Advances in Subsurface Modeling Using Seismic Refraction Data

Standard geophysical survey practice using seismic refraction techniques has predominantly produced two-dimensional cross sections of the subsurface. The state-of-the-practice for nearly three decades has been to process refraction data with layer reconstruction techniques using the generalized reciprocal method, time-intercept and other similar techniques. Within the past decade, advancements in the computer technology and the development of tomographic modeling algorithms have greatly increased the ability to detect subsurface anomalous features, increase lateral and vertical resolution, and provide better graphical presentation of the data. Recently, 2D finite-element modeling of seismic data has proven successful to image discrete anomalies such as voids.

This paper presents recent developments in a new approach for processing refraction data, the presentation of subsurface data, and the use of these data after geophysical modeling is complete. The approach adapts numerical modeling using the discrete element method and particle flow code (DEMPFC). The procedure is termed the Geostructural Analysis Package (GAP) which, in its initial stages of development has been optimized for geotechnical applications, such as 2D and 3D seismic refraction data processing and presentation on engineering projects. Although GAP has not been primarily created for seismic refraction, this paper will illustrate significant advancement in refraction data processing. Currently, using GAP for seismic applications represents an innovative approach that includes improved data analysis processes and produces more functional result for the end users. For the application illustrated in this paper the end users are typically civil or geotechnical engineers. The value of using this approach for seismic applications is its ability to produce 2D, 2.5D and 3D models to assist engineers or geologists extract additional information from the geophysical data (e.g., material properties), or perform static and dynamic stress analysis. This paper makes the point that mapping the top of bedrock may be the objective of a geophysical survey, but it is not the engineering purpose for the site investigation (e.g., construction of a critical facility, design of a foundation for a structure, etc.). With high-quality calibrated 2D, 2.5D and 3D DEM-PFC models, not geophysical images, engineers are more likely to use the seismic results by incorporating them directly into their engineering analyses.

Results from two case histories are presented showing the benefit of assessing seismic refraction data using the DEM-PFC numerical modeling approach. In the first example, standard 2D refraction data were analyzed and the interpolated results were presented as a 3D model. The second example is a 3D surface tomography reconstruction of four slightly offset 2D refraction shot lines.