1 INTRODUCTION

Delay timing is an important and critical blast design

parameter which may have a direct impact on frag-
mentation, muckpile characteristics, backbreak and
ground vibrations (Lang 1979, Anderson et al. 1985;
Chiappetta 1997,1998; Cunningham 2000).
As discussed by Jimeno et al. (1995), delays
should permit the succession of the following
events:
- Propagation of the compression and tensile waves
from the blasthole to the free face.
- Readjustment of the initial field of tensions, due to
the presence of radial cracks and the effect of the re-
flection of the shock wave on the free face.
- Acceleration of the fragmented rock by the action
of gases up to a velocity that assures an adequate
horizontal displacement.
The onset of rock mass movement depends on the
material response in conjunction with the stress and
gas pressure stimulus generated by the explosive.
There is usually a discrete element of time that has
elapsed from the time of explosive detonation to
mass burden displacement. This time is designated
as the minimum response time (T
min
) and is depend-
ent on the burden mass, explosive and dynamic ma-
terial response to the explosive stimulus. Generally,
but not always, T
min
can be decreased by employing
small burdens, using higher energetic explosives or a
combination of both.
Figure 1 illustrates the dependency of T
min
on
blast geometry given by the burden to blasthole di-
ameter ratio and the interaction between the explo-
sive and rock mass. This figure is a modified version
of the data published by Chiappetta (1998). As
shown, in surface blasting conditions, the onset of
rock mass movement may range from a few milli-
seconds to hundreds of milliseconds depending upon
the material blasted.
Selection of inter-hole and inter-row timing for surface blasting - an
approach based on burden relief analysis
I.Onederra & S.Esen
Julius Kruttschnitt Mineral Research Centre -The University of Queensland, Qld, Australia





ABSTRACT: It is widely acknowledged that delay timing (inter-hole and inter-row), if optimised, can have a
positive impact on blast outcomes such as fragmentation, muckpile characteristics, backbreak and ground vi-
brations. Available or acclaimed methods for selecting suitable timing combinations are often not conven-
iently packaged for direct application by site engineers during the design process. In surface free face blasting,
a key parameter that should facilitate the process of selecting appropriate inter-hole and inter-row delays is the
concept of minimum response time (T
min
). This is the time which elapses between explosive detonation and
rock mass movement and is a function of explosive type, rock properties and blast design parameters. This
paper describes a convenient methodology for calculating the minimum response time or T
min
to enable the es-
timation of optimal inter-hole and inter-row delay times for a given condition. The approach and subsequent
model has been developed and validated using high speed film/video data of full scale tests gathered over a
relatively wide range of blast designs, explosives and rock types. For convenience or ease of use, the model
has been incorporated into an interactive computer based approach for the selection of appropriate inter-hole
and inter-row delays. This approach is based on detonation simulations and a probabilistic burden relief analy-
sis concept which is also introduced in this paper. A demonstration case study is used to describe the applica-
tion of the proposed model and the burden relief analysis concept.
0
10
20
30
40
50
60
70
80
90
100
110
120
0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80
B/d
M
i
n
i
m
u
m

r
e
s
p
o
n
s
e

t
i
m
e

(
m
s
)
Overburden 2 IRON ORE A IRON ORE B Overburden 1 Phosphate GRANITE COAL


Figure 1. Minimum response time (T
min
) versus burden to blasthole diameter ratio for a number of blasting environments (Chiap-
petta, 1998).


2 DELAY SEQUENCE DESIGN - CURRENT
PRACTICE

In current mining practice, the selection of appropri-
ate delay times (e.g. inter-hole and inter-row) in-
volves the implementation of preliminary design
guidelines or more quantitative methods based on
knowledge of the minimum response time (T
min
).
Preliminary design guidelines are empirical in na-
ture and include the application of simple rules such
as those documented by Jimeno et al. (1995). The
list includes recommendations by Bauer who sug-
gests delay times of 5-7 ms/m of burden in blasts
with holes of 38 to 311 mm diameter; Bergman et al.
and Fadeev et al. suggesting 3-6 ms/m and 4-8
ms/m, respectively and Konya & Walter proposing
delay times of 6-7 ms/m, 4-5 ms/m and 3-4 ms/m of
burden for low, medium and high strength rocks re-
spectively.
As discussed by Chiappetta (1998), a more objec-
tive methodology can be adopted from knowledge of
the minimum response time. For example, the delay
time between holes in a row should be less or equal
to the minimum response time, to encourage positive
interaction for improved breakage and fragmenta-
tion. However the delay time between rows should
be of the order of 1.5 to 3.0 times the minimum re-
sponse time to maximise material displacement
and/or create a loose muckpile.
Literature shows that there are only a handful of
methods available to estimate minimum response
time. These include direct measurements with the
application of high-speed video analysis techniques
and the use of mechanistic and numerical modelling
tools. The use of high speed video analysis tech-
niques is site specific, can be both time consuming,
difficult to implement and costly. Numerical codes
offer a promising route but they are still evolving
and are not readily available to engineers on site.
This has driven the development of an empirical ap-
proach that allow engineers to make preliminary es-
timates of the time at which rock mass movement is
expected to occur, given a particular geometry and
explosive/rock mass combination. The framework of
the proposed model is discussed next.
3 A MODEL TO ESTIMATE MINIMUM
RESPONSE TIME

The model building process involved the compila-
tion and back-analysis of several case studies in
which comprehensive blast monitoring campaigns
were undertaken. The current database is composed
of 19 case studies, covering a wide range of geo-
technical and blasting conditions. A detailed exami-
nation of the data available showed that minimum
response time (T
min
) is strongly dependent upon the
blast geometry (i.e. burden and hole diameter), stiff-
ness of the rock mass and explosive/rock interaction.
This is in agreement with what has been previously
found in the blasting literature (Lang, 1979 and
Chiappetta, 1998). From this analysis, the following
empirical relationship has been derived:


( )
1
1
]
1

|
|

\
|
|
|

\
|
=
b
mass
mass
ERI K d
B
a ERI K T
1
min
(1)
where T
min
is the minimum response time assumed
to occur at the centre of the explosive charge (ms);
K
mass
is the rock mass stiffness (GPa) which is a
function of the rock mass dynamic Youngs modulus
(E
d
) and Poissons ratio (v
d
): K
mass
=E
d
/(1+v
d
); B is
the burden (m); d the hole diameter (m) and ERI is
an explosive rock interaction term. This term has
been based on the explosive performance term intro-
duced by Bergmann (1983) and is given by the fol-
lowing expression:
( )
e
CJ
p p
e
D
D
v
D
v
D
D
ERI

+
+ =
2
2
2
1
36 . 0
(2)

where ERI is the explosive-rock interaction term;
e

is the density of the explosive (g/cm
3
); D is the ac-
tual (non-ideal) detonation velocity (km/s); v
p
is the
Pwave velocity of the intact rock (km/s) and D
CJ
is
the CJ detonation velocity (km/s).
In Equation 1, there are two fitting constants,
namely a and b. For the current database, they are
determined as 2.408 and 1.465, respectively. In rec-
ognising the empirical nature of the proposed model,
predictions are restricted to geometries covering B/d
ratios of 12 to 45 and assume that blastholes are
fully coupled and properly stemmed.
Preliminary validations of the proposed approach
have shown encouraging results. For example evi-
dence documented by Guest et al. (1995) and Vassie
(1991) suggested that the minimum response time of
trough rings fired in Tuffisitic Kimberlite Breccia
(TKB) was of the order of 30 ms. This corresponded
to a ring geometry consisting of 102 mm diameter
blastholes on a 3.6 m burden configuration, charged
with a pumpable emulsion product.
For similar conditions, the proposed model esti-
mates the minimum response time to be in the range
of 22ms to 33ms with the most likely time calculated
at 27ms. These estimates were based on a pumpable
emulsion product with a density of 1.18 g/cm3; Ideal
VOD of 5438 m/s and an estimated confined VOD
of 4916 m/s. TKB properties included an unconfined
compressive strength (
c
) of 32 MPa, Dynamic
Youngs modulus (E
d
) of 17 GPa and a rock mass
stiffness (K
mass
) of 10.2 GPa.
The proposed model has also been tested for its
ability to realistically predict changes of minimum
response time given by changes in geometry (i.e.
namely burden distance). Measurements documented
by Mishra & Gupta (2001) were used in this assess-
ment. Because rock mass parameters were not pro-
vided, K
mass
was estimated to be 4 GPa, from as-
sumptions of P- and S-wave velocities of the order
of 2100 and 1100 m/s respectively and a rock den-
sity of 2000 kg/m
3
. This rock mass can be classified
as a low strength and energy absorbing overburden.
Blasting parameters were given by Mishra & Gupta
(2001) and these allowed the estimation of the ex-
plosive-rock interaction term (ERI).
The estimation of rock mass stiffness (K
mass
) did
not affect the ability to check whether the model was
sensitive to changes in burden, as this was kept con-
stant. Results from this particular analysis are shown
in Figure 2. From an engineering point of view and
based on the data collected by Mishra & Gupta
(2001), modelling results appear to effectively cap-
ture the dependency of minimum response time on
burden for a constant explosive/rock mass combina-
tion.

Figure 2. Comparison between measured minimum response time and model predictions for a constant explosive/rock mass combi-
nation.
0
10
20
30
40
50
60
70
80
90
100
110
0 1 2 3 4 5 6 7 8
Burden (m)
T
m
i
n

(
m
s
)
MINE I - Jharia coal field MOST LIKELY Tmin (Model)
+20% Tmin (Model) -20% Tmin (model)

4 APPLICATION OF THE T
MIN
MODEL

Minimum response time can be used as a key input
in the definition of inter-hole and inter-row delay se-
quences in a free face blast. As discussed earlier,
guidelines such as those proposed by Chiappetta
(1998) can be directly applied (e.g. 1.5 to 3.0 times
the minimum response time to maximise material
displacement). There are, however, limitations to
this approach, as the choice of delay times will also
depend on the precision of the available down-hole
and surface delay detonators.
To assist with the selection of inter-hole and in-
ter-row delays, taking into account both T
min
and de-
lay scatter, the concept of burden relief was devel-
oped by research engineers from the Julius
Kruttschnitt Mineral Research Centre (JKMRC,
1990 & Riihioja, 2003). The approach is based on
the assumption that for an explosive deck to be suc-
cessful, as far as fragmentation and displacement is
concerned, a free face towards which the rock can be
blasted must be available. For a free face to be avail-
able, it is required that at least two adjacent holes in
front of the hole in question have detonated at a de-
fined time earlier, to provide sufficient burden relief.
4.1 Burden relief analysis
Figure 3 illustrates the burden relief analysis concept
for a staggered pattern. For hole 4 to be considered
as having detonated successfully, holes 1 and 2 or 1
and 3, need already to have detonated successfully at
an earlier time interval, given in this case by the ex-
pected minimum response time (T
min
).

S
B
free face
X
1 2
3
4
5


Figure 3. Staggered pattern hole layout.

In general, the approach simply estimates the likeli-
hood of a deck being over-burdened at the instant of
detonation. To estimate the probabilities of success
or failure of a blasthole, a simple burden relief cri-
teria is checked for a number of detonation simula-
tions. The criteria follows that for a given relief dis-
tance (i.e. the largest distance between holes, x in
Figure 3), there is a minimum number of explosive
charges that must go off before a hole can detonate
successfully (two is the default number), these
charges should have also detonated at an earlier time
interval defined by the relief time or expected mini-
mum response time.
In order to compile the statistics necessary to es-
timate detonation scatter and probabilities of success
or failure, it is necessary to run a number of detona-
tion simulations. For a given simulation, an explo-
sive deck is 100% successful when the whole crite-
ria is met, if one fails, then the deck is assumed to
fail and subsequent detonations which rely on this
hole will also fail. Experience has shown that it is
possible to choke a whole section of a blast if just
one hole fires out of sequence.
During the simulation of the detonation sequence,
the inherent variability of pyrotechnic delays is mod-
elled with a normal probability distribution function
given by the nominal detonation time (mean) and a
scatter factor which defines the standard deviation.
For each simulation, Monte Carlo sampling is used
to estimate the likely detonation time of each deck.
The simulations also consider the velocity of detona-
tion of the explosive charge, down-hole shock tubes
and any type of surface connectors (i.e. shock tube or
detonating cord).
To demonstrate the use of the burden relief analy-
sis concept for design and optimisation purposes, a
simple case study is presented in the following sec-
tion.
4.2 Demonstration case study
This section demonstrates the use of the burden re-
lief analysis concept. All algorithms have been
coded into a blast design package designated as
JKSimBlast (JKTech, 2003). As will be shown, the
output of the burden relief concept is graphical, dif-
ferent colour scales are used to identify the probabil-
ity of success or failure of a deck for a pre-defined
number of simulations. This graphical output allows
the engineer to quickly assess areas of a blast that
could be affected by over-confinement and thus take
controlling measures. The process is done interac-
tively by selecting delays from a pre-defined set of
accessories with unique characteristics (i.e. depend-
ing upon supplier). If pyrotechnic delays are being
used, the engineer is able to run what if scenarios for
any number of scatter (variability) factors and thus
identify the risks of adopting a specific tie-up con-
figuration using pyrotechnics and at the same time
assess the benefits of increased detonator precision.
The following demonstration case study is based
on a regular staggered pattern of eight rows with ap-
proximately 20 holes each. Nominal design parame-
ters are described in Table 1.








Table 1. Nominal design parameters.
______________________________________________
Burden (m) 8
Spacing (m) 9
Blasthole diameter (mm) 241
Blasthole length (m) 15.2
Charged length (m) 9.1
Explosive type WRANFO
Explosive density (g/cm
3
) 0.99
Ideal VOD (m/s) 5829
Confined VOD (m/s) 5029
P-wave velocity (intact) (m/s) 2902
Rock mass stiffness (GPa) 3.2
_____________________________________________

Based on the parameters summarised in Table 1 and
by applying the empirical approach described earlier,
the estimated minimum response time is 57ms to 86
ms, with the most likely time being 71ms.
For this case, the following burden relief crite-
rion was established: For a blasthole to be 100%
successful, two charges that are within a distance of
10m must detonate successfully at a time interval
within 71ms.
For demonstration purposes, the analysis has in-
volved the assessment of two design sequences (Ta-
ble 2).








Table 2. Proposed delay configurations.
______________________________________________
Design 1 Design 2
__________________________________________
Inter-hole (ms) 42 65
Inter-row (ms) 65 150
Down-hole (ms) 500 500

The design 1 configuration was based on the rule of
thumb of 6-7 ms/m of spacing for low strength rock
environments and design 2 on the premise that 0.9
times the minimum response (T
min
) would be ade-
quate to increase the likelihood of inter-hole coop-
eration for breakage, and that 2 times T
min
would be
adequate for effective burden relief to take place. In
both cases, downhole delays of 500ms were consid-
ered necessary in order to allow the initiation front to
precede the detonation front, and thus reduce the
likelihood of surface cut offs. This can be visually
assessed during the playback of the detonation se-
quence in the JKSimBlast software.
In the above design sequences, 1000 Monte Carlo
simulations were conducted with an assumed delay
scatter of 4 %, which as discussed earlier, defines the
standard deviation to be 4% of the nominal delay.
Results from these two simulations are given in Fig-
ures 4 and 5 for designs 1 and 2, respectively.






Figure 4. Simulation results for design 1: Inter-hole = 42ms, Inter-row = 65ms, Down hole = 500ms; Simulations = 1000.
Scatter factor = 4%.


Figure 5. Simulation results for design 2: Inter-hole = 65ms, Inter-row = 150ms, Down hole = 500ms; Simulations = 1000.
Scatter factor = 4%.


As shown in Figure 4, the adopted sequence is
clearly not successful, as most blastholes fail to meet
the pre-defined burden relief criteria. This however
was not unexpected, as the estimated minimum re-
sponse time of 71ms is clearly greater than the
adopted 65ms inter row delays. Figure 5 shows an
improved outcome. In theory the use of 150ms inter
row delays should provide the necessary relief to
make the sequence successful, however towards the
back and corners of this blast, blastholes fail the
burden relief criterion, this may translate to the over-
confinement of back rows. Practical experience
shows that the over confinement of back rows or
corners is expected in blast patterns with a large
number of holes and rows, and particularly when us-
ing pyrotechnic delays.
A closer examination of the results given in Fig-
ure 5 indicated that the failure of blastholes was
mainly attributed to the inherent variability of the
pyrotechnic delay elements used (i.e. 4% scatter). To
confirm this hypothesis and to test the sensitivity of
the burden relief analysis algorithms, new simulation
runs were conducted for design case 2, but this time
with a reduced scatter factor (i.e. down to 0.1% for
all delay elements). Results from this analysis are
shown in Figure 6. With the exception of the corner
regions of the blast, the majority of holes appear to
have adequate relief and match the criterion of suc-
cess, hence demonstrating the benefits of increased
detonator precision.
This demonstration case study has shown that the
use of minimum response time as an input for con-
ducting burden relief analysis, can help identify re-
gions of a blast that may be subject to over-
confinement. This can help engineers make a rapid
assessment of the tie-up configuration adopted and,
if need be, provide corrective action prior to imple-
mentation.
It is also important to recognise that in terms of
controlling ground vibration amplitudes and fre-
quencies, the choice of delay times plays a vital role.
For conditions in which vibration must be con-
trolled, the burden relief analysis should be supple-
mented with specific analyses conducted with vibra-
tion modelling and monitoring methods.






Figure 6. Simulation results for design 2: Inter-hole = 65ms, Inter-row = 150ms, Down hole = 500ms; Simulations = 1000.
Scatter factor = 0.1%.


5 CONCLUSIONS
Minimum response time can be used as a key input
parameter in the design of appropriate inter-hole and
inter-row delays under free face blasting conditions.
An empirical model that allows the estimation of
minimum response time has been introduced. Pre-
liminary validations show that the model is able to
capture the dependency of minimum response time
on burden, explosive type and rock mass conditions.
It is however recognised that further validation work
is still required to improve its predictive capabilities.
The concept of burden relief analysis has been in-
troduced. The application of minimum response time
and burden relief analysis as a tool to assist in the se-
lection inter-hole and inter-row delays has been
demonstrated with a simple case study.
Burden relief analysis has shown that when a
theoretically acceptable delay sequence is chosen,
the inherent variability of pyrotechnics can still in-
fluence the final outcome. When precision is im-
proved, then the burden relief criterion is clearly
met. This highlights the benefits of delay precision
as a way of having true control over the sequence of
detonation of a blast. This allows greater control
over the degree of confinement (relief) of blastholes,
prior to detonation. In practice, the provision of ade-
quate relief has been shown to have a marked impact
on the final degree of uniformity of fragmentation,
muckpile looseness and back-break.
In terms of controlling ground vibration amplitudes
and frequencies, the choice of delay times is also
crucial. For conditions in which vibration must be
controlled, the burden relief analysis should be sup-
plemented with analyses conducted with specific vi-
bration modelling and monitoring methods.
6 ACKNOWLEDGEMENTS
The authors would like to acknowledge the support
of the Julius Kruttschnitt Mineral Research Centre
(JKMRC) and in particular the contributions of Mr.
Kai Riihioja, principal developer of the JKSimBlast
software.
REFERENCES
Anderson, Winzer & Ritter. 1985. The relationships between
rock structure and delay timing, and fragment size and dis-
tribution in explosively-loaded rock. Proceedings of frag-
mentation by blasting. SEM first edition: 41-62.
Bergmann, O.R. 1983. Effect of explosive properties, rock type
and delays on fragmentation in large model blasts, Proceed-
ings of the First International Symposium on Rock Frag-
mentation by Blasting, Lulea, Sweden: 71-78.
Chiappetta, R.F., Vandenberg, B. & Pressley, J.R. 1997. Port-
able, multi-channel and continuous velocity of detonation
recorders, Seventh High-Tech Seminar, BAI Inc., Florida,
USA: 787-865.
Chiappetta, R.F. 1998. Blast Monitoring Instrumentation and
Analysis Techniques, With an Emphasis on Field Applica-
tions, FRAGBLAST-International Journal of Blasting and
Fragmentation,Vol.2, No. 1, A.A.Balkema, Rotterdam: 79-
122.
Cunningham, C.V.B. 2000. The effect of timing precision on
control of blasting effects. Proceedings of Explosives and
Blasting Technique, Holmberg (ed.). Balkema, Roterdam:
123-127.
Guest, A.R., Chitombo, G.P. & Grobler, H.P. 1995. Blast
optimization for efficient extraction of a block cave
undercut Case studies at De Beers Consolidated Mine
Ltd., South Africa. Proceedings of EXPLO 95.
AUSIMM:75-80.
Jimeno, C.L., Jimeno, E.L. & Carcedo, F.J.A. 1995. Drilling
and Blasting of Rocks, A. A. Balkema, Rotterdam, Nether-
lands:235-286.
JKMRC 1990. AMIRA Project P93D Final report. JKMRC,
Australia.
JKTech 2003. www.jksimblast.com
Lang, L. C. 1979. Buffer blasting techniques in open pit mines.
Proceedings of the 5th conference on explosives and blast-
ing techniques. SEE annual meeting: 66-80.
Mishra,, A.K. & Gupta, R.N. 2002. Design of blast using high
resolution camera. Proceedings of the Seventh International
Symposium on Rock Fragmentation by Blasting, Metallur-
gical Industry Press, Beijing, China: 378-389.
Riihioja, K. 2003. Personal communication.
Vassie, R.S. 1991. High speed photographic analysis of ring
blasting at The Finsch and Koffiefontein mines. A Bauer &
Crosby Pty Ltd Mining Engineers project report to De
Beers. June.