China commissions a closed-loop CO2 geothermal heating project

At a site in Zhengzhou, carbon dioxide goes into the ground at 8 degrees Celsius and comes back at 22 degrees, drawing heat from rock 2,500 meters below the surface. China Huaneng Group commissioned the system in May 2026 as China's first supercritical CO2 closed-loop geothermal demonstration, and it could heat more than 18,000 square meters—roughly the equivalent of 180 apartments.

The trick is to stop looking for underground hot water. Traditional hydrothermal geothermal depends on naturally occurring reservoirs, which limits development to volcanic and tectonically active regions. In a closed-loop system, a sealed pipe circulates fluid through hot rock without touching groundwater. The site needs sufficient underground temperature, not a naturally heated aquifer.

CO2 helps the system move heat with less mechanical effort. Above 31.1 degrees Celsius and about 73 atmospheres, it becomes a supercritical fluid: neither a conventional liquid nor a gas, but a single phase with lower viscosity. Heating also makes it less dense, so cold CO2 sinks and warm CO2 rises, creating a natural thermosiphon. In Zhengzhou, the approach cut pumping loss by 10%, increased heat-extraction capacity by about 20%, and reduced per-unit heating energy consumption by 10%.

So what changes, concretely? Geothermal heating could reach places far from the high-temperature resources concentrated in southern Tibet, western Yunnan, and western Sichuan. The Zhengzhou project is still a demonstration, but it shows a path toward using ordinary hot rock as a steady heat source for district heating—and eventually for power generation that does not depend on sunlight, wind, or cloud cover.

The constraint has moved from geology to economics. Closed-loop electricity currently costs roughly five to seven times more than conventional generation, making drilling the central obstacle. China is testing the other lever: a 3,515.5-meter U-shaped well in Henan has achieved heat-extraction capacity five times that of the previous design. The technology has reached the field, but its next chapter depends on getting more heat from each well at a lower drilling cost.