Research & Publications

Version 1.1 · 2026 · six chapters and appendices · CC BY 4.0

Certifying Carbon Removal in Taiwan

Benchmarking the EU CRCF and the Case for an Independent Certification Framework

DOI: https://doi.org/10.5281/zenodo.21618472

Key findings

The gaps sit at three levels and cannot substitute for one another. Taiwan's NDC 3.0 sets a 2035 reduction target but does not disaggregate removals into a separate quantified commitment. Every approved or pending removal methodology addresses natural sinks; engineered and material-based removals are entirely absent. The verification regime recognises organisational, project and product-footprint categories, with no removal category capable of handling permanence monitoring, reversal risk or the termination of monitoring obligations.

The EU framework is closed to Taiwan by legal design, not by timing. Quality criteria apply only to activities carried out within the Union; all units count toward the EU's own NDC and may not be counted toward a third country's; and at methodology level, biochar production facilities and storage locations must both be located within the Union. Waiting does not eventually yield a transplantable framework — it yields a framework that stays closed.

The legal threshold for action is substantially lower than commonly assumed. Of six recommendations, five require no legislative amendment: a removal verification category can be created by administrative notice under an existing catch-all provision, and engineered removal methodologies fall squarely within the competent authority's existing approval powers. Only one — excluding removal units from carbon fee offsetting — requires amending subordinate legislation. None requires amending the parent act. The obstacle is priority, not authority.

Durability thresholds are policy choices, not scientific constants. Across six major biochar standards, H/Corg thresholds differ by a factor of 1.75 and assessment horizons by an order of magnitude, while the underlying literature is largely shared. Taiwan should therefore choose rather than wait, prioritising international compatibility, and should invest in the two gaps common to every existing standard: decay parameters calibrated for high-temperature soils, and models for non-soil applications.

Back-casting from a 2030 target for first issuance of removal units, methodology work had to begin in the second half of 2026.

Cite as

Version 1.0 · 2026 · twenty-four chapters and eight appendices · CC BY 4.0

Fulfilling Ongoing Emissions Responsibility

An Analysis of the OER Framework under the SBTi Corporate Net-Zero Standard V2.0 and Its Implications for Taiwan

DOI: https://doi.org/10.5281/zenodo.22869799

Key findings

The OER recognition programme does not require carbon removals. Eligible verified mitigation outcomes include value-chain-external reductions and the restoration of natural sinks, and contribution budgets may fund research, adaptation or loss and damage. Mandatory removals appear only in the post-2035 requirement, which the standard itself labels illustrative, and in net-zero neutralisation.

The 2035 long-lived removal quota is far smaller than commonly reported. The denominator in C45.4 is covered emissions attributable to long-lived GHGs — not total ongoing emissions. Combined with the 1% coverage level set for 2035, the minimum obligation is on the order of one thousandth of ongoing scope 1–3 emissions. The burden is in the slope, not the starting point: two linear ramps multiply into quadratic growth, so a company at ninety tonnes in 2035 reaches ninety thousand by 2050.

Buying now to hold for 2035 does not work. Read together, three criteria foreclose it: outcomes already used for OER cannot be reused for post-2035 or net-zero requirements; eligible outcomes must have occurred within five years of the reporting year; and qualifying removals must be delivered within the same reporting period as the emissions they cover. The constraint is shelf life, not price.

For Taiwan, three institutional facts converge. Taiwan is classified as a high-income economy and is therefore subject to the standalone 10,000 tCO₂e scope 1–2 trigger for Category A — below the 25,000 tCO₂e threshold of the domestic carbon fee, so the standard reaches further than the fee does. Unit prices are comparable while aggregate obligations are not: the Advanced contribution budget of US$20/t converts to roughly NT$637, below the rate committee's NT$1,200–1,800 benchmark for the post-2030 general rate, but OER is calculated on five-year cumulative scope 1–3 emissions with no exemption threshold. And Taiwan's voluntary reduction methodologies contain no durable removal category, so no qualifying domestic supply exists — an absence of classification rather than of capacity.

Responsibility travels down the supply chain. A company may share scope 3 coverage responsibility with a value-chain partner that reports the same emissions, provided at least one party explicitly assumes it and a written agreement is filed. The only condition the text places on the assuming party is that it reports the same emissions — it need not participate in the recognition programme, hold a validated target, or maintain a full annual scope 3 inventory. A Taiwanese supplier that has never engaged with the SBTi can therefore receive a sharing request from 2027 onwards.

Cite as