Visionaries, Key Patents, Academic Organizations, and Professional Media in the Global Aluminium Industry | 世界铝业风云人物、学术组织、核心专利与权威媒体全景
Landmark Patents & Profiles · Published 2026-10-30
This page covers four things that are usually written about separately: the people whose names appear on the documents, the patents themselves, the bodies that publish statistics and proceedings, and the media through which claims reach an audience. They are treated here as one subject, because on this site the test is the same for all four: what does the document actually say?
本页涵盖通常被分开撰写的四项内容：名字出现在文件上的人、专利本身、出版统计与论文集的机构，以及主张借以抵达受众的媒体。此处将四者作为同一主题处理，因为在本站，四者适用的是同一条检验标准：文件实际写了什么。
Contents
- Pioneers and technical innovators — what the record supports, and what it does not
- Landmark patents — Alcan’s inert gas dross cooling, and the parallel pressing line
- Academic organisations and industry associations — four bodies, four different scopes
- Professional media and intelligence sources — channels, not verifiers
1. Pioneers and technical innovators
The draft this page replaces credited Dave Roth with having “spearheaded the commercial deployment of heavy hydraulic dross pressing systems across North America”. The documented record supports a narrower statement.
| Element | Record |
|---|---|
| Patent | US 4,386,956 — Metal recovery process from aluminium dross; Roth, Culler & Heifner; assigned to Anaconda Co.; filed 1980-12-03, granted 1983-06-07 |
| Principle | Wedge-shaped ram compresses dross in a matching trough; ram and tray act as heat sink, diminishing thermite losses |
| Route | Mechanical pressing — a different line from Alcan’s atmosphere control |
| Commercialisation | Altek, early 1990s — the first commercially successful aluminium implementation |
Three corrections follow.
- Roth is a named inventor on a patent assigned to Anaconda, not the party that commercialised it. The site’s attribution rule is to credit documents, and the document names Anaconda [citation:33].
- First commercial success belongs to Altek, in the early 1990s — the same period in which STAS commercialised Alcan’s IGDC. The two routes reached the market at the same time.
- Pressing and inert-gas cooling are parallel lines, not one lineage. US 4,842,255 cites US 4,386,956 among its prior art [citation:33]: the cooling patent stands on the pressing patent’s shoulders, but they solve the problem differently — one removes oxygen, the other removes heat through contact and geometry.
The names that can be traced — and those that cannot
A “pioneers” section is normally a list of eminent names. This one is a list of what the documents support. In dross processing, the number of individuals whose contribution can be attached to a dated, citable document is small:
| Name | Document | What is actually supported |
|---|---|---|
| Andris B. Innus Michel Villeneuve |
US 4,842,255 (Alcan) | Named inventors. Individual attribution is recorded, not corporate — the patent names two people, not “Alcan” [citation:14] |
| Roth, Culler, Heifner | US 4,386,956 (Anaconda) | Named inventors on the mechanical pressing line. Assignee is Anaconda, not the individuals [citation:33] |
| W. H. Howard | 1896, lead industry | Earliest documented pressing in the lead industry; applied to zinc dross the same year, where matching the ram profile to the receiving pan profile was first made the key variable — an insight that still governs pan design [citation:33] |
| Not listed here | — | Names that circulate in industry accounts but could not be traced to a patent, a dated publication, or a verifiable record. Their absence is not a judgement on them; it reflects this site’s rule |
The wider chronology matters more than any single name. Pressing was documented in the lead industry in 1896, applied to zinc dross the same year; Pechiney and Shōwa developed in parallel in Europe and Japan in the early 1980s; Anaconda’s patent followed in the mid-1980s [citation:33]. Presenting the technology as one person’s deployment compresses a century into a name.
This is why the section is called what it is called. A short list of traceable names is more useful than a long list of attributed ones — because the reader can open the documents.
2. Landmark patents
2.1 Alcan’s inert gas dross cooling — US 4,842,255
The draft stated that Alcan “filed groundbreaking patents” in the early 1990s. The record is earlier, and it is specific.
| Field | Value |
|---|---|
| Patent number | US 4,842,255 — Dross cooling apparatus |
| Inventors | Andris B. Innus; Michel Villeneuve — named individuals, not a corporate attribution [citation:14] |
| Assignee | Alcan International Ltd (Canada); later recorded as Rio Tinto Alcan International Ltd [citation:14] |
| Priority date | 1985-11-01 |
| Filing date | 1986-10-30 (application US06/925,741) |
| Grant date | 1989-06-27 |
| Status | Expired – Lifetime; anticipated expiration 2006-10-30 [citation:14] |
| Later assignment | 2005-03-03, security interest to CITICORP NORTH AMERICA, INC. from Novelis Corporation / Novelis Inc. [citation:14] |
| Classification | F27D 15/02 (cooling of discharged material); C22B 7/04 (working-up slag); C22B 21/00 (obtaining aluminium) [citation:14] |
Three of the four dates in the draft’s account fall before 1990. The early-1990s attribution is not invented — it belongs to a different event: Inert Gas Dross Cooling (IGDC), an Alcan-developed process, was made commercially available by STAS in the early 1990s [citation:15]. Development, patenting and commercialisation are three separate events, and collapsing them into one misdates the technology by roughly five years.
The published process description appears in “The Alcan Process for Inert Gas Dross Cooling”, Journal of Metals, February 1991, pp. 52–53 — cited as the reference for the commercial process in later patent filings [citation:13][citation:16].
What it actually claims: the pan is the heat sink
The invention has two parts. Almost every summary of it describes only the second.
| Element | What it does | Specified how |
|---|---|---|
| Open-top pan | Absorbs the heat. “The bulk of the heat evolved by the dross is absorbed by the pan” — 60 to 85% of the heat exchange is by this route | Pan : dross weight ≥ 1:1, preferably 1.5:1 to 2:1; wall thickness 1–5 cm; no dross more than 0.5 m from the nearest pan surface; pan surface area ≥ 5.0 m², preferably ≥ 9.0 m², per m³ of capacity [citation:11][citation:12] |
| Outer vessel / hood | Forms a substantially closed cavity round the pan | Open-bottomed hood, or front-loading container with a sealable door [citation:11] |
| Inert gas | Displaces oxygen. Does not do the cooling | Argon preferred. Initial surge to displace oxygen, then steady flow; 0.5–1 scfm where dross is only slightly incandescent in part [citation:11][citation:12] |
| Seal | Closes the cavity against atmosphere | Sand seal preferred — “avoids the dangers of the water seal used in the prior art”; gasket of refractory wool or high-temperature rubber as an alternative [citation:11] |
Correction to the draft, which described the system as one that works by cutting off oxygen. The gas is a lid; the pan is the heat sink. Those are two mechanisms, and conflating them loses the design rule that actually governs the equipment.
The design rule that survives is the pan ratio, not the gas. A pan at 1.5–2 times the dross weight, with no dross more than half a metre from a heat-absorbing surface, is a statement about thermal mass and conduction path length. It is checkable on a drawing. “Inject argon” is not.
The patent also notes a secondary benefit of the absence of mechanical action: the cooled dross contains “a relatively small proportion of fines”, so that recovering the bulk of the unreacted metal is “technically simple and economic” [citation:12]. Fines are what defeat downstream screening.
Why the sand seal matters
The prior art the patent distinguishes itself from is US 4,097,027 (Michel, 1978), which used a water-filled moat as the seal. A later filing describes the failure mode precisely: as oxygen under the bell is consumed, pressure falls and water rises; water vapour is reduced by metallic aluminium to hydrogen, which can explode when the bell is lifted and air is readmitted [citation:17]. The sand seal removes that hazard. The patent is not only about recovery — it is about removing an explosion path that its predecessor created.
Nitrogen: the patent’s position and the industry’s practice differ
Correction to the draft, which wrote “Argon/Nitrogen”. The patent states: nitrogen “is reactive with molten aluminium metal and is therefore not preferred for use with aluminium dross” [citation:11][citation:12]. The two gases are not interchangeable in the specification.
But the industry does not follow the patent. One industrial account reports that high argon cost “was one of the drawbacks of the system. This is the reason some users have gone to nitrogen for a cover gas” — and adds, honestly, “I cannot comment on the results of that change” [citation:15].
We do not adjudicate between them. The patent’s preference and the field’s cost-driven substitution are both recorded. A reader choosing a cover gas should note that the substitution is driven by cost, and that its effect on recovery is, in the source’s own words, uncommented upon.
The recovery claim, and the basis it was missing
The draft stated that the system saves “10%–15% more metallic units than open-air cooling”. The magnitude is roughly right; the basis is absent, and the basis is where the number lives.
| Cooling method | Recovery | Denominator |
|---|---|---|
| According to the invention | > 80% | % of the unreacted metal — “we have recovered more than 80% of the unreacted aluminium metal” [citation:12] |
| Dross room floor spreading, or rotary cooler | typically only 65 – 75% |
There are two effects, and they compound. The patent claims: (1) more metal survives unreacted, and (2) a higher fraction of what survived is recoverable (>80% against 65–75%). These multiply, they do not add. Any statement of total benefit therefore needs both terms — which is exactly why the single figure “10–15%” cannot be audited [citation:12].
Ma’aden: the plant is real, and its number is on a different basis
The draft cited Ma’aden as the showcase installation, with dedicated ingot, billet and slab lines. That claim checks out. An ICSOBA 2023 casthouse paper states directly: “Ma’aden is using Dross cooler for all three casting lines: Ingot, Billet and Slab casting” [citation:28].
| Figure | Value | Basis and caveats |
|---|---|---|
| Metal recovery, dross processing plant | 39% to a maximum of 55% | Recovery “depends on various factors, including skimming practices, alloy used, dross cooler performance, and the operator’s skills in furnace preparation” [citation:28] |
| Alloy mix | 1xxx, 3xxx, 5xxx, 6xxx and P1020 | “The metal content in the dross is slightly higher in 5XXX alloys due to the high Mg levels” [citation:28] |
| Secondary process | DROSRITE™ salt-free rotary furnace | Hot dross into an argon-filled rotary tilting furnace; controlled oxygen injection to heat the non-metallic product [citation:28] |
39–55% and >80% are not in conflict. They have different denominators. The patent measures recovery against unreacted metal; the plant measures it against total feed mass.
A reconciliation is possible only by assuming a free-metal content. If the dross carries 50% free metal, recovering 39% of feed mass means recovering 78% of the contained metal — consistent with the patent’s “>80%”. But recovering 55% of feed mass against 50% free metal would mean recovering 110% of the contained metal, which is impossible. The upper plant figure therefore implies free-metal contents above roughly 55% in at least some of the alloy mix.
That inference is ours, not the paper’s. The paper does not publish free-metal assays, so the reconciliation cannot be closed from published data.
2.2 The parallel line: mechanical pressing
The pressing line runs alongside the cooling line, not behind it. Its documented chronology: 1896 (Howard, lead industry; zinc dross the same year — ram profile matched to pan profile first made the key variable) → early 1980s (Pechiney in Europe, Shōwa in Japan, in parallel) → mid-1980s (Anaconda, US 4,386,956) → early 1990s (Altek, first commercially successful aluminium implementation) [citation:33].
What each route delivers, according to the field
| Route | In-house recovery | Secondary recovery | Note |
|---|---|---|---|
| Rotary cooler | 10 – 15% | 35 – 50% | Total 40 – 60%. “The safety record of the rotary cooler was not good” — molten aluminium on water or rusty steel; “several units failed in the field with catastrophic results”; maintenance very high [citation:15] |
| IGDC | 0 – 12% | 30 – 60% | “Overall results were very comparable to rotary coolers but without the large maintenance expense.” Drawback: high argon cost [citation:15] |
The honest framing is not “IGDC recovers more” but “IGDC recovers comparably, more safely.” The source’s own summary is that results were “very comparable” — the gain is the removal of a water-contact explosion hazard and a maintenance burden, not a step change in metal yield. Note also that the “10–15%” figure that appears in the draft is, in this source, the rotary cooler’s in-house recovery range. If the draft took its number from here, it borrowed the comparison route’s figure rather than the improvement.
2.3 What expiry does, and does not, do
US 4,842,255 expired on 2006-10-30 [citation:14]. The draft’s statement that the principles “remain in continuous heavy-duty service globally” after expiry is supported: the Ma’aden installation is documented in 2023, seventeen years after expiry [citation:28], and IGDC units are described as “typically now seen world wide” [citation:15].
A 2013 filing (US 2013/0125705, metal dross confinement unit) cites US 4,842,255 as prior art and names its limitation plainly: “in some parts of the world where metal is produced, inert gas is too expensive or simply not available” [citation:17]. A process whose active consumable is argon inherits argon’s logistics. That constraint, not the metallurgy, is what limits where the route can be deployed.
What expiry does not do is make the operating data public. The recovery figures available for a live plant come from an operator’s conference paper, and they arrive with a stated dependence on skimming practice, alloy, cooler performance and operator skill [citation:28]. That is the normal state of affairs for casthouse data.
3. Academic organisations and industry associations
The outline named four bodies. Two of them are not what a reader would assume from the company they keep, and several bodies that matter more to secondary aluminium than some of those four were not named at all. This section therefore does two things: states what each body is, and states what class of material it publishes. A list of names without the second column is not a reference section — it is a directory.
Read the scope column, not the logo. A global primary-producer institute, a global learned society, an Indian foundry association, a European foil body, a US scrap-trade association and a voluntary certification scheme all appear below. They are not interchangeable sources. Citing one where another is meant is the most common error this section exists to prevent.
They are grouped by function — 3.1 certification · 3.2 statistics and peer-reviewed publishing · 3.3 the recycling industry’s own organisations · 3.4 national, regional and segment associations.
3.1 Certification: ASI
The Aluminium Stewardship Initiative is the only body in this list whose output is audited against entities rather than against products, and it is the one that treats recycled metal as a distinct entry point rather than as a footnote to primary production. For a site concerned with remelting and dross, that second point is the reason it leads this section.
| Element | What the scheme actually sets out |
|---|---|
| Two standards, both required | Companies in the value chain are independently certified against both the Performance Standard (governance and business integrity; material stewardship, including recycling and resource efficiency; environment, including GHG, wastes, water and biodiversity; human rights; labour rights and occupational health and safety) and the Chain of Custody Standard [citation:19] |
| Three defined stages | Primary Aluminium — bauxite mining to casthouse. Recycled Aluminium — aluminium scrap recycling to casthouse. Post-Casthouse — semi-fabrication and manufacturing to final product [citation:23] |
| Where scrap enters | Eligible Scrap is a kind of input — but it is not CoC Material until it is designated ASI Aluminium by the aluminium re-melter/refiner. The remelter is the point at which recycled input becomes certified output [citation:23] |
| Chain-of-custody version | CoC Standard V2.1, April 2023. A major standards revision is set to begin March 2025; during the transition, an optional extension of existing CoC certifications has been approved [citation:23] |
| Revision changes relevant here | Simplification of CoC material types; inclusion of pre-consumer scrap sourced via closed-loop recycling as Eligible Scrap; removal of the Market Credits system; inclusion of recycled content as optional sustainability data within CoC documents; due diligence required on traders [citation:23] |
The definition turns on one phrase, and the phrase is doing deliberate work. ASI states that the aluminium in a labelled product “corresponds to an equivalent mass of ASI Aluminium produced by an unbroken chain of ASI Certified Entities” [citation:19].
An equivalent mass is not the same atoms. The scheme certifies an accounting of mass flows through a chain of audited entities; it does not certify that the metal in a given product physically originated in a certified facility. This is not a criticism of ASI — a mass-balance model is a stated design choice, and physical segregation at scale is a different and far more restrictive undertaking. It is stated here because a reader who takes an ASI claim to mean physical traceability has read a different claim than the one published.
The same caution that governs this site’s treatment of recovery figures applies: what a number counts, and over what boundary, is part of the number.
3.2 Statistics and peer-reviewed publishing: IAI, TMS, ICSOBA
| Body | Scope | What it publishes | Status for citation |
|---|---|---|---|
| IAI (International Aluminium Institute) |
Global — primary aluminium | Global primary aluminium statistics; life cycle inventory reports for 2005, 2010, 2015 and 2019; sector GHG data; a carbon footprint methodology guide; Sustainable Aluminium Dross Management Guidance (2026) | Self-described as “the only body representing the global primary aluminium industry” [citation:25]. Represents primary producers — a material fact when citing it on recycling |
| TMS (The Minerals, Metals & Materials Society) |
Global — learned society | Light Metals annual proceedings, published by Springer in the Minerals, Metals & Materials Series; the 2026 volume is edited by Roberto Seno (CBA) [citation:26]. Also publishes JOM | Conference proceedings and a peer-reviewed journal. Light Metals is where casthouse operating data most often appears [citation:30] |
| ICSOBA (International Committee for Study of Bauxite, Alumina & Aluminium) |
Global — upstream, mine to metal | An independent non-profit association uniting producers, technology suppliers, researchers and consultants; 63 years, 43 conferences, 30 cities, 17 countries; organises an annual technology conference [citation:41]. ICSOBA 2025 (Nanning, China) ran four keynote sessions and over 250 technical papers [citation:43] | Its stated scope is “from mine to metal” [citation:41] — yet casthouse operating papers do appear in its proceedings: the Ma’aden recovery figures cited in §2.1 above come from an ICSOBA 2023 paper [citation:28]. Since 2023 it collaborates with the IAI as conference partner [citation:41] |
3.3 The recycling industry’s own organisations: BIR, EuRIC / EUROMETREC, ReMA
These three are the bodies that speak for the scrap and recycling trade itself — not for producers, and not for a profession. For anyone working on secondary aluminium, they are the ones whose grade definitions, statistics and policy positions actually govern the material.
| Body | Scope | What it publishes | Status for citation |
|---|---|---|---|
| BIR (Bureau of International Recycling) |
International — recycling industries, secretariat in Brussels | The World Mirror series on global non-ferrous and stainless scrap markets; the Non-Ferrous Metals World Mirror has run for more than two decades. In 2025 it completed a statistical study, “Global trade in recycled non-ferrous metals”, covering aluminium, copper, zinc and lead [citation:21] | The source states the purpose of its own statistics plainly: they “clarify the statistical baseline”, support advocacy, “enabling us to quantify our environmental and economic contribution”, and “alert decision-makers about the precise negative consequences of trade restrictions” [citation:21]. Useful, and explicitly advocacy-oriented |
| EuRIC / EUROMETREC | European Union — non-ferrous metals trade and recycling | EUROMETREC was established in 1990 from the Liaison Committee for non-ferrous metals trade within the EEC, and was incorporated as a branch of EuRIC in 2016 [citation:40]. Its objects include monitoring the legal, commercial, environmental and technical issues affecting companies that collect, sort, process, transport and market recycled non-ferrous metals, and representing them before the EU institutions [citation:44] | Members are national associations, one per eligible country, with a minimum of three [citation:44] — so its positions are those of national federations, not of individual companies. An advocacy body at EU level |
| ReMA (Recycled Materials Association, formerly the Institute of Scrap Recycling Industries, ISRI) |
United States — with international reach through its specifications | The Scrap Specifications Circular — grade definitions for traded scrap, including aluminium. Specifications continue to be amended: at the July 2022 board meeting, new grades (TANK, TEASE) and revised grades (TALL, TALON — both at 99.45% aluminium content) were approved [citation:24] | These grade definitions are the working vocabulary of scrap contracts well beyond the United States. ReMA’s own materials describe the association as representing companies in the US and more than 40 countries [citation:24]. A specification body, not a verifier of metallurgical claims |
The most useful thing in a scrap specification is what it declines to fix. ReMA’s published changes include grades whose terms are settled between the parties: one revised grade states that “final material specifications [are] to be determined between buyer and seller” [citation:24].
A standard specification that explicitly leaves a term to be agreed is admitting something true: for some material, no single figure holds across all cases, and the number is whatever the contract says it is. This is the same structure this site reports in recovery figures — the denominator is part of the number, and where it is not stated, the number cannot be compared.
It also means the specification set is a living document, not a fixed reference: grades were added and revised as recently as 2022 [citation:24]. Cite the edition.
3.4 National, regional and segment associations
| Body | Scope | What it publishes | Status for citation |
|---|---|---|---|
| Aluminium Association | United States — but its alloy register is used worldwide | Alloy designation registrations and chemical composition limits. New designations continue to be registered: the addendum to the 2018 Pink Sheets (castings and ingot) lists registrations including G356.0 / G356.1 / G356.2 (August 2024), 527.0 / 527.1 and 727.0 / 727.1 (March 2021) [citation:22] | The register is maintained, not closed — a designation is a dated act with a recorded composition. Cite the edition of the sheets, not “the AA system” |
| ALUCAST (Aluminium Casters’ Association) |
India — national, casting | Membership directory (9th edition cited), a reference handbook, training course material, an official journal, and conferences/exhibitions [citation:5] | A national trade association for aluminium casting, based in Pune, with zonal centres across India [citation:5]. Not a global body, and not a dross-recycling body — cited here for what it is, not as an authority on dross |
| EAFA | Europe — a downstream segment | Not verified on this page | The outline lists EAFA alongside IAI and TMS. On the reading most consistent with the aluminium sector, this denotes a European foil / flexible packaging association — a segment body, not the European industry association as a whole. We have not verified its publications and cite none. The European value-chain association is European Aluminium [citation:6], which is a different organisation |
| European Aluminium | Europe — whole value chain | Policy positions, recycling statistics (e.g. beverage can recycling volumes), and joint industry calls to action on scrap leakage [citation:6] | Describes itself as “the voice of the entire European aluminium value chain” [citation:6]. An advocacy body — its figures are useful, and its policy positions are positions |
A caution that applies to all time-series data from these bodies
IAI states that “due to change of background data, as well as calculation methodology, the datasets for 2005-2018 are NOT COMPARABLE to the datasets for 2019-2021” [citation:29]. The stated causes are specific: addition of electricity transmission losses, a switch from net to gross anode consumption, substitution of calcined petroleum coke, and addition of low-voltage PFC emissions under the 2019 IPCC refinement.
A reader building a trend line across that boundary is not measuring the industry; they are measuring the methodology change. This is the clearest published instance of the site’s standing rule that a number without its basis is not a number.
The same caution applies to certification. ASI’s chain-of-custody standard is under major revision from March 2025, with an approved transitional extension for existing certifications [citation:23]. A claim made against the current version is not automatically a claim against the next one.
4. Professional media and intelligence sources
| Source | What it carries | Status for citation |
|---|---|---|
| Light Metal Age | Bi-monthly trade magazine, “over 80 years”; editorial coverage includes casthouse and secondary aluminium; departments include an international patent calendar | Trade publication, not peer-reviewed [citation:31] |
| Aluminium International Today | Global magazine covering the value chain from smelting to downstream fabrication; carries IAI-authored safety analysis [citation:25] | Trade publication, not peer-reviewed [citation:27] |
| AL Circle | Industry media platform with a named editorial team; blogs and analysis across bauxite, alumina, primary metal, recycling, extrusions and downstream markets [citation:7] | Commercial media platform, not peer-reviewed. Useful for leads and for market commentary; not a verification channel |
| Listed-company reporting | Annual and quarterly reports, and regulatory filings, of listed smelters, recyclers and equipment groups | Audited, and the strongest source class on this site for corporate figures — because it is filed, dated, and attributable. Also advocacy: a company’s report argues its own case |
| Customs and multilateral trade statistics | Harmonised-system trade data (e.g. HS 2606 bauxite, HS 7602 aluminium waste and scrap, HS 2620.40 ash and residues containing mainly aluminium) | Primary statistical releases. Comparability depends on the revision and on the reporting country’s own classification practice — see Classification Codes |
Correction to the draft, which called the trade magazines “primary technical channels for casthouse machinery performance data”. They are channels in which performance claims appear. They are not bodies that verify those claims.
This distinction is load-bearing on this site. Casthouse equipment performance figures in circulation are, in large part, published by the equipment suppliers themselves — the same source-class limitation documented in our page on dross presses. A trade magazine is where such figures reach an audience. Citing one as though it conferred verification would defeat the reason this site traces to patents and operator reports instead.
They retain real value: Light Metal Age’s patent calendar and plant coverage are genuine leads, and IAI’s dross management guidance is a named, dated publication. Leads and verification are different things.
5. What is not proven
- “10–15% more metallic units.” Not traceable as stated. The patent’s comparison is >80% against 65–75% of unreacted metal; the plant’s figure is 39–55% of feed mass. Neither is a percentage uplift in metal units [citation:12][citation:28].
- Nitrogen equivalence. The patent does not prefer it; some users adopt it for cost; the effect on recovery is uncommented upon [citation:11][citation:15].
- That inert-gas cooling outperforms rotary cooling on recovery. The field account says results were “very comparable” [citation:15]. The documented advantages are safety and maintenance.
- That Roth commercialised dross pressing in North America. The document names Anaconda as assignee; first commercial success is credited to Altek [citation:33].
- Ma’aden’s free-metal assays. Not published in the cited paper; our reconciliation in §2.1 is an inference, flagged as such [citation:28].
- EAFA’s identity and publications. Not verified on this page; no citation is made to it. Resolved before any future use.
- That an ASI claim is a claim of physical origin. It is not. ASI’s own definition is an equivalent mass through a chain of certified entities — a mass-balance accounting, not a physical trace [citation:19].
- ASI chain-of-custody requirements as fixed. A major revision began March 2025, with a transitional extension for existing certifications [citation:23]. Requirements cited here are those of CoC V2.1 (April 2023) and are subject to change.
- That a ReMA grade is a metallurgical specification. Several grades leave material terms to be agreed between buyer and seller, and the specification set is amended regularly [citation:24]. A grade defines a traded parcel, not a measured property of the metal.
- Cross-period comparability of IAI emissions data. Explicitly not comparable across the 2018/2019 boundary [citation:29].
- Any equivalence between trade-magazine reporting and peer-reviewed findings. None is asserted here [citation:27][citation:31][citation:7].
6. What this page is not
- Not a technology recommendation. No supplier, licensor or equipment manufacturer is recommended, and none is linked. See Mission & Independence.
- Not a patent-legal opinion. Expiry dates and assignee records are reported as published; “expired – lifetime” is the status recorded on the patent, and freedom to operate depends on jurisdiction and on other unexpired rights. See Terms of Use §6.
- Not design data. Pan ratios, flow rates and clearances are reported from the specification to explain the mechanism, not to size equipment.
- Not a ranking of people or bodies. Named individuals appear because their work is documented in patent records or published proceedings; named bodies appear because their publications are cited. This site confers no standing.
Related: Landmark Patents & Profiles · Hot Dross Recovery · Cold Dross & Salt Cake · Presses, Pans and Sow Moulds · Dross Routes Compared · Editorial Standards
中文摘要 · 世界铝业风云人物、学术组织、核心专利与权威媒体全景
本页涵盖通常被分开撰写的四项内容：名字出现在文件上的人、专利本身、出版统计与论文集的机构，以及主张借以抵达受众的媒体。此处将四者作为同一主题处理，因为在本站，四者适用的是同一条检验标准——文件实际写了什么。
1. 风云人物与技术开拓者。本页所替换的原稿称 Dave Roth “推动了重型液压铝渣压榨机在北美铸造车间的商业化普及”。有据可查的记录支持一个更窄的陈述。US 4,386,956《从铝渣中回收金属的工艺》，Roth、Culler 与 Heifner，受让于 Anaconda Co.，1980-12-03 申请，1983-06-07 授权；原理为楔形压头在匹配的槽内压缩铝渣，压头与承接盘充当吸热体以抑制铝热损失；所属路线为机械压渣，与 Alcan 的气氛控制是两条不同的线；商业化归于 Altek，1990 年代初——首个商业成功的铝用实现。由此有三处更正：其一，Roth 是一项受让于 Anaconda 的专利的具名发明人，不是将其商业化的一方；本站的归属原则是凭据归属，而该文件写的是 Anaconda。其二，首个商业成功属于 Altek（1990 年代初）——与 STAS 商业化 Alcan 的 IGDC 是同一时期，两条路线同时进入市场。其三，压渣与惰性气体冷却是并行线，不是一条谱系。US 4,842,255 把 US 4,386,956 列为其引证在先技术之一：冷却专利站在压渣专利的肩上，但二者解决问题的方式不同——一个移除氧气，另一个通过接触与几何移除热量。能够追溯的名字——以及不能追溯的那些。“先驱”一节通常是杰出姓名的罗列，而本节的列表是文件所支持的东西。在铝渣处理领域，其贡献能附着于一份具日期、可引用文件的个人为数不多：Andris B. Innus 与 Michel Villeneuve（US 4,842,255，Alcan）——具名发明人，记录的是个人归属而非公司归属；Roth、Culler、Heifner（US 4,386,956，Anaconda）——机械压渣线的具名发明人，受让人是公司而非个人；W. H. Howard（1896，铅工业）——有据可查的最早压渣记录，同年被用于处理锌渣，并首次把”压头轮廓与承接盘轮廓相匹配”作为关键变量——这一洞见至今支配渣锅设计。未列入者：在工业界记述中流传、但无法追溯至专利、具日期出版物或可核验记录的名字。他们的缺席不构成对他们的评判，它反映的是本站的规则。更宽的年代记比任何一个名字都更要紧。压渣于 1896 年见于铅工业，同年用于锌渣；Pechiney 与昭和铝于 1980 年代初在欧洲与日本并行发展；Anaconda 的专利随后于 1980 年代中期出现。把这项技术呈现为某一个人的部署，等于把一个世纪压缩成一个名字。这也是本节如此命名的原因：一份可追溯姓名的短名单，比一份归属而来的长名单更有用——因为读者可以打开文件。
2. 里程碑式核心专利。US 4,842,255《铝渣冷却装置》；发明人为具名个人 Andris B. Innus 与 Michel Villeneuve，不是笼统的公司归属；受让人为 Alcan International Ltd（加拿大），后登记为 Rio Tinto Alcan International Ltd；优先权日 1985-11-01；申请日 1986-10-30（申请号 US06/925,741）；授权日 1989-06-27；状态 Expired – Lifetime，预计届满 2006-10-30；2005-03-03 曾以担保权益形式由 Novelis Corporation / Novelis Inc. 转让给 CITICORP NORTH AMERICA, INC.。原稿叙述中的四个关键日期有三个早于 1990 年。“90 年代初”这个说法并非凭空捏造——它属于另一件事：惰性气体铝渣冷却（IGDC）是 Alcan 开发的工艺，由 STAS 于 1990 年代初商业化。开发、申请专利、商业化是三件分开的事，把它们压成一个，等于把这项技术的时间点错移了约五年。已发表的工艺描述见《The Alcan Process for Inert Gas Dross Cooling》，Journal of Metals，1991 年 2 月，第 52–53 页。专利实际主张什么：吸热体是渣锅。敞口渣锅：吸收热量。专利原文”铝渣放出的大部分热量由渣锅吸收”，其中 60% 至 85% 的热交换经由这一路径；锅重与渣重之比不低于 1:1，优选 1.5:1 至 2:1；锅壁厚 1–5 cm；任何渣块距最近的锅面不超过 0.5 m；每立方米容积对应的锅体暴露面积不小于 5.0 m²，优选不小于 9.0 m²。惰性气体：置换氧气，不承担降温。优先使用氩气；若铝渣仅局部微红，0.5–1 scfm 的初始与稳态流量已获成功应用。密封：优选砂封——专利原文称其”避免了在先技术中水封的危险”。对原稿的更正：原稿把它描述为一个靠”切断氧气”工作的系统。气体是盖子，渣锅才是吸热体。这是两个机理，混为一谈会丢掉真正支配设备设计的那条规则。留存下来的设计规则是锅重比，而不是气体。一只重量为渣重 1.5–2 倍、且任何渣块离吸热面不超过半米的渣锅，是关于热质量与导热路径长度的陈述，可以在图纸上核对；”注入氩气”则不能。砂封为什么重要：该专利所区分的在先技术是 US 4,097,027（Michel，1978），它用水沟作密封。一份后来的申请把这个失效模式讲得很清楚：罩内氧气被消耗后压力下降、水面上升，水蒸气被金属铝还原为氢气，在提起钟罩、空气重新进入时可能爆炸。砂封消除了这一危险。这项专利不只关乎回收率，它移除的是其前身所制造的一条爆炸路径。氮气：专利的立场与行业的做法并不一致。专利原文写明，氮气”会与熔融铝反应，因此不优选用于铝渣”。但行业并没有照专利执行。一份工业界记述称，氩气费用高”是该系统的缺点之一。这正是一些用户改用氮气作覆盖气体的原因”——并且诚实地补了一句”我无法评论这一改变的结果”。本站不在二者之间作裁决。回收率主张，以及它缺失的那个分母。原稿称该系统”较传统敞口冷却多保全 10%–15% 的金属铝”。量级大致不差，但分母缺失，而数字正活在分母里。专利给出的对比是：按本发明冷却，回收率 > 80%；在渣房地面摊开冷却、或用回转冷却器冷却，通常只能经济地回收 65%–75%——两者的分母都是未反应金属。而且这里有两个效应，它们是相乘的：其一，更多金属以未反应状态存活；其二，存活下来的那部分中可回收的比例更高（>80% 对 65–75%）。二者相乘而非相加，因此任何关于总收益的陈述都必须同时给出两项——这也正是”10–15%”这个单一数字无法被审计的原因。Ma’aden：工厂是真的，但它的数字在另一个分母上。ICSOBA 2023 年一篇铸造车间论文直接写明：“Ma’aden 在三条铸造线上均使用铝渣冷却器：铸锭、铸棒与板坯铸造”。但该文也给出了工厂级数字：铝渣处理厂金属回收率 39%，最高 55%；回收率”取决于多种因素，包括扒渣作业、所用合金、铝渣冷却器性能，以及操作者在炉况准备上的技能”。39–55% 与 >80% 并不冲突，它们的分母不同。专利以未反应金属为分母，工厂以炉料总质量为分母。若渣中游离金属为 50%，回收 39% 的炉料质量意味着回收了所含金属的 78%，与专利相容；但回收 55% 在游离金属仅 50% 时意味着 110%，物理上不可能。因此工厂的上限数字意味着至少在该合金组合中的某些料上，游离金属含量高于约 55%。这个推断属于本站，不属于该论文。并行线：机械压渣。压渣线与冷却线并行，而非在其之后。其有据可查的年代记为：1896（Howard，铅工业；同年用于锌渣，首次把压头轮廓与承接盘轮廓相匹配作为关键变量）→ 1980 年代初（欧洲 Pechiney 与日本昭和并行发展）→ 1980 年代中期（Anaconda，US 4,386,956）→ 1990 年代初（Altek，首个商业成功的铝用实现）。按现场的说法，各路线交付了什么：回转冷却器——厂内自由沥出回收 10–15%，二次回收 35–50%，合计 40–60%；”回转冷却器的安全记录不好”，熔融铝浇到水或生锈钢材上，”现场有多台设备以灾难性后果失效”，维护费用很高。IGDC——厂内回收 0–12%，二次回收 30–60%；”总体结果与回转冷却器非常接近，但没有那笔巨大的维护开支”；缺点是氩气费用高。诚实的说法不是”IGDC 回收更多”，而是”IGDC 回收相当，但更安全”。另请注意：原稿中出现的”10–15%”这个数字，在本来源中是回转冷却器的厂内回收区间。届满做了什么，以及没做什么。US 4,842,255 于 2006-10-30 届满。Ma’aden 的装机在 2023 年被记录，即届满十七年之后。2013 年的一份申请（US 2013/0125705）把该专利列为在先技术，并直白指出其局限：“在世界上某些生产金属的地区，惰性气体过于昂贵，或者干脆无法获得”。一项以氩气为活性耗材的工艺，继承了氩气的物流约束。限制这条路线部署范围的是这一点，而不是冶金原理。但届满并没有让运行数据变成公开的。一座在产工厂可获得的回收率数字来自运营方的会议论文，且带有所声明的依赖条件。这是铸造车间数据的常态。
3. 学术组织与行业协会。原大纲点名的四个机构并不是同一类——其中两个并非读者会以为的那样，而若干个对再生铝比那四个中某些更要紧的机构，大纲根本没有点名。因此本节做两件事：说明每个机构是什么，以及说明它出版哪一类的材料。只有名字而没有第二列的清单不是参考章节，那是通讯录。读范围栏，不要读标识。一个全球原铝生产商协会、一个全球学会、一个印度铸造协会、一个欧洲铝箔机构、一个美国废铝贸易协会、以及一个自愿性认证体系，都在下面。它们不是可以互换的来源；引用甲而当乙，正是本节要防止的最常见错误。机构按职能分组：3.1 认证 · 3.2 统计与同行评议出版 · 3.3 回收产业自身的组织 · 3.4 国家级、区域级与细分领域协会。3.1 认证：ASI（铝业管理倡议）。本表中唯一一个其产出是针对实体而非针对产品进行审计的机构，也是唯一把再生金属当作独立入口而非原铝附表的机构。对一个关注重熔与铝渣的站点，第二点正是它排在首位的原因。两项标准，两者皆需：价值链中的公司须同时通过绩效标准（治理与商业诚信；材料管理，含回收与资源效率；环境，含温室气体、废物、水与生物多样性；人权；劳工权利与职业健康安全）与监管链标准的独立认证。三个界定阶段：原铝——铝土矿开采至铸造车间；再生铝——废铝回收至铸造车间；铸造车间之后——半成品加工与制造至最终产品。废铝从哪里进入：合格废铝是一种投入，但在重熔/精炼厂将其指定为 ASI 铝之前，它不是监管链材料。重熔厂是再生投入转变为认证产出的那个点。监管链版本：CoC 标准 V2.1，2023 年 4 月；重大标准修订定于 2025 年 3 月启动，过渡期内已批准对现有 CoC 认证的可选延期。与本页相关的修订内容：简化 CoC 材料类型；将通过闭环回收取得的消费前废铝纳入合格废铝；取消市场信用体系；将回收含量作为可选可持续性数据纳入 CoC 文件；要求对贸易商进行尽职调查。定义取决于一个短语，而这个短语在刻意起作用。ASI 称，带标产品中的铝”对应于由不间断的 ASI 认证实体链所生产的等效质量的 ASI 铝”。等效质量不等于同一批原子。该体系认证的是经由受审计实体链的质量流账目，而不认证某产品中金属是否实际产自某认证设施。这不是对 ASI 的批评——质量平衡模型是载明的设计选择，而规模化下的物理隔离是另一回事，且限制大得多。之所以写明，是因为把 ASI 主张理解为物理可追溯的读者，读到的是一条与已发布者不同的主张。本站处理回收率数字的同一条告诫在此同样适用：一个数字所计的是什么、跨越什么边界，是这个数字的一部分。3.2 统计与同行评议出版：IAI、TMS、ICSOBA。IAI（国际铝业协会）：范围全球——原铝；出版全球原铝统计、2005/2010/2015/2019 年生命周期清单报告、行业温室气体数据、碳足迹方法学指南，以及 2026 年《可持续铝渣管理指南》；其自我描述为”唯一代表全球原铝产业的机构”——代表的是原铝生产商，在引用它谈再生铝时这是一个实质性的事实。TMS（矿物、金属与材料学会）：范围全球——学会；出版《Light Metals》年度会议论文集（Springer 出版，2026 卷由 Roberto Seno 主编）与《JOM》；《Light Metals》恰是铸造车间运行数据最常出现的地方。ICSOBA（国际铝土矿、氧化铝与铝研究委员会）：范围全球——上游，矿山到金属；是联合生产商、技术供应商、研究者与顾问的独立非营利机构，63 年、43 届会议、30 座城市、17 个国家，每年举办技术会议；ICSOBA 2025（中国南宁）设 4 场主旨报告与超过 250 篇技术论文。其自述范围是”从矿山到金属”，然而其论文集中确实出现铸造车间运行论文——上文第 2.1 节引用的 Ma’aden 回收率数字即出自 ICSOBA 2023 年一篇论文。自 2023 年起，它与 IAI 建立会议伙伴合作关系。3.3 回收产业自身的组织：BIR、EuRIC / EUROMETREC、ReMA。这三家是为废铝与回收贸易本身发声的机构——不是为生产商，也不是为某个职业。对从事再生铝工作的人，它们才是其牌号定义、统计与政策立场真正支配物料的机构。BIR（国际回收局）：范围国际——回收产业，秘书处设于布鲁塞尔；出版全球有色与不锈钢废金属市场的 World Mirror 系列，其中《有色金属 World Mirror》已运行二十余年；2025 年完成统计研究《全球再生有色金属贸易》，涵盖铝、铜、锌、铅。该来源把自己的统计目的说得很明白：它们”澄清我们运作所依据的统计基线”，支持倡导工作，”使我们得以量化自身的环境与经济贡献”，并”提醒决策者注意贸易限制的准确负面后果”。有用，且明确以倡导为导向。EuRIC / EUROMETREC：范围欧盟——有色金属贸易与回收；EUROMETREC 于 1990 年由欧洲经济共同体（EEC）内有色金属贸易联络委员会改组成立，2016 年并入 EuRIC 成为其分支；其宗旨包括监督影响回收企业的法律、商业、环境与技术问题，并在欧盟机构面前代表其利益；成员为国家协会，每个合格国家一个，最低不少于三个——因此其立场是各国联合会的立场，而非单个公司的立场。欧盟层面的倡导机构。ReMA（再生材料协会，前身为美国废料回收工业协会 ISRI）：范围美国——但其规格具有国际效力；出版《废铝规格通告》，即交易废料的牌号定义，含铝。规格持续修订：2022 年 7 月董事会会议批准新增牌号（TANK、TEASE）与修订牌号（TALL、TALON，两者均为 99.45% 铝含量）。这些牌号定义是远超美国范围的废铝合同中的通用词汇；ReMA 自身材料称其代表美国及40 多个国家的企业。是规格机构，不是冶金主张的核实机构。一份废铝规格里最有价值的东西，是它拒不定死的那些。ReMA 公布的修订中，包含条款由双方商定的牌号：某一修订牌号写明“最终材料规格由买卖双方确定”。一份明确把某项留给约定的标准规格，是在承认一件真实的事：对某些物料，没有任何单一数字在所有情形下都成立，这个数字就是合同所说的那个。这与本站报道回收率数字时所见的同一结构——分母是数字的一部分，未载明分母的数字无法比较。这也意味着规格集是一份活的文档，不是固定参照：牌号直到 2022 年仍在新增与修订。引用时请注明版本。3.4 国家级、区域级与细分领域协会。Aluminium Association（美国铝业协会）：范围美国——但其合金注册被全球使用；出版合金牌号注册与化学成分限值。新牌号持续登记：2018 年版《Pink Sheets》（铸件与铸锭）增补本所列注册包括 G356.0 / G356.1 / G356.2（2024 年 8 月）、527.0 / 527.1 与 727.0 / 727.1（2021 年 3 月）。该注册是持续维护的，不是封闭的——一个牌号是一项带记录成分的具日期行为。引用时应注明该 Sheets 的版本，而非笼统的”AA 体系”。ALUCAST（铝铸造协会）：范围印度——国家级，铸造；出版会员名录（所引为第 9 版）、参考手册、培训教材、官方期刊，并举办会议与展览；是总部位于浦那的印度铝铸造行业协会，在印度各地设有区域中心；不是全球机构，也不是铝渣回收机构——此处按其本相引用，不作为铝渣问题的权威。EAFA：范围欧洲——下游细分领域；本页未核实其出版物，也未引用任何一条。原大纲将 EAFA 与 IAI、TMS 并列；按与铝行业最一致的解读，它指的是欧洲铝箔／软包装协会——一个细分领域机构，而非整个欧洲产业协会。覆盖欧洲全价值链的是 European Aluminium（欧洲铝业协会），那是另一个组织。European Aluminium：范围欧洲——全价值链；出版政策立场、回收统计（如饮料罐回收量）以及与钢铁业联署的关于废铝外流的行动呼吁；自我描述为”整个欧洲铝价值链的声音”——是一个倡导机构，其数字有用，而其政策立场是立场。一条适用于这些机构所有时间序列数据的警示：IAI 明确声明“由于背景数据与计算方法的变更，2005–2018 年的数据集与 2019–2021 年的数据集不可比较”。其所列原因很具体：加入电力传输损耗、净阳极消耗改为毛阳极消耗、石油焦改为煅后石油焦、以及依 2019 年 IPCC 精细化指南加入低压 PFC 排放。跨越这个边界画趋势线的读者，量到的不是产业变化，而是方法变更。这是”没有分母的数字不是数字”这条本站通则最清晰的已发表实例。同样的告诫适用于认证：ASI 监管链标准自 2025 年 3 月起处于重大修订之中，并为现有认证批准了过渡性延期。针对现行版本成立的主张，不自动针对下一版本成立。
4. 权威媒体与情报来源。Light Metal Age：双月刊行业杂志，”80 余年”历史，编辑报道范围包含铸造车间与再生铝，栏目含国际专利日历——行业媒体，非同行评议。Aluminium International Today：覆盖从冶炼到下游加工的价值链的全球杂志，刊载 IAI 人员撰写的安全分析——行业媒体，非同行评议。AL Circle：拥有具名编辑团队的行业媒体平台，博客与分析覆盖铝土矿、氧化铝、原铝、回收、挤压与下游市场——商业媒体平台，非同行评议；可用作线索与市场评论，不是核实渠道。上市公司报告：上市冶炼商、回收商与装备集团的年报、季报与监管备案——经审计，是本站最强的一类企业数字来源，因为它是备案的、具日期的、可归属的；同时也是倡导：一家公司的报告为自身立场辩护。海关与多边贸易统计：协调制度贸易数据（如 HS 2606 铝矿、HS 7602 铝废碎料、HS 2620.40 主要含铝的灰及残渣）——一手统计发布，但其可比性取决于修订版本与报告国自身的归类实践。对原稿的更正：原稿称前两本杂志是”铸造车间装备性能数据的主要技术渠道”。它们是性能主张出现于其中的渠道，而不是核实这些主张的机构。这个区分在本站是承重的。流通中的铸造车间装备性能数据，很大一部分是由装备供应商自己发布的——与本站压渣机页面所记录的同一来源类别局限。行业媒体是这些数据抵达受众的地方。把它当作赋予了核实效力来引用，会抵消本站转而追溯专利与运营方报告的全部理由。它们仍有真实价值：Light Metal Age 的专利日历与工厂报道是货真价实的线索，IAI 的铝渣管理指南是具名且具日期的出版物。线索与核实是两件事。
5. 什么还没有被证明。“多保全 10%–15% 的金属铝”——按所述方式无法追溯；专利的对比是 >80% 对 65–75%（分母为未反应金属），工厂的数字是 39–55%（分母为炉料总质量），两者都不是金属量的百分比增幅。氮气等效性——专利不优选它；部分用户因成本采用；对回收率的影响无人置评。惰性气体冷却在回收率上优于回转冷却——现场记述称结果”非常接近”，有据可查的优势是安全与维护。Roth 在北美商业化普及了铝渣压榨——文件写明的受让人是 Anaconda，首个商业成功归于 Altek。Ma’aden 的游离金属化验值——所引论文未公布；第 2.1 节的对账属本站推断，已作标注。EAFA 的身份与出版物——本页未核实，未引用其任何内容；日后若使用，须先解决。IAI 排放数据的跨期可比性——在 2018/2019 边界上明确不可比。行业媒体报道与同行评议结论之间的任何等效性——本站不作此主张。
6. 本页不是什么。不是技术推荐——不推荐任何供应商、许可方或设备制造商，也不链接其中任何一家，见《使命与独立性》。不是专利法律意见——届满日期与受让人记录按已公布内容报道；”Expired – Lifetime”是专利上记载的状态，而自由实施取决于辖区及其他未届满的权利，见《使用条款》第 6 节。不是设计数据——锅重比、流量与间距引自说明书，用以解释机理，不用于设备选型。不是人物或机构排名——具名个人出现，是因为其工作记载于专利记录或已发表论文集中；具名机构出现，是因为其出版物被引用。本站不授予任何地位。