The landscape below our feet is unseen and uncertain – so any construction project involving major subsurface excavation will unearth a range of complexities and risks. A comprehensive site investigation allows us to proactively manage risk by providing insight into the likely conditions to be encountered underground. This is the key to a solid start and a stable project with a lower likelihood of unwanted surprises.
Those unwanted surprises can be very serious. Instability can result in financial and social consequences, from damage to equipment and disruption of the project through to negative impacts on the health and safety of workers. In the project development stage, it is therefore essential to develop a detailed understanding of the project’s geological and hydrogeological setting using a phased site investigation approach.
Underground construction projects will naturally face different challenges depending on their location and setting. When it comes to works involving tunnelling, we need to understand the strength of the ground and how it will behave so that we can determine an appropriate design.
However, stability analysis and ground support design for underground structures are very challenging areas of practice, needing careful consideration and specialist expertise.
A complex range of instabilities
For both rock and soil tunnelling, investigation into a complex range of instability mechanisms is needed.
For rock tunnels, there are three key variables when assessing instability issues: rock mass strength, stress, and water. At shallower depths, the stability of the rock mass is generally governed by fractures in the rock. At greater depth, the behaviour of the rock mass is generally governed by stress. The presence of groundwater pressure in the rock fractures adds to stress and instability in the rock mass.
Depending on the fracturing and stress levels in the rock mass, the following types of instability can occur for underground excavations:
- structurally controlled instability caused by pre-existing discontinuities (defects such as beddings, joints, and faults),
- instability caused by rock fracturing or displacement under stress, and
- a combination of a stress-induced mechanism and pre-existing discontinuities.
With such a variety of potential instabilities, it is important to develop a geotechnical and hydrogeological investigation program to obtain the data necessary to determine and assess instability, and to inform the design of a stable structure.
In the first category of instability (structurally controlled), the intersection of pre-existing discontinuities forms blocks, which might fall or slide if they are cut by the excavation boundaries.
The movement of these rock blocks is highly dependent on their geometry and the frictional resistance of their surfaces.
The second and third categories of instability are controlled by stress-induced rock fracturing and associated movement. Underground excavations disturb the equilibrium of existing stresses and impose a new distribution of stresses. Depending on the excavation geometry and the characteristics of the existing and new stresses, the magnitude and distribution of stresses will vary around the excavation. When the new distributed stress exceeds the splitting strength of the intact rock, fracturing will occur. If the stress-induced rock fractures intersect with the excavation boundary or pre-existing discontinuities, removable blocks may form.
Digging into data yields deeper insights
A desktop study is the first step in any site investigation program. The study should focus on identifying potential geotechnical, hydrogeological and contaminated land issues and should involve a review of the literature and existing data. If done well, data gathered at this stage can have a significant beneficial impact on the required extent and cost of the final site investigation.
Ultimately, the information needed from a site investigation includes identification of the soil and rock units likely to be encountered, their geotechnical and hydrogeological characteristics, the magnitude and orientation of the in-situ stress field, potential groundwater levels and pressures at depth, and existing soil and groundwater quality.
When designing a site investigation program for an underground construction, it is important to be aware of the limitations of what we can test and fully understand. As we cannot test the large volume of ground that surrounds an underground excavation to measure its strength, we test small samples gathered from the site investigation, and then scale up these properties based on empirical correlations and classification schemes.
We also cannot measure the magnitude or orientation of in-situ stress directly. We measure other properties and make assumptions about how stress affects them. For properties that are difficult to measure directly, we make assumptions based on correlations from other tests. Laboratory testing must be broad enough to infer strength from the range of in-situ stress conditions expected at the tunnelling horizon.
We also need to be aware that site investigation activities can disturb the material we want to test. For example, the disturbance imposed during rock stress measurement may affect the reliability of the results, which requires that extra thought be given to how to minimise this potential impact at the planning stage.
Disturbance of soil samples also requires extra attention to ensure quality testing results. This includes preserving the sample’s in-situ moisture content, which should ideally remain unchanged throughout sampling and testing. Soil strength testing must also consider whether the testing should be completed under drained or undrained conditions. Taking measurements in-situ using instruments such as cone penetrometers, piezocones and pressuremeters can help overcome such testing scale and sample disturbance effects.
New tools improve underground ‘vision’
Our ability to ‘see’ into the subsurface and understand ground conditions will be greatly aided by the rapid development of non-destructive methods, core scanning and digital technologies. Seismic tomography, geoelectrical resistivity imaging, advanced hyperspectral imaging technology and 3D visualisation methods hold great potential. Tomography methods cover a much larger volume of the ground than can be exposed via boreholes, and do not disturb the soil and rock units. Their results can be used to characterise the ground both qualitatively and quantitatively. In addition, core scanning and borehole geophysical imaging such as acoustic televiewer (ATV) and sonic logs are useful means to improve interpretation of the conditions encountered in the recovered core, as well as in the walls of the boreholes and immediately surrounding ground.
Embracing such advances in methods and technologies will increase our understanding of a project’s geotechnical and hydrogeological setting, but there will still be inherent limitations to our knowledge of what lies down below. A well-designed site investigation is an important investment in managing this risk of the unknown and is a vital step towards delivering a successful underground construction project.
About the Authors
Les McQueen is a Principal at WSP* based in Sydney, Australia. He has 40 years' experience in rock mechanics tunnel support design and engineering geology for civil and mining projects. Les has consulted on transportation, hydroelectric and utility tunnels from feasibility planning, tender and detailed design stages through to construction (TBM, roadheader or drill and blast), which has included projects in Australia, Hong Kong, Indonesia and the Philippines.
Mahdi Zoorabadi is a Principal Geotechnical Engineer at WSP* based in Newcastle, Australia. Consulting in the areas of ground control and rock engineering, he specialises in rock testing, rock stress measurement, groundwater studies, numerical modelling, rock slope stability, primary and secondary support design for underground structures. Mahdi is a highly experienced numerical modeller who has provided rock engineering consulting services to civil and mining projects for over 18 years. He is actively involved in research activities through collaboration as adjunct senior lecturer with the School of Mining Engineering, University of New South Wales.
* This work was performed by Golder professionals who joined WSP in an acquisition completed in 2021.
