Effect of Weathering on the Gravitational Stability of Questa Rock Piles
By Murray Fredlund
February 25, 2009
The Questa Rock Pile Weathering and Stability Project (QRPWASP) is a scientific research project whose purpose was to determine how and to what extent weathering affects the gravitational stability of the Questa mine-rock piles in 100 and 1000 years. The project studied a theoretical composite rock pile taking characteristics from several on site, and was not a site-specific engineering or regulatory analysis of a specific rock pile. Gravitational stability refers to the static stability evaluated along circular (and other-shaped) failure surfaces.
The Questa molybdenum mine (Chevron Mining Inc., formerly Molycorp, Inc.) is located in the Sangre de Cristo Mountains in north-central New Mexico and is on southward facing slopes at elevations of approximately 2290 to 3280 m. During the period of open-pit mining (1969-1982), approximately 317.5 million metric tons of overburden rock was removed and deposited onto mountain slopes and into tributary valleys, forming nine rock piles surrounding the Questa open pit.
SoilVision Systems Ltd. was entrusted with the numerical modeling of the hydrological and slope stability aspects of the waste rock pile. The research aspect of the project required the furthering of current state-of-practice in order to employ new technologies for the analysis of rock piles. The project required the inclusion of unsaturated aspects of slope stability analysis and therefore a corresponding calibrated hydrological model in order to accurately describe the distribution of pore-water pressures in the rock pile. The interaction between processes involved in the waste rock is illustrated in the figure to the right.
Numerical modeling of a rock pile at this level of sophistication had not previously been attempted. Therefore the project involved new challenges, which needed to be resolved through the implementation of new technologies in software. Some of the specific challenges and findings are outlined in the following sections.
Gravitational stability analysis of the rock pile involved several unique numerical modeling challenges. A traditional total stress stability analysis would not account for the effects of suction on stability. There were also the questions of: "Can a weak layer compromise the stability of a slope?" and "What strength properties would a weak layer need to demonstrate in order to cause failure conditions to occur?"
Specific technical challenges for the slope stability model included:
Specific new technologies employed in the project included:
A unique calibration opportunity arose as the rock pile slope was being "flattened" by bulldozers. A crack formed during the "pushing down" process and provided a unique opportunity to perform a 2D back-analysis, shown in the diagram on the right.
The purpose of the hydrological numerical modeling was primarily to perform a calibration to the existing flow in the waste rock system. Once a calibrated model was achieved it could then be further used: i) as the basis for prediction of future pore-water pressure stress states for input into slope stability modeling and ii) the model would then form the basis for future geochemical modeling.
The primary numerical modeling challenges were flow in a waste rock system being i) unsaturated and ii) the material was primarily coarse-grained. These two aspects typically lead to a high degree of non-linearity in solving Richards flow equation. New techniques were employed in resolving these numerical obstacles.
Specific technical challenges of the hydrological model were:
A series of conceptual models were set up in SVFlux and solved. A number of improvements were implemented in the software package in order to i) improve numerical convergence and ii) handle the upper crust boundary condition in a reasonable manner. Automatic mesh refinement was used in order to handle the challenges of infiltration resulting from precipitation events along the upper boundary of the model. A methodology for handling uncertainty in the hydrological modeling was also established.
A key aspect of the project was that it was impossible to obtain a properly calibrated numerical model unless the crust at the top of the rock pile was properly represented in the numerical model. Several methodologies for handling crusts were developed as a result of this project.
Some of the finite element models solved for this project may be seen in the screenshots shown above.
The project was completed in December, 2008 and represents a landmark study which establishes the methodology for the analysis of future rock piles.