Model for predicting drainage quality from mine wastes Muhammad Muniruzzaman

Transcription

1 Model for predicting drainage quality from mine wastes Muhammad Muniruzzaman Kuopio

2 Introduction Quantitative modelling is instrumental in effective predictions of water quality from mine wastes Prior understanding about the future series of events in waste settings may help performing more robust risk assessment and designing remediation measures Increasingly recognized not only in data interpretation but also in long-term predictions Received a great deal of research attention and led to the development of a variety of modelling approaches However, there is still a lack of attention towards utilizing predictive models at early (e.g., planning) phases of a mine 2

3 Governing Equations Water Flow Solute Transport Gas Transport Mineral Reactions 3 ( ) w i r w w s Q h k t S t h S S = + K θ ( ) ( ) ( ) = = = + = g m r N r g ig N m m im N r r ir i i w i i w R R R C S C C S t υ υ υ θ θ D q ( ) ( ) g i g i g i p g g i g g i g Q C S C C S t + =, ) ( D q θ θ ( ) = j j j j j j K IAP C k R 1 υ γ

4 Modelling Approach 4

5 Modelling Approach (cont d) Structure of the Model Numerical Method: cell-centered Finite Volume Method 5

6 Example 1: Särkiniemi Waste Rock Pile 6 Mineral Mineral Content Surface Rate Reference Area, A coefficient, k [wt%] [mol L -1 w ] a [m 2 L -1 w ] [mol m -2 s -1 ] Biotite Nagy (1995) Hornblende Palandri & Kharaka (2004) Serpentine Declercq & Oelkers (2014) Albite Palandri & Kharaka (2004) Chlorite Palandri & Kharaka (2004) Pyrrhotite Williamson & Rimdstidt (1994) Anthophyllite Palandri & Kharaka (2004) Pentlandite Williamson & Rimdstidt (1994) Pyrite Williamson & Rimdstidt (1994) SiO 2 (a) Rimstidt & Barnes (1980)

7 Example 1: Särkiniemi Waste Rock Pile (cont d) Drainage Chemistry: NPR = NAG ph = Solid Lines Model Dashed Lines - Data 7

8 Example 1: Särkiniemi Waste Rock Pile (cont d) Transport of oxygen: 8

9 Example 1: Särkiniemi Waste Rock Pile (cont d) Evolution of minerals: Primary minerals Secondary minerals 9

10 Example 1: Särkiniemi Waste Rock Pile (cont d) Long-term predictions and cover scenario: 10

11 Example 2: Kylylahti Waste Rock Pile 11 Mineral Mineral Content Surface Rate Coefficient, k Reference Area, A [wt%] [mol L -1 w ] b [m 2 L -1 w ] [mol m -2 s -1 ] Mg-biotite Nagy (1995) Phlogopite Nagy (1995) Pyrite c Williamson & Rimdstidt Tremolite (k 1 ) Palandri & Kharaka (2004) (k 2 ) Albite Palandri & Kharaka (2004) Calcite d (k 1 ) Plummer et al. (1978) (k 2 ) (k 3 ) Pyrrhotite c Williamson & Rimstidt (1994) Sphalerite c Domenech et al. (2002) Pentlandite c Williamson & Rimdstidt (1994) Dolomite d (k 1 ) Plummer et al. (1978) (k 2 ) (k 3 )

12 Example 2: Kylylahti Waste Rock Pile (cont d) Evolution of Drainage Chemistry: Blue Lines lower oxidation rate Orange Lines higher oxidation rate 12

13 Example 2: Kylylahti Waste Rock Pile (cont d) Gas transport Evolution of minerals 13

14 Example 2: Kylylahti Waste Rock Pile (cont d) Sensitivity Analysis: Reaction rate Recharge rate Gas diffusion Domain size 14

15 Example 2: Kylylahti Waste Rock Pile (cont d) Long-term simulation and cover scenario: 15

16 Example 3: Pyhäsalmi Tailings Facility Mineral Mineral Content Surface Area, A Rate Coefficient, k [wt%] [mol L -1 w ] b [m 2 L -1 w ] [mol L -1 w s -1 ] Reference Pyrite c Williamson an & Rimstidt (1994) Hornblende Palandri & Kharaka (2004) Tremolite (k 1 ) Palandri & Kharaka (2004) (k 2 ) Biotite Nagy (1995) Albite Palandri & Kharaka (2004) Serpentine Declercq & Oelkers (2014) K-feldspar Palandri & Kharaka (2004) Dolomite d (k 1 ) Plummer et al. (1978) (k 2 ) (k 3 ) Chlorite Palandri & Kharaka (2004) Calcite c (k 1 ) Plummer et al. (1978) (k 2 ) (k 3 ) Anthophyllite Palandri & Kharaka (2004) Phlogopite Nagy (1995) Pyrrhotite c Williamson an & Rimstidt (1994) Chalcopyrite c Williamson an & Rimstidt (1994) Sphalerite c Domenech et al. (2002)

17 Example 3: Pyhäsalmi Tailings Facility (cont d) Predicted Drainage Quality: measured seepage water ph = 2.87 NAG ph =

18 Example 3: Pyhäsalmi Tailings Facility (cont d) Transport of Gases: 18

19 Example 3: Pyhäsalmi Tailings Facility (cont d) Evolution of mineral phases: 19

20 Example 3: Pyhäsalmi Tailings Facility (cont d) Long-term prediction of drainage water quality: 20

21 Example 3: Pyhäsalmi Tailings Facility (cont d) Cover scenario: 21

22 Major Sources of Uncertainties Heterogeneity: a problem cannot be solved with more computational resources! Existing analogues can be used Stochastic concepts can be applied to estimate the likelihood Reaction parameters and microbial mechanisms Mineral reaction parameters should be obtained from the static and kinetic test results Seasonal fluctuations Water saturation and moisture profiles Retardation and sorption like processes Dimensionality 1D, 2D or 3D dominated system! Coupling of gas and heat transport 22

23 Conclusions 23 The proposed model seem to be capable of predicting the future key events Simulation outcomes correspond to the static test results and the measured drainage quality Predictive modelling can be performed in an early stage of a mine with limited data Prediction accuracy is only as good as the conceptual understanding about involved processes Objective of such modelling would be achieve the right order of magnitude instead of obtaining the exact values Effects of uncertainty and the model limitations should be properly acknowledged especially when such prediction results are used in decision making processes Predictive models should be continuously updated with the increasingly collected data during the operational phases

24 Thank you for your attention! Contact: Dr. Muhammad Muniruzzaman Research Scientist Geological Survey of Finland 24