WHEN Lawrence Livermore researchers developed a means to detect tunnels in Korea in the early 1970s, they had no idea that American oil and gas companies in the late 1990s might turn to a derivative of that technology to gain a more informative underground picture of their oil fields.
Yet crosshole electromagnetic (EM) induction technology, developed jointly by Lawrence Livermore and Lawrence Berkeley researchers with scientists at Schlumberger (an international provider of oil production services), may hold the key to significant increases in the extraction of oil and gas from existing reservoirs. What's more, the technology also may prove to be a cost-effective means for environmental site characterization and remediation monitoring.
"It is becoming increasingly difficult and expensive to locate new petroleum deposits, particularly in the U.S., so we must devise ways to get more oil out of existing fields," says Michael Wilt, a geophysicist and leader of the Laboratory's crosshole EM induction technology development effort. "Improved reservoir characterization is the key to extracting more oil and gas from oil fields. Producers are missing large pockets of oil because existing exploration methods, such as surface seismic reflection, sample too coarsely, and borehole logging techniques only measure rocks and fluids in the immediate vicinity of wells."
Wilt explains that crosshole EM induction technology is designed to provide high-resolution images of oil, gas, and water-bearing rock layers between existing wells up to 1,000 meters apart. A series of such images over time can also provide insight to changes in the field caused by oil production or by steam or water injection, which is used for enhanced oil recovery.

Determining the Electrical Resistivity
The electrical resistivity (resistance to the flow of electrical current) of rock formations can be determined from EM induction measurements. Rock formations containing salt water, clay, or metallic minerals conduct current readily and therefore have low resistivity, while rocks containing fluids such as oil or gas have high resistivity. These resistivity differences can be used to distinguish between oil- and water-saturated sands, for example.
Measuring the electrical resistivity near a borehole has long been used to determine production zones in oil and gas fields and to map sand and shale layers. These measurements are usually made with an EM induction logging device that determines the resistivity within a meter or so around the borehole. Crosshole EM induction offers the ability for the first time to map subsurface resistivity between wells, on a reservoir scale (some 100 times greater an area than before), thereby locating bypassed concentrations of oil and gas.
The technology is an outgrowth of radar experiments conducted by Laboratory researchers in the early 1970s. The idea was to transmit high-frequency radio waves (greater than 20 megahertz) through the ground to detect tunnels in Korea. Although the technique proved effective for hard rock, the high-frequency signals could not propagate for more than a few meters in soft rock environments such as oil fields.

However, it was discovered that at frequencies in the kilohertz range, it is possible to propagate signals up to a kilometer through a typical oil field. The penalty is that at low frequencies, the signals are dispersive--i.e., they get smoother as they propagate--making them impossible to image with traditional techniques. Recently, new tools were developed at the University of California at Berkeley and Schlumberger that can determine the resistivity between boreholes from these crosshole EM measurements.
The crosshole system consists of a transmitter tool deployed in one well and a receiver tool deployed in a second well (see above figure). The transmitter uses a vertical-axis coil wrapped with 100 to 300 turns of wire tuned to broadcast a single low-frequency sinusoidal signal that induces currents to flow in surrounding rocks.
The optimum operating frequency depends on borehole separation and background resistivity, but generally the frequency ranges between 40 hertz and 100 kilohertz. A frequency that is too low limits the resolution, while one too high limits the range of the measurement.
At the receiver borehole, a custom-designed coil detects the total magnetic field, consisting of the magnetic field from the induced currents as well as the primary magnetic field generated by the transmitter. The receiver section consists of a magnetic field sensor and a commercial lock-in amplifier located at the surface. The lock-in amplifier operates like a radio by measuring only those signals that are coherent with the transmitted signal while rejecting incoherent background noise.
By positioning both the transmitter and receiver tools at various levels above, below, and within the zone of interest, researchers can create an image of the resistivity distribution between the wells. The EM data are interpreted by computer modeling in which the rock between the wells is divided into thousands of small, two-dimensional, square blocks 1 to 5 meters on a side. Each block is assigned an electrical resistivity value, estimated from the borehole resistivity log (if available). The computer then modifies the resistivity of these blocks until the calculated and measured data agree to within the measurement error. This process usually requires 10 to 12 hours per data set on a 50-megahertz workstation to produce a detailed image of the underground strata.

Application in Oil Fields
After two years of development, a series of tests was conducted at Mobil Oil leases at Lost Hills in Central California in 1993 and 1994. The experiments were made to demonstrate the technology for characterizing oil reservoirs and monitoring steam floods. Two fiberglass-cased boreholes were drilled about 55 meters apart near a steam injector in shallow, heavy oil sands. Steam was injected at depths of 65, 90, and 120 meters, corresponding to upper, middle, and lower layers of the target Tulare formation.
The resulting crosshole EM induction images in the figure below, collected before steaming and 6 months and 11 months after (a, b, and c), clearly show the distribution of the high resistivity oil sands (blue and green) and the intervening shale layers (red). (The arrows indicate points of steam injection.) The image in (a) indicates that the upper oil sand is a thick unit dipping gently eastward. The middle and lower sands are thinner and more discontinuous between the wells. The images shown in (b) and (c) are visibly different only at depths below 70 meters, where the resistivity has decreased significantly due to the steam injection. In all other parts of the image, the before and after resistivity values are unchanged. The resistivity decline is caused by the temperature increase and the replacement of oil by water and steam.
Images (d) and (e) are "difference" images made by subtracting the baseline image (a) from the other two images (b) and (c), respectively to show the percentage of change in resistivity. These images highlight the parts of the section that have changed during the steam flooding. Images (d) and (e) show that the resistivity has decreased dramatically in the middle and lower oil sands, indicating the presence of substantial steam there. The images also indicate that almost no steam has gone into the upper oil sand. The steam also seems to preferentially flow to the west in the middle sand but to the east in the lower unit.
The steam is clearly not as successful in moving the very thick upper layer of oil. The technology showed the steam flood to be much less uniform than the operator anticipated, providing valuable information on the progress of the flood and the parts of the reservoir affected by the steaming.

Future Developments
For the past three years, the focus of R&D work centered on a recently concluded Cooperative Research and Development Agreement with Schlumberger. During the developmental work, Schlumberger developed an advanced computer code for imaging the data, and the Laboratory refined the hardware. The end result was a complete prototype system designed for application in deep oil field environments. This system, which is a significant upgrade over existing equipment, will be tested in the fall of 1996.
In addition to the cooperative work with Schlumberger, EM induction research is also proceeding at other national laboratories (Lawrence Berkeley and Sandia) and at other companies (Western Atlas and Oyo Corporation) to improve the quality of field data and to sharpen the image resolution. Among the most pressing problems is the application of the technology through steel well-casing. Although the steel casing dramatically attenuates EM signals, recent field measurements have shown that the subsurface resistivity may still be obtained under the right conditions. Steel-cased wells make up the vast majority of oil field drillholes, so this improvement in the technology could have dramatic effects on its widespread application.
EM induction technology has also been applied to environmental site characterization with good success. It is presently included in the Laboratory's arsenal of geophysical tools for site cleanup and monitoring. Research is also in the early stages for applying the technology to the exploration for mineral deposits.

Key Words: crosshole electromagnetic (EM) induction, electrical resistivity, oil-field imaging.

For further information contact Michael Wilt (510) 422-3152 (mjwilt@llnl.gov).
Back to August 1996