Drilling into Faults

SAFOD rig, August 2004
Drilling rig at SAFOD site is 60 meters tall, August 2004. Photo (c) Andrew Alden, licensed to About.com (fair use policy)

Geologists are daring to go where they once could only dream of going—right to the places where earthquakes actually happen. This article describes three projects that have taken us into the seismogenic zone. As one report put it, projects like these put us "at the precipice of quantum advances in the science of earthquake hazards."

Drilling the San Andreas Fault at Depth

The first of these drilling projects made a borehole next to the San Andreas fault near Parkfield, California, at a depth of about 3 kilometers.

The project is called the San Andreas Fault Observatory at Depth or SAFOD, and it's part of the much larger research effort EarthScope.

Drilling began in 2004 with a vertical hole going down 1500 meters, then curving toward the fault zone. The 2005 work season extended this slanting hole all the way across the fault, and was followed by two years of monitoring. In 2007 drillers made four separate side holes, all on the near side of the fault, that are equipped with all kinds of sensors. The chemistry of fluids, microearthquakes, temperatures and more are being recorded for the next 20 years.

While drilling these side holes, core samples of intact rock were taken that cross the active fault zone giving tantalizing evidence of the processes there. Scientists kept up a website with daily bulletins, and if you read it you'll see some of the difficulties of this kind of work.

SAFOD was carefully placed at an underground location where regular sets of small earthquakes have been happening.

Just like the last 20 years of earthquake research at Parkfield, SAFOD is aimed at a part of the San Andreas fault zone where the geology seems to be simpler and the fault's behavior more manageable than elsewhere. Indeed, the whole fault is considered to be easier to study than most because it has a simple strike-slip structure with a shallow bottom, at about 20 km depth.

As faults go, it is a rather straight and narrow ribbon of activity with well-mapped rocks on either side.

Even so, detailed maps of the surface show a tangle of related faults. The mapped rocks include tectonic splinters that have been swapped back and forth across the fault during its hundreds of kilometers of offset. The patterns of earthquakes at Parkfield have not been as regular or simple as geologists had hoped, either; nevertheless SAFOD is our best look so far at the cradle of earthquakes.

See some pictures of the project in my Parkfield photo-tour.

The Nankai Trough Subduction Zone

In a global sense the San Andreas fault, even as long and active as it is, is not the most significant type of seismic zone. Subduction zones take that prize for three reasons:


  • They are responsible for all the largest, magnitude 8 and 9 earthquakes we have recorded, such as the Sumatra quake of December 2004 and the Japan earthquake of March 2011.
  • Because they are always under the ocean, subduction-zone earthquakes tend to trigger tsunamis.
  • Subduction zones are where lithospheric plates move toward and underneath other plates, on their way into the mantle where they give rise to most of the world's volcanoes.

    So there are compelling reasons to learn more about these faults (plus many more scientific reasons), and drilling into one is just within the state of the art. The Integrated Ocean Drilling Project is doing that with a new state-of-the-art drillship off the coast of Japan.

    The Seismogenic Zone Experiment, or SEIZE, is a three-phase program that will measure the inputs and outputs of the subduction zone where the Philippine plate meets Japan in the Nankai Trough. This is a shallower trench than most subduction zones, making it easier for drilling. The Japanese have a long and accurate history of earthquakes on this subduction zone, and the site is only a day's ship travel away from land.

    Even so, in the difficult conditions foreseen the drilling will require a riser—an outer pipe from the ship to the sea floor—to prevent blowouts and so that the effort can proceed using drilling mud instead of seawater, as previous drilling has used.

    The Japanese have built a brand-new drillship, Chikyu (Earth) that can do the job, reaching 6 kilometers below the sea floor.

    One question the project will seek to answer is what physical changes accompany the earthquake cycle on subduction faults. Another is what happens in the shallow region where soft sediment fades into brittle rock, the boundary between soft deformation and seismic disruption. There are places on land where this part of subduction zones is exposed to geologists, so results from the Nankai Trough will be very interesting. Drilling began in 2007. 

    Drilling New Zealand's Alpine Fault

    The Alpine fault, on New Zealand's South Island, is a large oblique-thrust fault that causes magnitude 7.9 earthquakes every few centuries. One interesting feature of the fault is that vigorous uplift and erosion have beautifully exposed a thick cross-section of the crust that provides fresh samples of the deep fault surface. The Deep Fault Drilling Project, a collaboration of New Zealand and European institutions, is punching cores across the Alpine fault by drilling straight down. The first part of the project succeeded in penetrating and coring the fault twice just 150 meters below the ground in January 2011, then instrumenting the holes. A deeper hole is planned near the Whataroa River in 2014 that will go down 1500 meters. A public wiki serves past and ongoing data from the project.