Stanford University
Lava channels

Geothermal at the foot of Kilauea

Geothermal engineer Roland Horne discusses geothermal energy in the face of natural hazards and a way to tap the earth’s heat far from volcanoes in the future.

BY Josie Garthwaite

Lava from Hawaii’s Kilauea volcano crept this week within a few hundred yards of the geothermal plant that supplies more than a quarter of the Big Island’s electricity. While the plant has been shut down since soon after eruptions began in early May, workers plugged the last of nearly a dozen wells only in recent days as authorities warned that a lava breach could cause heightened levels of toxic hydrogen sulfide gas in the air.

By Wednesday, May 23, lava spattered from a vent close to the plant had begun to cool and harden into a natural barrier against further encroachment by molten rock. But with wells reaching nearly a mile into the earth, the plant’s exposure at the surface is only part of the story. Stanford University energy resources engineer Roland Horne, director of the Stanford Geothermal Program in the School of Earth, Energy & Environmental Sciences (Stanford Earth) described why underground volcanic activity is just as important as lava flows at the surface, how geothermal plants fare in the face of natural hazards and how geothermal energy could be tapped far from volcanoes in the future.

Local authorities warned that if lava were to reach a well at the Puna Geothermal Venture plant, hydrogen sulfide could be released. Where would that come from?

The source of hydrogen sulfide gas is the volcano itself. In normal operation, steam and hot water circulate through the plant and are reinjected without being released to the atmosphere. However, a breach of an unprotected well could allow the release of the volcano’s gases from the wells. 

What other problems could occur with a geothermal plant in the midst of an active volcano?

Wells can be damaged even if the surface lava doesn’t reach the plant, if an eruption shears off the wells underground. If the wells are damaged, they could blow out. They would release steam that could contain hydrogen sulfide and CO2, and whatever else is underground. That happened at a place called Krafla in Iceland in the 1970s. In fact, they had some wells that produced lava just like fissures. 

If the well is damaged and it releases steam, it’s not particularly hazardous if you’re not standing next to it. They’d have to cap the well, otherwise it would blow down the geothermal reservoir, and that’s how they’re producing electricity. 

How common is it for geothermal power plants to store tens of thousands of gallons of the chemical pentane onsite, as they did at Puna? 

Every binary power plant has it - probably 100 of them worldwide. Binary plants like Puna produce water from the subsurface and run it through heat exchangers. Instead of using steam, they boil pentane and run it through the turbine. Water goes back in the ground. The problem is that pentane is flammable. But they have drained the pentane at Puna and taken it offsite. 

Large quotation punctuation icon

If the wells are damaged, they could blow out. That happened at a place called Krafla in Iceland in the 1970s. In fact, they had some wells that produced lava just like fissures. 

What are the advantages of a binary plant design?

You can run a binary geothermal plant with a pentane cycle at a lower temperature than steam. They also release no hydrogen sulfide or other gases from the subsurface. If the temperature is higher, a steam plant is more efficient and less expensive. But you can’t make enough steam to make a steam turbine work well if the temperature is too low. 

Geothermal resources are often found in volcano country. Is there anything that can be done with plant design to minimize vulnerabilities in the event of an eruption?

Nope. But the risk is pretty small. There are hundreds of geothermal plants worldwide and they’re almost all in volcanic regions. To have a volcanic eruption that also happens to be close to a geothermal power plant is not that common. In the case of Krafla, the eruption sheared off some wells. They capped them, they went on with life. 

Kilauea fissure 22

Lava erupts from fissure 22 on the morning of May 24. Photo credit: USGS/M. Patrick

Would you have any reservations about siting a geothermal plant around a highly active volcano?

Not really. Geothermal plants work best where it’s hot, so a volcano is kind of a good place to put them. And many natural disasters affect power plants. Here in California, fires in 2015 burned cooling towers at the Geysers complex, the world’s largest geothermal facility. In the Philippines, the typhoon that wiped out Tacloban city in 2013 took out three geothermal plants. Blew the cooling towers away. 

In Japan, the 9.0 earthquake and tsunami that caused the Fukushima accident in 2011 knocked nuclear power plants off nationwide, and those that weren’t knocked off were shut down. Many of the geothermal plants are right there in the Tohoku region, where the disaster centered, and they kept running. 

Enhanced geothermal systems, which are used in commercial plants in Europe but not yet in the U.S., don’t have to be close to volcanoes because they’re made artificially. If you have a place where the rock is hot, but there’s not water in it to bring steam to the surface to run the plant, you can create fractures in the rock and circulate water through it to bring the heat to the surface. 

You do your best to make the fractures you want, but it’s not like you’ve stamped them out with a cookie cutter. Our particular research program here at Stanford is associated with how you determine the structure of the fractures you want to create, make them, and then determine what you have made. 

Roland Horne is the Thomas Davies Barrow Professor in the School of Earth, Energy & Environmental Sciences and a Senior Fellow at the Precourt Institute for Energy.

Media Contacts

Josie Garthwaite

School of Earth, Energy & Environmental Sciences

josieg@stanford.edu, (650) 497-0947

Roland Horne

School of Earth, Energy & Environmental Sciences

roland.horne@stanford.edu, (650) 723-9595

maillinkedindouble carrot leftarrow leftdouble carrotplayerinstagramclosecarrotquotefacebooktwitterplusminussearchmenuarrowcloudclock