ECS/FuzzyExpert - CementMillOptimisation General Description

FLS Automation
ECS/FuzzyExpert - CementMillOptimisation
- a dedicated high-level control system for optimisation of power EED
T
consumption and productivity AN
R
A
GU

What is wrong? When to act?

A Typical operator interface of an ECS/Fuzzy CementMillOptimizer contains the most useful information
needed for the daily operation control of the mill.

Change-over control Application Features

During production of the different types of cement R Abnormal Situation Management
the control packet will store data for average fresh R Stabilization
feed, separator speed, recirculation factor etc. for R Optimization What to do?
the periods which represent good mill operation. R So3 Control Act !
These learned data will be used as starting points R Fineness Control
when production changes from one type to another. R Recipe Control
R Recipe selection and download to the
In addition to the learned values, the system holds process
default settings for the different types of cement. R Production change-over control
The system may use the default values for the new R Measurement validation
type of cement, or the system may be configured to
use the learned values. In other words, the Software Features
FuzzyExpert-CementMillOptimisation system R Graphical control strategy design
maintain two sets of recipes, i.e. user specified data R Pre/defined building-blocks for fast
in one set and automatically learned data in the configuration purpose
second recipe system. R ECS/OpStation-based operator interface
R Advanced trend and report generation
The change-over control strategy together with the R OPC interface support
customized recipe-selection mimics enables the R Communication capabilities to PLC systems
operators an easy, safe, and controlled initiation of a from the major suppliers
new bath production.

FLS Automation
a Division of F.L. Smidth

Høffdingsvej 34 Telephone +45 36 18 27 00

DK-2500 Valby Telefax +45 36 18 27 99 ECS/FuzzyExpert - Cement Mill Optimisation Product Profile
Copenhagen e-mail [email protected] rev January 2003/1 General Description
Denmark www.flsautomation.com
ECS/FuzzyExpert - CementMillOptimisation General Description

Application new as well as old cement mills, and the configuration can
The optimisation of cement mill processes has gained an be completed in a few months from project start.
increased attention in the recent years. The optimisation Depending on the specific plant, the minimum guaranteed Obviously, a typical operation area of a process
controlled by human operator is a based on a P2
involves both the environmental aspects as well as benefits are normally a combination of:
“conservative” control strategy to avoid production Optimum operation point
productivity aspects. Reduction of the specific energy
interrupts. However, when the optimum operation
consuption, increased productivity with improved • 3% increase in production. point is at the process limits, production benefit are Limits
production quality are typical examples of these • 4% reduction in specific electrical energy not maximized by manual control. The Operator Control
requirements. ECS/FuzzyExpert - Cement Mill consumption. ECS/FuzzyExpert system aims at forcing the ECS/FuzzyExpert
Optimisation is a control package to meet these • Minimum changeover time between different types of operation trends towards the optimum point.
requirements. cement P1
• Improved operation stability
The control strategy combines the best from fuzzy logic, • Payback time within 1 year.
hierarchical control organisation and model based
technologies. The packet consists of standard building Determination of the above guarantee figures are based on
Quality control Control actions are executed on either the separator speed
blocks, which easily can be adapted to a given cement mill qualified analysis of plant data in order to conclude realistic
Quality control includes control of cement fineness based or the fresh feed. The control strategy may be configured
installation. The optimisation packet can be installed in estimates of the potential improvements. to work with either the separator speed or the fresh feed, or
on either manual sampling every hour or every two hours,
or it may use on-line sampling which provides a it may be setup to work with a combination of the two.
measurement every 10 minutes, or more frequently. The
amount of gypsum will be controlled on the basis of a Fineness control in the ECS/FuzzyExpert-
Production control CementMillOptimisation scheme can reduce the daily
regular SO3 analysis, and other additives like fly ash and
The production control of a cement mill can be subdivided into three main categories (See the figure below). standard deviation on fineness by up to 35%
limestone may also be proportioned based on quality
Optimization analysis inputs.
SO3 control
The top level of the production control pyramid includes a learning feature, which will drive the mill towards the optimal
recirculation factor. An index will be calculated for the filling degree based on mill sound, elevator load etc., and optimal Fineness control The basic control of the percentage of gypsum is a simple
operation will be located within the constraints of the maximum mill filling degree. For the different types of cement Fineness control is based on statistical quality control proportioning calculation including a dead band. In
including different sets of constraints, the control strategy will learn optimal target values and set-points. The learned (SQC) methods. Blaine measurements, Surface Area, and addition, the control strategy considers changes in SO3 in
values will be saved and used as starting points when production of the particular type of cement starts again. Constantly sieve residues can be used. The basic SQC methods combination with moving average values to make the SO3
driving the mill towards optimal targets and set-points will increase production and reduce the specific electrical energy applies the newest measurement of the fineness, and the content as constant as possible. Also for So3, the daily
consumption. calculations result in adjustments, which should bring the standard deviation will be reduced by 35% or more as a
next analysis to either the set-point or within an certain result from applying the new control scheme.
interval around the set-point. FLSA has extended the basic
SQC control scheme to include the trends of the analysis
Stabilization in combination with validation rules for the fineness
For the amount of cement and for the measurements.
Optimisation cement quality it is crucial to run the
mill as stable as possible. Stable
operation is produced by the control
objectives in the center of the
production control hierarchy. The
stability control objectives can be
setup to work with both separator
speed and fresh feed in combination Mill Inlet
MillFlow
Flow Inlet
without reducing the cement quality. Low overfill
overfill22
Low
More stable operation also results in a
major improvement of the cement
quality. Outlet
Outlet Feed Optimization Max
Feed Optimization Max
Blockage Stabilisation of
Blockage Stabilisation ofFineness
of
offineness
Fineness
fineness Throughput
Throughput
Production optimisation can not be achieved before stable operation
conditions are established. Normalisation of the process is therefore the
Inlet
Inlet Product Mill
Mill SO3
SO3
first step towards operation stability. ECS/FuzzyExpert is designed to Product
overfill
overfill11 Change Cooling
Cooling Control
Control
follow this principle. Change

Abnormal Situation Management A production and quality control scheme in ECS/FuzzyExpert is highly customized for every cement mill plant.However,
The production control hierarchy is shown in the figuer below. The bottom level consists of the so-called Upset Control the achievement of the various objectives are in accordance with the intuitive interpretation of a practical control
Objectives, which are specialized control schemes for control of the abnormal situations. The purpose is to implement philosophy: Keep the process in stable conditions which are necessary for optimisation tasks. In emergency conditions,
dedicated control algorithms that will bring the mill back to normal operation as fast and safe as possible. If for instance the ignore the quality targets and save the process by urgent actions... This means that the available actuator, e.g. Fresh Feed
mill starts to overfill, then the production control strategy will switch automatically to a control mode that will bring the may be used in different control strategy to achieve a certain result for the given situation. The necessary co-ordination
mill back to normal filling degree with only minor disturbance of the stability and the cement quality. between the production and quality task is carried out by the so/called Priority Manager in ECS/FuuzyExpert
applications. The priority/based structuring of the objectives improves the user’s understandings for future maintenance of
the implemented control strategy.