Over the years, four programs of rock stress measurement and interpretation have been completed at the Ok Tedi Mine in Papua New Guinea, each providing different interpretations of the in situ field stresses. Measurements were taken at different locations and depths, and in different rock types. The hydro-fracturing method inferred a hydrostatic model, the acoustic emission method presented a highly anisotropic horizontal stress regime, and overcoring measurements using ANZI strain cells inferred a rocktype dependent but more benign intermediate stress regime.
The principal stress directions are not necessarily horizontal and vertical; perhaps impacted by rugged topography.
Even if the pre-mining in situ stress regime was uniquely consistent, considerable redistribution of stress directions and magnitudes were expected to develop due to interaction between open pit and underground excavations.
2D and 3D numerical modelling, using finite element, finite difference and distinct element methods, was used to evaluate, in terms of stressstrain response:
- The stability of the 1000m high West Wall at each stage of its proposed cutback
- The stability and long-term ground support requirements for an existing drainage tunnel beneath the pit
- The stability and sequencing of underground excavations for sublevel open stoping beneath the East Wall of the pit, as well as the stability of pit wall with the progression of the underground mining
The modelling involved a multistage approach; interpretation of the most appropriate ‘starting’ stress conditions in each of the mine areas under assessment was attempted following initial modelling phases. Comparisons of the effects of the various in situ stress field interpretations were made in the context of the continuing and progressively enlarging and deepening mining operations.
As the modelled field stresses changed through the mining sequence, the Hoek-Brown failure curves for the main material types were more closely approached or exceeded (predicting yielding of the rock mass). From observations made within the existing excavations with regards to rockmass conditions, failures and performance of ground support, it became apparent that certain stress fields may have been overestimating the stress conditions at the time of excavation.
Two stress field alternatives, the modelling of which resulted in strain that more closely approximated the observed conditions in various locations, were subsequently selected as sensitivity cases for input into further analyses for mine design.
The response and stability of open pit slopes was not particularly sensitive to the adopted pre-mining stress regime: the alternative in situ stress models produced a variation in factor of safety for the pit walls in the order of 3-5%. However, stress redistribution and concentration between surface and underground mine workings was highly dependent on the initial in situ stress regime and extraction sequence, with a 100-150% variation for the alternative in situ stress models in the near-excavation concentrated stresses for the underground openings.
