ARBOK-Aluminium

Overview

Every ton of alumina produced by the Bayer process leaves behind ~1.75 tons of red mud — caustic, fine-grained, pH 11–13, loaded with iron, titanium, aluminum, scandium, gallium, yttrium and the lanthanide series. Roughly 4 billion tons are already stockpiled across the planet behind earthen dams, ~150 million tons are added every year, and only 2–5% is put to any use. There is no commercial-scale remediation standard anywhere; the industry's answer for a century has been to build another pond, raise the dam and wait.

Ajka, Hungary, October 2010: an alumina plant's dam ruptured and ~1 million m³ of caustic mud at pH ~13 swept through the surrounding villages. 9 dead, 150+ chemical burns, 40 km² of farmland erased, the Marcal river killed in 24 hours, the plume reaching the Danube — Hungary's worst environmental disaster, and the slow, predictable failure of a single pond among more than 500 active red-mud ponds worldwide.

ARBOK processes the same mud into technical water, marketable caustic soda, and a dry mineral fraction carrying iron, alumina, titanium and rare-earth elements — no membranes, no chemicals, no reagents, no secondary waste. The framing shifts from cost center to revenue line.

Applications

• Primary use cases: fresh bauxite residue from operating alumina refineries, and legacy red-mud ponds (roughly 1 million tons per site).
• Outputs: technical/process water returned to the refinery loop; marketable caustic soda (NaOH) back into production; dry, inert, transportable mineral cake usable like clay in construction materials and as road sub-base; critical-materials concentrate (scandium, gallium, yttrium, lanthanides, plus iron, titania, alumina and silicate phases).
• Adjacent feedstocks named in the sources: steel slag from abandoned plants such as Bethlehem Steel (~8 million tons accumulated); mine water containing up to 12% salt and heavy-metal salts including radioactive contamination; tailings and acid mine drainage; brines.
• Geography: every alumina-producing country — Australia (largest stockpile, ~600 million tons), China (~500 million tons of new residue per year), India (~50 million tons/year and rising), Brazil (~30 million tons/year), Guinea (growing fast as Chinese refiners build out), Russia (RUSAL ponds), the Caribbean, the US (Alcoa legacy sites), Latin America (reference project).

Operating Principle

Red mud enters the module at 40–50% moisture. A proprietary capillary matrix pulls pore water and gas-phase volatiles by directional capillarity into the vacuum section. Water evaporates, condenses, and lands in the condensate loop together with caustic soda and the soluble metal fraction. The treated material leaves the unit as a dense, dry mineral cake at 15–18% moisture — inert, transportable, marketable. There are no membranes, no flocculant, no filtration, no chemicals of any kind; the only input is electricity. Nothing is discharged — closed balance, no tailings, Zero Waste Discharge.

The stream splits three ways: technical water, marketable caustic soda returned to production, and a dry mineral fraction (iron, alumina, titanium, rare earths) that is feedstock rather than waste.

Process conditions — the sources differ. The deep-dive text describes operation under mild vacuum at <0.5 atm and 60–95 °C, explicitly framed as moderate heat against Bayer process temperatures of ~250 °C already available on site. The LinkedIn and post texts describe deep vacuum at ambient temperature — no heat supplied — with heat recovery up to 98%.

> Расхождение в источнике: режим процесса описан двумя способами — «mild vacuum, <0.5 atm, 60–95 °C» (deep dive) и «deep vacuum, ambient temperature» (LinkedIn-пост и посты EN/RU). Обе формулировки приведены как есть.

Limits stated in the sources: ordinary thermal drying of red mud, the incumbent alternative, eats ~150 kWh per tonne and pays for no one on chemical effluent. Lime neutralization recovers nothing, reduces no mass, removes no heavy-metal load and emits ~0.08 tons of CO₂ per ton of mud neutralized. Reverse osmosis cannot clean this waste at all.

Key Parameters

| Parameter | Value |

|---|---|

| Feed moisture | 40–50% |

| Product cake moisture | 15–18% — dense, dry, inert, transportable, marketable |

| Vacuum / temperature (deep dive) | <0.5 atm, 60–95 °C |

| Vacuum / temperature (posts) | Deep vacuum, ambient temperature, heat recovery up to 98% |

| Energy (deep dive) | 80–120 kWh/ton |

| Energy (post EN/RU) | ~1–2 kWh/tonne |

| Energy (LinkedIn) | 2–3 kWh/t |

| Reference — ARBOK cold-cracking of lighter effluents | 2–3 kWh/t |

| Reference — ordinary drying | ~150 kWh per tonne |

| Nominal modular line throughput | 1,440 tons/day of input |

| Containerized module throughput (LinkedIn / RU article) | 200 tons/day per unit |

| JUMBO configuration | 4,800 tons/day |

| Water recovery (deep dive line) | ~41% — up to 593 t/day per line back into the refinery loop |

| Caustic soda recovery (deep dive line) | 7.5–8.0 t/day per line at typical Na₂O concentrations of 10 g/L |

| Consumables | None — no membranes, catalysts, reagents or chemicals; electricity only |

| Secondary waste | None — no tailings, no brine, closed balance (ZWD) |

| Footprint | 20–40 ft containers |

| Staff | 2–3 operators per shift |

| CAPEX | $2.5–3.5 million per module |

| Payback | 1.5–2.5 years (under 30 months) |

| Red mud composition (RU article) | Up to 20% iron, 5–10% titanium, 0.1–1% REE (lanthanum, cerium); nearly 28% caustic soda in the dumps |

| Red mud pH | 11–13 (pH >12 and ~13 cited for pond material) |

| Scandium in red mud | 50–150 ppm from key bauxite sources (Jamaica, Greece, Australia, India) — roughly 10–20x crustal average |

| Service life | [требует уточнения из базы] |

> Расхождение в источнике: удельная энергия названа тремя разными числами — 80–120 кВт·ч/т (deep dive), ~1–2 кВт·ч/т (посты EN/RU) и 2–3 кВт·ч/т (LinkedIn). Все три приведены как есть.

> Расхождение в источнике: извлечение каустической соды и воды дано двумя несовпадающими наборами — «7.5–8.0 т/сут на линию при Na₂O 10 г/л, возврат воды ~41% (до 593 т/сут)» и «~420 т соды и ~1,080 т воды в сутки при 1,500 т/сут и 28 % NaOH». Русская статья при тех же 1,500 т/сут даёт «~41 т каустической соды с одной установки при концентрации раствора 28 г/л». Все три приведены как есть.

Architecture and Components

Containerized modules (20–40 ft) installed directly at the pond edge or inline at an operating refinery, run by 2–3 operators per shift. Core element is a proprietary capillary matrix that moves pore water and volatiles by directional capillarity into the vacuum section; the condensate loop collects water together with caustic soda and the soluble metal fraction; the solid path discharges a dense dry mineral cake. Capacity configurations named in the sources: 200 t/day container unit, 1,440 t/day nominal modular line, and the ARBOK-JUMBO configuration at 4,800 t/day for legacy ponds. Stage-level component list: [требует уточнения из базы].

Advantages

• Technical: processes both fresh residue and accumulated stockpile; no membranes, flocculant, filtration or chemicals; no secondary waste; the dry cake is inert and transportable rather than a new pond feed. Reverse osmosis cannot process this stream at all.
• Versus lime neutralization: lime recovers nothing, reduces no mass, removes no heavy-metal load, emits ~0.08 t CO₂ per ton of mud neutralized, and still leaves a cake that needs another pond.
• Versus more pond capacity: a modern double-liner pond costs $15–25 per m³ of capacity, a typical refinery needs $200–400 million of new pond capacity over a decade of operation, and every pond eventually leaks.
• Economic: three revenue streams — soda, water, and alumina residue as building material; the refinery stops buying fresh caustic and cuts its carbon footprint; energy is a fraction of a percent of revenue.
• Environmental: removes the dam-failure exposure, eliminates groundwater contamination risk, frees and reclaims land — the existing field at the Latin American site is cleared in 2–3 years and fresh mud at 1,500 t/day is processed with no rise in field level, so the need for new ponds disappears.
• Strategic: recovers scandium, gallium (98% of which currently transits China), yttrium and lanthanides — all on the EU's 2024 CRMA list of strategic raw materials, none of which currently come out of red mud.
• Commercial: BOOM model (Build–Own–Operate–Maintain) on an off-take basis — the pond owner invests no capital, pays for hazardous-waste disposal, and buys water and caustic soda on site.

Integrations

Direct sibling entry for the same feedstock: ARBOK RED MUD. Platform and configurations: ARBOK-VC (Vacuum Cracking), ARBOK-JUMBO (4,800 t/day legacy-pond configuration). Recovered critical-materials basket links to ARBOK-Scandium-REE and ARBOK-Germanium-Gallium; the caustic soda product line to ARBOK-SODA. Adjacent mining-water streams named in the sources: ARBOK-CHEMILAKE-PURI, ARBOK-Copper-Waters.

Deployment & Operation

A containerized ARBOK unit is installed right at the pond or inline at the refinery and runs on electricity alone, staffed by 2–3 operators per shift. Two operating modes: inline processing of fresh residue (reference: 1,500 t/day with no rise in field level), and drawdown of the accumulated stockpile (~1 million tons per site; the Latin American reference field is cleared in 2–3 years, with part of the recovered water used to mix the old dried residue). Business model is BOOM on an off-take agreement: ARBOK builds, owns, operates and maintains; the client invests no capital, pays for hazardous-waste disposal and buys back water and caustic soda on site. Commissioning timeline: [требует уточнения из базы].

Technology Readiness Level (TRL)

TRL 9 — proven at full line scale (stated in the deep-dive source). Supporting evidence cited: a delivered project at a Latin American site where the task was to process fresh mud and dry it for road sub-base with no room left to expand the mud fields ("task solved"); field tests in Europe in 2023–2024 with independent expert assessments for the related mine-water application.

RUSAL engagement — documented (added 2026-08-17). No longer a bare claim. The evidence set is filed at Documents/TECHNOLOGIES/Arbok/Tests/PANKEMI 2016 — RUSAL before-after/:

• Rusal request pond water.xls — RUSAL's own filled questionnaire with live plant DCS tags (E:00ATS102.PV total soda, E:00AOX100.PV sodium oxalate, E:34HPVOL.MI pond volume, E:34HPVOLP.C pond percent full, e:34tothpr.pv pond return, E:28ATS105.PV 5th-stage soda), data window 2010-06-01 to 2010-08-09.
• Three PANKEMI LAB reports (Larnaca, Cyprus), customer ARBOK INTERNATIONAL CORP., samples in 2016-02-23: C093116 waste water before treatment, C093316 liquid component after treatment, C093216 hard component after treatment.
• Transmittal: results sent 2016-03-08 from alexander@aquatiq.us to Aleksandr.Grebennikov2@rusal.com, subject "Analyzing Rusal before and after treatment"; forwarded by D. Ponomarenko to M. Vischmidt 2023-10-31.

Pond liquor composition (RUSAL data): total soda 27.647 g/L; caustic 8.842 g/L; causticity 33.231 %; sodium oxalate 0.907 g/L. Manual averages over 13 points in 2010: caustic 8.842 g/L, TTS 26.667 g/L, causticity 33.231 %.

Pond overflow: volume grew 632,170 → 828,432 m³ and fill 77.0 % → 100.9 % between 2010-06-01 and 2010-08-03; design capacity ≈ 821,000 m³. The pond went past its design volume — this is what triggered the enquiry.

Incumbent cost, stated by RUSAL: evaporation of 160 m³/h consumes 63 t/h steam (steam factor 13 t steam per t fuel oil at $450/t) plus 2.49 MW electrical at $0.30/kWh → $2,180.77/h steam + $747.00/h power = $2,927.77/h, i.e. $18.299 per m³ evaporated. In specific energy that is ≈ 0.394 t steam/m³ (≈ 295 kWh thermal/m³) plus 15.56 kWh electrical/m³, ≈ 311 kWh/m³ thermal-equivalent. Against 0.72 kWh/m³ net this is 432× on energy and 102× on cost at $0.25/kWh. At 160 m³/h continuous (1,401,600 m³/year) the delta is ≈ $25.4 million per year on one site. RUSAL prices are 2010-vintage; re-tariff before any project model.

Product spec required by the client: "process to hose water composition (that is 5th stage liquor)" — 5th-stage soda averaged 102.910 g/L over the 2010 window.

Measured separation performance (PANKEMI, ISO 11885:2009 for metals): conductivity 96,745 → 54.0 µS/cm; TDS 67,722 → 37.8 mg/L; pH 8.05 → 5.98; methyl-orange alkalinity 49,743 → 13.3 mg/L as CaCO₃; Na 20,860 → 52.3 mg/L; Al 69.8 → 0.14 mg/L (99.799 %); Fe 75.5 → 0.12 mg/L; SO₄²⁻ 329 → 2.57 mg/L; P 17.6 → 0.17 mg/L; V 0.06 → <0.0088 mg/L; Cl⁻ 707 → 174 mg/L (only 75.389 %). Dry fraction in dry matter: moisture 22.8 %, pH 9.05, total CaCO₃ 48,000 mg/kg, CO₃²⁻ 28,800 mg/kg, Na 229,400 mg/kg, Al 75.9 mg/kg, Fe 670 mg/kg, P 52.0 mg/kg, V 0.87 mg/kg.

Two caveats on this set. Chloride is the one weak line (75.389 % versus 99 %+ elsewhere) and condensate pH fell to 5.98 — both point to volatile acid carrying over with the vapour; sulphate by contrast stayed behind (128× reduction), so the phosphogypsum balance claiming 18.5 % H₂SO₄ in the product stream runs on a different mechanism than carry-over and must not be explained by the same sentence. Aluminium mass balance does not close: 69.8 mg/L in, 0.14 mg/L in product water, ≈ 5.1 mg in the dry residue at ~68 g/L yield — ≈ 64.5 mg (92 %) unaccounted, so either a third fraction exists or the solids yield differs. Do not publish this as a closed material balance until resolved.

Jamaica + Aughinish set (added 2026-08-17). Filed at …/PANKEMI 2016 — RUSAL before-after/Jamaica + Aughinish/: Caustic soda Calculation Jamaica Project.xls, AQUATIQ — RUSAL Discussion (Executive Summary, Nov 2016).pdf (Strictly Confidential), plus the two Certificates of Analysis lifted from slides 9–10. Sample origin: Rusal Aughinish Alumina Ltd, Aughinish Island, Askeaton, Co. Limerick, Ireland.

Certificate of Analysis — sludge, 5 samples (C372216, C372316, C372416, C372516, C372616), w/w %: soda 48 / 49 / 44 / 47 / 46 (avg 46.8); TOC 1.0 / 1.2 / 0.8 / 0.9 / 0.9; total carbon 3.4 / 2.9 / 7.3 / 2.9 / 2.8; carbonate 12.0 / 8.5 / 32.5 / 10.0 / 9.5; alumina 9.1 / 8.7 / 8.7 / 9.9 / 8.9 (avg 9.06); sulphate 4.7 / 4.8 / 4.4 / 4.7 / 4.6; CaO 0.03 / 0.06 / 0.03 / 0.04 / 0.03; silica 0.06 / 0.06 / 0.06 / 0.08 / 0.07; iron 0.01 / 0.04 / 0.01 / 0.07 / 0.02; moisture 26.3 / 37.5 / 13.3 / 16.0 / 21.2.

This is not classic red mud. Iron 0.01–0.07 % and silica 0.06–0.08 % rule out bauxite residue (normally 30–40 % iron oxide). At 44–49 % soda with 8.5–32.5 % carbonate it is a sodium salt cake out of Bayer liquor, and the 9.06 % alumina in it is crystallised sodium aluminate. So 9.06 % is a measured Al₂O₃ figure for the Aughinish pond-liquor cake and may be used as such; the oxide composition of the solid bauxite residue itself is still not determined anywhere. [требует уточнения из базы]

Cake consumption per tonne of coagulant at Al₂O₃ 9.06 %: aluminium sulfate 1.90 t; PAC (30 % Al₂O₃) 3.31 t; sodium aluminate 6.87 t — the latter carrying 3.16 t of soda against the 0.488 t required, i.e. a sixfold surplus.

Certificate of Analysis — water: suspended solids <2 mg/L; pH 7.51; free residual chlorine <0.05 mg/L, total 0.03; SO₄ 2.7 mg/L; copper 133 µg/L; lead 15 µg/L; total iron 1.1 mg/L; aluminium 230 µg/L; total hardness 11 mg/L CaCO₃; bicarbonate alkalinity 18.4 mg/L CaCO₃; DO 8.5 mg/L (96.0 %); magnesium 0.61 mg/L; mercury 12.3 µg/L; calcium 2.9 mg/L; silicon as Si 93.5 µg/L.

This is recycle water, and that was the brief. The refinery needs water back in the process, not potable water, and the 2016 briefing frames it exactly so — safe water flow-back and re-use. The client's own written requirement was "process to hose water composition (that is 5th stage liquor)", i.e. wash water; this result meets it.

Output quality is set by the specification, not by a technology ceiling. The same unit is taken to whatever level is bought — recycle water, potable, or injection-grade — the difference being stage count and specific energy, not principle. Fitting potable-grade treatment at an alumina refinery is pointless: nobody there pays for it and nobody needs it.

Rule for materials: cite the Aughinish certificate as recycle water and do not hold it against drinking limits. Mercury 12.3 µg/L, aluminium 230 µg/L and lead 15 µg/L sit above potable thresholds because a potable duty was never set. For the potable claim there is separate evidence — the Cyprus PANKEMI run on RUSAL samples (Al 0.14 mg/L, Fe 0.12 mg/L). Different samples, different labs: never merge the two result sets, but it is the second pair that speaks to drinking grade.

Jamaica project calculation: red mud 1,440 t/day at 28 g/L → 40.32 t/day caustic at $455/t = $18,345.60/day; slurry lake 1,000,000 t at 28 g/L → 28,000 t caustic at $456/t = $12,768,000; clean water 1,400 t/day at $15/t evaporation saving = $21,000/day; daily throughput 2,880 t/day; lake drawdown 695 days; 29 units at 100 t/day in the file (the narrative uses 7 units for 1,440 t/day, i.e. the standard 200 t/day module); tax 33 %. All arithmetic in the file checks out.

Concentration is 28 g/L, not 28 %. At ~1.03 density that is 2.7 % by mass. The figure "~420 t/day of soda at 1,500 t/day and 28 % NaOH" that previously stood in this card and in the article set is overstated 10.4× and must be removed everywhere. The Russian article's "~41 t of caustic soda per unit at a solution concentration of 28 g/L" was the correct one. Separately: the 2010 questionnaire gives total soda 26.667 g/L against caustic 8.842 g/L at 33.231 % causticity — the Jamaica 28 g/L matches total soda, not caustic. Computed on caustic the yield is 12.73 t/day, not 40.32, a 3.2× gap. Fix which product is actually sold before any project model.

Caustic price — two incompatible values. The Jamaica file uses $455–456/t (market dry caustic). The business narrative states RUSAL was prepared to buy recovered caustic at $3,000/t. Ratio 6.6×, and the whole economics turns on it: at $455/t the site yields $39,346/day, $14.4 M/year, $27.32 per m³ of pond water; at $3,000/t it yields $141,960/day, $51.8 M/year, $98.58 per m³. $3,000/t is far above world caustic pricing and most likely reflected delivered fresh reagent into Jamaica with logistics and hazardous handling rather than a commodity quote. Not to be used in public material until confirmed in writing by the client.

Specific energy — historical spread. The November 2016 briefing claims net power consumption below 3 kWh/m³. Canonical today is 0.72 kWh/m³ net. Not a contradiction but different equipment generations; cite the 2016 number as historical only.

2016 briefing, corroborating record: readiness to serve any RUSAL plant; successful test results on RUSAL samples; six-year R&D programme by Israeli scientists; B.O.O.M. model. Timeline: 2009 first prototype and testing of evaporation, purification and compound separation; 2010 R&D base in Cyprus; 2012 molecular-separation concept validated and full specification complete; 2014 desalination and supply pilot in Cyprus with water delivery from November 2014; 2015 approval by the PAWD Chairman in the Philippines and comprehensive testing by the Cyprus Ministry of Agriculture leading to national acceptance; 2016 contract for 1 Mn m³/year with Nature's Spring (Philippines), contract for 5.5 M m³/year with the Cyprus Ministry of Agriculture & Resources, and a five-sample test for RUSAL. Process chain as stated: Phase Change → Separation → Concentration → Extraction → Purification.

Why the project died: US sanctions against Russia — the RUSAL office in Cyprus was relocated to Moscow and all Cyprus projects, this one included, were closed.

Not for public content: equipment and project cost — CAPEX is not mentioned in any form; the payback and IRR figures carried in the file — canonical payback wording is 5–7 years and no other value is quoted; the $3,000/t caustic price until documented by the client; the Option A / Option B project valuations (negotiation material).

Market Potential

Roughly 4 billion tons of red mud are stockpiled worldwide behind more than 500 active ponds, with ~150 million tons added every year and only 2–5% put to use. Australia holds the largest stockpile at ~600 million tons; China generates ~500 million tons of new residue per year; India ~50 million tons/year and rising; Brazil ~30 million tons/year; Guinea is growing fast. Russia's RUSAL ponds have been studied since the early 2010s, and their scandium content was one of the reasons the West paid attention to Russian REE supply long before 2022.

The critical-materials angle: bauxite from Jamaica, Greece, Australia and India concentrates scandium at 50–150 ppm in red mud, 10–20x the crustal average. A single 1.5 million ton/year refinery's residue carries 75–225 tons of scandium oxide per year against a global scandium market of ~30 tons/year. If even 10% of European refinery residue were processed through ARBOK cake, Europe could meet its entire defense-grade scandium demand from its own waste streams without opening a single new mine. The same physics catches gallium (radar, 5G, defense optoelectronics — 98% currently transiting China), yttrium (laser systems, superalloys) and the lanthanide pivot points of modern motors and sensors.

Regulatory drivers: the EU CRMA mandates that 25% of strategic raw material consumption be sourced from secondary, waste-derived streams by 2030, and after January 2027 tightens reporting on hazardous mining residues; the US is moving the same way under the proposed reform of the Bevill Amendment; the US Energy Act of 2020 and the IRA together fund $500 million of REE-from-waste extraction grants. China — which already controls 70% of the global REE supply chain — has been extracting scandium from red mud at industrial scale at three sites since 2022.

Typical Project Economics

Pricing basis used in the sources: NaOH ~$500/ton, water $2/m³ (or $2/t), ~330 operating days per year.

• Inline at an operating refinery, 1,500 t/day fresh residue (~28% NaOH content): ~420 tons of recovered caustic soda per day and ~1,080 tons of recovered water per day → roughly $70 million per year per refinery. The plant stops buying fresh caustic; carbon footprint drops.
• Legacy pond, ARBOK-JUMBO at 4,800 t/day: ~134 tons of recoverable soda per day plus water → ~$24 million per year per site, year after year until millions of tons of stockpiled mud have been processed into transportable cake. A pond that has sat as a $2 billion liability for 30 years becomes a $500 million revenue stream over its remediation life.
• CAPEX and payback: $2.5–3.5 million per module; payback 1.5–2.5 years (under 30 months). Energy is a fraction of a percent of revenue.
• Metal-value scenario (RU article), 1 million tons of red mud processed per year: up to $1 billion of annual revenue per project — 200,000 tons of iron (20% of volume) ≈ $120 million at $600/ton; 50,000 tons of titanium (5% of volume) = $500 million at $10,000/ton; 5,000 tons of rare earths such as lanthanum and cerium = $250 million at $50,000/ton.
• Adjacent case: steel slag at Bethlehem Steel, ~8 million tons accumulated → up to $258.75 million per year.
• Cost of the incumbent alternatives: chemical neutralization of 1 million tons of red mud can cost up to $10 million annually; double-liner ponds $15–25 per m³ of capacity, $200–400 million of new capacity per refinery over a decade; ordinary drying ~150 kWh per tonne.
• Water cost: the RU article states purified water at $10–15 per ton as the figure that makes the process economically viable, since the water is reused rather than evaporated to the sky.

> Расхождение в источнике: вода фигурирует и как стоимость получения ($10–15 за тонну, русская статья), и как цена реализации в экономической модели ($2/т). Обе цифры приведены как есть.

Risk Factors

• Process conditions, specific energy and recovery rates are stated inconsistently across the source documents (see the discrepancy blocks in sections 4, 5 and 12) — an authoritative parameter set must be fixed from the base before any project model is built.
• Red mud is a hot, heavy stream: the deep-dive source itself notes that 80–120 kWh/ton is far above the 2–3 kWh/t the platform runs on lighter effluents, justified by the mass and load.
• The feed is caustic (pH 11–13, nearly 28% caustic soda in the dumps) and carries heavy metals (arsenic, cadmium, chromium) plus natural radioactivity (uranium, thorium) — handling, worker safety and product qualification are governed accordingly.
• Recovery of the scandium / gallium / yttrium / REE basket is presented as latent potential in the cake ("none of them currently come out of red mud; all of them could"), not as a demonstrated separated product with stated grades — [требует уточнения из базы] for the refining step and yields.
• Market and offtake: the global scandium market is only ~30 tons/year, so a single refinery's 75–225 t/year of contained scandium oxide has no existing outlet at that scale.
• Competitive: China already extracts scandium from red mud at industrial scale at three sites and controls 70% of the REE supply chain.
• Institutional inertia — the sources describe a century of "build another pond" practice, accounting that hid red mud as stable inventory, and resistance from expert opinion anchored in older technology.
• Regulatory timing risk: the pond-economics collapse depends on the EU CRMA reporting tightening after January 2027 and on the proposed US Bevill Amendment reform actually landing.

> Расхождение в источнике: масштаб аварии в Айке описан по-разному — «8 villages» (deep dive) и «3 деревни / three villages» (русская статья и её английский перевод). Обе цифры приведены как есть. Так же расходятся выход шлама на тонну глинозёма (~1.75 т в deep dive и постах против 1–1.5 т в LinkedIn-версии) и годовой прирост (~150 млн т против ~170 млн т).

Related Technologies

ARBOK RED MUD · ARBOK-JUMBO · ARBOK-VC (Vacuum Cracking) · ARBOK-Scandium-REE · ARBOK-Germanium-Gallium · ARBOK-SODA · ARBOK-CHEMILAKE-PURI · ARBOK-Copper-Waters