The FORGE Architecture — Post 3: The Inola Proof

The FORGE Architecture  ·  FSA Critical Minerals Policy Series Post 3

The FORGE Architecture

Demand-Side Architecture for Domestic Critical Minerals Processing

The Inola Proof

The $4 billion aluminum smelter at the Tulsa Port of Inola is the first new primary aluminum production plant built in the United States since 1980. It is also an accidental proof-of-concept for the entire FORGE architecture. Five factors made Inola viable: multimodal logistics on the McClellan-Kerr Arkansas River, dedicated industrial power, strategic sovereign capital, a performance-based state incentive package, and price protection from the 50 percent Section 232 aluminum tariff that Century Aluminum's CEO called "fundamental to the future of the industry." Four of those five factors exist or can be replicated for rare earth processing on the same corridor. The fifth — price protection — is precisely what FORGE is designed to provide. Inola is not an analogy for the rare earth argument. It is a direct precedent. The tariff was the floor. The smelter was financed against it.

Randy Gipe  ·  Claude / Anthropic  ·  2026  ·  Trium Publishing House Limited  ·  Forensic System Architecture

Series Position — Post 3 of 5 Post 1 established the floor problem. Post 2 documented FORGE as the proposed solution — a plurilateral reference price and adjustable tariff architecture announced February 4, 2026. Post 3 provides the proof of concept: a $4 billion investment already committed to the same corridor, using the same logistics infrastructure, financed against the same mechanism FORGE proposes to apply to rare earths. The Inola smelter did not require a unique set of conditions. It required a specific combination of five factors. This post documents each of those factors and maps them, one by one, to what a rare earth processing facility on the Arkansas River would need.

The aluminum industry spent thirty years dying in the United States before it came back to Oklahoma. At the peak of American aluminum production in the 1980s, the United States operated more than thirty primary aluminum smelters. By 2024, four remained. The collapse was not a failure of geology — the United States has abundant bauxite trading relationships and alumina supply chains. It was not a failure of technology. It was a failure of economics: high energy costs, foreign competition pricing at levels that domestic producers operating at market-rate power contracts could not match, and the steady attrition of the industrial policy environment that had originally made aluminum smelting viable in postwar America. The U.S. aluminum industry went from thirty smelters to four for the same reason Mountain Pass went from the world's largest rare earth mine to a bankruptcy estate: a competitor willing and able to price below Western production costs held the market long enough for the domestic industry to shut down around it.

What brought aluminum back to Oklahoma was not a reversal of the underlying economics. It was a policy decision — the Section 232 aluminum tariff, imposed at 25 percent initially and raised to 50 percent under the current administration — that changed the price at which foreign aluminum could enter the American market. Fifty percent tariffs on aluminum imports are, in functional terms, a price floor: they guarantee that domestic production is not priced out by foreign producers selling below cost, because the tariff closes the gap between the foreign price and the domestic production cost. Once that floor existed, the investment thesis for a new American smelter became calculable. EGA of Abu Dhabi committed $4 billion. Century Aluminum signed on as a 40 percent partner. The Department of Energy committed up to $500 million. Oklahoma committed $275 million in performance-based incentives. JPMorgan, Goldman Sachs, and the project finance market followed. The floor came first. The capital followed the floor.

"Before President Trump came into office in his first term, the aluminum industry in the United States was on its knees. We've gone from 30 smelters in this country down to just four." Jesse Gary, CEO Century Aluminum — Fox Business News, January 2026

$4B

Total Investment — EGA / Century Joint Venture

Largest single investment in U.S. aluminum history. Construction begins late 2026; first metal by 2029–2030.

750K

Tonnes/Year Aluminum Capacity

More than doubles current U.S. production. First new primary aluminum plant in the U.S. since 1980.

50%

Section 232 Aluminum Tariff — The Floor

The mechanism that made the investment thesis calculable. This is what FORGE's adjustable tariffs are designed to replicate for rare earths.

I. The Five Factors

What Made Inola Work — Each Element Documented

The Inola smelter did not succeed because Oklahoma is special. It succeeded because a specific combination of five factors converged at one site at one moment in the industrial policy cycle. Understanding each factor precisely — what it provided, how it was secured, and what it cost — is the analytical work that makes the Inola model transferable rather than exceptional. Four of the five factors are replicable on the same Arkansas River corridor for rare earth processing. The fifth was provided for aluminum by Section 232. For rare earths, FORGE is the proposed equivalent. The factors follow.

Factor 1 of 5 Multimodal Logistics — The McClellan-Kerr Arkansas River Navigation System Confirmed for Inola · Available for REE

The Tulsa Port of Inola sits on the McClellan-Kerr Arkansas River Navigation System — a 445-mile waterway connecting the Arkansas and Verdigris Rivers to the Mississippi River system and, through it, to the Gulf Coast. The port encompasses 2,200 acres of industrial park land with direct rail and barge access. The smelter site occupies 350 acres within that park. Alumina — the feedstock for aluminum smelting — travels by ship from bauxite-producing nations to Gulf Coast ports, transloads to barge on the Mississippi, and moves up the Arkansas River to Inola at a cost per ton that no combination of truck and rail can approach. The same movement, in reverse, carries aluminum products to Gulf Coast export terminals. The logistics economics are fundamental to the investment thesis: a landlocked Oklahoma site competes with coastal alternatives because the river makes it, effectively, a coastal site for bulk commodity movement.

For rare earth processing: The same corridor, the same port, the same barge access applies. Rare earth concentrate from Energy Fuels' White Mesa mill in Utah, or from allied sources via Gulf Coast import terminals, moves by rail to Inola and by barge outbound. The logistics infrastructure that Inola proves is viable is already built. It does not need to be replicated. It needs to be used.

Factor 2 of 5 Industrial Power at Scale — PSO and the Dedicated Supply Architecture Confirmed for Inola · Addressable for REE

Aluminum smelting is among the most energy-intensive industrial processes in existence. The Inola facility will consume roughly the same amount of electricity as the city of Boston. Securing reliable, competitively priced industrial power at that scale is not a procurement exercise — it is a parallel infrastructure project. Public Service Company of Oklahoma, which serves the Inola area, acquired a 795-megawatt Green Country gas-fired power plant in part to meet the smelter's anticipated demand. EGA and Oklahoma negotiated discounted industrial power rates as part of the state incentive package. The power dimension of the Inola investment is as significant as the capital expenditure: a smelter without a long-term power contract at a viable rate is not a smelter. It is a construction project waiting for a stranded asset designation.

For rare earth processing: Rare earth separation is energy-intensive, but at a fraction of aluminum smelting's demand. A 5,000–10,000 tonne/year separation facility requires substantial power for solvent extraction chemistry and thermal processing, but not at Boston-scale demand. Oklahoma's industrial power infrastructure — enhanced by the Inola negotiations — is more than adequate for REE processing. The power constraint that aluminum had to solve at great cost is a solved problem for rare earth processing on the same corridor.

Factor 3 of 5 Strategic Sovereign Capital — EGA, Mubadala, and the UAE Investment Architecture Confirmed for Inola · Positioned for REE via Pax Silica

EGA — Emirates Global Aluminium — is owned by two UAE sovereign entities: Mubadala Investment Company of Abu Dhabi and Investment Corporation of Dubai. The $4 billion Inola commitment is, in structural terms, sovereign capital: patient, strategically motivated capital that is not solely optimizing for the quarterly return profile that constrains public market investors. EGA's willingness to commit to a 45-year industrial gap in U.S. aluminum production reflects a calculation that extends beyond the aluminum market — it reflects the UAE's strategic interest in being positioned as an essential industrial partner in the American supply chain reshoring moment. The sovereign capital layer is what makes the commitment durable. A public market aluminum company, exposed to quarterly earnings pressure, does not commit $4 billion to a project that does not produce metal for three to four years.

For rare earth processing: Mubadala — the same sovereign entity that owns EGA and is building Inola — is also a Pax Silica participant. The sovereign capital that proved the Inola model is already positioned in the rare earth supply chain architecture. It is not hypothetical alignment. It is the same investor, in the same geography, across two adjacent industrial investments. When Pax Silica sovereign capital looks for rare earth processing to invest in alongside FORGE reference price protection, the Inola corridor is already on the map.

Factor 4 of 5 State and Federal Incentive Architecture — Performance-Based, Not Permanent Subsidy Confirmed for Inola · Replicable for REE

Oklahoma's incentive package for the Inola smelter totals more than $275 million in state funds, tax exemptions, and power discounts, structured as performance-based obligations — EGA cannot begin drawing the annuity payment until 2030, and only if it has spent at least $2 billion and created 700 jobs by then. The package was vetted by the bipartisan Legislative Economic Advancement and Development (LEAD) Committee and codified in the Reindustrialize Oklahoma Act (ROA-25). The $500 million DOE Industrial Demonstrations Program grant adds federal support. The incentive architecture is a complement to, not a substitute for, commercial viability: the performance conditions mean Oklahoma is not paying for a facility that doesn't get built or staffed. The incentive is structured to activate on the delivery of the industrial outcome the state is paying for.

For rare earth processing: A comparable incentive architecture is available. The DOE's Industrial Demonstrations Program that supported Inola; the Defense Production Act Title III authority that underpins DoD critical minerals investments; the IRA's advanced manufacturing production credits; and Oklahoma's own demonstrated willingness to structure performance-based incentive packages for transformational manufacturing investments — all of these instruments are as applicable to rare earth separation as they are to aluminum smelting. The incentive infrastructure that Oklahoma built for Inola is the template, not the exception.

Factor 5 of 5 Price Protection — Section 232 as the Aluminum Floor, FORGE as the REE Equivalent Essential for Inola · Missing for REE without FORGE

The Section 232 aluminum tariff — 25 percent initially, raised to 50 percent under the current administration — is the mechanism that closed the gap between domestic aluminum production costs and the price at which Chinese and Russian producers (who together supplied 65 percent of global aluminum) were willing to sell into the American market. Century Aluminum's CEO Jesse Gary stated it directly: the tariffs are "fundamental to the future of the industry." EGA's investment thesis, the DOE grant, Oklahoma's incentive package, the private financing — all of it is predicated on a market environment in which the landed cost of foreign aluminum reflects something closer to real production costs than Chinese state-subsidized export prices. The tariff is the floor. The smelter was financed against it. Remove the tariff, and the investment math changes. The floor is not an incidental feature of the Inola model. It is the load-bearing wall.

For rare earth processing: This is precisely the missing factor. Rare earth oxides and processed materials face the same structural pricing problem aluminum faced: Chinese state-supported producers price at levels that Western processors cannot match at market-rate capital costs. Section 232 tariffs on rare earths have been discussed but not structured as the comprehensive, stage-by-stage price floor that FORGE's reference price and adjustable tariff mechanism proposes. Without that mechanism, rare earth processing on the Arkansas River corridor faces the same arithmetic that killed thirty aluminum smelters: a foreign competitor willing to price below cost indefinitely, and no floor to prevent it from doing so.

Series Analytical Insight — The Tariff Precedent

The Inola smelter is not an analogy for the FORGE argument. It is a direct precedent. The Section 232 tariff that enabled Inola is structurally identical to the adjustable tariff mechanism that FORGE proposes for rare earths: a border adjustment that prevents foreign state-subsidized producers from undercutting domestic production at below-cost prices. The aluminum industry required a 50 percent tariff to make a new smelter financeable. Rare earth processing requires the equivalent mechanism, applied at each stage of the supply chain — oxide, metal, magnet — rather than as a single commodity tariff. FORGE is the proposal to build that mechanism at plurilateral scale. The Inola smelter proves it works. The question is whether the political architecture to extend it to rare earths can be built and held.

II. The Mubadala Thread

When the Same Sovereign Capital Appears in Both Architectures

The connection between the Inola smelter and the FORGE rare earth architecture is not only structural — it is institutional. Mubadala Investment Company, the Abu Dhabi sovereign wealth fund that co-owns EGA and is committing $4 billion to the Inola smelter, is also a Pax Silica participant. Pax Silica — the companion initiative to FORGE focused on the silicon-AI supply chain — brings sovereign wealth funds into the critical minerals equity investment architecture once FORGE price floors make processing facilities commercially bankable. Temasek of Singapore and Mubadala of Abu Dhabi are the two sovereign wealth funds most prominently associated with the Pax Silica framework.

This is not coincidental alignment. Mubadala's investment posture — patient, strategic, willing to take the long view on industrial infrastructure in allied markets — is precisely the capital profile that critical minerals processing requires. A $1 to $1.8 billion rare earth separation facility has a development timeline, a permitting timeline, and an operating ramp-up timeline that public market investors are poorly suited to finance at early stages. Sovereign wealth funds, with their multi-decade investment horizons and strategic rather than purely financial mandates, are structurally better suited to the patient capital role. Mubadala proved it is willing to commit that capital to American industrial infrastructure — on the Arkansas River, at the Tulsa Port of Inola, for the first new American aluminum smelter in 45 years. The same entity is positioned in Pax Silica to make the analogous commitment for rare earth processing once FORGE provides the price floor that makes the investment bankable.

The node-control principle clarifies why this matters. Mubadala does not control the Inola smelter because it found a favorable investment in rural Oklahoma. It controls a node — the conversion point at which Gulf Coast alumina becomes American primary aluminum — because controlling that node positions it at the intersection of the U.S. industrial reshoring agenda, the tariff-protected aluminum market, and the downstream demand for American-made aluminum products from the aerospace, automotive, and defense sectors. The Pax Silica positioning represents the same logic applied to rare earths: not a financial bet on the rare earth market, but a strategic position at the node where allied supply chain architecture and sovereign capital intersect.

"The same sovereign capital that proved the Inola model — Mubadala, $4 billion, on the Arkansas River — is positioned in Pax Silica to make the analogous commitment for rare earth processing. The investor already knows the corridor. The investor already understands the mechanism. What is missing is the floor." The FORGE Architecture — Post 3

III. The Delta

Four Factors Present, One Missing — and What Fills It

The Inola comparison is analytically useful precisely because it is not a perfect analogy. Rare earth processing is more chemically complex than aluminum smelting, requires radioactivity handling for monazite feedstocks, involves a more fragmented supply chain, and faces a more volatile price history than the London Metal Exchange-listed aluminum market provides. These differences matter. The rare earth investment thesis is harder to build than the aluminum investment thesis was, even with all five Inola factors in place. Pretending otherwise would be the kind of analytical flaw the FSA methodology exists to prevent.

But the differences in complexity do not change the structure of the argument. They change the magnitude of the floor required, not the necessity of having one. If aluminum needed a 50 percent Section 232 tariff to become investable, rare earth processing needs an equivalent or stronger mechanism — because its price volatility is greater, its supply chain complexity is higher, its radioactivity handling requirements add permitting costs, and its Chinese competitor is more deliberately pricing below cost than the aluminum market has faced. The floor required for rare earth processing is higher and more structurally sophisticated than the Section 232 flat tariff that served aluminum. FORGE's stage-by-stage reference pricing and adjustable tariff mechanism is more sophisticated than Section 232 precisely because it needs to be.

Post 4 models the numbers: what a 5,000 to 10,000 tonne per year NdPr-focused separation facility near the Tulsa Port of Inola corridor actually requires in capital expenditure, what floor price produces investment-grade returns, what the Arkansas River logistics savings do to the operating economics, and how Project Vault's buyer-of-last-resort function completes the financial architecture. Post 3's job is the proof of concept. Post 4's job is the math. The proof of concept is documented. The corridor works. The capital is positioned. The logistics infrastructure is already being built for aluminum. The five factors are four present and one missing. FORGE is the one that is missing.

FSA Framework — Post 3: The Inola Proof

Source

The Inola Investment as Documented Industrial Policy Outcome The $4 billion EGA/Century Aluminum smelter at the Tulsa Port of Inola is a primary-source documented outcome of a specific combination of industrial policy instruments: Section 232 tariff protection, federal DOE grant support, state performance-based incentives, sovereign capital commitment, and multimodal logistics infrastructure. It is not a projection or a model. It is a real investment committed, construction timed to begin, permits being secured. The series uses it as a primary source for what works — not as a perfect template, but as documented proof that the five-factor combination produces the capital commitment the supply chain requires.

Conduit

The Arkansas River Corridor as Physical Proof The McClellan-Kerr Arkansas River Navigation System — connecting Inola to the Mississippi system, the Gulf Coast, and global shipping — is the conduit through which the Inola model's logistics advantage flows. The M-KARNS is already in service. The Tulsa Port of Inola's 2,200-acre industrial park is already operational. The rail connections are already built. The barge channel is already navigable. The conduit that the rare earth processing argument requires does not need to be constructed. It needs to be used. That is what distinguishes the Inola corridor argument from a greenfield proposal: the physical infrastructure is a documented present reality, not a future capital requirement.

Conversion

Section 232 → Investment Bankability → $4 Billion Committed The conversion mechanism in the Inola case is precisely the mechanism FORGE proposes: a tariff-enforced price floor converts an otherwise commercially nonviable domestic industrial investment into a bankable one. Section 232 at 50% → aluminum landed cost reflects real production costs rather than Chinese state-subsidized export prices → EGA's investment thesis becomes calculable → $4B committed → first new U.S. aluminum smelter in 45 years. The conversion fired. The mechanism worked. FORGE's proposal is to run the same conversion for rare earth oxides, metals, and magnets using a more sophisticated stage-by-stage reference price and adjustable tariff architecture.

Insulation

The Complexity Delta Between Aluminum and Rare Earths The insulation layer in the Inola comparison is the genuine complexity difference between aluminum smelting and rare earth separation. Rare earth processing requires radioactivity handling (for monazite feedstocks), more complex chemistry (solvent extraction for individual element separation), a more fragmented and opaque supply chain, higher price volatility, and a domestic regulatory environment less experienced with these specific processing requirements than with aluminum. These differences raise the floor required — they do not eliminate the case for having one. The FSA Wall on the rare earth processing complexity delta: the series documents the structural parallel; it does not minimize the genuine additional challenges. Post 4 models the numbers that reflect those challenges.

FSA Documentation — Five-Factor Comparison: Inola Aluminum vs. Hypothetical REE Processing, Arkansas River Corridor

Factor	Inola Aluminum — How Secured	REE Processing — Status	Gap / Resolution
Multimodal logistics	Tulsa Port of Inola on M-KARNS; rail spur + Arkansas River barge; 2,200-acre industrial park	Same port, same waterway, same rail infrastructure — already built for aluminum	No gap. Infrastructure present.
Industrial power	PSO long-term contract; PSO acquired 795MW Green Country plant; discounted industrial rates negotiated in state MOU	REE separation requires substantial power but far below Boston-scale aluminum demand; Oklahoma industrial power grid enhanced by Inola negotiations	Addressable. Power infrastructure expanded for aluminum; REE processing needs less.
Strategic sovereign capital	EGA (Mubadala + ICD ownership); $4B commitment; patient 45-year industrial gap investment	Mubadala (same entity) is Pax Silica participant; Temasek also positioned; sovereign capital aligned with allied supply chain architecture	Addressable. Same investor, adjacent architecture.
State + federal incentive architecture	$275M state (performance-based, ROA-25); $500M DOE IDP grant; LEAD Committee vetting process	DOE IDP, DPA Title III, IRA advanced manufacturing credits, Oklahoma incentive infrastructure — all available for REE processing	Addressable. Oklahoma template is replicable; federal instruments applicable.
Price floor / market protection	Section 232 aluminum tariff at 50% — "fundamental to the future of the industry" (Century CEO); closes gap between domestic cost and foreign subsidized price	MISSING. No equivalent stage-by-stage price floor for rare earth oxides, metals, or magnets exists outside bespoke DoD bilateral contracts	FORGE's reference price and adjustable tariff mechanism is the proposed resolution. Without it, the investment math does not close.
FSA Wall	The Inola investment figures — $4B total, $275M state incentives, $500M DOE grant, 750,000 tonnes/year capacity, EGA 60%/Century 40% JV structure — are drawn from primary sources including EGA press releases, Oklahoma Department of Commerce announcements, the Bond Buyer analysis, and Globe Newswire. The "city of Boston" power consumption comparison is drawn from the Aluminum Association via Bond Buyer reporting. The claim that Mubadala is a Pax Silica participant is drawn from published reporting on the Pax Silica initiative; the specific capital commitment Mubadala has made to Pax Silica is not publicly documented at the precision required for a financial claim, and the FSA Wall is declared on Mubadala's specific Pax Silica commitment level.

FSA Wall · Post 3 — The Inola Proof

The Jesse Gary quotation — "Before President Trump came into office in his first term, the aluminum industry in the United States was on its knees. We've gone from 30 smelters in this country down to just four" — is drawn from Fox Business News coverage of the EGA/Century joint development agreement announcement, January 2026. The characterization of Section 232 tariffs as "fundamental to the future of the industry" is attributed to Gary in the same reporting and in the okenergytoday.com coverage of the joint venture announcement. Both attributions draw on secondary reporting of Gary's remarks; the series does not have access to a transcript of his Fox Business appearance.

The state incentive package figure of "more than $275 million" is drawn from the AGBI analysis (July 2025) and the Newson6/okcommerce.gov reporting. The AGBI analysis specifies $20M from an executive fund and $255M redirected from existing legislative appropriations. The Bond Buyer analysis describes the package as including TIF bonding. The total figure varies slightly across sources; "more than $275 million" reflects the consistent floor across all published accounts. The $500M DOE figure is drawn from the okenergytoday.com report confirming Phase 1 of the DOE IDP award negotiations entered in early 2025.

The characterization of the Section 232 tariff as the "load-bearing wall" of the Inola investment thesis is analytical framing by the series — it is not language used in EGA, Century, DOE, or Oklahoma government documents. The structural parallel to FORGE's adjustable tariff mechanism is the series' original analytical contribution, not a claim found in existing coverage of either the Inola project or the FORGE initiative.

The rare earth processing comparison — the five-factor analysis and the complexity delta — draws on the FORGE Section Outline document provided in series development and on the series' own analytical framework. Specific REE processing cost figures are modeled in Post 4 from published industry comparables; they are not reproduced from primary project documents for any specific rare earth facility.

Primary Sources & Documentary Record · Post 3

1. EGA (Emirates Global Aluminium) — Press release: "EGA progresses plans to build first new primary aluminium production plant in the US since 1980, in Oklahoma," May 2025; press release on Century Aluminum joint venture, January 26, 2026 (media.ega.ae, public)
2. Globe Newswire — "Century Aluminum Joins EGA Project to Build First U.S. Smelter in Almost 50 Years," January 26, 2026; JV terms, capacity, EGA 60%/Century 40% structure (GlobeNewswire.com, public)
3. Oklahoma Department of Commerce — EGA investment announcement; Governor Stitt MOU; incentive package terms; port and waterway access (okcommerce.gov, public)
4. Oklahoma House of Representatives — "House Passes Historic Economic Legislation Backing $4 Billion Aluminum Smelter Project"; Reindustrialize Oklahoma Act (ROA-25)/HB 2781 passage, May 2025 (okhouse.gov, public)
5. Bond Buyer — "Oklahoma incentives land a $4 billion aluminum smelter," July 16, 2025; TIF structure; DOE grant; PSO power negotiations; incentive package detail (BondBuyer.com, public)
6. AGBI — "EGA's Oklahoma smelter set to outlast US trade policies," July 1, 2025; $20M executive fund + $255M legislature breakdown; performance conditions; construction and production timeline (AGBI.com, public)
7. OK Energy Today — "Aluminum plant proposed for Inola becomes a joint venture," January 27, 2026; DOE $500M grant confirmation; PSO Green Country plant acquisition; Century CEO Jesse Gary on tariffs (okenergytoday.com, public)
8. Southern Economic Development Council — Century/EGA joint venture summary; M-KARNS logistics detail (SEDC.org, public)
9. Light Metal Age Magazine — EGA Oklahoma investment overview; MOU terms; LEAD Committee vetting process (LightMetalAge.com, public)
10. Recycling Today — Century/EGA partnership announcement; capacity; hub development potential (RecyclingToday.com, public)
11. The FORGE Architecture — Post 1: The Floor Problem; Post 2: FORGE Anatomy — Trium Publishing House Limited, 2026 (thegipster.blogspot.com) — floor problem framing; FORGE mechanism documentation; Section 232 parallel
12. Hidden Arteries: FSA Inland Waterways Architecture Series — Trium Publishing House Limited, 2026 (thegipster.blogspot.com) — M-KARNS and Arkansas River logistics infrastructure primary source

← Post 2: FORGE Anatomy Sub Verbis · Vera Post 4: The Oklahoma Model →