Paste Back Filling Propertis

8/4/2019 Paste Back Filling Propertis

1/14

1 INTRODUCTION

The use of cemented paste backfill (CPB) is an in-creasingly important component of undergroundmining operations and is becoming a standard prac-tice for use in many cut-and-fill mines around theworld (Landriault et al. 1997, Naylor et al. 1997).Backfill material is placed into previously minedstopes to provide a stable platform for the miners towork on and ground support for the walls of the ad-

jacent adits as mining progresses by reducing theamount of open space which could potentially be fillby a collapse of the surrounding pillars (Barret et al.1978). The use of underground paste backfill pro-vides ground support to the pillars and walls, butalso helps prevent caving and roof falls, and en-hances pillar recovery, which enhances productivity(Coates 1981). Thus, the CPB placement provides anextremely flexible system for coping with changesin geometry of the orebody, that result in changingstope width, dip, and length (Wayment 1978). Themethod of the fill delivery depends upon the amountof energy required to deliver the backfill material

underground which depends on its distribution cone(Arioglu 1983). The CPB is usually transported un-derground through reticulated pipelines.

Paste backfill is composed of mill tailings gener-ated during mineral processing which are mixedwith additives such as Portland cement, lime, pul-verized fly ash, and smelter slag. The purpose of the

binding agents is to develop cohesion within CPB sothat exposed fill faces will be self-supporting whenadjacent stopes are extracted (Mitchell 1989). Withthe current low metal prices, the survival of manymines depends on their ability to maximize produc-

tivity while minimizing costs. At underground cut-and-fill operation, the costs associated with backfill-ing must be looked at critically so that potential costsavings can be identified (Stone 1993). Backfillingis expensive in some ways, but indispensable formost underground mines to provide ground supportfor mine safety and mining operations. Therefore,the fill should be cost effective and capable ofachieving the desired ground support and stability.

Analysis of the fill stability must consider the

geometric boundaries of the fill for the best eco-nomic use of CPB. Mine openings and exposed fillfaces in large underground mines vary in shape fromhigh and narrow to low and wide. Additionally, wallrock next to the backfill may be either steeply dip-

ping or relatively flat-lying (Mitchell 1989). Thestoping sequence can be modified to reduce thenumber of CPB-filled stopes, or the stope geome-tries could be revised to reduce the strength requiredof CPB exposures (Stone 1993).

This paper is an overview of the use of CPB forunderground ground support in mining operations,from preparation to placement underground. The pa-

per will first briefly introduce the notion of archingeffects and their importance in stability analysis offilled stopes. This will be followed by presenting thedesign of the required fill strength from reviews ofvarious current design methods. The paper will dis-cuss the optimization of CPB-mix designs (as ameans to reduce costs and improve fill strength) fol-lowed by a discussion on the rheological propertiesof CPB. Finally, the paper will discuss CPB deliverysystems and underground placement of CPB.

An overview on the use of paste backfill technology as a ground supportmethod in cut-and-fill mines

T. Belem, M. BenzaazouaUniversit du Qubec en Abitibi-Tmiscamingue, Dpt. des Sciences appliqus, Rouyn-Noranda, Canada

To cite this paper:

Belem T., Benzaazoua M. 2004. An overview on the use of paste backfill technology as a ground support method in cut-and-fill

mines. Proceedings of the 5th Int. Symp. on Ground support in Mining and Underground Construction. Villaescusa & Potvin (eds.),

28-30 September 2004, Perth, Western Australia, Australia, Tayler & Francis Group, London, pp. 637 650 .

ABSTRACT: This is a general overview on the use of paste backfill for ground support in underground min-ing operations and as such, concerns backfill design parameters (internal pressure development, requiredstrength and mix optimisation), its pumping qualities (consistency and rheological) and delivery to an under-ground operation through pipelines. Emphasis was placed on the optimization of the paste backfill mix for

backfill design, work safety and expense to the mining operation. This is because of the 20% representativecosts related to backfilling, 15% represents the binder costs.

637


2/14

2 DESIGN OF THE HORIZONTAL PRESSUREON THE FILLED STOPE SIDEWALLS

In general, self-support stresses govern backfill de-sign and the traditional design has been that of a freestanding wall, requiring a uniaxial compressivestrength (UCS) equal to the overburden stress at the

bottom of the filled stope. However, in many cases,the adjacent rock walls can actually help support the

fill through boundary shear and arching effects.Therefore, the backfill and rock walls can be mutu-ally supporting (Mitchell 1989). In backfilled stopes,when arching occurs (which is the case in manymines) the vertical pressure at the bottom of filledstope, an analogy similar to a trap door, is less thanthe weight of overlying fill (overburden weight) dueto horizontal pressure transfer (Martson 1930, Ter-zaghi 1943). This pressure transfer is due to fric-tional and/or cohesive interaction between the filland wall rock. When the pillars or stope walls beginto deform into the filled opening the fill mass will

provide lateral passive resistance. Passive resistanceis defined as the state of maximum resistance mobi-lized when force pushes against a fill mass and themass exerts resistance to the force (Hunt 1986).

The magnitude of pressure transferred horizon-tally to the sidewalls can be included into the designfor the required fill strength. Horizontal pressures af-fected by the fill arching can be evaluated by fouranalytical solutions which account for the existenceof cohesion at the fill-sidewall interface and/or thefrictional sliding along the sidewalls. These solu-tions are the Martson's model and its modified ver-sion, the Terzaghi's model and a proposed 3D

model.

2.1 Martson's cohesionless model

Martson (1930) developed a two-dimensional archsolution to predict the horizontal pressure (h) alongthe sidewalls of the pillars as follows:

=B

HKB ah

'2exp1

'2(1)

ahv K= (2)

( )2/45tan 2 =aK (3)

where = fill bulk unit weight (kN/m3);B = width ofstope (m); H= total height of filled stope (m); '=tan ' = coefficient of sliding friction between filland sidewalls; '() = angle of wall friction (may beassumed between 1/3 to 2/3); = angle of internalfriction of fill (degree); v = vertical pressure at thefloor of the stope (kPa); Ka = coefficient of activeearth pressure (see Eq. 3);.

2.2 Modified Martson's cohesionless model

Aubertin et al. (2003) proposed a modified versionof the Martson's two-dimensional arch solutionwhich was originally defined using active earth pres-sure (Ka) and wall sliding friction. The modified ver-sion for predicting the horizontal pressure (hH), at adepthH, along the sidewalls of the pillars is given asfollows:

=

BKHB f

f

hH'tan2exp1

'tan2(4)

KhHvH = (5)

where = fill bulk unit weight (kN/m3);B = width ofstope (m); H= total height of filled stope (m); 'f =fill effective angle of internal friction (degree); vH=vertical pressure at the floor of the stope (kPa); K=coefficient of fill pressure. K will correspond tothree different states (Ka, K0, Kp) given by the fol-lowing relationships:

( )( )

+==

==

==

2/'45tan

2/'45tan

'sin1

2

2

0

fp

fa

f

KK

KK

KK

(6)

where K0 = coefficient of fill pressure at rest or inplace (0.4 to 0.6);Ka = coefficient of active fill pres-sure (0.17 to 1);Kp = coefficient of passive fill pres-sure (1 to 10).

However, in a filled stope the active fill pressurecondition (Ka) seems improbable. In Equation 6, the

coefficient of fill at rest pressure can alternatively beevaluated using this well known relationship as fol-lows:

=1

0K (7)

where = Poisson's ratio of the fill (0.3 0.4).

2.3 Terzaghi's cohesive model

Terzaghi (1943) also developed a two-dimensionalarch theory for predicting the horizontal pressure

(h) along the pillar walls and this is given by:

( )

=

B

KHcBh

tan2exp1

tan2

2(8)

Khv = (9)

( )[ ]+= 2tan211K (10)where = fill bulk unit weight (kN/m3); c = cohesivestrength of fill (kPa); B = width of stope (m); H=depth below fill toe (m); tan = coefficient of inter-

638


3/14

nal friction of fill; = angle of internal friction offill (degree);K= coefficient of fill pressure (see Eq.10).

2.4 Proposed 3D model

Belem et al. (2004) proposed a three-dimensionalmodel (companion paper) which implicitly takesinto account the arching effects to predict the hori-zontal pressures, both the longitudinal pressure (x)and the transverse pressures (y). The model is givenas follows:

( )( )

+

=

B

zh

LB

zHH mmx

)(2exp1

3(11)

( )

+

=

B

zh

LB

zHH mmy

)(2exp1

185.0(12)

( ) yvz == (13)

where = bulk unit weight of the fill (kN/m3);Hm =total height of filled stope (m);z= elevation point ofmeasurement (m):z= 0 at the floor of the stope, z=

Hm at the fill toe (zhHm);B = width of stope;L= strike length of stope (m).

3 DESIGN FOR CPB REQUIRED STRENGTH

The required strength for paste backfill dependsupon its intended function. For a ground supportrole, the required uniaxial or unconfined compres-sive strength (UCS) of the fill should be at least 5

MPa whereas, for free-standing fill applications, theUCS can be commonly lower than 1 MPa (Stone1993). Previous work indicates that the UCS of thefill mass can range to between 0.2 MPa and 5 MPa,while the UCS of the surrounding rock mass is be-tween 5 MPa and 240 MPa.

3.1 Vertical support of backfill

The mechanical effects of fill are different fromthose of primary ore pillars. Research and in situtesting have shown that fill is incapable of support-ingc the total weight of overburden (v) and actsonly as a secondary support system (Cai 1983). Thefill rigidity can range from 0.1 GPa to 1.2 GPa whilethe surroundding rock mass rigidity varys from 20GPa to 100 GPa. As discussed by Donavan (1999), itis possible to assume that any vertical loading will

be a result of roof deformation (Fig. 1) and that thedesign UCS can be estimated by the following rela-tionships:

FSH

HEUCS

p

p

pdesign

= (14)

where Ep = rock mass or pillar elastic modulus(kPa); Hp = strata length variation (m);FS= factorof safety.

v

Open

stope

Open

stopePillar

Fill

block

Fill

block

Fill

block

H_pillar

H_fill

v

Open

stope

Open

stopePillar

Fill

block

Fill

block

Fill

block

H_pillar

H_fill

Figure 1. Schematic representation of vertical loading of a pil-lar into the fill mass.

When the stope walls deform before backfilling, themaximum load will probably never approach the to-tal weight of the deformed overlying strata(Donavan 1999) and the design UCS can be esti-mated by following relationships:

( )FSHkUCS ppdesign = (15)

where k = scaling constant which must vary from

0.25 to 0.5; p = strata unit weight (kN/m3); Hp =strata height below surface (meter); FS = factor ofsafety.

Numeric modeling can also be used to determine therequired stiffness or strength of a CPB to preventsubsidence due to the roof deformation. The resultscan be very useful in indicating the amount of the

paste backfill desired. Modeling can be done witheither of the FLAC (2D and 3D) codes. Physicalmodeling, such as with a centrifuge, also can offeran alternative to numeric modeling, but its applica-

tion is usually limited to simple gravitational modelswithout high tectonic or in situ horizontal stresses(Stone 1993).

3.2 Development through backfill mass

When one wants to cut an access gallery to a neworebody through the paste backfill (Fig. 2), it is nec-essary to consider the original design criteria. Thisdesign considers a fill mass to be more than twocontiguously exposed faces after blasting adjacent

pillars or stopes. As a result, the walls confining the

639


4/14

fill are removed and the fill block is subjected togravity loading similar to a uniaxial compressionsample (Yu 1992). The design UCS can be estimated

by the following relationships:

( )FSHUCS ffdesign = (16)

where f = fill bulk unit weight (kN/m3);Hf = to-

tal fill height (m);FS= factor of safety.

Filled

stope

Filled

stopeFill

GalleryFilled

stope

Filled

stopeFill

Gallery

Figure 2. Digging an access gallery through the fill mass.

3.3 Pillar recovery

In order to maximize ore recovery, it is very com-mon to return for mine pillars after primary ore re-covery. While this is being done, large verticalheights of massive paste backfill may be exposed.For delayed paste backfill, as used in open stopingoperations, the fill must be stable when free-standingwall faces are exposed during pillar recovery (Fig.

3). It is necessary that the fill has sufficient strengthto remain free-standing during and after the processof pillar extraction by resisting the blast effects. Fig-ure 3 illustrates a possible failure mechanism whichcan occur after a stope blast. Depending upon themining schedule, high strength for such engineeringmaterials may not be required for the short term(Hassani & Archibald 1998).

In the absence of numeric modeling, many mineengineers still rely on two-dimensional limit equilib-rium analyses along with a calculated safety factor(FS) to determine fill exposure stability. Theseanalyses typically result in an over-conservative es-timate of the limiting strength (Stone 1993) whichincrease the costs of backfilling operations.

In recent years, however, 2D- and pseudo-3D empirical models have been developed to account forarching effects, cohesion and friction along side-walls (Mitchell et al. 1982, Smith et al. 1983,Arioglu 1984, Mitchell 1989a &b, Mitchell & Ro-ettger 1989, Chen and Jiao 1991, Yu 1992). Allthese design methods use the concept of a confinedfill block surrounded by the wall rock.

ore

Fill block

PillarFreeface

Possible failure plane

Secondary

stoping

ore

Fill block

PillarFreeface

Possible failure plane

Secondary

stoping

Figure 3. Fill block failure mechanism during secondary stopemining.

3.3.1 More than two exposed facesEquation (16) should be used if there are more thantwo contiguously exposed faces after blasting adja-cent pillars or stopes (Fig. 4).

ExposedfaceExpo

sedface

H

vv

ExposedfaceExpo

sedface

H

vv

Figure 4. Schematic of a fill block with three exposed faces.

3.3.2 Narrowly exposed fill faceThis design method accounts for arching effects onconfined fill by adjacent stope walls (Fig. 5) usingTerzaghi's vertical pressure model (Eq. 9). Based on2D finite element modeling, Askew et al. (1978)

proposed the following formula to determine the de-sign fill compressive strength:

FSB

HK

B

c

K

BUCSdesign

=tan2

exp12

tan2

25.1

(17)where B = width of stope; K = coefficient of fill

pressure (see Eq. 10); c = cohesive strength of fill(kPa); = angle of internal friction of fill (degree); = bulk unit weight of the fill (kN/m

3); H = total

height of filled stope (m);FS= factor of safety.The fill cohesion (c) and its angle of internal friction() can be obtained from triaxial tests performed onlaboratory orin situ backfill samples.

640


5/14

H

L

v < H

Pillar

h

0

H

Fill

h

h

B

Expose

d

face

Arching

effect

H

L

v < H

Pillar

h

0

H

Fill

h

h

B

Expose

d

face

Arching

effect

Figure 5. Stability analysis of a narrowly exposed fill face.

3.3.3 Exposed frictional fill faceThis design refers to an exposed fill where both op-

posite sides of the fill are against stope walls (Fig.6). By assuming that there is shear resistance be-tween the fill and stope walls due to the fill cohe-sion, the design UCS can be evaluated by the fol-lowing relationship (Mitchell 1982):

( )FS

L

BHcL

UCSdesign

+

+

=2

45sin2

45tan2

2(18)

where = fill bulk unit weight (kN/m3); c = cohesivestrength of fill (kPa);L = strike length of stope (m);

B = width of stope (m);H= total height of fill (m); = angle of internal friction of fill (degree);FS= fac-

tor of safety.

Again, the fill cohesion (c) and its angle of internalfriction () can be obtained from triaxial tests per-formed on laboratory orin situ backfill samples.

L

B

H

He

Wall shear resistance

= cBHe (kN)

Direction of sliding

along failure plane

= tan2

BHHe

245

+=

(He)LB

= block weight

L

B

H

He

Wall shear resistance

= cBHe (kN)


along failure plane

= tan2

BHHe

245

+=

(He)LB

= block weight

Figure 6. Confined block with shear resistance mechanism (af-ter Mitchell et al. 1982).

3.3.4 Exposed frictionless fill faceThe compressive strength of paste backfill is mainlydue to the binding agents and any strength contrib-uted from friction can be considered negligible for

the long term (i.e. = 0). For a frictionless material(Fig. 7), cohesion is assumed to be half of the UCS(c = UCS/2). Thus, the design UCS can be evaluated

by the following relationship proposed by Mitchellet al. (1982):

FSL

BHcL

UCSdesign

)45sin(2

)2(

= (19)

where = fill bulk unit weight (kN/m3); c = cohesivestrength of fill (kPa); B = width of stope (m); L =strike length of stope (m); H = total height of fill(m);FS= factor of safety (ca. 1.5).

In Equation 19, the fill cohesion (c) can be obtainedfrom laboratory confined compression tests on back-fill samples.

The stability of a free standing backfill (Fig. 7)can also be determined from physical model tests.Based on centrifugal modeling tests, Mitchell (1983)

proposed a formula for evaluating the design UCS

which is given by:

FSHL

LHUCSdesign +

= (20)

where = fill bulk unit weight (kN/m3); L = strikelength of stope (m);H= total height of fill (m);FS=factor of safety.

L

B

H

He


along failure plane

= tan

2

BHHe

245

+=

No wall shear

resistance(He)LB

= block weight

L

B

H

He


along failure plane

= tan

2

BHHe

245

+=

No wall shear

resistance(He)LB

= block weight

Figure 7. Confined block without shear resistance mechanismof frictionless fill (adapted from Mitchell et al. 1982).

3.4 Ground support

After passive resistance has been mobilized by thefill, the strength increase in the surrounding pillarswill be equal to the magnitude of the passive fill

pressure. So, the main stabilizing effect of the fill isto give increased lateral confinement pressure to the

pillars (Fig. 8). The compressive strength of the pil-lar increases according to the following formula(Guang-Xu & Mao-Yuan, 1983):

[ ] ) ppfppp KKqHUCSUCS ++=' (21)

641


6/14

+=

245tan 2

f

fpK (22)

+=

245tan 2

p

ppK (23)

where UCS'p = pillar compressive strength with fill(kPa); UCSp = pillar strength before the stope filling

(kPa); = fill bulk unit weight (kN/m3

); q = sur-charge loading (kPa);H= total height of fill (m); f= angle of internal friction of fill (degree); p = an-gle of internal friction of pillar (degree);Kp-f = coef-ficient of passive pressure of the fill; Kp-p = coeffi-cient of passive pressure of the pillar.

Fill block pillar

h_fill

h_pillar

Fill block

Fill block pillar

h_fill

h_pillar

Fill block

Figure 8. Schematic of pillar confinement by the fill block.

3.5 Working platform

For cyclic backfilling operations, as in cut-and-fill

stoping, the fill in each operation must serve as a platform for both mining equipment and personneland typically requires high strength development forthe short term. A standard bearing capacity relation-ship that has been developed from civil engineeringtechniques for design of shallow foundations has

been found to be applicable to paste backfill. The filltop surface bearing capacity, Qf (kPa), can be deter-mined using Terzaghi's expression (Craig 1995):

cf cNBNQ 2.14.0 += (24)

) = tan18.1 qNN (25)

)

=

tan

1qc

NN (26)

( )

+= tanexp2

45tan 2qN (27)

where = bulk unit weight of the fill (kN/m3); c =cohesive strength of fill (kPa); B = width of squarefooting at surface contact position (m); N = unitweight bearing capacity factor; Nc = cohesion bear-

ing capacity factor;Nq = surcharge bearing capacityfactor; = angle of internal friction of fill (degree).

Equation 24 assumes that backfill bearing is by asquare footing, which is a reasonable representationof the footprint of a mine vehicle tire (Hassani &Archibald 1998). Equation 25 was developed byHansen (1968). For the mine vehicles (Fig. 9), thecontact width, B, corresponds to the tire contact

width and can be determined by the following rela-tionship (Hassani & Bois 1992):

p

FB t= (28)

whereFt = tire loading force (kN); p = tire air pres-sure (kN/m2).

B B

Fill block

B B

Fill block

Figure 9. Schematic of working platform (adapted from Has-sani & Bois 1992).

4 OPTIMIZATION OF PASTE BACKFILL MIX

DESIGNS

Once the required strength has been determined, themix variables can be optimized to provide the de-sired mix, which achieves the target strength withthe lowest cement usage. The mix variables underconsideration include the binder content and type,mill tailings grain size distribution and mineralogy,solids concentration, and the mixing-water chemis-try. For the design of a certain uniaxial compressivestrength (UCSdesign), these variables can be adjustedto produce an optimal mix design (Stone 1993).

The other essential requirement is that backfillmust be inexpensive. Typical costs of backfill rangefrom $2 to $20 per cubic meter, depending on theservice required. These costs can be a significantcontribution to the operating costs of the mine.Where cemented backfill is used, these costs tend to

be between 10 and 20% of the total operating cost ofthe mine and cement represents up to 75% of thatcost (Grice 1998). Optimization of CPB-mix designscan reduce binder usage and can offer significantcost savings (Fall & Benzaazoua 2003).

642


7/14

4.1 Laboratory optimization of CPB mix designs

Figure 10 shows the main components which can af-fect the quality of cemented paste backfill such asthe binding agents, mill tailings mineralogy, milltailings grain size, the density and solids percentageof tailings and finally, the mixing-water geochemis-try (Benzaazoua et al. 2002).

TailingsSulfides %

Grain size distribution

Density, specific gravity

Binding agents

CaSO4

%

Paste

backfill

Mixing waterSO

42- content

Soluble lime

Additives

(SiO2

+Al2O

3) /

(CaO + MgO)

Binder %(3 wt.% - 7 wt.% )

Water content(expected slump)

Solids %, Cw(78% -85%)

TailingsSulfides %

Grain size distribution

Density, specific gravity

Binding agents

CaSO4

%

Paste

backfill

Mixing waterSO

42- content

Soluble lime

Mixing waterSO

42- content

Soluble lime

Additives

(SiO2

+Al2O

3) /

(CaO + MgO)

Binder %(3 wt.% - 7 wt.% )

Water content(expected slump)

Solids %, Cw(78% -85%)

Figure 10. Schematic diagram illustrating the different compo-

nents of cemented paste backfill (Benzaazoua et al. 2002).

Each component plays an important role for the backfill transportation, its delivery and its strengthdevelopment in the course of curing time. Typical

binder percentages are 3 to 7% by weight of thepaste fill. Numerous laboratory test results have re- ported that the backfill strength is a function ofbinder content for a given curing age, (Fig. 11), butthis relationship is specific to each mine (e.g. Ben-zaazoua et al. 1999, 2002, 2004).

1.5 3 4.5 6 7.5

0

500

1000

1500

2000

2500

3000

35004000

4500

Binder content (wt.%)

UCS(kPa)

14 days

28 days

56 days

91 days

118 days

Binder type:

20% cement + 80% slags

1.5 3 4.5 6 7.5

0

500

1000

1500

2000

2500

3000

35004000

4500

Binder content (wt.%)

UCS(kPa)

14 days

28 days

56 days

91 days

118 days

Binder type:

20% cement + 80% slags

Figure 11. Example of variation of UCS as a function of binder

content at different curing times of 14, 28, 56, 91 and 118 days(after Benzaazoua et al. 2003).

4.1.1 Cement and others bindersHardening of the fill occurs as bonds are formed be-tween fill particles at grain contact points. Many dif-ferent types of binding agents are used, but the mostcommon is ordinary Portland cement (OPC). Ad-mixtures with pozzolanic materials are also used tocurb costs by reducing the amount of Portland ce-ment needed for hardening. Fly ash (FA) and smelterground blast furnace slags (BFS) are the most popu-lar pozzolans used as admixtures. The results of ce-

ment dissolution tests performed by Benzaazoua etal. (2004) showed that in either concrete or mortar,the hardening processes within the pastefill are notonly due to the cement hydration but also to the pre-cipitation of hydrated phases from the pore water ofthe paste. Figure 12 illustrates that paste backfillhardening occurs in two main stages: the first stage(dissolution-hydration) which is dominated by thedissolution reactions and the second stage (precipi-

tation and hydration) which is characterized by the precipitation reactions and direct hydration of thebinder. More details on this subject can be found inBenzaazoua et al. (1999, 2002, 2004).

Increasing curing time

Dissolvedcement

Hydrates

Precipitates

Anhydrouscement

Dissolvedcement

Hydrates

Precipitates

Anhydrouscement

Phase I Phase II

Hydrates

Dissolved cementdissolution

precipitation

hydration

cement

Increasing curing time

Dissolvedcement

Hydrates

Precipitates

Anhydrouscement

Dissolvedcement

Hydrates

Precipitates

Anhydrouscement

Phase I Phase II

Hydrates

Dissolved cementdissolution

precipitation

hydration

cement

Figure 12. Schematic illustration of the time-depending impor-tance of the reactions of dissolution and precipitation of thebinder in the hardening process of the pastefill (after Benzaa-zoua et al. 2004).

Water is necessary to ensure that proper hydration ofthe cement occurs. If proper hydration of the cementdoes not occur, the fill will not meet its requiredstrength and stiffness. Since tailings backfill is fairly

saturated to begin with and additional water is usu-ally required to pump it underground, the water con-tent of tailings backfill is always in far excess ofwhat is required for hydration of the Portland ce-ment. The main concern then is the pH of the waterand the amount of sulfate salts present in the water.Acidic water and sulfate salts can attack the cement

bonds within the fill, leading to a loss of strength,durability, and stability (Benzaazoua et al. 2002,2004).

Figure 13 shows that when using cement-slag binder with the same tailings sample mixed with

three different waters, the strength development isslow for all three waters for a curing time of 14 days(Benzaazoua et al. 2002). Beyond this curing dateand at a curing time of 28 days UCS reached a valueof about 1600 kPa with the sulfate-free waters (mu-nicipal and lake waters) and only 1000 kPa with thesulfate-rich water (mine A process water).

643


8/14

0

200

400

600

800

1000

1200

1400

1600

1800

0 7 14 21 28Curing time (days)

UCS(kPa)

Binder type

5 wt.% (30:70 of OPC-Slag)

Municipal water

Lake water

Mine A process water

Key

0

200

400

600

800

1000

1200

1400

1600

1800

0 7 14 21 28Curing time (days)

UCS(kPa)

Binder type

5 wt.% (30:70 of OPC-Slag)

Municipal water

Lake water

Mine A process water

Key

Figure 13. Effect of the mixing water on strength developmentwithin paste backfill mixtures with mine A tailings (after Ben-zaazoua et al. 2002).

4.1.2 Mixing processSome amount of water is added to set the resultant

paste backfill to attain the desired slump value. Theslump must vary from 12.4 to 25.4 cm (5 to 10 in)which correspond to solids concentrations of 78% to82% by weight. Slump is a measure of the drop inheight a material undergoes when it is released froma cone-shaped slip mold. Determination of slump

provides a way to characterize a material's consis-tency that can be related to its transportability.

The resultant paste backfill mixtures were pouredinto plastic cylinders 10.25 cm in diameter and 20.5cm height, sealed and cured in a humidity-controlledchamber at approximately 90%-100% relative hu-midity (similar to underground mine working condi-tions). The pastefill samples were then subjected touniaxial compression tests for periods of 7, 14, 28,56 and 91 days.

4.2 CPB preparation at the backfill plant

Figure 14 shows a typical flow chart for a backfillplant. The final mill tailings are first fed to a high-capacity thickener to increase their solids percentageto approximately 55% to 60% by weight. To aid fil-tration some flocculent is added. The thickened tail-ings are then pumped from the thickener to a high-capacity holding tank (after cyanide destruction).From the surge tank, the thickened tailings are fed

by a gravity circuit to disc filters operating alone or

in parallel to produce a filter cake with a solids per-centage of approximately 70% to 82%. The filtercake is then discharged onto a belt (or reversible)conveyor and is then fed to a screw feeder forweighing. Filter cake batches are mixed in a spiral(or screw) mixer with cement and water is added to

produce a paste with a specified slump (127 to 254cm). The mixed paste is dropped into a surge hopperand discharged underground under vacuum (bygravity or using concrete pump).

Figure 14. Paste backfill plant flow sheet at Louvicourt mine inCanada (after Cayouette 2003).

5 RHEOLOGICAL PROPERTIES OF CPB

Paste backfill consists of the full size fraction of thetailings stream prepared as a high slurry density. Theslurry behaves as a non-Newtonian fluid, whichmeans that it requires an applied force to commenceflowing (Fig. 15).

a) No flow

(Vertical head

= yield stress)

Hopper

open

stope

Pressure

(2 MPa)4002 00 6 00

8000

Hopper

400

2 00 6 00

8000

Pressure(3 MPa)

Stope

b) Flow

(Vertical head

> yield stress)

a) No flow

(Vertical head

= yield stress)

Hopper

open

stope

Pressure

(2 MPa)4002 00 6 00

8000

400

2 00 6 00

8000

Hopper

400

2 00 6 00

8000

400

2 00 6 00

8000

Pressure(3 MPa)

Stope

b) Flow

(Vertical head

> yield stress)

Figure 15. Schematic illustrating the yield stress of paste back-fill flowing in pipeline (after Revell 2000).

Toothpaste is an example of a Non-Newtonian fluidthat is commonly used and the yield stress ( appliedforce ) explains why you have to squeeze the tooth-

paste out of the tube. The paste has a higher viscos-ity and exhibits plug flow when transported in a

pipe. The outer portions of the slurry shear againstthe sidewall of the pipe and the central core travelsas a plug (Grice 1998). The flow of paste backfill in

pipeline is entirely governed by their rheological properties. Rheology is the science about flow and

deformation of matter.

5.1 Rheological models of CPB

The main mode for paste backfill flow in pipelines isthe full-fall. Full-pipe flow refers to the situationwhere the flowing paste forms a continuum andthere is no air-filled gap or discontinuities (vacuumholes) anywhere in the pipeline segment underconsideration (Li et Moerman. 2002).

The most fundamental relationship in the rheologyof a Non-Newtonian fluid is that between the shear

644


9/14

rate, & (s-1) and pipe wall shear stress, w(Pa). Oncethis relationship is known, the behaviour of the fluidin all flow situations can be deduced. All Non-

Newtonian fluid rheology can be derived from themost general Herschel-Bulkley model given by:

n

n

w kdr

dVk +=

+= &00 (29)

where 0 = yield stress (Pa), k= consistency parame-ter or viscosity (Pa.s), (dV/dr) = paste angular veloc-ity or shear rate (s

-1); r = point of velocity profile

(m),R = radius of the pipe (m), V= paste linear ve-locity (m/s), n = flow parameter.

For the Newtonian fluids, 0 = n = 0; for the pseudo-plastic fluids, 0 = 0 and n < 0; for dilatant fluids, 0= 0 and n > 0; for Bingham plastic fluids, 0 > 0, n =0 and k= = plastic viscosity in Pa.s (Fig. 16); foryield pseudoplastic fluids, 0 > 0 and n > 0 (Fig. 16);for yield dilatant fluids, 0 > 0 and n > 0.

Wallshearstressw

(Pa)

0

Shear rate & (1/s)

B

p

0

Bingham plastic

Pseudoplastic

fluid

Wallshearstressw

(Pa)

0

Shear rate & (1/s)

B

p

0

Bingham plastic

Pseudoplastic

fluid

Figure 16. Rheology models for time-independent fluids.

Paste backfills are non-Newtonian fluids and theirrheology is approximately time-independent duringits transport in pipeline. Most paste backfill show anappreciable yield stress and are Herschel-Bulkleyfluids (Eq. 29). Some paste backfills are Bingham

plastic in limited shear rate ranges. Others are yieldpseudoplastic or yield dilatant, with the former morecommon than the latter.

The relationship between the pseudo shear rate,8V/D, and the shear stress at the pipe wall, w, isgiven by:

14

00

4

3

14

3

41

8

4

+

=

PD

L

PD

L

D

V

L

PDw

(30)

where 0 = yield stress (Pa), = paste plastic viscos-ity (Pa.s), P= differential pressure in the pipe (Pa);

D = internal pipe diameter (m),L = pipe length (m);V= paste laminar velocity (m/s).The effective pipes diameter (D) for paste backfilltransport is ranged between 10 cm and 20 cm (4 and8 in). Paste flow velocity varies from 0.1 m/s to 1m/s. The practical pumping distance of paste can

reach 1000 m longitudinally (Lh) and unlimited ver-ticallyLv).

5.2 Standard measurements of the CPB consistency

In practice, it is not easy to obtain the truerheological properties of pastes due to thecomplexity of the experimental devices. This makesdifficult, even impossible, the determination or the

prediction of a pastes viscosity which depends onseveral factors. That is why the standard slump test(used in concrete experiments) is widely used, dueto its simplicity, to determine paste backfillconsistency. Slump is a measure of the drop inheight of a material when it is released from a trun-cated metal cone, open at both ends and sitting onhorizontal surface (Fig. 17). Determination of theslump provides a way to characterize a material'sconsistency that can be related to its transportability(Clark et al. 1995). According to Landriault et al.(1997), the ideal slump of the paste must be in arange between 150 mm (6 in) and 250 mm (10 in) to

facilitate the flow of cemented paste backfill by itspumping underground.

(a) (b)

H0

S(mm)

= slump

Before

After

H0

S(mm)

= slump

Before

After

Figure 17. Paste backfill consistency measurement by slump

tests: a) slump cone mold; b) schematic view of the slump test.

Solids concentration is often used to compare thecomposition of mixes, particularly in batch. Al-though solids percentage does not provide a directindication of a material's consistency, in some casesit can be correlated to the slump, which does.

In order to achieve the same mix consistency from batch to batch, consistency can be measured bymonitoring the electrical power used by a motorturning the paddles of a mixer. The mixer is startedand water is added until the power required by themotor corresponds to the target power for the mixconsistency desired (Brackebusch 1994, Landriault& Lidkea 1993). Using this arrangement requiresonly that slump be correlated to consistency andconsistency be correlated to power. It is also possi-

ble to predict what pressure gradient a mix will pro-duce based on power once a correlation has been es-tablished between slump and pressure loss.

645


10/14

5.3 Alternative methods for rheological factormeasurements

To correctly define the rheology of paste backfillboth the yield stress (0) and the viscosity () needto be measured. Most current tests measure only onerheological factor. The relationship between the fac-tor measured and either of the two fundamentalrheological parameters is not obvious. In most cases,0 and cannot be calculated from the factor meas-ured, but can only be assumed to be related. Accord-ing to Ferraris (1999), slump, penetrating rod and K-slump tests are related to the yield stress (0) be-cause they measure the ability of paste to start flow-ing. The remolding test, LCL apparatus, vibratingtesting apparatus, flow cone, turning tube viscome-ter, filling ability and Orimet apparatus are related tothe viscosity because they measure the ability of

paste to flow after the applied stress (vibration orgravity) exceeds the yield stress.

Recently, a modification of the slump cone wasdeveloped to allow the measurement of viscosity

(Ferraris & de Larrard 1998). As mentioned earlier,the standard slump test can only be correlated withthe yield stress (0). The modification consists ofmeasuring not only the final slump height but alsothe speed at which the concrete (or paste backfill)slumped. The method consists of measuring the time(T) for a plate resting on the top of the concrete toslide down with the concrete (or paste backfill) adistance of 100 mm (Fig. 18).

Figure 18. Schematics of the modified slump cone test basedon slumping time Tmeasurement (after Ferraris & de Larrard1998).

The yield stress, 0, can be calculated from the finalslump (S), using the following empirical equation

proposed by Ferraris & de Larrard (1998):

( ) ba

SH +=0 (31)

where = paste density (kg/m3); S = final slump(mm); a, b = material constants,H= 300 mm is thecone height. For the concrete paste, a = 347 and b =212.

From a range of paste backfill slump values (130 250 mm), the viscosity can be determined from the100 mm slump time (T) using an empirical equationthat was developed by Ferraris & de Larrard (1998):

Tk= (32)

where = viscosity (Pa.s); k= material constant (k=0.025 for concrete); = paste density (kg/m3); T=slumping time (s).

Other authors (Nguyen & Boger 1985) have sug-gested adapting the laboratory vane shear test for themeasure of paste yield stress (0). This test allowsobtaining a torqueangular deformation curve of the

paste whose peak corresponds to the maximumtorque (m). If these these parameters are known, theyield stress can then be calculated by the followingrelationship:

+

=

3

1

2

30

D

HD

m (33)

where 0 = paste yield stress (Pa), m = maximumpeak torque value (N.m),D = vane diameter (cm);H= vane height (cm).

6 CPB TRANSPORT BY PIPELINES

6.1 Type of underground distribution systems

There are three possible configurations for movingfill material from a point on the surface to the un-derground stopes as shown on Figure 19 (Thomas1979).

Pump

Pump ?

Open stope

Gravity / pump

system

Backfill plant

Mixer

Gravity

system

Pump / gravitysystem

Paste

backfill

Pump

Pump ?

Open stope

Gravity / pump

system

Backfill plant

Mixer

Gravity

system

Pump / gravitysystem

Paste

backfill

Figure 19. Basic configurations for paste backfill distributionsystems (adapted from Thomas et al. 1979).

As discussed by Thomas (1979), the "gravity/pump"system (Fig. 19) has the advantage of being totallycontained underground, thus causing no disruptionto surface activities. Furthermore, the ratio of thevertical to horizontal distance is usually so favour-able that little or no pumping energy is required.

The "gravity" system (Fig. 19) has the advantageof by converting vertical head to horizontal pressure

progressively which allows shorter and lighter pipesto be used. The pressure at the take-off points aremoderate and line failures, if any, do not disrupt the

646


11/14

main shaft or main level of operation. The circuitcan be developed progressively as the mine expands.

The "pump/gravity" system (Fig. 19) has the ad-vantage of easy installation, inspection and mainte-nance, with no special underground level require-ments and no disruption of the main shaft. However,such a system makes the filling operation dependentupon a pumping operation and requires a long bore-hole to place fill underground which results in a high

pressure take-off point.

6.2 CPB transport underground

The paste backfill is delivered by pipeline to the dis-posal point in the stope and the friction factors gen-erated require that high pressure pipelines be used totransport the pastefill. Pressures typically exceed5MPa for this type of laminar flow system. Earlysystems used high pressure reciprocating pumps butexperience has shown that pastefill can be readilytransported by gravity alone, provided that the re-ticulation geometry is favourable (Grice 1998).

6.2.1 Flow-loop tests of the CPBFor a given mine, a fully instrumented pipes for

paste backfill flow-loop tests must be performed todetermine the paste transport characteristics. Usuallythis is an instrumented, closed-circuit pipeline sys-tem powered by a diesel engine positive-displacement pump. The instrumentation on the

paste flow-loop tests provides essential engineeringdata such as flow rate (Q), friction head loss per unitlength of pipe (f= P/L), shutdown and restart ca-

pabilities, and power consumption needed to design

full-scale pipelines. Figure 20 is an example of pasteflow-loop tests performed at the USBM's SpokaneResearch Center (Clark et al. 1995).

Figure 20. Pastefill flow-loop tests configuration and pressuremonitoring locations (after Clark et al. 1995).

The calculation of the friction head loss (P/L) willallows determination of the running pressures of the

paste distribution system: type of volumetric dis-placement pump, choice of pipe diameters (D), flowrate (Q), and paste flow velocity (V). For a Bingham

plastic fluid flowing in laminar regime (pastefill),the friction head loss (f) is given by the following re-lationship:

+

=

=

4

00

2

3

1

3

41

32

ww

B

D

V

L

Pf (34)

where f = friction head loss (Pa/m); B = Bingham plastic viscosity (Pa.s); 0 = yield stress (Pa); w =wall shear stress in Pa (wDP/4L);D = pipe di-ameter (m); P = differential pressure in the pipe(Pa).

The use of rheological models such as Equation 33requires the a priori knowledge of the pasteBingham plastic viscosity () which is very difficultto predict because it depends on several factors. Thatis why it is important to relate the slump value to the

plastic viscosity as the relationships (Eqs. 30 & 31)proposed by Ferraris & de Larrard (1998). The pipe

diameters often used vary between 100 mm (4 in)and 200 mm (8 in). For example, a paste backfillwith a slump value of 180 mm (7 in) can betransported by gravity at a flow rate of 100 ton/hourin boreholes/pipes system with a 150 mm (6 in)diameter.

6.2.2 Horizontal transport distanceThe horizontal transport distance (Lh) generated by astanding column of material is obtained by dividingthe pressure at the bottom of the standing column(Pbottom) by the frictional pressure gradient or pres-sure loss (Clark et al 1995). The pressure at the bot-tom of a standing column is obtained by taking thedifference between the pressure imparted by gravityand pressure lost through frictional pressure gradi-ent, so that horizontal transport distance (Lh) is given

par the following relationship (Fig. 21):

P

LHL

PH

f

PmL bottomh

==)( (35)

where = fill bulk unit weight (kN/m3); H= maxi-mum free-fall height of the paste in the paste (m);

P/L = friction head loss (Pa/m).

647


12/14

Pbottom

H

Lh (horizontal distance)

Paste flowing in

the pipe

Pbottom

H

Lh (horizontal distance)

Paste flowing in

the pipe

Figure 21. Schematic illustrating the calculation of the horizon-tal distance of paste flow.

7 BACKFILL DELIVERY IN THE STOPES

Once all the transport parameters are correct, the paste backfill can be delivered to undergroundopenings through pipelines. Figure 22 shows ageneral outline of a backfilled stope with its variouscomponents (fill mass, barricade, rock mass,adjacent filled stope) as well as the stress fielddistribution.

Rock pilebarricade

Filled

stope

h_filled stopeh_fill

h_fill

v_fill

h_pillar

Paste discharge

Rock

mass

barricade

Rock pilebarricade

Filled

stope

h_filled stopeh_fill

h_fill

v_fill

h_pillar

Paste discharge

Rock

mass

barricade

Figure 22. Schematic showing the components of a backfilledstope and the stress field distribution.

After the stope is backfilled with CPB its mechani-cal integrity can be threatened by several macro-scopic factors (in opposition to the hydration proc-ess) which are going to influence the mechanical

strength of the CPB and the structural stability of thefilled stope. These factors which result from interac-tions between CPB and rock walls are, fill settle-ment and the drainage of its excess water, fill con-solidation, stope volume, stress field distributionwithin the backfill mass (pressures at the floor of thestope and on the barricade), wall convergenceagainst the fill mass, shrinkage and the arching ef-fect.

Drainage and settlement will favour the develop-ment of a high mechanical strength of the CPB(Belem et al. 2001, 2002). On the other hand, the fill

mass will be stable due to the development of arch-ing effects depending upon the stope dimensions.

The pressures at the floor of the stope and on thebarricade will have a harmful effect on the stabilityof the filled stope when these pressures are too high(see more details in Belem et al. in the companion

paper). Consequently, it is necessary to understandthese various factors which influence stope stability

to ensure better ground control.The knowledge of the magnitude of the pressureson the barricade will allow better planning of themining sequences. The knowledge of the stress fieldwithin the fill mass will facilitate its stability analy-sis when it is considered that one of its faces may beexposed or when one wants to cut an access galleryto a new orebody through the CPB.

8 CONCLUSION

This paper is a general overview on the use of ce-

mented paste backfill, from its design to its under-ground delivery. When a mining method uses paste

backfill, initially one must determine the limitingstrength and the pressures which will be developedin the fill according to the geometry of the openedstopes.

To meet these criteria, laboratory optimization ofpaste backfill mix design will be essential to deter-mine the ideal mixture to acheive the desired limit-ing strength. But before beginning the stope filling,it would be necessary to know the rheological prop-erties of the fill material. For that purpose, one will

select a rheological model of paste backfill behav-iour (Bingham or Pseudo-plastic) to determine thetwo essential parameters, yield stress and viscosity.

The pumpability of the paste backfill can be alsoestimated using the standard or modified slumptests. This last would allow relating the slump andthe "slumping time" to the yield stress and the plas-tic viscosity. According to existing distribution sys-tem at the mine concerned (e.g. gravity, pumping,etc.), paste flow-loop tests are necessary to estimatethe friction head loss of the pipelines for better con-trol of the operating pressures.

With this last parameter, it would be also possibleto calculate the maximum horizontal distance for the

paste flow without any additional pressure. Once thepaste backfill is transported underground through thepipelines to the open stopes, it will interact with thestopes and pillar walls and its initial physical andmechanical properties will evolve in the course of itscuring time.

648


13/14

ACKNOWLEDGMENTS

This research was supported by the IRSST and partsof NSERC and NATEQ. The authors gratefully ac-knowledge their support. The authors would alsolike to thank our mining partner, Cambior Inc. (MineDoyon) for their collaboration in the completion ofthis work.

REFERENCES

Arioglu E. 1983. Design of Supports in Mines, Wiley and Sons,New York, 248.

Arioglu, E. 1984. Design aspects of cemented aggregate fillmixes for tungsten stoping operations. J. of Mining Sci-ences and Technology, No. 1, pp. 209-214.

Askew J.E., McCarthy P.L., Fitzerald D.J. 1978. Backfill re-search for pillar extraction at ZC/NBHC. Proceedings of12

thCan. Rock Mech. Symp., pp. 100-110.

Aubertin M., Li L., Arnoldi S., Belem T., Bussire B., Benzaa-zoua M., Simon R. 2003. Interaction between backfill androck mass in narrow stopes. Soil and Rock America 2003:12th Panamerican Conference on Soil Mechanics and Geo-

technical Engineering and 39th U.S. Rock Mechanics Sym-posium, 22-26 June 2003, Boston, Massachusetts, USA,Vol. 1, Verlag Gckauf GmbH (VGE), Essen, pp. 1157-1164.

Barret J.R., Coulthard M.A., Dight P.M. 1978. Determinationof Fill Stability. Mining with Backfill 12th CanadianRock Mechanics Symposium, Canadian Institute of Miningand Metallurgy, Quebec, Special Vol. 19, 85-91.

Belem T., Benzaazoua, M., Bussire B., Dagenais A.-M. 2002.Effects of settlement and drainage on strength developmentwithin mine paste backfill. Proceedings of Tailings andMine Waste'02, 27-30 January 2002, Fort Collins, Colo-rado, Balkema : Rotterdam, pp. 139-148.

Belem T., BussireB., BenzaazouaM. 2001. The effect of mi-crostructural evolution on the physical properties of pastebackfill.Proceedings of Tailings and Mine Waste'01, Janu-ary 16-19, Fort Collins, Colorado, A.A. Balkema, Rotter-dam, pp. 365374.

Belem T., Harvey A., Simon R., Aubertin M. 2004. Measure-ment of internal pressures of a gold mine pastefill duringand after the stope backfilling. Proc. 5th InternationalSymp. On Ground support in mining and underground con-struction, 28-30 September 2004, Perth, Western Australia(to be published).

Benzaazoua M., Belem T., Bussire B. 2002. Chemical aspectof sulfurous paste backfill mixtures. Cement and ConcreteResearch, Vol. 32 (7), pp. 11331144.

Benzaazoua M., Fall M., Belem T. 2004. A contribution to un-derstanding the hardening process of cemented pastefill.

Minerals Engineering. 17-2 : 141-152.BenzaazouaM., Ouellet,J., Servant,S., Newman,P., Verburg,

R. 1999. Cementitious backfill with high sulfur content:physical, chemical and mineralogical characterization. Ce-ment and Concrete Research, 29: 719 725.

Brackebusch F.W. 1994. Basics of Paste Backfill Systems.Mining Engineering, Vol.46, October, pp. 1175-1178.

Cai S. 1983. A Simple and Convenient Method for Design ofStrength of Cemented Hydraulic Fill. Proceedings of Inter-national Symposium on Mining with Backfill, A.ABalkema, Rotterdam, 405-412.

Chen L., Jiao D. 1991. A design procedure for cemented fillfor open stoping operations. J. of Mining Science andTechnology, No. 12, pp. 333-343.

Clark C.C., Vickery J.D., Backer R.R. 1993. Transport of totaltailings paste backfill: results of full-scale pipe test loop.Pumping tests. Report of investigation, RI 9573, USBM,37p.

Coates D.F. 1981. Caving, Subsidence, and Ground Control. Rock Mechanics Principles, Minister of Supply and Ser-vices Canada, Chapter 5.

Craig R.F. 1995. Soil mechanics. Fifth Edition, Chapman andHill Publishing, London, 427 p.

Donovan J.G. 1999. The effects of backfilling on ground con-trol and recovery in thin-seam coal mining. M.Sc. Thesis,

Virginia Polytechnic Institute and State University, Blacks-burg, Virginia, 148 p.

Douglas E., Malhotra V.M. 1989. Ground granulated blast fur-nace slag for cemented mine backfill:Production andevaluation. CIM Bulletin, Vol.82, No.929, pp. 27-36.

Fall M., Benzaazoua M. 2003. Advances in predicting per-formance properties and cost of paste backfill. ProceedingsonTailings and Mine Waste'03, October 12-15, Vail, USA,Balkema:Rotterdam, 2003, p. 7385.

Ferraris C.F. 1999. Measurement of the Rheological Propertiesof High Performance Concrete: State of the Art Report. J.Res. Natl. Inst. Stand. Technol. Vol. 104, pp 461-478.

Ferraris C.F., de Larrard F. 1998. Modified slump test to me-sure rheological parameters of fresh concrete. Cement,Concrete and Aggregates, CCAGDP, Vol. 20, 2: 241247.

Grice T. 1998. Underground Mining with Backfill. The 2nd Annual Summit Mine Tailings Disposal Systems, 24-25November 1998, Brisbane, Australia, 14p.

Guang-Xu T., Mao-Yuan H. 1983. Assessment on SupportAbility of the Fill Mass.Proceedings of Int. Symp. on Min-ing with Backfill, A.A Balkema, Rotterdam, 433-442.

Hansen J.B. 1968. A revised extended formula for bearing ca-pacity.Danish Geotechnical Institute Bulletin, No. 28.

Hassani F, Bois D. 1992. Economic and technical feasibilityfor backfill design in Quebec underground mines. Final report 1/2, Canada-Quebec Mineral Development Agree-ment, Research & Development in Quebec Mines. Contractno. EADM 1989-1992, File no. 71226002.

Hassani F., Archibald J.H. 1998. Mine Backfill. CIM, CD-ROM.

Hunt R.E. 1986. Geotechnical Engineering Analysis andEvaluation. McGraw Hill, New York, 729.

Landriault D.A., Verburg R., Cincilla W. and Welch D. 1997.Paste technology for underground backfill and surface tail-ings disposal applications. Short course notes, Canadian Institute of Mining and Metallurgy, Technical workshop april 27, 1997, Vancouver, British Columbia, Canada,120p.

Landriault, D.A., Lidkea, W. 1993. Paste fill and high densityslurry fill. In Proceedings of the International Congress onMine Design, Queens University, Kingston, Ontario, Can-ada, August.

Lawrence C.D. 1992. The influence of binder type on sulfateresistance. Cement and Concrete Research, Vol. 22, pp.1047-1058.

Li M., Moerman A. 2002. Perspectives on the scientific andengineering principles underlying flow of mineral pastes.Proc. 34th Ann. Meet. of CMP, Ottawa, Canada. Paper No.35, pp. 573-595.

Marston A. 1930. The theory of external loads on closed con-duits in the light of latest experiments. Bulletin No. 96,Iowa Engineering Experiment Station, Ames, Iowa.

Mitchell R.J. 1983. Earth structures engineering. London, Al-len & Unwin. Chapter 6.

Mitchell R.J., Olsen R.S., Smith J.D. 1982. Model studies oncemented tailings used in mine backfill. Can. Geotech. J.,Vol.19, No.1, pp. 14-28.

649


14/14

Mitchell R.J., RoettgerJ.J. 1989. Analysis and modelling of sillpillars.Innovations in mining backfill technology. Balkema,Rotterdam, pp. 53-62.

Mitchell R.J. (1989a. Stability of cemented tailings backfill.Computer and physical modelling in geotechnical engi-neering. Balkema, Rotterdam, pp. 501-507.

Mitchell R.J. 1989b. Model studies on the stability of confinedfills. Canadian Geotechnical Journal, Vol. 26, pp. 210-216.

Nguyen Q.C., Boger D.V. 1985. Direct yield Stress Measure-ment with the Vane method. Journal of Rheology, 29 (3):

335 947.Revell M. 2000. Cannington backfill taking the pig out of

paste. Paste Technology Seminar, April 2000, Perth, Aus-tralia : 112.

Smith J.D., Dejongh C.L., Mitchell R.J. 1983. Large scalemodel tests to determine backfill strength requirements forpillar recovery at the Black Mountain Mine.Proceedings of Int. Symp. Mining with Backfill, Lulea, June 7-9, pp. 413-423.

Smith J. D., Dejongh C. L., and Mitchell R. J. 1983. LargeScale Model Tests to Determine Backfill Strength Re-quirements for Pillar Recovery at the Black MountainMine, Proceedings, International Symposium on Miningwith Backfill, Lulea, June 79, pp. 413423.

Stone D.M.R. 1993. The optimization of mix designs for ce-

mented rockfill. MINEFILL93. Proc. fifth Int. Symp. on Min-ing with Backfill Johannesbourg, SAIMM,pp. 249-253.

Terzaghi K. 1943. Theoretical soil mechanics. John Wiley &Sons, New York.

Thomas E.G. 1979. Fill Technology in Underground Metalli-ferous Mines. International Academic Services Ltd., King-ston, Ontario, 293p.

Wayment W.R. 1978. Backfilling with Tailings A New Ap-proach. Mining with Backfill 12th Canadian Rock Me-chanics Symposium, Canadian Institute of Mining and Met-allurgy, Quebec, Special Vol. 19, 111-116.

Yu T. R. 1992. Mechanisms of Fill Failure and Fill StrengthRequirements. Proceedings, 6th Canadian Rock MechanicsSymposium, June 1992, pp. 16.

Paste Back Filling Propertis

Documents