BM–031 | TECHNOLOGICAL OPTIONS FOR A NEW ALUMINA REFINERY

Author: Rajendra Kunwar, Bayer Process Expert

1. Executive engineering perspective

The Bayer process is a coupled hydrometallurgical system. Digestion determines extraction and liquor loading; desilication and residue washing influence soda losses and scale formation; clarification protects precipitation; precipitation governs hydrate yield, particle-size distribution and liquor productivity; and calcination converts hydrate into the required alumina product. Heat recovery, evaporation, steam generation, electrical power, water management and residue disposal link these sections together. The technology route for a greenfield Alumina refinery project must be developed from the bauxite mineralogy and variability, product specification, recovery target, water and energy balance, residue-management strategy, site constraints, operability and whole-life economics. The best solution is the one that delivers reliable production and saleable alumina at the lowest sustainable cost while meeting environmental and safety obligations.

 

2. Establish the design basis before choosing technology

·     Bauxite characterisation: representative composites by mine domain and blend; total and available Al₂O₃; reactive and total SiO₂; Fe₂O₃, TiO₂, organic carbon, sulfur, phosphorus and trace elements; loss on ignition; moisture; particle-size distribution; and mineralogical proportions of gibbsite, boehmite, diaspore, kaolinite and other silica-bearing phases.

·    Variability and geo-metallurgy: define normal, adverse and extreme feed envelopes, not just one average assay. Test the effect of seasonal moisture, changing mine faces, blending limits and stockpile management.

·   Product and capacity: metallurgical-grade alumina or specialty hydrate/alumina; annual saleable tonnes; operating days; stream factor; ramp-up allowance; and product quality limits.

·      Site utilities and constraints: fuel and steam options, electricity reliability and tariff, cooling-water availability, raw-water quality, rainfall and flood design, ambient temperature, altitude, land availability, logistics and grid connection.

·   Environmental basis: residue classification, permitted storage/disposal route, water recovery, dust/noise, emissions, caustic containment, storm-water separation and closure obligations. Confirm applicable permits and current local requirements at project definition.

·       Economic basis: installed CAPEX, owner's costs, contingencies, sustaining CAPEX, energy and reagent prices, caustic make-up, bauxite cost, residue handling, product value, financing assumptions and sensitivity cases.

Design rule: freeze a signed Design Basis Memorandum and a bauxite test work plan before equipment sizing. Any process guarantee—recovery, liquor productivity, steam rate, fuel rate or energy consumption—must state the bauxite envelope, product basis, boundary limits, measurement method and operating conditions.

3. Technology-selection matrix 

Section	Options to assess	Engineering selection considerations
Bauxite crushing	Double-roll crusher; hammer crusher; impactor	For the size reduction from about 120 mm to 20 mm, compare availability, wear, moisture sensitivity, tramp-metal protection and product-size control. Double-roll crushers can offer robust, controlled reduction for suitable feed, but selection must be confirmed by testwork and vendor guarantees.
Bauxite grinding	Dry grinding; wet open-circuit; wet closed-circuit; single- or two-stage wet grinding	Wet grinding is generally attractive for dust control and direct slurry preparation. Single-stage open circuit may be a simple greenfield candidate, but final choice depends on grind target, ore hardness, throughput turndown, power, classification needs and downstream digestion performance.
Digestion	Atmospheric; medium-pressure; high-pressure; double digestion; tube digestion; two-step digestion	Gibbsite-rich bauxite often favours lower-temperature digestion; medium-pressure digestion may improve extraction and reduce bauxite demand versus atmospheric operation. Boehmitic/diasporic content can require higher severity or staged treatment. Confirm with digestion tests and a complete heat balance.
Desilication	Pre-desilication; post-desilication; combined or staged schemes	Choose from reactive silica mineralogy, silica loading, residence-time needs, DSP formation, soda loss and scale risk. The location and duty of each stage should be established by testwork rather than a single universal silica cut-off.
Residue separation and washing	Conventional settlers; deep thickeners; CCD washing; filtration-assisted options	Optimise clarification area, underflow density, flocculant use, washing efficiency, liquor entrainment, soda recovery and residue rheology together. Deep thickeners can reduce footprint, but performance depends on feed solids and settling characteristics.
Residue handling	Wet storage; thickened residue; dry stacking / filtered residue	The preferred route is jurisdiction- and site-specific. Where wet impoundment is prohibited or unacceptable, evaluate thickened or filtered residue with water recovery, transportability, stack stability, rainfall management and closure design.
Security filtration	Horizontal leaf filters; auto-dumping vertical filters; other pressure filtration designs	Prioritise filtrate clarity, solids capture, cycle reliability, maintainability, safe cake discharge and minimal liquor loss. Select using liquor chemistry, suspended-solids load, required redundancy and vendor trial data.
Liquor heat recovery	Flash-based heat recovery; plate heat exchangers; shell-and-tube exchangers; hybrid systems	Compare recoverable duty, approach temperature, fouling/scaling, pressure drop, cleanability, footprint, metallurgy, maintenance access and integration with spent-liquor heating. PHEs can be compact; flash systems may suit particular pressure/temperature and vapour-reuse schemes.
Precipitation	Batch; single-stage continuous; two-stage continuous; multi-zone arrangements	Two-stage agglomeration and growth concepts can improve control of particle size and productivity when seed quality, supersaturation, temperature, residence time and classification are properly managed. Pilot or operating data should validate performance.
Hydrate classification and washing	Hydrocyclones/classifiers; thickeners; vacuum or pressure filtration; combinations	Design for product PSD, seed recycle, liquor recovery, soluble soda removal, cake moisture and filtrate routing. Hydrate washing and filtration should be integrated with precipitation and calcination quality targets.
Calcination	Rotary kiln; stationary fluidised-bed or other stationary calciner	Stationary calciners can offer high thermal efficiency and strong control of product properties; rotary kilns may retain advantages for particular capacities, feed variability or operating preferences. Compare fuel, emissions, product quality, turndown, maintenance and lifecycle cost.
Evaporation and utilities	Multi-effect evaporation; vapour recompression where justified; alternative steam/power configurations	Optimise steam economy against scaling, temperature limits, liquor concentration, condensate quality and availability. The number of effects is not an isolated choice: it must match the site steam system and whole-refinery energy integration.

 

4. Unit-operation options

4.1 Crushing and grinding

The crushing circuit should deliver a controlled top size without excessive fines, unstable throughput or avoidable wear. For a nominal reduction from approximately 120 mm to 20 mm, compare double-roll crushers, hammer crushers and impactors on a common basis: design capacity, feed-size envelope, moisture and clay sensitivity, installed power, wear-part life, maintenance access, dust extraction, noise, tramp-metal handling and availability. A nominal particle size alone is not a complete specification.

Grinding increases reactive surface area and supports digestion kinetics. Wet grinding has largely displaced dry grinding in many refinery applications because it reduces airborne dust and can feed the slurry circuit directly. Nevertheless, the choice between open and closed circuit, and between one and two stages, should be driven by grindability tests, target size distribution, power consumption, classification efficiency and the consequences of coarse particles or excessive fines. Specify a full particle-size distribution (for example, P80 plus oversize limits), not just a single nominal size.

4.2 Digestion and extraction

Digestion selection is primarily a mineralogy-and-economics decision. Atmospheric digestion may reduce pressure-equipment CAPEX and simplify operation, but lower extraction can increase bauxite consumption, residue generation and liquor flow. Medium-pressure digestion is often a strong candidate for gibbsitic feed where the incremental extraction justifies its equipment and steam-system requirements. Boehmitic or mixed bauxite may require higher temperature, longer residence time or staged digestion; double, tube and two-step configurations should be assessed using representative laboratory and, where needed, pilot testwork.

·       Confirm extractable alumina and extraction kinetics across the full bauxite envelope.

·       Calculate digestion temperature and pressure together with caustic concentration, liquor-to-bauxite ratio, residence time and heat losses.

·       Model slurry heating, flash-steam recovery, condensate return, pressure control, erosion and flashing-related wear.

·       Check design cases for low extraction, high silica, high moisture, upset feed and start-up/shutdown—not only the nominal case.

·       Define the basis of recovery clearly: extraction across digestion is not the same as overall refinery alumina recovery.

4.3 Desilication, residue thickening and washing

Reactive silica consumes available caustic and forms desilication products (DSP), which can increase residue mass and soda losses. Desilication strategy must reflect silica mineralogy, reaction kinetics, liquor composition, temperature and residence time. A fixed silica threshold should not be applied universally without confirming whether it refers to total or reactive silica, the sampling point and the specific bauxite blend.

Residue separation is not simply a clarifier-sizing exercise. Settling tests must establish overflow clarity, settling rate, underflow density, compression behaviour, flocculant demand and rheology. Washing stages should be optimised against soluble soda recovery, wash-water use, liquor dilution, hydrolysis risk and the final residue solids target. Deep thickeners can reduce footprint, while CCD trains and filtration-assisted arrangements may be justified where water recovery, footprint or residue transport requirements dominate.

Residue disposal technology must be selected against current regulatory permissions and site conditions. Thickened or filtered residue can reduce free water and support water recovery, but it requires geotechnical design, reliable filtration/thickening, conveying or pumping, drainage, storm-water management, dust control and a closure plan. Do not assume that a nominal solids percentage alone establishes dry-stack suitability.

4.4 Security filtration and liquor heat recovery

Security filtration protects downstream precipitation from suspended solids that can impair hydrate quality and equipment performance. Specify allowable filtrate solids, filter-cycle time, cake discharge reliability, liquor loss, redundancy and maintenance philosophy. Auto-dumping vertical filters are a common modern option, but final selection should be based on liquor chemistry and supplier test results rather than age of technology alone.

Hot clarified aluminate liquor and returning spent liquor offer an important opportunity for heat recovery. Plate heat exchangers can provide compact installation and high heat-transfer coefficients, while flash-based systems can recover sensible heat and generate vapour for reuse. A hybrid arrangement may be appropriate. Compare systems using a consistent heat-and-mass balance, including approach temperatures, pressure drop, fouling allowance, scaling tendency, cleaning strategy, gasket/material compatibility, space, installed cost, downtime and the value of recovered steam.

4.5 Precipitation, hydrate finishing and calcination

Precipitation is a crystal-engineering operation as much as a chemical conversion step. Batch, single-stage continuous and two-stage continuous configurations should be compared on hydrate yield, liquor productivity, particle-size distribution, attrition, seed demand, residence-time distribution, cooling duty, tankage, control complexity and operability. In a two-stage concept, agglomeration and crystal growth conditions are managed separately to improve control of the final particle population. Performance depends on supersaturation, temperature profile, seed surface area and quality, agitation, impurities and classification/recycle balance.

Hydrate classification, washing and filtration determine the split between product and seed, soluble soda in the cake, mother-liquor recovery, cake moisture and the load presented to calcination. The system should be modelled as an integrated recycle circuit, including filtrate return and the impact of water addition on evaporation duty. Calciner selection then balances fuel use, emissions, product phase and surface area, loss on ignition, dust handling, maintenance, turndown and reliability. Stationary calciners are often thermally efficient, but the final choice must be based on the required alumina quality and project-specific lifecycle economics.

5. Indicative performance targets: use as screening values, not guarantees

The source paper listed the following achievable figures for its selected process route. They are reproduced here as preliminary screening targets, not universal industry guarantees. Before use in a feasibility study or EPC contract, define the bauxite assay and mineralogy, product basis, recovery definition, plant boundary, fuel quality, steam conditions, utility imports/exports and measurement method. Some values may reflect a particular historical plant or operating basis.

Parameter	Source-paper figure	Required qualification before adoption
Overall alumina recovery	94%	Define recovery boundary and whether calculated from bauxite total, available or extractable alumina; state bauxite blend and accounting method.
Liquor productivity	90 g/L Al₂O₃	Define the liquor stream and measurement basis, operating temperature, dissolved alumina and caustic concentrations, and whether this is a design target or sustained operating result.
HFO consumption	73 kg/t Al₂O₃	Fuel-specific figure; validate against calciner technology, fuel calorific value, moisture, product specification and whether other fuels are included.
Process steam	1.80 t/t Al₂O₃	Define steam pressure/quality, battery limits, condensate return, evaporation duty, cogeneration integration and whether the figure is net or gross.
Total energy	8.5 GJ/t Al₂O₃	Define primary versus delivered energy, electrical conversion factor, fuel basis and plant boundary to prevent double counting.

 

These figures should be stress-tested against adverse bauxite, reduced availability, start-up, turndown, fouling and seasonal utility conditions. A credible performance guarantee should identify the test protocol, correction factors, exclusions and consequences of feed or utility conditions outside the agreed envelope.

6. Integrated design checks that are frequently missed

·       Close the overall mass balance: bauxite, caustic make-up, lime and other reagents, water, product alumina, residue, emissions and all recycle streams.

·       Close the soda balance by stream and by form: soluble caustic, total soda, soda in residue moisture, soda in hydrate cake, spills, purge streams and analytical/accounting losses.

·       Close the heat balance: digestion heating, flash recovery, liquor heat exchange, precipitation cooling, evaporation, calcination, steam generation and condensate return.

·       Model water balance for dry and wet seasons, including residue wash water, cooling systems, rainwater, storm-water segregation, condensate quality and zero/low-discharge constraints where applicable.

·       Size equipment for both normal and limiting cases: maximum throughput, minimum turndown, high-solids operation, degraded heat-transfer coefficients, one train out of service and start-up inventory.

·       Use realistic design margins. Avoid indiscriminate oversizing that increases capital cost, residence time, dead zones, power consumption or control difficulty.

·       Specify metallurgy from actual liquor chemistry, temperature, pressure, erosion, corrosion and stress-corrosion/caustic-embrittlement risk. Material selection and weld/stress-relief requirements require specialist review.

·       Design for maintainability: lifting access, exchanger bundle or plate removal, filter internals, pump pull-space, isolation, drains, flushing, sampling points, safe access and online instrumentation.

·       Apply process control and instrumentation from the process model: flow, density, temperature, pressure, level, solids, liquor chemistry, vibration, filter differential pressure and energy metering.

·       Assess reliability quantitatively: critical equipment redundancy, common-cause failures, spares, maintainability, mean time to repair and impact of downtime on annual saleable production.

7. CAPEX, OPEX and technology trade-off

Decision dimension	Questions to answer	Typical economic consequence
Extraction vs. pressure severity	Does higher extraction reduce bauxite and residue cost enough to justify pressure equipment and steam duty?	A higher-CAPEX digestion circuit can lower long-term raw-material and residue costs.
Grinding circuit	What are power, wear, availability and grind-control costs at the guaranteed throughput?	A simple circuit may save CAPEX but can cost more if grind quality or uptime is poor.
Residue washing	What is the marginal value of soda recovered versus extra wash water, pumps, thickeners and evaporation duty?	Improved washing can reduce caustic make-up; excessive dilution can increase energy cost.
Heat recovery	What is the annual recoverable duty after fouling, cleaning and availability allowances?	The best option is determined by net annual savings and maintainability, not nominal exchanger efficiency alone.
Precipitation	What product price or downstream benefit results from better PSD, yield and productivity?	More controlled precipitation may reduce tankage or improve product value, but requires disciplined process control.
Calcination	What are fuel, electricity, emissions-control, maintenance and product-quality costs over the project life?	Thermal efficiency matters, but reliability and product specification can dominate the lifecycle result.
Residue route	What are water recovery, transport, storage footprint, geotechnical, closure and long-term monitoring costs?	A low initial-cost route may carry high sustaining and closure liabilities.

 

Use discounted cash flow and sensitivity analysis rather than comparing equipment purchase prices alone. At minimum, test bauxite quality, caustic price, energy tariffs, fuel quality, exchange rate, availability, production ramp-up, residue cost and product premium/discount. Quantify the cost per tonne of saleable alumina and the annual cash impact of each major design choice.

8. Recommended project workflow

1.    Complete representative bauxite sampling, mineralogy and variability assessment.

2.    Carry out digestion, desilication, settling/washing, precipitation and calcination testwork appropriate to the ore and product.

3.    Develop at least two credible process configurations and prepare consistent mass, soda, water and energy balances.

4.    Screen major equipment options using performance guarantees, reliability, maintenance, footprint, CAPEX and OPEX.

5.    Prepare an integrated process flow diagram, equipment list, utility summary, control philosophy and residue-management concept.

6.    Run operability, HAZID/HAZOP, environmental and maintainability reviews before freezing the design basis.

7.    Set measurable performance guarantees with clear feed, utility, product and battery-limit definitions.

8.    Protect the design through commissioning plans, operator training, performance testing and a structured optimization programme.

9. Conclusions

No single technological configuration is optimal for every bauxite deposit or project site. The preferred route emerges from the interaction of mineralogy, extraction, silica behaviour, soda recovery, liquor productivity, energy integration, product quality, residue management, reliability and life-cycle economics. Public-domain technology offers a sound foundation, provided that it is adapted to the actual feed and validated by test work, engineering calculations and operating evidence.

For a new refinery, the most valuable design decision is often not choosing the newest equipment; it is ensuring that every major unit operation is selected on a common, technically defensible basis and integrated with the rest of the plant. A refinery should be optimized as one system—from ROM bauxite to saleable alumina and managed residue—not as a collection of independent equipment packages.

 

CETI ENTERPRISES, INDIA

CETI Enterprises supports alumina industry professionals through technical training, process audits, calcination performance assessment, energy optimization, process troubleshooting, equipment design review, debottlenecking and refinery improvement programmes. For further details please visit our website at https://ceti.co.in

About the Author: Rajendra Kunwar is Bayer Process & Alumina Technology Specialist. Through CETI Enterprises, he offers expert technical consultancy in customized online / onsite team training, process for production of specialty grade ATH & Alumina, efficiency enhancement, quality improvement, energy reduction, cost reduction, troubleshooting and plant debottlenecking. To explore how we can support your plant operations or upcoming expansion, please contact us at rajendra@ceti.com.