LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO

NOVEL TECHNIQUES.

G.F. Brent and L.W. Armstrong.

ICI Explosives Technical Centre,
George Booth Drive,
Kurri Kurri, NSW,
2327, Australia.

ABSTRACT

It is recognised that presplitting is a special case in blasting which requires the reduction of
conventional explosive pressures in order to avoid excessive rock damage. Two routes to
achieving this are the use of reduced density explosives or the practice of decoupling the
explosive from the blasthole walls. Applications of two such systems in large diameter, deep
blastholes are described. Both systems were tested in presplits alongside sections charged
with decks of conventional explosives.

The first system comprised an explosive formulation with an unusually low bulk density of
0.2 g/cm (12.5 lb/ft ). Following unconfined characterisation trials, a field trial at a large
open cut coal mine was performed in 311 mm (12 ¼ in) diameter vertical holes at depths of
about 45 m (150 ft). In-hole velocities of detonation (VOD) averaged 2200 m/s (7660 ft/s).
The resultant highwall after excavation revealed a half barrel factor of 62 % for the test
section compared with only 32 % for the adjacent section which had been charged
conventionally. This demonstrated the success of this approach for large diameter
presplitting.

The decoupled system comprised charges of 150 mm (6 in) diameter and 6 m (20 ft) length
containing an emulsion bulk explosive at a nominal density of 1.2 g/cm (75 lb/ft ). These
charges were employed in a presplit blast in a large open cut coal mine in 270 mm (10 5/8
in) diameter inclined holes which were up to 70 m (244 ft) deep and contained up to 60 m
(200 ft) of water. The presplit blast was successful and VOD measurements averaged 5800
m/s (20 200 ft/s). After excavation, the resultant highwall was seen to be smoother in the test
section in comparison to the adjacent section where conventional charges had been used,
demonstrating that this system also provided improved presplit results.

INTRODUCTION

The goal in presplit blasting is to form a single planar crack between the presplit holes
without causing extensive radial cracking and damage. The quality of this crack plane will
define the new highwall and for improved wall stability it is important that damage within
the wall structure is minimised. This requirement makes presplitting a special case in
blasting, because its major aim is to prevent fragmentation, while still ensuring the formation

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 511 1 of 10
of one preferential crack.

Many models and some simple formulae for presplit design have been proposed in the
literature (see for example the overview by Brent, 1995), but it remains a complex dynamic
blasting situation which has not been adequately modelled to date.

Most of the literature on presplit blast design recognises that the detonation pressures
produced by explosives, typically 10(exp 9) - 10(exp 10) Pa, are considerably in excess of
rock strengths. Unconfined static compressive and tensile strengths are typically 10(exp 6)
10(exp 8) Pa, although dynamic strengths are greater. This means that explosive pressures
exerted on the borehole walls need to be reduced in order to avoid failure and extensive
crack formation around the blasthole.

Two common means of achieving lower pressures which act on the borehole walls are:
- reducing the explosive density; and
- decoupling the explosive charge from the borehole wall.

Decoupling has been extensively used in small diameter (usually < 130 mm) applications
and is the accepted method for presplitting in these situations. However, there appears to be
limited previous work to test either decoupled or low density systems in large diameter
presplitting. This work describes applications of these systems in large diameter production
scale presplits.

ULTRA LOW DENSITY EXPLOSIVE

Reduction of the explosive density has the effect of reducing both the velocity of detonation
(VOD) and the peak pressure generated by the explosive. This effect may be demonstrated
by the theoretical ideal detonation velocities and pressures generated by ANFO of varying
density, as shown graphically in Figure 1 which also shows quadratic fits to the data. The ICI
ideal detonation code, IDeX (Braithwaite et al., 1996), which is based on a fundamental
intermolecular potential equation of state, was used for these calculations.

In real non-ideal detonations (for example, Sheahan and Minchinton, 1988), the VODs, and
hence detonation pressures, are lower than those calculated from ideal detonation codes.
However, they will still decrease as the density decreases.

Another advantage of density reduction in presplitting is the increase in the length of the
explosive column for a given charge mass. A longer charge provides a better distribution of
the explosive along the borehole than a high density, shorter charge.

The successful use of reduced density explosives in large diameter presplitting has been
reported by Stachura and Cumerlato (1995). They used a watergel with a density of 0.5
g/cm³ in 270 mm diameter presplit holes and compared its performance with conventional
bulk ANFO and emulsion-ANFO blends. They reported less visible overbreak around the
charges and a higher half cast factor for the low density explosive. The half cast factor, or, as

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 512 2 of 10
used here, the half barrel factor (HBF), is defined as the length of visible hole remnants,
called the "half barrels", in the highwall expressed as a percentage of the total length of
presplit holes drilled.

EXPLOSIVE SYSTEM AND CHARACTERISATION

The use of explosive blends using low density fillers has been reported for many years, for
example Greef (1977), Heltzen and Kure (1980) and Hunter et al. (1993).

For presplitting, it was considered that densities lower than those usually achieved with such
blends would be required for deriving the most benefit in terms of reduced damage and good
presplit performance. A blend of nominal density 0.2 g/cm³ was formulated. It was possible
to achieve this ultra-low density in a composition which maintained good flow properties
without any segregation of the components.

Unconfined test charges were made in a range of charge diameters and were fired in order to
measure VOD and to determine the minimum priming requirements so that the robustness of
the formulation could be ascertained before field trialling. Measurement of VOD was done
using a multichannel timer triggered by the light signal from optical fibres inserted into the
charges. The unconfined VOD results were fitted to the following relationships:

VOD = 2.6 - 223/d for d>200mm

VOD= 1.9-85/d for 100<d<200mm.
where VOD = Unconfined detonation velocity (km/s)
d = Unconfined charge diameter (mm)

The unconfined critical charge diameter, below which detonation was not sustained, was
found to be approximately 100 mm. It was found that the explosive was adequately initiated
by a 120 g primer.

The above characterisation trials confirmed the robustness of this ultra-low density
formulation for use in a large scale production presplit trial.

PRESPLIT TRIAL

Trial Set-up

A trial was conducted at a large open cut coal mine in the Hunter Valley of New South
Wales, Australia. The presplit at this mine comprised vertical blastholes of 311 mm diameter
on a 4 m spacing. Each hole was usually charged with two 100 kg charges of emulsion-based
explosives at a nominal density of 1.2 g/cm³. One charge was located at the toe while the
second was located halfway up the holes. The hole depths ranged from 39-46 m. Initiation
was by 5 g/m detonating cord trunkline and downlines with 400 g primers in each explosive
charge. Groups of ten holes were separated by 45 ms delays to reduce vibration levels.

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 513 3 of 10
The test was conducted in 19 holes within an otherwise conventionally-loaded presplit
consisting of about 100 holes in total. The ultra-low density explosive was loaded as a single
200 kg toe charge in 11 of the test holes and as two charges of 100 kg each in the remaining
8 holes. The total explosive column lengths in each test hole were thus about 13 m compared
with just over 2 m for the conventional explosive. Each charge was initiated by a 400 g
primer on detonating cord downlines, except for 3 holes which were used to measure the
VOD. These holes were initiated by signal tube downlines. A VOD measurement system
based on continuous time domain reflectometry was used (Brown et al., 1993).

Trial results

Inhole VODs were successfully recorded in all three holes and an example trace is shown in
Figure 2. The average VOD from the three holes was 2200 m/s.

After the blast, there were large open presplit cracks visible at the hole collars. These cracks
were more pronounced in the test section, with no radial fracturing evident. Figure 3 shows
examples of these.

Subsequent to firing and excavation of the production blast in front of the presplit, the
highwall was exposed some months later. Figure 4 shows the resulting highwall, with the
test section, comprising about 80 m strike length of presplit, marked. Although the overall
appearance of the highwall was excellent everywhere, there appeared to be a smoother face
in the test section, with less rock "clinging" in front of the presplit plane.

Photographs taken perpendicular to the highwall were used to measure the HBF. The section
of highwall directly adjacent to the test section was used as the control. The results,
summarised in Table 1, show that the HBF was nearly doubled in the test section.

LARGE DIAMETER DECOUPLED CHARGE

Decoupling the explosive from the blasthole walls provides an annular space for the
expansion of the detonation products, thus allowing a reduction in the pressure exerted on
the blasthole walls. It may be likened to reducing the overall density of the explosive within
the blasthole. An additional factor is that the explosive detonates in an essentially
unconfined state, thus leading to increased non-ideal behaviour and a consequent reduction
in the VOD and detonation pressure.

Decoupled explosives are commonly used in small diameter presplitting applications.

Generally, explosive cartridges are lined either internally or externally with detonating cord.
This ensures that they are continuously side-initiated by the cord and that
cartridge-to-cartridge propagation failures or disruption due to the channel effect (see for
example Udy, 1977) are avoided.

The use of larger diameter charges in large diameter, usually deep, presplit holes is generally
restricted by the handling difficulties which arise as the weight of the charge column

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 514 4 of 10
increases. Additional problems of large magazine inventories also arise.

EXPLOSIVE SYSTEM AND CHARACTERISATION

A system was designed for use in very deep, large diameter presplits which contained water.
In order to address the problems mentioned above, it was based on a bulk explosive in a
charge container which was to be lowered down the hole using a mechanical winch mounted
on a truck. Its use in water-filled holes was also expected to avoid problems from the
channel effect.

The charge container was a plastic pipe of 150 mm diameter and 6 m length. It was capped at
one end and primed with a 900 g primer which was initiated by an instantaneous signal tube
detonator. An emulsion explosive which had been specifically formulated to withstand high
hydrostatic pressures encountered in deep, wet holes was selected. It was pumped from an
explosives truck into the charge container which was initially suspended from a rope tied to
a stake at the hole collar. Approximately 120 kg of explosive at a nominal density of 1.2
g/cm³ was used in each container. Once filled, the charges were slowly lowered down to the
bottom of the holes using the winch.

A preliminary trial using six charges was conducted in 270 mm diameter presplit holes. The
holes were inclined at 20° to the vertical and the depths ranged from 36 m to 46 m. No
stemming was used. The signal tube downlines were initiated by a detonating cord surface
trunkline. The trial proved that the methods used for loading and lowering of the charges
were safe and practical. All the charges appeared to fire as evidenced by the water ejection
from the holes. The VOD was measured in one of the charges at 5840 m/s.

PRESPLIT TRIAL

Trial Set-up

The system was implemented in a full scale production presplit in a large open cut coal mine
in the Bowen Basin of Queensland, Australia. The presplit was designed to produce a
highwall of 65-70 m. This meant that the holes were much deeper than those conventionally
used (up to 55 m). Fast flowing water was present at the southern end of the presplit with
water levels in some holes as much as 60 m.

It was considered important to achieve a good presplit in a wall of this height in order to
maximise wall stability. It was found that decks of conventional explosive at the bottom of
the holes in the southern section were washed away by the water flow. Furthermore, because
of the high water levels, it was not possible to locate the standard self-inflating airbags down
these holes, which were usually used at the desired depth to hold the upper explosive deck
charges. These factors meant that the decoupled charge system was required at the southern
end.

The southern endwall which had a strike length of about 45 m and the first 160 m at the

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 515 5 of 10
southern end of the main highwall were loaded with decoupled charges. In total, 52 holes
were loaded with the decoupled charge system. Most of the holes were loaded with one 120
kg charge at the bottom and a second charge of 100 kg suspended 25 m from the hole collar.
All charges were initiated by 900 g primers on instantaneous signal tube downlines. The
downlines were of equal length to ensure simultaneous initiation of the charges, except for
two holes which were monitored for VOD. In these holes, the signal tube lengths were
arranged so that the upper deck would initiate about 10 ms after the bottom deck. A
detonating cord trunkline was used on the surface to initiate the downlines. No stemming
was used.

The rest of the presplit was conventionally charged with a bottom deck of 150 kg of
emulsion explosive and an upper deck of 90 kg of a heavy ANFO explosive. All of these
charges were initiated by 400 g primers on detonating cord downlines with a surface
detonating cord trunkline. No stemming was used.

Trial Results

Video records of the presplit blasts showed that all the holes appeared to detonate
successfully. The VOD records showed that the decoupled charges detonated successfully at
an average VOD of about 5800 m/s. Figure 5 shows the VOD trace obtained from one hole.
The detonation of both decks can be clearly distinguished.

Several months later, after the production blast in front of the presplit had been fired and
excavated, the resulting highwall was photographed. While the overall appearance was
impressive, the first 150 m of the southern end where the decoupled charges had been used
was visibly smoother. Figure 6 shows a view of the highwall. Access restrictions meant that
it was not possible to take perpendicular photographs of the highwall for determination of
the HBF.

CONCLUSIONS

The potential theoretical benefits of reduced density and decoupling in presplitting have been
obtained in large diameter presplitting trials. Improved highwalls were obtained through the
use of two novel systems:

- an ultra-low density explosive yielded a doubling in the half barrel factor when
compared to conventional explosive in a trial in a large open cut coal mine; and
- a large diameter decoupled charge based on a bulk explosive was used to produce a
smoother highwall than conventional decks of explosive in another large open cut
coal mine.

In the latter case, the presence of fast flowing water and its excessive depth in one section of
the presplit precluded the use of conventional explosive decks and the decoupled charge
system provided a unique solution.

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 516 6 of 10
ACKNOWLEDGEMENTS

The authors wish to acknowledge the substantial assistance of personnel from Ulan Coal
Mine, NSW and Newlands Coal Mine, Queensland in conducting these trials. Both mines
are thanked for their assistance and permission to publish this paper. The authors also
acknowledge the assistance of ICI Explosives personnel, in particular the late Dr. R.M.
Sheahan.

REFERENCES

Braithwaite, M., Byers Brown, W. and Minchinton, A., 1996, "The use of ideal detonation
computer codes in blast modelling", Proceedings of the 5th International Symposium on
Rock Fragmentation by Blasting-Fragblast 5, Montreal, Balkema, pp. 37-44.

Brent, G.F., 1995. "The design of Presplit Blasts", in Proceedings of. Explo'95,
Brisbane:AusIMM, pp. 299-305.

Brown, C. F., Spathis, A.T., Felice, J.J. and Henley, K.N., 1993 "Continuous velocity of
detonation measurements in full scale blast environments", Proceedings of 6th Mining,
Drilling and Blasting BHP Seminar, Cairns, HP.

Greef, P, 1977, "Use of Isanol at Roseberry Mine", Proceedings of AusIMM Conference

(Conference Series No.6), Parkville, AusIMM, pp. 249-255.

Heltzen, A. M. and Kure, K, 1980, "Blasting with ANFO/Polystyrene Mixtures",

Proceedings of 6th Conference on Explosives and Blasting Technique, Tampa, SEE, pp.
105-116.

Hunter, C., Fedak, K. and Todoeschuck, J.P., 1993, "Development of Low Density
Explosives with Wall Control Applications", Proceedings of the 19th Conference on
Explosives and Blasting Technique, San Diego, ISEE, pp. 549-555.

Sheahan, R. M. and Minchinton, A., 1988, "Non-ideal explosive performance prediction

using the CPeX model", Proceedings of Explosives in Mining Workshop-Explo '88,
Melbourne, AusIMM, pp. 17-22.

Stachura, V. J. and Cumerlato, C., 1995, "Highwall damage control using presplitting with
low-density explosives", Proceedings of the 21st International Conference on Explosives and
Blasting Technique, Nashville, ISEE, pp. 187-197.

Udy L.L., 1977, "The influence of detonation generated plasmas upon the performance of
semi-confined explosive cartridges", Proceedings of the 3rd Conference on Explosives and
Blasting Techniques, Pittsburgh, SEE, pp. 90-109.

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 517 7 of 10
Figure 1. Variation of ideal VOD and detonation pressure of ANFO with density.

Figure 2. A VOD trace from one of the ultra-low density charges.

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 518 8 of 10
Figure 3. Well formed presplit cracks at the hole collars in the test section.

Figure 4. View of the highwall after excavation, showing the test section where the ultra-low
density explosive was used.

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 519 9 of 10
Figure 5. VOD trace from one of the presplit holes containing two decoupled charges.

Figure 6. View along the 70 m highwall with the test section where the decoupled charges
were used in the foreground.

Table 1. Results of Half Barrel Factor Photographic Analysis

Section No. of Total Total Half Barrel

Holes Length of Length of Factor (%)
sampled visible Half visible
Barrels Highwall
(mm) (mm)
Control - Conventional 18 305 968 32
Explosive
Test- Ultra-Low 18 526 846 62
density Explosive

Copyright © 2000 International Society of Explosives Engineers

1998G - LARGE DIAMETER PRESPLITTING IMPROVED THROUGH TWO NOVEL TECHNIQUES. - P 520 10 of 10