CSI GPC vs SPC Permanence Classes
How CSI Global Biochar C-Sink v3.3 splits soil-applied biochar carbon into geologically persistent (GPC) and semi-persistent (SPC) pools, and how Ro / HyPy unlock the upper class.
Primary: Global Biochar C-Sink 3.3 PDF Ch. 3 · Formulas PDF · science context Permanence Science for Credits.
Definitions (v3.3 glossary)
| Term | Meaning |
|---|---|
| GPC | Fraction expected to persist >1000 years and enter the geological carbon cycle |
| SPC | Fraction expected to decay within the first 1000 years after soil application (temporary C-sink) |
| CINK_H / C-Sink_H | Annual average mass of biochar carbon stored over horizon H years (t aCO₂e) — integral average of the remaining SPC (and display conventions for GPC) |
| TCE | Total Climate Effect — time-integrated radiative forcing (tCO₂e·yr); positive = warming, negative = cooling |
Source: v3.3 glossary.
CSI’s framing shift: not “how stable is biochar?” but “what size is the GPC fraction?” once applied to soil (v3.3 §3.2).
Three operative classes (v3.3)
| Class | Gate | % GPC | % SPC |
|---|---|---|---|
| SPC-only | H/C_org > 0.40 | 0 | 100 |
| Lower persistence | H/C_org ≤ 0.40, upper criteria not met | 75 | 25 |
| Upper persistence | H/C_org ≤ 0.40 and Ro or HyPy gate | 90 | 10 |
(v3.3 §1.1, §3.5–3.6; formulas §2.1.2)
Upper-class analytical gates (either suffices)
- Mean random reflectance Ro ≥ 3.8%, with:
- ≤ 3% of measured points with Ro < 2%
- ≥ 500 measured points
- Gaussian distribution
- heterogeneity score ≤ 8
- BC_HyPy ≥ 90% of total organic carbon (hydropyrolysis-resistant fraction)
Ro resolves microscale heterogeneity / mixed batches; HyPy isolates a condensed-carbon operational fraction. CSI accepts either to keep lab access practical (v3.3 §3.6).
Sampling rule: Ro/HyPy must be on the same representative EBC/WBC batch sample; separate campaigns are rejected. Retroactive Ro/HyPy requires the sample registered in the Global Biochar Tool plus elemental analysis + SEC per EBC Table 1 (v3.3 §3.7).
Worked example (standard Table 2)
100 t biochar DM, 80% C → 80 t C ≈ 293.3 t CO₂e total.
| Row | H/C_org | Ro | GPC (t CO₂e) | SPC (t CO₂e) | CINK_100 from SPC | CO₂-offsettable GPC |
|---|---|---|---|---|---|---|
| Upper | 0.15 | 3.90% | 264 | 29.3 | 15.4 | 264 |
| Lower | 0.30 | 2.80% | 220 | 73.3 | 38.5 | 220 |
| SPC-only | 0.60 | 1.80% | 0 | 293.3 | 153.9 | 0 |
(v3.3 Table 2). Online SPC calculator referenced on Carbon Standards site from the same table caption.
SPC decay function
From the formulas companion (4000115EN §2.1.2):
[ C_{\mathrm{remain}}(y) = M_{BC}\cdot C_{\mathrm{cont}}\cdot\Big(%GPC + %SPC\cdot\big(0.1787,e^{-0.5337 y} + 0.8237,e^{-0.00997 y}\big)\Big) ]
SPC is described as having a minimum half-life of ~50 years (MRT ~83 years) in the narrative intro (v3.3 §1.1).
Climate-service use of SPC: CINK_H (e.g. CINK_20, CINK_100, CINK_200) for methane compensation and temporary offsetting — transparently distinct from geological CO₂ offsetting, which CSI restricts to the GPC pool (v3.3 §3.4, §3.7). See also CSI MRV and Emission Portfolio for CH₄ TCE matching.
Proxies and science stance
| Proxy | Role in CSI | Caveat in-standard |
|---|---|---|
| H/C_org | Class gate (0.40); bulk aromatic condensation proxy with longest incubation/field empirical base | Average property; ash-rich chars can under/over-estimate; mineral-bound H; see Azzi / Lebrun Thauront citations in §3.3 |
| Ro | Upper-class gate (≥3.8%); microscale | Needs endorsed lab + ≥500 points + distributional QC |
| BC_HyPy | Alternate upper-class gate (≥90%) | Operational fraction; availability differs by lab |
| Soil temperature (Woolf-style) | Not used as a global adjustment in v3.3 ops | Footnote: lab vs field stabilizing mechanisms; data insufficient for global T effect (v3.3 Ch. 3 footnote) |
Literature cited in-standard includes models soil GHG from biochar (DOI), inertinite permanence benchmark (DOI), Schmidt et al. 2025 (DOI), Azzi et al. 2024 (DOI).
Forward look: v3.3 says classes will be refined in the 2026/27 update; higher GPC for a batch may be adjusted retrospectively under future rules (§3.7). Draft v4.0 existed in consultation (closed 30 Apr 2026) but v3.3 is the live standard (service page).
Materials vs soil
| Pathway | Persistence treatment (high level) | Source |
|---|---|---|
| Soil / geological entry | GPC/SPC split by class; SPC decays from year 0 | Ch. 3, 12.1 |
| Composites / plastics (matrix positive list) | Entire C persists while the material exists; removed from registry at destruction/disposal | §1.1; matrix list |
| Concrete / asphalt / mineral binders | 100% persistence during structure life; EoL landfill can start soil-like curves after a protected period (concrete narrative: ≥100y protected then curves — see Ch. 12.3–12.4) | Ch. 12.3–12.4; §3.7 |
Matrix positive list header (v4_17, 28 May 2026) restates the same GPC/SPC splits for H/C_org < 0.40 (4000078EN).
Cross-standard comparison (permanence only)
| Program | Horizon / claim | Quantifying gate closest to CSI upper class |
|---|---|---|
| CSI v3.3 | GPC >1000y / SPC ≤1000y | Ro ≥ 3.8% or HyPy ≥ 90% → 90% GPC |
| Isometric | 200y or 1000y | Ro ≥ 2% path → 1000y module (protocol) |
| Puro Ed. 2025 | Several centuries (CORC200+) | Ro encouraged, not quantifying (permanence post) |
| Verra VM0044 | 100 years | H/C_org → Table 3 temperature factors |
Chooser: Verra vs Isometric vs Puro · Puro vs Other Standards. Science camps: Permanence Science for Credits.
Practical lab note: CSI’s 3.8% Ro bar is materially stricter than Isometric’s 2% 1000-year threshold — budget petrography early if upper-class GPC is the commercial product.
Primary links
- C-Sink 3.3 PDF
- Formulas & EFs
- Matrix positive list
- Sanei et al. 2024 — https://doi.org/10.1016/j.coal.2023.104409
- Schmidt et al. 2025 — https://doi.org/10.1111/gcbb.70092
- Woolf et al. 2021 — https://doi.org/10.1021/acs.est.1c02425