China's MIIT puts carbon, not energy, at the centre of industrial compliance: reading the 2026–2030 dual-control plan

MIIT published, on 21 October 2026, the Implementation Plan for Accelerating the Construction of the Carbon Emission Dual-Control Work System in Industry and Information Technology (2026–2030). The document itself has not been widely circulated in full text, but the ministry's own summary is specific enough to establish what changes for industrial operators — and, for steel, what it likely means.

The headline is institutional rather than technological. China is converting the slogans of the 15th Five-Year Plan into a measurable compliance architecture in which the unit of account is carbon, not energy.

What was actually issued

The Plan is framed as implementing three upstream documents: the CPC/State Council General Office opinion on higher-quality energy-saving and carbon-reduction work; the Measures for Comprehensive Evaluation and Assessment of Carbon Peaking and Carbon Neutrality; and the State Council General Office work plan for building the dual-control institutional system. Its mandate, in MIIT's words, is the "steady implementation" of five strands already named in the 15th FYP Outline: local carbon assessment (地方碳考核), sectoral carbon control (行业碳管控), enterprise carbon management (企业碳管理), project carbon evaluation (项目碳评价) and product carbon footprint (产品碳足迹).

That five-layer stack is the real substance. Each layer answers a different question, and together they close the gaps that the previous energy-intensity regime left open.

The announcement organises delivery into four work streams that map onto this stack: key industries (enforcement, diagnostics, technology, standards); enterprises and parks (zero-carbon factories, zero-carbon computing facilities, zero-carbon parks, low- and zero-carbon supply chains, with green power procurement and digital MRV as the named pathways); industrial products (standards as lead, database as support, application scenarios as the pull); and localities and projects (carbon accounting mechanisms, project-level emission substitution, assessment and local capacity building).

The four numbers that matter

Three points deserve emphasis.

First, the intensity gap. A 17% carbon cut against a 10% energy cut implies roughly 7–8% of the reduction must come from decarbonising the energy actually consumed — cleaner electricity, fuel switching, process change — rather than from efficiency alone. That is a materially different ask from previous five-year cycles, when efficiency did most of the work.

Second, the denominator problem. Both targets are value-added-denominated. For cyclically depressed sectors such as steel, flat or falling value-added mechanically tightens the required absolute reduction. Illustratively (my calculation, not a policy statement): if steel value-added were 10% lower in 2030 than 2025, meeting a 17% intensity cut would require an absolute CO2 reduction of roughly 1 − 0.83 × 0.90 ≈ 25%. Falling prices and volumes therefore convert an intensity target into something close to an absolute cap — an uncomfortable feature of value-added-based metrics in deflationary industrial conditions.

Third, the zero-carbon count. The ~500 figure covers factories and computing facilities combined, not 500 of each. Given the scale of China's data-centre build-out and the political priority attached to "zero-carbon computing," the split between the two categories will be a useful indicator of where MIIT's green-power and grid-connection attention is actually directed.

Why project carbon evaluation is the sleeper provision

Of the five strands, project carbon evaluation carries the most immediate consequences for heavy industry. The Plan calls for carbon emission substitution and evaluation for "two-high" (energy-intensive, high-emission) industrial projects.

In practice, this is the mechanism through which provincial authorities have been rationing carbon space for new capacity. Substitution — replacing emissions elsewhere before a new project is approved — has been applied unevenly across provinces and sectors, often with replacement ratios above 1:1. Elevating it into a national work system implies greater standardisation of how baselines, ratios and verification are set, and a firmer legal footing for refusing approvals.

Steel is not named in the MIIT announcement, and readers should treat the following as analysis rather than reported fact. But steel is the archetypal two-high sector, and project-level carbon evaluation would sit directly in front of new EAF plants, DRI/hydrogen projects, coking and sintering units, and blast furnace relines or conversions. The practical question for project developers is not whether the policy applies, but what substitution ratio and baseline methodology their province will use — details that will emerge in provincial implementing rules, not in the MIIT document.

The product footprint layer is where steel meets trade policy

Four hundred product-category rules by 2030 is an ambitious expansion. For steel, the commercially decisive question is alignment with the EU's CBAM methodology — system boundaries, embedded emissions, treatment of precursors such as pig iron and ferroalloys, and the acceptability of default versus verified values.

A Chinese steel footprint standard that diverges methodologically from CBAM expectations would leave exporters running two parallel accounting systems, or paying default values that penalise them regardless of actual performance. A converging standard, backed by a credible national database, would be an export asset. The MIIT announcement establishes the architecture — standards, database, application scenarios — but says nothing verifiable about methodological alignment. That gap is the single most important thing to watch for steel.

Supply-chain benchmarking as a transmission channel

The Plan's inclusion of "low- and zero-carbon supply chains" alongside zero-carbon park and factory benchmarks is easy to skim past. Its significance is that it creates a domestic, policy-endorsed route by which downstream buyers — automakers, appliance makers, machinery OEMs — can cascade footprint requirements onto steel and metals suppliers, independent of export market pressure. Combined with digital energy-carbon management requirements at enterprise level, the infrastructure for supplier-level carbon data requests is being built now.

What remains unresolved

• The full Plan text, including annexes, sector lists, whether steel carries specific numeric targets, and the certification criteria for "national-level" zero-carbon factories. The MIIT summary is a policy brief, not the operative document.
• Provincial implementing rules (Hebei, Jiangsu, Shandong, Liaoning), which historically set the substitution ratios that determine project economics.
• CBAM methodological alignment for steel product rules.
• Industry reaction, including from CISA, which had not surfaced at the time of writing.

The judgment

The plan is best read not as new ambition but as new machinery. China already had targets; what it lacked was a coherent chain running from provincial scorecards through plant-level data to product-level footprints that trading partners can recognise. Building that chain — with 400 standards and a national database as its backbone — is a multi-year institutional project in which the metals industries are both the largest test case and the most exposed.

For steel specifically, the sequence to watch is: provincial substitution ratios, then the steel product-category rule and its CBAM compatibility, then corporate and buyer-driven footprint requirements arriving through supply-chain benchmarking. Each is more commercially immediate than the last headline number.