting to process conditions.
The controller will take advantage of
stable conditions to increase production,
while under upset conditions – such
as high vibration – the controller will
simultaneously act to reduce vibration,
protect process stability, reduce and then
restore production while maintaining
product quality.
Since the controller solves the
optimisation problem online, updates
can be made to the objective function
(to change the operational behaviour of
MAY 2012 INTERNATIONAL CEMENT REVIEW
LMmaster_forVRMs.indd 3 19.06.12 11:46

4. PROCESS CONTROL
LM-Master-based
the process), to the model (to change an interesting challenge in demonstrating
parameter values), and to process limits in how a software-based solution could
optimisation
response to external factors, eg tyre wear deliver additional performance benefits
in a VRM. over a mechanically-sound, well-operated
advantages
These changes are made with the mill.
controller running, so the effect of the The project was structured with discrete
tuning changes can be evaluated in real steps:
time. • historical data was analysed to • Fast process dynamics are
Many of the challenges in designing determine mill baseline performance controlled by a control solution
controllers which can handle multiple • important control variables were with high frequency.
uncertainties are embedded in this online identified for the mill • Multiple inputs and outputs
optimisation routine. This results in • process tests of the main variables were are tightly controlled over the
controllers which are robust over time conducted to determine mill response complete mill circuit.
and which require less maintenance post- • controller was constructed and • Competing objectives of
commissioning. connected to the PLC control network throughput, quality and energy
The controller resides on a desktop or • calculated control actions were initially consumption are balanced
rack server and uses an OPC interface checked in Read-Only mode optimally.
to connect to the control network. • variables were activated in Read-Write • Continuous display of control
When online, LM-Master writes set mode one by one and tuned performance aids the operators.
points directly to the PLC and out to the • multiple variables were activated in • Stable operation reduces mill
individual field devices. The architecture is Read-Write mode and interactions tuned vibration problems and allows
shown in Figure 3. • run time was accumulated in full closed- operators to focus on kiln
A browser-based interface allows loop control to fine tune. operation.
multiple users to access the system and By integrating the customer’s operators
monitor performance. An operator log-on into the process and promoting close
screenshot is shown in Figure 4. teamwork, buy-in of the concept was cent, from 505tph to 528tph
achieved and ensured an excellent final • Specific power consumption
Project case study solution. was reduced almost five per cent
The development and testing of The development and implementation • Mill vibration was reduced by 17 per
the LM-Master control solution was was achieved in 2-3 calendar months and cent
conducted on a 500tph mill. This was the overall schedule, including planning • Controller utilisation was greater than
considered to be already operating well by and post-project evaluation of results was 90 per cent
the customer and therefore represented approximately four months as shown in • Operators are freed up to better
Figure 5. monitor kiln performance
• Reduced CO2 emissions from the plant
Project results (produces its own electric power)
The project showed • Project Return on Investment was less
impressive results, than 12 months.
particularly given the
customer’s assertion Conclusion
that the mill was already The specially-developed roller mill
operating well prior controller LM-Master has been
to the LM-Master demonstrated to improve the operating
implementation. environment for control room operators
• Production rate was and increase mill performance. The
increased almost five per solution leads to better utilisation of
Figure 4: operator log-on screen
personnel by automating the routine
control of roller mills and freeing
operators to focus on kiln operation.
On the roller mill, the solution delivers
higher production rates, lower specific
power consumption, more stable mill
operation leading to improved product
consistency and lower vibration levels.
This can have additional benefits such
as avoiding vibration trips with a loss
of production time and reducing wear
and costly damage to major mechanical
Figure 5: project Implementation schedule components such as gearboxes. ____ I
INTERNATIONAL CEMENT REVIEW MAY 2012
LMmaster_forVRMs.indd 4 19.06.12 11:46