Neil Freeman

Technical Specialist Mineral Processing

Honeywell Hi-Spec Solutions

Joe Spirito

Australian Manager Mining Minerals and Metals

Honeywell Limited


The production of alumina is a highly intensive process involving the interactions of numerous processes. The ability to make this process economically viable relies on good process design together with an understanding of the process dynamics. The present economic climate requires that alumina producers are able to make good quality alumina at the right price and in a timely manner. Best in class alumina producers are aware of this and embrace every opportunity to gain an advantage. Optimisation of the business can provide a significant profit improvement to those companies willing to embrace the new technologies.

This paper investigates the aspects of business optimisation available within the process control domain (ie process performance, asset management and unified information systems). Examples are provided to highlight the benefits to be derived.


Neil Freeman and Joe Spirito


More than any single measure, strong corporate financial performance separates industry leaders from all other companies in today's highly competitive environment. And stock prices reflect this leadership position. As might be expected, the industry leaders have more market share, better growth opportunities, and world-class organisations. Their healthy financial performance is derived from a solid business foundation.

These successful business managers have learned to balance the need for profit to improve income statement results with the need to preserve assets for sustained balance sheet performance, while integrating corporate objectives with plant operations. This ongoing process is known as business optimisation.

Business optimisation is a paradigm shift for most organisations — away from simply simulation, control, and optimisation of one or more process units at an operating site — to the overall performance and contribution of multiple process units at multiple sites within the total enterprise. As the accompanying illustration shows, a business optimisation approach should balances the month-to-month demand for maximum profit with the longer-term need for sustained asset capability, while meeting corporate objectives.


1.1 More Profit to Improve Income Statement Results

Process Performance Solutions help meet ongoing profit objectives through control and optimisation of process units throughout the enterprise. Solutions are linked to site and corporate targets in real time to optimise overall process operations. These solutions entail implementing applications that are complementary and unified to work together effectively.

Financial benefits from these advanced applications can be as much as 10 percent of feedstock value. Payback on investment is normally measured in months rather than years. Because of the rapid return, these solutions are typically the most attractive projects available at a site.

1.2 Preserving Valuable Assets for Sustained Balance Sheet Performance

Asset Capability Solutions deliver sustained plant capabilities through automated monitoring and testing of process equipment and process operations. Early detection and alert of individual problem areas are combined with plant operating data to determine probable causes and recommended actions to sustain optimal asset performance.

Financial benefits can be substantial. For example, avoidance of process incidents and shutdowns can save a company millions of dollars annually. It is estimated that in the hydrocarbon processing industry (HPI), in just the U.S., companies experience as much as $10 billion in lost profits each year from such causes. The solution to this should incorporate a total solution addressing the needs of all the assets - people, equipment and process. This will enable the process performance gains to be fully achieved.

1.3 Integration with Corporate Objectives, Availability of Plant-wide Data

Unified Information Solutions help ensure that corporate objectives are an integral part of the operational decisions and planning for each plant and throughout the enterprise. In order to do this the solutions must make the wealth of plant-wide data available as relevant, consistent information for use by plant and management staff to improve decision making and operational effectiveness. The ideal solution being a series of interoperable applications that distil data into the information needed for optimum management of operating units within a plant. In addition provision should be made for real-time information exchange with enterprise resource planning systems such as SAP, allowing the ongoing exchange of information between plant operations and business systems.

Improvement in plant operations results in better utilisation of invested capital and increased return for shareholders. Companies in a variety of process industries — hydrocarbon processing, chemicals, pulp and paper, power generation, metals processing — are typically able to increase production by 2 to 5 percent and decrease costs by 0.5 to 1 percent.


The challenge to any alumina refinery is to minimise the cost of production per tonne of alumina consistent with safety and environmental considerations. This translates to maximising the production of alumina (plant flow and yield) and minimising the energy costs per tonne of alumina. For the liquor circuit this equates to maximising the liquor flow whilst maintaining liquor concentration. In addition liquor inventories and surge volume must be maintained within high and low limits. Similarly mud washer dilution must be carefully controlled to maximise caustic recovery whilst minimising the dilution.

The Bayer circuit poses unique problems for control. Recovery of energy and caustic are what makes the Bayer Process economically viable. This causes the unit processes to be highly interactive. Combined with long dead times this causes problems for conventional control. Advanced control techniques in the form of multivariable predictive control, such as the Honeywell Hi-Spec Solution’s RMPCT, solve these problems. The robust nature of this controller handles model incongruencies experienced as a result of process variations such as scale build up or feed stock variations.

This type of control considers an entire process at a time, such as digestion or heat exchange. In this way the control objectives become those of the process. The process is kept within operational limits by the manipulation of variables that are not at their limits. Similarly the process is optimised by pushing the process to the operational constraints.

RMPCT provides responsive control and optimisation of processes even when there are large interactions between controlled variables. The controller incorporates a model of the process dynamics that it uses to predict future behaviour of the process and to determine how to adjust the controller’s outputs in order to bring all process variables to setpoints or within limits. If there are any degrees of freedom remaining, the controller adjusts the process to optimise operations, for example by maximising total product value.

RMPCT has the unique ability to maintain good control of a process with highly interacting variables even though the model of the process may have significant errors.

RMPCT can be viewed as a tool to keep interacting process variables at specified setpoints. But often the real value of RMPCT is that it lets you consider an entire process as a single entity rather than as a collection of isolated control loops. RMPCT then becomes a tool to keep the process within operational constraints while optionally optimising some performance measure.

There are a number of areas within a typical alumina refinery that are well suited to the application of advanced process control. Significant benefits can be derived from these by continuously pushing the various process limits. If a 1 million tonne per annum alumina refinery is considered then approximate benefits of the application of advanced control can be determined as below:

Area Benefits Benefits Description $ per annum





Power saving





Yield improvement




Digester yield


Heat Exchange


Precipitator fill temperature




Energy saving


Mud Washing


Caustic loss


Liquor Stock Control


Overall circuit throughput





    1. 1.Milling throughput assumes that the main Bayer circuit can take more bauxite
    2. Assuming 6 MW power at a cost of 3cents per kW/hr
    3. Yield improvement in precipitation due to tighter alumina to caustic ratio control.
    4. Assumes that undigested bauxite is disposed of to mud lakes
    5. 1 Degree C precipitator fill temperature improvement assumed to represent 1% yield improvement
    6. Assume energy consumed in calcination is 4 GJ per tonne at a cost of 3 cents per GJ.
    7. Assume caustic loss to mud lakes is 200 g per tonne at a cost of $200 per tonne
    8. Assumes improvement in digester throughput is production and value of alumina is $200 per tonne.
    9. Cost of production assumed to be $150 per tonne

The cost associated with advanced control is very effective when compared to capital equipment costs. For instance a new 1 million tonne per annum refinery might cost of the order of 1 billion dollars. This equates to 10 million dollars per percent of production. Advanced control techniques can generally achieve this same increase for less than 1 million dollars making it an attractive investment.


The assets of an organisation (people, equipment and the process) are key to the continued successful exploitation of the resource. Irrespective of the improvements gained from advanced control these can be wasted if the assets are not correctly managed. The assets are considered to be:


Operations personnel are the key to successful operations of the plant


Equipment reliability enables full advantage to be made of process improvement solutions.


Abnormal situations can result in huge losses of potential revenue notwithstanding potential safety issues.

Successful use of the assets provides the assurance that the plant is operated safely and profitably.

3.1 People

The major asset that any organisation has is the people. It is the people and the way that they operate the plant that is key to the profitability of the organisation. In order to fully utilise the operations personnel it is vital that they be adequately trained. The training requirements are really three-fold:

3.1.1 Control System

Operations personnel need to be totally familiar with the control system and how to use it. This should be considered a vital tool in the execution of their duties, which successfully deployed can alleviate considerable effort.

3.1.2 Process

An understanding of the process and chemistry will enable the operators to make correct decisions relating to the optimal operation of the plant.

3.1.3 Controlling the Process

To correctly operate it is necessary to combine the knowledge of the process with knowledge of the control system and to understand the dynamics of the operation. This will enable the avoidance of possible abnormal situations as well as provide an ability to optimise the operation under varying conditions.

The only effective means of training in all three aspects is through the use of a direct connect dynamic simulation which replicates the environment in which the operator works. This being analogous to the flight simulators which are an integral part of pilot training.

One such dynamic simulator is the Honeywell Trainer, which is a state-of-the-art dynamic process simulator designed specifically for the process industries. It is the most technologically advanced simulator on the market today, and has a wide range of applications that make it an extremely useful tool throughout the life cycle of a plant, from the design stage to full operation. Trainer provides a view of the changing state of a process in operation by calculating future states of a process as it reacts to changing conditions. As these changes occur throughout the system, they can be observed and tracked, providing the "look and feel" of real-world operations.

Trainer comprises of simulation software, models of the process, controls and logic, training features and various hardware components. The software and models reside in a desktop computer connected to an engineer’s or instructors console, and most importantly, to a console of a distributed control system operator. Users interact with the Trainer System through intuitive graphical user interfaces that feature tools and utilities specific to the task at hand.

The Trainer model is built from first principals based on rigorous chemical and thermodynamic principles. The simulation engine is a modular program used to perform real-time dynamic simulation calculations of a process and control system. It solves a set of algebraic and differential equations and calculates the state of a process and control system at a defined time frame into the future.

Through the use of Trainer applied to a new process constructed within an existing alumina refinery the company reported benefits in the following areas:

3.1.4 Operator Training

The overall result of this being an improvement in start-up time of 25% with a resultant reduction in off specification of 50%. This translated to many millions of dollars in revenue improvement providing a payback in months.

3.2 Equipment

The Bayer process provides a harsh environment in which equipment is required to operate. Hot caustic solutions containing abrasive suspended and dissolved solids create problems either with wear or precipitation. This means that there is a continual requirement to monitor the health of equipment such as valves, pumps and heat exchangers. Optimally maintaining equipment will result in not only an improvement in the operations of the plant but also a reduction in the cost of maintenance. In order to do this the plant maintenance engineers require tools to enable them to focus their efforts. The Honeywell Equipment Health Management (EHM) is one such tool set.

EHM is a set of solutions that focus on plant maintenance and reliability. EHM balances reliability, predictive, and preventive maintenance strategies, allowing the most effective use of plant assets. As an open solution, EHM gathers data from any specified plant asset for analysis and action. Typical plant assets include heat exchangers, control valves, compressors, pumps, motors, analysers, instruments and pressure relief valves. Reliability and maintenance engineers are provided with a consistent desktop environment that gathers data from a variety of sources, such as engineering documents, plant historians, maintenance records, field sensors, and control systems.

There is the need to collect and transform thousands of bits of data. DocuMint provides a basis for automating the field instrumentation inspection, testing and simplifying field data collection. Whilst the Alert Manager transforms thousands of data points - collected from plant control systems, historians, sensor networks and other plant databases - into symptom-based fault isolation. Unique in its ability to understand symptoms and faults, Alert Manager provides directed troubleshooting and gathers the associated process, mechanical and maintenance data necessary to analyse problems.

In addition the EHM Virtual Testing continuously monitors the health of equipment such as control valves and heat exchangers. For heat exchangers, special algorithms allow, for the first time, the measurement and display of fouling and sensor drift.

A complete EHM solutions includes:

3.3 Process

It has been estimated that in the USA alone over $20 billion are lost each year in the process industries through abnormal situations. Generally it is perceived that production is only lost when there is an alarm, however the diagram below illustrates that this is merely the tip of the iceberg.

It can be seen that there is a huge potential for increased profitability through adequate management of abnormal situations.

One major developmental program currently is the Abnormal Situation Management (ASM) Joint Research and Development Consortium. The ASM Consortium, led by Honeywell, was formed to enhance the ability of plant operations personnel to control abnormal situations. It is comprised of seven of the largest refiners in the world, along with several academic partners, and six Honeywell organisations.

The goal of the ASM Consortium is to demonstrate the technical feasibility of reducing the economic impact of abnormal situations by a factor of ten. This will be done by developing technologies that enable better methods for preventing them, for informing operations personnel more rapidly when they do occur, and for helping minimise their economic, safety and environmental impact.

3.4 Integrated Information Systems

In order to compete successfully in the global marketplace, manufacturers must be able to instantly understand the consequences of each business option in order to best manage the relationship between supply and demand. This requires companies to rethink all aspects of their business, including how resources are tracked and managed, and how products are taken to market. Cost savings realised directly impact the bottom line.

However in order to fully realise the benefits companies are integrating the production environment with higher level business systems such as SAP R/3. The benefits of this level of integration are many, including faster and more accurate planning, reduced inventories, accurate production cost tracking, accurate material tracking and genealogy, and better utilisation of resources and personnel.

The challenge lies in integrating the disparate systems currently used to run the business. Control, maintenance, laboratory, accounting and marketing all have their own systems, often with separate technologies and methods for data storage, data entry, and reporting. In addition, the detailed real-time information contained in the control system must be transformed into relevant transaction-based information that is meaningful to a business system. An in-depth understanding of the information flows and timing requirements is essential for a smooth integration of the production environment into the business system.

The requirement is for a "common data repository" approach to storing and managing all of the different types of data throughout the plant environment. This provides a rich environment to store, manage, and relate the many pieces of data required for both plant-level and business-level decision support. Once the many disparate sources of information are collated, a modelled view of the manufacturing facility can be used to apply structure to the data stored in the system. This plant data model presents a single view of the plant to the enterprise system, providing the foundation for integrating plant-level data with the enterprise system.

One such production management system is the Honeywell Uniformance product. Uniformance establishes a single, open, unified architecture for plant-wide data collection, storage, and management. The common data infrastructure supports and unifies all plant data, business data, applications, and procedures. Uniformance promotes consistent use of applications in each of the three solution areas, provides for their effective co-operation, and simplifies the implementation effort and ongoing software management.

In addition there are a rich suite of production management applications to provide for the day to day requirements of plant personnel.

Uniformance provides a unified architecture for data collection, storage, and management, as well as a common execution environment. This unique architecture combines the capabilities of a real-time data historian with a relational database for transactional data. The flexible Uniformance system has an inherent capability to manage real-time data, transaction data, reconciled data, planned data, model data, and many other types of data. Through the application of the system data model, structure and context is applied to the data to allow the enterprise-to-plant system mapping to occur.

The TPS Uniformance system offers a "common data repository" approach to storing and managing all of the different types of data throughout the plant environment. The powerful combination of a world-class historian and its associated relational repository provide a rich environment to store, manage, and relate the many pieces of data required for both plant-level and business-level decision support. This data repository approach allows the Uniformance system to support both the analytical requirements of local plant personnel and business system integration requirements. This integration infrastructure provides the "glue" that binds the disparate plant data together. This infrastructure helps improve the efficiency of local plant personnel as well as facilitate integration with business systems such as SAP R/3.

Integration projects incorporate information from implementation of the SAP business system, Distributed Control System Production Tracking System, Laboratory Information Management System (LIMS) , PLC networks , etc . The project is designed to provide Business Information down to the Plant control level as well as real-time information up to the SAP business system, unifying the control and business domains. Typical integration of information is from the SAP PP-PI, QM and PM modules with MM, HR also providing useful information exchanger. The information transfer involves the extraction of financial information from other modules within SAP via the above Plantcentric SAP modules (PP-PI, PM, QM, HR, MM).

The integrated system ensures that;


There are considered to be three main areas for business optimisation:

Each of these can add considerably to the overall profitability of an alumina operation. However, business optimisation is a paradigm shift for most alumina producers — away from simply operation of the plant — to the overall performance optimisation and contribution of a multitude of equipment and process units within the total enterprise.

Some alumina refineries are starting to take advantage of the benefits that business optimisation can contribute. In general the benefits to be derived provide for huge paybacks with returns being measured in months rather than the years required for capital project paybacks. Business optimisation can provide benefits exceeding $5 per tonne, significantly improving the profitability of organisations.