China's PVC sector enters 2026 with the tightest supply backdrop in over a decade: only ~800,000 tonnes of new nameplate capacity is scheduled across 2026-2028, and the 2027 line-up currently shows zero additions, leaving operating rates and inventory drawdown sensitive to any unplanned outage [S2][S3].
The chlorine-alkali linkage defines the upstream floor. Each tonne of caustic soda co-produces ~0.88 t of chlorine, and PVC absorbs roughly 46% of that chlorine output, making PVC the single largest sink for chlor-alkali chlorine and the structural backstop for any chlor-alkali operating-rate decision [S2]. On the downstream side, the same report maps PVC into construction pipe, profiles, wire and cable sheathing, synthetic leather, and decorative film — the demand basket that decides whether an integrated plant idles a line or runs it flat out [S2]. For the broader synthetic resin chain context, this PVC block sits among the five major commodity resins, ranking third by volume behind polyethylene and polypropylene [S2].
Upstream: Ethylene vs Calcium Carbide Routes, Caustic-Chlorine Balance
Two process routes feed global PVC: the ethylene route (EDC/VCM → PVC) used predominantly in the US, Western Europe, and Northeast Asia, and the calcium-carbide route (CaC₂ → acetylene → VCM → PVC) dominant in western China where coal-derived power and limestone co-locate [S2][S3]. Geopolitical conflict has pushed up overseas ethylene-route PVC costs, leading to operating rates below 七成 (seven-tenths) at Japan, South Korea and European facilities, a factor in the global PVC price formation chain [S3].
The chlor-alkali balance ties PVC economics to caustic soda demand rather than to PVC itself. When caustic demand is weak, integrated producers face a choice: cut chlor-alkali rates and starve PVC, or run full and build PVC inventory. The 0.88 t Cl₂ per tonne NaOH ratio and the ~46% chlorine consumption share for PVC are the two numbers that govern that decision tree [S2]. For spec engineers sourcing PVC-U pipe feedstock, the practical consequence is that spot PVC offers track caustic soda export volumes more closely than they track crude oil, a non-obvious linkage that often catches Western buyers off guard [S2].
Midstream: 2026-2028 Capacity Pipeline, Regional Concentration
The Chinese capacity pipeline thins to roughly 800,000 tonnes cumulatively over 2026-2028, with 2027 currently showing no announced additions — a posture last seen in the 2015-2016 capacity rationalisation window, before the next investment wave broke in 2017-2020 [S3]. That 80-figure includes both ethylene and calcium-carbide additions, and the report explicitly contrasts the new-build drought against an operating-rate environment where integrated producers have already cut runs to defend cash margin [S3].
Regionally, calcium-carbide capacity clusters in Xinjiang, Inner Mongolia, Shaanxi, and Ningxia — provinces where coal feedstock cost and captive power are the binding constraints — while ethylene-route capacity sits largely in coastal provinces (Shandong, Jiangsu, Zhejiang, Tianjin) where VCM import logistics and ethylene availability dominate the spec [S2]. The new 80-figure is heavily weighted toward the coal-belt cluster, which means the next marginal tonne of Chinese PVC will be a carbide-route tonne priced off coal and electricity rather than off Brent [S3]. For procurement teams mapping synthetic resin exposure, that regional concentration is the structural reason why domestic Chinese carbide PVC consistently under-prices imported ethylene-route PVC by a working spread, even after freight and import duty.
Downstream: Construction, Pipe, Profiles, Wire and Cable
PVC's downstream basket is construction-heavy and price-sensitive. Building materials, rigid pipe, window and door profiles, wire and cable insulation, synthetic leather, and decorative film are PVC's main application areas [S2]. The pipe sub-segment is the largest single sink and feeds directly into water supply, drainage, irrigation, conduit, and cable ducting, with PVC-U pipe the dominant specification for non-pressure building services lines.
Profiles and wire-and-cable sheathing are the next two demand pillars. Window and door profile demand tracks property completions with a 6-12 month lag, which means the 2025-2026 property weakness feeds into a softer 2026 profile order book; wire and cable sheathing, by contrast, is tied to grid capex and EV charging rollout, both of which are counter-cyclical and partly offset the construction drag [S2]. Downstream converters operate on typically thin 3-8% gross margins, so any PVC price step above mid-single-digit percent usually forces a working-capital destock at the converter level before the resin price itself rolls over [S2].
Trade Flow: China Overtakes US as #1 Exporter
China exported a record 3.8 million tonnes of PVC in 2025, equivalent to roughly 16% of domestic production, and overtook the United States to become the world's largest PVC exporter in the same year [S3]. That 16% ratio is the structural release valve: in a flat domestic demand environment, every incremental export tonne keeps operating rates above the cash-cost floor and prevents inventory from building into a price-busting overhang [S3]. The export pull is what distinguishes the 2026-2028 outlook from a pure domestic balance-sheet story.
Geopolitical disruption has reshaped the competitive map. Overseas ethylene-route PVC plants in Japan, Korea, and parts of Europe are running at less than 70% utilisation as their oil-linked feedstock cost curve has stepped up, which opens price room for Chinese carbide-route exports into South-east Asia, India, Turkey, and Latin America [S3]. The risk to that flow is freight, anti-dumping exposure, and any reversal in the oil-vs-coal spread — a reversal that would re-price the export arbitrage within a quarter rather than a year. For PVC-U pipe projects outside China, the working assumption through 2026 is that Chinese export offer remains the marginal price-setter.
Comparison: Ethylene-Route vs Calcium-Carbide PVC on Four Buyer Criteria
For a buyer or spec engineer weighing the two routes, four criteria do most of the work. Feedstock exposure: ethylene-route PVC follows the crude/naphtha/ethylene chain; carbide-route follows the coal/power/lime chain — and they decouple during oil spikes, which is exactly the 2026 set-up [S3]. Carbon intensity: carbide-route carries materially higher process CO₂ per tonne of PVC and is increasingly exposed to Chinese domestic carbon-cost pass-through, while ethylene-route intensity tracks cracker configuration. Product consistency: ethylene-route typically yields tighter molecular-weight distribution and lower residual VCM, which is why downstream medical-grade and high-clarity film converters prefer it; carbide-route dominates rigid pipe and profile grades where colour and clarity tolerance are wider [S2]. Cash-cost position: at today's coal-to-oil price ratio, Chinese carbide-route sits at or near the global cash-cost floor, with the export channel as the price-formation mechanism [S3].
Process Signals and Where the Chain Breaks Next
Three signals are worth tracking through the second half of 2026. First, the 2027 new-capacity column — any project sanction against the current zero-addition baseline re-prices the entire 2027-2028 balance, and the calcium-carbete provinces are the ones to watch [S3]. Second, chlor-alkali operating rates, which lead PVC output by 4-8 weeks; a sustained drop below 80% utilisation usually means PVC supply tightens within a month, while a climb above 90% typically presages inventory build [S2]. Third, export volume to South-east Asia and India: monthly customs prints above 350 kt run-rate confirm the export-lifeline narrative, while a drop below 250 kt points to arbitrage fatigue or freight friction [S3]. For deeper sourcing context across adjacent chains, see the spec map for epoxy resin supply and the broader PVC resin supply chain 2026 working map.
For the relevant spec sheets and selection criteria, see resin sand line.