Canaan Launches Bitcoin Mining Heat Project for Sustainable Greenhouse Operations in Canada

Hardware manufacturer Canaan has introduced a new proof-of-concept project that combines Bitcoin mining operations with agricultural applications. The initiative deploys liquid-cooled computing equipment to capture and reuse waste heat in a commercial greenhouse located in Manitoba, Canada. Announced on Tuesday, the pilot aims to demonstrate how high-density computing can serve as a reliable supplemental heat source in cold climates instead of releasing excess energy into the atmosphere.

The 3-megawatt project is being conducted in collaboration with Bitforest Investment, which will host the equipment at its greenhouse facility. Canaan plans to install 360 Avalon A1566HA-460T liquid-cooled servers within four containerized cooling modules under a 24-month agreement. Bitforest focuses on year-round tomato production and has committed to maintaining 95% uptime for the mining hardware.

Advancing Dual Purpose Bitcoin Mining Applications

Heat from the mining servers will be recovered through a closed-loop heat-exchange system and directed to preheat water for the greenhouse's electric boilers. Canaan estimates that approximately 90% of the electricity used by the servers can be converted into usable heat, though final figures will depend on real-world performance once the system is fully operational. This approach allows the company to test critical metrics such as heat recovery efficiency, system reliability, and ongoing maintenance requirements in an agricultural setting.

The design also offers potential cost advantages over traditional liquid-cooled data centers. By integrating directly with the greenhouse's existing boiler system, the setup eliminates the need for separate industrial cooling towers. Canaan reports an all-in power cost of about $0.035 per kilowatt-hour, covering electricity, operations, and maintenance, with possible additional savings if Bitforest engages in demand-response programs or grid power sales.

Large greenhouse operations often depend on fossil-fuel boilers for heating, particularly in regions with carbon pricing policies like Canada. The combined mining and heating system is projected to circulate up to one million tonnes of hot water annually by reusing energy that would otherwise be lost. This recycling process supports greater site efficiency and helps lower the overall environmental footprint of both computing and agriculture.

Canaan Chairman and CEO Nangeng Zhang described the program in a statement as an opportunity to measure and scale heat recovery solutions for colder climates. He noted that liquid cooling produces hot water above 75 degrees Celsius, making it directly suitable for greenhouse needs. Zhang emphasized that the project aligns with broader efforts to integrate computing infrastructure with sustainable energy practices across various sectors.

This initiative builds on Canaan's earlier work in dual-purpose mining technology. Last January, the company released the Avalon Mini 3, a compact Bitcoin mining device designed to function simultaneously as a home heater. Zhang highlighted at the time that such innovations create value from computing while reducing waste through multi-purpose applications.

Other manufacturers have pursued similar paths in the space. Heatbit began shipping its own dual-purpose mining heaters in 2022, and various third-party developers have modified units from competitors like Bitmain for heating uses. Meanwhile, mining operators continue to explore new directions, as seen in recent moves by firms like Bitfarms to divest certain assets and shift focus toward high-performance computing and artificial intelligence opportunities in North America.

The Manitoba pilot represents a practical step toward integrating Bitcoin mining with real-world energy needs. By capturing heat that is typically discarded, the project illustrates how computing power can contribute to operational efficiency in energy-intensive industries. Results from the 24-month trial will help determine the viability of scaling this model to additional agricultural or industrial sites.