A proactive approach to in situ stress characterization in underground slate mine.
Accurate knowledge of the in situ stress state is a cornerstone of underground design. Whether the objective is to size large chambers, define support requirements or assess stability, the mechanical environment that exists before excavation fundamentally governs ground response. Yet in many engineering projects, especially in mining, this information is only obtained after substantial excavation has already taken place, limiting its usefulness for proactive design.
Our work addresses this challenge in the context of slate mining in Galicia, Spain, where modern extraction increasingly relies on large underground chambers with spans of tens of metres. Roofing slate is an economically important dimension stone, and Spain accounts for a significant share of its global supply. In this environment, the transition from open pit to deep underground mining has introduced new demands for reliable stress data before large volumes of rock are removed.
The full research is published in the Journal of the Southern African Institute of Mining and Metallurgy and can be accessed via the DOI: https://doi.org/10.17159/2411-9717/1916/2022
Traditionally, stress ratios such as the k-ratio defined as the ratio of horizontal to vertical stress have been inferred after mining, using observations from newly formed openings or from instrumentation installed once the chamber is already excavated. While this post-excavation approach can provide valuable information, it is inherently retrospective and offers limited value for the design of the chamber itself and the choice of support systems. In situations where large volumes are to be excavated at significant cost and risk, a method that provides stress estimates before excavation is clearly preferable.
To address this gap, we proposed a methodology that combines two powerful yet complementary tools: direct stress measurements using the flat jack technique and numerical back-analysis through finite element modelling. The flat jack method involves inserting a hydraulic jack into a slot cut into the rock and measuring the pressure required to restore the original spacing of reference points. This pressure corresponds to the in situ stress acting perpendicular to the slot’s axis, providing a direct and local estimate of stress components in the rock mass.
In the mine studied, access adits were available well before chamber excavation, making them ideal locations for preliminary stress measurement. The geological quality of the rock mass in the adits was similar to that expected in the chamber area. This allowed us to measure stress values that closely represented the virgin field, without the confounding effects of excavation-induced relaxation.
Conducting flat jack tests requires careful selection of measurement locations to minimize the influence of blast damage and existing fractures. In our case, readings were taken along the roof and sidewalls of an access adit at a depth of approximately 300 metres, where structural continuity and rock quality were favourable. Each flat jack measurement captured deformation restored by the device and the corresponding pressure provided by the jack, from which stress components were inferred.
However, direct readings alone are not sufficient to define the stress state confidently. Measurements are affected by experimental error and the presence of localized disturbed zones caused by excavation. To navigate these uncertainties, we used a numerical back-analysis procedure based on a two-dimensional elastic finite element model created in Rocscience RS2. By comparing the ratios of stresses derived from flat jack data with numerical solutions across a range of assumed stress ratio values, we identified the best match between measured and modelled stress states. This approach effectively normalizes experimental deviations and isolates the underlying virgin stress distribution.
The results were illuminating. The stress ratio obtained from flat jack readings, when interpreted through the numerical back-analysis, corresponded closely to values that had been previously inferred only after chamber excavation through instrumentation data. In other words, the proactive method was validated by comparison with independent post-excavation observations, confirming its practical reliability. This alignment not only enhances confidence in the methodology but also highlights the potential for improved design decisions before excavation begins.
From a practical standpoint, the implications are significant. Underground chamber dimensions in slate mines are large, sometimes on the order of 150 000 m³, and support requirements vary substantially with assumed stress conditions. Overestimating stresses can lead to overdesign, increased cost and unnecessary support installation. Underestimating them risks instability and safety concerns. By providing a method to accurately estimate the stress state before chamber excavation, engineers can optimize support design, reduce uncertainty and improve safety outcomes.
At a conceptual level, the study demonstrates the value of integrating field measurements with numerical modelling. The flat jack method gives localized, direct stress data that reflect actual conditions in situ. Numerical back-analysis situates these measurements within broader mechanical behaviour, accounting for geometric and material responses. Together, they offer a more complete picture than either could alone.
This synergy echoes broader trends in geomechanics and underground engineering, where hybrid approaches that combine empirical observations with computational methods increasingly define best practice. As measurement technologies and modelling tools continue to evolve, such integrated methods will likely become standard in challenging environments where stress conditions are critical to performance.
In the specific context of slate mining and similar rock masses of good geotechnical quality, the methodology offers a blueprint for extending stress characterization into early project phases. It is especially relevant in operations transitioning from surface to underground extraction, where knowledge gaps about pre-excavation conditions are largest.
The broader lesson also touches on how engineering disciplines should frame uncertainty. In situ stress is one of the most variable and influential parameters in underground design, yet it is also among the most difficult to measure directly. By combining direct and modelled data, engineers can reduce uncertainty in ways that single methods cannot achieve. This not only improves design quality but also aligns engineering decisions with realistic representations of subsurface conditions.
In summary, the development and validation of a proactive stress estimation method based on flat jack measurements and numerical back-analysis represent a meaningful advance in geomechanical practice. It enhances our ability to design safer, more efficient underground chambers by providing reliable stress information before excavation begins, thereby aligning technical knowledge with practical needs in the field.
The full paper is available in the Journal of the Southern African Institute of Mining and Metallurgy:
https://doi.org/10.17159/2411-9717/1916/2022
More of my research on rock mechanics, underground design and stress characterization can be found on my Springer Nature author profile:
https://communities.springernature.com/users/antonio-alonso-jimenez