David Heath

Worsley Alumina Pty Ltd

Gastaldo Road, Collie, WA 6225


The Bayer process for alumina refining, although not as energy intensive as smelting, derives a significant proportion of the cost of production from energy consumption. Worldwide there continues to be an increased push to reduce ozone-depleting emissions from all sources, culminating in the signing of the Kyoto Protocol (1997). Those refineries that actively contain and improve their energy efficiency relative to competitor alumina producers achieve both reductions in the unit cost of production thereby progressing down the alumina industry cost curve and also minimising their emissions of greenhouse gases.

Current strategies and tools used to improve the energy efficiency of the Worsley Alumina refining process are discussed followed by opportunities for further reductions in the unit production cost of alumina and hence emission reduction post-3.1 mtpa expansion project.


energy efficiency, energy management, greenhouse gases, cogeneration, greenhouse challenge




David Heath


The Worsley Alumina refinery is located in the south west of Western Australia processing bauxite mined from the Mt Saddleback deposit within the Darling Range since 1984. To maintain it’s cost competitiveness all inputs to the refining process must be critically reviewed, understood and controlled. The expansion project currently under construction will increase refinery production from 1.8 million tonnes per annum to 3.1 million tonnes per annum and further enhance Worsley Alumina’s ranking within the lowest quartile of the alumina producers cost curve (King, 1995).

The Bayer process for alumina refining is a highly energy intensive in nature. As such the cost borne by refiners in meeting this energy requirement represents a significant percentage of the unit cost of producing a tonne of alumina (Industry Com, 1998). Worsley currently derives its utility steam and electrical power from a coal-fired powerhouse and the hydrate produced is calcined within gas-fired circulating fluid bed calciners.

Worldwide there as been an increased push for all nations to curb their emission of gases detrimental to the ozone layer, commonly known as greenhouse gases. This culminated in the signing of the Kyoto Protocol in 1997, of which Australia is a signatory.

This paper will discuss some of the strategies employed at Worsley Alumina to both decrease the total cost of satisfying the energy needs of the Bayer process and to actively decrease the total emission of Greenhouse gases.


During the 1992 Earth Summit in Rio, under the UN Framework Convention on Climate Change, greenhouse gas emissions were acknowledged to produce a significant threat to the near-earth surface temperature (global warming) of the Earth. In 1997 an agreement was reached known as the Kyoto Protocol whereby Annex B (developed nations) committed to decreasing net emissions. Issues to be resolved include ratification by the United States government, and ratification of the protocol as a whole, the exact voluntary restrictions made by developing countries and the framework for any emissions trading scheme (Pritchard, 1998).

The Australian Government, as part of it’s National Greenhouse Strategy (AGO, 1998), initiated the Greenhouse Challenge program. The ‘Challenge’ involves cooperative agreements between industry and the Commonwealth Government to voluntarily decrease the net emission of greenhouse gases. Of relevance to the Australian alumina refining industry is the cooperative agreement entered into by the Australian Aluminium Council (AAC) to reduce the emission of carbon dioxide by 7% on a per tonne of production basis (1990 baseline). Worsley, as a joint signatory to this agreement and aware of the cost benefits of saving energy, has continued to actively target improved energy efficiency of the refining process.


The current utility steam and power requirements are met by operation of the refinery powerhouse where steam is raised via three coal-fired water tube boilers. The bulk of the refinery energy demand is met by this means. The steam is expanded through three extraction condensing turbines. Process steam is extracted at 1300 kPa for digestion heating and 450 kPa steam for all other process heating applications. Provision is made for gas ignition and fuel oil firing of the boilers for transient periods of operation only.

The calcination of the aluminium tri-hydrate filter cake is carried out within four circulating fluid bed calciners which completes the alumina extraction process. Although originally designed for firing with fuel oil, these are now fired with natural gas.


The original refinery was designed with a nameplate capacity of one million tonnes per annum. With capital expansion projects for debottlenecking and advances in the understanding of the Worsley Bayer process chemistry production rates have increased to 1.8 million tonnes per annum. This has been largely achieved without any major change to process equipment and has pushed the refinery energy balance away from the original design conditions. Due to Worsley’s favorable position within the world alumina cost curve and its desire to remain a low cost producer the challenge is to both maximise production while minimising energy inefficiencies both now and after completion of the 3.1 mtpa Expansion Project.

Different tools have been employed to improve energy efficiency of the current plant, some of which are generic in nature and as such apply further afield than being specific to the alumina industry. Many of the current energy bottlenecks have been addressed within the expansion design process.

Projects completed are discussed below.

4.1 Pinch Analysis

Studies in the past have modeled the current plant using process integration techniques to identify opportunities to reduce steam consumption via improved heat recovery, commonly known as pinch analysis. Pinch technology (Linhoff, 1982) defines the minimum steam load required to satisfactorily heat the total process and the minimum heat rejection load required to satisfactorily cool process streams.

Heat & material balances were completed to define a model case of the refinery operation. The various cold and hot process streams that require heat interchange were then meshed into hot and cold composite curves. Bringing together the heating and cooling profiles allows construction of the refinery grand composite curve that defines the steam - process - cooling water interaction as a function of temperature. The pinch point is defined as the point at which heat exchange is most constrained and the actual temperature difference is denoted D Tmin. The practical significance of this point is any further decrease in D Tmin will result in reduced heating and cooling duties. Once the current refinery pinch point and production rate benefits associated with improving this bottleneck were determined resources were directed to de-bottlenecking the refinery pinch point.


Figure 1

Example Grand Composite Curve Derived by Pinch Analysis Techniques

4.2 Energy Key Performance Drivers

The old adage "you can’t control what you can’t measure" is very apt when attempting to contain energy costs within any energy intensive process. Indeed the same effect is achieved if measured energy performance information is not imparted succinctly to those that can take remedial action to improve the energy efficiency of the refining process.

A series of critical key performance drivers were identified to aid in optimising the overall energy efficiency of the refining process. This information is plotted on a weekly and quarterly basis for management and technical staff to ascertain where longer-term strategic resources must be placed to improve energy efficiency. These performance drivers, to be effective, are presented in a manner that are readily understood by the target audience. Figure 2 below is indicative of the plots circulated.

Figure 2

Example of an Energy Key Performance Graph.






Std Dev






In the above example, control of Green Liquor temperature is seen as crucial for many reasons, notably precipitation yield and product quality control. In addition to Bayer chemistry issues, control of the heat interchange facility is crucial to the refinery energy balance as it is at this point that the Worsley Refinery pinch point is located. Any problems with this facility adversely impact on the energy balance.

The energy key performance plots circulated are effective for longer term planning however this information suffers from being reactive in nature with a long lead time from the initial perturbation occurring to remedial action being taken. As with the optimisation of any process variable, significant advantages lie in the management of real time data collection resources.

Although previously real time data collection of variables pertinent to the energy efficiency of the refinery had been collated within the process history database, this information had not been consolidated into a clear summary of process energy performance. In order to improve the accessibility of key parameters impacting on the ‘health’ of the refinery energy balance a central high-level summary was created. Process variables have deviation alarms and this summary acts as a pathway for staff to drill down, further targeting specific unit operations.


Figure 3

Example of Energy SCADA Summary for Production & Technical Staff

4.3 Steam Reticulation Audit

An audit of the refinery steam reticulation and condensate recovery system was completed to quantify all steam leaks, steam trap condition and to identify all instances of inadequate insulation of the reticulation system. Although the Worsley refinery was benchmarked favourably against other industries, notably petrochemicals, oil & other mineral processors, the audit identified instances where repair of current installed plant would generate significant savings. Factors contributing to Worsley’s favourable comparison to other sites included a relatively new plant, a low pressure steam distribution system and a workforce that already had a high degree of awareness regarding equipment maintenance.

In addition resources have been made available to further improve and refine the maintenance practices on the reticulation system which will derive further benefits with time.

4.4 Fugitive Emissions

Although carbon dioxide is by far the most prevalent greenhouse gas emitted in the refining process slight emissions of another greenhouse gas pertinent to the is methane (CH4). Methane has a global warming potential 24.5 times that of carbon dioxide (CO2) (Pritchard, 1998).

As previously mentioned the major consumer of methane are the gas fired circulating fluid bed calciners for conversion of hydrate to product specification alumina. The natural gas reticulation system also extends to ignite the powerhouse boilers and for a heat treatment kiln within the main workshop.

Although calciners are gas tested prior to being brought on line for leaks it was considered timely to audit the entire reticulation system. Any potential methane leaks not only constitute inefficient use of a raw material of production already paid for but also a very serious safety concern to both the workforce and the environment. Some leaks were detected and attended to immediately.

4.5 Process Control Loop Retuning

A study was initiated to review the current control methodologies employed on the utility generation and reticulation system. The major driving force for this project was the reduction in cycling steam demands that on occasions have induced instability into the three boilers operating at maximum capacity. Also boilers producing to a stable steam demand can be operated at a higher thermal efficiency.

Process control areas reviewed have included simple PID controllers within the facilities consuming the steam through to boiler control algorithms such as those controlling boiler feed pumps or drum level. To effect any energy-related advanced control strategy over a sub-optimally tuned group of PID controllers would negate any benefits of the advanced control project. This was considered a prerequisite to further energy-related advanced control projects.

4.6 Boiler and Calciner Thermal Efficiency Tests

Efficiency checks were completed on the boilers and calciners to ensure that both the boilers and calciners were operating at least to their original design capacity. Test conditions were controlled such that the equipment under test was maintained at steady operating conditions whilst the other units throughput were varied as per the refinery production demand at the time.

Boiler efficiencies were calculated by the heat loss method. This involved sampling temperature, gas compositions, velocities, ash and coal feed samples. Boiler tests were completed both at the original design capacity and at their current operating throughput. Representative calciner efficiency checks were also completed to determine whether their efficiencies (GJ gas/t alumina product) were again operating at an acceptable level of thermal efficiency.


4.7 Refinery Production Rate

When process equipment is properly utilised the operation results in an optimum energy consumption level however when this equipment is underutilised the same fixed energy costs are not diluted by the lost production attributable to the underutilised equipment. For example reduced precipitation yield or increased hydrate returning with spent liquor to digestion severely impacts on the total unit energy consumption as the entire liquor stream is reheated through the digestion process. Also baseline process steam requirements and electrical costs associated with equipment operation (e.g. pumping) must still be borne by the refinery irrespective of the production rate. The marginal increase in total fuel costs consists mainly of increased calcination fuel consumption to calcine the marginal increase in production.


5.1 Pinch Analysis Techniques Applied to the Proposed Expansion Design

A brief overview of Process Integration has already been discussed. It is considerably cheaper to incorporate into the design of a large major capital expansion with enhanced energy efficiency than retrofitting heat recovery equipment at a later date. Therefore at an early stage of the Expansion conceptual design process, improving the thermal and electrical energy efficiency of the process was considered critical to the success of the project by both the Worsley Joint Venture owners and the Expansion design team. To do this the heat and material balance flow sheet of the expanded plant was modeled to identify heat recovery opportunities.

Given the physical layout of the refinery, identified energy recovery opportunities were incorporated into the final design. As previously mentioned it is at the refinery pinch point where real gains in thermal efficiency can be made by pushing the hot and cold composite curves together thereby reducing the total utility requirements necessary for the Expansion project. To achieve this, the Worsley heat interchange facility is to be rebuilt utilising plate heat exchangers, a technology already proven within the alumina industry. This will open up the available driving temperature forces for efficient heat exchange. Due to the removal of the flash evaporation duty of the current facility, extra evaporation capacity is incorporated into the design for refinery water balance considerations.

5.2 Gas-fired Co-generation Plant to Satisfy Additional Utility Requirements

In order to satisfy the increased utility requirements of the Expansion process flow sheet a gas-fired turbine and heat recovery steam generator (GT/HRSG) owned by an independent power producer is to be constructed. This will produce 120MW of power and enough steam to satisfy the incremental steam requirements for the Expansion project. A proportion of the electrical power produced will be consumed by Worsley to partially provide for the increased electrical demand associated with the Expansion project. For a more detailed explanation of the co-generation selection considerations please refer to Gaynor (Gaynor, 1998).

Gas-fired combined heat and power plants located at a heat sink source are a tremendous opportunity to reduce the total emissions of greenhouse gases whilst still partially satisfying a consumer’s requirement for electricity. The table below details the significant reductions in emissions gained by the installation of the GT/HRSG. Although there is an increase in the localised total carbon dioxide emissions at the Worsley refinery site this is due to the increased electrical power generated for export to the state utility’s transmission grid. This increase in emissions is considerably less than if the incremental increase in generating capacity was produced via a conventional coal-fired boiler.


Reduction in Emissions at the Worsley Site Relative to an Additional 4th Coal Fired Boiler Sized Solely for Worsley Expansion Requirements

Overall Reduction in Emissions Combining the Worsley Site and Statewide Power Production Requirements





NOx (as NO2)




- 3 900

18 600







Figure 4

Reduction in Emissions due to GT/HRSG Inclusion in Expansion Design

5.3 Other Future Opportunities Post - Expansion

As would be expected with comissioning a project of the magnitude of the Worsley Expansion, significant resources will be directed to ensuring that the expanded plant is successfully running at or above design capacity. Optimisation of the new and existing plant will directly lead to improving the overall energy efficiency of the process.

In addition to expansion commissioning, other areas that are earmarked to enhance the overall energy efficiency of the refining process include the following: -


Increasing the energy efficiency of a mineral processing facility produces the dual benefits of lowering the unit cost of producing a tonne of product and lowering a site’s total greenhouse gas emissions. To achieve this end at Worsley Alumina, various short and long-term strategies, both employing additional capital expenditure and changes to operating practices are being utilised.


King, J., The Market for Alumina Current Trends and Future Prospects, August 1995, Northumberland, England.

Industry Commission, Micro Reform – Impacts on Firms: Aluminium Case Study, Research Paper, March 1998, AusInfo, Canberra, Australia.

Pritchard, R., Energy Council of Australia: The Impact of Climate Change on the Energy Sector, Pritchard-Udovenya, September 1998, Sydney, Australia.

Australian Greenhouse Office, The National Greenhouse Strategy, 1998, Canberra, Australia.

Linhoff. B., User Guide on Process Integration for the Efficient Use of Energy, 1982, IChemE, Rugby, England.

Gaynor, B., Energy Developments at Worsley Alumina, Energy in Western Australia ’98 conference proceedings, March 1998, pp 126-139.