Site Conceptual Modelling for Tailings Storage Facility Seepage Assessment

Site Conceptual Modelling for Tailings Storage Facility Seepage Assessment

The importance of assessing seepage risks

Seepage management is critical in reducing failure events which can have unintended consequence associated with water balance and social operating licence (ANCOLD, 2019a). Hydrogeological risks associated with Tailings Storage Facilities (TSF) seepage are connected to impacts on groundwater levels and reduction in environmental water quality. Of concern is the potential migration of contaminants offsite. Seepage emanating from the TSF can migrate in both the vadose and saturated zone along preferential pathways.

In general terms, two seepage pathways are present, one through the embankment and a second associated with seepage through the footprint area. This article deals with the later as TSFs are commonly underlain by sedimentary or alluvial aquifers in which the permeability is primarily associated with the lithologic unit and result in variable infiltration zones.

Under such circumstances, the treatment of the foundation material as a homogeneous porous media would not be appropriate. The consequence is an increase in the risk to the design due to incorrect assumptions relating to groundwater vulnerability. Representative conceptual models relevant to site conditions are essential to reduce risks associated with seepage and migration of contaminants in groundwater.

We provide an overview of the elements of developing a site conceptual model appropriate for assessing seepage risks associated with the TSF footprint area. Our framework for assessment ensures we document key elements in the process and develop a firm understanding of potential risks to a TSF project.

TSF footprint seepage conceptual models

There are many definitions of conceptual models. For this article, concerning TSF seepage problems, a conceptual model is an evolving (non-mathematical) hypothesis identifying key features, processes and events controlling fluid flow and contaminant transport of consequence at a specific field site in the context of a recognised problem (NRC, 2001).

Worthy of note in this definition is that it is non-mathematical. The purpose of the conceptual model is to provide the basis for mathematical modelling. It is critical that the conceptual model is representative so that errors and omissions at this stage are not carried through into the analysis. Secondly, it is constantly evolving as information is obtained from various stakeholders. The conceptual model should be

  • developed at the onset of the project and collate available data to aid in understanding potential data gaps in the design
  • challenged as soon as possible in a workshop setting with the client, project hydrogeologists, geotechnical engineers and regulatory or community leaders

The benefit of a site conceptual model is the use in planning activities, such as field investigations, cut-off trenches and design criteria.

Key elements of a TSF conceptual model are:

  • Regional groundwater setting
  • Hydrogeological and geochemical properties of tailings, foundation material and the underlying country rock. The availability of information on local faults, shears, joints, fractures or ore veins should be noted and incorporated in the conceptual model.
  • TSF geometry, deposition strategy, and decant pond, especially free water surfaces, should be included in the conceptual model.
  • Clay liner design and operational concept
  • Footprint topography and potential modifications due to footprint preparation
  • Water balance and groundwater levels within the TSF development area

Framework for TSF footprint seepage conceptual models

The complexity and extent of the conceptual model depend on the problem being solved. The focus of the conceptual model will also depend on the objective of the investigation and outputs required for design or regulatory purposes.

To improve the quality and consistency of the conceptual model, the following framework steps are recommended: problem definition – initial conceptual model development – conceptual model validation – presentation of results and interpretation.

Problem definition: The purpose of the conceptual model is to answer a specific question or set of questions. In this case, the problem to be investigated is the quantification of footprint seepage or could be associated with a trade-off between two liner options. The problem definition is the primary factor in deciding appropriate simplifications and assumptions for the model and factors that could negatively affect the project.

Initial conceptual model development: With a problem definition, a relevant conceptual model can be constructed. Field data are assembled and analysed to define the hydrogeologic system. It is recommended to develop multiple conceptual models at an early stage to account for uncertainty in describing the field setting. A site visit is recommended to allow understand the hydrogeological setting, and to contextualise the assignment of parameter values. A factor that impacts on seepage rates in the foundation is the variability of the underlying TSF footprint. The stratigraphy of the TSF footprint should be discretised at a finer scale on bore logs but should not exceed a resolution of more than 1 m in the vertical interval to represent the rapid changes in lithology. The units are assigned hydraulic conductivity based on stratigraphy using appropriate gridding algorithms.

Conceptual model validation: Hydrogeological conceptual models contain a high level of uncertainty due to simplifications made to represent a complex hydrogeological system. Although hydrogeological conceptual models are intended to be non-mathematical, the validity of the assumptions adopted should be tested through simple analytical models. Where data is available, alternative versions of the hydrogeological conceptual models should be tested during calibration.  The favoured conceptual representation is retained for full scale predictive numerical modelling.

Presentation of results: The details of the conceptual model are presented in the modelling report.  The hydrogeological report should document the process followed, presenting the hydrogeological conceptual model(s) and states and limitations. The report should include details on the hydrogeologic setting (both regional and local scale), and explanation of the data and assumptions used to formulate the conceptual model. It should also highlight the extent of the specific purpose of the model to limit potential errors in analysis.

Examples of TSF footprint conceptual model elements

ATC Williams is involved with numerous TSF designs which require seepage assessments of the footprint to understand liner or footprint preparation required. Following are examples of conceptual models and associated elements applied in these projects.

Diagram 1 shows an idealised weathering profile in which the regional stratigraphy is validated against site-specific bore logs.


Diagram 2 shows stratigraphic conceptual models developed from assessing the frequency of the occurrences of the different soil types at 2 m depth intervals from ground surface to the bedrock surface for the geology unit assessed.


Diagram 3 shows a typical conceptual model for a TSF developed in this environment. Initial percolation into the foundation passes through the unsaturated zone located between the foundation of the facility and the underlying rock. If infiltration continues for a long time, a perched water table will develop at the top of the residual soil resulting in a saturated mount within the tailings. A partially saturated wetting front may advance downwards through the residual soils to the region water table to form a second mound.


Diagram 4 illustrates seepage patterns assessed as part of conceptual modelling.


ATC Williams experience in hydrogeology offers a unique perspective in the design of tailings dams and underground infrastructure in general. In assessing seepage potential ATC Williams applies a combination of analytical and numerical modelling, and Monte Carlo analysis to assess and simulate multiple factors that can influence seepage rates in large structures.

Key Contacts

Freeternity Rusinga on 0423405894 or via email FreeternityR@atcwilliams.com.au
ideon Steyl on 0424974473 or via email GideonS@atcwilliams.com.au