Google and Terradot in Brazil:
How Rice Fields and Basalt Are Creating a
Carbon Removal Value Chain

What the Google–Terradot Deal Changes

Terradot was already operating in Brazil before the new agreement. In December 2024, Google announced a separate purchase of 200,000 tons of carbon-removal credits from Terradot and an equity investment in the company, while Frontier buyers committed US$27 million for roughly 90,000 tons of removals to be delivered between 2025 and 2029.

The September 2026 agreement changes the scale and the operating model. Instead of treating enhanced rock weathering as a stand-alone carbon-removal pathway, the new structure places it on the same rice fields as an irrigation intervention intended to cut methane emissions earlier.

Terradot reports that its wider ERW platform has already deployed more than 500,000 tonnes of rock and treated 20,000 hectares. That is evidence of prior operating activity, but it should not be read as evidence that the newly announced 200,000-hectare rice project has already been deployed.

Why Rio Grande do Sul Is More Than a Convenient Location

The project starts in one of Brazil's most concentrated rice-producing regions. Rio Grande do Sul's state rice institute, IRGA, reported 970,216 hectares of irrigated rice sown in the 2024/25 season and 8.76 million tonnes of production. That scale matters because a carbon project tied to farm operations needs a large, repeated land base rather than a one-off industrial site.

The second advantage is geological. Google's project announcement says basalt quarries are located close to the farmland. Brazil's Geological Service separately maps soil-remineralization potential in the Serra Geral geological group in Rio Grande do Sul, while Brazilian law has recognized mineral-based soil remineralizers as an agricultural-input category since 2013.

Those two systems — irrigated rice and accessible volcanic rock — create the physical basis for the commercial model. The closer the quarry is to the farm, the less material must be moved over long distances. That affects both operating cost and the net carbon value of the removal because mining, grinding, transport and spreading all consume energy.

A New Value Chain Is Forming Between Quarry and Corporate Buyer

A rice field normally sits inside an agricultural value chain: seed, inputs, water, machinery, harvest, storage, milling and sale. The Terradot model adds a second commercial layer to the same land.

Basalt must be sourced, ground and transported. Farmers must change water management and allow rock application. Samples and field data must be collected. Removal and methane reduction must be quantified under credible methodologies. Credits or other contracted climate outcomes then need verification, registration and a buyer willing to pay for them.

One Field, Two Different Climate Products

The Google agreement combines two mechanisms that should not be treated as the same product.

Alternate Wetting and Drying periodically drains rice paddies rather than keeping them continuously flooded. Reduced anaerobic conditions can lower methane generation. Google measures the contracted methane component using a 20-year global-warming-potential basis, or GWP20, and says the result should provide earlier temperature impact.

Enhanced Rock Weathering works on a different timeline. Finely ground silicate rock is applied to soil. As minerals dissolve, they react with carbon dioxide and generate bicarbonate that can enter long-lived geochemical storage. The removal develops over time rather than at the moment the rock is spread.

Research from Embrapa-linked authors is directly relevant to the rice-management side. Field measurements in Capão do Leão, Rio Grande do Sul, compared continuous flooding with Alternate Wetting and Drying over four crop seasons from 2019 to 2023. The study confirms that water management materially affects methane dynamics in the same rice-growing environment where the Terradot project begins.

The Economics Are Still Being Built

The new Google contract does not disclose a price. The clearest public price reference comes from Terradot's earlier Frontier transaction: US$27 million for approximately 90,000 contracted tons of carbon removal between 2025 and 2029.

That implies an average contracted value of roughly US$300 per ton of CO₂ removal. This is an Econosur calculation based on Frontier's disclosed contract value and contracted volume, not a quoted market price. Frontier also states that the contract price includes measurement, reporting and verification.

Frontier says Terradot projects a 35% reduction in removal cost by 2030. The proposed levers are concrete: shorter quarry-to-farm transport, lower sampling and laboratory requirements through new MRV methods, and software that optimizes spreading and sampling.

The September 2026 project introduces another cost idea. Reuters reported that Terradot plans to delay some third-party verification rather than repeatedly verifying very small amounts of enhanced-weathering removal early in the process. Methane reduction is intended to provide earlier contracted impact while the rock continues to weather.

Verification Is an Industrial Layer of Its Own

Carbon removal cannot be inspected like a shipment of rice or a machine leaving a factory. The buyer is purchasing a quantified environmental outcome. That makes measurement, reporting and verification part of the production system.

Terradot already has one useful proof point in Brazil. Isometric's registry shows that Project Carcará, a Terradot enhanced-weathering project using finely ground basalt from local quarries on agricultural land, is validated under the Enhanced Weathering in Agriculture protocol. Isometric's issuance record shows that 204.95 certificates were issued on March 17, 2026 for a reporting period from October 16, 2024 to August 18, 2025.

Project Carcará is not the new Google rice project. Its relevance is narrower: it shows that Terradot has already moved from rock deployment to third-party verified issuance in Brazil. The scale contrast is substantial: 204.95 certificates have been issued from Carcará, while the new Google agreement targets 1 million metric tons of permanent carbon removal by 2040. The comparison shows the size of the scaling challenge; it does not imply that the projects are operationally identical.

Terradot's February 2026 acquisition of Eion adds another signal. The company says the combined platform now integrates operating data, intellectual property and MRV capabilities across Brazil and the United States. That suggests the market is beginning to consolidate around measurement capability as well as physical deployment capacity.

Brazil Is Building a Second Market Layer Around the Project

The Google–Terradot agreement sits in the voluntary carbon market. At the same time, Brazil is implementing its regulated Sistema Brasileiro de Comércio de Emissões, or SBCE, under Law 15,042/2024.

The law creates Brazilian Emission Quotas and Certificates of Verified Emission Reduction or Removal, known as CRVEs. Carbon credits can be recognized as CRVEs only when they follow an accredited methodology, are independently verified and are registered in the SBCE central registry. The national allocation plan will also set limits on how CRVEs can be used for regulated compliance.

The system is not yet fully operational. In July 2026, Brazil's Ministry of Finance opened consultation on a phased MRV schedule, with the first group of covered sectors proposed to begin monitoring, reporting and verification obligations in 2027. Primary agricultural production is not listed in any of the three proposed MRV-entry stages, which cover sectors including heavy industry, oil and gas, electricity, food and beverage manufacturing, waste and transport. That omission does not by itself determine the final regulatory treatment of agricultural carbon projects, but it separates rice farms from the sectors currently proposed for direct MRV entry.

That creates a potentially important future interface between voluntary projects and regulated carbon assets, but there is no public evidence that the Google–Terradot project will qualify as an SBCE CRVE. The methodologies, registration rules and treatment of specific project types still matter.

This broader capital and policy architecture also connects with Econosur's analysis of Brazil's Eco Invest model, where the central question is how new financial structures translate into investable projects rather than remaining policy announcements.

Four Business Questions That Matter Next