CVRD Inco

EXPLORATION

SAMPLING CHALLENGES
IN
LATERITES
PT Inco Laterite Meeting
April 2007

Waheed Ahmad
Revised: March 22, 2007
CVRD Inco
EXPLORATION OBJECTIVE OF SAMPLING

• The objective of sampling is to reproduce the same

statistical parameters within the sampled population that
exist within the target population
§ Mean, Mode, Median,
§ Standard Deviation,
§ Skewness, Kurtosis, and
§ correlations among various components

• The sampled population can only duplicate the statistical

parameters of the target population if the samples are
truly representative of the target population
CVRD Inco
EXPLORATION TYPE OF SAMPLING IN LATERITES

• In situ sampling of undisturbed laterite:

§ Drilling: Core, Percussion drilling, augering
§ Test pits: channel sampling; sampling of excavated material
§ Trenches: channel sampling; sampling of excavated material

• Sampling of broken ore:

§ Mine face sampling (for grade control)
§ Stockpile sampling
§ Truck sampling
§ Conveyor belt sampling
CVRD Inco
EXPLORATION REQUIREMENTS OF LATERITE SAMPLING

• IN SITU SAMPLING (Drilling, Test pitting, Trenching)

1. Sample should be representative of the area drilled
2. The device should sample the entire laterite profile
3. The entire drilled or pitted sample should be recovered
4. Sample integrity should be maintained
5. Sample reductions should be carried out appropriately

• SAMPLING OF BROKEN ORE (Stockpile, belt, etc.)

1. Sample should be representative of the material
(in assays, mineralogy, fines/coarse contents, etc)
2. Sample reductions should be carried out appropriately
CVRD Inco INFORMATION SOUGHT IN LATERITE
EXPLORATION
SAMPLING

• Assay values:
§ Metal contents: Ni, Co, Fe
§ Chemistry: Fe, Si, Mg, Al, Cr, Mn, Ca, LOI (Loss-on-Ignition)

• Ore mineralogy
• Bulk Density
• Moisture content
• Upgrading characteristics:
§ Dry screening
§ Wet screening

• Screen characterisation
• Slurry rheology
CVRD Inco
EXPLORATION SAMPLING TOOLS FOR LATERITES

• DRILLING
§ Hand Auger
§ Light-weight Winkie
§ Power Auger
§ Percussion / RC
§ Core (HQ, PQ, 20-cm diameter)
• TEST PITTING
§ 1.25 X 1.25 metre
• CHANNEL SAMPLING
§ Backhoe trenches, test pits, mine face, mine bench
• SAMPLING OF BROKEN MATERIAL (Bulk sampling)
§ Stockpile sampling
§ Sampling for upgrading tests
CVRD Inco
EXPLORATION CHALLENGES POSED

• DRILLING
1. High lateral variability in laterite formations
2. Penetration of the entire laterite profile
3. Laterites made of mixed hard and soft materials
4. Nickel values often occur in fines fraction prone to washing
5. Limonite and soft laterite subject to extrusion

• TEST PITTING
1. Limitation on depth penetration due to safety issues and water
logging

• SAMPLING OF BROKEN MATERIAL

1. Segregation due to mechanical handling
2. Extreme variation in particle size (mixed rock & clayey material)
3. Sample reduction issues
CVRD Inco1. ISSUE OF EXTREME LATERAL VARIABILITY
EXPLORATION

Red Laterite

Limonite

Saprolite

Bedrock pinnacle
CVRD Inco
EXPLORATION Sample Representivity

100m 1m
100m

HQ core size: 63.5mm diameter

1m Core sample: 0.003167 cu m.
HQ 100x100x1m volume: 10,000 cu m
core hole
Representation ratio: 1 : 3.2 million

To improve sample representivity:

(1) Increase the size of sample per metre
(2) Increase the sampling density
CVRD Inco Sample volume as a ratio of total volume
EXPLORATION

PROVEN PROBABLE INF.

Sample 6.25 12.5 25 50 100 150 450
Spacing M M M M M M M
NQ 22 88 351 1.4 5.6 12.6 114
47.6mm K K K M M M M
HQ 12 49 197 789 3.2 7.1 64
63.5mm K K K K M M M
PQ 7 28 110 441 1.8 4.0 35.7
85.0mm K K K K M M M
LD [ERTK] 1 5 20 80 318 716 6.4
200mm K K K K K K M
TRIV 50 199 796 3 13 28.6 258
1000mm K K K K
Test Pit 21 83 333 1 5 12 108
1.25x1.5m K K K K
CVRD Inco
EXPLORATION 2. Issue of penetrability of laterite profile

• Light-weight Hand Auger and Winkie drills have very

limited penetration (lack of power)
• Power Auger drills still fail to penetrate hard saprolite in
unserpentinised peridotite terrain (risk of breaking the
rods)
• Test pits generally fail to reach bottom due to safety
issues and water logging problems
• Percussion drills can generally reach the bottom except
in very hard dunite
• Core drills can penetrate the entire laterite profile
CVRD Inco
EXPLORATION Sorowako West Block – high grade zone
CVRD Inco
EXPLORATION 3. Issue of mixed rocky and soft material

• Auger sampling devices cannot break and sample the

rock
§ Auger samples give higher Fe, Ni, Co values

§ Auger samples give lower SiO2, MgO values

• Percussion sampling pushes the floating boulders aside

and over-samples the fines

• Core sampling is better but still not perfect

§ Frequent core loss due to mixed hard and soft profile
CVRD Inco
EXPLORATION Sorowako West Block mining profile
CVRD Inco
EXPLORATION 4. Issue of fines losses during core drilling

• Garnieritic fines and limonite are subject to washing during high

water pressure
• Difficult to maintain appropriate water pressure when drilling mixed
rocky and fines material
• Examples from Sorowako West Block test block studies:

Hasan Test Block Koro S. Test Block Watulabu South

Core Actual Core Actual Core Actual
Ni 2.28 2.47 1.95 2.21 1.82 2.42
Fe 24.4 22.2 30.8 20.3 23.9 19.4
SiO2 30.9 33.1 26.6 38.5 34.9 40.0
MgO 12.3 12.2 9.0 13.0 12.3 13.3

• Core drilling loses Ni and SiO2 and gains Fe

CVRD Inco
EXPLORATION
CVRD Inco
EXPLORATION 5. Issue of core extrusion
• Core recovery in laterites can be:
§ < 100% due to core loss
§ > 100% due to extrusion
§ A combination of the above two that masks true core recovery

• Two possible explanations (models) for

>100 core recovery:
§ Sample mass is correct; sample volume is increased (shaking)
• Bulk density = sample weight / drilled volume
§ Sample mass is increased; sample volume is increased
• Bulk density = sample weight / sample volume
§ PT Inco uses the second model assuming that extraneous
material has been pushed into the core barrel

• Extrusion occurs in limonite and soft saprolite

• Extrusion also occurs in Power Auger sampling
CVRD Inco Extrusion in Limonite
EXPLORATION
Konde Hill in West Block
CVRD Inco
EXPLORATION 6. Issues in test pitting

• Test pits cannot penetrate the entire laterite profile

§ Due to pit safety issues
§ Due to water logging
§ Due to hard bedrock

• Test pits return excellent sample in the interval sampled

• However, they under-represent the hard, rocky portion of
the saprolite
• Test pit data yields higher Fe and fines recoveries and
lower SiO2 and MgO values. Ni content is not affected
seriously (in most cases)
CVRD Inco
EXPLORATION ISSUE OF SAMPLE INTEGRITY

• Only core drilling with full core recovery ensures

physical integrity of the sample
• Auger sampling does not recover rocky fragments and
mixes the fines
• Percussion or RC drilling breaks up the rocky fragments.
These samples cannot be used for:
§ Screen recovery tests or upgrading characteristics

§ RQD estimates

§ Other structural information

CVRD Inco FACTORS TO CONSIDER DURING
EXPLORATION
LATERITE SAMPLING
• Pattern of sampling:
§ Systematic grid: square; rectangular; diamond-shaped
§ Random: contour-based; random pattern
• Method of sampling:
§ Boreholes: Core drilling, Auger drilling, RC drilling
§ Test Pitting: channel sampling; sampling entire excavated material
§ Trenching: channel sampling; sampling excavated material
• Size of sample:
§ Small vs. large [Core size, Test Pit size, Trench size]
• Sampling Density (Frequency of sampling)
§ 400m; 200m; 100m; 50m; 25m; 12.5m
• Objective of sampling:
§ Reconnaissance level — Augers; 400-500m centres [Greenfield]
§ Resource definition — Coring; 200m / 100m centres [Pre-Feasibility]
§ Reserve definition — Coring; 100m / 50m centres [Feasibility; Production]
CVRD Inco
EXPLORATION Various sampling patterns

Square Grid Rectangular Grid Low variability

Variability
High
200m
1m
14

Staggered grid
diamond pattern

Random locations
following contour lines
CVRD Inco FACTORS TO CONSIDER DURING
EXPLORATION
LATERITE SAMPLING
• Pattern of sampling:
§ Systematic grid: square; rectangular; diamond-shaped
§ Random: contour-based; random pattern
• Method of sampling:
§ Boreholes: Core drilling, Auger drilling, RC drilling
§ Test Pitting: channel sampling; sampling entire excavated material
§ Trenching: channel sampling; sampling excavated material
• Size of sample:
§ Small vs. large [Core size, Test Pit size, Trench size]
• Sampling Density (Frequency of sampling)
§ 400m; 200m; 100m; 50m; 25m; 12.5m
• Objective of sampling:
§ Reconnaissance level — Augers; 400-500m centres [Greenfield]
§ Resource definition — Coring; 200m / 100m centres [Pre-Feasibility]
§ Reserve definition — Coring; 100m / 50m centres [Feasibility; Production]
CVRD Inco
EXPLORATION
SAMPLING TOOLS USED AT PT Inco

H.Auger (WB)
Winkie (WB)
High
H.Auger (EB)
Sampling Bias

Winkie (EB)
P.Auger (WB)
B. Hammer
Medium (WB)
P.Auger (EB)

Core hole (NQ)

Low Core hole (PQ)
T.Pit (EB) Core hole (LD)

Low Medium High

Depth of Laterite Penetration

CVRD Inco FACTORS TO CONSIDER DURING
EXPLORATION
LATERITE SAMPLING
• Pattern of sampling:
§ Systematic grid: square; rectangular; diamond-shaped
§ Random: contour-based; random pattern
• Method of sampling:
§ Boreholes: Core drilling, Auger drilling, RC drilling
§ Test Pitting: channel sampling; sampling entire excavated material
§ Trenching: channel sampling; sampling excavated material
• Size of sample:
§ Small vs. large [Core size, Test Pit size, Trench size]
• Sampling Density (Frequency of sampling)
§ 400m; 200m; 100m; 50m; 25m; 12.5m
• Objective of sampling:
§ Reconnaissance level — Augers; 400-500m centres [Greenfield]
§ Resource definition — Coring; 200m / 100m centres [Pre-Feasibility]
§ Reserve definition — Coring; 100m / 50m centres [Feasibility; Production]
CVRD Inco Representivity of sample as a function of
EXPLORATION
Sample size and sample spacing
100,000,000
10,000,000
NQ
1,000,000
R a tio d rille d to c o re v o lu m e

100,000
ERTK Trivelsonda
10,000
T. Pit
1,000
100
10 NQ HQ PQ
1 ERTK Trivelsonda Test Pit
0
0 25 50 75 100 125 150 175 200
DRILL SPACING, Metres
CVRD Inco ISSUES IN SAMPLING OF BROKEN
EXPLORATION
MATERIAL
• Issue of material segregation
§ Limonite & saprolite tend to segregate due to mining sequence
§ High and low grades tend to segregate due to mining sequence

• Mixture of large boulders and clayey fines makes proper

sampling very challenging
• JIS consider the following for Bulk Sampling:
§ Size of the Lot
§ Size and number of sub-lots
§ Maximum particle size (95% passing)
§ Dimensions of each increment
§ Number of sampling increments
§ Intralaminar variation (standard deviation)
CVRD Inco
EXPLORATION JAPANESE INDUSTRIAL STANDARDS

LOT Min. 10,000 tonnes

2–4
SUB-LOTS
Min. 5,000 t

JIS-recommended
Sampling Scoop a

d b
e
c
CVRD Inco IMPOSITION OF
EXPLORATION
CONDITIONAL PROBABILITY

• Any condition imposed during sampling leads to

probabilistic errors due to variance between estimation
and reality
§ Imposition of a nickel cut-off grade for plant feed

§ Imposition of Silica to Magnesia ratio in pyrometallurgy

§ Imposition of Mg ceiling in HPAL processing

§ Imposition of a screen size for waste/ore separation

CVRD Inco VARIANCE BETWEEN
EXPLORATION
ESTIMATION AND REALITY

1.5% Ni
Estimated: <1.5% Estimated: >1.5%
True Nickel Grade

Actual: >1.5% Actual: >1.5%

Not mined / Loss Mined as “ore”

1.5% Ni

Estimated: >1.5%
Actual: <1.5%
Mining Dilution

Estim.: <1.5%
Actual: <1.5%
Mined as “Waste” Estimated Nickel Grade
CVRD Inco
EXPLORATION CONCLUSIONS

• Laterites are very heterogeneous over short distances

(in elemental values, size fractions, mineralogy, etc.)

• Mixed rock and clay profile presents additional sampling

challenges

• Their relatively soft nature and easy penetrability leads

many persons to use simple but inadequate sampling
devices: auger drilling, test pitting, shovel sampling

• Sample reductions are generally carried out inadequately

• Frequently, sampling is not consistent with objectives

(generally too little; sometimes too much)