Explainer

How Do You Find Underground Oil?

First, get a vibrating truck.

Hey. Did the Earth move for you, too?

The Senate voted yesterday to allow oil drilling in Alaska’s Arctic National Wildlife Refuge. According to estimates made by the U.S. Geological Survey, around 10.4 billion barrels of oil could be recovered from the area. How can they tell?

By using geological information about the area and what they’ve learned from drilling elsewhere. No one knows for sure how much oil there is in a given area until people drill there, but geologists can make an educated guess. First, they examine seismic and aeromagnetic data to determine the likely features of underground rock formations, and compare these estimates with what’s been found in nearby wells.

For oil to accumulate underground, there must be what geologists call a “source,” a “trap,” and a “reservoir.” Source material is rock with organic content that can be converted into oil or natural gas (“cooked”) by heat and pressure. Generally, the oil will just disperse, seeping up toward the surface over millions of years. But if a hard, dense material forms in the right shape underground (say, an inverted bowl), it can trap the oil in place. Even so, there won’t be very much in the trap unless the substance under the bowl is porous enough to store the oil. These reservoirs can consist of rock that contains cave-like bubbles, or of a porous material such as sandstone, which can soak up oil like a sponge.

Geologists can measure the density of underground rocks by sending radio waves into the ground and measuring the speed with which they reflect back to the surface. This can be done quickly and easily from a plane (which obtains what’s called “aeromagnetic” data), or it can be done from the surface. Environmental concerns in the wildlife refuge made aeromagnetic surveys an obvious choice.

But far better data comes from seismic waves. Surveyors can send vibrations into the earth using a vibrating (“vibroseis“) truck. An array of sensitive geophones picks up readings which computers then use to produce a map of the underground terrain. In modern three-dimensional seismic imaging, the data are converted into a textured image that can be rotated and examined at multiple levels.

Interpreting seismic maps is something of an art. You could look at the 3-D data on a regular computer screen, but some geologists use 3-D projectors and glasses to inspect the graphs, and others wear virtual-reality helmets. The only seismic data available for the wildlife refuge was collected in the 1980s, however, and can only be examined in two dimensions, which makes interpretation more difficult.

A team of around 10 geologists examined the data from the refuge, looking at seismic and aeromagnetic data, traces of oil that had seeped up from the ground, and information from nearby wells. One well, called KIC#1, has been drilled in the refuge already, by private oil companies in partnership with the local Inupiat Eskimos. But information from that well is being kept secret (even from the USGS), and those in the industry refer to KIC as “the tightest hole in North America.”

Then the geologists split up into pairs, and each pair looked at one of the sites of potential oil reserves. Based on the data, they try to describe around 10 features, including the porousness of the rock, the size of the reservoir, and the degree to which the reservoir is saturated with water instead of oil. Then they present their guesses (in the form of probability distributions) to a roomful of their colleagues—about 18 or 20 geologists—and defend their reasoning.

When everyone has presented and defended their estimates, the distributions for each variable are plugged into a computer, which then creates a plausible description of what’s under the ground and how much oil you might find there. They run the simulation 9,999 more times and look at the results. Ninety-five percent of the time, the computer said there was between about 6 billion and 16 billion barrels of oil in the refuge. The average value from all the simulations together was the 10.4 billion barrels reported in the press.

Next question?

Explainer thanks Dave Houseknecht of the U.S. Geological Survey, Philip Budzik of the Energy Information Administration, and Larry Nation of the American Association of Petroleum Geologists.