Session 1 | Track B | Reclamation and Remediation
Abstract
Complex numerical models are commonly used to predict the performance, and often to refine the design, of mine waste covers. However, such modeling is fraught with uncertainty due to its reliance on heterogeneous and temporally variable inputs. This includes climate inputs that vary over the short-term (weather), medium-term (seasons), and long-term (climate change).
Understanding the climatic trends is critical because similar magnitude rainfall events can result in vastly different cover responses, depending on the precedent moisture conditions and soil macrostructure. This is most evident for store-and-release type covers, where the net percolation in the rainy season is a direct function of moisture storage at the end of the preceding dry season.
Modeling these climatic trends is often difficult as site-specific climate data are often not available, or are not of the quality required, for defining soil-plant-atmosphere boundary conditions. The current state of practice to address this issue is to conduct sensitivity analysis. This approach is however somewhat limiting in that sensitivity analysis without recognition of probabilities results in an equal weighting of outcomes which may be misleading. Furthermore, percolation modeling – specifically surface flux boundary modeling – is mathematically complex and this often becomes a limiting factor in how much analysis can be done. Finally, presenting model results from individual sensitivity runs as “absolute” values, or even worse averages could be misleading and inadequate for decision making as it can obscure transient phenomena.
An alternative framework to address uncertainties in climatic records is the use of synthetic datasets developed using stochastic techniques. With the use of modern computing tools, unique algorithms and routines can be developed that allow for comprehensive stochastic percolation modeling to be completed in a fraction of the time it used to take, and as a result, present percolation as a probabilistic input. This greatly improves our ability to assess climatic risks in cover design and adjudicate the value of proposed covers. Further, it can facilitate stakeholder and regulator communication both in the short-term as well as long-term through the presentation of cover performance to different future climatic events.
This presentation covers new algorithms and routines developed to conduct stochastic cover design and presents specific case studies to demonstrate the value of the approach against the traditional sensitivity-based design.
Authors
- Iozsef Miskolczi | Senior Geotechnical Engineer | SRK Canada
- John Mayer | Senior Hydrogeologist | SRK Canada
Click here to view other SRK presentations at this event
