The short version: On 8 September 2026, POSCO broke ground on No. 8 Continuous Galvanizing Line (8CGL) at its Gwangyang Works — a KRW 480 billion, 450,000 t/y automotive hot-dip galvanizing line targeted for completion in 2028. It is POSCO's first new CGL in nine years, and its defining feature is not the zinc pot but the furnace upstream of it: a heat-treatment system explicitly engineered to process steel made from electric-furnace hot metal. The commercial significance is obvious. What is conspicuously absent is the contracting chain — no EPC award, no line OEM, no equipment package value has been publicly confirmed.

What is actually being built

POSCO's own release is the primary record. The 8CGL at Gwangyang Works in South Jeolla Province will have nameplate capacity of 450,000 metric tons per year and is scheduled for completion in 2028. Once operational, POSCO's total hot-dip galvanized (HDG) capacity rises to 4.95 million tonnes per year — a roughly 10% increment on the existing base.

The product mix is the interesting part, and it is narrower than a generic galvanizing line:

  • Extra-wide automotive exterior panels. POSCO frames 8CGL as its answer to the shift in global demand toward mid-size and larger SUVs and RVs, which require wider exposed body panels (hoods, doors, roofs, tailgates) from a single coil with minimal welding. This is a capability play, not purely a volume play.
  • Lower-carbon automotive sheet. The line is designed to take steel produced via POSCO's Haptang (합탕) route, in which molten iron from the new Gwangyang electric arc furnace is mixed with blast-furnace hot metal.
  • Premium exposed-panel quality. POSCO positions 8CGL as a "POSCO-developed facility" (포스코형 자동차강판 특화 설비) — meaning the line configuration is presented as proprietary rather than a reference plant.

The nine-year gap matters

POSCO has not built a new CGL since 2017. In the intervening period, global auto-sheet competition has moved on two fronts simultaneously: the specification race toward higher-strength and better surface quality for exposed panels, and the decarbonization race toward lower-embodied-carbon steel. 8CGL is POSCO's attempt to answer both in one asset rather than sequentially. That compression is why the furnace specification, not the pot capacity, is the headline engineering question.

The EAF connection: a separate project, already finished

One thing worth stating plainly, because it is easy to conflate: the associated electric arc furnace is not part of 8CGL. It is a separate, completed project.

The Gwangyang EAF was started in February 2024, inaugurated on 17 June 2026 — roughly three months before 8CGL's groundbreaking — and represents about KRW 600 billion in investment for 2.5 million tonnes per year of capacity, described as South Korea's largest single EAF. Tenova supplied the furnace with its Consteel continuous scrap charging and preheating system and Consteerrer electromagnetic stirring. POSCO cites up to ~75% lower carbon emissions versus the blast-furnace route on a crude-steel Scope 1 and 2 basis (against a 2017–2019 baseline), a figure that is contingent on scrap supply quality and the power generation mix.

The strategic logic chains together as follows:

LayerFacilityStatusRole in the chain
Low-carbon primary steelGwangyang EAF, 2.5 Mt/yCompleted June 2026Produces EAF hot metal; blended via Haptang
Downstream processingGwangyang 8CGL, 450 kt/yGroundbreaking Sept 2026, completion 2028Converts EAF-compatible substrate into exposed-grade HDG
End marketAutomotive OEMs (domestic and export)—Low-carbon, extra-wide exposed panels

POSCO's stated target is mass production of automotive and electrical steel sheets from EAF-based premium steel by 2030, with these products designated among its "Eight Strategic Products." 8CGL is best understood as the downstream half of a two-stage bet whose upstream half is already standing. A CGL that cannot process EAF-derived substrate would strand the EAF as a long-products or commodities asset; 8CGL is what converts the EAF from a decarbonization statement into an automotive-qualified product line.

Why the furnace is the hard part

Hot-dip galvanizing is a mature process. Coil is unwound and continuously cleaned, annealed, immersed in molten zinc, cooled, temper-rolled, surface-treated and recoiled, with air knives (high-pressure gas jets) metering coating weight. The metallurgical challenge in POSCO's case sits in the annealing and heat-treatment stage.

EAF-based steel carries a different residual-element profile — copper, tin, nickel and other tramp elements from scrap — and different thermal history than blast-furnace steel. Achieving the surface cleanliness and mechanical properties required for exposed automotive panels (Class A surface, where any defect is visible after paint) from this substrate is materially harder than for unexposed parts. The ultra-high-temperature heat treatment system POSCO describes is designed around that problem.

Notably, POSCO has not disclosed the maximum temperature or residence-time specifications. That omission is itself informative: this is where the competitive differentiation lives, and POSCO is treating it as proprietary process know-how rather than marketing headline material.

The second disclosed technology layer is control and automation:

  • AI-based process control drawing on what POSCO describes as more than 30 years of operational data, analyzing material, process and quality parameters in real time and adjusting equipment to reduce quality variation and defect rates.
  • AI and robotics for high-risk tasks, specifically dross removal from the molten zinc bath and sample collection — the two routine operations most associated with heat exposure and ergonomic risk on a CGL.

This is a plausible and, in the broader industry, increasingly common automation package. The distinguishing claim is data depth: three decades of Gwangyang operating history feeding the models, rather than a greenfield digital twin built from scratch.

The contracting picture: strong signals, no confirmation

The public record on who is building 8CGL is thinner than one would expect for a KRW 480 billion project with a two-year construction window.

What is established: POSCO is owner, developer and operator. That is unambiguous.

What is strongly indicated but not formally confirmed: POSCO E&C is running the "Gwangyang 8CGL New Construction Site." Two independent hiring postings on the Korean recruitment platform Saramin carry POSCO E&C's corporate name and the 8CGL site designation:

  • A quality manager posting for the 8CGL site within Gwangyang Works, with a stated contract term of 16 March 2026 to 31 October 2028 (one-year renewable), posted 11 February–13 March 2026.
  • A civil works site manager posting covering civil construction supervision, schedule/quality/safety compliance, and coordination of subcontractors and materials, posted 26 May–15 June 2026.

The March 2026 start date matters: it places POSCO E&C personnel on site roughly six months before the September groundbreaking ceremony. Site mobilization ahead of a public ceremony is normal practice, but it is worth noting that the visible construction timeline is longer than the ceremonial one.

A building permit filing for "Geumho-dong 8CGL main plant" (금호동 8CGL본공장) in Gwangyang, dated 2026, corroborates a project approval in that year. A caveat worth flagging: the aggregator listing associated with this filing gives a floor area of 1,422,966 m², which is implausible for a 450 kt/y galvanizing line and is almost certainly a data-entry artifact. Treat the permit existence as corroborative and the area figure as unreliable.

What is genuinely unknown:

ElementStatus
Main EPC contractor and contract valueNot publicly disclosed
Joint-venture / consortium compositionNot disclosed
Line OEM (furnace, pot equipment, automation)Not identified in any source
Package-level awards (e.g., electrical)A "Gwangyang 8CGL new main works electrical works 1" package appears in a construction database; contract parties and value not verifiable
Steel procurement (structural, plate, sections)No evidence of tender or award found

On the equipment side, searches surface plausible CGL technology suppliers — Primetals Technologies for CGL automation and coating-weight control, ANDRITZ for continuous galvanizing line furnaces, Tenova for strip processing — but no source links any of them to Gwangyang 8CGL. Tenova's relationship with POSCO Gwangyang is confined, on the available evidence, to the EAF and its Consteel and Consteerrer systems. Anyone inferring a Tenova CGL award from the EAF relationship is drawing a conclusion the record does not support.

Why the silence is worth noting rather than reading into

POSCO describing 8CGL as a "POSCO-developed facility" is consistent with several scenarios, and the available evidence does not distinguish between them:

  1. A deliberately in-house-engineered line with selective equipment sourcing — plausible given the proprietary heat-treatment requirement.
  2. Conventional OEM supply with POSCO retaining integration and control IP — the most common arrangement in the industry.
  3. Disclosure simply deferred to a later announcement cycle.

What can be said with confidence is that a KRW 480 billion facility with a 2028 completion target requires its major equipment orders to be placed well before this article's publication window. The absence is a disclosure gap, not evidence of an empty supply chain.

Reading the market signal

Three things make 8CGL more than a routine capacity addition.

First, capacity discipline. POSCO is adding roughly 10% to HDG capacity after a nine-year pause. Korean and broader Asian auto-sheet markets have faced persistent overcapacity concerns; a 450 kt/y addition is calibrated rather than aggressive. The framing around capability — extra-wide exposed panels — rather than tonnage suggests the investment case rests on product mix and price premium, not volume.

Second, the low-carbon premium is being industrialised, not piloted. The Gwangyang EAF is built; 8CGL is the finishing step that makes its output saleable to automotive OEMs with supply-chain emissions targets. This matters because automotive qualification for exposed panels is slow — typically 18–36 months of trials, OEM approvals and part-level validation. A 2028 mechanical completion plus qualification implies meaningful commercial volumes of EAF-based exposed auto sheet in the 2029–2030 window, which aligns with POSCO's stated 2030 mass-production target.

Third, the steel-on-steel demand question. A 450 kt/y CGL is itself a substantial steel consumer during construction — structural sections, plate, rebar, process piping, and large volumes of stainless and specialty alloys in the zinc pot, snout, rolls and ductwork. For the domestic construction supply chain, this is a meaningful but unglamorous demand signal that has received essentially no public attention. POSCO E&C's civil works hiring in mid-2026 suggests that phase is under way or imminent, though no steel procurement tenders have surfaced.

What to watch

  • The EPC award. Whether POSCO E&C's apparent site role is formalised as a disclosed contract, and whether the scope is sole-contractor or consortium. A joint-venture structure would open the field to the major Korean industrial contractors.
  • The line OEM. Watch for announcements from Primetals, ANDRITZ, Tenova or a domestic supplier. The heat-treatment furnace specification will be the tell — a supplier naming POSCO Gwangyang 8CGL in a reference list would resolve the question immediately.
  • Haptang ratio and product qualification. POSCO has not disclosed what proportion of EAF hot metal the 8CGL substrate will contain, nor which automotive OEMs have committed to qualification trials. That ratio determines how much of a genuine decarbonization claim the line supports versus a marginal blending story.
  • Construction steel procurement. Any tender activity for structural steel, plate or sections tied to the Gwangyang 8CGL site would be a leading indicator that civil and structural works are scaling.
  • Pohang HyREX. POSCO's hydrogen-reduction demonstration (300,000 t/y, commercialization technology targeted by 2030) is the longer-horizon decarbonization leg. If HyREX progresses, the question becomes whether CGL capacity built for EAF substrate can also absorb hydrogen-reduced substrate. Designing 8CGL for flexibility now would be the strategically coherent choice.

Assessment

8CGL is a strategically coherent project whose public disclosure is unusually thin on the commercial side. The technical intent — a galvanizing line that can turn electric-furnace steel into Class A exposed automotive panels — is the correct answer to where both automotive procurement and steel decarbonization are heading. The KRW 480 billion price tag and 450 kt/y scale are consistent with a capability-focused rather than volume-focused investment.

The principal uncertainty is not whether the project will be built — site mobilization evidence and permit filings suggest it is already under way — but who is building it and with whose equipment. For suppliers and contractors in the steel plant ecosystem, that is the actionable gap in the public record. For competitors, the number to watch is not 450,000 tonnes; it is the heat-treatment specification POSCO has chosen not to publish.

  1. 2026-09-08POSCO announced an approximately KRW 480 billion investment to establish a 450,000-ton-per-year automotive hot-dip galvanized steel plant at Gwangyang Works.
  2. 2026-09-08Construction of the 8CGL continuous galvanizing line began at POSCO’s Gwangyang Works.
Recorded development timeline for POSCO 8CGL automotive hot-dip galvanized steel line at Gwangyang Works.Credit: MetalMate · Original diagram (MetalMate)