PART VII: BOREHOLE GEOPHYSICS AND LOGGING
A 2-D oil-field survey and a 3-D environmental study illustrate the problems of collecting and interpretating electromagnetic data in crosshole and surface-to-borehole geometries. The first example employs both methods to map the development of a steam flood in a shallow heavy oil reservoir at the Lost Hills oil field in central California. Measurements were made from observation wells that straddle the steam injector in a plane orthogonal to the regional strike, and from a surface-to-borehole line that extends between the observation boreholes and into an undeveloped portion of the field. In a second example, a crosswell experiment was designed to image 250 000 liters of saltwater injected into a shallow aquifer near the University of California at Berkeley campus. A five-spot well pattern was drilled for the test, and saltwater was injected in the central well at a depth of 30 m. Crosswell data were collected, using all five boreholes, before and after injection.
Bias and uncertainty of calibration are at least as important as random noise in these data. The bias can be caused by ground loops, cultural noise, systematic drift, and errors in sensor placement. Repeatability and reciprocity tests can often quantify this measurement error.
Imaging schemes also introduce errors through geometrical assumptions. The Lost Hills data were imaged in a 2.5-D model (point sources in 2-D geometry), which appears to represent the background geology well, but begins to break down within the steam injection zone. Although a 3-D code was used to image the data collected at Richmond, artifacts were still present in the image. Subsequent analysis has shown that these are likely due to sensor placement errors as a result of borehole deviations.