Waste Rock Dumps water balance




  • Avatar
    Brink, Nicholas

    Hi Camilo, I unfortunately don't have any example models that I could readily send, but in general we tend to model WRDs similar to leach pads - we would estimate runoff using hydrological approaches (although for sizing collection drains it is common to assume no runoff and to route all precipitation into and through the rock pile).  Depending on the complexity of the situation, we may also set up a two-dimensional saturated/unsaturated seepage model to estimate seepage rates while modeling precipitation at the top of the WRD (requires some knowledge of the WRD material characteristics).  These models also feed into geotechnical stability analyses, etc. 

    In GoldSim we have used an attenuation relationship to simulate the time delay between precipitation entering the top of the pile and it exiting through the seepage collection infrastructure.  The attenuation relationship can be calibrated based on seepage model results, or using historical precipitation and seepage collection records (i.e. by looking at the delay between a precipitation event and a corresponding increase in seepage collection rates).

    I apologize for the late response, but I hope this helps!  Cheers!

    Comment actions Permalink
  • Avatar
    Camilo Martinez

    Thank you Nicholas, this is very helpful.


    Comment actions Permalink
  • Avatar
    Saeed Golian

    Hi Camilo and Nicholas,

    Following this topic, have you found any model or succeeded in developing a Goldsim model for WRD simulation? 

    Thanks if you share your experience,



    Comment actions Permalink
  • Avatar
    Nguyen Anh Thao Ngo

    Hi all,

    Have you found any model example for waste rock simulation?

    I am interested in this topic as well. I am trying to develop a spoil pile component in mine pit water balance, to calculate spoil runoff, infiltration and spoil seepage contributing into the pit lake. As I know, most GoldSim model practices in mining industry build a spoil or waste rock as a sub-catchment type in AWBM (along with other catchment like natural, highwall, etc.), which mean spoil seepage is the "baseflow_runoff", and spoil runoff is the "surface_runoff" in AWBM.

    I think the problem is there is a lack of data on waste rock properties, and water table in the spoil, to build in the simulation and people just simplify it by using AWBM.

    Do you think the "Spoil_Heap_Runoff" example file is a good reference for this case?



    Comment actions Permalink
  • Avatar
    Nicholas Brink

    A good starting point would probably be some of Jason's infiltration/runoff models.  The "Spoil_Heap_Runoff" model could be a good example too.  I haven't dug into those examples yet but from what I've seen they probably will get you started on the right track.  Thao - it sounds like your question is more related to hydrology of the spoils piles in which case I agree the biggest hurdle is lack of information.  If the piles already exist, the best approach is often to develop a simplistic GoldSim model that accounts for the key aspects listed later in this reply, which can be calibrated.  That requires knowing something about the seepage rates, phreatic levels, or similar within that pile.  In the absence of any useful information, you probably need to dig into literature estimates of saturated and unsaturated properties.  I highly recommend at least some quick/simplistic 2D seepage models to help inform your GoldSim model, but if little information is available those models should not be too much of the focus of your time (I would think).

    As some additional background, key components to capture (in my opinion) are summarized below.  I'll list below the "detailed approach" requiring lots of characterization and modeling, applicable for models needing high accuracy and precision.  Probably you can take the concepts below and apply them more conceptually to develop a more high-level or simplified model.  This is just a list of items we like to consider in these sorts of models.

    • Infiltration versus runoff estimates - this can vary widely from one site to the next based on climate conditions, and based on properties of the waste pile/dump.  If open rock with little or no fines (rare occurrence) then it may be  possible to just assume 100% infiltration.  Also influenced by local climate.  In very wet regions there will likely be more immediate runoff (not in all cases, but may help to conceptually get a model started).
    • Change in storage - pore volume in these piles is often large, even in piles consisting of mostly topsoil.  So there must be a storage component modeled to simulate how much water remains within the pile versus seeps through it.  Change in storage will express as a change in the phreatic surface elevation.  That in turn will lead to changes in seepage rate at the base of the pile (i.e. greater driving head).  It is advantageous in most models to develop a stage/storage curve for the pile in question which shows the relationship between phreatic surface elevation and water stored in pore spaces.  This becomes an input to your model, and allows you to track the phreatic surface elevation over time.  Likely you can trigger new seepage rate calculations as the phreatic surface elevation changes - for example if the phreatic surface rises to a level where it encounters a drain, or encounters a topographic feature that changes the drainage regime, the seepage rate may change.
    • Delay in seepage release - the time between a rainfall event, and seepage appearing at the bottom of the pile, may be hours or months.  It depends on the hydraulic conductivity of the pile, and the unsaturated conductivity can play an important role here.  In GoldSim it may be convenient to use a delay element which also considers storage to accommodate this bullet and the previous. If a high level of detail and precision is needed, it becomes necessary to perform transient seepage models to evaluate this.  And, it becomes necessary to sample material from the pile and test for saturated and unsaturated seepage properties.  Literature values may be considered to start with, but again if a high level of precision is needed then sampling and testing is absolutely required, along with advanced unsaturated seepage modeling.  Often those models can be coupled with vadose models to get better estimates of infiltration rates as well with site-specific climate conditions considered.
    • Infiltration into native foundation - this is hard to estimate, unless the pile is fully lined underneath in which case infiltration is likely close to zero.  Most piles are unlined.  If you have characterized permeabilities and groundwater levels in the foundation, you can use those data in a seepage model to help estimate infiltration rates.  Infiltration rates will also be influenced by phreatic surface elevation. 
    • Seepage Exits - Most piles are designed with drainage features to keep the phreatic surface below the face of the pile to avoid surface seeps, for both physical stability and for environmental reasons.  Seepage through these features can be highly three-dimensional and can be difficult to simulate in a 2D seepage model (therefore difficult to simulate in GoldSim).  Calibration to known conditions, in an existing pile, is my preferred approach where possible.  In extreme cases it may be possible to do 3D seepage analyses, though I've never encountered a situation requiring that level of detail.  As an alternative, you can create a GoldSim model that limits the elevation of the phreatic surface to a maximum elevation associated with the drainage feature, and then calculates the seepage rate through that feature that would be required to maintain the phreatic surface at that elevation.  A simple approach is to limit the maximum storage in your delay/storage element to the water storage value from your stage/storage curve, associated with the elevation of the drainage feature.  The storage element in GoldSim will then have an outflow node which you can link to, and which will automatically calculate that seepage rate for you. 

    The above items assume the pile in question has not been constructed yet.  If it already exists, it may be possible to forego some or all of the modeling discussed above (again, depends on the level of detail/precision needed) and instead develop a more simplistic model in GoldSim that can be calibrated.  Calibration is highly recommended if the pile in question already exists, and if high-quality seepage and climate data are available.  Often that is not the case, but if you have reliable seepage measurements at the toe of the pile, or piezometers within the pile, you can use those data along with local climate station information to calibrate and improve your GoldSim simulation.


    Comment actions Permalink

Please sign in to leave a comment.