- Email: [email protected]
Determination of consolidation behaviour of clay slurries
International Journal of Mining Science and Technology xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Mining Science and Technology journal homepage: www.elsevier.com/locate/ijmst
Determination of consolidation behaviour of clay slurries Khan Faseel S., Azam Shahid ⇑ Environmental Systems Engineering, University of Regina, Regina S4S0A2, Canada
a r t i c l e
i n f o
Article history: Received 2 July 2015 Received in revised form 9 August 2015 Accepted 15 October 2015 Available online xxxx Keywords: Clay slurry Consolidation Conventional oedometer test Bench-top centrifuge
a b s t r a c t The main objective of this study was to determine the consolidation behaviour of clay slurries. A finegrained clay with high consistency limits (wL = 180%, wP = 120%) was investigated using conventional oedometer and bench-top centrifuge tests. Results indicated that the slurry had an apparent preconsolidation (due to initial conditions, electrochemical interactions, tortuous drainage, and thixotropic strength) from e = 5.7 to e = 5.5 followed by virgin compression. Likewise, the low hydraulic conductivity (1010–1012 m/s) was due to low porosity (small pore throats) and high tortuosity (long flow paths). Unlike consolidation of soils, the cv and mv decreased with increasing r0 but increased with increasing e and k. The data from the two tests correlated well in the range of r0 = 10–65 kPa, e = 5.5–3.86, k = 1.7 1010–5 1011 m/s, Fc = 1–40 MN. New equations were developed to correlate the consolidation parameters (e, r0 , k) with Fc. The deviation of k beyond 40 MN (e = 4.65) was due to deviation from the initial straight line portion of the settlement curve in the centrifuge test. Ó 2016 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
1. Introduction Several mining operations generate waste tailings that contain variable amounts of clay fraction thereby affecting the consolidation behaviour of such water-rich slurries [1]. The consolidation properties are important for slurry containment facilities and when the slurries are used as backfills [2]. The water release rate and amount is governed by the following factors: (i) self weight due to gravity and applied loading due to upper layer deposition [3–5]; (ii) initial and boundary conditions including the prevalent climate, side and/or bottom drainage and containment geometry [6–8]; (iii) electrochemical interactions between the clay and the pore fluid [9–11]; and (iv) microstructure and thixotropy derived from the above [12–14]. The determination of consolidation properties is based on volume compressibility and hydraulic conductivity relationships obtained from laboratory tests. Table 1 provides a summary of the various types of consolidation tests used for clay slurries. The first three tests require long time to complete but directly determine the constitutive relationships. In the conventional oedometer test, the hydraulic conductivity measurement can cause consolidation in the low effective stress range if the seepage force is greater than the applied stress. Likewise, the slurry consolidation test uses large samples thereby affecting the spatial variability in material properties. Similarly, the seepage induced test uses a complex equipment setup but is ⇑ Corresponding author. Tel.: +1 306 3372369. E-mail address: [email protected] (S. Azam).
useful in the low effective stress range. Although the continuous loading test and the controlled gradient test require less time, these tests are based on the conventional consolidation theory thereby limiting these tests to small strain applications. Finally, the bench-top centrifuge test generates consolidation data similar to the conventional oedometer test, the seepage induced test, and the continuous loading test [32,33]. There is a need to understand the results of the bench-top centrifuge test (requires the least time) in conjunction with the conventional oedometer test (well established test). The main objective of this study was to determine the consolidation behaviour of clay slurries using laboratory tests. Initially, the geotechnical index properties were determined for preliminary soil assessment. Next, the conventional oedometer test and the bench-top centrifuge test were conducted. Finally, the test results were compared to convert the former test data to the later test data thereby determining the volume compressibility and hydraulic conductivity relationships.
2. Test methods The clay was retrieved from a local geological deposit of high plasticity. The samples were collected in situ, secured in sealed plastic bags, and transported to the Geotechnical Testing Laboratory at the University of Regina where these were stored at a temperature of 25 °C. The geotechnical index properties of the active clay were determined according to standard ASTM test
http://dx.doi.org/10.1016/j.ijmst.2015.12.014 2095-2686/Ó 2016 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
2
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
Table 1 Summary of consolidation tests. Test type and reference Conventional oedometer test [15–18]
Slurry consolidation test [9,19–23] Seepage induced test [16,19,24,25]
Continuous loading tests [19,26–29]
Controlled gradient test [30,31]
Bench-top centrifuge test [32–37]
Test equipment and analysis procedure
Limitations
D = 60–63 mm, H = 20–25 mm e–r0 : Measured by incremental loading and data analyzed using SSC theory or LSC theory k: Directly measured using constant/falling head method or indirectly determined using square root of time method D = 40–495 mm, H = 150–1800 mm e–r0 : Measured by incremental loading and data analyzed using LSC theory k: Directly measured or indirectly determined by inversion of LSC theory D = 50–150 mm, H = 20–40 mm e–r’: Measured by application of hydraulic gradient and data analysis does not require any theory k: Directly measured using constant head method D = 60–254 mm, H = 20–1000 mm e–r0 : Measured by constant rate of stress or strain and data analyzed by inversion of SSC theory or LSC theory with a constant cv k: Indirectly determined by inversion of SSC theory or LSC theory D = 25 mm, H = 63 mm e-r’: Measured by controlled gradient by continuously adjusting loading rate and data analyzed by inversion of SSC theory k: Indirectly determined by inversion of SSC theory D = 26–58 mm, H = 20–125 mm Centrifuge maximum radius: 300–350 mm, rpm: 300–3000 e–r0 : Indirectly determined and data analyzed using scaling laws, LSC models or comparison of results with a different test type k: Indirectly determined from initial straight line portion of settling curve
Long test duration Direct determination of k may consolidation Scatter in consolidation constants
cause
Long test duration if k is directly measured Spatial variation in material properties in large samples Long test duration Requires sophisticated instrumentation Not applicable for high r0 range e altered due to sample rebound Compressibility and hydraulic conductivity cannot be validated Theoretical limitations render it unsuitable for large strains Loading rate is unknown at the start of the test Theoretical limitations render it unsuitable for large strains Particle segregation is more likely to occur No standard procedure to convert centrifuge data to consolidation data
Note: e–r0 – compressibility; e – void ratio; r0 – effective stress; k – hydraulic conductivity; SSC – small strain consolidation; LSC – large strain consolidation; D – diameter; and H – height.
methods as follows: (a) specific gravity (Gs) by the Standard Test Method for Specific Gravity of Soil Solids by Water Pycnometer (ASTM D854(10)); (b) liquid limit (wl), plastic limit (wp) and plasticity index (Ip) by the Standard Test Method for Liquid Limit, Plastic Limit, and Plasticity Index of Soils (ASTM D4318(10)); and (c) soil classification by the Standard Practice for Classification of Soils for Engineering Purposes (Unified Soil Classification System) (ASTM D2487(11)). Fig. 1 gives the schematic of the conventional oedometer (Model S-450). The test was conducted as per the ASTM Standard Test Methods for One-Dimensional Consolidation Properties of Soils Using Incremental Loading (D2435). The clay (up to 20 mm aggregates) was soaked in tap water and was frequently stirred with a steel rod to ensure homogeneous water distribution. The resulting slurry, at a water content (w = wl), was poured in the oedometer ring (63.5 mm diameter and 25.4 mm height). Presaturated porous stones above and below the sample allowed upward drainage, the geotextile precluded porous stone clogging, and the outer casing prevented surface tilting. The loading frame was lowered (touching the steel ball on the loading head to minimize eccentricity) and was fixed in-place using clamps. The combined stress of the loading head, porous stone, geotextile and steel ball measured 1.5 kPa. Subsequent effective stress (r0 ) increments were pneumatically applied using an air compressor and were maintained through a digitally controlled pressure regulator. A linear vertical displacement transducer (LVDT) was used to record deformations at specified time intervals. A 20 mm water head was maintained on top of the sample to ensure complete saturation throughout the test. The hydraulic conductivity (k, m/s) after each r0 increment was measured using the falling head method. Upward drainage was allowed (to prevent consolidation) using the hydraulic gradient between a bottom injection tube and the water level in the oedometer ring [38]. The hydraulic gradient was kept less than the critical hydraulic gradient to preclude internal erosion. Several k measurements were made to ensure a reliable value at a given void ratio (e). The test was stopped when two consecutive readings showed no significant change. The k was calculated from knowl-
edge of the internal standpipe area (a, mm2), sample length (L, mm), sample cross-sectional area (A, mm2), elapsed time (Dt, s), initial water head (h1, mm) and final water head (h2, mm), according to the following version of Darcy’s law:
k ¼ 0:23
aL h1 log10 A Dt h2
ð1Þ
Fig. 2 shows schematic of the bench-top centrifuge test. This setup has been recently used at the University of Regina for oil sand fine tailings [39,40]. The digitally controlled unit (Sorvall Thermo Scientific Biofuge Primo R) comprised of a swinging bucket rotor with four buckets and had a maximum radius of 127.4 mm as shown in elevation (Fig. 2a) and plan view (Fig. 2c). Graduated acrylic tubes (24.2 mm internal diameter and 106 mm height) were filled with the slurry (prepared as before) and placed in ceramic tube-holding adapters in two separate buckets (Fig. 2b). The slurry was filled up to the 40 mm mark which gave a ratio of sample height to centrifuge radius of 1:3, as suggested by McDermott and King [36]. The test was conducted at a constant temperature of 20 °C and an angular velocity of 4000 r/min. The centrifuge was stopped at specified time intervals and the interface height of the sediment was noted (Fig. 2d). The measurements were taken from both tubes and the average value was recorded. Using Ht (mm) for the interface height, Ho (mm) for the initial interface height and eo for the initial void ratio, the e was calculated using the following equation [33]:
e¼
Ht ð1 þ eo Þ 1 Ho
ð2Þ
Likewise, the centrifugal force, Fc (MN) was computed using the following equation [32]:
F c ¼ mr x2
ð3Þ
where m is mass (kg) of the slurry, r is radius (m) that is the centroid of the slurry mass and w = (rpm/60)2p is the angular velocity (1/s). The centrifugal force was applied to the specimen in 1 s and
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
3
Loading frame
Leads to readout unit
Moveable frame
Linear variable differential transducer
Head
Moveable frame
Standpipe
Loading frame
Porous stone and geotextile
Sample
Base plate
Low Low pressure regulator
Leads to readout unit
Steel ring
Loading head Outer casing
Loading frame
Steel ball
High High pressure regulator Unload
Load
Fig. 1. Schematic of the conventional oedometer test.
Transparent cover Axis of rotation Tube cap During flight
At rest
(a)
Ceramic adapter
(b) Tube cap
Plastic bucket
Plastic bucket
Plastic bucket
Released water
Sediment height (Ht)
Graduated acrylic tube
Plastic bucket
Protection walls
(c)
Graduated acrylic tube
At rest
Plastic bucket
Motor
Plastic bucket
. Initial height (Ho)
(d)
Fig. 2. Schematic of the centrifuge test: (a) side view; (b) section of the plastic bucket; (c) plan view; and (d) close up of the graduated tube.
the total centrifugal force for a specific duration was obtained as the product of Fc and the number of seconds per loading level. The k was calculated from the settlement curves according to the following equation [37]:
k¼S
cw
Nc0
ð4Þ
where S is the settlement rate, cw is the unit weight of water, c0 is the submerged unit weight, N is the acceleration in terms of
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
4
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
multiples of 9.81 m/s2 (gravitational acceleration) and was taken at a distance of 0.59 times the sample height from the base [36]. The following equation was used for calculating the c0 at a given Ht:
6
c0 ¼
4
Gs qw e g cw 1þ Gs 1þe
5
Void ratio
ð5Þ
3 2
3. Results and discussion
Test data
1 R2 = 0.98
0 0 10
101 102 103 Effective stress (kPa)
10-9 Test data R 2 = 0.99 10-10
10-11
10-12
0
Soil property
Value
Specific gravity, Gs Material finer than 0.075 mm (%) Material finer than 0.002 mm (%) Liquid limit, wL (%) Plastic limit, wP (%) Plasticity Index, wPI (%) USCS symbol
2.85 97 72 180 60 120 CH
1.5 kPa
Vertical deformation (mm)
0
15 kPa
35 kPa
66 kPa
2
3 4 Void ratio
5
6
and inaccessible pores in the clay fabric; and (iv) thixotropic strength which is highest at wl [44–47]. Likewise, the steep portion of this curve is governed by intrinsic material properties (specific gravity, grain size distribution, and clay mineralogy) and the rate of expulsion of water [46]. The test data best fitted the following Wiebull function:
ð6Þ
where A, B, E and F are dimensionless constants having values of 5.7, 5.45, 20 and 0.69 respectively. These values are similar to oil sand fine tailings [13]. Fig. 5 gives the hydraulic conductivity relationship (k–e). The k was not measured at the end of the test (1440 kPa) because of time constraint. The k varied by two orders of magnitude from 1010 to 1012 m/s when the void ratio decreased from 5.7 to 0.92. In this e range, similar values of k have been reported for phosphatic clays and for oil sand fine tailings [13,43]. The low k values are attributed to low porosity (small pore throats) and high tortuosity (long flow paths) of clay slurries. Furthermore, the data scatter at high void ratios may be attributed to possible occluded air in the specimen [48]. The k–e relationship was a typical power function as follows:
k ¼ CeD
110 kPa 5
1
Fig. 5. Hydraulic conductivity in the conventional oedometer test.
e ¼ A B expðEr0F Þ Table 2 Summary of geotechnical index properties of the investigated clay.
104
Fig. 4. Volume compressibility in the conventional oedometer test.
Hydraulic conductivity (m/s)
Table 2 gives the geotechnical index properties of the clay. The Gs was 2.85 which is typical of sedimentary clays (that is, between 2.4 and 2.95 [41]. Material finer than 0.075 mm measured 97% along with a clay size fraction (material finer than 0.002 mm) of 72% indicating the fine grained nature of the investigated soil. The high consistency limits (wL = 180%, wP = 120%) depicted high water adsorption and retention characteristics. Overall, the material was classified as highly plastic clay (CH). Fig. 3 gives the vertical deformations from the conventional oedometer test. The slurry did not settle under the seating stress of 1.5 kPa over a period of 4 days (5760 min). Subsequent r0 increments took 5 days to 19 days for the sample to enter the secondary consolidation stage which was an average deformation rate of 1/50 mm a day (secondary compression index (cae = 0.15). The vertical deformations under each r0 were found to be in the range of 1–3 mm such that the cumulative settling at the end of the test was 18 mm (70% strain). Fig. 4 gives the volume compressibility relationship (r0 –e). The e decreased from 5.7 at r0 = 1.5 kPa to 0.92 at r0 = 1440 kPa. Such a large variation in e produced a non-linear volume compressibility curve and consisted of two distinct parts: an initial flat portion (from e = 5.7 to e = 5.5) pertaining to apparent pre-consolidation followed by a steep portion (from e = 5.5 to e = 0.92) corresponding to virgin compression. Apparent pre-consolidation has been observed in other fine grained slurries such as dredged materials, phosphatic clays and oil sand tailings [13,42,43]. This phenomenon should be attributed to the following reasons: (i) initial slurry conditions such as void ratio and water content; (ii) electrochemical interactions in the colloid–liquid mixture; (iii) tortuous flow paths
ð7Þ
180 kPa 360 kPa
10
720 kPa 1440 kPa
15
20 0
50 100 Time (×103 min)
150
Fig. 3. Vertical deformation versus time in the conventional oedometer test.
where C is a constant with units of velocity (m/s) and D is a dimensionless constant. The values of C and D being 1.12 1012 m/s and 2.74, respectively. Once again, the parameters C and D were similar to phosphatic clays [43]. Fig. 6 plots the coefficient of consolidation (cv) and the coefficient of volume change (mv) with respect to the above mentioned measured properties (r0 , e, k). The cv, which governs the rate of consolidation, was calculated using the square root of time method. Likewise mv, which is the inverse of modulus of elasticity, was calculated from the slope of the virgin compression curve
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
5
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
Fig. 6. Consolidation parameters from the conventional oedometer test.
5
high eo. This was proceded by a gradual decrease in slope such that it became nearly horizontal after 6000 min. This is attributed to a reduced pore size (reduced e) due to expulsion of water through consolidation. The test data were fitted using a 3rd order polynomial equation that was differentiated to determine S and subsequently used in calculating k according to Eq. (4). The estimated k decreased from 1010 to 1012 m/s with decreasing interface height. Fig. 8 gives the volume compressibility relationship (e–Fc) from the bench-top centrifuge test. The e was calculated from Eq. (2) and the total Fc was calculated from Eq. (3). The initial Fc was arbitrarily assigned a value of 1 MN. The e decreased from 5.50 to 3.86 and the total Fc reached 162.8 MN. A non linear curve was obtained with two distinct portions; an initial flat portion (at e = 5.5) and a lower steep portion (e = 5.5–3.86). This behaviour is similar to that determined through the consolidation test that showed apparent preconsolidation followed by virgin compression. Once again, a Weibull function best fitted the volume compressibility data and was found to be of the following form:
4
e ¼ A B expðEF Fc Þ
40
Interface height (mm)
38 36 34 10 -11
32 30
Test data
28
3rd order polynomial R 2 = 0.99 S = - 0.004 +12.3 × 10 -7t - 9.2×10 -11 t 2
26
0
2000
4000 Time t (min)
6000
Hydraulic conductivity k (m/s)
10 -10
10-12 8000
Fig. 7. Interface height versus time in the centrifuge test.
Void ratio
6
ð8Þ
The above equation is similar to Eq. (6) with r replaced with Fc such that the values of the constants are A = 5.5, B = 5.0, E = 6.0 and F = 0.33. Fig. 9 shows the correlation of conventional oedometer test data with the centrifuge test data. Fig. 9a plots the applied centrifugal force (Fig. 8) versus the measured effective stress (Fig. 4) to correlate the two tests. The comparison between the two tests was made in the void ratio range of 5.7–3.86. The data best fitted the following power function: 0
3 2 1 0 -1 10
Test data e = 5.5 - 5exp (-6 Fc -0.18 ) R 2 = 0.94
10 0 10 1 10 2 Total centrifugal force (MN)
10 3
Fig. 8. Void ratio versus total centrifugal force in the centrifuge test.
F c ¼ 0:028r02:6 2
2
(Fig. 4). The cv ranged from 0.007 m to 0.077 m per year which is similar to active clays (0.019–0.09 m2 per year) [46]. Likewise, the mv varied from 0.002 m2/kN to 0.00007 m2/kN which is similar to dredged slurries [49]. As expected, both cv and mv decreased with increasing r0 but increased with increasing e and k. Power functions were used to fit the relationships and were found to correlate well with the test data as indicated by the corresponding coefficients of determination (0.99 P R2 P 0.73). The power functions demonstrate that the small strain consolidation theory does not apply to the dewatering behaviour of clay slurries. Fig. 7 gives the settlement curve from the bench-top centrifuge test. The interface height decreased from 40 mm to 30 mm (25% strain) in 6000 min (4 days). Initially, the interface decreased following a steep slope (S) due to the high initial k corresponding to a
ð9Þ
The above equation captured the large changes in void ratio of the slurry as opposed to a linear relationship for soils [32]. Fig. 9b plots Fc (derived from Eq. (9)) versus k (from both tests). The k values measured in the range of Fc = 1–40 MN correlated well with k from the conventional oedometer test according to the following exponential equation:
F c ¼ 2950 expð5:6 1010k Þ
ð10Þ
The results of the centrifuge test departed from the conventional oedometer test above Fc = 40 MN or below e = 4.65, as depicted in Fig. 9c that plots e versus k. This digression is attributed to deviation from the initial straight line portion of the settlement curve (Fig. 7) and confirms previous work [50]. Fig. 9d gives a comparison between r0 of the conventional oedometer test and the calculated r0 (using Eq. (9)) from the
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
6
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
Fig. 9. Correlation between conventional oedometer test and centrifuge test: (a) total centrifugal force versus effective stress; (b) total centrifugal force versus hydraulic conductivity; (c) void ratio versus hydraulic conductivity; and (d) void ratio versus effective stress.
bench-top centrifuge test. A good agreement was obtained between the two tests in the range of r0 = 10–65 kPa and e = 5.5–3.86. 4. Summary and conclusions The dewatering behaviour of clay slurries is important for tailings management. A fine-grained clay with high consistency limits (wL = 180%, wP = 120%) was used. The conventional oedometer test was carried out alongside a bench-top centrifuge test to understand the consolidation process of such materials. The main conclusions of this study are as follows: The clay slurry showed apparent preconsolidation (due to initial conditions, electrochemical interactions, tortuous drainage, and thixotropic strength) from e = 5.7 to e = 5.5 followed by virgin compression. Likewise, the low hydraulic conductivity (1010– 1012 m/s) is attributed to low porosity (small pore throats) and high tortuosity (long flow paths) of the clay slurries. Unlike consolidation of soils, the cv and mv of the clay slurry decreased with increasing r0 but increased with increasing e and k. The behaviour of the slurry determined from the oedometer test correlated well with that from the centrifuge test in the range of r0 = 10–65 kPa, e = 5.5–3.86, k = 1.7 1010–5 1011 m/s, Fc = 1–40 MN. New Eqs. (8)–(10) were developed to correlate the consolidation parameters (e, r0 , k) with Fc. The deviation of k beyond 40 MN (e = 4.65) is due to deviation from the initial straight line portion of the settlement curve in the centrifuge test. Acknowledgments The authors would like to acknowledge the University of Regina for providing laboratory space and the Natural Science and Engineering Research Council of Canada for financial assistance.
References [1] Leitea A, Dimitrakopoulos R. Stochastic optimization of mine production scheduling with uncertain ore/metal/waste supply. Int J Min Sci Technol 2014;24(6):755–62. [2] Liu Z, Dang W, Liu Q, Chen G, Peng K. Optimization of clay material mixture ratio and filling process in gypsum mine goaf. Int J Min Sci Technol 2013;23 (3):337–42. [3] Been K, Sills GC. Self-weight consolidation of soft soils: an experimental and theoretical study. Géotechnique 1981;31(4):519–35. [4] Gibson RE, Schiffman RL, Cargill KW. The theory of one-dimensional consolidation of saturated clays. II. Finite nonlinear consolidation of thick homogeneous layers. Can Geotech J 1981;18(2):280–93. [5] Scott JD, Dusseault MB, Carrier III WD. Large-scale self-weight consolidation testing. In: Consolidation of soils: testing and evaluation, vol. 10. ASTM; 1986, p. 500–15 [6] Fredlund MD, Donaldson M, Gitirana GG. Large strain 1D, 2D and 3D consolidation modeling of Tailings. In: Proceedings, tailings and mine waste, Banff; 2009. [7] Gjerapic G, Johnson J, Jeffrey C, Znidarcic D. Determination of tailings impoundment capacity via finite-strain consolidation models. In: Geo congress, New Orleans; 2008. p. 798–806. [8] Ito M, Azam S. Large strain consolidation modeling of mine waste tailings. Environ Syst Res 2013;2(7):1–12. [9] Azam S. Large strain settling behavior of polymer-amended laterite slurries. Int J Geomech 2011;11:105–12. [10] Caughill DI, Morgenstern NR, Scott JD. Geotechnics of nonsegregating oil sand tailings. Can Geotech J 1993;30:801–11. [11] Demoz A, Mikula RJ. Role of mixing energy in the flocculation of mature fine tailings. J Environ Eng 2012;138(1):129–36. [12] Bajwa T, Simms P. Evolution of microstructure during desiccation of oil sands mature fine tailings. In: Proceedings of the 18th international conference on soil mechanics and geotechnical engineering, Paris; 2013. [13] Jeeravipoolvarn S, Scott JD, Chalaturnyk RJ. 10 m standpipe tests on oil sands tailings: long-term experimental results and prediction. Can Geotech J 2009;46(8):875–88. [14] Kochmanova N, Tanaka H. Role of microstructure in the mechanical behaviour of clay. In: Proceedings, soil behavior and geo-micromechanics, Shanghai; 2010. p. 166–71. [15] Farkish A, Fall M. Consolidation and hydraulic conductivity of oil sands mature fine tailings dewatered by using super absorbent polymer. In: 66th Canadian geotechnical conference, Montreal; 2013. [16] Sridharan A, Prakash K. Simplified seepage consolidation test for soft sediments. Geotech Test J 1999;22(3):235–44.
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx [17] Terzaghi K. Theoretical soil mechanics. New York: John Wiley & Sons, Inc.; 1943. [18] Taylor DW. Fundamentals of soil mechanics. New York: John Wiley & Sons, Inc.; 1948. [19] Bo MW. Compressibility of ultra-soft soil. New Jersey: World Scientific Publishing; 2008. [20] Moore T, Zhang C, Sobkowicz J. Laboratory investigation of the large strain consolidation behaviour of oil sands fluid fine tailings treated with DuPont ParticlearTM. In: 66th Canadian geotechnical conference, Montreal; 2013. [21] Dromer J-B, Aubertin M, Kennedy G, Pedroni L, Bussière B. A new testing system to investigate the sedimentation and consolidation of sludge and slurry. In: Proceedings of the 57th Canadian geotechnical conference, Quebec; 2004. [22] Pedroni L. Measurement and simulation of the large strain consolidation of AMD treatment sludge. In: 66th Canadian geotechnical conference, Montreal; 2013. [23] Suthaker NN, Scott JD. Measurement of hydraulic conductivity in oil sand tailings slurries. Can Geotech J 1996;33(4):642–53. [24] Gan J, Ebrahimi-Birang N, Haug MD. Consolidation and hydraulic conductivity testing of slurries. In: Proceedings, tailings and mine waste, Vancouver, B.C.; 2011. [25] Berilgen SA, Berilgen MM, Ozaydin K. Determination of consolidation behaviour of Halic dredged material by using a seepage-induced consolidation test. Manage Environ Qual 2005;16(6):681–90. [26] Yune CY, Chung CK. Consolidation test at constant rate of strain for radial drainage. Geotech Test J 2005;28(1):1–8. [27] Bo MW, Choa V, Arulrajah A, Na YM. One-dimension compression of slurry with radial drainage. Japanese Geotech Soc 1999;39(4):9–17. [28] Sheahan TC, Watters PJ. Using an automated Rowe cell for constant rate of strain consolidation testing. Geotech Test J 1996;19(4):354–63. [29] Umehara Y, Zen K. Constant rate of strain consolidation for very soft clayey soils. Soils Found 1980;20(2):79–95. [30] Lowe J, Jonas E, Obrcian V. Controlled gradient consolidation test. J Soil Mech Found Div, ASCE 1969;95(1):77–98. [31] Gorman CT, Hopkins TC, Drnevich VP. Constant-rate-of-strain and controlledgradient consolidation testing. Geotech Test J 1978;1(1):3–15. [32] Kayabali K, Ozdemir A. Assessing the practicality of the centrifuge method for 1-D consolidation. Bull Eng Geol Environ 2012;71:735–45. [33] El-Shall H, Moudgil B, Bogan M. Centrifugal modeling of the consolidation of solid suspensions. Miner Metall Process J 1996;13(3):98–102.
7
[34] Fox PJ, Lee J, Qiu T. Model for large strain consolidation by centrifuge. Int J Geomech 2005;5:267–75. [35] Robinson RG, Tan TS, Lee FH. A comparative study of suction-induced seepage consolidation versus centrifuge consolidation. Geotech Test J 2003;26 (1):92–101. [36] McDermott IR, King AD. Use of a bench-top centrifuge to assess consolidation parameters. In: Proceedings of the tailings and mine waste conference, Rotterdam; 1998. p. 281–8. [37] Takada N, Mikasa M. Determination of consolidation parameters by selfweight consolidation test in centrifuge. In: Consolidation of soils: testing and evaluation. ASTM; 1986. p. 548–66. [38] Pollock WG. Large strain consolidation of oil sands tailings sludge. Edmonton: University of Alberta; 1988. [39] Azam S, Rima US. Oil sand tailings characterisation for centrifuge dewatering. Environ Geotech 2013;1(3):189–96. [40] Owolagba J, Azam S. Geotechnical properties of centrifuged oil sand fine tailings. Environ Geotech 2015;2(5):309–16. [41] Terzaghi K, Peak RB, Mesri G. Soil mechanics in engineering practice. New York: John Wiley & Sons Inc.; 1996. [42] Liu JC, Znidarcic D. Modeling one-dimensional compression characteristics of soils. J Geotech Eng Div, ASCE 1991;117(1):162–9. [43] Abu-Hejleh AN, Znidarcic D, Barnes BL. Consolidation characteristics of phosphatic clays. J Geotech Eng 1996;122(4):295–301. [44] Scully RW, Schiffman RL, Olsen HW, Ko HY. Validation of consolidation properties of phosphatic clay at very high void ratios. In: Proceedings, symposium on sedimentation consolidation models: predictions and validation, San Francisco; 1984. p. 158–81. [45] Pane V. Sedimentation and consolidation of clays. Boulder: University of Colorado; 1985. [46] Mitchell JK, Soga K. Fundamentals of soil behaviour. New York: John Willy and Sons; 2005. [47] Seng S, Tanaka H. Properties of very soft clays: a study of thixotropic hardening and behavior under low consolidation pressure. Soils Found 2012;52 (2):335–45. [48] Olsen HW, Nichols RW, Rice TL. Low gradient permeability measurements in triaxial system. Geotechnique 1985;35(2):145–57. [49] Imai G, Yano K, Aoki S. Applicability of hydraulic consolidation test for very soft clayey soils. Soils Found 1984;24(2):29–42. [50] Liu JC. Determination of soft soil consolidation. Boulder: University of Colorado; 1990.
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
Contents lists available at ScienceDirect
International Journal of Mining Science and Technology journal homepage: www.elsevier.com/locate/ijmst
Determination of consolidation behaviour of clay slurries Khan Faseel S., Azam Shahid ⇑ Environmental Systems Engineering, University of Regina, Regina S4S0A2, Canada
a r t i c l e
i n f o
Article history: Received 2 July 2015 Received in revised form 9 August 2015 Accepted 15 October 2015 Available online xxxx Keywords: Clay slurry Consolidation Conventional oedometer test Bench-top centrifuge
a b s t r a c t The main objective of this study was to determine the consolidation behaviour of clay slurries. A finegrained clay with high consistency limits (wL = 180%, wP = 120%) was investigated using conventional oedometer and bench-top centrifuge tests. Results indicated that the slurry had an apparent preconsolidation (due to initial conditions, electrochemical interactions, tortuous drainage, and thixotropic strength) from e = 5.7 to e = 5.5 followed by virgin compression. Likewise, the low hydraulic conductivity (1010–1012 m/s) was due to low porosity (small pore throats) and high tortuosity (long flow paths). Unlike consolidation of soils, the cv and mv decreased with increasing r0 but increased with increasing e and k. The data from the two tests correlated well in the range of r0 = 10–65 kPa, e = 5.5–3.86, k = 1.7 1010–5 1011 m/s, Fc = 1–40 MN. New equations were developed to correlate the consolidation parameters (e, r0 , k) with Fc. The deviation of k beyond 40 MN (e = 4.65) was due to deviation from the initial straight line portion of the settlement curve in the centrifuge test. Ó 2016 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
1. Introduction Several mining operations generate waste tailings that contain variable amounts of clay fraction thereby affecting the consolidation behaviour of such water-rich slurries [1]. The consolidation properties are important for slurry containment facilities and when the slurries are used as backfills [2]. The water release rate and amount is governed by the following factors: (i) self weight due to gravity and applied loading due to upper layer deposition [3–5]; (ii) initial and boundary conditions including the prevalent climate, side and/or bottom drainage and containment geometry [6–8]; (iii) electrochemical interactions between the clay and the pore fluid [9–11]; and (iv) microstructure and thixotropy derived from the above [12–14]. The determination of consolidation properties is based on volume compressibility and hydraulic conductivity relationships obtained from laboratory tests. Table 1 provides a summary of the various types of consolidation tests used for clay slurries. The first three tests require long time to complete but directly determine the constitutive relationships. In the conventional oedometer test, the hydraulic conductivity measurement can cause consolidation in the low effective stress range if the seepage force is greater than the applied stress. Likewise, the slurry consolidation test uses large samples thereby affecting the spatial variability in material properties. Similarly, the seepage induced test uses a complex equipment setup but is ⇑ Corresponding author. Tel.: +1 306 3372369. E-mail address: [email protected] (S. Azam).
useful in the low effective stress range. Although the continuous loading test and the controlled gradient test require less time, these tests are based on the conventional consolidation theory thereby limiting these tests to small strain applications. Finally, the bench-top centrifuge test generates consolidation data similar to the conventional oedometer test, the seepage induced test, and the continuous loading test [32,33]. There is a need to understand the results of the bench-top centrifuge test (requires the least time) in conjunction with the conventional oedometer test (well established test). The main objective of this study was to determine the consolidation behaviour of clay slurries using laboratory tests. Initially, the geotechnical index properties were determined for preliminary soil assessment. Next, the conventional oedometer test and the bench-top centrifuge test were conducted. Finally, the test results were compared to convert the former test data to the later test data thereby determining the volume compressibility and hydraulic conductivity relationships.
2. Test methods The clay was retrieved from a local geological deposit of high plasticity. The samples were collected in situ, secured in sealed plastic bags, and transported to the Geotechnical Testing Laboratory at the University of Regina where these were stored at a temperature of 25 °C. The geotechnical index properties of the active clay were determined according to standard ASTM test
http://dx.doi.org/10.1016/j.ijmst.2015.12.014 2095-2686/Ó 2016 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
2
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
Table 1 Summary of consolidation tests. Test type and reference Conventional oedometer test [15–18]
Slurry consolidation test [9,19–23] Seepage induced test [16,19,24,25]
Continuous loading tests [19,26–29]
Controlled gradient test [30,31]
Bench-top centrifuge test [32–37]
Test equipment and analysis procedure
Limitations
D = 60–63 mm, H = 20–25 mm e–r0 : Measured by incremental loading and data analyzed using SSC theory or LSC theory k: Directly measured using constant/falling head method or indirectly determined using square root of time method D = 40–495 mm, H = 150–1800 mm e–r0 : Measured by incremental loading and data analyzed using LSC theory k: Directly measured or indirectly determined by inversion of LSC theory D = 50–150 mm, H = 20–40 mm e–r’: Measured by application of hydraulic gradient and data analysis does not require any theory k: Directly measured using constant head method D = 60–254 mm, H = 20–1000 mm e–r0 : Measured by constant rate of stress or strain and data analyzed by inversion of SSC theory or LSC theory with a constant cv k: Indirectly determined by inversion of SSC theory or LSC theory D = 25 mm, H = 63 mm e-r’: Measured by controlled gradient by continuously adjusting loading rate and data analyzed by inversion of SSC theory k: Indirectly determined by inversion of SSC theory D = 26–58 mm, H = 20–125 mm Centrifuge maximum radius: 300–350 mm, rpm: 300–3000 e–r0 : Indirectly determined and data analyzed using scaling laws, LSC models or comparison of results with a different test type k: Indirectly determined from initial straight line portion of settling curve
Long test duration Direct determination of k may consolidation Scatter in consolidation constants
cause
Long test duration if k is directly measured Spatial variation in material properties in large samples Long test duration Requires sophisticated instrumentation Not applicable for high r0 range e altered due to sample rebound Compressibility and hydraulic conductivity cannot be validated Theoretical limitations render it unsuitable for large strains Loading rate is unknown at the start of the test Theoretical limitations render it unsuitable for large strains Particle segregation is more likely to occur No standard procedure to convert centrifuge data to consolidation data
Note: e–r0 – compressibility; e – void ratio; r0 – effective stress; k – hydraulic conductivity; SSC – small strain consolidation; LSC – large strain consolidation; D – diameter; and H – height.
methods as follows: (a) specific gravity (Gs) by the Standard Test Method for Specific Gravity of Soil Solids by Water Pycnometer (ASTM D854(10)); (b) liquid limit (wl), plastic limit (wp) and plasticity index (Ip) by the Standard Test Method for Liquid Limit, Plastic Limit, and Plasticity Index of Soils (ASTM D4318(10)); and (c) soil classification by the Standard Practice for Classification of Soils for Engineering Purposes (Unified Soil Classification System) (ASTM D2487(11)). Fig. 1 gives the schematic of the conventional oedometer (Model S-450). The test was conducted as per the ASTM Standard Test Methods for One-Dimensional Consolidation Properties of Soils Using Incremental Loading (D2435). The clay (up to 20 mm aggregates) was soaked in tap water and was frequently stirred with a steel rod to ensure homogeneous water distribution. The resulting slurry, at a water content (w = wl), was poured in the oedometer ring (63.5 mm diameter and 25.4 mm height). Presaturated porous stones above and below the sample allowed upward drainage, the geotextile precluded porous stone clogging, and the outer casing prevented surface tilting. The loading frame was lowered (touching the steel ball on the loading head to minimize eccentricity) and was fixed in-place using clamps. The combined stress of the loading head, porous stone, geotextile and steel ball measured 1.5 kPa. Subsequent effective stress (r0 ) increments were pneumatically applied using an air compressor and were maintained through a digitally controlled pressure regulator. A linear vertical displacement transducer (LVDT) was used to record deformations at specified time intervals. A 20 mm water head was maintained on top of the sample to ensure complete saturation throughout the test. The hydraulic conductivity (k, m/s) after each r0 increment was measured using the falling head method. Upward drainage was allowed (to prevent consolidation) using the hydraulic gradient between a bottom injection tube and the water level in the oedometer ring [38]. The hydraulic gradient was kept less than the critical hydraulic gradient to preclude internal erosion. Several k measurements were made to ensure a reliable value at a given void ratio (e). The test was stopped when two consecutive readings showed no significant change. The k was calculated from knowl-
edge of the internal standpipe area (a, mm2), sample length (L, mm), sample cross-sectional area (A, mm2), elapsed time (Dt, s), initial water head (h1, mm) and final water head (h2, mm), according to the following version of Darcy’s law:
k ¼ 0:23
aL h1 log10 A Dt h2
ð1Þ
Fig. 2 shows schematic of the bench-top centrifuge test. This setup has been recently used at the University of Regina for oil sand fine tailings [39,40]. The digitally controlled unit (Sorvall Thermo Scientific Biofuge Primo R) comprised of a swinging bucket rotor with four buckets and had a maximum radius of 127.4 mm as shown in elevation (Fig. 2a) and plan view (Fig. 2c). Graduated acrylic tubes (24.2 mm internal diameter and 106 mm height) were filled with the slurry (prepared as before) and placed in ceramic tube-holding adapters in two separate buckets (Fig. 2b). The slurry was filled up to the 40 mm mark which gave a ratio of sample height to centrifuge radius of 1:3, as suggested by McDermott and King [36]. The test was conducted at a constant temperature of 20 °C and an angular velocity of 4000 r/min. The centrifuge was stopped at specified time intervals and the interface height of the sediment was noted (Fig. 2d). The measurements were taken from both tubes and the average value was recorded. Using Ht (mm) for the interface height, Ho (mm) for the initial interface height and eo for the initial void ratio, the e was calculated using the following equation [33]:
e¼
Ht ð1 þ eo Þ 1 Ho
ð2Þ
Likewise, the centrifugal force, Fc (MN) was computed using the following equation [32]:
F c ¼ mr x2
ð3Þ
where m is mass (kg) of the slurry, r is radius (m) that is the centroid of the slurry mass and w = (rpm/60)2p is the angular velocity (1/s). The centrifugal force was applied to the specimen in 1 s and
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
3
Loading frame
Leads to readout unit
Moveable frame
Linear variable differential transducer
Head
Moveable frame
Standpipe
Loading frame
Porous stone and geotextile
Sample
Base plate
Low Low pressure regulator
Leads to readout unit
Steel ring
Loading head Outer casing
Loading frame
Steel ball
High High pressure regulator Unload
Load
Fig. 1. Schematic of the conventional oedometer test.
Transparent cover Axis of rotation Tube cap During flight
At rest
(a)
Ceramic adapter
(b) Tube cap
Plastic bucket
Plastic bucket
Plastic bucket
Released water
Sediment height (Ht)
Graduated acrylic tube
Plastic bucket
Protection walls
(c)
Graduated acrylic tube
At rest
Plastic bucket
Motor
Plastic bucket
. Initial height (Ho)
(d)
Fig. 2. Schematic of the centrifuge test: (a) side view; (b) section of the plastic bucket; (c) plan view; and (d) close up of the graduated tube.
the total centrifugal force for a specific duration was obtained as the product of Fc and the number of seconds per loading level. The k was calculated from the settlement curves according to the following equation [37]:
k¼S
cw
Nc0
ð4Þ
where S is the settlement rate, cw is the unit weight of water, c0 is the submerged unit weight, N is the acceleration in terms of
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
4
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
multiples of 9.81 m/s2 (gravitational acceleration) and was taken at a distance of 0.59 times the sample height from the base [36]. The following equation was used for calculating the c0 at a given Ht:
6
c0 ¼
4
Gs qw e g cw 1þ Gs 1þe
5
Void ratio
ð5Þ
3 2
3. Results and discussion
Test data
1 R2 = 0.98
0 0 10
101 102 103 Effective stress (kPa)
10-9 Test data R 2 = 0.99 10-10
10-11
10-12
0
Soil property
Value
Specific gravity, Gs Material finer than 0.075 mm (%) Material finer than 0.002 mm (%) Liquid limit, wL (%) Plastic limit, wP (%) Plasticity Index, wPI (%) USCS symbol
2.85 97 72 180 60 120 CH
1.5 kPa
Vertical deformation (mm)
0
15 kPa
35 kPa
66 kPa
2
3 4 Void ratio
5
6
and inaccessible pores in the clay fabric; and (iv) thixotropic strength which is highest at wl [44–47]. Likewise, the steep portion of this curve is governed by intrinsic material properties (specific gravity, grain size distribution, and clay mineralogy) and the rate of expulsion of water [46]. The test data best fitted the following Wiebull function:
ð6Þ
where A, B, E and F are dimensionless constants having values of 5.7, 5.45, 20 and 0.69 respectively. These values are similar to oil sand fine tailings [13]. Fig. 5 gives the hydraulic conductivity relationship (k–e). The k was not measured at the end of the test (1440 kPa) because of time constraint. The k varied by two orders of magnitude from 1010 to 1012 m/s when the void ratio decreased from 5.7 to 0.92. In this e range, similar values of k have been reported for phosphatic clays and for oil sand fine tailings [13,43]. The low k values are attributed to low porosity (small pore throats) and high tortuosity (long flow paths) of clay slurries. Furthermore, the data scatter at high void ratios may be attributed to possible occluded air in the specimen [48]. The k–e relationship was a typical power function as follows:
k ¼ CeD
110 kPa 5
1
Fig. 5. Hydraulic conductivity in the conventional oedometer test.
e ¼ A B expðEr0F Þ Table 2 Summary of geotechnical index properties of the investigated clay.
104
Fig. 4. Volume compressibility in the conventional oedometer test.
Hydraulic conductivity (m/s)
Table 2 gives the geotechnical index properties of the clay. The Gs was 2.85 which is typical of sedimentary clays (that is, between 2.4 and 2.95 [41]. Material finer than 0.075 mm measured 97% along with a clay size fraction (material finer than 0.002 mm) of 72% indicating the fine grained nature of the investigated soil. The high consistency limits (wL = 180%, wP = 120%) depicted high water adsorption and retention characteristics. Overall, the material was classified as highly plastic clay (CH). Fig. 3 gives the vertical deformations from the conventional oedometer test. The slurry did not settle under the seating stress of 1.5 kPa over a period of 4 days (5760 min). Subsequent r0 increments took 5 days to 19 days for the sample to enter the secondary consolidation stage which was an average deformation rate of 1/50 mm a day (secondary compression index (cae = 0.15). The vertical deformations under each r0 were found to be in the range of 1–3 mm such that the cumulative settling at the end of the test was 18 mm (70% strain). Fig. 4 gives the volume compressibility relationship (r0 –e). The e decreased from 5.7 at r0 = 1.5 kPa to 0.92 at r0 = 1440 kPa. Such a large variation in e produced a non-linear volume compressibility curve and consisted of two distinct parts: an initial flat portion (from e = 5.7 to e = 5.5) pertaining to apparent pre-consolidation followed by a steep portion (from e = 5.5 to e = 0.92) corresponding to virgin compression. Apparent pre-consolidation has been observed in other fine grained slurries such as dredged materials, phosphatic clays and oil sand tailings [13,42,43]. This phenomenon should be attributed to the following reasons: (i) initial slurry conditions such as void ratio and water content; (ii) electrochemical interactions in the colloid–liquid mixture; (iii) tortuous flow paths
ð7Þ
180 kPa 360 kPa
10
720 kPa 1440 kPa
15
20 0
50 100 Time (×103 min)
150
Fig. 3. Vertical deformation versus time in the conventional oedometer test.
where C is a constant with units of velocity (m/s) and D is a dimensionless constant. The values of C and D being 1.12 1012 m/s and 2.74, respectively. Once again, the parameters C and D were similar to phosphatic clays [43]. Fig. 6 plots the coefficient of consolidation (cv) and the coefficient of volume change (mv) with respect to the above mentioned measured properties (r0 , e, k). The cv, which governs the rate of consolidation, was calculated using the square root of time method. Likewise mv, which is the inverse of modulus of elasticity, was calculated from the slope of the virgin compression curve
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
5
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
Fig. 6. Consolidation parameters from the conventional oedometer test.
5
high eo. This was proceded by a gradual decrease in slope such that it became nearly horizontal after 6000 min. This is attributed to a reduced pore size (reduced e) due to expulsion of water through consolidation. The test data were fitted using a 3rd order polynomial equation that was differentiated to determine S and subsequently used in calculating k according to Eq. (4). The estimated k decreased from 1010 to 1012 m/s with decreasing interface height. Fig. 8 gives the volume compressibility relationship (e–Fc) from the bench-top centrifuge test. The e was calculated from Eq. (2) and the total Fc was calculated from Eq. (3). The initial Fc was arbitrarily assigned a value of 1 MN. The e decreased from 5.50 to 3.86 and the total Fc reached 162.8 MN. A non linear curve was obtained with two distinct portions; an initial flat portion (at e = 5.5) and a lower steep portion (e = 5.5–3.86). This behaviour is similar to that determined through the consolidation test that showed apparent preconsolidation followed by virgin compression. Once again, a Weibull function best fitted the volume compressibility data and was found to be of the following form:
4
e ¼ A B expðEF Fc Þ
40
Interface height (mm)
38 36 34 10 -11
32 30
Test data
28
3rd order polynomial R 2 = 0.99 S = - 0.004 +12.3 × 10 -7t - 9.2×10 -11 t 2
26
0
2000
4000 Time t (min)
6000
Hydraulic conductivity k (m/s)
10 -10
10-12 8000
Fig. 7. Interface height versus time in the centrifuge test.
Void ratio
6
ð8Þ
The above equation is similar to Eq. (6) with r replaced with Fc such that the values of the constants are A = 5.5, B = 5.0, E = 6.0 and F = 0.33. Fig. 9 shows the correlation of conventional oedometer test data with the centrifuge test data. Fig. 9a plots the applied centrifugal force (Fig. 8) versus the measured effective stress (Fig. 4) to correlate the two tests. The comparison between the two tests was made in the void ratio range of 5.7–3.86. The data best fitted the following power function: 0
3 2 1 0 -1 10
Test data e = 5.5 - 5exp (-6 Fc -0.18 ) R 2 = 0.94
10 0 10 1 10 2 Total centrifugal force (MN)
10 3
Fig. 8. Void ratio versus total centrifugal force in the centrifuge test.
F c ¼ 0:028r02:6 2
2
(Fig. 4). The cv ranged from 0.007 m to 0.077 m per year which is similar to active clays (0.019–0.09 m2 per year) [46]. Likewise, the mv varied from 0.002 m2/kN to 0.00007 m2/kN which is similar to dredged slurries [49]. As expected, both cv and mv decreased with increasing r0 but increased with increasing e and k. Power functions were used to fit the relationships and were found to correlate well with the test data as indicated by the corresponding coefficients of determination (0.99 P R2 P 0.73). The power functions demonstrate that the small strain consolidation theory does not apply to the dewatering behaviour of clay slurries. Fig. 7 gives the settlement curve from the bench-top centrifuge test. The interface height decreased from 40 mm to 30 mm (25% strain) in 6000 min (4 days). Initially, the interface decreased following a steep slope (S) due to the high initial k corresponding to a
ð9Þ
The above equation captured the large changes in void ratio of the slurry as opposed to a linear relationship for soils [32]. Fig. 9b plots Fc (derived from Eq. (9)) versus k (from both tests). The k values measured in the range of Fc = 1–40 MN correlated well with k from the conventional oedometer test according to the following exponential equation:
F c ¼ 2950 expð5:6 1010k Þ
ð10Þ
The results of the centrifuge test departed from the conventional oedometer test above Fc = 40 MN or below e = 4.65, as depicted in Fig. 9c that plots e versus k. This digression is attributed to deviation from the initial straight line portion of the settlement curve (Fig. 7) and confirms previous work [50]. Fig. 9d gives a comparison between r0 of the conventional oedometer test and the calculated r0 (using Eq. (9)) from the
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
6
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx
Fig. 9. Correlation between conventional oedometer test and centrifuge test: (a) total centrifugal force versus effective stress; (b) total centrifugal force versus hydraulic conductivity; (c) void ratio versus hydraulic conductivity; and (d) void ratio versus effective stress.
bench-top centrifuge test. A good agreement was obtained between the two tests in the range of r0 = 10–65 kPa and e = 5.5–3.86. 4. Summary and conclusions The dewatering behaviour of clay slurries is important for tailings management. A fine-grained clay with high consistency limits (wL = 180%, wP = 120%) was used. The conventional oedometer test was carried out alongside a bench-top centrifuge test to understand the consolidation process of such materials. The main conclusions of this study are as follows: The clay slurry showed apparent preconsolidation (due to initial conditions, electrochemical interactions, tortuous drainage, and thixotropic strength) from e = 5.7 to e = 5.5 followed by virgin compression. Likewise, the low hydraulic conductivity (1010– 1012 m/s) is attributed to low porosity (small pore throats) and high tortuosity (long flow paths) of the clay slurries. Unlike consolidation of soils, the cv and mv of the clay slurry decreased with increasing r0 but increased with increasing e and k. The behaviour of the slurry determined from the oedometer test correlated well with that from the centrifuge test in the range of r0 = 10–65 kPa, e = 5.5–3.86, k = 1.7 1010–5 1011 m/s, Fc = 1–40 MN. New Eqs. (8)–(10) were developed to correlate the consolidation parameters (e, r0 , k) with Fc. The deviation of k beyond 40 MN (e = 4.65) is due to deviation from the initial straight line portion of the settlement curve in the centrifuge test. Acknowledgments The authors would like to acknowledge the University of Regina for providing laboratory space and the Natural Science and Engineering Research Council of Canada for financial assistance.
References [1] Leitea A, Dimitrakopoulos R. Stochastic optimization of mine production scheduling with uncertain ore/metal/waste supply. Int J Min Sci Technol 2014;24(6):755–62. [2] Liu Z, Dang W, Liu Q, Chen G, Peng K. Optimization of clay material mixture ratio and filling process in gypsum mine goaf. Int J Min Sci Technol 2013;23 (3):337–42. [3] Been K, Sills GC. Self-weight consolidation of soft soils: an experimental and theoretical study. Géotechnique 1981;31(4):519–35. [4] Gibson RE, Schiffman RL, Cargill KW. The theory of one-dimensional consolidation of saturated clays. II. Finite nonlinear consolidation of thick homogeneous layers. Can Geotech J 1981;18(2):280–93. [5] Scott JD, Dusseault MB, Carrier III WD. Large-scale self-weight consolidation testing. In: Consolidation of soils: testing and evaluation, vol. 10. ASTM; 1986, p. 500–15 [6] Fredlund MD, Donaldson M, Gitirana GG. Large strain 1D, 2D and 3D consolidation modeling of Tailings. In: Proceedings, tailings and mine waste, Banff; 2009. [7] Gjerapic G, Johnson J, Jeffrey C, Znidarcic D. Determination of tailings impoundment capacity via finite-strain consolidation models. In: Geo congress, New Orleans; 2008. p. 798–806. [8] Ito M, Azam S. Large strain consolidation modeling of mine waste tailings. Environ Syst Res 2013;2(7):1–12. [9] Azam S. Large strain settling behavior of polymer-amended laterite slurries. Int J Geomech 2011;11:105–12. [10] Caughill DI, Morgenstern NR, Scott JD. Geotechnics of nonsegregating oil sand tailings. Can Geotech J 1993;30:801–11. [11] Demoz A, Mikula RJ. Role of mixing energy in the flocculation of mature fine tailings. J Environ Eng 2012;138(1):129–36. [12] Bajwa T, Simms P. Evolution of microstructure during desiccation of oil sands mature fine tailings. In: Proceedings of the 18th international conference on soil mechanics and geotechnical engineering, Paris; 2013. [13] Jeeravipoolvarn S, Scott JD, Chalaturnyk RJ. 10 m standpipe tests on oil sands tailings: long-term experimental results and prediction. Can Geotech J 2009;46(8):875–88. [14] Kochmanova N, Tanaka H. Role of microstructure in the mechanical behaviour of clay. In: Proceedings, soil behavior and geo-micromechanics, Shanghai; 2010. p. 166–71. [15] Farkish A, Fall M. Consolidation and hydraulic conductivity of oil sands mature fine tailings dewatered by using super absorbent polymer. In: 66th Canadian geotechnical conference, Montreal; 2013. [16] Sridharan A, Prakash K. Simplified seepage consolidation test for soft sediments. Geotech Test J 1999;22(3):235–44.
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014
F.S. Khan, S. Azam / International Journal of Mining Science and Technology xxx (2016) xxx–xxx [17] Terzaghi K. Theoretical soil mechanics. New York: John Wiley & Sons, Inc.; 1943. [18] Taylor DW. Fundamentals of soil mechanics. New York: John Wiley & Sons, Inc.; 1948. [19] Bo MW. Compressibility of ultra-soft soil. New Jersey: World Scientific Publishing; 2008. [20] Moore T, Zhang C, Sobkowicz J. Laboratory investigation of the large strain consolidation behaviour of oil sands fluid fine tailings treated with DuPont ParticlearTM. In: 66th Canadian geotechnical conference, Montreal; 2013. [21] Dromer J-B, Aubertin M, Kennedy G, Pedroni L, Bussière B. A new testing system to investigate the sedimentation and consolidation of sludge and slurry. In: Proceedings of the 57th Canadian geotechnical conference, Quebec; 2004. [22] Pedroni L. Measurement and simulation of the large strain consolidation of AMD treatment sludge. In: 66th Canadian geotechnical conference, Montreal; 2013. [23] Suthaker NN, Scott JD. Measurement of hydraulic conductivity in oil sand tailings slurries. Can Geotech J 1996;33(4):642–53. [24] Gan J, Ebrahimi-Birang N, Haug MD. Consolidation and hydraulic conductivity testing of slurries. In: Proceedings, tailings and mine waste, Vancouver, B.C.; 2011. [25] Berilgen SA, Berilgen MM, Ozaydin K. Determination of consolidation behaviour of Halic dredged material by using a seepage-induced consolidation test. Manage Environ Qual 2005;16(6):681–90. [26] Yune CY, Chung CK. Consolidation test at constant rate of strain for radial drainage. Geotech Test J 2005;28(1):1–8. [27] Bo MW, Choa V, Arulrajah A, Na YM. One-dimension compression of slurry with radial drainage. Japanese Geotech Soc 1999;39(4):9–17. [28] Sheahan TC, Watters PJ. Using an automated Rowe cell for constant rate of strain consolidation testing. Geotech Test J 1996;19(4):354–63. [29] Umehara Y, Zen K. Constant rate of strain consolidation for very soft clayey soils. Soils Found 1980;20(2):79–95. [30] Lowe J, Jonas E, Obrcian V. Controlled gradient consolidation test. J Soil Mech Found Div, ASCE 1969;95(1):77–98. [31] Gorman CT, Hopkins TC, Drnevich VP. Constant-rate-of-strain and controlledgradient consolidation testing. Geotech Test J 1978;1(1):3–15. [32] Kayabali K, Ozdemir A. Assessing the practicality of the centrifuge method for 1-D consolidation. Bull Eng Geol Environ 2012;71:735–45. [33] El-Shall H, Moudgil B, Bogan M. Centrifugal modeling of the consolidation of solid suspensions. Miner Metall Process J 1996;13(3):98–102.
7
[34] Fox PJ, Lee J, Qiu T. Model for large strain consolidation by centrifuge. Int J Geomech 2005;5:267–75. [35] Robinson RG, Tan TS, Lee FH. A comparative study of suction-induced seepage consolidation versus centrifuge consolidation. Geotech Test J 2003;26 (1):92–101. [36] McDermott IR, King AD. Use of a bench-top centrifuge to assess consolidation parameters. In: Proceedings of the tailings and mine waste conference, Rotterdam; 1998. p. 281–8. [37] Takada N, Mikasa M. Determination of consolidation parameters by selfweight consolidation test in centrifuge. In: Consolidation of soils: testing and evaluation. ASTM; 1986. p. 548–66. [38] Pollock WG. Large strain consolidation of oil sands tailings sludge. Edmonton: University of Alberta; 1988. [39] Azam S, Rima US. Oil sand tailings characterisation for centrifuge dewatering. Environ Geotech 2013;1(3):189–96. [40] Owolagba J, Azam S. Geotechnical properties of centrifuged oil sand fine tailings. Environ Geotech 2015;2(5):309–16. [41] Terzaghi K, Peak RB, Mesri G. Soil mechanics in engineering practice. New York: John Wiley & Sons Inc.; 1996. [42] Liu JC, Znidarcic D. Modeling one-dimensional compression characteristics of soils. J Geotech Eng Div, ASCE 1991;117(1):162–9. [43] Abu-Hejleh AN, Znidarcic D, Barnes BL. Consolidation characteristics of phosphatic clays. J Geotech Eng 1996;122(4):295–301. [44] Scully RW, Schiffman RL, Olsen HW, Ko HY. Validation of consolidation properties of phosphatic clay at very high void ratios. In: Proceedings, symposium on sedimentation consolidation models: predictions and validation, San Francisco; 1984. p. 158–81. [45] Pane V. Sedimentation and consolidation of clays. Boulder: University of Colorado; 1985. [46] Mitchell JK, Soga K. Fundamentals of soil behaviour. New York: John Willy and Sons; 2005. [47] Seng S, Tanaka H. Properties of very soft clays: a study of thixotropic hardening and behavior under low consolidation pressure. Soils Found 2012;52 (2):335–45. [48] Olsen HW, Nichols RW, Rice TL. Low gradient permeability measurements in triaxial system. Geotechnique 1985;35(2):145–57. [49] Imai G, Yano K, Aoki S. Applicability of hydraulic consolidation test for very soft clayey soils. Soils Found 1984;24(2):29–42. [50] Liu JC. Determination of soft soil consolidation. Boulder: University of Colorado; 1990.
Please cite this article in press as: Khan FS, Azam S. Determination of consolidation behaviour of clay slurries. Int J Min Sci Technol (2016), http://dx.doi. org/10.1016/j.ijmst.2015.12.014