Slurry Transport in Blind Backfilling Technique 4 3 17 - 8 17 PM

Slurry Transport in Blind Backfilling Technique
Susmita Panda
Slurry Transport in Blind Backfilling Technique
Thesis submitted to IIT Kharagpur
for the Award of the Degree of
Doctor of Philosophy
by
Susmita Panda
Under the supervision of
Prof. S K Pal
Mining Engineering
&
Prof. P V S N Murthy
Mathematics
Indian Institute of Technology Kharagpur
2017 Susmita Panda. All rights reserved.
April, 2017
Dedicated to my parents
Abstract
Blind hydraulic backfilling is a commonly used technique to control subsidence of the

strata over unapproachable water-logged underground mines. In this investigation, stud-
ies have been carried out in a transparent scaled model of underground mine roadway us-
ing simple gravity blind backfilling method. The objectives of the thesis are to investigate
effects of flow conditions on different flow parameters through experimental and theoret-
ical studies. Since sudden jamming is a major hurdle in filling-up of a large area from one
borehole, an attempt has also been made to evaluate a pre-jamming indication parameter.
The study has revealed that the basic process of filling occurs by sand transportation
along one or more self-formed meandering channels. The relative influences of flow con-
dition on maximum volume of sand throughput, maximum length of bed advancement,
flow pressure loss per unit length etc. have been explored and cross-validated with previ-
ous field results. Theoretical relationships have been developed about channel shape,
channel area, flow pressure loss etc. for different flow conditions and obtained result has
been cross-validated with experimental work. The developed relationships among various
parameters have been found to be useful for practical implementation of gravity blind
backfilling process.
A pre-jamming indication parameter has also been evaluated using stored pressure-time
data. The pre-jamming parameter, so developed, may be used to indicate the arrival of
final stage of filling after which not more than 26% filling will be possible with the same
flowrate and concentration. In order to prolong the filling process either the concentration
needs to be reduced or the flowrate may be increased.
Keywords: Gravity blind backfilling method, flow pressure loss, channel cross-section,
bed advancement, pre-jamming indication parameter, maximum volume of sand through-
put.
ii
APPROVAL OF THE VIVA-VOCE BOARD
Date:
Certified that the thesis entitled Slurry transport in blind backfilling technique sub-
mitted by Susmita Panda to Indian Institute of Technology, Kharagpur, for the award of
the degree of Doctor of Philosophy has been accepted by the external examiners and that
the student has successfully defended the thesis in the viva-voce examination held today.
_____________________ _____________________ _____________________
Prof. K U M Rao (MI) Prof. A Bhattacharjee (MI) Prof. G P Rajasekhar (MA)
(Member of the DSC) (Member of the DSC) (Member of the DSC)
_____________________ _____________________ _____________________
Prof. S K Pal (MI) Prof. P V S N Murthy (MA)
(Supervisor) (Supervisor) (External Examiner)
_____________________
Prof. K. Pathak
(Chairman)
iv
CERTIFICATE
This is to certify that the thesis entitled Slurry transport in blind backfilling tech-
nique submitted by Ms. Susmita Panda to the Indian Institute of Technology, Kharag-
pur, is a record of bona fide research work carried out under my supervision and is wor-
thy of consideration for the award of the degree of Doctor of Philosophy of the Institute.
_________________________ _________________________
Supervisor Supervisor
Prof S K Pal, Professor Prof P V S N Murthy, Professor

Department of Mining Engineering Department of Mathematics
Date: Date:
vi
DECLARATION
I declare that
the work contained in this thesis is original and has been done by me under the
guidance of my supervisor, Professor S K Pal and Professor P V S N Murthy;
the work has not been submitted to any other Institute for any degree or diploma;
I have followed the guidelines provided by the Institute in preparing the thesis;
I have conformed to the norms and guidelines given in the Ethical Code of Con-
duct of the Institute;
whenever I have used materials (data, theoretical analysis, figures and text) from
other sources, I have given due credit to them by citing them in the text of the the-
sis and giving their details in the references. Further, I have taken permission
from the copyright owners of the sources, whenever necessary.
_____________________ ________________________
Date Susmita Panda
viii
ACKNOWLEDGEMENTS
Ph.D days are truly memorable days in my life and it would not have been possible to
complete Ph.D work without the support and guidance that I received from many people.
To start with, I would like to express my sincere gratitude towards my Ph.D supervisors
Professor S K Pal and Professor P V S N Murthy who introduced me to the area. I am
obligated towards their valuable guidance, support and care throughout my research peri-
od. I am grateful to them, for all the thinking, research experience, reading and writing
skills that I learnt from these years. It is my great privilege to be a member of their enthu-
siastic research group.
I am grateful to Professor K. Pathak (present Head of the Department) his constant inspi-
ration. It is my pleasure to thank to my DSC members, Prof. K U M Rao, Prof. A
Bhattacharjee, Prof. G P Rajasekhar for their constant encouragement and support. I must
thank all other faculty members for their valuable suggestions and nonteaching staffs of
the department for their cooperation.
Words would be insufficient to express my feeling for my lab mates and friends. The love
and support I got from them throughout my Ph.D work. I must thank Ms. Tuli Bakshi,
Mr. Tapas Dey, Mr. Biswajit Manna, Mr Sourav Jain, Dr. Devi Ranjan Tripathy, Dr.
Sourav Kumar Dandapat, Dr. Suman Babu Patteti, Dr. Sneha Rani, Dr. Radhakanta
Konar for their constant help and suggestion. A big part of my thank goes to my friends
and colleagues in IIT Kharagpur for their cooperation during my research period. I am
also grateful to all the staff members and research scholars of our department for their
cooperation. The acknowledgement is extended to the technical staffs Mr. Narayan Dolai,
Mr. Chandan Raksit, Mr. Samir Das, Mr. Suren Doloi, Ms. Anita Mondal, Ms. Chaitali
Mazumder, Mr. Sujit Das and Mr. Anshuman Dolai for their help during my experi-
mental work.
x
I must express my deep sense of gratitude to my mother, Mrs Satadal Panda and all of my
four sisters and brothers-in-law. I must convey my regards to my other family members
for their love and care. I must express my gratitude to my father, Late Subhendu Kumar
Panda. I must thank all my teachers who helped and guided me during the journey.
IIT Kharagpur Susmita Panda

Contents
Abstract ........................................................................................................................... ii
APPROVAL OF THE VIVA-VOCE BOARD ................................................................. iv
CERTIFICATE ............................................................................................................... vi
DECLARATION ........................................................................................................... viii
ACKNOWLEDGEMENTS .............................................................................................. x
Contents ........................................................................................................................ xii
List of Tables ............................................................................................................... xvi
List of Figures ............................................................................................................ xviii
List of symbols, abbreviations and formula ............................................................... xxiv
1 Introduction .................................................................................................................. 2
1.1 Preamble ............................................................................................................... 2
1.2 Background ........................................................................................................... 2
1.3 Motivation ............................................................................................................. 3
1.4 Objectives ............................................................................................................. 4
1.5 Scope of this thesis................................................................................................ 5
1.6 Novelties of this thesis .......................................................................................... 5
1.7 Organization of this thesis .................................................................................... 5
1.8 Methodology ......................................................................................................... 6
xii
Contents
2 Literature Survey ......................................................................................................... 8
2.1 Literature review on backfilling............................................................................ 8
2.2 Literature review on sediment transport related work ........................................ 16
3 Experimental Investigation ........................................................................................ 32
3.1 Introduction ......................................................................................................... 32
3.2 Experimental setup.............................................................................................. 33
3.3 Experimental procedure and measurement ......................................................... 35
3.4 Empirical relationships ....................................................................................... 40
3.5 Discussions ......................................................................................................... 54
4 Progress of deposited sand bed .................................................................................. 56
4.1 Introduction ......................................................................................................... 56
4.2 Bed slope angle ................................................................................................... 56
4.3 Bed advancement ................................................................................................ 63
4.4 Discussions ......................................................................................................... 80
5 Channel Cross-Section and flow Pressure Loss......................................................... 82
5.1 Model formulation .............................................................................................. 82
5.2 Computational procedure for supporting variables............................................. 87
5.3 Solution method for the model ........................................................................... 90
5.4 Channel shape ..................................................................................................... 94
5.5 Concentration profile in the moving bed layer ................................................... 98

5.6 Velocity in the upper layer ................................................................................ 100
5.7 Pressure drop ..................................................................................................... 101
5.8 Discussions ....................................................................................................... 105
6 Pre-jamming indication parameter ........................................................................... 108
6.1 Introduction ....................................................................................................... 108
6.3 Results ............................................................................................................... 110
6.5 Discussions ....................................................................................................... 118
7 Conclusions and Future scope of the work .............................................................. 120
7.1 Conclusions ....................................................................................................... 120
7.2 Future Scope of work ........................................................................................ 124
Curriculum Vitae ........................................................................................................ 126
References ................................................................................................................... 128
xiv
List of Tables
List of Tables
Table 3.3.2.1 Repeatability experiments for maximum sand throughput
Table 3.4.2.2 Maximum flow pressure loss in channels at jamming
Table 3.4.3.2 Bed advancement rates for different flow conditions
Table 3.4.4.1 Chanel height and width ratio for 15 lpm flowrate
Table 4.2.1 Variation of maximum, minimum and average bed slope angle
Table 4.3.1.1 Percentage of error in the average measured value of
Table 4.3.2.1 Values of coefficient values using SPSS
Table 4.3.3.1 Comparison of percentage errors values computed using linear and ex-
ponential profiles
Table 5.7.1 Comparison of experimental and theoretically modeled pressure loss for dif-
ferent flow conditions
Table 6.1.1: Area of filling for different flow conditions.
xvi
List of Figures
List of Figures
Figure 1.8.1 Schematic work flow diagram of the present study
Figure 2.1.1 Classification of backfilling methods
Figure 2.1.2 Pneumatic pipe feeder systems
Figure 2.1.3 Schematic diagram of pneumatic backfilling
Figure 2.1.5 Typical sections showing boreholes and configuration of fill material placed
in mine void as point support technique
Figure 2.1.6 Schematic depiction of pumped slurry backfill injection
Figure 2.2.1 Schematic depiction of gravity blind backfill technique
Figure 2.2.2 Sediment deposition in river plume when channel water enters the lake
Figure 2.2.3 Particle trajectory in open channel
Figure 2.2.4 Particle trajectory within a straight pipe
Figure 2.2.5 Particle trajectory in pipe bend
Figure 2.2.6 Area and Surface in three layer model
Figure 2.2.7 Transverse profile of channel cross section
Figure 2.2.8 Longitudinal Profile of Seafloor Stratigraphy
Figure 3.2.1 Schematic diagram of the experimental set up for gravity blind backfilling.
Figure 3.3.1.1 Channel formation of filling process - arrow shows the path of channel.
Figure 3.3.2.1 Variation in volume of sand throughput for different flow conditions
xviii
List of Figures
Figure 3.4.1.1 Maximum volume of sand throughput, versus
Figure 3.4.1.2 Volume of sand throughput versus
Figure 3.4.1.3 Volume of sand throughput for different sand flowrate and water flowrate
Figure 3.4.2.1 Pressure loss versus length of the channel for 25 lpm flowrate and 6%
concentration
;
Figure 3.4.2.2 Plot of pressure loss versus
Figure 3.4.2.3 Flow pressure loss curve for different flow condition (applicable for field
condition)
Figure 3.4.3.1 Advancement of sediment bed during filling process (a) Top view (b) Side
view.
Figure 3.4.3.2 Bed advancement during filling process for different flow conditions
Figure 3.4.3.3 Rate of bed advancement versus
Figure 3.4.4.1 Channel cross-section
Figure 3.4.4.2 Width and height ratio relationship
Figure 3.4.5.1 Pressure-time curve recorded by the data acquisition system during the
filling process.
Figure 4.2.1 Sequence of the filling process
Figure 4.2.2 Cyclic (a) deposition and (b) collapse during the experiment
Figure 4.2.3 Variation of bed slope angle with time for different flow conditions
Figure 4.2.4 Maximum and minimum bed slope angle for different flow condition
Figure 4.2.5 Maximum bed slope angle versus 1/Qs

List of Figures
Figure 4.2.6 Minimum bed slope angle versus 1/Q
Figure 4.2.7 Average bed slope angle versus
Figure 4.3.1 Direction of bed advancement
Figure 4.3.2 Longitudinal profile of bed advancement
Figure 4.3.1.1 Side view of deposited sand bed.
Figure 4.3.2.1 A plot of with
Figure 4.3.2.2 Comparison of experimentally measured and semi-empirically computed

values of bed advancement for different flowrates under 6% concentration

values multiplied with swelling factor of bed advancement for different flowrates under
10% concentration

values of bed advancement for different flowrates under 14% concentration
Figure 4.3.2.5 Top view of the right side of deposited sand bed showing the difference
Figure 4.3.2.1 Exponential longitudinal profile of slanted part of bed edge
Figure 4.3.2.2 Plot of ( ) with
Figure 4.3.3.1 Error comparison of different longitudinal profile for bed advancement
with measured value
Figure 4.3.4.1 Longitudinal profile for 10% concentration with different flowrate
Figure 4.3.4.2 Longitudinal profile for 9% concentration with 6200 lpm flowrate
Figure 5.1.1 Top view of the filling process and channel formation
xx
List of Figures
Figure 5.1.2 Side view during filling process
Figure 5.1.3 Geometry of flow channels in blind Backfilling
Figure 5.1.4 Cross-sectional view of two layer model
Figure 5.3.1 Flow chart for the evaluation of pressure drop for a particular flow condition
Figure 5.3.2 Evaluation procedure for settling velocity of hindered particle
Figure 5.3.3 Evaluation procedure for friction coefficients
Figure 5.3.4 Flowchart for evaluating stationary bed height
Figure 5.3.5 Evaluation procedure for shear stress
Figure 5.4.1 Variation of channel height with slurry flowrate for different concentration
Figure 5.4.2 Variation of channel width with slurry flowrate for different concentrations
Figure 5.4.3 Variation of channel area of moving bed layer with concentration for differ-
ent flowrate
Figure 5.4.4 Channel shape for different flowrate with 10% concentration
Figure 5.4.5 Channel shape for 25 lpm flowrate and different concentrations
Figure 5.4.6 Channel cross-section for different flowrate with 10% concentration for
field application
Figure 5.5.1 Concentration profile at different depth (mm) of the channel for different
flowrate with 10% concentration
Figure 5.5.2 Concentration profile at different depth of the channel for different concen-
tration with flowrate 25 lpm
Figure 5.6.1 Variation of velocity (m/sec) for moving bed layer with flowrate (lpm)
List of Figures
Figure 5.7.1 Pressure loss per unit length for different flow condition
Figure 5.7.2 Comparison of theoretically modelled flow pressure loss with 8 times exper-
imental pressure loss for (a) 15 lpm (b) 20 lpm (c) 25 lpm and (d) 30 lpm flowrates
Figure 6.3.1 Pressure loss curve with bed advancement for different flow conditions
Figure 6.3.2 Plot of slope S versus
filling process.
Figure 6.3.2.2 Plots of coefficient of variation of inlet pressure for a slurry flowrate of 20
lpm and varying concentration of 6%, 10% and 14%.
Figure 6.3.2.3 Plots of coefficient of variation of inlet pressure for different slurry con-
centrations and flowrate of 25 lpm.
xxii
List of symbols, abbreviations and formula
Symbol Meaning
Inlet area or total area (m2)
Cross sectional area of moving bed layer respectively (m2)
Cross sectional area of stationary bed layer respectively (m2)
Width of void space (m)
Swell factor due to swelling of sand (=1.055)
Length curve part of the flow-channel (m)
Length straight part of the flow-channel (m)
Slurry volumetric concentration percentage
Drag coefficient depend on particle Reynolds number
Inlet slurry concentration in fraction
Lift coefficient
Average concentration of moving bed layer volume fraction
stationary bed concentration in fraction
is inlet pipe diameter
Particle diameter(m)
Flow pressure loss per unit length experimental (m/m)
Diameter of a particle
Hydraulic diameter of moving bed layer
Pressure loss per unit length using two layer formation
=diffusion coefficient
Slurry friction coefficients
Friction co-efficient for pure liquid flow
Drag force produced by the fluid flowing past it
Friction coefficient of moving bed layer and top cover
Friction coefficient between stationary layer and moving bed
layer
Ratio of inertial to the viscous terms
Gravitational constant
h Available head
xxiv
H Height of void space
, , Lift force coefficient in , , directions respectively
Lift force produced by the fluid flowing past it
Particle mass
Viscosity of feeding material (kg/ms)
Coefficient of static friction
Viscosity of water (kg/ms)
2 . /3
Average bed slope angle

Minimum bed slope angle
Maximum bed slope angle
Settling velocity of a particle
Terminal settling velocity of hindered particle
Slurry flowrate (lpm)
Sand flowrate (lpm)
Water flowrate (lpm)
Particle Reynolds number
Reynolds number of moving bed layer
Reynolds number of hindered particle
Liquid density
Solid density
Density of moving bed layer
S Slope for flow pressure loss and bed advancement
= Specific gravity
Surface in between the moving bed layer and stationary bed

layer
Surface in stationary bed layer
Surface in moving bed layer
Time (s)
Maximum time of filling for a particular flow condition
; Maximum duration of filling under field condition
Shear stress in moving bed layer
Shear stress along the interface of stationary bed layer and mov-
ing bed layer
Height of moving bed layer
Height of stationary bed layer
Average velocity of moving bed layer
Velocity in the direction respectively
Shear velocity
Mean velocity of mixture of slurry flow within a pipe( m/sec)
Magnitude of relative velocity
Maximum volume of sand throughput (litre)
Weight of particle
Component particle acceleration, velocity and position in di-
rection.
Maximum bed advancement
; Maximum bed advancement for any experimental flow condi-

tion
Maximum bed advancement under field condition
Length of bottom part of the deposited sand bed
Length of top part of the deposited sand bed
Particle acceleration, velocity and position in direction
Height of moving bed layer
Height of stationary bed layer
xxvi
Formula
= lift force coefficient in direction
( ), ( ) ( )- ( ), ( ) ( )-
( ), ( ) ( )- ( ), ( ) ( )-
( ), ( ) ( )- ( ), ( ) ( )-
,*( ), ( ) ( )- ( ), ( ) ( )-+
*( ), ( ) ( )- ( ), ( ) ( )-+
*( ), ( ) ( )-
(
), ( ) ( )-+ -
Abbreviation
VMA Vanguard Motion Analyzer
LAP Least action principle
Terminology:
SEDFLUX: A basin-fill model, able to simulate the delivery of sediment and their accu-
mulation over time scales.
Plume: Generated by a river draining its suspended sediment load into a receiving basin.
Grouting: Grout is a fluid form of concrete.
Magnus effect: A force generated by a spinning object traveling through a viscous fluid.
Saltating: The leaping movement of sand or soil particles as they are transported in a flu-
id medium over an uneven surface.
Friction velocity: Square root of the ratio of shear stress in an arbitrary layer of fluid and
density of the fluid
Aeolian: Wind activity in the study of geology and weather and specifically to the wind's
ability to shape the surface of the Earth
Rational regime model: The relations amongst governing equations traditionally applied
to predict the geometry of mobile-bed canals.
Hypopycnal: (surface) plumes were subject to a convective instability driven by some

combination of heat diffusing out of the warm, fresh, sediment-laden plume and particle
settling within it.
xxviii
Chapter 1: Introduction
1 Introduction
1.1 Preamble
Subsidence from old abandoned underground coal mines has become an everyday con-
cern of many people living in coal producing regions across the world. From the earlier
days of coal mining in India, coal seams at shallow depths were developed by bord and
pillar method of mining and conventional pillar-extraction was not undertaken. Very high
risk is associated with many important structures like railways, highways, building etc.
located in such area. In some places however, the pillars had been partially extracted and
mines were closed after filling the voids up to a certain extent. The gradual deterioration
of pillars, mine floor and mine roof resulted in the crushing of coal pillars and collapse of
strata above. Subsidence then occurs at the surface resulting in depression, cracking of
ground surface, development of sinkholes, potholes etc. Many of these old abandoned
workings are either fully or partially waterlogged.
Backfilling of mine voids is used to avoid the effects of surface subsidence. The pumped-
slurry, blind backfilling process has been particularly successful for stabilizing ground
over abandoned and inaccessible underground coal mines. With the help of this tech-
nique, it is possible to fill-up a large area using less number of injection boreholes. An-
other most simple and economic technique is the gravity blind backfilling method, which
may easily be adopted for the filling of old, abandoned mine workings. However, the
field implementation of this method often faces problems like premature jamming, inade-
quate area of filling from one borehole, lack of technological knowledge about the extent
of filling along different directions etc.
1.2 Background
The ground subsidence caused in abandoned mine depends on the type of overlying stra-
ta, the depth of mine and the support left during the mining operation. When one subsid-
ence occurs due to depreciation of mine pillars, it causes the overlying strata to cave into
the mine voids and this process propagates upwards causing ground subsidence.
2
Introduction
Simple gravity blind backfilling is used when the mine workings are inaccessible to
workers. With this approach, slurry of backfill material is fed through a borehole into
mine until the borehole fails to accept any additional backfill material. Gutfeld et al.
(2000) described that at least two boreholes are required in this method of filling process.
To minimize the cost, the number of boreholes should be minimized. This method is
more cost effective than any other filling processes, since the mixture flows freely inside
the mine cavity covering a large area and no additional equipment are required. Pal S K
(2001) demonstrated that for lower concentration of slurry fill up larger area but it takes
more time to fill up which is often undesirable. Very little amount of theoretical work
have been done on blind backfilling in mines by others researchers worldwide. Therefore
the present work aims at development of mathematical models and empirical relation-
ships in the area of simple gravity blind backfilling, development of pre-jamming indica-
tion parameter to avoid frequent jamming.
1.3 Motivation
Till date not much of theoretical analysis on blind backfilling (Chugh et al. 2005; Thill et
al. 1982; Baeckblom et al. 2003; Thomas E G 1985; Gamble and gray 1988) process have
been done on national or international levels. During the filling process of gravity blind
backfilling technique, every sand particle makes a trajectory inside the flow channel be-
fore its final deposition. To describe the path of solid particles in feeding slurry, a critical
study on the trajectory of solid particles in various applications like wind flow, open
channel flow and pipe flow etc. is required to be done. During initial phase of backfilling
the solid particles in sand-water slurry, upon its entrance into a much larger space, tend to
settle down to the floor of the mine. The thickness of the deposited sand bed increases
leaving a small gap at the top for flow of the slurry at a relatively higher velocity and car-
ry the sand particles to the end of the bed for final deposition. This process usually con-
tinues till the end of the filling process. In this process, the solid particles exist in two
prominent layers: one, where the particles do not have any movement in the deposited
bed, which is termed as lower layer or stationary-bed layer and the other, where parti-
cles move with the fluid in a suspended state, termed as upper layer or moving heteroge-
neous layer. In this upper layer, solid particles move forward by creating channels of
variable cross-sections depending on slurry flow conditions. To estimate the effect of dif-
ferent parameters like cross-sectional area of channel, flow pressure drop across the
channel and solid concentration profile in the heterogeneous layer on different flow con-
ditions, a study on similar form of slurry transport within a pipe (Doron and Barnea 1983;
Turian and Yuan 1977; Thomas A D 1979; Kaushal and Tomita 2002) or in open chan-
nels (Parker Gary 1978; Chen et al. 2009) is also essential for the present study. Geome-
try of the self-formed alluvium channels (Schuaam S A 1961; Ikeda and Izumi 1991; Cao
and Knight 1998) is also to be examined. To evaluate change in longitudinal concentra-
tion profile at the end of a channel during filling process, a critical study needs to be car-
ried out using advection diffusion equations.
Towards the end, the filling process stops abruptly due to the onset of jamming, this oc-
curs without any prior indication. Due to the sudden and pre-mature jamming, the field
implementation of gravity blind backfilling technique often leads to unsatisfactory per-
formance. Evaluation of a pre-jamming indication parameter through pattern analysis
(Box et al. 2015) of inlet pressure variation can be helpful in reducing the possibility of
sudden jamming and can prolong the duration of filling process from a single borehole.
1.4 Objectives
From the aforementioned motivation, the following objectives on blind backfilling tech-
nique are drawn:
I. To examine the physical process involved in gravity blind backfilling method and
obtain different empirical relationships connecting the variables like maximum
volume of sand throughput, rate of bed advancement, pressure loss per unit
length, longitudinal profile of bed edge etc. with inlet slurry flowrate and sand
concentration.
II. To evaluate, theoretically the shape, cross-sectional area, concentration profile of
flow channels and pressure drop per unit length by using the concept of layer
formation during filling process.
III. To verify the theoretical results with experimentally obtained values of the differ-
ent parameters.
4
Introduction
IV. To evaluate pattern of inlet pressure variation during filling process and thereby
determination of the pre-jamming indication parameter.
Initially some laboratory experiments on a scaled mine gallery carried out to understand
the physical phenomenon of the filling process. Based on this physical process witnessed
in a transparent laboratory model during experimentation, theoretical concepts are to be
developed to meet different objectives of the thesis. Evaluation of pre-jamming indication
parameter can also be done using pattern of inlet pressure variation.
1.5 Scope of this thesis
Since enough theoretical work has not been done on blind backfilling technique, theoreti-
cal development on this study will bring new prospects for mining industry. As sudden
jamming is a major problem in filling process, evaluation of pre-jamming indication pa-
rameter will make the filling process cost effective in certain areas.
1.6 Novelties of this thesis
This thesis establishes a reliable theoretical model which matches closely with experi-
mentation results. This thesis also provides a cost effective filling mechanism by evaluat-
ing pre-jamming indicator.
1.7 Organization of this thesis
Apart from the introduction chapter, this thesis consists of one chapter on literature re-
view, four contributory chapters and one chapter on summary and future scope of work.
Chapter 2 is the global literature review, whereas Chapters 3 to 6 deal with the contrib-
uted work of this thesis. Chapter 7 summarizes the findings of the present study.
Chapter 2 provides a detailed literature survey exposing the reader to important but gen-
eral literature within the scope of the present study. It is important to mention here that
the literature that is closely related to a particular problem has been discussed in individ-
ual chapters. Literature reviews mainly cover the area of backfilling techniques and other
research area which has some similarities to the physical processes involved in blind
backfilling.
Detailed discussions on experimental set-up and experimental procedure on model stud-

ies of blind backfilling have been done in Chapter 3. From the obtained experimental
data, empirical relationships have been formed and comparisons have been made with
field results.
In the Chapter 4, semi-empirical equations have been framed to evaluate the process of
bed advancement assuming different bed edge profiles. Evaluated relationships on bed
advancement with time have been compared with experimental results.
The main aim of Chapter 5 is to evaluate different flow parameters from a theoretically
developed model in which an assumption has been made about formation of two layers,
namely stationary bed layer and moving bed layer in the filling process.
Since sudden jamming is a major problem during filling process, evaluation of pre-
jamming indication parameter is discussed in details in Chapter 6.
Finally, Chapter 7 summarizes the major findings in the preceding and suggestions for
scope of the present study are discussed for future research.
1.8 Methodology
The workflow of the present investigation has been presented in the following Figure
1.8.1.
6
Introduction
Literature review Problem formulation
Objectives
Construct experimental set-up Formation of theoretical

equations and solve
Pre-jamming indicator Experiments for dif- Validate with experi-

from pressure data ferent flow conditions mental results
Empirical relationships
for different parameters
Validate with field

results
Conclusions
Figure 1.8.1 Schematic work flow diagram of the present study

Chapter 2: Literature Survey
2 Literature Survey
In earlier days, most coal mines in India have been worked under shallow cover and were
partially extracted using bord and pillar system of working. Saro et al. (2010), Galve et
al. (2009a, 2009b) evaluated mine subsidence hazard using statistics approach. Wolkers-
dorfer C (2008) discussed several problems associated with mine water aquaculture. Ay-
dan and Ito. (2015) presented the effect of the depth and ground water on the formation
of ground subsidence associated with abandoned room and pillar lignite mines under stat-
ic and dynamic conditions and discussed the effects on the areas above abandoned lignite
mines.
To avoid ground subsidence, backfilling technique is used to fill-up the void space of
abandoned mine (Li and Yang. 2015; Li. L 2013; Andreyev and Koons 2010; Wissmann
and Gale 2013; Karfakis M G 2016). Mhlongo and Amponsah-Dacosta (2015) reviewed
problems and solutions associated with abandoned mines in South Africa.
To achieve the objectives of the present study, there are many angles from which litera-
ture review is approached. Searches of relevant database, internet sites and publications
search are carried out on two different fields: (i) mine backfilling and (ii) other areas
which are very similar to the filling process of blind backfilling.
2.1 Literature review on backfilling
Placement of backfill material in underground mines may be accomplished using, manu-

al, mechanical, pneumatic and hydraulic methods. Out of the above methods manual and
mechanical methods of backfilling are seldom practiced in modern mines. Placement of
backfill material by mechanical means using sling packing machines or high speed belt is
sometimes applied in mines for packing rejected rocks from within the mine. Some re-
searchers have worked on the pneumatic and hydraulic placement of backfill material in
underground mines and these have been discussed in the following sections. Figure 2.1.1
shows a detailed classification of different backfilling techniques.
8
Literature Survey
2.1.1 Pneumatic Backfilling:
Backfill material is transported into an abandoned mine through a borehole in a stream of

repeatedly flowing air, either in a vacuum or in pressure as shown in the Figure 2.1.2 and
Figure 2.1.3. When a dense phase approach is used, the pipeline is nearly filled with
material that is moved as a fluid with low velocity air pressure in the form of slug flow.
Walker J S(1993), Sand et al. (1990) showed that when common dilute phase approach
is used, an air/backfill mixture typically consisting of less than 5 % backfill material has
been moved through a pipeline at relatively high velocity as compared to the fluid. These
approaches used for dry mines or water scarce mines.
2.1.2 Hydraulic Backfilling:
Filling of mine voids with backfill material by pumping the material as slurry (hydraulic
sandfill or sand backfill) through a borehole or large diameter pipeline into the mine is
used for Hydraulic backfilling.
Placement methods
Pneumatic Hydraulic Manual Mechanical

backfilling backfilling
Controlled Blind
flushing backfilling
Pneumatic blind Hydraulic blind

backfilling backfilling
Air-assisted gravity flushing Pump slurry injection Simple gravity flushing
Figure 2.1.1 Classification of backfilling methods

Figure 2.1.2 Pneumatic pipe feeder systems (Source Burnett, 1995)
Figure 2.1.3 Schematic diagram of pneumatic backfilling (Source Sand et al., 1990)
10
Literature Survey
2.1.2.1 Controlled Flushing:
Controlled flushing is used in mines where workers can safely work in the extraction
zones of active mines during the filling operations. When this approach is used, partitions
may be built in mine passages around the perimeter of the area to be filled. One or more
boreholes are drilled and cased from the surface to the upper portion of the mine work-
ings to be filled (Sutter Gold Mining 1998) as shown in Figure 2.1.4.
Figure 2.1.4 Schematic of controlled flushing using hydraulic backfilling (Source Sutter
Gold Mining 1998)
2.1.2.2 Blind Flushing
Where the operation consists of drilling boreholes to underground openings of an un-

approachable mine, which are expected to exist according to available mine maps. Due to
flooding or extensive caving, mines became inaccessible and the openings were flushed
blindly in the past. The gravity-feed method simply builds a conical pile beneath the un-
derground opening of the flushing hole. When the apex of the cone builds up to the mine
roof, no more fill material enters the mine opening. Depending on the carrier fluid the
blind backfilling method can be divided into two types:
a) Pneumatic Blind Backfilling:
In pneumatic blind backfilling method, backfill material is transported into the abandoned
mine through a well or pipeline in a stream of continually flowing air under pressure with
a complete circular rotation of injecting direction (Sand et al. 1990). Chugh et al. (2005)
developed a model to identify the flow pattern during filling process of pneumatic blind
backfilling using fly ash.
b) Hydraulic Blind Backfilling:
Hydraulic flushing and grouting of abandoned mines using remote filling from single or
multiple boreholes, are the most commonly used methods for the placement of backfill
material. Walker J S (1993) described the various techniques for backfilling of aban-
doned mine voids by hydraulic flushing from single or multiple boreholes. The hydraulic
blind backfilling of abandoned mines is used with two variants of ground supporting
technique.
Point Support Technique: Point support technique is used in areas where analysis
of subsurface condition indicates that insitu support would not be sufficient for
long term stability of surface features. In this method the filling material like ce-
ment grout is fed by gravity and it forms a conical pile touching the roof of the
mine. For further injection of backfill material, other boreholes are required to be
drilled and used as shown in Figure 2.1.5. This method is very cost effective
when applied correctly. The different types of point support techniques are: grout
column, fly ash grout injection and fabric formed concrete. These methods are
typically used with the intent of supporting a small area such as a structure or a
specific surface feature.
12
Literature Survey
Figure 2.1.5 Typical sections showing boreholes and configuration of fill mate-
rial placed in mine void as point support technique (Source Collaizzi et al. 1981)
Area-wide Technique: Area-wide backfilling is commonly practiced by sending

solids like sand, fly ash, or mine refuse underground through a large diameter
borehole by hydraulic means. This technique is used to fill large portion of the
mine with fill material, from multiple injection boreholes.
The hydraulic blind backfilling is classified in three different types depending on the
method of injection into underground mine voids.
i) Air -assisted gravity blind backfilling:
Solid-water mixture is sent to fill underground voids through a larger diameter pipe and
compressed air is fed through a smaller diameter pipe placed inside the larger diameter
pipe (Saxena et al. 1984). In this method the buoyant force of air inside sand-water slurry
aids the driving force for the channel flow and therefore the total area of filling from a
single borehole can be increased.
ii) Pumped slurry injection:
Filling is done hydraulically by pumping of backfill-slurry down the borehole (Whaite et

al. 1975). This method is helpful for filling a large area from a single feeding hole.
iii) Simple gravity blind backfilling:
Sand water mixture is fed through a borehole under the action of gravity when mine
workings are not accessible to workers. The quantity injected down a single well depends
on the conditions of the underground, such as the height, slope, and the proximity of pil-
lars in the mine workings (Pal et al. 2001).
Studies on backfilling of abandoned mines have been carried out by various researchers
mainly through experiments on different transparent scaled models. In these model stud-
ies, the physical phenomenon of filling process is critically observed during filling by hy-
draulic means. The general observations in these model studies indicate that the filling
occurs through multiple meandering channels along the top of the deposited sand bed.
The sand particles in the flow-channels are kept in suspension due to the turbulence cre-
ated by water and is transported away along the direction of flow upto the end of deposit-
ed sand bed. It was also observed in these transparent models that four to eight small
channels existed at the beginning of the experiment, but ultimately after some time only
one, two or at most three channels continued. When sand transportation along a routed-
channel continued for some duration, the pressure within the channel increased sufficient-
ly to cause a puncture, thereby forming a new channel and the old channel gets plugged
with sand. The filling continues in this way till the hydraulic head available is mainly uti-
lized for overcoming the flow-path resistance, and too little additional head available for
puncturing a new path (Pal S K 2003). Figure 2.1.6 shows the sequential progress of
filled up sand bed during this blind backfilling process. Thill, et al. (1983) described the
process of backfilling the flooded mines using sand or crushed mine refuse.
14
Literature Survey
Figure 2.1.6 Schematic depiction of pumped slurry backfill injection (Source Whaite
et al. 1975)
The blind backfilling process in un-approachable underground workings cannot be wit-

nessed. Therefore, the exact phenomenon that occurs during the blind backfilling process
can only be visualized through model studies (Whaite et al. 1975) using fully transparent
Perspex sheets.
It is seen from the model studies that the mode of transport in hydraulic blind backfilling
is similar to slurry transport in alluvium channels of water resources engineering. So, the
available literature on open-channel sediment transport, slurry transport through pipeline,
sediment deposition, shape of alluvial channel etc. have been reviewed. A detailed
knowledge in this area will assist in developing the theoretical basis for finding shape of
transport channels and movement of backfill particles in these channels.
2.2 Literature review on sediment transport related work
Discharge Transport Entrance Transport Discharge

zone zone zone zone zone
Figure 2.2.1 Schematic depiction of gravity blind backfill technique
In general the gravity blind backfilling process can be divided in three zones, namely,
entrance zone, transport zone and discharge zone (Figure 2.2.1). In the entrance
zone the slurry from the inlet enters a large void space and the velocity drops down to a
very low level. As a result the solid particles tend to settle at the bottom void space. The
scouring effect of the high velocity inlet slurry creates a conical turbulent solid-liquid
mixture just below the inlet point. The slurry then exits this zone through a small gap
near the top of the deposited sand bed, termed as flow channel. In the transport zone
the slurry moves along this flow channel over the deposited sand bed at an uniform rate
as a heterogeneous mixture. Finally the slurry enters the discharge zone located at the
sloping portion of the deposited sand bed. As the slurry progress forward in this zone the
cross-sectional area of flow increases and the velocity drops, thereby depositing sand par-
ticles along the slope of the bed.
The entrainment, transport and deposition of sand in reality and in model depends partial-
ly on the properties of sand particles, fall velocity, sand concentration, and dynamic vis-
cosity of sand-water mixture. During model studies, it has been observed that the sand
particles start moving along the bed in the form of rolling, sliding and making small
jumps. So the literature review of sediment transport in open channel, slurry transport
within the pipe, channel geometry of alluvial channel, sediment deposition from river
plume (Figure 2.2.2) etc. are helpful for studying the filling process in blind backfilling
method.
16
Literature Survey
Water Flow
River water
River Bottom
Lake Water
Lake Bottom
Figure 2.2.2 Sediment deposition in river plume when channel water enters the lake
The available literature on open-channel sediment transport, slurry transport through

pipeline, sediment deposition, shape of alluvial channel, particle trajectory in fluid medi-
um etc. have been reviewed in this section.
2.2.1 Particle trajectory in fluid flow
During filling process of blind backfilling technique, it is generally observed that sedi-
ment movement and deposition is effected through two layers, the first one is where the
sediment particles drop down and settle to form a stationary bed, and the other is a pas-
sage of moving slurry called a moving bed layer which lies above the layer of stationary
bed. The solid particles in the flowing slurry of heterogeneous layer make several small
jumps and ultimately reach at the end of the transport zone as shown in Figure 2.2.1. The
following section deals with the research work of several researchers in the field of tra-
jectory of the solid particles in moving fluid.
2.2.1.1 Particle trajectory in open channel flow
Motion of a single particle in an open channel is similar to blind backfilling in flow-

channels excepting that a top cover is present in the upper portion of fluid medium in the
present study. Comparison of filmed trajectory and theoretical trajectories of glass beads
(of diameter 350-710 m and density 2.5g/cm3) has been made in an environmental tun-
nel by White and Schulz (1977). The motion of particles was analyzed using VMA (Van-
guard Motion Analyzer) and the obtained trajectories are compared with analytical result.
When only drag force is considered, it was found that filmed particle consistently reach
higher heights than the predicted value but it has a good agreement when magnus effect
is included. For finding average analytical trajectories of particles in saltating flows in-
formation of the relationships between the impact and lift-off angles, velocities and rota-
tional speeds of the particle is compulsory. Relationships among the above parameters
are established based on the experimental data.
White et al. (1975) has done both numerical and experimental studies on the trajectory of
a solid particle in atmosphere. An experimental investigation has been performed for the
flow field around a model crater in the atmosphere. Numerical investigation has been
made for both two- and three-dimensional equations of a single particle under the influ-
ence of turbulent boundary layer as shown in Equation (2.2.1). Figure 2.2.3 shows the
approximate motion of a particle in a fluid medium.
Three dimensional equations of motions, which is used to define the particle trajectory,
are as suggested by White et al. (1975) are:
( )
[ ]
( ) ( )
[ ]
( )
[ ] }
The above system of equation is solved numerically to find out the motion of the particle
around the crater.
Philips M (1980) adopted force balance to evaluate particle trajectory within a fluid
stream. The obtained equation then compared with that given in Flecher B (1976), Bag-
nold R A (1951) and found to have a good agreement with experimental data. Wiberg and
Smith (1985) solved the differential equation of motion for a sediment grain near a non-
cohesive bed numerically for particle trajectory as a function of time. Yamamoto F
18
Literature Survey
(1986) framed equations and solved numerically by Runge Kutta Gill method for finding
the trajectory of a spherical particle within a pipe. It has been shown that numerical solu-
tion of equations is in good agreement with experimental result when drag coefficient
lift coefficient vary in between and .
Figure 2.2.3 Particle trajectory in open channel
Nalpanis et al. (1993) has developed a new statistical approach to evaluate velocities of
ejection and impact of sand particles (median diameters 118 and 188 m) in a turbulent
air flow above a sand bed. Multiple-image photographs above a sand bed in a wind tunnel
is used to measure ejection, impact velocity, trajectory length and maximum height of
rise of sand grains. The mean angle of ejection is found to be nearly from the hori-
zontal with mean vertical ejection velocity of about 2 , where , is the friction velocity.
Computed theoretical trajectories of saltating grains, using measurements of the initial
ejection velocities and the mean velocity profile of the air flow, largely agree with his
experimental measurements, and those of other researchers like Willets and Rice (1985)
and Araoka and Maeno (1981), but disagrees with the results of White and Schulz (1977).
2.2.1.2 Particle Trajectory in Pipe flow
During filling process of blind backfilling technique it has been seen that flow path being
irregular and meandering in nature, the slurry moves along straight as well as bent paths
in the heterogeneous layer below a non-moving top cover. This type of flow can be con-
sidered to be similar to the flow of slurry in a straight and bent pipe. In the following sec-
tion available literature related to the above field has been reviewed.
Salman et al. (2005) has described numerical simulation of a particle motion in a horizon-
tal pipe using variation of aerodynamic forces. The results from the simulation agree
closely with experimental time-of-flight measurements. Approximate particle trajectory
within a straight channel has been shown in the Figure 2.2.4.
Figure 2.2.4 Particle trajectory within a straight pipe
Govan et al. (1989) has described numerical simulation, which treats the particle motion
as a series of interactions with discrete eddies. This model was used to study the behavior
of the particles used in the experiments. The observed trajectories are mainly unidirec-
tional and it is concluded that the particle motion is strongly dependent on inlet condi-
tions. The simulation model has also been used to investigate the importance of profile
effects in the gas flow field in determining particle dispersion and deposition rates. It was
found that the relatively large particles used in this study had much flatter mean and ve-
locity profiles than the gas-phase profiles and deposition rates can be calculated using a
simple description which ignores profile effects. On the other hand, it was found that lift
forces due to the gas velocity profile produce a very large reduction in deposition rates.
Salman et al. (2005) has described numerical simulation of a particle motion in a horizon-
tal pipe using variation of aerodynamic forces. The results from the simulation agree
closely with experimental time-of-flight measurements. Approximate particle trajectory
within a pipe bent has been shown in the Figure 2.2.5.
20
Literature Survey
Figure 2.2.5 Particle trajectory in pipe bend
Wadke et al. (2005) has worked on a bend pipe instead of a straight pipe in a similar con-
cept as done by Salman et al. (2005) and found numerical simulations of particle motion
within a pipe-bend. It has been found that presence of bend can cause an increase in mean
particle velocity as compared to that in a horizontal pipe. Results show that the number of
impacts in the bend decreases as the velocity of the particle increases. The results from
the simulation agree closely with the experimental time-of-flight measurements.
Experimental and numerical investigation has been made by Peng et al. (1996) to develop
the motion of a single particle in flowing gas medium along a horizontal smooth pipe.
The observations made are: i) a single particle finally moves to the bottom of pipe, ii)
when gas velocity is higher, the height of the spherical particle is also higher, and iii) par-
ticle rotation is related to the friction co-efficient of the particle.
Morsi and Alexander (1971) developed a theoretical approach for (i) the response of a
spherical particle to a one-dimensional fluid flow, (ii) the motion of a spherical particle in
a uniform two-dimensional fluid flow about a circular cylinder and (iii) the motion of a
particle about a lifting aerofoil section. In all three cases the drag force on the particle is
allowed to vary with instantaneous Reynolds number by using an analytical approxima-
tion to the standard experimental Drag-Reynolds-number relationship for spherical parti-
cles.
William G (1963) made an experimental set-up of wind tunnel for investigating the
changes in size frequency distribution of sedimentary particle trapped at successive
heights above an eroding surface. For re-analyses of Williams experimental data, Jensan
and Sorenman (1986) developed a mathematical model for analyzing observed Aeolian
sand transport rate profiles.
The lift coefficient is based on a re-analysis of data produced by Chepil W S (1958), and
the other parameters of the problem are set using standard fluid mechanical relationships,
the initial velocity and position are specified with a separate model.
2.2.2 Transport of solid particles in open channels and pipelines
It is observed that there is some similarity of the blind backfilling process with the solid
transport through open channels and pipelines. Stationary bed layer and heterogeneous
layer are observed during backfilling process (Figure 2.2.1). The sediment particles are
subjected to deposit due to gravitational forces. Suspension of sediment and its forward
deposition is caused by turbulent forces. The existence and uniqueness of the solution of
the mathematical model (nonlinear, free boundary layer, first order PDE) for the sedi-
mentation of particle in cylindrical axisymmetric flow is established by Angelis and
Mancini (1997). Richardson and Jaki (1954) described a method of calculating sedimen-
tation velocity for a suspension, or fluidization velocity in a solid-liquid system, as a
function of the free falling velocity of a single particle and particles concentration. The
main objective of his study is to evaluate effect of suspended particle concentration on
settlement of particle. Doron and Barnea 1993,Doron et al. 1987), Peker and Helvaci
(2011) considered continuity and momentum equations to obtain flow parameters such
as, height and velocity of every layer, pressure gradient etc.
2.2.2.1 Pressure drop during slurry transport
When liquid starts to flow, energy dissipation occurs, which in turn, appears in the form
of a pressure drop. When is pressure drop across a channel length , the loss of en-
ergy per unit distance (P/L), often called as head loss, is proportional to , is the
flow velocity and ( ) is an index. It is seen that theoretical analysis based on layer-
wise approach as suggested by Newitt et al. (1955) can be applied for solid liquid
transport. Doron et. al (1987) calculated pressure drop in inclined pipe considering three
layer distribution i.e., stationary bed layer, moving bed layer and heterogeneous layer.
22
Literature Survey
Turian and Yuan (1977) and Turian et. al (1971) determined pressure drop correlation in
pipelines, which was developed using experimental analyses for various flow regimes
like saltation flow, heterogeneous flow and homogeneous flow. For computing total pres-
sure drop across a pipe line Wasp et al. (1977) developed a method, which is based on the
assumption that the total pressure drop in two phase flow can be split into two parts: pres-
sure drop due to homogeneously distributed particles and additional pressure drop due to
heterogeneously distributed particles. Total pressure drop is the sum of the above men-
tioned two components. Theoretical prediction of pressure drop is quite complex for any
type of flow.
A physical mechanistic model for the prediction of pressure drop and flow patterns for
the flow of settling slurries in horizontal pipes was presented by Doron et al. (1987). The
analysis was based on a two-layer model: a stationary or moving bed at the bottom of the
pipe and a heterogeneous mixture of solid particles with carrier fluid at the upper part.
One of the main assumptions underlying their analysis was the absence of slip between
the solid and liquid phases. This assumption seems to be quite unrealistic, especially re-
garding the bed layer, and may hamper the validity of the model. In order to examine the
effect of the no-slip assumption on the model results, the bed layer is considered by
Doron and Barnea (1992) as a modified porous medium, where the velocities of the sol-
ids and the liquid in the bed are no longer identical. The results obtained by using the
modified model shows that the previous no-slip assumption for the bed layer is reasona-
ble. Xiaofeng et al. (2013) studied a two-layer steady model to depict the annular pres-
sure drop for transport of solid particles in the annular space of a gas pipe. It has been
shown that various parameters like inclination, pipe eccentricity, gas density and viscosi-
ty influence the pressure drop for transport of solid particles. Out of these factors, pipe
eccentricity is the most important parameter which influences for the design of transport
system.
Televantos et al. (1979) used a lock hopper feeding system for transport of gravel and
coal-shale slurries at high concentrations through a pipe. Measurements of velocity, de-
livered concentration, hydraulic gradient, in-situ concentrations and particle velocities
have been performed. The above parameters, predicted theoretically using two-layer
(Figure 2.12) model, have been found to be in good agreement with the measured values.
Motousek (2002) discussed a relation between the flow friction and the flow pattern in
pipeline transport of sandwater mixture. The study is based on laboratory experiments.
Kayiem et al. (2013) assumed two-layer approach to estimate the shear forces as well as
dry friction force in a solid-liquid flow through a horizontal pipe. The developed mathe-
matical model is tested against one available model Ramadan et al. (2005) that also ap-
plies the pseudo hydrostatic pressure method using three-layer approach. Based on the
calculated results, there is lack of agreement between Kayiem et al. (2013) and Ramadan
et al. (2005). Wilson (1989) demonstrated and made experiment to check that the friction
behavior of mobile beds at high shear stress is inherently different from either of the two
classical cases, i.e., smooth-wall and rough-wall cases. For those two cases, the character-
istic lengths for friction are based on viscosity and grain size, respectively.
Tanaka and Tsuji (1991) used Lagrangian Simulation to evaluate the effect of inter-
particle collision in the vertical pipe. They showed that irregular bouncing is not effective
enough to produce particle diffusion in a vertical pipe. Inter-particle collisions have a de-
cisive effect on the particle diffusion even at a small solid loading. The obtained results
for distribution of solid concentration and velocity, matched well with experimental data
of Tanaka et al. (1989). Worster and Denny (1955) find pressure drop in vertical and pipe
bend experimentally. Lin and Ebadian (2008) was introduced a simplified 3-D algebraic
slip mixture (ASM) model to obtain the numerical solution in sand-water slurry flow. It
was found that the difference of friction factors between single- and double-species slur-
ries increase along flow direction in most part of the entrance region.
Gillies et al. (1991) predicted head losses for coarse-particle or settling slurries using an
improved version of the two layer model. Ramadan et al. (2005) has developed and ap-
plied a three-layer solid transport model to simulate solid particle transport in inclined
channels. The results from a series of experiments are compared with the model predic-
tions. Turian Yuan (1977) developed pressure drop correlations for flow of slurries in
24
Literature Survey
pipelines for each of the four flow regimes: flow with a stationary bed, saltation flow,
heterogeneous flow and homogeneous flow.
In gas-liquid transport the flow-pattern maps are widely are used. Doron and Barnea
(1995) have been demonstrated the potential of flow pattern maps as a tool for research
and design of slurry pipelines. The transitions between the flow patterns are then predict-
ed by means of the mechanistic three-layer model of Doron and Barnea (1993) and com-
pared with experimental data.
Continuity and momentum equations proposed by Doron and Barnea (1993) to obtain
flow parameters described below.
Continuity equations for the two phases i.e., liquid and solid particles are
For the solid particles,
... ... ... (2.2.5)
and, for the liquid phase,
( ) ( ) ( ) ... ... ... (2.2.6)
Momentum equations are given as follows
Force balance equation in heterogeneous layer:
... ... ... (2.2.7)
Force balance equation for the moving bed layer:
... ... ... (2.2.8)
Bagnold (1954) considered the collisions between particles of a rapidly sheared granular
material consisting of identical spherical grains. He also claimed that Due to momentum
exchanged in a collision and the frequency of collisions are proportional to the mean rate
of shear, the shear stress and the normal stress must both be proportional to the square of
the mean shear rate.
According to the stationary bed layer by Doron et al. (1987):
... ... ... (2.2.9)
Figure 2.2.6 Area and Surface in three layer model
2.2.2.2 Critical velocity
Critical Velocity refers to the velocity, below which deposition of sediment occurs, and
above this velocity no deposition takes place i.e., at critical velocity the flow rate is bare-
ly fast enough to keep the solid particles in suspension. Stationary bed is formed when
fluid velocity is below the critical velocity and in this bed no sediment transport occurs.
Oroskar and Turian (1980), Turian et al. (1986) established a correlation between critical
velocity and solid concentration by using large amount of experimental data and found
that the maximum critical velocity corresponds to concentration range of 25% to 30%.
Ramadan et al. (2003) presents a mechanistic model to predict the critical velocity and
shown that the model predictions and experimentally measured data show satisfactory
agreement with predicted velocity. The critical velocity of solids in an inclined channel is
significantly influenced by the particle size in relation to the viscous layer thickness.
Xiaofeng et al. (2013) evaluated rolling velocity and lifting velocity for the solid particle
26
Literature Survey
in a deposited layer and then defined critical velocity as minimum of the above two ve-
locities.
2.2.3 Bed formation
During filling process of blind backfilling method a large area is filled-up by sand
transport in channels of meandering nature. The sand particles move through these chan-
nels formed over the previously deposited sand bed. These channels are similar to self-
formed meandering alluvial channels. It is required to study the literature related to trans-
verse profile (Figure 2.2.7) of self-formed stable alluvial channels. Sand particles carried
along these channels in the transport zone get deposited in the discharge zone (Figure
2.2.1). The sand deposition phenomenon is similar to advancement of seafloor stratigra-
phy as depicted in Figure 2.2.8. Therefore relevant literature about the progress of sea-
floor stratigraphy has also been reviewed in this section.
2.2.2.1 Transverse profile
The cross-section of channel during filling process of blind backfilling is similar to allu-
vial channel. Cao and Knight (1998) designed the hydraulic geometry of alluvial channel
using the definition of friction between bank and solid particle on bank and compared the
observed depth of channel with theoretical depth of the channel. Ikeda and Izumi (1991),
Parker G. (1978) developed cross sectional using momentum balance equation. Eaton and
Millar (2004) predicted alluvial channel width using rational regime model and compared
it with alluvial channel obtained from both laboratory experiments and natural channels.
Using 48 data points, Davidson and Hey (2011) developed empirical relationship for av-
erage values of bankfull width, mean depth, slope, meander arc length and sinuosity in
terms of bankfull discharge and associated bed material load, the caliber of bank vegeta-
tion density and valley slope. To determine the stability of alluvial-channel formation
Huang et al. (2002) applied physical principle of least action. Principle of least action is
an approach to characterize the entire pattern of a motion using scalar quantities like en-
ergy and work, without reference to all the forces acting on or within the system. Using
integral analysis, it examines the possible trajectories of paths of motions and demon-
strates that motion occurring in nature. There are many expressions of LAP, of which
Hamiltons LAP and minimum potential energy (MPE) condition is used by the author.
Leopold and Maddock (1953) expressed average height, average width, average velocity
and average slope in the form of power function of discharge. Using conservation princi-
ples and sediment transport laws Smith (1974) evaluate hydraulic geometry. Julien and
Wargadalam (1995) evaluated hydraulic geometry of alluvial channels by solving flow
rate, resistance to flow, secondary flow and particle mobility equations.
Type A Type B
z ;
a

o y
Figure 2.2.7 Transverse profile of channel cross section (Source Cao S 1995)
2.2.2.2 Longitudinal profile
The progress of longitudinal profile of deposited sand bed in blind backfilling process as
shown in the discharge zone of Figure 2.2.1 is similar to the advancement of seafloor
stratigraphy as observed in Figure 2.2.8. Physics-based numerical models provide an
ideal platform to examine the response of morphology to changing sedimentary and geo-
logical conditions. OGrady and Syvitski (2001) has shown profile is stable which is
evaluated from numerical experiments with the multi process, basin-fill model SedFlux.
Syvitski et al. (1998) adopted two-dimensional advection-diffusion equation describing a
28
Literature Survey
turbid hypopycnal plume emanating from a river, is solved by ANSI-standard FORTRAN
77 model. Hutton and Syvitski (2008) used sedflux 2.0, the newest version of sedflux
basin filling model. This introduces a series of new process models and the ability to op-
erate in one of two models to track the evolution of stratigraphy in either two or three di-
mensions. Peckham (2008) established a mathematical model and its solution method for
advancement of seafloor stratigraphy for different initial bathymetry profiles. Perron and
Royden (2013), Royden and Perron (2013) used an integration of the steady-state form of
the stream power equation to evaluate bedrock river profile. It is observed in the litera-
ture, that river bed profile forms are exponential, logarithmic, power function with a
degree of negligible error.
Figure 2.2.8 Longitudinal Profile of Seafloor Stratigraphy (Source Pckham 2008)
2.2.4 Concentration profile in longitudinal direction
In the discharge zone as shown in Figure 2.2.1 the progress of sand bed occurs through
gradual deposition of sand particles. In this zone the sloping edge of the deposited sand
bed causes gradual increase in cross-sectional area of flow, which decreases the flow ve-
locity. The sand concentration in the flow also gets reduced as the distance increases. In
this section literature relevant to the variation in concentration in the field of hydrology
has been reviewed. Syvitsky et al. (1998) has solved steady two-dimensional advection-
diffusion equation (ADE) describing a turbid hypopycnal plume emanating from river
mouth. Kumar et al. (2010) has solved one-dimensional ADE with variable co-efficient
under both uniform and increasing nature of input concentration. Jaiswal et al. (2011),
Kumar et al. (2009) have developed analytical solution of one dimensional ADE for the
evaluation of concentration profile when both the co-efficients, namely dispersion co-
efficient and flow-domain solid transportation co-efficient, are time dependent. Jha et al.
(2011, 2012) have analytically solved advection-diffusion equation in cytosolic calcium
concentration under variation of biophysical and physiological parameters.
Blind backfilling process in actual mines mainly consists of flow channels similar to self-
formed channels that occur in loose soil. The transport mechanism is similar to the flow
of particles in a heterogeneous mixture that forms during flow below the critical velocity.
When sediments of rivers get discharged into sea the advancement of shoreline is similar
to the bed advancement in backfilling process. Therefore in the present work the theoreti-
cal modeling of the backfilling process will be carried using similar procedures of sedi-
ment transport in river flow.
30
Chapter 3: Experimental Investigation
3 Experimental Investigation
3.1 Introduction
Hydraulic blind backfilling technique is commonly used for subsidence control of the
strata over unapproachable water-logged underground excavations. In air-assisted gravity
backfilling process, the injected air, if not released completely from the rise side holes,
may cause troubles by way of creating cracks and potholes on the surface. The pumped-
slurry technique has shown filling up of a relatively larger area at a faster rate, especially
when high-volume, low-pressure method was selected. The simple gravity hydraulic
backfilling has been found to be an equally effective method as slurry pumping or air-
assisted gravity backfilling, especially when flow rates are high (Walker J S 1993, Boger
D V 2012).
There exists number of differences between backfilling methods in active mines and
those used in blind backfilling of abandoned mines. The main difference is the lack of
access to the abandoned mine voids. The backfilling operation in active mines can be di-
rectly observed. The backfilling operation in such cases can be controlled at the location
of underground mine where the fill material is required to be deposited. In abandoned
mines due to lack of access, all the work must be done from the surface in remotely con-
trolled fashion (Pal S K 2001- 2002, Sand et al. 1990).
In hydraulic gravity blind backfilling, slurry of backfill material is gravity fed through a
borehole (either a drilled borehole or the mine shaft) into the water-logged underground
mine until the borehole will not accept any additional backfill material. The quantity that
can be fed by gravity through a single borehole depends not only on the conditions of un-
derground, such as the slope, height, and the proximity of pillars in the mine workings,
but also on backfill material flow characteristics such as slurry flow rate and sand con-
centration in the slurry (Whaite et al. 1975) etc. In this research, some experimental stud-
ies have been carried out on simple gravity hydraulic blind backfilling in a fully water-
logged transparent model of a section of underground coal mine worked by a straight un-
derground mine gallery.
32
Experimental Investigation
3.2 Experimental setup
Fully transparent 14 mm thick perspex sheets have been used to construct the horizontal
model of 4.9 m X 0.4 m X 0.26 m size to simulate pattern of the filling process. The ge-
ometric scale of the model is 10:1 and it is used in a gravity backfilling set-up as shown
in Figure 3.2.1. The complete transparent rectangular model is fabricated with water-
tight joints. One inlet hole of 25 mm diameter is provided to feed sand-water mixture into
the model and two outlet holes are kept for water outflow from two ends. The height of
the inlet feeding pipe along with its top funnel has been kept as 1.5 m in order to limit the
maximum head of slurry column to 1.5m. The total arrangement of the experimental set-
up is shown in Figure 3.2.1.
Figure 3.2.1 Schematic diagram of the experimental set up for gravity blind backfill-
ing.
3.2.1 Water feeding, metering and recirculation arrangement
A small capacity centrifugal pump is used to deliver water through a turbine type flow
meter into the mixing tank. The water-feeding arrangement is of recirculation type. Vol-
umetric flow rate of water is varied by means of a manually operated throttle valve. The
water flow rate is displayed by the flow meter through an LED display unit located in the
control panel.
3.2.2 Variable-speed sand feeding conveyor and metering arrangements
River sand is sieved initially through 2 mm sieve to remove oversize particles and then
through ASTM No. 140 sieve to remove finer particles below 100 . The main purpose
for removing finer particles is to allow quicker settling of sand particles and provide a
clear view of the filling process in the transparent model. The sieved sand of -2 mm +
0.105 mm size is finally stored in a tank in sufficient quantity to permit complete filling
of the model. The stored sand is transported to the mixing funnel using a variable-speed
bucket elevator type conveyor. Sand flow rate is measured by calibrating sand quantity
transported with the rpm of the conveyor.
To make sand-water slurry of any desired concentration, metered amounts of sand and
water are fed into the mixing tank with a tapered-funnel like bottom. As the fed sand and
water gravitate down the inlet feeder pipe, they get mixed up and form homogeneous
slurry.
3.2.3 Data acquisition system for continuous recording of inlet pressure of the slurry
and sand-water flow rates
The static pressure of the slurry at the inlet point of the model is transferred to a pressure
transducer by a capillary tube inserted on the wall of the inlet pipe close to the inlet point.
The pressure transducer measures the static pressure of the inlet slurry column and in fact
this pressure is responsible for causing slurry transport inside the model. An automatic
data acquisition system connected to a computer receives output signals from the pressure
sensor, water flowmeter and the rpm meter of sand conveyor. The data acquisition soft-
ware is calibrated to receive signals in terms of respective units of the measured parame-
ters and record them every second. The installed data acquisition software has the capa-
bility to display time-variations of the measured parameters in graphical form in separate
windows as well as to store the measured data in an Excel file.
34
3.3 Experimental procedure and measurement
Initially few experiments have been conducted in the transparent model to critically ex-
amine the filling process and to justify the viability of the gravity filling method. The fill-
ing process is described in Section 3.3.1. Several experiments are carried out to establish
various empirical relationships relating to flow parameters and it is discussed in Section
3.3.2.
3.3.1 Filling Process
At first the slurry from the mixing funnel gravitates down through inlet pipe to the water-
logged void space of the model. As soon as the slurry enters a large void space of the
model, its velocity gets reduced quickly and as a result the kinetic energy of slurry be-
comes insufficient to keep the solid particles in suspension. Thus, the solid particles drop
down by getting separated from the slurry. In this way a conical-shaped heap is formed at
the bottom of the model directly below the inlet point and continues to grow in size. As
the height of the heap approaches the roof, the gap between the roof and the fill material
at the top of the heap gradually reduces and the slurry velocity in this gap increases suffi-
ciently enough to keep the solid particles in suspension again and finally push them to the
sloping edge of the conical heap. The entering slurry from the inlet hole impacts on the
top of the conical heap just under the bottom of inlet hole. The turbulence created by the
water helps the sand particles to be in suspension and transports them through the small
gap between the top of the conical heap and underside of the top cover of the model and
finally the particles continue to get deposited along the slope of the conical heap. In this
way, the slanted surface of the conical heap advances almost equally towards both left
and right side of the inlet hole of the model. It is observed that at the beginning, four
channels existed, but ultimately after some time the number of channels gradually re-
duced to one (Figure 3.3.1.1), thereby transporting solids to any one side of the model at
one time. When the length of the deposited bed in any one direction increases sufficient-
ly, so that the pressure required for transporting the solids along the long channel in that
direction becomes very high, a new shorter route in the opposite direction is created by
puncturing the top layer of the deposited sand bed. As a result, the older channel slowly
gets blocked and the deposited sand bed continues to grow in the other direction. In this
way, the sand bed advances almost equally towards both directions in an alternate manner
(Figure 3.3.1.2). Finally, jamming occurs when the maximum possible slurry head avail-
able in the model is not sufficient to puncture a new alternative flow path.
Four-channel flow Three-channel flow
Inlet Inlet
Two-channel flow Two-channel flow
Inlet
Inlet
Single-channel flow Single-channel flow
Inlet
Inlet
Figure 3.3.1.1 Channel formations of filling process - arrows show the path of
channel.
36
Figure 3.3.1.2 Sequence of the filling process
3.3.2 Studies on effect of slurry flowrate and sand concentration on maximum area
of filling from a single inlet pipe
3.3.2.1 Experimental studies on repeatability of blind backfilling process
It is perceived that the filling process being a natural system can be replicated by repeat-
ing the experiment. In order to test this repeatability a few trials have been carried out by
repeating the same experiment with same slurry flow rates and volumetric sand concen-
trations. It is found that, there is negligible change in total volume of sand throughput
(Table 3.3.2.1). To study the effect of slurry flow rate Q and sand concentration C on
maximum volume of sand throughput Vs, different experiments with varying slurry flow
rate and sand concentration have been conducted.
Table 3.3.2.1 Repeatability experiments for maximum sand throughput
From the above Table 3.3.2.1 it has been seen that average standard deviation of is
6.22 repeated observation. Table 3.3.2.2 shows maximum volume of sand throughput
for different flow rate and sand concentrations. When sand volume concentration 6% for
all flow rates except 15 lpm flow rate it has been seen that the model is fully filled-up.
3.3.2.2 Experiments with varying flow rate and concentration
Further experimental studies on hydraulic gravity blind backfilling are conducted with
slurry flow rates of 15, 20, 25 and 30 litre/min and sand concentrations of 6 %, 10 % and
14 % by volume. From these studies, it is observed that with decrease in sand concentra-
tion, the area of filling increases. The results of total volume of sand throughput are
shown in Table 3.3.2.2. It may also be noticed that the area of filling can be increased by
increasing the total slurry flowrate at any given sand concentration. In other case the area
38
of filling can also be enhanced by diminishing the sand concentration for a fixed slurry
flow rate. Therefore, in order fill-up a large area in a minimum possible time, it is advised
that the filling should start with very high flow rate and high concentration and then a
gradual reduction of concentration should be carefully done with progress in filling. Fig-
ure 3.3.2.1 shows variation in sand throughput versus concentration and flow rate for dif-
ferent flow rate and different sand volume concentration respectively.
Fully Filled-up
Figure 3.3.2.1 Variation in volume of sand throughput for different flow conditions
Table 3.3.2.2: Maximum Volume of Sand Throughput,
3.4 Empirical relationships
3.4.1 Maximum volume of sand throughput
In the previous section it has been observed that the maximum volume of sand through-
put ( ) is function of both slurry flowrate ( ) and sand concentration ( ). After remov-
ing the values for full filled-up condition of the model, a relationship among , and
has been evaluated from the experimental data. During this process different functions
;
of and have been tried and finally a relationship between and as shown in
Figure 3.4.1.1 is obtained with a correlation coefficient of 0.78. The general form of this
relationship is given in Equation (3.4.1).

, (litre) (3.4.1)

40

Figure 3.4.1.1 Maximum volume of sand throughput, versus

Equation (3.4.1) suggests that the maximum volume of sand throughput ( ) can be in-
creased either by increasing the flowrate, (keeping concentration constant) or by de-
creasing the sand concentration, (keeping flowrate constant). Again, from the very na-
ture of this relationship, it can be easily understood that reduction in concentration ( )
can increase the maximum volume of sand throughput ( ) more effectively than increase
in flowrate ( ).
For field experiments it is easier to measure water flowrate ( ) and sand flowrate ( )
using measuring instruments and therefore another attempt is made to obtain a relation-
ship of in terms of and . After several attempts for ( ) a very simplified
equation is obtained with a correlation coefficient of 0.78 (Figure 3.4.1.2). This relation-
ship is given as
, (litre) (3.4.2)

(litre)

Figure 3.4.1.2 Volume of sand throughput versus

A surface plot for the relationship given in Equation (3.4.2) is shown in Figure 3.4.1.3
where the values of continuous variable is taken in the range 500 to 2500 litre/min and
the same for is kept in the range of 500 to 1500 litre/min. The above values of and
can be used for actual filling operations in abandoned mines. The value of can thus
be predicted for any set of chosen and values. The predicted value of obtained
from Equation (3.4.2) can be used to estimate the area that could be filled up in the mine.
This will assist mining engineers to plan for the filling scheme. On critical observation of
the Figure 3.4.1.3 it can inferred that the water flowrate ( ) of 12,000 litre/min and/or
higher in combination with sand flowrate ( ) of 1,200 litre/min or lower (indicated by
the arrow lines in the Figure 3.4.1.3) may be selected for backfilling in the actual mines.
This can produce quicker rise in the value of sand throughput ( ) from a single input
borehole and thereby improve the efficiency of filling.
42
Figure 3.4.1.3 Volume of sand throughput for different sand flowrate and water
flowrate
3.4.2 Flow pressure loss
The effect of slurry flow rate and sand concentration on pressure loss per unit length is
studied by measuring the mean flow pressure loss during progress of sand bed. As shown
in Figure 3.4.2.1 the average slope of the plot between pressure loss ( ) and length of
channel ( ) indicates the pressure loss per unit length ( ) for sand transport through
sediment channel. The data for all the experiments is furnished in Table 3.4.2.1.
Pressure Loss (m)
Length (m)
Figure 3.4.2.1 Pressure loss versus length of the channel for 25 lpm flowrate
and 6% concentration
Table 3.4.2.1 Pressure Loss for different flow conditions
44
In order to correlate the unit pressure loss ( ) with a single flow parameter various
combinations and have been attempted and finally a moderately good correlation
;
co-efficient as is obtained by plotting against as shown in Figure
3.4.2.2. The general form of this relationship may be given as
( )
(3.4.3)

Pressure loss (m/m)
;
Figure 3.4.2.2 Plot of pressure loss versus ;
This relationship given in Equation 3.4.3, shows that pressure loss is reduced by reduc-
ing the sand flowrate or by increasing the water flowrate. It may also be shown that de-
crease in pressure loss can be more effectively achieved by decrease in sand concentra-
tion as compare to increase in slurry flowrate. A continuous surface plot of the Equation
3.4.3 is shown in Figure 3.4.2.3 which indicates pressure loss that will occur during
backfilling under any combination of water and sand flowrates.

Figure 3.4.2.3 Flow pressure loss curve for different flow condition (applicable for
field condition)
Table 3.4.2.1 shows for different flow conditions. This maximum pressure loss can
be obtained by multiplying with maximum bed length at the time of jamming. Ta-
ble 3.4.2.2 shows this maximum flow pressure loss under different flowrates and sand
concentrations where jamming have taken place. It may be noted that the average value
of maximum flow pressure loss is 0.1614 m. The maximum head available in the experi-
mental set-up is 1.5 m out of which the mean value of maximum pressure loss for send-
ing slurry through the flow channels just before jamming is only 0.1614 m. Therefore, a
safety factor of more than 1.5/0.1614, i.e., 9.294, needs to be provided to assist punctur-
ing of a new flow-path, failing which jamming will occur.
46
Table 3.4.2.2 Maximum flow pressure loss in channels at jamming
The above concept of safety factor of 9.294 can be assumed to be valid for field condi-
tions also. To validate the obtained results, a field experiment at Kenda area of Eastern
Coalfields Limited (ECL), carried out by Pal S K (2012) has been considered, in which
about 6641.43 lpm flowrate with 14.61% concentration was used for backfilling opera-
tion. It is found that the height of water table in Kenda area is about 8 m which means
that the maximum slurry head available for filling operations in that area is 8 m. using the
concept of safety factor 9.294, it can be said that the maximum flow pressure loss at the
time of jamming would be 8/9.294=0.86077. From the Figure 3.4.2.2 as well as from the
Equation 3.4.3 it has been seen that =971.6 lpm and =5669.4 lpm, the unit pressure
loss is 0.00604 m/m. In that case maximum length of channel at the time of jamming is
0.86077/0.00604=142.5 m. Then maximum volume of sand throughput is
3
3x2.5x142.5=1069 m . In the field application the maximum volume of sand throughput
was 1427 m3 under identical flowrate and concentration (Pal S K 2012). Further model
study on pressure loss during slurry flow has been presented in Chapter 5.
3.4.3 Bed Advancement
Progress of sand bed occurs initially towards both directions and after progress of suffi-
cient length the flow occurs towards one direction at one time and in this way the model
is filled up through alternate filling in each direction. At the end of the experiment the
advancement of deposited sand bed has been found to be nearly symmetrical on either
side of the inlet hole as shown in the Figure 3.4.3.1. If and denote the lengths of
top part and bottom part of the deposited sand bed from centre of inlet hole respectively.
Several experiments have been performed in the laboratory with different flow condi-
tions, and measurements for , are taken at intervals of 5 or 10 minutes depending
upon the rate of bed advancement. Length of deposited sand bed ( ) along the top part
has been shown in the Table 3.4.3.1 for different flow conditions. Figure 3.4.3.2 shows
bed advancement of the top part during filling process of different flow conditions. A lin-
ear trend line has been made all the slope has been shown in the Table 3.4.3.2. It is seen
from this table as flowrate increases slope increase and also as concentration slope is also
increases, which leads to correctness of experimental flow condition (More details in
Chapter 4).
(a)
(b)
Figure 3.4.3.1 Advancement of sediment bed during filling process (a) Top view (b)
Side view.
48
Table 3.4.3.1 Bed advancement during experimental study for different flow conditions
Figure 3.4.3.2 Bed advancement during filling process for different flow conditions
Figure 3.4.3.2 shows bed advancement during filling process for different flow condi-
50
tions. A linear fit has been made and it is seen from the figures that slopes are increased
as flowrate and/or concentration are increased. The slope indicates the rate of bed ad-
vancement and its values are furnished in the Table 3.4.3.2.
Table 3.4.3.2 Bed advancement rates for different flow conditions

Q (l/min)
C (%) 15 20 25 30
6 9.44 11.77 14.56 19.6
10 16.9 23.84 26.94 28.78
14 23.09 31.93 37.64 35.6
Rate of bed advancement
(mm/min)

Figure 3.4.3.3 Rate of bed advancement versus
From the Figure 3.4.3.3 it can be conclude that a relationship for rate of bed advance-
ment in terms of and is
, (mm/mi )
() 9 (3.4.4)
Rate of bed advancement =
The above Equation (3.4.4) can be modified for any gallery of height, in metre and
width in metre as
() 9
Rate of bed advancement = , (mm/mi ) (3.4.5)

3.4.4 Channel cross-section
The measurements of height and width of channels have been undertaken at different
points along its length by opening the top cover of the experimental model after stopping
the experiment temporarily. These measurements should be done when only one channel
is present, once at the starting of single channel flow and the next during the last phase of
the experiment. It has been seen that increase in slurry concentration often increase chan-
nel depth by scouring action. Height to width ratio has been evaluated during the experi-
ment for 15 and 30 lpm flowrates under 6%, 10% and 14 % concentrations (Tables
3.4.4.1 and 3.4.4.2).
Figure 3.4.4.1 Channel cross-section
15 lpm flowrate
Measurement Measurement Measurement
C Ratio C Ratio C Ratio
Side Side Side
6 Left Side 27.65 10 Left side 19.81 14 Left Side 12.2
6 Left Side 28.75 10 Left side 17.95 14 Left Side 11.87
6 Left Side 28 10 Left side 19.2 14 Left Side 12.1
Average 28.13 Average 18.99 Average 12.06
6 Right Side 28.5 10 Right Side 18.5 14 Right Side 11.54
Average 28.44 Average 18 Average 12.5
Mean of both side 28.29 Mean of both right side 18.49 Mean of both side 12.28
52
30 lpm flowrate
Measurement Measurement Measurement
C Ratio C Ratio C Ratio
Side Side Side
6 Left side 27 10 Left side 17.5 14 Left side 12.5
6 Left side 29 10 Left side 18 14 Left side 11.6
6 Left side 31 10 Left side 20 14 Left side 12.8
Average 29 Average 18.5 Average 12.3
6 Right Side 28 10 Right Side 17.4 14 Right Side 12.2
6 Right Side 28.5 10 Right Side 18 14 Right Side 13.5
6 Right Side 27 10 Right Side 19 14 Right Side 12
Average 27.83 Average 18.13 Average 12.6
Average of left Average of left Average of left
28.42 18.32 12.43
and right side and right side and right side
From the Table 3.4.4.1 and Table 3.4.4.2 the empirical relationship for height width ratio
has been evaluated.
Width and height ratio
Concentration in fraction
Figure 3.4.4.2 Width and height ratio relationship
From the Figure 3.4.4.2 it can be conclude that the width and height ratio relationship is
(3.4.7)

where is the inlet slurry concentration in percentage.
3.4.5 Pressure signature and Flow pressure loss
Continuous recording of inlet pressures was made using automatic data acquisition sys-
tem. Figure 3.4.5.1 shows a pressure-time curve recorded by the data acquisition system.
Healthy phase Unhealthy

phase
filling process.
During experiment pressure fluctuation has been recorded using DAS. The pressure sig-
nature obtained during the experiments is observed to have two distinct phases:
(i) Healthy/ normal phase where the pressure fluctuations are low, and
(ii) Unhealthy/ abnormal phase where the pressure fluctuations are severe in na-
ture.
3.5 Discussions
It has been found that the process of gravity backfilling is a natural phenomenon with
nearly 6% accuracy of repeatability. Two empirical relationships, and
54
have been developed to predict the maximum volume throughput for a

given flow condition. A general relationship, has been established among
unit pressure loss . /, sand flowrate ( ) and water flowrates ( ). All the relationships
indicate that increase in or decrease in can be achieved by reducing the sand
flowrate or by increasing the water flowrate. It may also be shown that increase in or
decrease in can be more effectively achieved by decreasing the sand concentration in-
stead of increasing the slurry flowrate. 3D-plots as shown in Figures 3.4.1.3 and 3.4.2.2,
for finding volume of sand throughput and as well as pressure loss per unit length under
wide range of sand and water flowrates have been plotted to obtain values of these pa-
rameters ( a ) directly from the 3D-plots for any set of flow variables. To
simulate field conditions with = 971.6 lpm and = 5669.4 lpm, the unit pressure loss
is obtained as 0.00604 m/m, whereby the maximum length of channel at the time of
jamming is found to be 142.5 m. In that case the maximum volume of sand throughput is
3x2.5x142.5=1069 m3, which nearly matches the maximum volume of sand throughput,
i.e., 1427 m3 in actual field application under identical flowrate and concentration (Pal S
K 2012).
Table 3.4.3.2 shows the empirical rate of bed advancement during filling process. Which
shows that as flowrate and/or concentration is increased rate of bed advancement is also
increased which shows that correctness of the flow conditions. Relationship for rate of
bed advancement for any gallery of height, in metre and width in metre as
:
(mm/mi ). Width-to-height ratio of the channel is given by, This in-
;
dicates that this ratio is influenced mainly by slurry concentration.
From the Figure 3.4.5.1 it is observed that the pressure time curve can be divided into
two sections, one is healthy phase and another is un-healthy phase. The distinguishing
between the two phases is done from the frequency of pressure data fluctuation.
Chapter 4: Progress of deposited sand bed
4 Progress of deposited sand bed
4.1 Introduction
In the experimental investigation on filling-up of a long straight water-logged gallery,

which has been carried out using gravity blind backfilling technique in a fully transparent
scaled model, the progress of sand bed occurs by deposition of transported sand particles
at its sloping edge in the discharge zone (Figure 2.2.1). This dynamic process gives rise
to a continuous and systematic progress of the deposited sand bed in the discharge zone.
The longitudinal profile of the deposited sand bed is similar to that of sediment transport
in various fields such as, alluvial channel stratigraphy (Bridge J S. 1979; Deutsch and
Tran. 2002; Mackey and Bridge. 1992; Leopold and Modak 1953), seafloor stratigraphy
(Peckham 2008; Syvitski et. al. 1999) etc.
In Figure 3.3.1.2 of Chapter 3, a three-dimensional view of filling process has been pre-
sented. The rates of bed advancement for different flow conditions have been estimated
from an empirical relationship and presented in Table 3.4.3.2. In this chapter attempts
have been made to find out other relationships for bed slope angle and bed advancement
using semi-empirical and analytical approaches.
4.2 Bed slope angle
Figure 4.2.1 represents the sequence of filling in two-dimension. From the measurements
of top and bottom edges of deposited sand-bed during gradual progress of filling, slopes
of the bed-edge have been determined. Slopes of the bed edge continuously change as the
bed progresses in length. Sand particles at the sloping edge get deposited initially at up-
per part of the bed-edge (Figure 4.2.2a) and thus the overall bed slope angle increases.
Then, due to sudden collapse of steeper and heavier top part of the slant edge (Figure
4.2.2b), overall bed slope angle gets reduced. These cycles of increase and decrease of
overall slope angle continue and the bed-edge advances gradually.
56
Progress of deposited sand bed
Figure 4.2.1 Sequence of the filling process
Figure 4.2.2 Cyclic (a) deposition and (b) collapse during the experiment
The above measurements for bed slope advancement have been carried out with varying
flowrates of 15, 20, 25 and 30 lpm under 6%, 10% and 14% sand concentration. Figure
4.2.3 shows variation of bed slope angle with time for different flow conditions. From
this figure it has been seen that there is cyclic increase and decrease in slope angle in total
course of filling process, due cyclic deposition and collapse of the sand bed-edge.
Figure 4.2.3 Variation of bed slope angle with time for different flow conditions
From these plots of variation of slope angles with time, the highest and lowest slope an-
gle values for each flow condition are noted and demonstrated in Figure 4.2.4. From this
58
figure it may be noted that maximum slope angle for any fixed concentration decreases as
flowrate is increased. At increased flowrate, higher discharge velocity at the sloping edge
causes deposition of sand particles at the middle or near the bottom of bed-edge. Thus the
maximum slope angle is reduced. Similarly with increase in sand concentration under any
given flowrate, the total sand flowrate increases at the discharge point, which has a ten-
dency to rapidly load the upper part of the slope causing sliding, collapse at short inter-
vals. Similarly, the minimum slope angle also reduces with increase in flowrate due to
higher discharge velocity causing throw of sand particles to longer distance from the dis-
charge point leading to flatter slope angle. However the minimum slope angle does not
vary considerably with variation in sand concentration since there is no significant
change in total flowrate or discharge velocity.
Figure 4.2.4 Maximum and minimum bed slope angle for different flow condition
Since there is continuous increase and decrease in bed slope angle during the filling pro-
cess, the maximum, minimum and average variation of bed slope angle has been evaluat-
ed for each flow condition and furnished in Table 4.2.1. However the minimum bed
slope angle can be assumed to indicate the most stable configuration of the slope for any
particular flow condition. The average of minimum bed slope angle, is calculated as
34.4 and is given in Table 4.2.1.
Table 4.2.1 Variation of maximum, minimum and average bed slope angle
Flowrate, Concentration, Maximum Minimum Average

Q C angle angle angle
15 6 45.2 37.1 40.9
20 6 44.2 34.4 38.8
25 6 42.6 33.8 37.8
30 6 39.1 32.5 36.2
15 10 44.2 37.1 41.8
20 10 42.8 34.6 39.4
25 10 40.6 33.5 35.7
30 10 38.4 32.2 35.9
15 14 42.8 36.5 39.2
20 14 40.4 35.1 37.6
25 14 38.3 33.6 34.9
30 14 37.2 32.5 35.4
Average Values 41.3 34.4 37.8
To evaluate the dependency of slope angle with slurry flowrate and sand concentration,
empirical relationships have been developed from Figures 4.2.5, 4.2.6 and 4.2.7.
60

Figure 4.2.5 Maximum bed slope angle versus
Figure 4.2.6 Minimum bed slope angle versus 1/

Average angle

Figure 4.2.7 Average bed slope angle versus

The empirical relationships for maximum, minimum and average bed slope angle are pre-
sented by the Equation (4.2.1), (4.2.2.) and (4.2.3).
() (4.2.1)
8 (4.2.2)
(4.2.3)

During variation of experimental conditions the range of maximum angle, changes

from 37.20 to 45.20 and minimum value, ranges from 32.20 to 37.10. The average
angle, lies inbetween and , and ranges from 34.90 to 41.80. The slope
62
angles may be used to evaluate bed advancement during filling process as discussed in
the next section.Maximum slope angle, is reduced at increased flowrate due to
higher discharge velocity at the sloping edge causes deposition of sand particles at the
middle or near the bottom of bed-edge. Similarly with increase in sand concentration un-
der any given flowrate, the total sand flowrate, increases at the discharge point, which
has a tendency to rapidly load the upper part of the slope causing sliding collapse at short
intervals. However the minimum slope angle does not vary considerably with variation in
sand concentration since there is no significant change in total flowrate or discharge ve-
locity.
4.3 Bed advancement
Throughout the experiment it is observed that sand particles are discharged at the top of
the slant edge of deposited sand-bed. The slope angle of slant edge increases and de-
creases in a cyclic manner causing gradual advancement of the deposited sand bed. The
Figure 4.3.1 shows the different layers of deposited sand bed after post-collapse stabili-
zation.
Figure 4.3.1 Direction of bed advancement
In this section the exact nature of the sloping bed-edge curve called longitudinal bed pro-
file is found out through semi empirical approach. Longitudinal bed profile (Figure
4.3.2) in different fields has been studied worldwide (Peckham 2008, Syvitski 1999, Syv-
itski 1989, Syvitski et al. 1998). In most of the cases the longitudinal profile has been as-
sumed to be exponential. In some other cases the longitudinal profile is assumed to be
linear or logarithmic. In the present study an attempt has been made to evaluate the bed
profile assuming linear longitudinal profile initially.
Bed advancement direction
Bed height in direction

Longitudinal
Profile
Bed length in direction

Figure 4.3.2 Longitudinal profile of bed advancement
4.3.1 Bed advancement with assumed linear profile in the slanted part
In case of linear longitudinal profile, the shape of deposited sand bed becomes trapezoi-
dal with , being distance of upper and lower edges of the deposited sand bed respec-
tively as shown in Figure 4.3.1.1.
Sand Bed
Figure 4.3.1.1 Side view of deposited sand bed.
64
Pal et al. (2001) assumed linear longitudinal profile with average slope angle of 320. In
the present case an elaborate exercise on different slope angles such as , , and
have been found out in Section 4.2 and furnished in Table 4.2.1. For evaluation of
for a given quantity of sand input in a gallery of known dimension and the slope
angle, has been used because it corresponds to the most stable condition of bed-
edge. By equating the volume of sand input during a given time with the volume occu-
pied by trapezoidal sand bed, a theoretical formula for computation of can be ob-
tained as
W (mm) (4.3.1)

where is swell factor due to swelling of sand in the deposited sand-bed in the fully
flooded model. The value of has been experimentally determined and its average value
has been found to be .
From the experimental data a graph of with is plotted for all concentrations as
shown in Figure 4.3.1.2.

(litre/min)
Figure 4.3.2.1 A plot of with
From Figure 4.3.2.1 an empirical relationship is obtained as
(4.3.2)
Combining Equations (4.3.1) and (4.3.2) the following general relationship is obtained

(4.3.3)

Figure 4.3.2.2, 4.3.2.3 and 4.3.2.4 represents comparison of measured values of bed ad-
vancement during filling process with computed values of the same using Equation
4.3.3. A good agreement between measured and computed values justifies the validity of
Equation 4.3.3.
66
Figure 4.3.2.2 Comparison of experimentally measured and semi-empirically com-

puted values multiplied with swelling factor of bed advancement for different
flowrates under 6% concentration
68

All of the Figures 4.3.2.2, 4.3.2.3 and 4.3.2.4 show a common trend that the computed
values are lower than measured values. These deviations between computed and meas-
ured values have been considered to be the errors in the averaged measured values of
when compared with the actual volume of sand throughput calculated using .
Table 4.3.1.1 shows percentage of error for all flow conditions.
Computed value is based on the occupied volume of sand in the flooded model. But
during averaging of three measured values in experimentation an error can occur as
depicted in Figure 4.3.2.5. The average measured value is usually higher than the com-
puted values.
Figure 4.3.2.5 Top view of the right side of deposited sand bed showing the difference
between computed and average (measured)
4.3.1.1 Maximum length of deposited sand bed
Maximum volume of sand throughput, which occurs during filling under any given
flowrate, and concentration, can be obtained by using Equation 3.4.1. By combining
Equations 3.4.1 and 4.3.3, the maximum bed advancement, for a given flowrate,
and concentration, C can be expressed as,
70
x
4 5 (mm) (4.3.5)

Measured and semi-empirical (Equation 4.3.5) bed advancement has been plotted in
Figure 4.3.2.5. It has been seen that the experimental results have shown a good agree-
ment with the semi empirical results. Owing to full fillingup of the model for 20, 25, 30
litre per minute flowrate and 6% concentration the measured points under these flow
conditions do not match with the semi-empirically computed results. However for all
other cases the semi-empirical relationship as given in Equation 4.3.5 can be used for
finding out the maximum length of the packed sand bed, under any given
flowrate, Q and concentration, C.
Figure 4.3.2.5 Maximum bed advancement versus flowrate for different flow conditions
4.3.2 Bed advancement with assumed exponential profile in the slanted part
In the previous section the profile of the bed-edge slope has been assumed to be linear
and the computed bed advancement is compared with measured bed advancement values.
Table 4.31.1 shows the error in the computed values of bed advancement. In order to re-
duce the mismatch between the computed and measured values of bed advancement a
more accurate profile for the bed-edge needs to be assumed. Syvitski (1989), Syvitski et
al. (1998) have considered the sediment deposition rate to be varying exponentially with
average concentration gradient and Peckham (2008) evaluated the profile which is expo-
nential when river entering to the receiving basin using the findings of Syvitski (1989),
Syvitski et al. (1998). Accordingly, in this section the slanted part of the deposited sand
bed, i.e., bed-edge profile, is assumed to be non-linear according to the equation
(;) (4.3.3)
The co-efficient a in the Equation (4.3.3) is evaluated using built-in Levenberg-

Marquardt algorithm in SPSS, by comparing theoretically calculated volume of deposited
sand bed with the actual volume computed from known sand flowrate multiplied by time.
The parameter may be determined by introducing two boundary conditions such as,
( ) and ( ) . Figure 4.3.2.1 shows the diagrammatic representa-
tion of curved longitudinal profiles of the bed-edge in dotted lines.
Figure 4.3.2.1 Exponential longitudinal profile of slanted part of bed edge
The equation of the curved slope thus obtained from Equation (4.3.3) which satisfy the
boundary conditions, ( ) and ( ) is
( ) (; )
(4.3.4)

So, the occupied volume of deposited sand bed inside the model can thus be calculated as
72

4 5 (4.3.5)

In Equation (4.3.5) putting , total length of upper part of the sand bed is given
by
( ) (;)

(4.3.6)

The Equation (4.3.6) can be simplified to

4 ( (;) )5
( ) , mm

(4.3.7)
In the Equation (4.3.7), the parameter ( ) needs to be evaluated in term of some

known controllable parameters, like flowrate, and concentration . In order to evaluate
the dependence ( ) with and an extensive work done by trial and error meth-
od and finally an empirical relationship of ( ) with has been arrived
and it is depicted in Figure 4.3.3.2.
)
(

Figure 4.3.2.2 Plot of ( ) with
The general relationship as obtained from Figure 4.3.3.2 may be expressed as
(4.3.8)
Combining Equations (4.3.7) and (4.3.8), the bed advancement, can be written as
3 7)
4 ( (
)5
( )
, mm

(4.3.9)
74
The parameter is unknown and it has been evaluated by equating the right hand side of
the Equation 4.3.9 with measured values for each flow conditions. The Table 4.3.2.1
shows all the obtained values of the parameter by using SPSS.
Table 4.3.2.1:Values of coefficient values using SPSS

C
Q 6 10 14
15 -0.047 -0.044 -0.061
20 -0.049 -0.093 -0.064
25 -0.038 -0.040 -0.042
30 -0.042 -0.027 -0.036
The range of values of varies from -0.027 to -0.093 for all flow conditions used in the
experimental work. As the range of variation of is small, an average value of -0.049
has been used in Equation 4.3.9 and compared with measured values to obtain error
in computed measurements.
4.3.3 Error comparison
Error comparison has been made to obtain the most appropriate longitudinal profile for
the bed-slope. Table 4.3.3.1 shows comparison errors between linear and exponential
profiles for the bed-slope. It is clearly observed that the error in predicted bed advance-
ment, can be considerably lowered by assuming an exponential bed-slope profile. The
average error with linear profile 21.9% whereas the same for exponential profile is only
6.4%.
Table 4.3.3.1 Comparison of percentage errors values computed
using linear and exponential profiles
Figure 4.3.3.1 shows a typical error comparison of measured bed advancement with pre-
dicted values of bed advancement using exponential and linear longitudinal profiles for
20 lpm flowrate and 10 % concentration.
Figure 4.3.3.1 Error comparison of different longitudinal profile for bed ad-
vancement with measured value
It may be clearly seen from the Figure 4.3.3.1 that prediction of bed advancement can be
more realistically done using Equation 4.3.9, which considers exponential shape of bed-
slope.
76
4.3.4 Visualization of bed advancement
Since it has been already convened that exponential profile for the bed-slope matches
well with the experimental results, the visualization of longitudinal profile of deposited
bed-slope has been carried out on the basis of exponential shape of bed-slope. The gradu-
al progress of bed-slope may be computed from Equation 4.3.7, where can be
replaced by . Then the resulting relationship may be written as

x 4 ( ( cot ) )5
( )
, mm

(4.3.10)
The bed advancement will stop when jamming occurs and at that time the maximum
length of bed advancement, is reached. As discussed in Section 3.4.1 the maximum
volume of sand throughput is dependent on flowrate, and sand concentration, . Table
3.3.2.2 shows volume of sand throughput for different flow condition. Using multivariate
regression analysis maximum volume of sand throughput can be expressed as
(4.3.11)

So, the maximum time of filling, for any flow conditions in a gallery can be given
as
(4.3.12)

Using this in the Equation (4.3.10) the maximum bed advancement for any given
flow condition can be computed as
;
99
, mm
4 ( ( cot ) )5
( )

(4.3.13)
For known flow conditions, visualization of bed advancement by plotting the bed-slope
profile using MATLAB program. Figure 4.3.4.1 shows progress of bed advancement for
10 % concentration with different flowrates under laboratory condition.
Figure 4.3.4.1 Final positions of filled-up bed for 10% concentration with different
flowrate
In the Figure 4.3.4.1 blue line represents the different positions of bed-edges during pro-
gression of filling process and red line represents final position of the bed-edge at jam-
ming. From the laboratory experiment, it has been seen that bed advancements are 1310
mm, 1197 mm, 982 mm and 581 mm, have occurred for 30, 25, 20 and 15 lpm flowrates
with 10% concentration, respectively and these values are nearly matching with the plot-
78
ted positions of the final bed-edge in Figure 4.3.4.1. The relative percentage of error be-
tween measured and computed values of have been presented Table 4.3.2.3 of previ-
ous section.
Equation (4.3.13) holds good in the laboratory experimental set-up where the maximum
available head at the time of jamming is limited to 1.5 m. In the field condition the avail-
able head is much more than 1.5 m. Therefore, in order to obtain the maximum duration
of filling under field condition, the time in the Equation (4.3.12) should be multi-
plied by a factor .

; 0 1 (4.3.14)

The time, t in the Equation (4.3.10) can now be substituted by ; to obtain max-
imum length of bed advancement under field condition. The final equation becomes
99 , mm

9x [ ] 4 ( ( cot ) )5
( )

(4.3.15)
Where h is the maximum available head at the time of jamming in field condition
The maximum length of bed advancement, as obtained from the Equation

(4.3.15) can now be compared with the same for the blind back-filling application in a
coal mine in Kenda area of Eastern Coalfields Limited (ECL) (Pal S K 2012), where the
flowrate was about 6200 lpm and 9% concentration. A plot of bed progress for this field
condition is presented in Figure 4.3.4.2.The maximum available head was 8m. Height
and width of mine gallery is 2.5 m and 3 m respectively. Figure 4.3.4.2 shows maximum
length of filled-up length to be approximately 82 m. This has been verified from the field
application work of Kenda mine of ECL, India (Pal S K 2012).
Figure 4.3.4.2 Longitudinal profile for 9% concentration with 6200 lpm flowrate
4.4 Discussions
In the discharge zone of filling process, sequential increase and decrease of bed-edge
slope takes place during continuous deposition of sand particle at this zone. When slurry
flowrate, Q is increased keeping the concentration, C fixed, the maxi mum bed-slope an-
gle, decreases due to longer distance of throw of sand particles caused by higher
velocity of slurry (Figure 4.2.4). When sand concentration, C is increased keeping the
slurry flowrate, Q fixed the maximum bed-slope angle , decreases due to rapid load-
ing of the upper part of the slope, leading to frequent collapse of the slope, thereby not
allowing the slope angle to steepen much (Figure 4.2.4). Similarly, the minimum slope
angle, also reduces with increase in flowrate due to higher discharge velocity caus-
ing throw of sand particles to longer distance from the discharge point, which led to flat-
ter slope angle (Figure 4.2.4). The developed empirical relationships for maximum, min-
imum and average bed slope angle as given in Equations (4.2.1), (4.2.2.) and (4.2.3).
Two empirical relationships have been developed for bed advancement using both linear
and exponential profiles for the bed-edge slope. These are as follows:
3 7)
x 4 ( (
)5
( )
, mm

where the constant a, is determined by comparing computed with measured values
80
for all flow conditions using SPSS. The average value of is obtained as -0.049. Com-
parison between the above two empirical relationships has shown that computation of bed
advancement, using exponential profile has a better agreement with its measured val-
ue. The average error in computation of bed advancement using exponential profile is
less than 1/3 rd that of linear profile. For visualization of bed advancement, the Equation
(4.3.10), as given below, may be used

x 4 ( ( cot ) )5
( )
, mm

For any time the length of bed advancement, can be obtained from the above rela-
tionship. At the time of jamming, the following equation can be used to compute the
maximum length bed advancement, .
99 , mm

x 4 ( ( cot ) )5
( )

The above equation holds good in the laboratory experimental set-up where the maxi-
mum available head at the time of jamming is limited to 1.5 m. In the field condition the
final length of bed advancement at the time of jamming may be obtained from the Equa-
tion (4.3.15) as
99 , mm

9x [ ] 4 ( ( cot ) )5
( )

where a=-0.049, =1.055 ,average of minimum bed slope angle degree

and is maximum available head at the time of jamming.
Chapter 5: Channel Cross-Section and flow Pressure Loss
5 Channel Cross-Section and flow Pressure Loss
On examination of the filling process of blind backfilling, it is revealed that the basic
process of filling occurs by sand transportation along one or more meandering channels.
During filling it has been seen that the maximum length of channel at the time of jam-
ming, depends on the flow parameters. Occurrence of jamming is connected with exces-
sive pressure loss in the channel flow, which is beyond the available pressure head. In
this chapter estimation of pressure loss has been attempted through theoretical analysis.
In this process, a detailed work on channel cross-section, concentration and velocity pro-
files in channel flow are established to arrive at the pressure loss estimation.
In this case, two prominent layers are found: one of them is stationary bed layer and other
is moving heterogeneous layer. Within the stationary bed layer there is no movement of
solid particles and within the moving bed layer sand particles are transported with the
flowing water. The larger sediment particles are subjected to deposition due to higher
gravitational forces and stationary bed is formed, whereas smaller particles remain in
suspension owing to the action of turbulent forces in the fluid medium and advances for-
ward within the moving bed layer. It has been detected that there is partial similarity pre-
sent in the filling process of the laboratory experiment with the solid liquid transport
through pipe line, when solid liquid mixture flows at a speed below critical velocity.
Doron et al. (1993, 1987) considered continuity and momentum equations to obtain flow
parametersi for slurry flow in pipe. Ramadan et al. (2005) also considered continuity and
momentum equations to evaluate flow parameter. In the present study computation of
channel geometry and pressure loss in the flow path has been attempted using different
layer formation during filling process. The variation of solid concentration in the hetero-
geneous layer is also found out.
5.1 Model formulation
Cross-section of the channels creation during filling process in laboratory experimental

model as well as in the fields is similar to that of alluvial channels geometry. Cao and
i
height and velocity of every layer, pressure gradient etc.
82
Channel Cross-Section and flow Pressure Loss
knight (1998) designed the geometry of alluvial channel using the definition of friction
between bank and solid particles. In the present study geometry of channel cross-section
has been evaluated. Geometrical representation of the channel design is given in the Fig-
ure 5.1.3.
During filling process of blind back filling the entrainment, transport and deposition of
sand in field as well as in the model is dependent partially on the sand properties, settling
velocity, sand concentration and dynamic viscosity of sand water mixture. During the ex-
periment it is observed that the sand particles drops down and start moving forward along
the bed. Due to bed formation the gap between top of the model and the deposited sand
bed decrease which cause increase in velocity of slurry flow all around the inlet hole and
finally this leads to the sand transportation through a limited number of channels formed
around the inlet hole as shown in Figure 5.1.1. The side view of a channel is shown in
Figure 5.1.2. The pressure drop in flow channels depends on some parameters like chan-
nel cross-section, velocity of flow, sand concentration etc. To evaluate flow pattern and
pressure drop using layer distribution, the geometry of channel cross-section should be
determined.
Figure 5.1.1 Top view of the filling process and channel formation
Figure 5.1.2 Side view during filling process
From the definition of static friction Cao and Knight (1998) determined the transverse
slopeii of self-formed channels, at each point of slopping side of flow-channels during
filling process as

( ) . /

(5.1.1)

( ) . /
}
where , y-axis is taken in transverse direction and z-axis is taken vertically up-
wards, . The shape of the channel cross-section is shown in Figure 5.1.3, where
the two sides are assumed to be of parabolic.

;
a
Figure 5.1.3 Geometry of flow channels in blind Backfilling
It has been noted from the data available on river cross-section, that the range of width to
depth ratio in river channel is much higher than that in the blind backfilling channel. This
is quite natural as in the present study the slurry having significantly higher concentration
can flow with a very strong erosive power, causing greater depth of flow. These blind
backfilling channels transports sediments, which have been observed to take place
ii
The transverse slope on either side of a channel will be symmetrical and adequate to prevent the accumu-
lation of water.
84
through self-formed channels. In the filling process two prominent layers are observed,
one is called a stationary bed layer and the other is a moving bed layer which consists
of a heterogeneous mixture of sand and water and is formed over the stationary bed layer.
The boundary between the two layers is delineated by the perimeter of the channel cross-
section (Figure 5.1.4). Therefore geometry of channel cross-section should be deter-
mined to obtain flow pattern and pressure drop in channels.
Figure 5.1.4 Cross-sectional view of two layer model

Geometrical properties which follow from the Equation (5.1.1) and Figure (5.1.4) can
be written as
(5.1.2a)
( )
(5.1.2b)

2 . /3

(5.1.2c)

:

( )

( ) (5.1.2d)
(5.1.2e)
(5.1.2f)
Where and are width and height of the water logged volume of the experimental set-
up. is height of the stationary bed layer.
At the inlet of a moving bed layer the slurry flowrate, is given by,
(5.1.3)
where and are cross sectional area and mean velocity of moving bed layer re-
spectively.
Force balance equation in moving bed layer gives (Figure 5.1.4),
(5.1.4)

Distribution of solid particles in the moving bed layer is assumed to be governed by well-
known one-dimensional diffusion equation
( ) ( ) (5.1.5)

where is a function of ( ) only and and are diffusion coefficient and terminal
settling velocity of the hindered particles respectively. The concentration distribution,
( ) can be obtained following Doron and Barnea (1993) by integrating Equation
(5.1.5) two times with suitable boundary conditions, as
86

( ) ;
(; ) (5.1.6)
The average concentration, in the moving bed layer can be given as

( )

Substituting the values of ( ) from Equation (5.1.6) and from Equation (5.1.1)
it follows that,

(; ) ( )
; (5.1.7)

and are hindered settling velocity and diffusion coefficient respectively. is inlet
slurry concentration in volume fraction, is average concentration in volume fraction
of moving bed layer and it is assumed that in the channels of transport
zone. ( ) , assuming tightly packed maximum cubic packing.
5.2 Computational procedure for supporting variables

For the numerical evaluation of dimension of flow channel and pressure-drop across the-
se channels it is required to evaluate all the necessary supporting parameters pertinent to
the experimental conditions.
For finding out the terminal settling velocity of hindered particles, in the Equation
(5.1.7), it is important to evaluate, at first, the value of un-hindered settling velocity of a
particle, inside the flowing fluid. Un-hindered settling velocity of a particle, inside
a fluid medium, is given by
( ) (5.2.1)

, drag co-efficient is dependent on particle Reynolds number and
;
when and for (Bird et
al. 1960). Figure 5.2.2 shows a re-iterative evaluation procedure for determination of un-
hindered settling velocity, .
Since filling is main aim so concentration is high. So cluster of particles is happens and
is hindered particle setting velocity where is evaluated using the procedure suggested
by Richardson and Zaki (1954) which is
( ) (5.2.2)
where is average solid concentration (in volume fraction) in upper layer, is un-
hindered settling velocity of particle and for and m
9 for , is Reynolds number of hinder particle. Figure 5.2.2 shows a
re-iterative evaluation procedure for determination of hindered settling velocity, .
In the Equation (5.1.7), is diffusion coefficient, evaluated according to Taylor (1954)

as
(5.2.3)
(6.3.13)
where is shear velocity as given by and is hydraulic diameter of
moving bed layer which may be written as . Televantos et al. (1979) as-
:
sumed layer formation during transportation of coarse particles of high concentration in

fluid medium and found a relationship for the evaluation of the friction coefficient,
which can be expressed as

(5.2.4)

where, the Reynolds number in the moving bed layer, .
88
The hydrodynamic shear stress in moving bed layer, and that along the interface of
moving bed layer and stationary bed layer, are given by
(5.2.5)
( ) (5.2.6)
is considered to be equal to zero since no movement is observed in the stationary bed

layer.
Density of the moving bed layer, can be expressed as
( ) (5.2.7)
Einstein A (1906) gave a formula to evaluate viscosity of feeding mixture which is set as
( ) (5.2.8)
kg/ms
where is the viscosity of flowing water, the value of which may be assumed as
; ;
x x kg/ms
Guth and Simha (1936) gave modified Einstein (1906) formula to get viscosity of feeding
mixture which can be recognized as
( ) (5.2.9)
The friction coefficient, at the top of the moving bed layer is evaluated from

(5.2.10)

where and =0.2, if the flow is turbulent and and =1 if the
flow is laminar as described by Doron et al. (1987). is hydraulic diameter of moving
bed layer.
5.3 Solution method for the model
Equations (3.4.7), (5.1.1), (5.1.3), (5.1.4), (5.1.7) are five equations with five unknowns
b, L, , , . Using various relationships given in Section 5.2 the different support-
ing variables like, , , , , , , etc. can be evaluated and used for the
evaluation of final flow parameters, such as, pressure drop, , channel height, and
channel width, c .
Stepwise procedure is given below:

1. Known: height and width of void space ( and W), particle size ( ), density of
sand ( ), density of water ( ). The maximum slope angle at the sides of the
flow channel has been assumed to be equal to 32.20, which is the lowest value of
minimum bed slope angles ( ) as may be observed in Table (4.2.1) of Chap-
ter 4.
2. For known flowrate and concentration, settling velocity of un-hindered particle
and hindered particles have been evaluated. The formulas for evaluation of set-
tling velocity have been shown in Equations (5.2.1) and (5.2.2).
3. Find height and width ( and ) of each layer using Equations (3.4.7), (5.1.1),
and (5.1.7). Since it is not possible to solve the Equation (5.1.7) analytically, so
evaluation has been done within a loop. In the loop every is possible to find
out.
4. Shear stresses have been evaluated using Equations 5.2.5 and 5.2.6.
5. Find out pressure drop, ,height, and width, , of the channel.
The detailed evaluation procedures for estimation of flow parameters relevant to the ex-
perimental condition are described through various flowcharts shown in Figures 5.3.1,
5.3.2, 5.3.3, 5.3.4 and 5.3.5.
90
Figure 5.3.1 Flow chart for the evaluation of pressure drop for a particular flow condition
In the flowcharts, shown in Figures 5.3.2 and 5.3.3 the evaluation procedure for settling
velocity of hindered particle and for actual stationary bed height is furnished. Figure
5.3.4 shows the evaluation procedure for shear stresses and . Using all these it is
possible to find out flow pressure loss as shown in Figure 5.3.1.
Figure 5.3.2 Evaluation procedure for settling velocity of hindered particle
92
Figure 5.3.3 Evaluation procedure for friction coefficients
Figure 5.3.4 Flowchart for evaluating stationary bed height

Figure 5.3.5 Evaluation procedure for shear stress
In the above example the data for river sand has been used for design of the model. In
many mining application alternative material to river sand is now being attempted. For
applicability of the developed model for backfilling with other alternative material, the
values of solid density ( ) and mean particle size ( ) needs to be incorporated.
5.4 Channel shape
In order to determine the channel shape, its width and height should be found out at first.
Evaluation procedure for the height of the stationary bed has been discussed in the Fig-
ure 5.3.3. Using Equations (3.4.7) and (5.1.1) it is possible to evaluate height and width
of the channel. Figures 5.4.1, 5.4.2 and 5.4.3 shows height, width and area of the chan-
nels for different flow conditions. From Figure 5.4.1 it is seen that height of channel in-
creases with increase in both concentration ( ) as well as slurry flowrate ( ). However
the width of channel increases with increase in flowrate but decreases with increase in
concentration. It may also be noted from the Figure 5.4.3, that increases in slurry
flowrate and/or decrease in sand concentration can increase the flow-area of channels
thereby make the flow process smoother. Decrease in concentration reduces the channel
height, but, at the same time, it increases the area of flow-channel by sufficient widening
the channels for ease in flow.
94
Figure 5.4.1 Variation of channel height with slurry flowrate for different concentration
Figure 5.4.2 Variation of channel width with slurry flowrate for different concentrations
Figure 5.4.3 Variation of channel area of moving bed layer with concentration for differ-
ent flowrate
In order to visualize the shape of flow channels a MATLAB code is developed to plot the
exact shape of the flow-channels under any given flowrate and concentration. Figure
5.4.4 shows 4 different shapes of channels at 10% concentration with 15, 20, 25 and 30
lpm flowrates. Figure 5.4.5 shows channel shapes with 25 lpm flowrate and 8 10 12 and
14 % sand concentration. From these figures it is concluded that with increase in flowrate
at any given concentration the width as well as depth of channels increase, whereas at any
fixed flowrate with decrease in concentration channel-depth decreases, but the channel
area can still increase by increasing the width of the channel sufficiently enough to over-
shadow the decrease in depth. From an experimental results of gravity blind back filling
in an inclined model (Pal and Mukhopadhyay, 2007) it has been found that for the value
of channel cross-section varied from 500 to 1100 mm2 for flowrate varying from 15 - 30
lpm and concentration ranging between 6 and 15%. From the Figure 5.4.3 it may be seen
that for similar flowrate and concentration regime the channel cross-section varies from
325 mm2 to 1275 mm2 which is very close to the experimental values.
96
Figure 5.4.4 Channel shape for different flowrate with 10% concentration
Figure 5.4.5 Channel shape for 25 lpm flowrate and different concentrations
The above findings have been extrapolated for high flowrates and concentrations to suit
field applications as carried out by Pal S K (2012) and Figure 5.4.6 shows the channel
cross-sections for a few sets of field condition. The channel widths and depths appear to
be practically feasible, but its field verification could not be done as none of the research-
ers, who have applied hydraulic blind backfilling process in mines, have not measured or
reported the channel dimensions.
Figure 5.4.6 Channel cross-section for different flowrate with 10%

concentration for field application
5.5 Concentration profile in the moving bed layer

The concentration profile in the moving bed layer of transport zone is given in Equation
; ( ; )
5.1.6 as ( ) . To obtain the concentration distribution along the
depth of the moving bed layer, the values of settling velocity of hindered particles, and
the diffusion coefficient, are to be found out. The procedure for determining values of
and have been described in Section 5.3. Using MATLAB code, distribution of sand
concentration in moving bed layer has been computed and shown in Figures 5.5.1, 5.5.2.
In the analysis it has been considered that the maximum solid concentration occurs at the
interface of moving and stationary bed layer, which is 74% and as the depth reduces the
concentration decreases. The rate of decrease is nonlinear and nearly similar for different
flowrates as long as the average concentration of slurry remains same (Figure 5.5.1). As
98
found from the MATLAB program, the depth of channel increases from 5.1 mm to 6.8
mm when flowrate is increased from 15 lpm to 30 lpm.
Concentration (fraction)
Channel depth (mm)

Figure 5.5.1 Concentration profile at different depth (mm) of the channel for different
flowrate with 10% concentration
As shown in Figure 5.5.2 the solid concentration of slurry decreases along the height of
the channel from the bottom but, the rate of change of concentration is high towards the
bottom for 10% concentration, whereas it is low for 14% concentration.
Figure 5.5.2 Concentration profile at different depth of the channel for different concentra-
tion with flowrate 25 lpm
In Figure 5.5.2 it has been seen that channel depth increases and concentration profile
gets more flat when concentration is increased from 10% to 14%.
5.6 Velocity in the upper layer
Due to sand deposition at the bottom of the model, the flow area decreases which causes
increase in flow velocity. Figure 5.6.1 represents change in slurry velocity in the moving
bed layer with flowrate.
100
Figure 5.6.1 Variation of velocity (m/sec) for moving bed layer with flowrate (lpm)
It may be seen from the figure that for 6% concentration there is mild increase in slurry
velocity of the order of 6.8% when flowrate is increased from 15 to 30 lpm. Similar in-
crease in velocity is observed for 10% and 14% concentrations. But, for any given
flowrate, the increase in velocity with increase sand concentration is very significant. As
an example, the increase in sand concentration from 6% to 14 % has led to an increase in
velocity by 104% when the flow rate is kept constant at 20 lpm. The velocity regime in
the blind backfilling model studies in an inclined model has been reported (Pal and
Mukhopadhyay 2007) to be varying between 0.23 m/s and 0.77 m/s for the flowrate vary-
ing from 15 lpm to 30 lpm with concentration ranging from 6 to 15%. Figure 5.6.1 as
obtained from theoretically developed relationships shows similar variation of velocity
values under identical flow conditions.
5.7 Pressure drop
When liquid starts to flow, energy dissipation occur which in turn, appears in the form of
a pressure drop. Doron et. al. (1987) calculated pressure drop in inclined pipe for three
layer distribution. Turian and Yuan (1977) determined pressure drop correlation in pipe-
lines. Wasp et al. (1977) method is based on the assumption that the total pressure drop in
two phase flow can be split into two parts: pressure loss due to homogeneously distribut-
ed particles and additional pressure loss due to bed formation.
The flowchart for the evaluation of pressure drop is already shown in the Figure 5.3.1.
Based on this flowchart, a MATLAB code has been generated to evaluate the relationship
of pressure drop with slurry flowrate and sand concentration. Figure 5.6.1 represents var-
iation in pressure loss per unit length with slurry flowrate for 6%, 10% and 14% sand
concentrations. It is observed from the figure that pressure loss per unit length decreases
marginally with increase in flowrate for any given concentration, whereas there is signifi-
cant decrease in pressure loss per unit length with reduction in sand concentration from
14% to 6%.
Figure 5.7.1 Pressure loss per unit length for different flow condition
The above findings have led to a conclusion that fillingup of larger area using high slur-
ry concentration, requires sufficiently large pressure loss which may not be available in
102
the field application of gravity backfilling process for filling-up of voids lying at shallow
depth. Therefore it is preferable to use gravity blind back filling method with low concen-
tration and high flowrates. In this section of the study the experimental pressure drop has
been compared with the theoretically modeled results. Table 5.7.1 shows a comparison
between flow pressure loss obtained from the above theoretical model and the experi-
mental flow pressure loss. From this table it is seen that the entire experimental pressure
loss values are much below the theoretical values. This may be due to the presence of air
during experimental process. Since it has been observed during experimental process that
air enters the experimental model in the form of bubbles which can steer the flow better
and thereby decrease in the pressure drop. It is already found in the Chapter 3 that de-
crease in the pressure drop make the flow smoother. Thus entrainment of air bubble in the
flow-path assist puncturing of new flow-paths and make the filling process healthy. To
match the two sets of values for pressure losses the experimental pressure loss needs to
be multiplied by a factor of 8. Figure 5.7.2 shows a graphical comparison of the two sets
of pressure loss values. It may be seen from the figure that the two pressure loss values
have a better match when flowrates are higher i.e., 25 and 30 lpm.
Table 5.7.1 Comparison of experimental and theoretically modeled pressure
loss for different flow conditions
Concen- Pressure loss from Experimental Experimental

Flowrate ./ ./
tration theoretical model
30 6 0.227 0.0326 0.2605
30 10 0.556 0.0593 0.4741
30 14 1.032 0.0948 0.7580
25 6 0.2355 0.0486 0.3890
25 10 0.5753 0.0810 0.6480
25 14 1.062 0.1130 0.9000
20 6 0.2455 0.0493 0.3947
20 10 0.5998 0.0978 0.7827
20 14 1.114 0.1092 0.8733
15 6 0.2587 0.0659 0.5275
15 10 0.632 0.1027 0.8219
15 14 1.175 0.1017 0.8136
104
(a) (b)
(c) (d)
Figure 5.7.2 Comparison of theoretically modelled flow pressure loss with 8 times
experimental pressure loss for (a) 15 lpm (b) 20 lpm (c) 25 lpm and (d) 30 lpm
5.8 Discussions
It has been seen from Section 5.4 as flowrate is increased channel cross-sectional area
has been increased. From Section 5.6 as increase in velocity at channel cross-section ar-
ranged increase in area of the cross-sectional portion which ended filling process to be
healthy, which prepared decrease in pressure drop. Also it has been seen depth of channel
is increased as concentration increase. When concentration is more particle is present
which can dig the bed which helps to increase depth of channel. When total flowrate is
increased depth of channel also increased. Area and width of channel is increased as
flowrate increased and/or concentration is decreased. As shown in Figure 5.5.2 the solid
concentration of slurry decreases along the height of the channel from the bottom but, the
rate of change of concentration is high towards the bottom for 10% concentration, where-
as it is low for 14% concentration. For change in flowrate with fixed concentration, con-
centration profile has same trend. It may be seen from the Figure 5.6.1 that for 6% con-
centration there is mild increase in slurry velocity of the order of 6.8% when flowrate is
increased from 15 to 30 lpm. Similar increase in velocity is observed for 10% and 14%
concentrations. But, for any given flowrate, the increase in velocity with increase sand
concentration is very significant. As an example, the increase in sand concentration from
6% to 14 % has led to an increase in velocity by 104% when the flow rate is kept con-
stant at 20 lpm. It has been found that theoretical values for cross-sectional area of flow
channels nearly match with the experimental results of previous researcher (Pal and
Mukhopadhyay 2007). Figure 5.6.1 as obtained from theoretically developed relation-
ships shows similar variation of experimental model studies in an inclined model of ve-
locity values under identical flow conditions.
Figures 5.7.1 shows pressure loss per unit length (m/m) versus mixture flowrate for
6% 10% and 14% input concentrations respectively. It is observed from the Figures 5.6.1
as concentration increases pressure gradient is also increases which show that difficulty
during filling process of blind backfilling technique. But for all concentration critical ve-
locity is same. Theoretical result favorably compare with experimental results. It is shows
from the Figures 5.6.1 increase in mixture velocity increases make pressure gradient de-
crease from where it is prominent that filling process is healthy. When mixture velocity
more than critical velocity pressure gradient increases which implies start in unhealthy
region. It is observed that as slurry velocity is increased pressure drop decreases. Increase
in slurry velocity in flow channel causes increase in its area of the cross-section due to
scouring effect, thereby produces reduced pressure loss per unit length. Theoretically
computed pressure drop has been found to be in good agreement with the experimental
results.
106
Chapter 6: Pre-jamming indication parameter
6 Pre-jamming indication parameter
6.1 Introduction
Subsidence from old abandoned underground water-logged coal mines has become an
everyday concern of many people living in coal producing regions. In earlier days, most
coal mines in India had been worked under shallow cover and were partially extracted
using bord and pillar system of working. Correct records of plans for many of these
mines are not available and also not reliable in many occasions. As time passes, the
strength of supporting coal pillars left behind in the abandoned mines gradually reduces
and collapse of several adjacent pillars often leads to sudden subsidence in the overlying
region. Deterioration of pillars can also occur due to the acid mine drainage in abandoned
mine as reported in a case study of an abandoned coal mine near Carbondale, Illinois,
USA (RoyChowdhury et al. 2016). Some authors have used probabilistic approach to
evaluate mine subsidence (Galve et al. 2009a; 2009b). To avoid problems of ground sub-
sidence, backfilling technique is used to fill-up the void space of abandoned mine (Ay-
dan, & Ito 2015, Li, & Pengyu 2015, Li 2013, Andreyev, & Koons 2010; Wissmann, &
Peterson 2013) without dewatering. Aydan, & Ito (2015) have presented the effect of the
depth and ground water on the formation of ground subsidence associated with aban-
doned room and pillar lignite mines under static and dynamic conditions and discussed
the effects on the areas above abandoned lignite mines. Dang et al. (2013) has described a
research work on the feasibility test of using phosphogypsum as a backfill aggregate, fo-
cusing on the mixture ratio of filling materials and the filling process.
In most of earlier mines lying at shallow depth depillaring with stowing was not practiced
due to several constraints and this has led to the high risk of subsidence in the areas lying
above those mines. Therefore, the coal mines which had been worked by caving method
became either fully or partially water logged and have now become unapproachable. De-
velopment of suitable methods to stabilize such unapproachable mine voids had been ne-
glected in past. The area to be filled-up cannot be directly controlled in such case. By us-
ing gravity blind backfilling technique, the occurrences of sudden, pre-mature jamming is
a major hurdle in filling-up of large area from a single borehole. To achieve successful
108
Pre-jamming indication parameter
filling of a large area from a single borehole, it is essential to evaluate a pre-jamming in-
dication parameter.
In the present research, it is aimed to find out a pre-jamming indication parameter

through a detailed experimental study on simple gravity blind backfilling method and in-
depth analysis of inlet pressure variation during the progress of filling.
A fully transparent model has been fabricated (detailed has been described in Chapter 3)
and used in a gravity backfilling set-up to critically examine the filling process and ana-
lyze the inlet pressure variation during filling-up of the model.
In this experimental study slurry flowrates of 20, 25 and 30 litre/min have been used and
sand concentrations have been kept at 6%, 10% and 14% by volume. Therefore a total
combination of nine experiments has been conducted and the volume filled-up by the de-
posited sand bed during each experiment is noted.
Table 6.1.1: Area of filling for different flow conditions

Flowrate, Q Concentration, C Maximum filled-up
(litre/min) (volume %) Area (mm 2)
20 6 14,66,667
20 10 6,66,666
20 14 5,60,000
25 6 14,66,667
25 10 12,25,000
25 14 6,41,666
30 6 14,66,667
30 10 9,16,200
30 14 6,58,000
Table 6.1.1 depicts the deposited area of sand bed at the end of each experiment. It may
be observed that with decrease in sand concentration, the area of filling increases. Also
the area of filling can be increased by increasing the total slurry flowrate keeping sand
concentration constant. Filling-up of large area from a single inlet hole is very beneficial
and also probably cost effective since filling-up of the same large area will require multi-
ple inlet holes when smaller area of filling occurs with higher concentrations and/or low-
er flowrates. In later case, the cost of drilling additional holes can make the total filling
job less costlier than the first case.
6.3 Results
6.3.1 Pressure Loss and Bed Advancement
The pressure loss data for different lengths of channel is obtained from the continuously
recorded inlet pressure data. A sample average of 20 pressure data is considered against
the corresponding bed length data. In this way a number of data sets of average pressure
loss and corresponding channel length have been obtained for a particular flowrate and
sand concentration. Figure 6.3.1 shows the plots of these pressure loss data with bed ad-
vancement under different flow conditions. It may be noticed from the best-fit line of the
data that pressure drop increases as bed advances because of increase in flow channel
length. In every case of linear fit, it is observed that slope of the line is always less than 1,
which indicates that for every mm progress of channel the additional pressure required to
continue the flow is less than 1 Pa under the present flow conditions.
110
Figure 6.3.1 Pressure loss curve with bed advancement for different flow conditions
Table 6.3.1: Slope of best-fit line between pressure loss and bed advancement.
Slope () of the linear fit equa-

S. No. Flowrate Concentration tion between pressure loss P R2
and bed advancement
1 20 6 0.44 0.92
2 20 10 0.45 0.85
3 20 14 0.88 0.74
4 25 6 0.44 0.93
5 25 10 0.34 0.79
6 25 14 0.86 0.55
7 30 6 0.34 0.86
8 30 10 0.70 0.84
9 30 14 0.85 0.81
From the Table 6.3.1, it may be seen that variation of slope is present for different flow
conditions. To find the healthy condition for the filling process, S is plotted against
and the best-fit linear plot is shown in Figure 6.3.2.

Figure 6.3.2 Plot of slope S versus

The relationship as obtained from the best-fit line is given in Equation (1)

, (Pa/mm) (1)
112
is in litre/min and is the volumetric concentration of solid in slurry in percentage.
This implies that lower flow pressure loss can occur with increase in flowrate or decrease
in concentration, which can lead to higher quantum of filling under a constant hydraulic
head.
6.3.2 Pre-Jamming Indication Parameter
Continuous recording of inlet pressures is accomplished by using the automatic data ac-
quisition system. Figure 6.3.2.1 shows a typical pressure-time curve recorded by the data
acquisition system for the slurry flowrate of 25 lpm and 14% sand concentration.
40000
30000
Pressure(Pa)
20000 J
A
M
M
I
10000
N
G
0
0 500 1000 1500 2000 2500 3000 3500 4000
Time(S)
Figure 6.3.2.1 Pressure-time curve recorded by the data acquisition system
during the filling process.
The restless nature in the pre-jamming condition creates a special signature pattern which
can be readily identified by human eye. It has been observed that the pressure fluctuation
in the initial phase of filling is low as compared to the same during the final phase of fill-
ing. The magnitude of pressure fluctuation and the exact shape of these curves differ
when flowrates and sand concentrations are varied. In order to eliminate bias on the esti-
mation of pre-jamming condition, a statistical analysis of pressure-signature curve is re-
quired to be performed to obtain a dependable pre-jamming indication parameter inde-
pendent of flow conditions. This parameter may also be used for obtaining an indication
about arrival of the pre-jamming phase in an automated backfilling system.
A closer inspection of pressure-time data by using different statistical parameters re-

vealed that the entire region of pressure-time curve can broadly be divided into two sec-
tions -
Healthy region, and

Unhealthy region.
In healthy region, pressure fluctuations are low, whereas in unhealthy region, pressure
fluctuations are comparatively high. There is a strong possibility of jamming of the back-
filling process in this unhealthy region and therefore it is essential to identify the arrival
of this unhealthy phase in the pressure time curve to indicate a pre-jamming condition.
Several trials have been made to identify the initiation of this unhealthy, pre-jamming
phase by determining different statistical parameters of the pressure-time data like, vari-
ance ratio, log likelihood ratio and ratio of coefficient of variation (Thill, Hulk, & Steg-
man, 1983) etc. Out of all of the above trials, a reasonably good success has been
achieved from the analysis of coefficient of variation.
The coefficient of variation is a dimensionless parameter and it measures the variability

of a series of numbers independently of unit of measurement used for these numbers. In
the present study the Coefficient of Variation (CV) of a variable window size is used as
the pre-jamming indication parameter. The analysis started after 210 seconds when first
210 pressure data are recorded and the CV for these initial 210 pressure data has been
computed and recorded against a time value of 210 second which is termed as CV210.
Next analysis of CV is performed after 10 seconds when a total of 220 pressure data have
been recorded and the second value of CV is computed and recorded against a time value
of 220 second. In this manner the CV values are computed and recorded against the cor-
114
responding time data at intervals of 10 seconds till the occurrence of jamming. . Figures
6.3.2.2, 6.3.2.3, 6.3.2.4 represent variation of CV for different slurry concentrations and
flowrates. On critical observation of the CV values from these figures, Table 6.3.2.1 is
prepared to show the percentage of filling that has taken place at the time when the CV
value reaches 1.3 times the CV of first 210 data-set (CV210). It may be concluded from
Table 6.3.2.1 that a minimum of 74% of maximum filled-up area could be achieved
using a cut-off CV value of 1.3 times CV210. Therefore 1.3CV210 can be used as a pre-
jamming indicating parameter for the blind backfilling process under all the flow condi-
tions adopted in the present study.
0.075
0.07
20 lpm flowrate 6 % concentration
20 lpm flowrate 10% concentration
0.065
CV (dimensionless)
0.06
0.055
0.05
0.045
0.04
0 0.5 1 1.5 2 2.5
Time (s) 4
x 10
Figure 6.3.2.2 Plots of coefficient of variation of inlet pressure for a slurry flowrate of 20
lpm and varying concentration of 6%, 10% and 14%.
0.22
0.2
0.18 25 lpm flowrate 6 % concentration

0.16
CV (dimensionless)
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Time (s) 4
x 10
116
0.09
0.085
30 lpm flowrate 6 % concentration

0.08 30 lpm flowrate 10% concentration
0.075
CV (dimensionless)
0.07
0.065
0.06
0.055
0.05
0.045
0.04
0 5000 10000 15000
Time (s)
Table 6.3.2.1 Percentage filling of different experiments with 1.3 coefficient of
variation cut-off.
S. No. Flowrate Concentration Percentage filled at 1.3 CV cut
(liters/min) (%) off level
1 20 6 95
2 20 10 99
3 20 14 99
4 25 6 74
5 25 10 97
6 25 14 85
7 30 6 79
8 30 10 97
9 30 14 84
6.5 Discussions
Occurrence of sudden pre-mature jamming is a major hurdle in ensuring the success of

gravity blind backfilling technique. Evaluation of a pre-jamming indication parameter is
an important step towards avoidance of sudden jamming of the filling process. Using the
developed pre-jamming indication parameter, one can avoid sudden jamming and fill-up
additional area by lowering flow path resistance. The developed pre-jamming indication
parameter for this model can further be applied in field conditions of some abandoned
mines to establish its efficacy in other flow conditions also. Thereafter the pre-jamming
indication parameter can be applied for designing an automated filling system which can
maximize the filling from a single feeding point by understanding the arrival of jamming
phase and thereby taking necessary steps to avoid pre-matured jamming of the gravity
backfilling process.
118
Chapter 7: Conclusions and Future scope of the work
7 Conclusions and Future scope of the work
7.1 Conclusions
A study has been made to describe the filling process of blind backfilling for a known
flow condition as well as develop a pre-jamming indication parameter to avoid sudden
jamming. Concurrently a theoretical model has been formulated for the blind backfilling
process to arrive at some semi-empirical relationships among various flow parameters.
Influence of flow condition on various parameters like maximum volume of sand
throughput, flow pressure loss, rate of bed advancement and channel cross-section have
been investigated empirically and the obtained relationships have been cross validated
with field data. During filling process, the deposited sand-bed advances with cyclic in-
crement and decrement in slope angle due to intermittent collapse of bed-edge caused by
unwarranted deposition of sand particles on its upper part. Relationships for bed ad-
vancement have been evaluated assuming the bed-edge profile to be either linear or ex-
ponential. Average error for bed advancement obtained from experimental study with lin-
ear profile is 21.9% while it is 6.4% for exponential profile. Equations are formulated
assuming that two prominent layers have been present during filling process. Solution of
which determine the shape of the channel cross-section, concentration profile with re-
spect to height within the channel as well as pressure drop during filling process for dif-
ferent flow conditions. Evaluation of pre-jamming indication parameter has been done to
avoid sudden jamming.
As an outcome of the present study the following conclusions are drawn:
1. In gravity blind backfilling process the entering slurry first gravitates down to the
water-logged void space and the solid particles get separated from the slurry and
drop down to the bottom of the void space. As the height of the heap of settled
solid particles approaches the roof, the narrowed gap between the roof and the de-
posited sand bed causes an increase in the velocity of slurry. The turbulence cre-
ated at the higher velocity of the slurry helps the sand particles to be in suspension
and transports them through the narrowed gap. Finally the particles get deposited
along the edge of conical heap. In this way, the slanted surface of the conical heap
120
Conclusions and future scope of the work
advances almost equally in both directions from the inlet pipe before jamming
takes place. The process of filling is found to be very similar to sediment transport
process in self-formed meandering channel.
2. For any given flowrate and sand concentration, the maximum volume of sand
throughput nearly remains same and is repeatable within a minor variation of
nearly 6%.
3. From different empirical relationships developed from the experimental results
the following inferences are made :
i. Maximum volume of sand throughput increases as flowrate increase
and/or concentration decrease. The very nature of their relationship sug-
gests that maximum volume of sand throughput ( ) can be more effec-
tively increased by decreasing the sand concentration ( ) rather than by
increase in the flowrate ( ).
ii. Flow pressure loss decreases when flowrate is increased and/or concentra-
tion is decreased. It may also be shown that decrease in pressure loss can
be more effectively achieved by decrease in sand concentration as com-
pare to increase in slurry flowrate.
iii. Channel height to width ratio mainly depends on sand concentration. Rate
of bed advancement increases with increase in both flowrate and concen-
tration.
iv. Since available head in the field condition is much higher than the availa-
ble head in model study, the obtained relationship is suitably adjusted for
head correction and the results are verified for matching with maximum
volume of sand throughput ( ) in field conditions.
4. From the detailed study on change of bed-edge angle during progress of filling the
following findings have been made:
i. During advancement of deposited sand-bed, increment and decrement in
measured bed-edge slope angle occurs in a cyclic manner, due to deposi-
tion of sand particles at the top of sloping edge, leading to heavier load on
upper part of the bed edge and then sudden collapse of slope occurs by
sliding.
ii. Dependence of maximum, minimum and average bed slope angle on slur-
ry flowrate, Q and sand concentration, C few empirical relationships have
been developed.
iii. When slurry flowrate, is increased keeping the concentration, fixed,
the maximum bed-slope angle, decreases due longer distance of
throw of sand particles caused by higher velocity of slurry. When sand
concentration, is increased keeping the slurry flowrate, fixed the max-
imum bed-slope angle, decreases due to rapid loading of the upper
part of the slope, leading to frequent collapse of the slope, thereby not al-
lowing the slope angle to steepen much.
iv. Similarly, the minimum slope angle, also reduces with increase in
flowrate due to higher discharge velocity causing throw of sand particles
to longer distance from the discharge point, which led to flatter slope an-
gle.
5. Another two semi-empirical relationships have been developed for bed advance-
ment, considering either linear or exponential profile for the bed-edge slope.
Comparison between the above two semi-empirical relationships has shown supe-
riority of the exponential profile due its stronger agreement with the measured
values of .
6. For visualization of bed advancement, may be used for any time the length of
bed advancement, can be obtained from the above relationship. At the time of
jamming, the following equation can be used to compute the maximum length bed
advancement, . The developed formula holds well in the laboratory ex-
perimental set-up where the maximum available head at the time of jamming is
limited to 1.5 m.
7. In the field condition the final length of bed advancement at the time of jamming
may be determined. The maximum length of the deposited bed considering expo-
nential profile of the bed-edge, been found to agree with the field results of previ-
ous researcher. In field condition maximum length of bed advancement at the
time of jamming for known flow conditions and available head can be obtained
from the equation is valid with field data.
122
8. A theoretical model has been developed assuming two layer formations during
solid transport process namely, stationary bed layer and moving bed layer to eval-
uate depth, width and area of flow-channels as well as flow velocity and flow
pressure loss.
i. Theoretically determined solution for channel cross section shows that its
area as well as width increases with increase in flowrate and decrease in
concentration, whereas depth of channel increases with increase in con-
centration due to digging by large number of solid particles present in the
slurry.
ii. The visualization of channel cross-sections has been done by plotting for
both experimental and field flow conditions.
iii. Flow velocity increases with increase in both flowrate and concentration.
Theoretically obtained velocity range for similar flow condition is
matched with the experimental results of previous researchers.
iv. For any given flowrate, the increase in velocity with increased sand con-
centration is significantly higher than that for increase in flow rate. As an
example, an increase in sand concentration from 6% to 14 % has led to an
increase in velocity by 104% when the flow rate is kept constant at 20
lpm, whereas the increase in velocity is only 6.8% when flowrate is in-
creased from 15 to 30 lpm, concentration being kept constant at 6%.
v. Decrease in flowrate and/or increase in concentration causes difficulty in
sand transport process by increasing the unit flow pressure loss. Theoreti-
cally evaluated pressure drop has been found to be nearly 8 times higher
as compared to the experimentally measured pressure drop. This is proba-
bly due to presence of entrained air in the experimental flow channels,
which tends to offer an additional assistance to flow channels due to its
buoyancy effect.
9. Using one-dimensional diffusion equation the distribution of solid concentration
of slurry along depth of the channel has been found to increases from top to the
bottom of the channel and the rate of change of concentration is also higher to-
wards the bottom.
10. Recorded inlet pressure data shows that the pressure loss per unit length continu-
ously increases along with the bed advancement. To maintain the pressure loss
within certain bound, a threshold based controller, called pre-jamming indicator,
has been developed. This pre-jamming indication parameter depicts the arrival of
pre-jamming phase of filling and ensures the completion of at least 74% of maxi-
mum area of filling. For further continuation of filling from the same borehole
one has to either increase the slurry flowrate keeping the same sand concentration,
or decrease the sand concentration keeping the same slurry flowrate. In this man-
ner the filling process can be prolonged sufficiently enough to fill-up a very large
area from a single inlet point. Finally, the above exercise can possibly lead to an
overall saving in cost of filling by reduction of the number of boreholes to be
drilled for filling-up of similar large area.
7.2 Future Scope of work
Due to high risk involved in residing over old, abandoned mine working lying at shallow
depth, the scientific development of the technology of blind backfilling process need im-
mediate attention. In the present work, some attempts have been made to develop theoret-
ical approaches on this aspect and its validation, through experiments as well as field da-
ta, has been done. Blind backfilling work being a new field, not many significant scien-
tific works exists all over the world, till date. Therefore, there is a vast scope of future
developments in this research field.
Motion of particles in transport zone as well as in discharge zone is similar to particle tra-
jectory within a fluid medium, excepting the presence of a top cover in the upper portion
of fluid medium. As sand particles make hopping movement during filling process it
touches the top of the void space and then either dragged along or dropped downward
and in this way it moves forward. Motion of a single particle within the transport zone as
well as within the discharge zone needs to be evaluated through solution of force balance
equations forming two dimensional equations of motion by analyzing all possible forces
acting on a particle. It is possible to transform two-dimensional equation to a system of
one-dimensional equation of motion, which can be solved by using R-K method to obtain
124
the particle trajectory. Interactions among particles during their motion in the channel
should be studied so as to visualize the dynamics of filling process.
Using Advection Diffusion Equation (ADE) it is possible to find out change concentra-
tion along the flow direction in the discharge zone. From the nature of variation in con-
centration it is possible to find-out sediment deposition rate at this zone. Using sediment
deposition rate, it is possible to find-out theoretical longitudinal profile with respect to
distance and time as well as bed advancement with respect to time. It can be compared
with experimental and previously computed theoretical results.
Pattern evaluation can also be attempted from the nature of autocorrelation and partial
autocorrelation initial pressure data and consequently a more efficient pre-jamming indi-
cation parameter may be derived.
Experimentation for faster filling rate can be tried with blind backfilling process using
dense phase slurry pumping and theoretical modeling of the same process may also be
tried. In all the aforementioned process the backfilling can only be feasible in fully flood-
ed abandoned mines. Another comparatively new technique, namely, paste backfilling
can also be considered for blind backfilling in dry abandoned mines through boreholes
from surface. Theoretical modeling of this process of fluid flow of non-Newtonian fluid
is also another new area which may be explored in future.
Curriculum Vitae
Susmita Panda completed her Bachelor of Science in Mathematics, Vidyasagar Universi-

ty, 2008. She completed her Masters of Science from Indian Institute of Technology,
Kharagpur in the year 2008. She qualified NET (June, 2009) and GATE (2010). In 2010,
she went on to pursue Doctor of Philosophy from IIT Kharagpur.
She has published few papers in international journals and conference proceedings.
Publications
Conference Publications
1. Panda S, Pal S K, Murthy, P V S N, (2014), Slurry feeding to Avoid Ground

Subsidence Civil Engineering Systems and Sustainable Innovations (pp 343-
347), New Delhi, Excellent publishing House. ISBN No. 978-93-83083-78-7.
2. Panda S, Pal S K and Murthy P V S N, (2014), Concentration Profile during Fill-

ing Process of Blind Backfilling, Mathematical Sciences for Advancement of
Science and technology 3,137-142. ISBN. 978-81-925832-2-8.
3. Panda S, Pal S K, Murthy P V S N, (2015), "Pressure Drop Evaluation During

Filling Process of Blind Backfilling Technique", In the proceedings of 8th Inter-
national Congress on Industrial and Applied Mathematics, pp - 353, Beijing, Chi-
na (Abstract).
4. Pal S K, Ghosh S, Pal A, Panda S ,(2015), A brief outline of the Indian coal in-
dustry with special Focus on Some Recent Research areas on Subsurface Coal
Mines Proceedings of international Symposium on Earth science and Technol-
ogy, 2-7. ISBN. 978-4-9902356-4-2.
5. Pal S K, Panda S, Tripathy A K (2015), Studies on Gravity Backfilling Method

for Blind Backfilling of Abandoned Coal Mines Proceedings of the 24th interna-
tional mining congress of turkey, 1453-1465. ISBN. 978-605-01-0705-0
126
Journal Publication
1. Panda S, Pal S K, and Murthy P V S N (2016). Evaluation of pre-jamming indica-

tion parameter during blind backfilling technique. Journal of Sustainable Min-
ing, 15(3), 100-107.
2. Pal, S K, Mukhopadhyay S K, Panda, S, and Tripathy, K A, (2016), Model

Studies on the Efficiency of Gravity Blind Backfilling Method and Evaluation of
a Pre-Jamming Indication Parameter International Journal of Earth Sciences
and Engineering, 1474-1479.
References
Al-Kayiem, H.H. and Jamil, I.E.A., 2013. Theoretical Modeling of Pseudo Hydrostatic
Force in Solid-Liquid Pipe Flow with Two Layers. Pertanika Journal of Science & Tech-
nology, 21(2).
Andreyev, N.E. and William, K., Andreyev Engineering Independent Drilling, LLC,
2010. Sand slurry injection systems and methods. U.S. Patent 7,832,962.
Araoka, K. and Maeno, N., 1981. Dynamical behaviors of snow particles in the saltation
layer. Memoirs of National Institute of Polar Research. Special issue, 19, pp.253-263.
Aydan, . and Ito, T., 2015. The effect of the depth and ground water on the formation of
sinkholes or ground subsidence associated with abandoned room and pillar lignite mines
under static and dynamic conditions. Proceedings of the International Association of Hy-
drological Sciences, 372, p.281-284.
Baeckblom, G., Christiansson, R., Hedin, A., Norman, F. and Lagerstedt, L., 2003. Re-
port on the access to the deposition areas of the repository. Shaft or Ramp? (No. SKB-R--
03-11). Swedish Nuclear Fuel and Waste Management Co..
Bagnold, R. A., 1951. The movement of a cohesionless granular bed by fluid flow over it.
British Journal of Applied Physics, 2(2), 29.
Bagnold, R. A., 1954. Experiments on a gravity-free dispersion of large solid spheres in a

Newtonian fluid under shear. In Proceedings of the Royal Society of London A: Mathe-
matical, Physical and Engineering Sciences (Vol. 225, No. 1160, pp. 49-63). The Royal
Society.
Bird, R. B., Stewart, W. E., & Lightfoot, E. N., 1960. Transport Phenomena John Wiley
& Sons. New York, 413.
Boger, D.V., 2012. Personal perspective on paste and thickened tailings: a decade on.
Mining technology, 121(1), pp.29-36.
128
Box, G.E., Jenkins, G.M., Reinsel, G.C. and Ljung, G.M., 2015. Time series analysis:
forecasting and control. John Wiley & Sons.
Bridge, J. S. 1979. A FORTRAN IV program to simulate alluvial stratigraphy. Comput-

ers & Geosciences, 5(3), 335-348.
Cao, S. and Knight, D.W., 1998. Design for hydraulic geometry of alluvial channels.
Journal of Hydraulic Engineering, 124(5), pp.484-492.
Cao, S., 1995. Regime theory and geometric model for stable alluvial channels (Doctoral
dissertation, University of Birmingham).
Chen, X., Ma, J. and Dey, S., 2009. Sediment transport on arbitrary slopes: Simplified
model. Journal of Hydraulic Engineering, 136(5), pp.311-317.
Chepil, W.S., 1958. The use of evenly spaced hemispheres to evaluate aerodynamic forc-
es on a soil surface. Eos, Transactions American Geophysical Union, 39(3), pp.397-404.
Chugh, Y.P., De, I. and Powell, E., 2005. A model study for blind pneumatic backfilling
of fly ash in abandoned underground mines. International Journal of Surface Mining,
Reclamation and Environment, 19(4), pp.295-311.
Colaizzi, G.J., Whaite, R.H. and Donner, D.L., 1981. Pumped-slurry backfilling of aban-
doned coal mine workings for subsidence control at Rock Springs, Wyo (Vol. 8846). US
Dept. of the Interior, Bureau of Mines.
Dang, W.G., Liu, Z.X., He, X.Q. and Liu, Q.L., 2013. Mixture ratio of phosphogypsum
in backfilling. Mining Technology, 122(1), pp.1-7.
Davidson, S.K. and Hey, R.D., 2011. Regime equations for natural meandering cobble-
and gravel-bed rivers. Journal of Hydraulic Engineering, 137(9), pp.894-910.
De Angelis, E., & Mancini, A., 1997. A model for the evolution of sedimentation beds in
the dynamic of a pipelined non-Newtonian fluid.Mathematical and Computer Model-
ling, 25(7), 65-78.
Deutsch, C. V., & Tran, T. T. 2002. FLUVSIM: a program for object-based stochastic
modeling of fluvial depositional systems. Computers & Geosciences, 28(4), 525-535.
Doron, P. and Barnea, D., 1993. A three-layer model for solid-liquid flow in horizontal
pipes. International Journal of Multiphase Flow, 19(6), pp.1029-1043.
Doron, P., & Barnea, D., 1992. Effect of the no-slip assumption on the prediction of sol-
id-liquid flow characteristics. International journal of multiphase flow, 18(4), 617-622.
Doron, P., & Barnea, D., 1996. Flow pattern maps for solid-liquid flow in pipes. Interna-
tional journal of multiphase flow, 22(2), 273-283.
Doron, P., Granica, D. and Barnea, D., 1987. Slurry flow in horizontal pipes-
experimental and modeling. International Journal of Multiphase Flow, 13(4), pp.535-547.
Eaton, B.C. and Millar, R.G., 2004. Optimal alluvial channel width under a bank stability
constraint. Geomorphology, 62(1), pp.35-45.
Einstein, A., 1956. Investigations on the Theory of the Brownian Movement. Courier
Corporation.
Fletcher, B., 1976. The incipient motion of granular materials. Journal of Physics D: Ap-
plied Physics, 9(17), p.2471.
Galve, J.P., Gutirrez, F., Lucha, P., Guerrero, J., Bonachea, J., Remondo, J. and
Cendrero, A., 2009. Probabilistic sinkhole modelling for hazard assessment. Earth Sur-
face Processes and Landforms, 34(3), pp.437-452.
Galve, J.P., Gutirrez, F., Remondo, J., Bonachea, J., Lucha, P. and Cendrero, A., 2009.
Evaluating and comparing methods of sinkhole susceptibility mapping in the Ebro Valley
evaporite karst (NE Spain). Geomorphology, 111(3), pp.160-172.
Gamble, J.C. and Gray, R.E., 1988. Mine subsidence control. In General Geology (pp.
592-599). Springer US.
130
Gillies, R. G., Shook, C. A., & Wilson, K. C., 1991. An improved two layer model for
horizontal slurry pipeline flow. The Canadian Journal of Chemical Engineering, 69(1),
173-178.
Govan, A. H., Hewitt, G. F., & Ngan, C. F., 1989. Particle motion in a turbulent pipe
flow. International journal of multiphase flow, 15(3), 471-481.
Gutfeld, V., Jacob, R., and Lien, S.H. 2000. Method and apparatus for filling liquid crys-
tal display (LCD) panels. International Business Machines Corporation, US Patent
6,055,035.
Guth, E., Simha, R. and Gold, O., Kolloid-Z., 74, 266. 1936. CrossRef| CAS| Web of
Science Times Cited, 220.
Huang, H.Q., Nanson, G.C. and Fagan, S.D., 2002. Hydraulic geometry of straight allu-
vial channels and the principle of least action. Journal of Hydraulic Research, 40(2),
pp.153-160.
Hutton, E.W.H. and Syvitski, J.P.M., 2008. SedFlux2. 0: New advances in the seafloor
evolution and stratigraphic modular modeling system. Comput. Geosci, 34, pp.1319-
1337.
Ikeda, S. and Izumi, N., 1991. Stable channel cross sections of straight sand rivers. Water
resources research, 27(9), pp.2429-2438.
Ikeda, S. and Izumi, N., 1991. Stable channel cross sections of straight sand rivers. Water
resources research, 27(9), pp.2429-2438.
Jaiswal, D.K., Kumar, A. and Yadav, R.R., 2011. Analytical solution to the one-
dimensional advection-diffusion equation with temporally dependent coefficients. Journal
of Water Resource and Protection, 3(01), p.76-84.
Jensen, J.L. and Srensen, M., 1986. Estimation of some aeolian saltation transport pa-
rameters: a reanalysis of Williams' data. Sedimentology, 33(4), pp.547-558.
Jha, B.K., Adlakha, N. and Mehta, M.N., 2011. Solution of advection diffusion equation
arising in cytosolic calcium concentration distribution. Int. J. of Appl. Math and Mech,
7(6), pp.72-79.
Jha, B.K., Adlakha, N. and Mehta, M.N., 2012. Analytic solution of two dimensional ad-
vection diffusion equation arising in cytosolic calcium concentration distribution. In In-
ternational Mathematical Forum (Vol. 7, No. 3, pp. 135-144).
Julien, P.Y. and Wargadalam, J., 1995. Alluvial channel geometry: theory and applica-
tions. Journal of Hydraulic Engineering, 121(4), pp.312-325.
Karfakis, M.G., 2016. Residual subsidence over abandoned coal mines. Surface and Un-
derground Project Case Histories: Comprehensive Rock Engineering: Principles, Practice
and Projects, p.451.
Kaushal, D.R. and Tomita, Y., 2002. Solids concentration profiles and pressure drop in
pipeline flow of multisized particulate slurries. International journal of multiphase
flow, 28(10), pp.1697-1717.
Kumar, A., Jaiswal, D.K. and Kumar, N., 2009. Analytical solutions of one-dimensional
advection-diffusion equation with variable coefficients in a finite domain. Journal of
Earth System Science, 118(5), pp.539-549.
Kumar, A., Jaiswal, D.K. and Kumar, N., 2010. Analytical solutions to one-dimensional
advectiondiffusion equation with variable coefficients in semi-infinite media. Journal of
Hydrology, 380(3), pp.330-337.
Leopold, L.B. and Maddock, T., 1953. The hydraulic geometry of stream channels and
some physiographic implications (Vol. 252). US Government Printing Office.
Li, L. and Yang, P., 2015. A numerical evaluation of continuous backfilling in cemented
paste backfilled stope through an application of wick drains. International Journal of
Mining Science and Technology, 25(6), pp.897-904.
132
Li, L., 2013. A new concept of backfill design-Application of wick drains in backfilled
stopes. International Journal of Mining Science and Technology, 23(5), pp.763-770.
Lin, C.X. and Ebadian, M.A., 2008. A numerical study of developing slurry flow in the
entrance region of a horizontal pipe. Computers & Fluids, 37(8), pp.965-974.
Mackey, S. D., & Bridge, J. S. 1992. A revised FORTRAN program to simulate alluvial
stratigraphy. Computers & Geosciences, 18(2), 119-181.
Matousek, V., 2002. Pressure drops and flow patterns in sand-mixture pipes. Experi-
mental thermal and fluid science, 26(6), pp.693-702.
Mhlongo, S.E. and Amponsah-Dacosta, F., 2016. A review of problems and solutions of
abandoned mines in South Africa. International Journal of Mining, Reclamation and En-
vironment, 30(4), pp.279-294.
Morsi, S. and Alexander, A.J., 1972. An investigation of particle trajectories in two-phase

flow systems. Journal of Fluid mechanics, 55(02), pp.193-208.
Nalpanis, P., Hunt, J.C.R. and Barrett, C.F., 1993. Saltating particles over flat beds. Jour-
nal of Fluid Mechanics, 251, pp.661-685.
Newitt, D.M., Richardson, J.F., Abbott, M. and Turtle, R.B., 1955. Hydraulic conveying
of solids in horizontal pipes. Transactions of the institution of Chemical Engineers, 33,
pp.93-110.
OGrady, D.B. and Syvitski, J.P., 2001. Predicting profile geometry of continental slopes
with a multiprocess sedimentation model. In Geologic Modeling and Simulation (pp. 99-
117). Springer US.
Oroskar, A.R. and Turian, R.M., 1980. The critical velocity in pipeline flow of slurries.
AIChE Journal, 26(4), pp.550-558.
Pal S K and Mukhopadhyay S 2007. Model studies on the efficiency of gravity blind
backfilling method and evaluation of pre-jamming indication parameter. A report from
Mining Engineering, IIT Kharagpur, of Coal S& T.
Pal S K., 2012. Re-application of the S&T Project Titled Model studies on the efficiency
of gravity blind backfilling method and evaluation of a efficiency of gravity blind back-
fill-ing method and evaluation of pre-jamming indication parameter in the Field. Report
from Mining Engineering, IIT Kharagpur, of Coal S&T.
Pal, S. K., 2001. System Optimization of Air-Assisted Backfilling of Inaccessible Water-

Logged Mine Working A Model Study Approach. Transactions Journal of MGMI, Vol.
97, No. 1 & 2, pp. 105 113.
Pal, S.K., 2003, November. Experimental studies on stabilization techniques for ground
over abandoned subsurface excavations. In International Symposium on Fusion Technol-
ogy of Geosystem Engineering, Rock Engineering and Geophysical Exploration, Seoul,
Korea (pp. 18-19).
Pal, S.K., Ray, S.C. and Barve, S.D., 2001. System Optimization of Air-Assisted Back-
filling of Inaccessible Water-Logged Mine WorkingA Model Study Approach. Transac-
tions Journal of MGMI, 97(1), pp.105-113.
Parker, G., 1978. Self-formed straight rivers with equilibrium banks and mobile bed. Part
1. The sand-silt river. Journal of Fluid Mechanics, 89(01), pp.109-125.
Peckham, S.D., 2008. On the form and stability of seafloor stratigraphy and shelf pro-
files: A mathematical model and solution method. Computers & Geosciences, 34(10),
pp.1358-1369.
Peker, S.M. and Helvaci, S.S., 2011. Solid-liquid two phase flow. Elsevier.
Peng, X., Tomita, Y., Tashiro, H. and Funatsu, K., 1996. Motion of a single particle in a
horizontal pipe. Advanced Powder Technology, 7(2), pp.101-109.
134
Perron, J.T. and Royden, L., 2013. An integral approach to bedrock river profile analysis.
Earth Surface Processes and Landforms, 38(6), pp.570-576.
Phillips, M., 1980. A force balance model for particle entrainment into a fluid stream.
Journal of Physics D: Applied Physics, 13(2), p.221-233.
Ramadan, A., Skalle, P. and Johansen, S.T., 2003. A mechanistic model to determine the
critical flow velocity required to initiate the movement of spherical bed particles in in-
clined channels. Chemical Engineering Science, 58(10), pp.2153-2163.
Ramadan, A., Skalle, P. and Saasen, A., 2005. Application of a three-layer modeling ap-
proach for solids transport in horizontal and inclined channels. Chemical engineering sci-
ence, 60(10), pp.2557-2570.
Richardson, J.F. and Zaki, W.N., 1954. The sedimentation of a suspension of uniform
spheres under conditions of viscous flow. Chemical Engineering Science, 3(2), pp.65-73.
RoyChowdhury, A., Sarkar, D., Deng, Y. and Datta, R., 2016. Assessment of Soil and
Water Contamination at the Tab-Simco Coal Mine: A Case Study. Mine Water and the
Environment, pp.1-7.
Royden, L. and Taylor Perron, J., 2013. Solutions of the stream power equation and ap-
plication to the evolution of river longitudinal profiles. Journal of Geophysical Research:
Earth Surface, 118(2), pp.497-518.
Salman, A.D., Gorham, D.A., Szabo, M. and Hounslow, M.J., 2005. Spherical particle
movement in dilute pneumatic conveying. Powder technology, 153(1), pp.43-50.
Sands, P.F., Boldt, C.M. and Ruff, T.M., 1990. Blind pneumatic stowing in voids in
abandoned mines (Vol. 9268). US Department of the Interior, Bureau of Mines.
Saro, L., Hyun-Joo, O. and Ki-Dong, K., 2010. Statistical spatial modeling of ground
subsidence hazard near an abandoned underground coal mine. Disaster Advances, 3(1),
pp.11-23.
Saxena, N.C., Parti, S.K., Kumar, B. and Singh, B., 1984. Blind Backfilling of Unap-
proachable Workings Underneath Surface Properties. Transactions Journal of MGMI,
81(2), pp.47-59.
Schumm, S.A., 1961. The shape of alluvial channels in relation to sediment type. US
Government Printing Office.
Smith, T.R., 1974. A derivation of the hydraulic geometry of steady-state channels from
conservation principles and sediment transport laws. The Journal of Geology, 82(1),
pp.98-104.
Sutter Gold Mining Company. 1998. Environmental Analysis of Underground Backfill-

ing at the Lincoln Project. Sutter Creek, California.
Syvitski, J.P.M., 1989. Modelling the fill of sedimentary basins. Statistical applications in
the earth sciences: Canada Department of Energy, Mines & Resources Paper, pp.89-9.
Syvitski, J.P, 1989. Modelling the fill of sedimentary basins, paper 89-9. In: Agterberg,
F., Bonham-Carter, G. (Eds.), Statistical Applications in the Earth Sciences. Geological
Survey of Canada, pp. 505515.
Syvitski, J.P., Pratson, L. and O'Grady, D., 1999. Stratigraphic predictions of continental
margins for the US Navy.
Syvitski, J.P., Skene, K.I., Nicholson, M.K. and Morehead, M.D., 1998. PLUME1. 1:
deposition of sediment from a fluvial plume. Computers & Geosciences, 24(2), pp.159-
171.
Syvitski, J.P., Skene, K.I., Nicholson, M.K. and Morehead, M.D., 1998. PLUME1. 1:
deposition of sediment from a fluvial plume. Computers & Geosciences, 24(2), pp.159-
171.
Tanaka, T. and Tsuji, Y., 1991. Numerical simulation of gas-solid two-phase flow in a
vertical pipe: on the effect of inter-particle collision. ASME/FED GasSolid Flows, 121,
pp.123-128.
136
Tanaka, T., Takagi, Y., Tsuji, Y. and Morikawa, Y., 1989. Measurements of gas-solid
two-phase flow in a vertical pipe. TRANS. JAPAN SOC. MECH. ENG.(SER. B).,
55(516), pp.2302-2309.
Televantos, Y., Shook, C., Streat, M. and Carleton, A., 1979. Flow of slurries of coarse
particles at high solids concentrations. The Canadian Journal of Chemical Engineering,
57(3), pp.255-262.
Thill, R.E., Huck, P.J. and Stegman, B.G., 1982. Monitoring of blind backfilling in aban-
doned mines. Prepr., Soc. Min. Eng. AIME;(United States), 82.
Thomas, A.D., 1979. Predicting the deposit velocity for horizontal turbulent pipe flow of
slurries. International Journal of Multiphase Flow, 5(2), pp.113-129.
Thomas, E.G., 1985. Preparation of mine fill material from coal washplant reject. Mining
Science and Technology, 3(1), pp.1-9.
Turian, R.M. and Yuan, T.F., 1977. Flow of slurries in pipelines. AIChE Journal, 23(3),
pp.232-243.
Turian, R.M., Hsu, F.L. and Ma, T.W., 1987. Estimation of the critical velocity in pipe-
line flow of slurries. Powder Technology, 51(1), pp.35-47.
Turian, R.M., Yuan, T.F. and Mauri, G., 1971. Pressure drop correlation for pipeline flow
of solidliquid suspensions. AIChE Journal, 17(4), pp.809-817.
Wadke, P.M., Pitt, M.J., Kharaz, A., Hounslow, M.J. and Salman, A.D., 2005. Particle
trajectory in a pipe bend. Advanced Powder Technology, 16(6), pp.659-675.
Walker, J.S., 1993. State-of-the-art techniques for backfilling abandoned mine voids
(Vol. 9359). US Department of the Interior, Bureau of Mines.
Wasp, E.J., Kenny, J.P. and Gandhi, R.L., 1977. Solid-liquid flow: slurry pipeline trans-
portation.[Pumps, valves, mechanical equipment, economics]. Ser. Bulk Mater.
Handl.;(United States), 1(4).
Whaite, R.H., Allen, A.S. and Carlson, E.J., 1975. Pumped-slurry backfilling of inacces-
sible mine workings for subsidence control. With an appendix on hydraulic model studies
for backfilling mine (No. BM-IC-8667). Bureau of Mines, Washington, DC (USA); Bu-
reau of Reclamation, Denver, Colo.(USA).
White, B.R. and Schulz, J.C., 1977. Magnus effect in saltation. Journal of Fluid Mechan-
ics, 81(03), pp.497-512.
White, B.R., Iversen, J.D., Greeley, R. and Pollack, J.B., 1975. Particle motion in atmos-
pheric boundary layers of Mars and Earth.
Wiberg, P.L. and Smith, J.D., 1985. A theoretical model for saltating grains in water.
Journal of Geophysical Research: Oceans, 90(C4), pp.7341-7354.
Willettsb,. B. & Rice,M .A ., 1985. Intersaltation collisions. In Proc. Intl Workshop on

the Physics of Blown Sand (ed. O. E. Barndorff-Nielsen), vol. I , pp. 83-100.
Williams, G., 1964. Some aspects of the eolian saltation load. Sedimentology, 3(4),
pp.257-287.
Wilson, K.C., 1989. Mobile-bed friction at high shear stress. Journal of Hydraulic Engi-
neering, 115(6), pp.825-830.
Wissmann, K.J. and Peterson, G.M., Geopier Foundation Company, Inc., 2013. Shielded
tamper and method of use for making aggregate columns. U.S. Patent 8,562,258.
Wolkersdorfer, C., 2008. Water management at abandoned flooded underground mines:

fundamentals, tracer tests, modelling, water treatment. Springer Science & Business Me-
dia.
Worster, R.C. and Denny, D.F., 1955. Hydraulic transport of solid material in pipes. Pro-
ceedings of the Institution of Mechanical Engineers, 169(1), pp.563-586.
Xiaofeng, S., Kelin, W., Tie, Y., Shuai, S. and Shizhu, L., 2013. Cuttings Transport with
Gas in the Highly-Deviated Horizontal Wells. Open Energy & Fuels Journal, 6.
138
Yamamoto, F., 1986. A study of motion of a sphere in air flow through horizontal pipe.
Bulletin of JSME, 29(253), pp.2055-2061.

Slurry Transport in Blind Backfilling Technique 4 3 17 - 8 17 PM

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Slurry Transport in Blind Backfilling Technique 4 3 17 - 8 17 PM

Hochgeladen von

Copyright:

Verfügbare Formate

Slurry Transport in Blind Backfilling Technique

for the Award of the Degree of

Under the supervision of

Indian Institute of Technology Kharagpur

2017 Susmita Panda. All rights reserved.

Blind hydraulic backfilling is a commonly used technique to control subsidence of the

_____________________ _____________________ _____________________

Prof. K U M Rao (MI) Prof. A Bhattacharjee (MI) Prof. G P Rajasekhar (MA)

(Member of the DSC) (Member of the DSC) (Member of the DSC)

_____________________ _____________________ _____________________

Prof. S K Pal (MI) Prof. P V S N Murthy (MA)

(Supervisor) (Supervisor) (External Examiner)

Prof S K Pal, Professor Prof P V S N Murthy, Professor

Date Susmita Panda

IIT Kharagpur Susmita Panda

APPROVAL OF THE VIVA-VOCE BOARD ................................................................. iv

DECLARATION ........................................................................................................... viii

Contents ........................................................................................................................ xii

List of Tables ............................................................................................................... xvi

List of Figures ............................................................................................................ xviii

List of symbols, abbreviations and formula ............................................................... xxiv

1.1 Preamble ............................................................................................................... 2

1.2 Background ........................................................................................................... 2

1.3 Motivation ............................................................................................................. 3

1.4 Objectives ............................................................................................................. 4

1.5 Scope of this thesis................................................................................................ 5

1.6 Novelties of this thesis .......................................................................................... 5

1.7 Organization of this thesis .................................................................................... 5

1.8 Methodology ......................................................................................................... 6

2 Literature Survey ......................................................................................................... 8

2.1 Literature review on backfilling............................................................................ 8

2.2 Literature review on sediment transport related work ........................................ 16

3 Experimental Investigation ........................................................................................ 32

3.1 Introduction ......................................................................................................... 32

3.2 Experimental setup.............................................................................................. 33

3.3 Experimental procedure and measurement ......................................................... 35

3.4 Empirical relationships ....................................................................................... 40

3.5 Discussions ......................................................................................................... 54

4 Progress of deposited sand bed .................................................................................. 56

4.1 Introduction ......................................................................................................... 56

4.2 Bed slope angle ................................................................................................... 56

4.3 Bed advancement ................................................................................................ 63

4.4 Discussions ......................................................................................................... 80

5 Channel Cross-Section and flow Pressure Loss......................................................... 82

5.1 Model formulation .............................................................................................. 82

5.2 Computational procedure for supporting variables............................................. 87

5.3 Solution method for the model ........................................................................... 90

5.4 Channel shape ..................................................................................................... 94

5.5 Concentration profile in the moving bed layer ................................................... 98

5.7 Pressure drop ..................................................................................................... 101

5.8 Discussions ....................................................................................................... 105

6 Pre-jamming indication parameter ........................................................................... 108

6.1 Introduction ....................................................................................................... 108

6.3 Results ............................................................................................................... 110

6.5 Discussions ....................................................................................................... 118

7 Conclusions and Future scope of the work .............................................................. 120

7.1 Conclusions ....................................................................................................... 120

7.2 Future Scope of work ........................................................................................ 124

Curriculum Vitae ........................................................................................................ 126

References ................................................................................................................... 128

Table 3.3.2.1 Repeatability experiments for maximum sand throughput

Table 3.4.2.2 Maximum flow pressure loss in channels at jamming

Table 3.4.3.2 Bed advancement rates for different flow conditions

___________ _ ___________

___________ _ ___________