The Observational Approach for geotechnical design and construction was proposed by Karl Terzaghi in 1945 and then reworked and discussed by Bjerrum and Terzaghi in the early 1960's as a means to cope with the inherent uncertainty of working and construction with and through Earth materials. While this process was successfully weaponized by Terzaghi in many significant engineering works, Ralph Peck wrote a seminal paper in 1969 on the advantages and limitations of the method (in soil mechanics). To use a tunnel example, the basic premise of the method is that rather than attempt to define and locate, a priori, all conditions along the tunnel, the project team conducts enough exploration to define the nature, pattern and range of properties, assess the most probable deviation from these conditions, establish a design based on most probable condition, with a toolbox of modifications to the design should there be a foreseeable deviation from this baseline. Measurable quantities during construction (wall displacement, ground settlement, support loads) are to be determined and monitored in as much as they reflect actual conditions. The as-built design is then modified to suit actual conditions. This method has served geotechnical engineering well, but as Peck observes, "“It is, of course, essential that the quantities to be observed should reflect the phenomena that will actually govern the behaviour of the works to be constructed....A mistaken preconception of the nature of the problem may lead to omission of observations of types that would have disclosed the real reasons for concern.” This last statement is never more a truism than in major tunnelling and underground construction projects in complex rock with extensive tectonic history and speaks to the critical importance of fundamental Engineering Geology as a both a precursor and a critical tactical tool for the implementation of observational design. While in many projects in rock engineering, observational design has been reduced to recording immediate tunnel performance or rock properties at the working heading and modifying the local design according to tabulated, prespecified approaches. In complex geological environments, this approach fails in a manner consistent with Peck's warning quoted above. This talk will highlight, through case histories on four continents, the critical importance of creating and updating the broader geological model, throughout construction, in addition to geotechnical monitoring of the immediate project envelope.

​

Mark Diederichs, PhD, PEng, FEIC, began his career with a BASc in Geological Engineering (1987) and completed an MASc in Rock Engineering (1989) at the University of Toronto. He moved on to a position as a field micro-seismic engineering technician at Creighton Mine in support of an industry-academic collaborative research program to improve monitoring and response in deep dynamic mining environments. From there, he worked with an R&D team at Laurentian University on a variety of projects including conventional ground support, support for large mining stopes, micro-seismic and rockburst research, and other tools in support of mining geomechanics. He is the author of a handbook on mine support and completed his PhD in brittle rock mechanics at the University of Waterloo in 1999. He has been a Professor of Geological Engineering at Queen’s University since 2001, has authored 400 refereed articles on a variety of subjects in rock mechanics, engineering geology, and rock engineering. He has been an active consultant in Canadian and international mining and tunnelling for over 25 years and currently acts as a senior international advisor and reviewer in site investigation, rock engineering and design, risk management and contractual issues for major underground and surface mining, tunnelling and cavern construction, hydroelectric power projects, nuclear repository geoscience, and engineering and rock mechanics including structural instability, squeezing and swelling ground, and brittle fracture and rockbursting.