Market sizing · pathway economics · where to play
Sizing India's agricultural carbon-removal opportunity on a base of 44.16 Mha of kharif rice. The useful finding is not the market size — it is that the two leading pathways are different businesses, and a cross-check against published farmer earnings is what shows it.
| Pathway | Gross $/ha/yr | Farmer at 50/50 | Implied developer share |
|---|---|---|---|
| Rice methane | $34.1 | $17.1 | -105% to 41% |
| Biochar | $450.0 | $225.0 | 84% to 96% |
Published smallholder pilots earn US$20–70 per hectare per year. A flat 50/50 split puts rice-methane farmers below that band and biochar farmers at roughly three times its top — so the same split cannot describe both.
The negative share is the finding. To pay a farmer $70/ha from rice methane, the developer would have to pay out more than the credit is worth: the whole credit generates $34/ha. Biochar's 84–96% is what a pyrolysis business looks like — the processor carries the equipment cost, so the credit cannot be split like a land-based one. Biochar is a processing business that sources from farms; rice methane is a farmer-aggregation business.
| Scenario | Rice methane | Biochar |
|---|---|---|
| bear — 1% of rice area | $7.5m | $178.8m |
| base — 3% of rice area | $22.6m | $536.5m |
| bull — 8% of rice area | $60.2m | $1,430.8m |
Developer-side revenue per year. The only Indian programme publishing a hectare figure covers 10,117 ha — 0.023% of the rice base. Even the bear case is roughly a 40× expansion on disclosed enrolment, so read these as if-then, not as a forecast.
Swing in rice-methane developer revenue, one lever at a time. Enrolment dominates — and it has the least evidence behind it. Price and yield can be bounded from published data; enrolment cannot, because nobody has demonstrated it at scale in India. That is where diligence belongs.
| Assumption | In this model | What operators disclose | Effect on developer revenue |
|---|---|---|---|
| Farmer share of credit revenue | 50% | 55% – 75% | Overstated by 10–50% |
| Credit yield, rice methane | 1.1 t/ha | 3.71 – 4.20 t/ha | Understated ~3.4× |
Of the six programmes screened, 3 disclose a farmer revenue share — Boomitra 55% of gross, Varaha 70% of net, Grow Indigo 75% of net — and all 3 sit above the 50% modelled here. Two are promised on net revenue with no published cost base, so they cannot be converted to a gross split at all. The yield range comes from a published 1.5–1.7 credits-per-acre estimate under VM0042, which is directional and site-dependent, so it bounds the assumption rather than replacing it. The two conflicts push in opposite directions, so neither is a one-line fix: the model keeps its sourced assumptions and records the disagreement.
Not one of the six publishes both hectares and credits issued, which is why credits per hectare — the number every sizing of this market rests on — still cannot be derived from any disclosed Indian programme. Full screen in the diligence memo; tracker and checks in diligence.py and test_diligence.py.
| Programme | Farmers | Credits | Hectares | Practice |
|---|---|---|---|---|
| Sow&Reap Agro (Telangana) | ~35,000 | 37,405 | — | Paddy, Gold Standard |
| Boomitra India | 12,000+ | — | 10,117 | Soil carbon |
| UP state programme | 25,000+ | — | — | Tree planting |
| Grow Indigo | — | — | — | Soil, biochar, rice, agroforestry |
| Varaha | — | — | — | Biochar |
Published figures only; dashes are undisclosed, not zero. Sow&Reap's 1.07 credits per farmer is the only bottom-up check available on the credit-yield assumption anywhere in this market — and it is credits per farmer, not per hectare. That one missing conversion is the weakest link in every sizing of this market, including this one.
Rice area from Ministry of Agriculture kharif sowing data via BigMint; emissions from Our World in Data; credit prices from the Senken Q4 2026 benchmark; programme figures from the Institute for Competitiveness. Every input is labelled fact or assumption in SOURCES.md. · Model, memo and tests on GitHub