In 2009 a drill crew in northeast Iceland went looking for very hot water and hit something that was still moving.
They were working at Krafla, a volcanic caldera about ten kilometres wide in one of the most active corners of Iceland. The plan was straightforward on paper: drill down more than four kilometres to reach water so hot and so compressed that it stops behaving like a liquid or a gas, a state scientists call supercritical.
They never got that far. At just over two kilometres down, the rock coming back up the drill pipe changed, turning to fresh volcanic glass. That meant one thing. The bit had punched straight into magma, molten rock at around 900 degrees Celsius. Not a pocket of hot water sitting nearby. Actual magma.
The 450-degree steam that would eventually make this well famous was still a year away, and only because of what happened next.
Drilling into magma is the one thing geothermal engineers are trained never to do.
What the crew decided to do with the hole
Nobody sealed the hole. Only one well anywhere had ever struck magma before, in Hawaii in 2007. This time Landsvirkjun, Iceland’s national power company, and its partners in the Iceland Deep Drilling Project decided to find out what they had.
Wilfred Elders, the University of California geologist who edited a special issue of the journal Geothermics on what happened next, described the workaround. A steel casing, perforated along the section closest to the magma, was cemented into place. Then came a slow reheat, months of it, before anyone touched the valve.
What came out set a record. IDDP-1 produced superheated steam above 450 degrees at around 140 bar of wellhead pressure, the hottest production well ever measured. Elders called it the first magma-enhanced geothermal system anywhere on Earth, and the first to draw heat straight from molten rock.
Why 450 degrees rewrites the arithmetic
Ordinary Icelandic wells produce fluid at around 250 degrees, arriving up top as a mix of water and steam that has to be separated before it gets near a turbine.
Above roughly 374 degrees that mixture stops existing. Water turns into superheated steam carrying far more thermal energy per kilogram, so a turbine extracts more work from every kilogram that reaches the surface.
Flow tests bore this out. Reviewing the data, the Clean Air Task Force noted that IDDP-1 could have generated up to 36 megawatts of electricity, five to ten times what a typical commercial well at lower temperatures manages. Björn Þór Guðmundsson, chief executive of the Krafla Magma Testbed, has put the same well at ten times the output of an average geothermal well across nearly two years of intermittent flow testing.
Fewer holes for the same power is the only version of geothermal economics that scales.
The well died of chemistry
Corrosion did most of the damage. Acid gas, sulphur and silica dust travelled up with the steam. The casing took a beating during the long stretch when the fluid was a wet mix rather than dry superheated steam.
Then in July 2012 several surface valves failed. The well had to be quenched with cold water. The innermost casing contracted and split under the thermal shock, and IDDP-1 has never flowed since.
That failure may be the most valuable thing the experiment produced. Casing depth, cement blends, alloy ductility, how long to let a well reheat before opening it: the list of things that broke has become a specification sheet for the next attempt.
Going back on purpose
That attempt belongs to a non-profit consortium funded by the International Continental Scientific Drilling Program, Iceland’s environment and energy ministry, Landsvirkjun, Reykjavik Energy and the Iceland Drilling Company. Their plan is to hit the same magma body deliberately.
Two wells are on the drawing board. As reported by GeoExpro, KMT-1 is designed to reach about 2,100 metres into magma expected to sit near 970 degrees, carrying sensors built to measure temperature and pressure inside molten rock for the first time. KMT-2 stops around 2,050 metres, just above the chamber at roughly 500 degrees, and exists to test whether an energy system can survive beside that heat. Guðmundsson told the American Association of Petroleum Geologists that drilling is scheduled to begin in 2027, with well integrity, thermal stress, corrosion and cementing at the top of his worry list.
The scientific payoff may outstrip the electricity. No geophysical method reliably finds magma chambers from the surface, so volcanologists read what is happening underground through ground deformation, gas emissions and seismic signals, always at one remove. Writing in Eos, the team behind the project described their ambition as a permanent magma observatory, closer in spirit to a telescope array than a power plant.
What one well can and cannot prove
One borehole is one borehole.
The ten-times number comes from a single well that was never engineered for the conditions it met, tested on and off, and shut in before anyone built a commercial plant around it. Feasibility work from the early 2000s predicted an order-of-magnitude gain from supercritical fluids, and IDDP-1 gestured in that direction without settling the question. IDDP-2 at Reykjanes reached supercritical conditions in 2017 and lost its casing as well.
Corrosion is the gate. Solve it once, in a hole beside a chamber whose depth is already known, and everything downstream becomes engineering that other volcanic countries can borrow. A lot of nations sit above exactly this kind of heat with no way yet to put it to work.
Krafla’s magma has spent four decades cooling quietly under a working power station, doing nothing for anyone. The next drill bit decides whether that changes.
