SME Annual Meeting

Feb. 19 - 22, 2012, Seattle, WA

Preprint 12-002
SOLUTIONS TO REDUCE BLAST-INDUCED ORE LOSS & DILUTION AT AHAFO GOLD MINE IN GHANA
W. Rogers, JK Tech Pty Ltd, Indooroopilly. QLD, Australia
S. S. Kanchibotla, JK Tech Pty Ltd, Indooroopilly. QLD, Australia
A. Tordoir, W H Bryan Research Centre (BRC), St Lucia, QLD, Australia
S. Ako, Newmont Ahafo Operations, Kenyasi, Ghana
E. Engmann, Newmont Ahafo Operations, Kenyasi, Ghana
B. Bisiaux, Newmont Ahafo Operations, Kenyasi, Ghana

ABSTRACT
Blast-induced rock mass displacement can have a significant
impact on ore loss & dilution. The mischaracterisation of the grade
boundaries both prior to, as well as a result of blasting can lead to
significant financial losses on account of ore loss and dilution. Ore
dilution occurs when waste material is miscategorised as ore and sent
for processing or stockpiling, diluting the run of mine head grade and
recovery.
Ore loss takes place when valuable mineral is
miscategorised as waste and sent to the waste dumps, thus never
being processed and recovered.

Extensive research has been conducted at the University of

Queensland over the last fifteen years to understand the impact of
blast movement on ore loss and dilution. Recent outcomes of this
research have developed innovative tools and techniques to measure
and model blast movement.

Horizontal movement increases as distance from the active

face increases (for current free face conditions) but stabilizes
after a certain distance behind the active face. Horizontal
displacement is maximum in the mid bench region of the blast
which has the highest energy concentration.
In well controlled blasts, movement is consistent within the
body of the blast except near the power trough, where the
movement is different to the body of the blast.
Mining ore with pre-blast boundaries can have significant
adverse economic impact as direct cost due to ore loss and
dilution and as indirect opportunity cost due to misclassification
of low grade ore as high grade ore.

These techniques have been implemented at Ahafo gold mine in

Ghana to understand blast movement dynamics and to implement
solutions to reduce blast-induced ore loss and dilution. These solutions
are validated through controlled trials and have been incorporated into
site standard operating procedures to sustain the benefits.
INTRODUCTION
A methodology using these techniques to minimise ore loss and
dilution in open pit mines is shown in Figure 1.This methodology has
been implemented at Newmont Ghana Gold, Ahafo operations and is
summarised as follows:

Understand the dynamics of blast movement with

comprehensive monitoring using high speed video and blast
movement markers.
Develop site specific models to predict the extent of movement
within the blast and its impact on grade control.
Develop alternative strategies to minimise ore loss and dilution.
Validate the alternative strategies with controlled trials and then
incorporate them into standard operating procedures to sustain
the gains.

Figure 1. Methodology to minimize ore loss and dilution.

Rogers and Kanchibotla et al., 2012, show the potential economic
impact of ore loss and dilution estimated for the current grade control,
blasting practices where ore polygons are not adjusted for blast
movement.

Three benchmarking trial blasts were monitored at Ahafo in

November 2010 to understand the blast movement dynamics. The
results fromthe measurement and modelling of blast movement in
these trial blasts have been discussed in Rogers and Kanchibotla et
al., 2012. The main findings of this work were:

The economic impact of dilution is calculated as the extra cost

associated with mining and processing the blasted tonnes of waste
that are treated as ore.
The economic impact of ore loss is calculated as the net profit lost
when ore is misclassified as waste and sent to the waste dump and
never recovered.

The magnitude and direction of rock movement in a blast is

influenced by the energy concentration, confinement, quality of
free face, timing and initiation pattern.
Uncontrollable and inconsistent blast movement take place due
to uneven free faces, uneven drill patterns, poor stemming
practices and excessive confinement along the centre lines and
at the back of the blast.
General direction of blast movement is perpendicular to the
timing contours with a degree of variation typically within 20.

The economic impact of misclassified is was calculated as the

opportunity cost when low grade material is processed through the mill
instead of high grade
For the three benchmarking trial blasts, the economic impact of
ore loss and dilution is estimated to be approximately half a million US
dollars and the opportunity cost due to misclassification is estimated to

Copyright 2012 by SME

SME Annual Meeting

Feb. 19 - 22, 2012, Seattle, WA
applying the most common horizontal movement vector for top flitch by
8m and bottom flitch by 12m (Figure 2).

be approximately 2.5 US million dollars. Ore loss, dilution and

misclassification percentages for the three blasts are shown in the
Table 1.
Table 1. Summary of ore loss, dilution and misclassified ore.
Top Flitch
Bottom Flitch
Trial Blast ID Dilution Ore loss Mis-classified Dilution Ore loss Mis-classified
AP1052_210
AP1052_211
AP1052_214

2.7
4.2
4.9

0.6
0
0.4

2.7
4.6
8.0

2.5
4.6
4.8

0
0
0

2.4
8.2
8.0

ALTERNATIVE STRATEGIES TO MINIMISE ORE LOSS AND

DILUTION
Based on the understanding of blast movement dynamics from
the benchmark blasts, alternative strategies were developed to reduce
blast induced ore loss and dilution at Ahafo. These strategies
included:

Blast designs to achieve consistent movement along the strike

of the ore body and minimise inconsistent movement from
edge effects, uneven free faces and cratering especially along
the ore / waste boundaries.
Adjusting the post blast ore boundaries to account for expected
blast movement and to ensure that excavation follows the
adjusted polygons.

Figure 2. Histogram of horizontal blast displacement from all three

trial blasts.
The direction of all ore polygon adjustments is made
perpendicular to the timing contours because it is the most common
direction measured during the trials (Rogers and Kanchibotla et al.,
2012). The ore polygons adjustment method is shown in Figure 3. A
single adjustment value is used for each flitch for the following
reasons:

Alternative blast designs

The utilisation of blast movement measurements to improve
grade control, by altering blast design and initiation sequencing, has
previously been discussed by a number of authors (e.g. Taylor et al,
1996,Harris et al, 2001, Firth et al, 2002).

To minimise ore loss & dilution at Ahafo; blast design and

implementation procedures were aimed at minimising edge effects
(e.g. power trough & free faces), uncontrolled movement (e.g.
cratering) and detrimental blast movements (i.e. ore moving into waste
or vice versa). Blast timing and initiation patterns were designed to
promote consistent movement parallel the strike of ore body.

To keep the procedure simple for the body of the blast; 80 / 20

rule
Ore body geometry in the main ore zone of Apensu pit (where
the trials were conducted) is large and the most common blast
movement is consistent and along the strike of the ore body.

Consistent and controlled blast movement along the ore strike is

the most effective way of minimising ore loss & dilution and this was
achieved by designing:

A shallow V initiation
Centreline away from ore / waste boundaries

Edge effects result in inconsistent and uncontrollable blast

movement and makes post blasts adjustments less effective. Edge
effects were minimised by:

Slowing down the timing in the back rows.

Avoid blasting ore / waste boundaries in the power trough
region.

Figure 3.
Ore polygons adjustments post blast with average
movement vectors.

Cratering and uncontrolled movement along ore / waste

boundaries increases ore loss & dilution. This was minimised by:

For the three trial blasts measured at Ahafo, applying average

movement adjustments to the ore polygons, to account for the
expected blast movement, would have reduced the cost of ore loss &
dilution by an estimated 33% (~ 160,000 $US) and opportunity costs
from misclassification by an estimated 56% (1,500,000 $ US).

Better QA / QC along the ore / waste boundaries.

Better quality and increased stemming near ore/waste
boundaries.
Designing centreline away from ore / waste boundaries.

This method of applying average movements vectors to adjust

post blast ore polygons is effective, timely efficient and alleviates the
necessity to continually monitor future blasts.

Post blast ore boundary adjustments

The method of using blast movement measurements to estimate
average movement vectors for various regions of the blast has
previously been discussed by Thornton et al., 2005. In this case,
movement templates were applied to pre-blast ore polygons to
account for blast movement.

VALIDATION TRIALS
During May 2011, two validation blasting trials were conducted to
confirm the effectiveness of the above alternative strategies. The
setup of the two validation blasts (i.e. AP-1028-211P and AP-1028210P) is shown in Figure and included the following:

Blast movement monitoring results (Rogers and Kanchibotla et

al., 2012) from the trial blasts at Ahafo showed that the direction and
magnitude of movement in the body of the blast is consistent when the
edge effects and inconsistent movements are kept to a minimum.
Therefore the ore polygons can be adjusted for blast movement by

A total of 16 additional monitoring holes for each blast with two

Blast Movement Monitors BMMs and a 2m long section of

Copyright 2012 by SME

SME Annual Meeting

Feb. 19 - 22, 2012, Seattle, WA

2012) suggesting that the predictions are accurate and reliable for the
current conditions (Table 2 and Figure 9).

PVC pipe for each hole. Details of the BMM can be seen in La
Rosa et al., 2004.
Increased stemming heights by 0.5 m along the ore / waste
boundaries.
Flatter V initiation timing used with 5 ms inter-hole & 67 ms
inter-row (AP-1028-211P, Figure 6) and 11 ms inter-hole & 81
ms inter-row (AP-1028-210P, Figure 8).
Initiation point (and hence centreline) located in centre of high
grade block away from grade boundaries (Figure 6 & Figure 8).
The timing in the back few rows of each shot was slowed down
using 119 ms inter-row delays.
Greater focus and control on blast QA/QC was applied around
low/medium & waste grade boundaries.

Figure 6.
vectors.

AP-1028-211 measured horizontal blast displacement

Figure 4. Setup of validation trials; AP-1028-210P & AP-1028-211P.

Results from validation trials
Controlled blast movement. First validation trial blast (Figure 5)
had a partial free face (top 4 m of the bench) in the bottom right-hand
corner resulting in edge effects during blasting. As the material in this
section of the blast was all high grade ore, its effect on blast-induced
ore loss and dilution was minimal. In this blast 7 to 8 holes cratered out
of about 450 blast holes. Cratering along the ore / waste boundaries
was observed even with the increased stemming (by 0.5 m). Reasons
for this cratering may be due to:

Figure 7. Validation blast AP-1028-210.

Inconsistent stemming due to manual shovelling of stemming

material because of stemming loader breaking down during the
trial.
Stemming material being wetter than usual.
The occurrence of wet holes particularly along the high wall
side of the blast.

Figure 5. Validation blast AP-1028-211.

Horizontal displacement, pre-blast ore boundaries (bottom flitch)
and timing contours for trial blast AP-1028-211, are shown in Figure 6.
Second validation trial blast (Figure 7) also had a partial free face
resulting in edge effects during blasting. This excessive movement at
the front of the blast is unlikely to have had an adverse effect on blast
induced ore loss and dilution, as it is predominately high grade material
being blasted onto other high grade material from the previous blast. In
terms of cratering, only 4 or 5 holes cratered (out of about 500 blast
holes) and most of these were confined to the back few rows of the
blast. This could have been the result of water in these holes or slight
over-confinement towards the back of the blast.

Figure 8.
vectors.

AP-1028-210 measured horizontal blast displacement

Table 2. Summary of measured horizontal movement compared to

predicted
Average
Validation
Validation
Predicted from
horizontal
Results APResults APbenchmarking trials
movement
1028-211
1028-210
Surface
4m
4.0 m
3.5 m
Top flitch
8m
9.0 m
7.3 m
Bottom flitch
12 m
12.9 m
11.1 m

Horizontal displacement, pre-blast ore boundaries (bottom flitch)

and timing contours for trial blast AP-1028-210, are shown in Figure 8.
Magnitude and direction of blast movement. Blast movement
measurements from the validation blasts are comparable with the
predictions from the benchmark blasts (Rogers and Kanchibotla et al.,

Copyright 2012 by SME

SME Annual Meeting

Feb. 19 - 22, 2012, Seattle, WA
polygon adjustments using average movement can be a reliable
method of accounting for blast movement for the current geological
and blasting conditions at Ahafo Gold mine.

Figure 9. Histogram of horizontal displacement for validation blasts.

The validation blasts are split into two zones either side of the
initiation point and the bearing perpendicular to the timing contours is
calculated for both sides (Figure 10) and compared against the
average displacement directions measured from the BMMs. A
summary of the results from this comparison is given in Table 3 and
indicates little difference (less than 1) between the measurements and
predictions.

Figure 11. Comparison of ore polygon adjustments applying average

movements and actual measurements.
CONCLUSIONS
The data analysis of bench mark trial blasts, modelling and
simulation suggest:

All dilution and ore loss occurs around the grade boundaries
hence, blasts should be designed to ensure movement is
consistent and parallel to the strike of the ore body.
Uncontrolled and inconsistent movements in a blast make it
difficult to control ore loss and dilution hence should be
avoided.
Flatter V initiation with the centre line away from the ore/waste
contacts can enhance consistent movement within the blast.
Consistent free face conditions and better confinement along
the ore waste boundaries can minimise inconsistent movement
within the blast.

The validation trial results demonstrated that ore loss and dilution
can be reduced by:

Implementing blast designs to promote movement along the

strike of the ore body and to minimise inconsistent movements
along the ore/waste boundaries
Adjusting the ore polygons for blast movement using the
average movement vectors measured.

Ahafo operations now adjust ore polygons to account for blast

movements. The alternative strategies developed from this study have
been incorporated in the standard site operating procedures.A
limitation identified in areas where the flat V centre line does not fall
within an entire ore or waste zone, is the design of ore blocks below
smallest mining unit (SMU). In such cases the centre line acts as a
boundary resulting in the design of smaller ore blocks below SMU. This
can be mitigated to some extent by overlooking the strict use of SMU
along the boundaries.

Figure 10. Direction of blast displacement in validation blast AP1028211P.

Table 3. Summary of measured direction of blast displacement
compared to predicted.
AP-1028-211P
AP-1028-210P
Direction of
Validation
Validation
Displacement
Predicted
Predicted
Results
Results
Zone 1
36
37
40
40
Zone 2
30
32
28
27

The blast movement study was carried out in 1 out of the 4

operating pits and therefore makes it difficult to quantify the direct
benefits as it is not practical to track material from this pit alone
through the mill. Site management agrees that there is definitely some
benefit from this study.

Comparison of ore polygon adjustments using average

movement vectors and the actual measurements
The final validation is comparison of ore polygons adjusted using
the average movement vectors against the adjustment from the direct
measurements, using the 2DMove blast movement model, for each
blast. Details of the 2DMove model can be seen in Tordoir, (2009) and
Rogers and Kanchibotla et al., (2012). Figure 11 shows the
comparison for the top flitch of blast AP1028-211 and the bottom flitch
of blast AP1028-210. The results show little difference (e.g. less than a
bucket width) between the adjusted ore polygons using the
measurements and by applying the average movement vectors. Based
on these results, it can be concluded that the recommended ore

Further tests are planned for the other pits and the expectation is
that benefits would show up over a long term (1 year) through
reconciliation reporting.
ACKNOWLEDGEMENTS
All members of staff at Ahafo, Newmont Ghana Gold, are
acknowledged for their assistance and support during the blast
movement monitoring process.

Copyright 2012 by SME

SME Annual Meeting

Feb. 19 - 22, 2012, Seattle, WA
REFERENCES
1.

Firth, I.R., Mousset-Jones, P. and Daemen, J. (2002), Blast

movement measurement for grade control, Proceedings of the
28th Annual Conference on Explosives and Blasting Technique.
Annual Conference on Explosives and Blasting Technique.
International Society of Explosives Engineers, Las Vegas,
Nevada, USA, pp. 55-68.

2.

Harris, G.W., Mousset-Jones, P. and Daemen, J. (2001),Blast

movement measurement to control dilution in surface mines CIM
Bulletin, 94(1047): pp. 52-55.

3.

La Rosa, D., Thornton, D., and Wortley, M. (2004), Blast

movement monitor and method for determining the movement of
a blast..US Patent number 7367269, Brisbane, Australia.

4.

Rogers, W., Kanchibotla, S., Tordoir. A., Ako, S., Engmann, E.,
Bisiaux, B. (2012), Understanding blast movement and its
impacts on grade control at Ahafo gold mine in Ghana Accepted
in 38th Annual Conference on Explosives and Blasting
Technique. International Society of Explosive Engineers,
Nashville, TN USA.

5.

Taylor, S.L., Gilbride, L.J., Daemen, J.J.K. and Mousset-Jones, P.

(1996), The impact of blast induced movement on grade dilution
in Nevada's precious metal mines In: B. Mohanty (Editor),
Proceedings of the Fifth International Symposium on Rock
Fragmentation by Blasting Fragblast-5. A.A. Balkema, Montreal,
Canada, pp. 407-413.

6.

Thornton, D., Sprott, D. and Brunton, I. (2005), Measuring blast

movement to reduce ore loss and dilution, Proceedings of the
Thirty-First Annual Conference on Explosives and Blasting
Technique. International Society of Explosives Engineers,
Orlando, FL, USA, pp. 189-200.

7.

Tordoir, A.E. (2009), 2DMove Blast Displacement Model

preliminary users manual, University of Queensland, Brisbane,
Australia.

Copyright 2012 by SME