Tailings and Mine Waste Conference

Selection of shear strength profile for desiccated tailings to support an upstream raise Williams, D.J.; Paterson, S.; Yau, R.; Goddard, D. Oct 31, 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
59368-Williams_D_et_al_Selection_shear_TMW_2015.pdf [ 1.6MB ]
Metadata
JSON: 59368-1.0314305.json
JSON-LD: 59368-1.0314305-ld.json
RDF/XML (Pretty): 59368-1.0314305-rdf.xml
RDF/JSON: 59368-1.0314305-rdf.json
Turtle: 59368-1.0314305-turtle.txt
N-Triples: 59368-1.0314305-rdf-ntriples.txt
Original Record: 59368-1.0314305-source.json
Full Text
59368-1.0314305-fulltext.txt
Citation
59368-1.0314305.ris

Full Text

Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015  Selection of shear strength profile for desiccated tailings to support an upstream raise D.J. Williams The University of Queensland, Australia S. Paterson, R. Yau SRK Consulting, Australia D. Goddard Cannington Mine, Australia ABSTRACT The shear strength profile of existing tailings is crucial for supporting an upstream embankment raise for a conventional surface slurried tailings storage both safely and with acceptable settlements. The existing tailings will need to be desiccated in order to provide adequate bearing capacity for the raise, and will require geotechnical investigation to demonstrate this. Typical geotechnical investigation methods include in situ vane shear testing and cone penetration testing (CPT), usually including pore pressure measurement and possibly including seismic measurements. The vane shear and CPT may use the same rig. Heavily desiccated tailings may allow direct access using a conventional rig. Otherwise, a rig with wide tracks may be required, or access roadways may have to be constructed over the tailings. The data from these tests are then interpreted to determine the shear strength profile for the desiccated tailings, to enable geotechnical slope stability and settlement analyses of the proposed upstream embankment raise. The paper describes a case study of the geotechnical investigation, test data interpretation and geotechnical design considerations for a proposed upstream embankment raise on heavily desiccated tailings. The focus is on test data interpretation, which is an area that is often deficient. Keywords: cone penetration test, desiccated tailings, shear strength, upstream raise, vane shear  1 INRODUCTION South32’s Cannington Underground Mine is located 250 km south-east of Mount Isa in northwest Queensland, Australia, and has produced silver, lead and zinc since 1997. The ore is ground to -180 micron, and processed to produce Sandy SILT to Silty SAND-sized tailings. The tailings are thickened in a high rate thickener and discharged at 65% solids by mass. Of the tailings produced by the processing of the ore, 60% is returned underground as cement paste backfill, and 40% reports to a surface tailings storage facility (TSF) comprising three cells. Tailings deposition into Cell 1 ceased about 4 years ago, and the embankment is proposed to be raised by the upstream method using borrow material located partially on heavily desiccated tailings to accommodate tailings from late 2017. The proposed upstream raise of Cell 1 is the optimal solution to cost-effectively and safely provide additional tailings storage capacity beyond the filling of Cell. This paper describes the geotechnical investigation, test data interpretation and geotechnical analyses for the proposed upstream Cell 1 embankment raise on heavily desiccated tailings, focusing on test data interpretation, which is an area that is often deficient. 2 DESCRIPTION OF SITE Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015  A description of the pertinent features of the Cannington Mine site is given in the following sections. 2.1 Surface tailings storage facility The surface TSF comprises three cells. Cell 1, with an area of 44 ha, and Cell 2, with an area of 44 ha, were constructed in the late 1990’s, and have each been raised a number of times to reach their current embankment heights. Tailings deposition into Cell 1 ceased about 4 years ago and deposition into Cell 2 ceased about 1 year ago. Cell 3, with an area of 58 ha, was completed in May 2014 and is currently the active cell. Cell 3 is expected to be filled by August to October 2017. A 20 September 2013 aerial view of the Cannington surface TSF, with Cell 2 still operational and Cell 3 under construction, is shown in Figure 1, and a view of the heavily desiccated surface of Cell 1 is shown in Figure 2.  Figure 1: Aerial view of Cannington’s surface TSF (Google Earth 20 September 2013). Figure 2: View of heavily desiccated surface of Cannington’s surface TSF Cell 1. 2.2 Site climatic setting and water balance Figure 3 shows yearly rainfall data since 1887 for Devoncourt Station (Bureau of Meteorology 2015), which is located halfway between Mount Isa and Cannington mines in Queensland, Australia. The mean yearly rainfall is about 400 mm, trending upwards over time. The 10th Cell 1 Cell 2 Cell 3 Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015  percentile and 90th percentile yearly rainfalls are about 160 mm and about 700 mm, respectively, representing a measure of the rainfall range from year to year.  Figure 3: Yearly rainfall data since 1887 for nearby Devoncourt Station. Figure 4 shows that monthly rainfall is concentrated between December and April for Devoncourt Station. Figure 4 also shows the 10th and 90th percentile and highest recorded monthly rainfalls, which show a tremendous range. The 10th percentile monthly rainfalls are minimal, at up to 12 mm, the 90th percentile monthly rainfalls range from 10 to 225 mm, and the highest recorded monthly rainfalls range from 40 to almost 700 mm, the latter equal to the 90th percentile yearly rainfall. Figure 5 shows the mean and highest daily rainfalls for Devoncourt Station. The mean daily rainfall ranges from 3.5 to 100 mm, while the highest recorded daily rainfall ranges from 30 to 215 mm. The current Cell 3 decant pumping capacity is about 240 m3/hr (or 10 mm/day over the 58 ha surface area of Cell 3), of which 100 m3/hr on average is required to meet the production of supernatant (or 4 mm/day over the 58 ha surface area, leaving 6 mm/day available to handle rainfall runoff). This decant pumping capacity is dwarfed by the maximum mean and highest daily rainfall totals of 100 mm and 215 mm, respectively, which would require 17 to 36 days to pump out with continued tailings deposition. The spillways are designed for the 1 in 10,000 year Annual Return Interval (ARI) rainfall event, which is a similar to the 1 in 1,000 year ARI rainfall event plus wave run-up, and approaches the Probable Maximum Flood (PMF). 2.3 Rationale for upstream raising of Cell 1 The rationale for the expansion of the TSF by the upstream raising of Cell 1 is that this would take advantage of the desiccation-induced strengthening of the tailings in Cell 1 during the 4 years since the last deposition of tailings. Further, it would only require the three perimeter walls of Cell 1 to be raised by the required 3 m to RL 272.0 m, since the wall adjoining Cell 2 is already at RL 272.0 m. Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015     Figure 4: Monthly rainfall data for nearby Devoncourt Station.  Figure 5: Daily rainfall data for nearby Devoncourt Station.  Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015    2.4 Risk level and stability of upstream raise of Cell 1 The factors affecting the risk level and stability of the proposed upstream raise of Cell 1 of Cannington’s surface TSF include the shear strength of the tailings and the level of the phreatic surface within the perimeter embankments, piping and erosion potential, earthquake-induced loading of the perimeter embankments, construction pore water pressures, and seepage through the perimeter embankments. 2.5 Borrow materials and compaction The upstream raise of the Cell 1 embankment is to be constructed using available borrow materials located to the north-east of the TSF, including alluvium and residual soil. About 107,000 m3 of clayey fill is required, plus about 24,500 m3 of rock fill from the underground mine as a source of rip rap, if this is required to protect the faces of the raise. Some of the available clayey soils have been assessed as potentially dispersive, giving the raise a potential to pipe and erode, although the testing was carried out using deionised water, representing a worst case. To address the potential risk of piping, it is proposed to keep supernatant water (and high hydraulic gradients) away from the upstream face of the raise. To address the potential risk of erosion of the upstream face of the raise during periods of prolonged heavy rainfall, leading to the ponding of flood waters against the upstream face, it is proposed to place an HDPE geomembrane against the face, as was done for the Cell 3 embankment. The decant is proposed to be located against the internal embankment separating Cells 1 and 2. A geotextile separation layer is proposed between the desiccated tailings and the Cell 1 upstream raise, serving to allow the pore water pressure-induced upward flow of tailings water to pass into the raise, while limiting the mixing of the underlying tailings and the placed borrow material. The 3.0 m raise will induce excess pore water pressures of up to 60 kPa, which could cause a rise in the perched water table (currently at an average depth of about 4.6 m below the tailings surface) within the desiccated tailings by up to 6 m, taking it well up into the raise. However, these will dissipate over time and result in a strengthening of the tailings. Field compaction of clayey fill is typically specified in terms of laboratory Standard compaction testing. However, this has in the past proved problematic at Cannington due to the low natural gravimetric moisture content of the available clayey borrow material and the difficulty experienced in wetting it up to the moisture content specified in terms of laboratory Standard compaction testing. An alternative is to specify field compaction in terms of laboratory Modified compaction testing, which achieves its higher Maximum Dry Density (MDD) at a lower Optimum Moisture Content (OMC), closer to the natural gravimetric moisture content of the available clayey borrow materials and largely avoiding the need to moisture condition this material. The (approximately 4.5 times) higher compactive effort applied in laboratory Modified compaction testing, compared with laboratory Standard compaction testing, results in a higher compacted dry density. This ensures that both compaction methods compact the material to similar degrees of saturation. Previous experience has been that the dry density specified based on laboratory Standard compaction testing is generally exceeded in the field, indicating that the same field compaction equipment should be capable of achieving the dry density specified based on laboratory Modified compaction testing. The appropriate compaction specifications are -1 to +3% of OMC for laboratory Modified compaction testing, and a minimum field dry density of 95% of Modified MDD. There is no advantage in using decant water to wet-up the available clayey borrow material to the required moisture content, since this is recycled to the plant, and would otherwise have to be made up using Great Artesian Basin (GAB) water, and there is a reluctance to use decant water outside the TSF.  Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015   3 GEOTECHNICAL INVESTIGATION AND INTERPRETATION The geotechnical investigations reported by SRK Consulting (2015) included Dynamic Cone Penetration Tests carried out in the desiccated Cell 1 and 2 tailings, which highlighted the far greater desiccation of the Cell 1 tailings that had been left to desiccate about 4 times as long as the Cell 2 tailings since the last deposition of tailings. Twelve Cone Penetration Tests with pore water pressure measurement (CPTu) and in situ vane shear testing were carried out in the heavily desiccated Cell 1 tailings. Laboratory characterisation and geotechnical parameter testing was carried out on tailings and borrow samples. 3.1 Interpretation of CPTu and vane shear results The consolidation state of a soil may simply be interpreted using the method of Schmertmann (1978), which was applied to all 12 CPTu cone resistance profiles with depth in the heavily desiccated Cell 1 tailings. The Schmertmann (1978) method involves extrapolating lines through the lower bounds of the cone resistance profiles to the surface. The lower bound line intersecting the surface to the right of the origin indicates under-consolidated tailings, passing through the origin indicates normally-consolidated tailings, and intersecting the surface to the right of the origin indicates over-consolidated (usually desiccated) tailings. In addition, the depth to the (presumed) perched water table within the tailings was estimated based on the measured pore water pressures. A typical interpretation is shown in Figure 6. The upper layer of the tailings in Cell 1 is seen in Figure 6 to have become heavily over-consolidated due to 4 years of desiccation since tailings deposition ceased. The layer below that is generally somewhat desiccated (over-consolidated) due to intermittent surface desiccation followed by re-wetting by fresh tailings, or normally-consolidated (maintained under water), and occasionally under-consolidated (unable to drain the self-weight-induced excess pore water pressures). The layer of tailings above the foundation is drained or desiccated. The spikes in cone resistance profile indicate sandy lenses. The range and average interpreted consolidation states and estimated perched water table depths estimated from the 12 CPTu profiles in Cell 1 tailings are summarised in Table 1. Based on CPTu dissipation testing, the horizontal coefficient of consolidation of the Cell 1 tailings was estimated to be in the range from 1,000 to 5,000 m2/yr, and the saturated hydraulic conductivity of the Cell 1 tailings was estimated to range from 0.3 x 10-7 to 1 x 10-6 m/s, with a median value of 3 x 10-7 m/s (SRK Consulting 2015). 3.2 Vane shear results and correlation with CPTu cone resistance Vane shear strength testing carried out in the Cell 1 tailings gave the data shown in Figure 7 (SRK Consulting 2015). There is a lack of vane shear strength data in the upper 4 m constituting the desiccated crust since the crust was too stiff to test. Correlating the vane shear strengths Su with the corresponding CPTu cone resistance values Qc gives the bearing capacity factor Nc : Nc = Qc / Su (1)  Values for the peak and remoulded bearing capacity factors are plotted in Figure 8. The reported vane shear strength data and the calculated bearing capacity factors are very scattered. The range of peak bearing capacity factors is 7.8 to 24.0, with an average value of 14.4, while the range of remoulded bearing capacity factors is 53 to 160, with an average value of 88.6. Conservative (lower bound) values of the peak vane shear strength are equivalent to 0.25 times the vertical effective stress, and conservative values of the remoulded vane shear strength are equivalent to 0.04 times the vertical effective stress. The average ratio of the remoulded to the peak vane shear strengths of 0.16 is low compared with ratios expected for tailings of about 0.5 below the water table and about 0.33 above the water table (Williams 2005). The remoulded Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015  vane shear strength is representative literally of the shear strength that applies on remoulding, such as that caused by loading the tailings so much and so rapidly as to cause “bow-wave” failure.   Figure 6: Typical interpreted consolidation state and depth to water table for a CPTu in heavily desiccated Cell 1 tailings.  Table 1: Interpreted consolidation states and estimated perched water table depths from CPTu profiles in Cell 1 tailings. ________________________________________________________________________________________________________    Layer description     Layer thickness (m)    Depth to perched water table (m) ________________________________________________________________________________________________________ Range  Heavily desiccated      1.8 to 4.8          2.6 to 7.0 Somewhat desiccated     0.0 to 7.5 Under-consolidated     0.0 to 5.4 Normally-consolidated    0.0 to 10.6 Drained/desiccated      0.0 to 3.8 Foundation        from 7.8 to 16.4 Average  Heavily desiccated       2.8            4.6 Somewhat desiccated      2.5 Under-consolidated      0.4 Normally-consolidated     3.9 Drained/desiccated       2.1 Foundation         from 12.0 ________________________________________________________________________________________________________ Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015    Figure 7: Peak and remoulded vane shear strengths with depth in Cell 1 tailings  Figure 8: Calculated peak and remoulded bearing capacity factors with depth in Cell 1 tailings Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015   Skempton and Henkel (1953) were the first to suggest an empirical relationship between undrained shear strength Su and vertical effective stress v’. For normally-consolidated (nc) soils (in the context of tailings, those always stored below water): (Su / v’)nc = 0.11 + 0.37 (IP/100) (2)  where IP is the Plasticity Index. For over-consolidated (oc) soils (in the context of tailings, those subjected to desiccation drying followed by re-wetting): (Su / v’)oc = (su / v’)nc .OCRm (3)  where OCR is the over-consolidation ratio and m is an empirical exponent, generally taken as 0.8. Inputting the values of the parameters for Cannington Cell 1 tailings, Equation (2) gives (su / v’)nc = 0.17, which is low compared with the value typical for normal soils of 0.25. It will be subsequently be shown that a value for (su / v’)nc of 0.25 is consistent with the CPTu and vane shear data, and can be adopted. The upper Cannington Cell 1 tailings are clearly over-consolidated due to their heavy desiccation, leading to a higher value for (Su / v’)oc. All 12 CPTu cone resistance profiles are plotted in Figure 9, together with the average profile. In Figure 10, the CPTu cone resistance data have been multiplied by the average peak bearing capacity factor of 14.4 (from Table 1) to provide the estimated profiles of peak shear strength with depth, again including the average profile, plus the “smoothed” lower bound of the average profile, ignoring the peaks caused by the sandy lenses.  Figure 9: CPTu cone resistance versus depth profiles. Also shown in Figure 10 are the approximate normally-consolidated, self-weight profile and the approximate over-consolidated profile, the latter made to match the smoothed average profile below the water table (from about 4.6 m depth). The average OCR is approximately 3.15 and, Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015  using Equation (3), (su / v’)oc ~ 0.63, 2.5 times the adopted normally-consolidated value of 0.25. The remoulded values are approximately 0.4 times these values, but will only be realised if remoulding is initiated.  Figure 10: Estimated peak shear strength versus depth profiles. In Figure 11, the smoothed average and self-weight profiles have been matched to the factored CPTu cone resistance versus depth profile, to obtain and estimate the average remoulded bearing capacity factor of 36 (rather than the average value of 88.6 given in Table 1), and a more reasonable average ratio of remoulded to peak vane shear strength of 0.4. 3.3 Laboratory characterisation and geotechnical parameter testing The Cannington tailings particle size distribution is Silty SAND to Sandy SILT-sized, with 40 to 70% sand-size (0.06 to 1.5 mm), 37 to 63% silt-size (0.002 to 0.06 mm), and 3 to 7% clay-size (<0.002 mm). The tailings are generally non-plastic, with a Liquid Limit of 14 to 16%. The specific gravity of the tailings is about 3.15, and the near surface settled and desiccated dry density is about 1.76 t/m3. The drained shear strength parameters reported for the tailings were an effective cohesion of 0 kPa, and an effective friction angle of 39o (SRK Consulting 2015). The available borrow material typically has the particle size distribution of Gravelly, Clayey, Sandy SILT, with <10% gravel-size (>2 mm), 10 to 35% clay-size, 15 to 35% sand-size, and 35 to 50% silt-size. The fines classify as intermediate to high plasticity. The specific gravity is generally 2.60 to 2.65, and the compacted dry density was generally 1.80.1 t/m3, corresponding to a void ratio of 0.4 to 0.45. Limited consolidated undrained and unconsolidated undrained triaxial testing of compacted borrow material suggested an effective cohesion of 12 kPa, and an effective friction angle of 22o. The drained shear strength parameters of the compacted borrow material were an effective cohesion of 0 kPa, and an effective friction angle of 29 to 31o. Hydraulic conductivity tests on compacted borrow specimens indicated values ranging from 0.073 to 1.9 x 10-9 m/s, with an average value of 0.5 x 10-9 m/s.  Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015    Figure 11: Estimated remoulded shear strength versus depth profiles. 4 GEOTECHNICAL DESIGN CONSIDERATIONS The geotechnical design considerations for the proposed upstream raise of Cell 1 of Cannington’s surface TSF include the shear strength of the tailings and the level of the phreatic surface within the perimeter embankments, piping and erosion potential, earthquake-induced loading of the perimeter embankments, construction pore water pressures, and seepage through the perimeter embankments. The discussion in Section 3.1 indicates that the heavily desiccated Cell 1 tailings provide a more than adequate shear strength to support the proposed 3 m raise. The available shear strength and low phreatic surface ensure adequate geotechnical stability of the proposed raise. The calculated factors of safety satisfy the ANCOLD (2012) minimum factor of safety of 1.3. Consolidation settlement of the average 12 m depth of tailings underlying the raise is estimated to be between 150 mm and 250 mm (1.25 to 2.1% of the depth of tailings), the range depending on the level of the perched water table. It is noted that any settlement of the raise itself would be minor and would likely occur largely during construction and not be seen. Earthquake-induced liquefaction is dependent on four conditions being met (Williams, 1988): 1. a sufficient earthquake-induced peak ground acceleration, generally above 0.13 g; 2. silty to fine-grained sand-size; 3. near to full saturation; and 4. a loose consistency. The maximum design peak ground acceleration estimated for the Cannington site is only 0.049 g, much lower than would normally be required to induce liquefaction of susceptible Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015  materials. Hence, earthquake-induced liquefaction of the Cannington tailings is extremely unlikely, as confirmed by SRK Consulting (2015). While the particle size distribution of the Cannington tailings is potentially liquefiable, they are unsaturated and relatively dense due to desiccation. Further, the deeper the water table within the tailings, the greater the overburden stress available to resist possible liquefaction. Construction of the Cell 1 upstream raise will induce excess pore water pressures in the underlying tailings, which may impact geotechnical stability in the short-term. It is therefore essential that the raise be constructed progressively using appropriately-sized equipment, over the entire length of the embankment, to allow the dissipation of these pore water pressures. Their dissipation will increase the shear strength of the tailings, enabling further embankment fill to safely be supported. Seepage through the perimeter embankments of the raised Cell 1 of Cannington’s surface TSF is not expected to increase as a result of the raise. Embankment seepage is known to be driven by tailings water during operations, not by wet season rainfall, although the last 3 years have been unseasonably dry and little seepage has occurred. Seepage control during and beyond operations is expected to include a seepage interception trench around the toe of the perimeter embankments, connected to sumps at low points, and a seepage monitoring plan. 5 CONCLUSION A case study has been described in which heavily desiccated tailings are shown to provide adequate bearing capacity to cost-effectively and safely support a proposed 3 m high upstream embankment raise on a conventional surface slurried tailings storage facility. Data from cone penetration and vane shear testing of the tailings were interpreted to determine their shear strength profile. The geotechnical design considerations for the proposed upstream embankment raise on heavily desiccated tailings were then discussed. The key message of the paper is the importance of undertaking careful test data interpretation to ensure the reliable design and construction of an upstream raise partially supported on desiccated tailings. 6 ACKNOWLEDGEMENT The management of South32’s Cannington Mine is acknowledged for providing access and granting approval to publish the tailings data contained in this paper. 7 REFERENCES ANCOLD. 2012. Guidelines on Tailings Dams – Planning, Design, Construction, Operation and Closure. Bureau of Meteorology. 2015. Available from <http://www.bom.gov.au> [Accessed on 6 August 2015]. Google Earth. 20 September 2013. Available from <https://www.google.com/earth/> [Accessed on 6 August 2015]. Schmertmann, J.H. 1978. Guidelines for cone penetration test, performance and design. Federal Highway Administration, Report FHWA-TS-78-209, Washington, July 1978. Skempton, A.W. and Henkel, D.J. 1953. The post-glacial clays of the Thames estuary at Tilbury and Shellhaven. Proceedings of the 3rd International Conference on Soil Mechanics and Foundation Engineering, Zurich, Switzerland, 16-27 August 1953, I, pp. 302-308. SRK Consulting. 2015. Cannington TSFE2 Cell 1 Tailings and Embankment Investigation – Factual and Interpretive Report. May 2015. Williams, D.J. 1988. Potential engineering risks in the earthquake hazard to the east coast of Queensland. IEAust, Civil Engineering Transactions, CE30/5, pp. 307-317. Proceedings Tailings and Mine Waste 2015 Vancouver, BC, October 26 to 28, 2015  Williams, D.J. 2005. Chapter 17: Placing covers on soft tailings. In: Ground Improvement-Case Histories, pp. 491-512, Eds B. Indraratna and Chu Jian. (Elsevier: Oxford). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.59368.1-0314305/manifest

Comment

Related Items