ADVANCES IN DRILL RIG DEPLOYED RADARS Mr Tim Sindle, ARCO/CRC Mining Imaging Lab, The University of Sydney Dr Carina Kemp, Business Development Manager, GEOMOLE 11 th SAGA Biennial Conference and Exhibition, 16-18 Septemb 1 SAGA, September 2009
Feb 25, 2016
SAGA, September 2009 1
ADVANCES IN DRILL RIG DEPLOYED RADARS
Mr Tim Sindle, ARCO/CRC Mining Imaging Lab, The University of SydneyDr Carina Kemp, Business Development Manager, GEOMOLE
11th SAGA Biennial Conference and Exhibition, 16-18 September 2009
SAGA, September 2009
2
Outline Introduction Method and Results
Survey Gear Minimisation Analyzing Drill Deployed Data Automatic Algorithm Development
Conclusions
3
SAGA, September 2009
Introduction – In-mine geophysics
Anticipate problems ahead of mining
Improve efficiency of mining operations
Bulky gear Time consuming
surveys cause delays in production
The Good The Bad
No matter how good the results, if any technique cannot be easily and reliably implemented in the mining environment, it will not be used mainstream.
4
SAGA, September 2009
Introduction – Borehole Radar (BHR)? Ground penetrating radar
(GPR) in a drillhole Reflections indicate a
contrast in the electrical properties of the rock.
BHR provides high detailed continuous reflections from lithology contacts and structures.
GeoMole BHR 10 – 124 MHz Bandwidth Resolution: less than1m Range: up to 50m or more
(depending on rock type)
Probe diameter: 32 mm BHR Profiling at ~10 m/min
Omnidirectional antenna
5
SAGA, September 2009
BHR then…. Survey trials of
BHR showed very promising results, but the gear let us down. 50 kg optical fibre
winch 20 kg push rods 10 kg probes
6
SAGA, September 2009
BHR Now - Minimal Gear Radar Tool
1.6m 3kg
Non-conductive spacers 1.5m 2kg each
Drill attachment PDA
Radar SpacersDrill
Attachment
+IQ
+PDA
SAGA, September 2009
7
BHR Now – Drill rig deployed
IQCore barrel
drill bit
spacers
Drill Rig Deployed Borehole Radar- Pumpdown Radar Tool
Spacers
The radar tool continuously records data.
The motion of the rods is discontinuous as the rods pulled and removed.
Depth Depth
Mea
sure
men
t (s
tati
on)
Mea
sure
men
t (S
tati
onar
y)Winch Survey OTR Survey
Stationary
Moving
Moving
Deployment Motion…
Winch Survey OTR Survey
Same 40m section of a horizontal borehole
Radar Data…
StationaryMoving
11
SAGA, September 2009
Raw Data Aim:
To understand the motion in order to work out how to recompress it.
Different motion for each type of drill-rig
Boart LM75 Diamond
Raw Data
Recompressed Data
Recompressing Radar Data..
13
SAGA, September 2009
Movement Log Logging procedure
tracks accurately the motion of the drill rig.
User records ‘MOVE’, ‘STOP’ and ‘ROD-CHANGE’ following the motion of the drill.
These events are time stamped and recorded for data processing
S SS MM MMSM
Rada
r Dat
a
Trace Number
Acce
lero
met
er D
ataR R
Ampl
itude
Accelerometers were installed in the radars to assist with movement logging
14
SAGA, September 2009
Time Log processed Data Vulnerable
to human error
SAGA, September 2009
15
Automatic Algorithm DevelopmentUsing the accelerometer data for
automatic processing: Statistical deviation measurement Fourier Spectrum Analysis Velocity integration calculations
SAGA, September 2009
16
Statistical Processed Data
700 705 710 715 720 725 730 735 740 745-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2Am
plitu
de
Traces
Standard DeviationThreshold
17
SAGA, September 2009
Accelerometer Processed Data
Suffers from random accelerometer events
18
SAGA, September 2009
Fourier Spectrum Analysis Examine the power in
various regions of motion
Difference observed between some moving and stopped traces by examining the higher frequency content.
However, drill vibrations cause wide band energy gains.
METHOD ABANDONED 0 5 10 15 20 25 30 35 400
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04Single-Sided Amplitude Spectrum of y(t)
Frequency (Hz)
Am
plitu
de
*Stopped with drill shockStart of moveStopped Constant velocity move
Frequency Spectrum
19
SAGA, September 2009
Velocity Processed Data Noisy environment causes
spurious accelerations and accurate velocity is hard to gather.
A high pass filter distributes the velocities aroundzero.
Then the mean representation of the velocity is calculated
700 705 710 715 720 725 730 735 740 745 750
-1
-0.5
0
0.5
1
Trace
Ampl
itude
velocityonaccelerati
Positive = Moving, Negative = Stopped
20
SAGA, September 2009
Velocity Processed Data Copes well with
the sharp drill shocks and vibrations as they often have equal positive and negative direction.
Captures the start and stop of the movement well.
Particularly violent jerks can cause a trace to be lost.
SAGA, September 2009
21
Comparison…
Raw DataTime LogAccelerometerVelocity
22
SAGA, September 2009
Conclusions… Drill deployed radars
can be run with minimal disruption to normal work flow.
Using the time log alone can be vulnerable to human error
Yet all automated methods investigated so far are vulnerable to sharp spurious drill movements.
A combination of a time log together with statistical and velocity methods will result in smooth “winch quality” images being produced.
Development in this
area continues
SAGA, September 2009
23
Conclusions The ultimate aim of a tool knowing its own
position automatically is theoretically possible, but only within well defined constraints, and there will always be the unknown events on the drill rig that can cause inaccuracies.
The above progress makes it possible for quick data turnaround from survey to seamless integration of BHR data into mine planning packages, to enable day to day mining decisions to be made using such tools.
SAGA, September 2009
24
Acknowledgements… The authors would like to thank DeBeers
Canada in particular Kevin Smith, for their ongoing feedback and use of the tool.
The funding contributions of ARCO, CRC Mining, and GeoMole are gratefully acknowledged.
Many thanks to the tireless work by Sydney University ARCO Lab members including; Andrew Bray, Steven Owens, and Phillip Manning.