Journal of Cleaner Production 14 (2006) 1168e1175

www.elsevier.com/locate/jclepro

Geotechnical considerations in mine backfilling in Australia

N. Sivakugan a,*, R.M. Rankine b, K.J. Rankine a, K.S. Rankine a
a
School of Engineering, James Cook University, Townsville 4811, Australia
b
Cannington Mine, BHP Billiton, P.O. Box 5874, Townsville 4810, Australia
Received 1 March 2004; received in revised form 2 June 2004; accepted 3 June 2004
Available online 26 April 2005

Abstract

Mine backfilling can play a significant role in the overall operation of a mine operation. In the Australian mining industry, where
safety is a prime consideration, hydraulic systems are the most common backfills deployed. Many accidents reported at hydraulic fill
mines worldwide have mainly been attributed to a lack of understanding of their behaviour and barricade bricks. This paper
describes the findings from an extensive laboratory test programme carried out in Australia on more than 20 different hydraulic fills
and several barricade bricks. A limited description of paste backfills is also provided, and the usefulness of numerical modelling as
an investigative tool is highlighted.
Ó 2005 Elsevier Ltd. All rights reserved.

Keywords: Hydraulic fills; Mining; Backfills; Paste fills; Geotechnical

1. Introduction can be studied using soil mechanics theories. A typical

example of uncemented backfilling is the use of
In the mining industry, when underground ore bodies hydraulic fills that are placed in the form of slurry into
are extracted, very large voids are created, which must be the underground voids. The second category, cemented
backfilled. The backfilling strategies deployed often make backfilling, makes use of a small percentage of binder
use of the waste rock or tailings that are considered such as Portland cement or a blend of Portland cement
by-products of the mining operation. This is an effective with another pozzolan such as fly ash, gypsum or blast
means of tailing disposal because it negates the need furnace slag.
for constructing large tailing dams at the surface. The The purpose of this paper is to analyse the findings
backfilling of underground voids also improves local and from an extensive laboratory test programme carried
regional stability, enabling safer and more efficient out in Australia on hydraulic fills and several barricade
mining of the surrounding areas. The need for backfilling bricks. Hydraulic fills are uncemented techniques, and
is a major issue in Australia, where 10 million cubic are one of the most widely used backfilling strategies in
metres of underground voids are generated annually as Australia. More than 20 different hydraulic fills,
a result of mining [1]. representing a wide range of mines in Australia, were
There are two basic types of backfilling strategies. studied at James Cook University (JCU). The grain size
The first, uncemented backfilling, does not make use of distributions for all of these fills lie within a narrow
binding agents such as cement, and their characteristics band as shown in Fig. 1. Along with them, the grain size
distribution curves for a paste fill and a cemented
hydraulic fill are also shown. It can be seen that the
* Corresponding author. Fax: C61 7 47751184. cemented hydraulic fill falls within the same band as the
E-mail address: [email protected] (N. Sivakugan). hydraulic fill. The addition of a very small percentage of

0959-6526/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jclepro.2004.06.007
 N. Sivakugan et al. / Journal of Cleaner Production 14 (2006) 1168e1175 1169

100 water, and also serve as drains when the hydraulic fill
90 Cemented hydraulic fill rises in the stope.
80
Percent finer by weight

70
2.1. Drainage considerations
60
50
Drainage is the most important issue that must be
40
considered when designing hydraulic fill stopes. There
30 Paste fill
have been several accidents (namely, trapped miners and
Australian hydraulic fills
20 machinery) worldwide caused by wet hydraulic fill
10 rushing through horizontal access drives (Fig. 2).
0 Several reasons, including poor quality barricade bricks,
0.1 1 10 100 1000 10000
liquefaction, and piping within the hydraulic fill are
Grain size (µm)
attributed to such failures [2]. Therefore, permeability of
Fig. 1. Typical grain size distribution curves for hydraulic fills, the hydraulic fill in the stope is a critical parameter in
cemented hydraulic fills and paste fills. the design; continuous effort is made during mining to
ensure that it is kept above a threshold limit in the
vicinity of 100 mm/h [3]. Larger permeability leads to
cement has a limited effect on grain size distribution. quicker removal of water from the stope, thus improv-
Paste fills generally have a much larger fine fraction than ing the stability of the fill contained within the stope.
hydraulic fills or cemented hydraulic fills, but have Permeability tests for mine fills and barricade bricks
negligible colloidal fraction finer than 2 mm. are discussed by Rankine et al. [4]. The constant head
and falling head permeability tests carried out on the
hydraulic fill samples give permeability values in the
2. Hydraulic backfills range of 7e35 mm/h. In spite of having permeability
values much less than the 100 mm threshold suggested
Hydraulic fills are simply silty sands or sandy silts by Herget and De Korompay [3], each of these hydraulic
without clay fraction, and are classified as ML or SM
under the Unified Soil Classification System. The clay
Slurry enters
fraction is removed through a process known as de- stope
sliming, whereby the entire fill material is circulated
through hydrocyclones and the fine fraction is removed
and then sent to the tailings dam. The remaining
hydraulic fill fraction is reticulated in the form of slurry
through pipelines to underground voids.
Over the past decade there has been a steady increase
in the solid content of the hydraulic fill slurry placed in
Horizontal
mines in an attempt to reduce the quantity of water that access drive
must be drained and increase the proportion of solids.
The challenge posed by a high solid content is that it
becomes difficult to transport the slurry through the
pipelines due to rheological considerations. Currently,
solid contents of 75e80% are common, although
even at 75% solid content, assuming a specific gravity
of 3.00 for the solid grains, 50% of slurry volume is Decant water
water. Therefore, there is opportunity for a substantial
amount of water to be drained from the hydraulic fill
stope.
To contain the fill, the horizontal access drives
created during mining are generally blocked by barri- Hydraulic fill Porous barricade
brick wall
cades constructed from specially made porous bricks
(Fig. 2). The access drives, which are made large enough
to permit the entry of machinery during mining, are
blocked by the barricades during filling. The drives are Horizontal
access drive
often located at more than one level. Initially, the drives
located at upper levels act as exit points for the decanted Fig. 2. An idealised stope with two sublevel drains.
1170 N. Sivakugan et al. / Journal of Cleaner Production 14 (2006) 1168e1175

fills has performed satisfactorily. Anecdotal evidences Densest possible

state
and back calculations using the measured flow in
the mine stopes suggest that the permeability of the 100
hydraulic fill in the mine is often larger than what is
measured in the laboratory under controlled conditions.
Kuganathan [5] and Brady and Brown [6] proposed
permeability values in the range of 30e50 mm/h, which

Relative Density, Dr ( )
are significantly larger than those measured in the
laboratory for similar fills. These values are much less
than the threshold limit prescribed by Herget and De
50
Korompay [3], suggesting that it is a conservative
Hydraulic fills
recommendation. in mines

2.2. Stability considerations

The stability of the hydraulic fill stope during and

after the drainage period depends on several parameters
that determine the strength and the stiffness of the 0
emaximum eminimum
hydraulic fill mass. These parameters can be measured in Loosest possible Void ratio, e
the laboratory using reconstituted samples or in the state
mine using in situ testing devices. Due to the difficulties
Fig. 3. Relative density of the hydraulic fills sedimented in the
and high costs associated with carrying the in situ testing laboratory.
rigs into the underground openings, laboratory tests are
the preferred alternatives.
Strength and stiffness are directly related to the relative
density of the fill. When the hydraulic fill is denser, the density values ranging from 44 to 66% at four different
relative density and friction angle are higher, and thus the mines.
fill is more stable. In geotechnical engineering, there are The laboratory exercise also showed that the hydrau-
several empirical correlations relating relative density to lic fill slurry settles to a dry density (g/cm3) of 0.6 times
the Young’s modulus and friction angle of a granular soil. the specific gravity (Gs) for a wide range of tailings with
specific gravity values ranging from 2.8 to 4.4. Dry
density (rd) and void ratio (e) are related by:
2.2.1. Maximum and minimum dry density tests
A larger void ratio does not always mean a looser Gs rw
granular soil. Relative density is a good measure of rd Z ð2Þ
1Ce
the density of the grain packing, and depends on the
maximum and minimum possible void ratios for the soil This implies that all the hydraulic fills settle to a void
whilst still maintaining intergranular contact. The ratio of 0.67 and porosity of 40%. The laboratory
minimum void ratio is generally determined by pouring sedimentation exercise verifies this.
the dry tailings from a fixed height so that the grains are
placed at a very loose state [7]. The maximum void ratio is 2.2.2. Oedometer tests
generally achieved by saturating the tailings and vibrating Oedometer tests are carried out on hydraulic fills to
them to attain dense packing [8]. These two extreme void determine the constitutive modelling parameters for the
ratios provide the lower and upper bound for the void Cam Clay model e one of the constitutive models that
ratios, and, depending on where the current void ratio can be adapted for hydraulic fills when analysed using
of the hydraulic fill is, the relative density is defined as: numerical modelling packages such as FLAC, FLAC3D
or ABAQUS. In addition, oedometer tests are useful in
emax
ecurrent determining the constrained modulus (D) from which,
Dr Z !100% ð1Þ
emax
emin Young’s modulus (E ) can be estimated for an assumed
value of Poisson’s ratio using the following equation.
Laboratory sedimentation exercises at JCU labora-
tories, during which hydraulic filling processes were ð1CyÞð1
2yÞ
EZ D ð3Þ
simulated, showed consistently that when slurry settles ð1
yÞ
under its self-weight, the relative density of the fill is in
the range of 40e70% (Fig. 3). Similar observations were Young’s modulus is a crucial parameter in deforma-
made by Pettibone and Kealy [9] at selected mines in the tion calculations using most constitutive models. The
United States. The in situ measurements showed relative oedometer tests on the hydraulic fills showed significant
 N. Sivakugan et al. / Journal of Cleaner Production 14 (2006) 1168e1175 1171

2.2.4. Placement property test

A placement property test for hydraulic fills was
proposed by Clark [10]. This is essentially a compaction
test, where the compactive effort is applied through
5 min of vibration on a vibrating table. Porosity at the
end of vibration is plotted against the water content.
Alternatively, dry density can be plotted against water
content, as shown in Fig. 5. Here a is the air content,
and the contours of a Z 0, 3, 10, 20 and 30% are shown
in the figure. The shaded region is where the hydraulic
fill can exist whilst maintaining intergranular contact.
The slurry follows a saturation line when settling under
its self-weight, with the density increasing with some
vibratory loading.
One of the main applications of the placement
property test, as in a compaction test, is to determine
Fig. 4. Scanning electron micrograph of a hydraulic fill sample.
optimum water content. In Fig. 5, the optimum water
content of the fill is 14%, with the maximum dry density
creep settlements that took place on the completion of of 2.42 t/m3. This water content can also be estimated
consolidation settlements. This has yet to be verified from a maximum dry density test and the saturation
quantitatively and on a full-scale stope. line as 12%. These curves are useful in assessing the
contractive or dilative behaviour of hydraulic fills at
2.2.3. Direct shear test various water contents. For example, when the fill in
Direct shear tests are carried out to determine the Fig. 5 is subjected to vibratory loading (e.g., due to
peak and residual friction angle of the hydraulic fill. The blasting) at 14% water content and a dry density of 2.0 t/
tests are carried out on reconstituted hydraulic fills m3, it will densify, whilst the same fill at 8% water
representing the in situ grain packing in the stope, which content and dry density of 2.2 t/m3 will become looser.
can be at relative densities of 40e70%. Since there is
no clay fraction, cohesion is zero. Direct shear tests 3. Barricade bricks for hydraulic fill mines
conducted at JCU reveal that the friction angles
determined from direct shear tests are significantly Barricade failure in underground mining operations
higher than those determined for common granular is a primary safety concern because of the potential
soils. This can be attributed to the very angular grains consequences of failure. Between 1980 and 1997, 11
that result from crushing the rock waste, which interlock barricade failures were recorded at Mount Isa Mines in
more than the common granular soils. The angular both hydraulic and cemented hydraulic fills [5]. In 2000,
grains can be seen in the scanning electron micrographs a barricade failure at the Normandy Bronzewing Mine
of the hydraulic fill samples (Fig. 4). in Western Australia resulted in a triple fatality, and two

2.8

2.6
Maximum dry density
Dry density (t/m3)

2.4
5 min vibration
2.2
< 5 min vibration
2
No vibration (free settling under self weight)
1.8
Minimum dry density
1.6 Intergranular
contact exists
1.4

1.2

1
0 10 20 30 40 50 60

Fig. 5. Placement property curve of a hydraulic fill sample.

1172 N. Sivakugan et al. / Journal of Cleaner Production 14 (2006) 1168e1175

permeable brick failures were reported later that same value for uniaxial compressive strength for the bricks in
year as a result of hydraulic fill containment at the the order of 10 MPa [11]. Kuganathan [5] and Duffield
Osborne Mine in Queensland [1]. et al. [11] have reported uniaxial compressive strength
The specialized barricade bricks often used for the values from 5 MPa to over 26 MPa.
containment of hydraulic fill in underground mines are A series of uniaxial compressive strength tests
generally constructed of a mortar composed of mixture undertaken on a large sample of brick cores have
of gravel, sand, cement and water at the approximate demonstrated the scatter of results, but more impor-
ratio of 40:40:5:1, respectively. Fig. 6 shows a photo- tantly, have highlighted a distinct variation in brick
graph of (a), a barricade brick and (b), an underground performance when saturated, as it would occur in the
containment wall constructed from bricks. Traditionally, mines. Two identical cylindrical cores were cut from 29
the walls have been constructed in a vertical plane, porous barricade bricks. One of the brick cores from
but the recent industry trend has been to increase wall each of the individual bricks was tested dry, and the
strength by constructing them in a curved manner, with other core was tested after having been saturated
the convex toward the hydraulic fill as shown in Fig. 6b. for either 7 or 90 days. The strength and deformation
Although it is known within the mining industry parameters (namely, the uniaxial strength, Young’s
that the porous bricks used in underground barricade modulus, and the axial failure strain) for the wet and
construction are prone to variability in strength prop- dry cores are shown in Figs. 7e9.
erties [5], the manufacturers often guarantee a minimum Firstly, the extreme scatter between all results reiter-
ates the significant deviation in brick quality. Fig. 7 shows
the average uniaxial compressive strength of dry bricks to
fall between 6 and 10 MPa, when the brick manufacturers
guarantee minimum of 10 MPa. It can also be seen from
this figure that there is a distinct loss of compressive
strength as a result of wetting the brick. There was no
significant difference between 7 and 90 days soaking,
implying that the strength loss occurs immediately
upon wetting. This loss appears to be in the order of
approximately 25%, which is notable considering that
bricks are generally exposed to saturated conditions when
placed underground, and all manufacturer strength
specifications are based on bricks that are tested dry.
The stiffness also appears to be reduced by wetting
(Fig. 8). The Young’s modulus of the dry cores ranged
between 1 and 3.5 MPa. The length of time the bricks
were wetted did not have a significant impact on
the magnitude of the reduction in stiffness. The peak
failure axial strain was not reduced by wetting (Fig. 9).
The cores in general failed under an axial strain of less
than 1%.
The porous bricks are designed to be free draining
and therefore, their permeability is at least an order of
magnitude greater than that of hydraulic fill. The
Uniaxial compressive strength

14 7 days
90 days
12
of wet core (MPa)

10
8
6
4
2
0
0 2 4 6 8 10 12 14
Uniaxial compressive strength of dry core (MPa)
Fig. 6. Porous brick barricade. (a) A brick, (b) brick barricade under
construction in a mine. Fig. 7. Uniaxial strength of dry and wet bricks.
 N. Sivakugan et al. / Journal of Cleaner Production 14 (2006) 1168e1175 1173

4 1
Young's modulus of wet core (GPa)

7 days
90 days 0.9 SLURRY PASTE CAKE

A Measure of Strength
3.5 (segregates) (non-segregating) (non-pumpable)
0.8
3 0.7
0.6
2.5 0.5
2 0.4
0.3
1.5 0.2
0.1
1
0
0.5 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
A Measure of Concentration
0
0 0.5 1 1.5 2 2.5 3 3.5 4 Fig. 10. Thickened tailings continuum [13].
Young's modulus of dry core (GPa)

Fig. 8. Young’s modulus of dry and wet bricks.

Since a desliming of the tailings is not undertaken,
there is a substantial fine content in paste fills (Fig. 1). A
barricade bricks have proven, over time, to satisfy the generic ‘‘rule of thumb’’ for the grain size distribution is
free-draining situation, and the reduction of permeabil- for a minimum of 15% of the material to be finer than
ity through mitigation of fines has not been recorded. 20 mm, which ensures that the surface area of the grains
Rankine et al. [4] carried out constant head and falling is large enough to provide adequate surface tension to
head permeability tests on several barricade bricks and ensure that the water is held to the solid particles and to
reported permeability values in the order of 3500 mm/h, provide a very thin, permanent lubricating film. Paste fill
three orders of magnitude greater than the permeability typically shows non-NewtonianeBingham plastic flow
of the tailings. characteristics, resulting in plug flow (batches flow in
solid slugs) characteristics of the paste.
As most of the early research performed on paste fills
4. Paste fill was on the transportation and deposition of the paste,
the majority of the definitions of the paste are based on
Like hydraulic fill, paste fill falls into the category of its rheological characteristics. Table 1 summarises some
thickened tailings. A conceptual framework to describe common characteristics of the thickened tailings con-
thickened tailings in terms of concentration and strength tinuum shown in Fig. 10 [14]. Hydraulic fills fall into the
is shown in Fig. 10 [12,13]. Paste fill is comprised of full thickened tailings profile. A significant difference to note
mill tailings with a typical effective grain size of 5 mm, is that the water content in paste fill is retained on
mixed with a small percentage of binder, in the order of placement, through the large surface area of the grains,
3e6% by weight, and water. It is the densest form of eliminating the need for the design of drainage of the fill
backfill in the spectrum of thickened tailings placed or barricades.
underground as a backfill material. The acceptance of The design requirements for paste filled stopes are then
paste backfill as a viable alternative to hydraulic slurry reduced to static and dynamic stability requirements. By
and rock fill did not truly occur until the mid- to late- designing the fill masses with sufficient strength to ensure
1990s with the construction and successful operation of the vertical faces of the back filled stopes remain stable
several paste backfill systems in Canada and the BHP throughout the mining of the adjacent stopes, the static
Billiton Cannington Mine in Australia. stability requirements are satisfied. If the paste becomes
unstable, the adjacent faces may relax and displace into
the open stope, causing high levels of dilution and loss of
1.2
7 days mining economies. The required strength of the backfills
90 days
1 is typically calculated using analytical solution techni-
0.8
ques [15e17]. More recently, numerical modelling
solutions [18,19] have been used to determine backfill
0.6 stability throughout the entire mining sequence.
0.4 The dynamic stability of the paste fill stopes is
addressed by designing the backfill mass to resist lique-
0.2
faction or other seismic activities. Due to the increased
0 residual moisture content of paste, there is an increased
0 0.2 0.4 0.6 0.8 1 1.2
liquefaction potential risk for the paste. Clough et al.
[20] showed that cemented sand with a uniaxial com-
Fig. 9. Axial failure strains of dry and wet bricks. pressive strength of 100 kPa was capable of resisting
1174 N. Sivakugan et al. / Journal of Cleaner Production 14 (2006) 1168e1175

Table 1
Material properties for thickened tailings continuum [14]
Material property Slurry Thickened tailings Paste
Particle size Coarse fraction only. No particles Some fines included (typically !15%), Additional/most fines
less than 20 mm. Segregation fines content tends to modify behaviour (typically 15% (min) O 20 mm
during transportation and or from slurry e i.e. rheological characteristics
placement is dependent only on more similar to paste, however does
the coarse fraction segregate when bought to rest. Segregation
during transportation and or placement is
dependent only on the coarse fraction
Pulp density (%) 60e72 70e78 78e82
Flow regimes/line Critical flow velocity. To maintain Critical flow velocity. To maintain flow No critical pipeline flow
velocities flow must have turbulent flow must have turbulent flow (vel O 2 m/s). velocity, i.e. no settling
(vel O 2 m/s). If vel ! 2 m/s If vel ! 2 m/s partial settling occurs in pipe
settling occurs
Newtonian flow Newtonian flow Laminar/plug flow
Yield stress No minimum yield stress No minimum yield stress Minimum yield Stress
Preparation Cyclone Cyclone end elutriation Filter/centrifuge
Segregation in stope Yes/high Slight/partial None
Drainage from Stope Yes Partial/limited None/insignificant
Final density Low Medium/high High
Supernatant water High Some None
Post placement shrinkage High Insignificant Insignificant
Rehabilitation Delayed Immediate Immediate
Permeability Medium/low Low Very low

a seismic activity measuring 7.5 on the Richter scale. fill and paste fill stopes, studying the developments of
This figure has been adopted by the mining industry as stresses and drainage within the fill [19,21].
the minimum design strength fill for any fill mass. The The intention of this paper has not been to detail the
strength of the paste satisfying the static stability findings from these simulations but rather to highlight
requirements are generally in excess of dynamic strength the potential these modelling tools have to dramatically
requirements. increase the confidence with which stope predictions
Barricades are designed as underground retaining may be made, ultimately leading to optimised mine
walls. The structural design and construction of the operation and safety.
walls may vary slightly to those designed for hydraulic
fills, due to the absence of drainage capabilities. The
barricades are designed as temporary structures in paste 6. Conclusions
fill stopes. The walls must be designed to retain the
liquid mass of the fill, until such time as it has cured Cemented backfilling and uncemented backfilling are
sufficiently to act as a plug at the base of the stope, thus the two strategies used in mine backfilling in Australia.
preventing the additional deposited paste from entering Hydraulic fills and paste fills are examples of unce-
the mine workings. mented and cemented backfills, respectively. A series of
laboratory tests carried out at James Cook University
on more than 20 different hydraulic fill samples suggest
the following:
5. Numerical modelling
 The hydraulic fill, placed in the form of slurry, settles
In large-scale underground mining operations, where to relative densities of 40e70%, comparing well with
in situ monitoring of stresses, strains, displacements and the field measurements.
pore pressures is often very difficult, expensive or not  Specific gravity of the hydraulic fill grains range
feasible at all, the use of numerical modelling techniques from 2.8 to 4.4.
becomes extremely valuable in understanding and pre-  All of the reconstituted hydraulic fill samples
dicting the behaviours of both the materials and the sedimented in the laboratory, simulating the slurry
systems being modelled. placement in the mine, settled to a void ratio of 0.67,
FLAC and FLAC3D are explicit, finite difference and porosity of 40%.
software packages specifically designed for solving geo-  From constant head and falling head permeability
technical and mining problems in two and three dimen- tests carried out on the hydraulic fill samples, the
sions, respectively. The research group at JCU has used permeability was measured to be in the range of
FLAC3D in simulating the filling operations in a hydraulic 7e35 mm/h.
 N. Sivakugan et al. / Journal of Cleaner Production 14 (2006) 1168e1175 1175

There is significant scatter in the uniaxial strength Townsville: The Australian Institute of Mining and Metallurgy;
and Young’s modulus of porous barricade bricks 2002. p. 161e5.
[7] ASTM D 4254e91. Standard test method for minimum index
measured in the laboratory. Uniaxial strength decreases density and unit weight of soils and calculation of relative density.
by about 25% as a result of wetting the brick. Since Annual book of ASTM standards. U.S.A.: American Society of
these bricks are always subjected to wet conditions Testing Materials; 1996.
within the mine, the strength and Young’s modulus [8] ASTM D 4253e93. Test method for maximum index density and
values of wet bricks should be used in the design of unit weight of soils using a vibratory table. Annual book of
ASTM standards. U.S.A.: American Society of Testing Materials;
barricade walls. 1996.
Paste fill contains at least 15% of grains finer than [9] Pettibone HC, Kealy CD. Engineering properties of mine tailings.
20 mm, and the effective grain size (D10) is in the order of Journal of Soil Mechanics and Foundations Division, ASCE
5 mm. The 3e6% binder improves the strength and thus 1971;97(SM9):1207e25.
stability significantly. The large fine content within the [10] Clark IH. The properties of hydraulically placed backfill. In:
Proceedings of backfill in South African mines. Johannesburg:
paste fill enables most of the water to be held to the SAIMM; 1988. p. 15e33.
surface of the grains, and therefore drainage is not a [11] Duffield C, Gad E, Bamford W. Investigation into the structural
concern in paste backfilling. behaviour of mine brick barricades. Institute of Engineers;
March/April, 2003. p. 45e50.
[12] Robinsky EI. Thickened discharge e a new approach to tailings
Acknowledgements disposal. Canadian Mining and Metallurgical Bulletin 1975;68:
47e53.
Several mines have contributed cash and in-kind [13] Jewell RJ. Introduction. In: Jewell RJ, Fourie AB, Lord ER,
editors. Paste and thickened tailings: a guide. Perth (Australia):
to the research discussed in this paper. Their support UniPrint; 2002. p. 1e8.
is gratefully acknowledged. Senior Technical Officers [14] Jones H. Designer Waste. PASTE TECHNOLOGY 2000, an
Mr. Warren O’Donnell and Mr. Stuart Petersen assisted international seminar on the production and disposal of
in most of the laboratory test work carried out on the thickened/paste tailings for mine backfill or on the surface. Perth
bricks and hydraulic fills. Our regular discussions with Australia; 2000. p. 1e6.
[15] Mitchell RJ, Wong BC. Behaviour of cemented tailings sands.
Mr. Richard Cowling of Cowling Associates were very Canadian Geotechnical Journal 1982;19(3):289e95.
valuable throughout our mining research. [16] Winch C. Geotechnical characteristics and stability of paste
backfill at BHP Cannington Mine. B.E. Hons. Thesis, James
Cook University, Townsville (Australia); 1999.
References [17] Aubertin M, Li L, Arnoldi S, Belem T, Bussiere B, Benzaazoua
M, Simon R. Modelling arching effects in narrow backfilled
[1] Grice T. Recent mine developments in Australia. In: Proceedings stopes with FLAC. In: Proceedings of the 3rd international
of the 7th international symposium on mining with backfill symposium on FLAC and FLAC3D numerical modelling in
(MINEFILL); 2001. p. 351e7. geomechanics. Ontario (Canada); 2003. p. 1157e64.
[2] Bloss ML, Chen J. Drainage research at Mount Isa Mines [18] Bloss ML. Prediction of cemented rock fill stability e design
Limited 1992e1997. In: Proceedings of Minefill ’98. Brisbane procedures and modelling techniques. PhD Thesis, The University
(Australia); 1998. p. 111e6. of Queensland, Australia; 1992.
[3] Herget G, De Korompay V. In situ drainage properties of [19] Rankine RM, Rankine KJ, Sivakugan N, Karunasena W, Bloss
hydraulic backfills. Research and innovations, CIM special ML. Geotehnical characterisation and stability analysis of BHP
volume; 1978. p. 117e23. Cannington pastefill. In: Proceedings of the 15th international
[4] Rankine KJ, Sivakugan N, Rankine KS. Laboratory tests for conference on soil mechanics and geotechnical engineering.
mine fills and barricade bricks. In: Farquhar G, Kelsey P, Istanbul (Turkey); 2001. p. 1241e4.
Marsh J, Fellows D, editors. Proceedings of the 9th Australia [20] Clough GW, Sitar N, Bachus RC, Rad NS. Cemented sands
New Zealand conference on geomechanics. Auckland; 2004. under static loading. Journal of Geotechnical Engineering Di-
p. 218e24. vision, ASCE 1981;107(6):799e817.
[5] Kuganathan K. Mine backfilling, backfill drainage and bulkhead [21] Rankine KJ, Rankine KS, Sivakugan N. Three-dimensional
construction e a safety first approach. Australia’s mining drainage modelling of hydraulic fill mines. In: Leung CF,
monthly February, 2001;58e64. Phoon KK, Chow YK, Teh CI, Yong KY, editors. Proceedings
[6] Brady AC, Brown JA. Hydraulic fill at Osborne mine. In: of the 12th Asian regional conference on soil mechanics and
Proceedings of 8th AUSIMM underground operators conference. geotechnical engineering. Singapore; 2003. p. 937e40.