You are on page 1of 5

SME Annual Meeting

Feb. 19 - 22, 2012, Seattle, WA

Preprint 12-009
EVALUATION OF BLAST FRAGMENTATION IN QUARRIES NEARBY ISTANBUL
T. Hudaverdi, Istanbul Tech. Univ., Istanbul, Turkey
C. Kuzu, Istanbul Tech. Univ., Istanbul, Turkey
A. Fisne, Istanbul Tech. Univ., Istanbul, Turkey

m that of E1 is coarser and equal to 0.20 m. By considering only the


mean fragment size, Blast A1 may be evaluated as a better blasting
operation than Blast E1. However, because the in-situ block size of A1
is twice that of E1, the fragmentation index (FI) values of A1 and E1
are 5.81 and 10.30, respectively. Therefore, E1 appears to be a more
efficient blast than A1.

INTRODUCTION
Control of the particle size distribution of a muckpile after blasting
is always an important subject for mining industry. Blasting has a
significant impact on downstream processes such as loading, crushing
and grinding. Improvement of blasting results provides increase in
loader and excavator productivity due to increased diggability capacity,
and increased bucket and truck fill factors. Suitable and uniform
particle size distribution results increase in crusher and mill throughput
and decrease in energy consumption in size reduction process.

The use of the fragmentation index as a tool to evaluate blasting


efficiency was first introduced by Aler and coworkers (4, 5). They
described the blasting index as a useful tool to identify the effect of
blast design parameters on the fragmentation. The researchers
performed their investigations in zinc and uranium mines in Spain. Aler
et al. (4) used the fragmentation index to compare the blasts
performed in different rock conditions.

Digital image analysis technique has been used for many years in
mining industry in order to determine blast fragmentation. The images
of muckpile surface can be used to determine the mean fragment size
(x50). The source of digital images can also be haul truck, leach pile,
waste dump, stockpile, and conveyor belt. The screening that is
conventional method to measure blast fragmentation is time
consuming and expensive. Digital image analysis is a cost effective
and faster method (1).

Jhanwar et al. (6) used fragmentation index to compare the


efficiency of conventional blasting and air-deck blasting in the DongriBuzurg openpit manganese mine situated in Central India. The rock
type in Dongri-Buzurg mine were micaceous and muscovite schist. The
hole diameter was 100 mm and bench height was 6-11 m. The burden
was between 2 to 2.5 m. The researcher used 0.9 m air decks in the
blast holes. In Figure 1, the efficiency of the performed blasts is
compared using fragmentation index. The fragmentation indices of airdeck blasts are obviously higher.

Mean fragment size (x50) is used as a measure of average


fragmentation. It indicates the mesh size which the half of the muckpile
passes. Mean fragment size is the most used parameters to
characterize the blast fragmentation. The main aim of most
fragmentation prediction approaches is to predict mean fragment size
(e.g. KuzRam, SVeDeFo (2, 3)). Mean fragment size is a single
parameter to characterize blast fragmentation and cannot be used to
compare the blasts performed under different conditions.
Fragmentation index which is the main subject of this paper is
calculated using in-situ block size of rock and mean fragment size.
Fragmentation index makes it possible to compare the fragmentation
efficiency of the blasts performed under different conditions. In this
paper, the blasts performed by different explosives were compared
using fragmentation index. By that way, the efficiency of the different
explosives was examined. Fragmentation index may also be used to
compare the blasts performed in different rock formations or to
evaluate the blasts performed by different initiation systems.
FRAGMENTATION INDEX CONCEPT
The discontinuity intersections into rock mass create in-situ
blocks. The in-situ block size plays a major role in creating the blasted
block size of a muckpile. Fragmentation index is a size reduction ratio
that can be used to compare the blast results. High fragmentation
index indicates more efficient blasting operation.

Figure 1. FI values calculated in respect of conventional and air-deck


blasts (6).

The fragmentation index is defined as:

FI =

XB
x50

Jhanwar et al. (6) also examined the relation between


fragmentation index, spacing-burden ratio (S/B), powder factor (Pf) and
Bieniawskis Rock Mass Rating (7). The Equation (2) given below is
the numerical expression of this relationship for the conventional
blasts.

[1]

FI = Fragmentation index
XB = Average in-situ block size (m)
x50 = Mean fragment size of the muckpile resulting from blasting (m)

FI= 0.03 (RMR Pf S/B) + 0.73

(r = 0.73)

[2]

Gama (8) introduced a new concept named Fragmentability.


Fragmentability is an index value related to rock properties like
blastability. According to Gama, Fragmentability is a minimum
explosive energy threshold needed to separate the in-situ blocks

As an example, Blast A1 and E1 in Table 2 (See Appendix) can


be compared to show the importance of use of fragmentation index.
While the mean fragment size of the muckpile (x50) of Blast A1 is 0.16

Copyright 2012 by SME

SME Annual Meeting


Feb. 19 - 22, 2012, Seattle, WA
through weakness planes. As seen in Equation (3), the Fragmentability
concept is slightly a different version of the fragmentation index
concept.

K =W

Sa
Sb

emulsion can be increased by adding small nitrogen bubbles or microspheres made of glass. Emulsion explosives may be produced in bulk
or cartridge form. Table 1 shows general characteristics of the
explosives used as main charge in the quarry.

[3]

Where; K is the Fragmentability, W is the explosive energy


consumed per ton of rock (kwh/ton), Sa and Sb are the representative
sizes of blocks after and before blasts, respectively.
Onederra (9) used fragmentation index to create a fragmentation
modeling framework for underground ring blasting application. In this
research, the fragmentation index (FI) was referred to as the
fragmentation factor (F) and calculated using the mean in-situ block
size (Xinsitu) and mean fragment size (x50).

F=

X insitu
1
x 50

[4]
Figure 2. A view of the northern side of the quarry.

Xinsitu is a measure of the mean size of in-situ blocks formed by


the rock mass discontinuity network in the local domain. If the F value
approaches zero, the value of x50 approaches the size of the in-situ
blocks. This situation indicates a low concentration of explosive
energy. Onederra also created an equation based on KuzRam model
to predict the fragmentation factor (9).

Table 1. The properties of explosives used in the quarry.


Specifications
ANFO
Emulsion
0.77-0.82
1.25-1.28
Density(gr/cm3)
3.90
3.05
Energy (mj/kg)
974
1016
Gas Volume (liter/kg)
2565
2009
Detonation Temperature (0C)
4000
4500-4800
V.O.D. (m/sec)
None
Perfect
Water resistance

FIELD STUDIES
Blasting Site
The field investigation was performed in the Akdaglar quarry
located at the European side of the city. The rock in the quarry is
sandstone. The quarry produces aggregate for concrete and asphalt
plants. Production capacity of the quarry is 5000 ton/day.

Insufficient blasting causes coarser fragmentation than expected


in the quarry. Also big boulders occur in the muckpile after blasting.
The fragmentation index approach was used to compare the efficiency
of ANFO and Emulsion explosives in terms of fragmentation.

Electrical initiation system is applied at Akdaglar Quarry. The


diameter of the drill holes is 89 mm. The number of the holes varies
between 10-25. The burden applied is 2-2.5 and the spacing between
holes is 2-3 m. The height of the benches is 5-8 m. ANFO is used as
column charge. Primer explosive is emulsion type. The primer is
placed at the bottom of the hole. If the benches are too high an
additional primer explosive is placed under stemming. After blasting,
muck is transported by trucks and the big boulders are broken by
hydraulic excavators.

Determinations of in-situ block size and mean fragment size


We followed four steps to investigate the fragmentation efficiency:
(1)
(2)
(3)
(4)

The formation of Cendere Region where investigation was done is


called Trakya Formation that commonly contains sandstone. Trakya
formation represents carbonipherous aged, fragmented sediments.
Generally, it is observed at European Side of Istanbul. Trakya
Formation was deformed, folded and fragmented by joint systems in
different locations during Hersinian and Alpine Orogenesis period. The
thicknesses of sedimentary rocks which form Trakya Formation are
between 5 cm and 50 cm. In some parts especially for sandstones, the
thickness reaches 50 to 100 cm.

Determination of the in-situ block size,


Determination of mean fragment size of the muckpile,
Calculation of fragmentation index for each measured blast,
The evaluation of fragmentation efficiency by the
fragmentation index.

There are several studies that use image processing technique to


determine rock discontinuities. Maerz (10) defined two dimensional
numerical images as a coordinate plane and tried to determine the
discontinuities by using tones of grey. Reid and Harrison (11)
developed a semi automatically method to detect discontinuity traces.
Kemeny (12) scanned the bench face by laser scanner and used Split
FX software to model discontinuities. Kwangmin (13) used Split
Desktop software to determine the fracture toughness of the benches.
The researcher detected the in-situ block size of the benches by Split
desktop software and tried to determine the effect of the fracture
toughness on fragmentation efficiency.

Sixty percent of the sandstone is formed by silicium. Density of


the rock is 2.70 kg/cm3. Mohs hardness value is 5-6. Average porosity
value is 1.03%. The compressive strength is 822-829 kg/cm2 and
average tensile strength is 57.08 kg/cm2. Elasticity modulus is 16.9
GPa. A view of the northern side of the quarry is presented in Figure 2.

In this research, WipJoint software used to determine in-situ block


size and joint orientation. The software determines variation of joint
spacing for 18 different directions. WipJoint performs joint and
apparent block size detection automatically. It also allows manual
corrections. Figure 3 shows the bench of the Blast E1 presented in
Table 2 (see Appendix). The white box in the middle of the bench was
used as scaling object.

In the quarries located northern Istanbul, one of the major


problem that affects blasting efficiency is presence of water in the drill
holes. The quarries generally use Ammonium nitrate/fuel oil mixtures
(ANFO) as main charge. ANFO has no inherent water resistance,
because of their solubility in water. As water content increases in
blastholes, detonation velocity and explosive efficiency decreases.
Water resistant explosives should be used in wet holes to use
explosive energy properly. The quarries in Cendere region tends to
use emulsion explosives in wet holes. Emulsion explosives contains
substantial amount of oxidizer (generally ammonium nitrate) dissolved
in the water, fuel (eg. Fuel oil) and wax. The dissolved ammonium
nitrate droplets are stabilized using various emulsifiers. Sensitivity of

During taking photograph for image processing of bench face, it


must be avoided to capture inclined images caused by inaccurate
holding of camera. If the bench face is wide it may be difficult to take
the picture of whole face at once. Multiple images may be merged to
create image of whole face. The closeness and distance to the face
affect resolution. The photos should be taken under direct sunlight.
Shadows create problems during image processing stage. The

Copyright 2012 by SME

SME Annual Meeting


Feb. 19 - 22, 2012, Seattle, WA
efficiency of software is better if the fractures are in same direction. As
the number of fractures increases, the success of program decreases,
so the need of manual correction increases.

There are two main sampling strategies: systematic and random.


In both cases decision about sampling location are made before
entering the muckpile area. In both cases, the muckpile surface is
divided up into manageable sections (grids). Then researcher decides
which sections will be sampled (14).

Figure 3. The bench of the Blast E1.


If the joint aperture is wide, image processing software detects
the fractures easily. The presence of water on rock surface may make
the fractures more visible. Figure 4 shows the apparent blocks
determined by the software on the bench face of the Blast E1. The
software paints each block using different colors to make the blocks
more visible.

Figure 5. Block size distribution of the bench of the Blast E1.


Generally, increase of the number of the images means increase
of the accuracy of the fragment size distribution. In the quarry, several
images were captured for each muckpile. These images were
analyzed and the results were combined. In Figure 6, a sample image
taken from the muckpile of the Blast E1 is represented. The fragment
size distribution of the muckpile of Blast E1 is shown in Figure 7.

Figure 6. Sample image used for digital image analysis.


Figure 4. The blocks determined for the bench face of the Blast E1.

In Figure 7, the mean fragment size is a central production


measure equivalent to 50% passing value. Mean fragment size (x50) is
shown by D50 symbol. Also, it is possible to see different size
fractions. Symbol D10 represents the fine size fraction or the 10%
passing value and D90 represents the coarse size fraction or the
90% passing value. The symbol N indicates uniformity index which is
the slope of the fragment size distribution curve.

Figure 5 shows block size distribution of the bench of the Blast


E1. The software creates a histogram for measured block dimensions.
It is possible to see measured minimum and maximum spacing
between joints. The orientation rosette on the upper right side of the
images is used to see the main orientation of discontinuities.
Fragment size distribution of the muckpiles was determined by
WipFrag image processing software. During taking photograph of
muckpile, sampling location is very important. A muckpile is
heterogeneous with respect to fragment size. The particle size
distribution may be more uniform at the one side of the muckpile than
the other side or big blocks may occur in some part of the muckpile.
Without a clear sampling strategy personal bias enters to image.

EVALUATION OF TEST BLASTS


Table 2 (see Appendix) shows 11 blasts performed in the quarry.
Blast design parameters, in-situ block size, mean fragment size of the
muckpile and fragmentation indices are presented for each blast. In the
table, the first five blasts indicated with H symbol are performed by
ANFO. The others indicated by E symbol were performed using
emulsion explosive. Holes were drilled as staggered two rows and the
inclination of the holes is 900. The average specific charge of the

Copyright 2012 by SME

SME Annual Meeting


Feb. 19 - 22, 2012, Seattle, WA
blasts performed by ANFO is 0.67 kg/m3. Average specific charge is
0.92 kg/m3 for the blasts performed by emulsion explosive.

REFERENCES
1.

Singh, B, Pal, Roy P. (1993), Blasting in Ground Excavations and


Mines, A. A. Balkema, Rotterdam.

2.

Cunningham, CVB. (1983), The KuzRam model for prediction of


fragmentation from blasting, Proceedings of the 1. International
Symposium on Rock Fragmentation by Blasting, Holmberg, Lulea,
Sweden, pp. 439-453.

3.

Hjelmberg, H. (1983), Some ideas on how to improve


calculations of the fragment size distribution in bench blasting,
Proceedings of the 1st International Symposium on Rock
Fragmentation by Blasting, Lulea, Sweden, pp. 469-494.

4.

Aler, J, Du Mouza, J. and Arnould, M. (1996), Measurement of


the fragmentation efficiency of rock mass blasting and its mining
applications, International Journal of Rock Mechanics and Mining
Sciences and Geomechanics Abstracts Vol. 33, No. 2, pp. 125139.

5.

Aler, J, Du Mouza, J. (1996), Measuring fragmentation efficiency


of mine blasts, Measurement of Blast Fragmentation Fragblast 5,
Montreal, Canada, pp. 257-263.

Fragmentation index is obviously higher for the blasts performed


by emulsion explosive. In other words, higher size reduction was
achieved in comparison to blasts performed by ANFO. The arithmetic
mean of the calculated fragmentation indices belongs to ANFO blasts
is 6.63. The mean of the fragmentation indices belong to the blasts
performed by emulsion explosive is 8.80. The mean of the
fragmentation indices of the blasts performed by emulsion explosive is
33% higher than the ANFO blasts.

6.

Jhanwar, JC, Jethwa, JL, Reddy, AH. (2000), Influence of airdeck blasting on fragmentation in jointed rocks in an open-pit
manganese mine, Engineering Geology, Vol. 57, No. 1-2, pp. 1329.

7.

Bieniawski, ZT. (1976), Rock mass classifications in rock


engineering, Proceedings of the Symposium for Exploration for
Rock Engineering (1), Johannesburg, South Africa, pp. 97106.

CONCLUSIONS

8.

Gama, CD. (1996), The concept of rock mass Fragmentability.


Fragblast 5, Workshop on Measurement of Blast Fragmentation,
Montreal, Canada, pp. 209-214.

9.

Onederra, I. (2004), A fragmentation modelling framework for


underground ring blasting applications, Fragblast, Vol. 8, No. 3,
pp. 177-200.

Figure 7. Particle size distribution of the Blast E1.

Fragmentation index is a very effective tool to compare the blasts


performed under different condition. The index is directly the measure
of the work done by explosive. In this research, the performance of the
explosives was assessed using fragmentation index approach.
Determination of the in-situ blocks size of rocks for fragmentation index
calculations is a difficult and timeconsuming process. The use of the
image analysis software made it easier to determine apparent in-situ
block size of the benches.

10. Maerz, N., (1990). Photoanalysis Of Rock Fabric, Ph.D. thesis,


Canada: University of Waterloo.
11. Reida, TR, Harrison JP. (2000), A semi-automated methodology
for discontinuity trace detection in digital images of rock mass
exposures, International Journal of Rock Mechanics & Mining
Sciences, Vol. 37 No. 7, pp. 1073-1089.

It is apparent that explosive and rock coupling increases the


blasting efficiency. The water content of the holes causes detonation
failure of ANFO. Detonation failure is the main reason of the poor
fragmentation. The use of the water resistant emulsion explosive
eliminates detonation failure and results better fragmentation.
Additionally, emulsion is a denser explosive than ANFO. The
detonation velocity and detonation pressure of emulsion explosive are
higher than ANFO. High explosive characteristics also played a role in
better fragmentation. The cost of emulsion explosive is higher than
ANFO. Nevertheless, the increase in efficiency of loading and crushing
operations after blasting makes the cost a secondary concern. The use
of the emulsion explosive provided a 19% decrease in the amount of
the boulders. Image analysis software used in this research was very
effective measurement tool for blast fragmentation. Development in
image analysis technique will enable to measure the blast
fragmentation more accurately.

12. Kemeny, J, Norton, B. and Turner, K. (2006), Rock slope stability


analysis utilizing ground-based lidar and digital image
processing, Felsbau Rock and Soil Engineering, Vol. 24, No. 3,
pp. 8-15.
13. Kwangmin, K. (2006), Blasting design using fracture toughness
and image analysis of the bench face and muckpile, MSc.
Thesis, Virginia Polytechnic Institute.
14. Maerz, NH. (1996), Image sampling techniques and
requirements for automated image analysis of rock
fragmentation, Measurement of Blast Fragmentation Fragblast 5,
Montreal, Canada, pp. 115-120.

Copyright 2012 by SME

SME Annual Meeting


Feb. 19 - 22, 2012, Seattle, WA
APPENDIX
Table 2. The test blasts performed in the quarry
B
S
H
L
T
Blast No
(m)
(m)
(m)
(m)
(m)
A1
2.2
2.5
7.0
7.5
2.5
A2
2.2
2.5
7.0
7.5
2.5
A3
2.0
2.5
6.5
7.0
2.5
A4
2.2
2.7
5.5
6.0
2.5
A5
2.0
2.5
7.0
7.5
3.0
E1
2.2
2.5
7.0
7.5
2.5
E2
2.5
2.8
7.0
7.5
2.5
E3
2.0
2.5
7.0
7.5
2.5
E4
2.2
2.7
5.5
6.0
2.5
E5
2.2
2.5
7.0
7.5
3.0
E6
2.0
2.5
6.5
7.0
2.5
B: Burden, S: Spacing, H: Bench height,
XB: Average apparent in-situ block size,
B

Charge
Number of
XB
U
3
PF (kg/m )
(m)
Height (m)
(m)
Holes
0.5
0.69
5
23
0.93
0.5
0.69
5
16
1.15
0.5
0.73
4.5
13
1.32
0.5
0.57
3.5
12
1.26
0.5
0.68
4.5
14
1.45
0.5
0.97
5
22
2.06
0.5
0.76
5
23
1.24
0.5
1.07
5
14
1.13
0.5
0.80
3.5
13
1.86
0.5
0.87
4.5
17
1.37
0.5
1.03
4.5
16
1.31
L: Hole length, T: Stemming, U: Subdrilling, PF: Powder factor,
x50: Mean fragment size of muck pile, FI: Fragmentation index
B

x50
(m)
0.16
0.17
0.21
0.17
0.21
0.20
0.15
0.14
0.21
0.15
0.16

FI
5.81
6.76
6.29
7.41
6.90
10.30
8.27
8.07
8.86
9.13
8.19

Copyright 2012 by SME

You might also like