Biochar: Carbon Removal, Soil Response, and the Accounting Question

A sustainability lead at a food brand asks whether the cooperative's biochar programme is “basically charcoal in the soil.” The agronomist on the call says no, the salesperson says yes, the carbon-finance lead says both, and the buyer leaves the meeting unsure whether the tonnes on the invoice are a removal, an amendment, or a marketing claim. The question is more answerable than the meeting suggests, but the answer travels with three numbers, not one.

The short answer to “isn't it basically charcoal?” is yes. Biochar is a charcoal, specifically the soil-application sub-class, produced under controlled pyrolysis conditions, classified by the H/Corg molar ratio, and certified against a contaminant ceiling. That sub-classification is what makes its carbon storage accountable as a removal. The 2026 standards cycle (Verra VM0044 v1.2 active since June 2025, the EU CRCF permanent-removal methodologies adopted in February 2026, the EBC guidelines now at version 10.5, and the Puro.earth Edition 2025 methodology) has converged on a small set of rules for what counts, who can claim it, and how the carbon-removal tonnage should be discounted for the part that will not stay put1234.

Charcoal, biochar, and the chemistry that decides

All charcoal is biomass pyrolysed in an oxygen-limited environment. Biochar is the sub-class of charcoal made specifically as a soil amendment and carbon sink: higher carbonisation, smaller particles, lower volatile-organic content, and a stable aromatic structure that resists microbial decomposition for decades to centuries5. Charcoal produced for combustion (the BBQ-grade, large-piece, energy-yield-optimised material most people picture) sits at the other end of the same production spectrum: same fundamental process, different specifications and a different intended fate.

The operational threshold that classifies a charcoal as biochar is the molar ratio of hydrogen to organic carbon, H/Corg. Pyrolysis at increasing temperature drives off hydrogen-rich volatiles and reorganises the residue into fused aromatic sheets. The H/Corg drops monotonically with pyrolysis temperature, and the IPCC 2019 Refinement, the IBI Biochar Standards, and the European Biochar Certificate all adopt H/Corg ≤ 0.7 as the upper bound for a material that can be classified as biochar; the EBC C-Sink and most credit protocols tighten this to H/Corg ≤ 0.4 for the most permanent grade163. Above 0.7 the material is, for accounting purposes, fuel-grade char or torrefied biomass: still a charcoal in the family sense, but not certifiable as biochar for soil or removal-credit purposes.

Two practical consequences follow. Pyrolysis temperature must exceed roughly 350 °C for the IPCC default permanence factors to apply, and the EBC additionally imposes ceilings on polycyclic aromatic hydrocarbons (PAHs ≤ 4 mg kg⁻¹ for the premium grade, ≤ 12 mg kg⁻¹ for the basic grade), heavy metals, and dioxins to keep contaminants out of the food chain3. Uncertified biochar can carry PAH loads an order of magnitude higher and is not a substitute, agronomically or for credit issuance.

Why most charcoal stays out of the field

The world produces tens of millions of tonnes of charcoal a year, the overwhelming majority of which is fuel-grade and has no business on a field. The gap between “charcoal” and “biochar” is not categorical in the chemistry; it is operational in the production controls, the feedstock rules, and what may or may not be added. Three classes of risk explain why the certification regime exists.

Contaminants from the production process. Low-temperature, partially oxidative, smouldering kilns (the dominant production mode across much of the world) produce a residue carrying polycyclic aromatic hydrocarbons (PAHs), the condensation products of incomplete pyrolysis. Hale and colleagues (2012) systematically characterised PAH and dioxin residues across commercial and traditional-kiln biochars and found Σ16-EPA-PAH loads varying by more than an order of magnitude between batches, with uncharacterised material routinely exceeding the EBC basic-grade ceiling18. PAHs are persistent, several of the EPA-listed species are demonstrated carcinogens, and once on cropland they accumulate in plant tissue and enter the food chain. Controlled, higher-temperature pyrolysis with proper gas treatment is what brings the PAH load down to certifiable levels, the same production conditions that drop H/Corg below 0.7.

Contaminants from the feedstock. Charcoal concentrates whatever the biomass contained. Pressure-treated lumber, painted or varnished wood, demolition material, sewage sludge, and biomass grown on contaminated soil all carry heavy metals (chromium, copper, arsenic, lead, cadmium) and synthetic compounds that survive pyrolysis and end up in the residue at multiples of their original concentration. EBC and the IBI Biochar Standards exclude these feedstocks categorically; certified biochar is restricted to a defined-source positive list (forestry residues, named agricultural residues, and a small set of waste streams under specific protocols). The heavy-metal ceilings published by EBC are calibrated against EU Regulation 2019/1009 (Fertilising Products) and the German Federal Soil Protection Ordinance, which are soil-protection thresholds, not industry-friendly defaults3.

Added materials. Commercial BBQ briquettes typically contain starch or lignin binders, paraffin or petroleum-based accelerants, sodium nitrate or borax for ignition behaviour, and sometimes flavourings. None belong in soil. A charcoal product without a full materials declaration cannot, by definition, be certified for soil application; the absence of a declaration is itself the disqualifier, even if a specific batch would have passed on test.

The operational answer to “which charcoals are good for soil?” is therefore not a chemistry test in isolation; it is a small, concrete checklist that any project should be working against from the start. The EBC publishes the list; its specific thresholds are the reference, but the structure of the checklist is what matters for procurement.

The short answer to “is biochar just charcoal?” is therefore: yes, biochar is a charcoal. The longer answer is that most charcoal is not biochar. The gap is not an accident. The certification regime exists because the failure modes (food-chain contamination, ecotoxicity, regulatory non-compliance, liability) are durable and expensive when they happen, and the cost of the tests that close the gap is small relative to the consequences of skipping them.

What the yield evidence actually says

The most cited single result is Jeffery et al. (2017): biochar elicits a 25 percent average yield increase in the tropics and no statistically detectable effect on average in temperate latitudes. The mechanism in the tropics is liming of acidic, low-CEC soils plus a co-fertilisation effect from biochar-borne ash; in temperate cropland, well-fertilised soils at near-neutral pH gain little because the limiting factors biochar addresses are not present7.

More recent meta-analyses refine this without overturning it. Ye et al. (2020), across 153 studies and over 1,100 paired observations, find a 16 percent mean yield response with biochar alone and a 17 percent response when biochar is co-applied with fertiliser, with the strongest gains on acidic, sandy, and low-organic-carbon soils8. The 2025 global meta-analysis of Aurangzeib and colleagues isolates the response on acidic soils (pH 5–6, coarse texture, tropical climate): +38 percent yield, with concurrent improvements in porosity (+23 percent), mean weight diameter (+46 percent), and a bulk-density reduction of −19 percent9.

Three honest qualifications travel with these numbers. Effective rates are high (most positive responses are at 10–40 t ha⁻¹, an order of magnitude above casual application). Wood feedstocks pyrolysed at 450–500 °C dominate the positive responses; manure and crop-residue biochars at low temperatures perform less consistently. And the between-study standard deviation in temperate trials is substantially larger than the mean, so a temperate project that assumes a tropical-magnitude response on its own soils is making a statistical bet that the literature does not support.

How much of the carbon actually stays put?

The carbon-removal claim rests on persistence. Biochar carbon decomposes more slowly than fresh organic matter because most of it is locked in fused aromatic sheets that microbial enzymes cannot easily cleave. Lehmann and colleagues (2021), reviewing the field after a decade of incubation and field-trial data, conclude that biochar mean residence time scales with the degree of carbonisation, and that the H/Corg molar ratio is the most robust single proxy5.

The IPCC 2019 Refinement formalised the linkage as a default Tier-1 method. The fraction of biochar carbon remaining after 100 years (Fperm) is computed from the H/Corg ratio as Fperm = 1.04 − 0.635 × (H/Corg), with the formula calibrated against incubation meta-analyses; a Tier-2 user can substitute regionally calibrated values1. Woolf and colleagues (2021) re-fit the underlying decay data with a two-pool exponential model and conclude that 63–82 percent of biochar carbon remains unmineralised at 100 years across the global cropland mean annual temperature, with the spread driven by pyrolysis temperature and feedstock10.

The calculator below applies the IPCC formula to user-set H/Corg and extrapolates to a 200-year horizon (the basis the EU CRCF adopted in February 2026 for biochar removal certification). The blue band is the carbon that gets counted at the chosen horizon; the white area above the curve is the carbon that is discounted in advance.

Two design implications follow. First, the slope from H/Corg = 0.4 (premium grade) to H/Corg = 0.7 (basic grade) costs roughly twenty percentage points of counted carbon at 100 years; doubling the horizon to 200 years compounds the penalty. Second, Petersen et al. (2025) caution that the linear H/Corg → Fperm relationship explains only a fraction of observed variance, and that programme designers building a project economic case off a single Fperm value without feedstock-specific data are extrapolating beyond what the underlying meta-analysis supports11. The 2026 EU CRCF methodology accordingly requires either analytical reflectance testing of the biochar or a calibrated decay model, not the formula alone, for the full certification grade2.

What “permanent” means in the 2026 regimes

Four standards converged through 2024–2026, each codifying permanence differently but agreeing on the general shape: a permanence factor tied to H/Corg, a life-cycle accounting boundary, and a chain-of-custody discipline to prevent double counting.

• Verra VM0044 v1.2 became active 27 June 2025 and was approved by the ICVCM as Core Carbon Principles-eligible. It requires investment-analysis additionality (not solely counterfactual baselines), credits removals to the pyrolysis operator, and assigns a discounted Fperm to the durable fraction of biochar carbon4.
• EU CRCF biochar methodology was adopted by the European Commission in February 2026 under Regulation 2024/3012. It scores permanence over a 200-year horizon (rather than 100), caps reporting uncertainty at ±20 percent, requires re-certification every five years, and only credits the fraction of biochar mass that passes either an analytical reflectance test or a calibrated decay model. The activity-period choice was tightened from 10 to 5 years to accommodate ongoing methodological refinement2.
• Puro.earth Biochar Methodology Edition 2025 uses cradle-to-grave system boundaries, classifies projects into new-built, retrofit, and charcoal-repurpose categories, and requires explicit baseline modelling for the alternative biomass fate, the alternative production assets, and the alternative end-use12.
• European Biochar Certificate (EBC) v10.5 sets the product-quality floor (H/Corg, PAHs, heavy metals, dioxins) that most credit programmes layer atop. The EBC C-Sink certificate, used by the Carbon Standards International registry, is the sister mechanism issuing the actual removal credit on EBC-certified material3.

A project designed to one of these standards is not automatically eligible under another. The EU CRCF's 200-year horizon and ±20 percent uncertainty cap, in particular, exclude a non-trivial fraction of biochars that comfortably pass VM0044 or Puro at 100 years. Choosing the protocol the buyer will be audited against (and designing to it from the start) is the operational decision that determines whether the tonnes survive verification.

The full life-cycle balance

The permanent-carbon number is only one component of the net-CO₂e balance. The published life-cycle assessments converge on a useful range: Roberts et al. (2010) reported net removals of −0.86 to −0.89 tCO₂e per dry tonne of feedstock for corn stover and yard waste, with the carbon-storage component contributing roughly two-thirds of the total13. Tisserant and Cherubini (2019), reviewing 34 biochar LCA studies, found a mean of −0.9 tCO₂e per tonne dry biomass with a range of −1.5 to 0 depending on feedstock, pyrolysis configuration, and bio-oil treatment14.

Five components dominate the balance, and the composer below lets you weight them. The biochar carbon stored is the largest negative term, but the counterfactual fate of the biomass (open burning, aerobic decay, or landfill with methane emissions) sets the avoided-emissions baseline; the pyrolysis energy balance can swing from a small benefit (when syngas and bio-oil displace fossil heat) to a meaningful cost (when external fuel supplements the reactor); transport adds up over hundreds of kilometres; and soil N₂O suppression contributes a small but real additional reduction.

Two findings from the composer matter for project economics. The counterfactual fate dominates the avoided-emissions side; a biochar project that displaces methane-emitting landfill (or open burning of agricultural residues) carries a larger net benefit than one that displaces aerobic composting, and a project that mis-classifies its counterfactual is the most common way the headline tonnage is overstated. Transport is rarely the decisive component below ~300 km, but it becomes one when the project ships biomass across continents to a centralised pyrolysis facility. The new generation of small, mobile units is in part a response to this geometry.

Who owns the removal?

The credit chain typically runs from biomass owner to pyrolysis operator to farmer (the applier) to off-take buyer. Each actor has a separate stake in the project, and each has a different permissible claim. The double-counting risk is not theoretical: a single tonne of biochar carbon, mis-attributed across the chain, can appear simultaneously in a buyer's retired-credit ledger, a farmer's Scope 3 inset, and a corporate sustainability report. The 2026 standards close this loophole by assigning the removal credit unambiguously to one actor and constraining what the others can say.

The diagram below shows the chain and the permissible claim for each actor under the three major protocols. Switch the toggle to compare how Verra VM0044, the EU CRCF, and Puro.earth resolve the ownership question. The credit is always assigned at one point in the chain; the other actors keep agronomic or co-benefit claims but cannot also count the removal itself.

Two operational disciplines follow. First, “insider” chains, where a food company contracts biochar production within its own supply shed for Scope 3 inset purposes, must satisfy a stricter chain-of-custody to avoid the same tonne showing up in both the company's emissions inventory and a separately retired credit. Verra and the EU CRCF have explicitly tightened this since 2024–202542. Second, farmers operating under a credit-backed application contract should treat the agronomic value (yield, soil pH, water retention) as the residual benefit they may legitimately claim, but not the tonne of CO₂ removed, which has already been retired against the buyer's target.

Side effects worth pricing in: N₂O, priming, PAHs

Biochar interacts with the soil's nitrogen and carbon cycles, and these second-order effects are large enough to belong in the inventory. Three deserve explicit treatment.

N₂O suppression. The Borchard et al. meta-analysis (2019) and successive updates report a mean reduction of soil N₂O emissions of 12 to 16 percent following biochar application on fertilised cropland, with the mechanism a combination of improved aeration (denitrification suppressed), increased abundance of the N₂O-reducing nosZ gene, and adsorption of NH₄⁺ and NO₃⁻ onto aromatic biochar surfaces15. This is a real co-benefit, but it is also small (around −0.05 to −0.08 tCO₂e per tonne biochar at typical N application rates), and it rarely tips the LCA verdict on its own.

Priming of native soil organic carbon. The concern that biochar accelerates mineralisation of existing soil carbon (positive priming) has not survived the meta-analytical evidence. Wang and colleagues (2016) and subsequent global meta-analyses report a net negative priming effect: biochar slightly slows decomposition of native SOC by roughly 3 to 4 percent, likely via physical encapsulation within aggregates and organo-mineral complexation16. The net effect on the soil carbon pool, summing the biochar carbon added and the small protective effect on native SOC, is positive.

Contaminants. PAHs, heavy metals, and dioxins are the most serious risk class for uncertified biochar. EBC-certified product is required to stay below 4 mg kg⁻¹ Σ16-EPA-PAHs at the premium grade and 12 mg kg⁻¹ at the basic grade; uncertified biochar from low-temperature or poorly controlled kilns can exceed these limits by an order of magnitude3. For any biochar destined for food-crop soils, the certification is not a marketing nicety but a regulatory floor, and a project that buys uncertified material is taking on a liability the credit revenue will not cover.

What survives an audit

Four numbers, supplied together, are the operational test for a credible biochar carbon claim under 2026 verification.

1. The H/Corg molar ratio measured on the actual product batch, with the pyrolysis temperature, residence time, and feedstock declared.
2. The applied permanence factor at the horizon the protocol requires (100 yr for IPCC and VM0044, 200 yr for the EU CRCF), and the test basis used (formula, reflectance, or decay model).
3. The cradle-to-grave LCA balance in tCO₂e per tonne biochar, with the counterfactual fate of the biomass explicitly identified and defended against leakage. A range under ±20 percent uncertainty is mandatory under the EU CRCF.
4. The credit-chain assignment showing who holds the issued credit, who has signed off that they will not also claim the removal, and the registry on which retirement happens.

A claim of “2.6 tonnes CO₂ removed per tonne of biochar” with no H/Corg, no LCA boundary, and no credit-chain attribution is rhetoric, not evidence. The same claim with H/Corg = 0.35 (premium), Fperm = 81 percent at 100 yr, net LCA of −2.3 tCO₂e per tonne biochar (counterfactual: aerobic decay), and the credit issued and retired under VM0044 against the buyer's target, is operational.

Key takeaways