Abstract:

When Can Four to Twelve Tests Be Sufficient for Engineering Design Parameters? A Tale of Two Papers

Keaton, Jeffrey, Wood, jeff.keaton@woodplc.com; Abdul Shakoor, ashakoor@kent.edu (Poster)

 “It depends…” is the only logical answer. The title of a 2019 Environmental & Engineering Geoscience paper posed the question, “Are three to ten tests enough to characterize a rock property?” and cited a 2014 paper entitled, “Selection of geotechnical parameters using the statistics of small samples,” with an example using four and 12 tests. Among the many references cited in the 2019 paper was one by three authors in 2005 with the title, “Determining the minimum number of specimens for laboratory testing of rock properties.” Among the few references cited in the 2014 paper was a second 2005 paper by the same three authors entitled, “A statistical approach for determining practical rock strength and deformability values from laboratory tests.” This contrasts science, seeking to explain and predict, with engineering, seeking to design with multiple constraints. The 2019 paper concludes that ten or fewer measurements are too few to capture the inherent variability of rock properties. The 2014 paper begins with “Selection of parameters for geotechnical calculations frequently is based on results of relatively few laboratory tests that are used to guide judgment of experienced engineers.” It concludes that rigorous statistical procedures provide a method for data analysis in which target values of confidence interval, precision index, and stability probability are treated explicitly, which provides a way of analyzing test results that allows uncertainty and variability to be quantified for use in probabilistic stability analyses and the value of additional test results to be demonstrated. The two 2005 papers were written by professors from a university department that includes mining engineering; mining companies use rigorous statistical analyses of mineral assay results for economic decisions. It should be no surprise that feasibility-level stability analyses of large open-pit mines use practical strength values from rigorous statistics on a handful of tests.

Bio:

Jeffrey R. Keaton specializes in quantifying hazardous natural processes for siting and design of all types of facilities in all geologic environments and is particularly interested in quantifying uncertainty and variability in geologic observation and characterization. He has degrees in geological engineering, geotechnical engineering, and geology from University of Arizona, UCLA, and Texas A&M University, and is licensed as an engineer and as a geologist. Keaton has been employed by consulting firms for over 45 years, and in Wood’s Los Angeles office since 2005 where he is a principal. He served until recently on the editorial boards of Engineering Sustainability Journal (ICE UK), Quarterly Journal of Engineering Geology and Hydrogeology (The Geological Society of London), Bulletin of Engineering Geology and the Environment (IAEG), and Environmental and Engineering Geoscience (AEG & GSA). He is past president and honorary member of the Association of Environmental & Engineering Geologists, and Fellow of the American Society of Civil Engineers and of the Geological Society of America. He a past member of ASCE’s Committee on Sustainability and of the steering committee of the Geotechnical Extreme Event Reconnaissance (GEER) Association. He recently served as an officer of the Engineering Accreditation Commission of ABET, and is an ABET Fellow.

CEU Evaluation Form