BM-036 | What Makes an Alumina Refinery Efficient?

20 Performance Parameters for Every Bayer Engineer of Alumina Refinery

Author | Rajendra Kunwar, Bayer Process Expert

In the Bayer process, operational excellence is defined by the fine balance between thermochemistry, mass transfer, and energy conservation. Optimizing an alumina refinery requires managing interdependent variables across digestion, clarification, precipitation, evaporation and calcination.

Here are the 20 key performance indicators and benchmark parameters that every Bayer chemical, production and process control engineer must master to evaluate refinery health, maximize alumina yield, and optimize specific energy consumption.

1. Available Alumina (Al₂O₃)

Typical Benchmark: 40% – 50% (depending on ore type: Trihydrate/Gibbsite vs. Monohydrate/Boehmite/Diaspore)

Impact: Defines the theoretical maximum mass of alumina recoverable per tonne of raw bauxite feed. It dictates refinery raw material throughput requirements and mud load across the clarification circuit.

2. Reactive Silica (SiO₂)

Typical Benchmark: 1.5% – 3.5%

Impact: Directly determines caustic soda consumption through the formation of Desilication Product (DSP, 3Na₂O·3Al₂O₃·6SiO₂·2H₂O). Every kilogram of reactive silica bound into DSP neutralizes approximately 0.6–0.8 kg of equivalent NaOH.

3. Bauxite Moisture

Typical Benchmark: 5% – 12% (As-Received basis)

Impact: High free moisture increases freight costs, causes handling-bin hang-ups in dry grinding circuits, and creates parasitic water input into the process liquor balance, placing an additional load on the evaporation station.

4. Grinding Particle Size Distribution (PSD)

Typical Benchmark: 80% – 90% passing 150 µm (+149 µm / +100 mesh < 5%)

Impact: Balances reaction kinetics against downstream solids-liquids separation. Finer grinding enhances alumina leaching rates and reduces erosion in digestion pipelines, but excessively fine grinding impairs red mud settling and filtration rates in the clarification circuit.

5. Pre-Desilication Efficiency

Typical Benchmark: >95% Reactive Silica converted to DSP prior to main digestion

Impact: Holding slurry at 95°C – 100°C for 6 – 8 hours promotes controlled nucleation and growth of sodalite-type DSP on solids. High efficiency prevents high-temperature scale formation (cancrinite/sodalite hard scale) inside digestion heat exchangers.

6. Digestion Temperature

Typical Benchmark: Gibbsitic Ore: 140°C – 150°C (Low-Temperature Digestion) | Boehmitic Ore: 240°C – 270°C (High-Temperature Tube/Autoclave Digestion)

Impact: Governs reaction thermodynamics and solubility limits. Operating below target temperature leads to incomplete extraction, while operating excessively high increases steam requirements and promotes unwanted reactions with non-reactive silicates.

7. Digestion Retention Time

Typical Benchmark: 10 – 15 minutes (High-Temp Tube Digestion) to 30 – 60 minutes (Autoclave Systems)

Impact: Provides the required kinetic contact time to reach near-saturation equilibrium (A/C target) without oversized reactor volume. Excessive residence time consumes parasitic heat and increases capital footprint.

8. Alumina Extraction Yield

Typical Benchmark: 94% – 99% of Available Alumina

Impact: Measures digestion recovery performance. Unextracted alumina remains trapped in the insoluble red mud phase as unreacted gibbsite/boehmite or reprecipitated hydrate, directly lowering total plant efficiency.

9. Residue (Red Mud) Generation

Typical Benchmark: 0.8 – 1.5 t red mud / t alumina produced

Impact: Reflects bauxite grade and mineralogy. Higher residue generation burdens clarification washer trains (Counter-Current Decantation), increases flocculant consumption, and demands larger residue storage facilities (RSF / Dry Stacking).

10. Residue Soda Loss

Typical Benchmark: <40 kg Na₂CO₃ equiv. / t alumina (Bound + Soluble)

Impact: Represents one of the largest operational variable cost drivers. It consists of bound soda (fixed chemically in DSP) and unwashed soluble soda remaining in the disposed cake moisture.

11. Spent & Pregnant Liquor Concentration

Typical Benchmark: 200 – 260 g/L Total Caustic Soda (TCS as Na₂CO₃)

Impact: Sets the liquor density and viscosity profiles. Higher caustic concentrations enhance alumina solubility and digestion driving force, but elevated liquor viscosity hinders red mud settling, filtration, and precipitation kinetic rates.

12. Digestion Outgoing A/C Ratio

Typical Benchmark: 0.62 – 0.70 (g Al₂O₃ / g Na₂CO₃ equiv.)

Impact: Represents liquor supersaturation potential exiting digestion. Maximizing the target A/C ratio increases the working capacity of the liquor stream, allowing higher alumina throughput per cubic meter of circulating liquor.

13. Precipitation Yield / Temperature Profile

Typical Benchmark: 70°C – 75°C (Green Liquor Inlet) down to 45°C – 52°C (Discharge)

Impact: Controls the thermodynamic driving force for trihydrate (Al(OH)₃) precipitation. The temperature gradient must be managed carefully to balance supersaturation, nucleation rate, and crystal growth velocity without generating ultra-fine gibbsite particles.

14. Seed Concentration

Typical Benchmark: 300 – 500 g/L (Fine + Coarse Seed charge in Precipitation)

Impact: Supplies the available surface area (m²/L) required to direct precipitation toward crystal growth rather than unseeded secondary nucleation. Proper seed area management prevents auto-nucleation and maintains target product sizing.

15. Precipitation Residence Time

Typical Benchmark: 24 – 48 hours (across continuous precipitator trains)

Impact: Gibbsite crystallization kinetics are notoriously slow. Adequate residence time allows the liquor A/C ratio to decay safely down to equilibrium (0.32 – 0.38), ensuring high alumina precipitation yield per pass.

16. Liquor Productivity

Typical Benchmark: 70 – 90 kg Al₂O₃ / m³ liquor

Impact: Defined as the delta in alumina concentration between Green Liquor and Spent Liquor (ΔA). It directly impacts specific pumping energy, heat exchange requirements, and equipment sizing across the entire plant loop.

17. Precipitation Efficiency

Typical Benchmark: 45% – 52% conversion of dissolved alumina in liquor

Impact: Measures how effectively the supersaturated pregnant liquor drops its alumina load before returning to evaporation and digestion. Low efficiency forces higher liquor circulation rates for the same target annual plant tonnage.

18. Evaporator Steam Economy

Typical Benchmark: 4.5 – 5.5 kg water evaporated / kg live steam supplied (5 to 6-Effect Falling Film Evaporators)

Impact: Quantifies thermal efficiency in managing the plant water balance. Evaporation removes dilution water introduced during mud washing and seal flushes to restore spent liquor caustic concentration back to digestion targets.

19. Calciner Specific Fuel Consumption

Typical Benchmark: 2.7 – 3.1 GJ / t calcined alumina (Stationary Fluidized Bed Calciners)

Impact: Calcination accounts for roughly 25% – 35% of total refinery energy costs. Modern gas-fired CFB calciners utilize multi-stage cyclone preheating and hydrate drying to minimize thermal losses in off-gas and product alumina cooling.

20. Overall Alumina Recovery

Typical Benchmark: 90% – 95% (Overall Plant Yield)

Impact: The macro metric combining chemical extraction, precipitation yield, and process losses (mud wash losses, calciner dust losses, liquor spillages). It represents total plant effectiveness in converting raw ore input into marketable sandy grade alumina (SGA).

Bayer process efficiency is not controlled by one parameter. It is the result of maintaining the entire process chain in the plant.

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.

 

BM-035 | Why Does Reactive Silica Matter in Efficient Operation of an Alumina Refinery?

Process Chemistry • Bauxite Quality • Soda Economy • Residue Management

Author | Rajendra Kunwar, Bayer Process Expert

1. EXECUTIVE OVERVIEW

 Reactive silica is not merely an impurity reported in a bauxite assay. Under Bayer digestion conditions, its mineralogical form and reactivity determine how much caustic soda and dissolved alumina are immobilized in desilication products (DSP), how much residue is generated, and how reliably the refinery can sustain extraction, liquor productivity and operating cost. Reactive silica is one of the most economically consequential constituents of bauxite processed by the Bayer process. Unlike much of the alumina-bearing mineral matter, reactive silica dissolves in hot caustic liquor and subsequently combines with sodium, aluminium and other species to form insoluble desilication products (DSP). The resulting reactions consume valuable process soda, bind recoverable alumina, increase residue burden and can aggravate scaling, filtration and washing constraints.

The operational impact depends on more than the total SiO₂ assay. Mineralogy, particle size, liberation, digestion temperature, caustic concentration, residence time, liquor composition and the effectiveness of pre-desilication all influence the fraction that reacts. A refinery that treats total silica as a single uniform number can therefore misjudge both its bauxite value and its process risk.

The central management objective is to minimise the quantity of reactive silica entering the high-temperature digestion circuit, convert it in a controlled manner where appropriate, and remove the resulting DSP under conditions that protect alumina recovery and soda economy.

2. WHAT IS REACTIVE SILICA?

In Bayer-process practice, reactive silica is the portion of bauxite silica that dissolves or reacts appreciably with caustic liquor under the refinery’s actual digestion and pre-desilication conditions. It is commonly associated with clay minerals and poorly crystalline or fine-grained silica phases. Kaolinite is a major source; other clay minerals and amorphous silica may also contribute.

Silica form	Typical Bayer behaviour	Process significance
Kaolinite and other reactive clays	Attack by hot caustic; silica enters liquor and forms aluminosilicate species	High DSP formation potential; soda and alumina losses
Amorphous / poorly crystalline silica	Often highly reactive; rate depends on structure and particle size	Can react rapidly and contribute materially to liquor silica
Quartz / crystalline silica	Generally much less reactive under conventional digestion; rate increases with severity, fineness and time	Often reports substantially to residue, but fine or activated quartz should not automatically be treated as inert
Total SiO₂	Analytical sum of reactive and relatively non-reactive silica	Not a direct substitute for a measured or modelled reactive-silica value

 

Important distinction: total silica is a chemical assay; reactive silica is a process-response parameter. A useful bauxite evaluation therefore combines X-ray diffraction (XRD), petrography or mineralogical interpretation, particle-size information, and a digestion-based reactivity test representative of the refinery flowsheet.

3. THE CHEMISTRY: From DISSOLVED SILICA TO DSP

During digestion, reactive silica is attacked by sodium hydroxide and forms soluble silicate species. In the presence of dissolved aluminate and sodium, these species can precipitate as hydrated sodium aluminosilicates. Depending on liquor chemistry, temperature, time and available anions, DSP may occur as sodalite-, cancrinite- or related phases. The actual phase assemblage is refinery-specific.

A simplified representation is:

Reactive aluminosilicate minerals + NaOH + sodium aluminate  →  insoluble DSP

This is a process-level representation, not a balanced reaction equation. DSP composition is variable: carbonate, sulphate, chloride and other anions may be incorporated, and the Na₂O/Al₂O₃/SiO₂ ratios differ by phase and operating conditions.

Two economic penalties arise simultaneously:

·       Soda immobilisation: sodium becomes chemically bound in DSP and leaves the liquor circuit with the solid residue unless recovered by residue treatment.

·       Alumina immobilisation: DSP incorporates aluminium, reducing the alumina available for hydrate precipitation and increasing the effective alumina loss to residue.

4. WHY REACTIVE SILICA MATTERS TO REFINERY PERFORMANCE

4.1 Soda consumption and caustic inventory

DSP formation is a major route by which process soda is transferred from circulating liquor into bauxite residue. The consequence is not limited to the purchase cost of replacement caustic. Higher soda loss also increases the requirement for make-up, affects liquor composition and can constrain the refinery’s ability to maintain target causticity and alumina concentration.

For preliminary screening, an empirical planning factor of approximately 0.65–0.70 tonne Na₂O equivalent per tonne of reactive SiO₂ may be used where it is supported by the refinery’s own balance or comparable operating data. It is not a universal stoichiometric constant. Actual soda fixation depends on DSP mineralogy, liquor chemistry, washing efficiency and residue-treatment practice.

4.2 Alumina loss and extraction efficiency

DSP incorporates alumina into an insoluble solid. The resulting loss reduces overall extraction and can increase the gap between theoretical digestion yield and saleable hydrate production. The penalty is especially important where the bauxite has a low available-alumina-to-reactive-silica ratio. A high total Al₂O₃ assay can therefore be misleading if a significant fraction of the alumina is offset by reactive-silica-related losses.

4.3 Residue generation, washing and disposal

Reactive silica contributes to the formation of additional DSP solids and changes residue chemistry, particle characteristics and liquor retention. Higher residue load can increase thickener, washer, filtration and residue-storage duties. The economic impact includes pumping power, wash-water demand, equipment capacity, residue area and the liquor and soda carried out with the washed solids.

4.4 Scaling and equipment availability

Silica-bearing liquor and DSP precipitation can contribute to deposits in digestion, flashing, heat-interchange, clarification and liquor-handling equipment. Scale composition and location vary; silica-related deposits should be confirmed by deposit analysis rather than assumed from appearance alone. Deposits reduce heat-transfer coefficients, increase pressure drop, disturb flow distribution and force cleaning outages.

4.5 Liquor productivity and plant capacity

When soda and alumina are tied up in DSP, the refinery’s effective liquor composition and available alumina inventory are affected. Higher residue production and scaling can also constrain circuit throughput. The net result may be lower liquor productivity, reduced extraction stability and less operating flexibility, even when the nominal equipment capacity is unchanged.

4.6 Product quality and process stability

Reactive silica is not normally a direct determinant of hydrate particle-size distribution in isolation. Its influence is mainly indirect, through liquor impurity loading, residue carryover, filtration performance and operating instability. A plant should distinguish genuine silica-related effects from those caused by oxalate, organics, calcium, iron, fine solids or precipitation control.

5. THE ROLE OF PRE-DESILICATION

Pre-desilication (PDS) deliberately exposes bauxite to caustic liquor under controlled conditions before the main digestion and clarification sequence. It promotes conversion of dissolved silicate into DSP in a location where the solids can be managed and separated as part of the intended flowsheet. The objective is not to eliminate the chemical penalty, but to control where and how the reaction occurs and to protect downstream liquor quality and equipment.

The choice between pre-desilication and relying primarily on in-digestion desilication is based on bauxite mineralogy, reactive-silica loading, digestion regime, plant configuration, residence time, residue handling and economics. A threshold used for screening must be treated as a project-specific decision criterion, not as a universal law.

Bauxite reactive SiO₂ (screening basis)	Indicative process consideration
Below about 1.5%	A flowsheet without a dedicated PDS stage may be technically assessable, subject to mineralogy, silica reactivity and plant-specific balance.
Above about 1.5%	Dedicated or intensified pre-desilication should be evaluated in the flowsheet and economic study.
Any level with unusually reactive silica	Mineralogy and digestion tests may justify PDS even when the bulk assay is below the screening threshold.

 

The approximately 1.5% criterion is a preliminary screening benchmark. It should be validated against the actual bauxite blend and the refinery’s digestion conditions. A single threshold cannot represent every orebody or Bayer flowsheet.

6. ILLUSTRATIVE ECONOMIC ASSESSMENT

Consider a refinery processing 10,000 tonnes per day of dry bauxite. Assume reactive SiO₂ is 3.0% by mass and use an indicative soda-fixation planning factor of 0.67 t Na₂O/t reactive SiO₂.

Parameter	Calculation	Illustrative result
Dry bauxite feed	Given	10,000 t/d
Reactive SiO₂ in feed	10,000 × 3.0%	300 t/d
Indicative Na₂O fixation	300 × 0.67	201 t/d Na₂O
NaOH equivalent	201 × (80/62)	259 t/d NaOH equivalent

 

The calculation illustrates the scale of the potential soda penalty; it is not a prediction of actual net refinery loss. The assumed factor must be replaced by a site-specific mass balance that accounts for DSP composition, residue washing, liquor recovery and any residue desilication or soda-recovery system.

For an economic model, multiply the validated net soda loss by the delivered replacement-caustic cost, then add the incremental cost of alumina loss, residue handling, washing, energy, maintenance and capacity constraints. Avoid double counting: soda reported as Na₂O and purchased caustic reported as NaOH must be converted consistently.

7. KEY OPERATING AND DESIGN VARIABLES

Variable	Why it matters	Recommended control / evaluation
Bauxite mineralogy	Determines the fraction and rate of silica dissolution	Characterise clay phases, quartz, amorphous silica and blend variability
Particle size and liberation	Fine reactive phases may dissolve faster; grinding changes exposed surface	Track size distribution and mineral liberation; avoid assuming total SiO₂ equals reactive SiO₂
PDS temperature and residence time	Control silica dissolution and DSP formation kinetics	Optimise using representative tests and validated residence-time distribution
Caustic concentration and liquor composition	Influence dissolution, supersaturation and DSP phase formation	Trend Na₂Oc, total soda, aluminate, carbonate and relevant anions
Digestion severity	Can increase attack on otherwise slower-reacting silica phases	Balance extraction benefit against soda fixation, scaling and energy
Residue washing efficiency	Controls soluble soda and liquor carried with residue	Monitor washer underflow density, wash ratio, overflow soda and residue losses
DSP / residue characteristics	Affect settling, filtration, rheology and disposal	Trend settling rate, mud density, filterability and mineralogical composition

8. MONITORING, TESTING AND RECONSILIATION

A robust silica-management programme must connect ore characterisation to plant performance. Recommended practices include:

·       Establish a bauxite-blend database with total SiO₂, reactive SiO₂, available alumina, mineralogy, particle size and source variability.

·       Use a consistent laboratory method for reactive silica. Record test temperature, caustic concentration, liquid-to-solid ratio, particle size and contact time; results from different methods are not automatically comparable.

·       Sample bauxite, PDS discharge, digestion slurry, clarified liquor, residue and washer streams on a coordinated basis.

·       Trend dissolved silica in process liquor where analytical methods and sampling practice are reliable; interpret trends with liquor composition and temperature.

·       Analyse DSP and residue periodically by XRD and, where useful, SEM-EDS or equivalent mineralogical methods.

·       Close a monthly Na₂O and Al₂O₃ balance. Separate chemically fixed soda from soluble soda lost in residue moisture and other losses.

·       Correlate silica indicators with extraction, liquor productivity, residue generation, washer performance, scale-cleaning frequency and caustic make-up.

9. COMMON MISINTERPRETATIONS TO AVOID

·       Using total SiO₂ as if it were entirely reactive silica.

·       Assuming quartz is always completely inert, irrespective of fineness, mineral activation and digestion severity.

·       Applying a single soda-fixation factor as a universal stoichiometric constant.

·       Treating pre-desilication as a way to eliminate soda and alumina losses; it primarily controls the reaction and its location.

·       Attributing every scale deposit or poor settling event to silica without deposit or residue analysis.

·       Optimising digestion temperature only for alumina extraction while ignoring the additional silica attack and DSP burden.

·       Comparing reactive-silica test results from different laboratories without checking method conditions.

10. PRACTICAL IMPROVEMENT ROADMAP

Priority	Action	Expected benefit
1.Ore control	Define reactive-silica limits for each bauxite source and blend; include penalties in ore valuation	Better feed predictability and sourcing decisions
2. Laboratory validation	Standardise reactive-silica test and correlate it with plant soda and alumina balances	Reliable design and operating basis
3. Flowsheet optimization	Evaluate PDS need, temperature, residence time, mixing and solids separation	Controlled DSP formation and reduced downstream burden
4. Washing optimization	Improve counter-current washing and minimise liquor entrainment	Lower soluble soda loss with residue
5. Equipment reliability	Identify silica-bearing scale by analysis; optimise cleaning and operating windows	Higher availability and stable heat transfer
6. Continuous reconciliation	Integrate ore, liquor, residue and caustic-consumption data in a monthly balance	Sustained cost and recovery improvement

 

11. CONCLUSIONS

Reactive silica matters because it converts a bauxite-quality issue into a refinery-wide process and economic issue. Its reaction with caustic liquor produces DSP, immobilising soda and alumina, increasing residue burden and potentially aggravating scaling and operating constraints. The magnitude of the penalty is controlled by mineralogy and process conditions—not by total SiO₂ alone.

An efficient alumina refinery should therefore treat reactive silica as a key bauxite acceptance and process-design parameter. The best results come from mineralogical characterisation, representative digestion testing, a flowsheet decision on pre-desilication, disciplined residue washing and a reconciled soda–alumina mass balance. The objective is not simply to tolerate silica, but to manage its reaction deliberately and minimise its net cost to the refinery.

“Bauxite quality sets the opportunity; reactive-silica management determines how much of that opportunity the refinery can convert into saleable alumina.”

Technical Note

All numerical factors and screening thresholds in this article are indicative engineering benchmarks. Final design and operating decisions should be based on representative bauxite testing, the actual Bayer flowsheet, validated plant mass balances and site-specific economic evaluation. Soda values are expressed as Na₂O unless explicitly stated otherwise.

This article has been prepared for the Bayer Process Masterclass knowledge series by CETI Enterprises, India.

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. For any technical services, we may be contacted at https://ceti.co.in and available services have been published at our website: https://ceti.co.in