Pr

ac
t
i
c
alRec
onc
i
l
i
at
i
on

Resource and Reserve

Reconciliation

Objectives of today
Understand reconciliation key issues and
terminology
Consider the components of a reconciliation
system
Activity: Brainstorm the features of an ideal
system
Reconciliation systems analysis
Activity: Case study systems analysis and
design
Activity: Charting your system
Review of best practice in reconciliation

Desired outcomes
Understand the elements of a reconciliation system
Take raw production data and generate meaningful
reconciliation measures
Differentiate an ideal system from what is practically
achievable
Understand the shortcomings of the various data
sources and measurements
Use a set of tools to analyse data flow and design a
system
Analyse the results and know what corrective action
to take
Produce a process map of current data flow
Understand best practice in reconciliation

Reconciliation
- Areas for consideration

What do you want to measure?

Resource model performance?

R
Reserve
model
d l performance?
f
?
Mining performance?
The true quantity of metal/material produced?
Mill performance?
Equipment
q p
p
performance?
Other items?

What do you want to communicate?

Resource/Reserve performance to site and head
office management
Mining performance site and head office
manager
Mill performance
Production trends
Reconciliation
R
ili ti ttrends
d

Stakeholders
Geology Superintendent performance of resource
model, KPI
Senior Mine Geologist performance of grade
control model
model, KPI of self and staff
Senior Mine Planning Engineer success of mine
plan, KPI for self and staff
Mining Manager performance of mining operation,
possible KPI for contract
Mill Superintendent performance of plant,
optimisation of feed, KPI for self
Resident Manager performance against target and
budget, KPI
Head office Commercial Manager performance
against budget responsible to directors and
shareholders
Managing Director are we meeting production and
cashflow targets? responsible to shareholders

Other measures/information required?

In groups, brainstorm possible measures/KPIs from

the mine/mill cycle which are available through the
reconciliation process
Time: 5 minutes

Generic model for reconciliation

Areas to consider the resource

Are the resource and
the mining
g areas
coincident?
Can the resource model
be subdivided on mining
benches/mining areas?
Issues with recoverable
resources
Does the resource
include dilution?

Areas to consider reserve and planning

How has the reserve been
derived from the resource?
Factors vs real model

How has dilution and ore loss

been incorporated into the
reserve?
Does the reserve cover the
entire mining area?
Can the reserve be easily
depleted with mining shapes?
At what levels does planning
take place (short-term, mediumterm long-term)?
term,
long term)?
How good was the mining to the
plan?
What models are used for the
planning horizons?
Do they all need to be
reconciled?

Areas to consider the grade control model

Is the grade control model
separate to the resource/reserve?
Are the budget/forecast/target
figures derived from the reserve or
the grade control model?
Is it easy to interface mining
shapes with the grade control
model?
Does the grade control model
represent what is planned to be
mined?
How
H
often
ft iis th
the grade
d control
t l
model updated?
Is resource (exploration) data used
in the grade control model?
Is grade control data used in the
resource model?

Areas to consider mining

Ore loading method
Production recording

Dispatch systems
Truck counts
Load cells
Databases

Intermediate stockpiles/ore
passes
Skip weighing
Moisture!!
Variable density of material
Frequency of survey
Ore ramps in pit
As planned versus asmined

As-planned versus as-mined

Areas to consider stockpiles

Stockpile management
systems

At the mine
At the mill (mill pad)
After the crusher
Etc.

Interface with dispatch systems

Who manages the stockpiles?
eque cy of
o survey
su ey
Frequency
compaction etc.
Stockpile grade models
Last in first off (LIFO) - loaders
First in First off (FIFO) feeders
Average

Areas to consider processing

Who manages input to the
plant?
p
Post-crusher stockpiles
Sampling of post-crushed ore
Measurement techniques
Sampling of tails
Accuracy and format of
information from plant
mine mill
Integrated mine-mill
databases flexibility of
reporting
Timing of plant information
Ease of generation of
relevant figures

Activity designing the

ideal reconciliation

10

Basics of
Reconciliation

Designing reconciliation systems

Reconciliation defined
C
Causes
off poor reconciliation
ili ti
Capturing the physical data - tonnes and grade
Terms and definitions
Coping with multiple ore sources

11

What is reconciliation?
Dictionary term: The act of making
compatible or consistent
In a mining sense - the act of comparing
various measures of production to make them
consistent
Comparing a prediction (eg. a grade control
estimate) with a measurement (ie. the millf t i d tonnes
factorised
t
and
d grade)
d )
The mill is the final arbiter! (you cant argue
with gold bars)

Three sides of reconciliation

GRADE CONTROL - The best (hopefully)
measure of tonnes and grade delivered to the
mill.
ORE RESERVES/RESOURCES - Estimated
annually - used for amortisation, corporate
valuations, etc.
MILL PRODUCTION - The final arbiter of
tonnes and grade.

12

The essence of reconciliation

Ore Reserve
Performance

Mill
Performance

MINE
CLAIMED
mine 1

ORE
RESERVES

MINE
BREAK

U/G
STOCKPILE

MINE
SURFACE
STOCKPILE

MINE
CLAIMED
mine 2

PRODUCT

MILL
STOCKPILE

MILL
HEAD
TAILS

MINE
CLAIMED
mine 3

Mine
Performance

CRITICAL AREAS

Why reconcile?
A measure of performance - allows assessment
of the grade/tonnage performance of various
mines
Especially important when comparing disparate
production sources

A valuation of assets - Ore reserves are a

companies main asset - reconciliation allows
surety
t in
i their
th i value
l (or
( not!)
t!)
A measure of improvement - Both grade control
and ore reserves should be as close to reality as
possible

13

Four measures of reconciliation

1. Ore Reserve performance - measures
budgeting and forecasting ability
2. Mining performance - measures the degree
of optimisation of extraction of metal, within
certain (safety) constraints
3. Processing performance - measures the
efficiency of the plant via the metallurgical
recovery
4. Global performance - measures the total
conversion of resource metal to final
product

Causes of poor reconciliation

Geological model
Significant nugget effect
Sampling and sub-sampling errors in resource
drilling
Analytical errors
Poor estimation practice
Top cuts overly severe
Incorrect
I
t density
d
it allocation
ll
ti
Incorrect boundary definition

14

Causes of poor reconciliation

Grade control

Significant
S
g ca t nugget
ugget e
effect
ect
Sampling/subsampling, eg. blastholes vs RC
Analytical errors
Poor angle of intersection of drilling
Methodology issues, eg. polygonal, averaging vs kriging
Incorrect ore boundary (ie. cut-off grade too high)
Poor grade estimation practice
Block layout inaccuracies
Database errors or inadequate database
Human errors pressure of production

Causes of poor reconciliation

Open pit mining
Mining direction parallel to mineralisation
Heave and throw of mineralisation during
blasting
Hole layout inaccuracy
Truck dispatch errors
Selective loss of fines during blasting
Excessive
E
i dilution
dil ti d
during
i bl
blasting
ti
Mining through underground workings- ore
loss/unknown openings

15

Causes of poor reconciliation

Underground mining
Development off ore/ore left behind in development
Additional stripping not captured in pick up
Excessive dilution or hang-ups (ore-loss) during
longhole stoping
Unknown ore profile between levels dilution/ore
loss
Irregular ore outline insufficient drilling
Bogging
gg g of fill
Poor face sampling practice
Poor sampling from sludge or longhole drilling
Overdrawing during sublevel caving
Poor sample quality from draw points

Causes of poor reconciliation

Processing plant

Metal
eta hang-ups
a g ups within
t
p
plant
a t
Analytical inaccuracy
Overcompensation by process control system
Instability in mill processes
Poor knowledge of process cycles
Mill not optimised for blend of ore supplied
Infrequent/non-calibration of flowmeters and
weightometers
i ht
t
Poor subsampling at laboratory
Reconciliation carried out over too short a timeframe
Poor/insufficient sampling of residue or tails

16

Capturing physical data

We need the volume of ore mined (excavation
volume)
We need the grade of that volume (from grade
control sources)
We need the equivalent ore reserve volume and
grade

Excavation volumes - pits

Based upon pick-ups

problems with merging
g gp
pick-ups
p
Mayy be p
May be problems if mining on multiple benches
May be problems if in-pit stocks exist
May be problems with fired but not excavated
benches
Need to delineate as
as-mined
mined outlines

17

Excavation volumes - underground

Combination of Development and Production (Stoping)
Development
Based upon surveys
Surveys need to be timely!
Beware stripping

Production
Can be easily measured, eg. Room and Pillar mining
May be impossible to measure
measure, eg
eg. Block Caving
CMS pick-up is the ideal situation, but may rely on fired volumes

Rule is to subtract development tonnes from break (truck

counts) production tonnage

Grades of excavation volumes

There should be a grade control model separate
to the ore reserve model
As-mined
As mined blocks should be interfaced with the
[Link] to give a weighted grade
May be comparison problems with mill if mining
on successive benches simultaneously
Underground: combination of face, blast, and
sample data for grade control purposes
May be difficult to avoid re-use of data with ring
holes defining ore reserve model

18

Ore reserve depletion volumes & grades

Ideally a separate ore reserve model from the

grade control model
Generally straightforward in open pits
Problems underground:

Mining outside the ore reserve

Sub-grade material inside the ore reserve limits
Not mining to the full width of the ore reserve
Mining wider than the ore reserve
Depicting remnant ore

Terms and definitions (1)

BREAK tonnes and grade - material
physically fragmented in a pit or in an
underground mine during the production period
May not be removed to a stockpile or the
surface
HOIST tonnes and grade - dry, broken
material delivered out of the underground mine
or pit, to whatever destination
Break tonnes and grade adjusted for the
change in u/g or in-pit stockpiles

19

Terms and definitions (2)

DELIVERED tonnes and grade - Dry ore
delivered to the mill pad from the mine
The hoist tonnage adjusted for the change in
surface stocks at the mine
May be factored if load cells are on trucks

CLAIMED material - tonnes or grade based

upon estimates - surveys, truck counts
ACTUAL values - tonnes or grade which has
been factored to represent measured
tonnages or grades through the plant

Terms and definitions (3)

MILLED TONNES - amount of ore entering
the milling circuit after primary crushing
Generally a series of weightometer or weighbridge
measurements

ASSAY MILL HEAD GRADE - grade of

material treated by the plant before milling or
metal extraction
Either a blend or a figure from each mine source

CALCULATED MILL HEAD GRADE - the

th
grade of ore treated by the processing plant,
adjusted for the change in stocks
The sum of product plus tails

20

Terms and definitions (4)

THE MINE CALL FACTOR - The ratio of the
calculated head grade and the claimed
delivered grade

Takes no account of mill tails or mill recovery

May be factored by milled tonnes and delivered tonnes
May be allocated on a mine-by-mine basis
Generally allocated uniformly to all sources

ORE RESERVE DEPLETION - The tonnage

and grade removed from the ore reserve
model by interaction with the mining outlines
Tonnes from survey or CMS
Grade from ore reserve block model
Independent of grade control values

Dont forget stockpiles!

Stockpiles can exist

Underground
In the pit
On the surface
At the mill
After crushing

Each figure needs to be adjusted for the

change in stockpiles,
stockpiles ie
ie.
Start-of-Month stockpile + Material delivered Material removed = End-of-Month stockpile
e.g. 3000t +1500t - 2000t =2500t

21

Reconciliation definitions
Mine reconciliation:
Milled tonnes x Mill Delivered grade
Delivered tonnes x claimed grade
This is a METAL reconciliation
Also known as the MINE CALL FACTOR

Problems - multiple sources

Invariably a mixture of sources

Multiple open pits

Multiple sources within an open pit
Open pits and underground
Hard ore/soft ore
Oxidised ore/fresh ore

We need to assess the contribution of each

source to
t monitor
it its
it profitability
fit bilit
How?

22

Solving multiple source problem

Two solutions:
1. Batch or campaign treatment of different
ore sources
Cons:
Plant needs a blend of material for optimum
performance
It may take time to purge the circuit
Any campaign must be limited
2. Separate dedicated crushing and
sampling plant
Cons:
Expensive!

Essence of reconciliation
Ore Reserve
Performance

Mill
Performance

MINE
CLAIMED
mine 1

ORE
RESERVES

MINE
BREAK

U/G
STOCKPILE

MINE
SURFACE
STOCKPILE

MINE
CLAIMED
mine 2

PRODUCT

MILL
STOCKPILE

MILL
HEAD
TAILS

MINE
CLAIMED
mine 3

Mine
Performance

CRITICAL AREAS

23

Objectives of any system

24

Reconciliation
what can go wrong?

Case study
Open pit and underground gold mine

Narrow
a o mineralisation
e a sat o
High grade
Lithological control on mineralisation
Open pit and underground mining commenced prior
to start-up of treatment plant
Small open cut, larger underground mine
Development of levels followed by longhole open
stoping
Claimed grade by face sampling of development
Snowden called in to review mine geology and
reconciliation practices

25

Existing reconciliation system

Overly convoluted
Failed to recognise which inputs were measured and
which were calculated
Total mill reconciled total mine claimed (No
stockpiles)
Grade control claimed higher tonnes and grade
compared to mill perception of ore loss occurring
More tonnes milled at slightly lower grade compared
to reserve

After review
Tonnage milled (weightometer) was 9% higher than
depleted reserve tonnage
Reconciled metal (gold bars plus tailings) was 2.1%
greater than depleted reserve (-5% grade)
Face sampling in upper levels showed a slight
positive grade bias compared to mill reconciled
grades (grade control claimed overstating grades)
This overstatement increased substantially with depth
with move from supergene to primary mineralisation

26

Reconciliation findings
Before review:
Perception of ore loss occurring

After review:
Mine is either generating additional material at
lower grade (mining dilution or incorrect
selectivity) or resource model is wrong

Observations:
Development is 5% wider and higher than reserve
on average
Grade control is overstating grades, potentially
leading to mining more tonnes which are really
lower grade

Lessons
Distinguish measurements from calculations in
system
Keep the system as simple as possible
Review face sampling direction versus mineralisation
orientation
Always need to question reconciliation inputs:

Resource model
Reserve conversion factors
Grade control sampling issues
Even mill figures (how are they derived?)

27

28

Reconciliation systems analysis

Objectives
Learn techniques for mapping your current system
Apply
A l these
th
to
t the
th production
d ti off a process map for
f
a gold mine
Input the reconciliation data and generate key
reconciliation results
Analyse the results obtained

29

Designing a system

Define the various stages of the mining process

Consider the inputs and outputs for each stage
Consider the measurement points
Consider the reports or data input and those
generated
t d
Analyse the use and distribution of those
reports
Iterate and refine

Process mapping approaches

Free-form Mind mapping

St t d Fishbone
Structured
Fi hb
diagrams
di
More structured process mapping
Hybrid approaches

30

Mindmap gather your thoughts

Mindmap
identify the inputs and relationships
take
t k a central
t l th
them and
d id
identify
tif th
the iinputs
t
use one word per branch

Brainstorming fishbone diagrams

31

The root cause of the problem

Why did your car stop?
Because it ran out of gas

Why did it run out of gas?

Because I didn't buy any gas on my way to
work

Why didn't you buy any gas this morning?

Because I didn't have any money

Why didn't you have any money?

Because I lost it all last night in a poker game.

Stage 2b Problem cause and effect

Fishbone (Ishikawa) analysis

32

Why? Why?
There is more than one reason

Why? Why? Why? Why? Why?

Reasons behind the reasons

33

Process mapping example

Process Map
Start end with
circles (or
ellipses)
Processes in
boxes
Decisions in
diamonds

34

Some useful symbols

Model resource/reserve, grade
control, scheduling model, etc.
Document budget, forecast,
schedule, etc.
Database Physical database of
information (eg. database of dig
plans), dispatch system, survey
system,or metallurgical accounting
system
Process any generic or specific
process
Stockpile temporary or permanent
storage which is measured on a
regular basis

Suggested approach
1.
2.
3.
4.
5.
6.
7.
8.
9.

Identify the key stages or activities

List the keyy processes
p
within each activityy
Identify the inputs to and outputs from each activity
Identify KPIs or reconciliation metrics (current or
desired)
Identify stockpiles, ore movement, measurement
points
Discriminate tonnes, grade, density, moisture (other?)
measurements
If possible, identify responsibilities for each process
and KPI
Identify recipients of information
Iterate and refine!

35

Activity process mapping and reconciliation

36

Best practice
reconciliation

Areas of best practice

Underlying data structure

D t sources and
Data
d exchange
h
Data quality and reliability
Accessibility
Processing of models and generation of KPIs
Automation of the system
y
Output and analysis provided
Integration of systems

37

A customisable, controllable system

Equivalence
Reconciliation
System

Plant control
System

We are aiming to generate a system which can be

monitored, controlled, and is as responsive as a
plant control system

Data structure
A spreadsheet is not a database
Spreadsheets
Spreadsheets > 1Mb in size have > 80% chance
of including serious errors (Panko, 2000)
Underlying engine is a database, which should be
industrial-strength
SQL server, Oracle, DB2, etc.
Redundancy, backup, checkpoints, rollbacks
Links in and out of databases (ODBC, etc.)

38

One suggested structure

Data sources and exchange

As many data sources as possible should be linked:

Resource models
Reserve models
Budgets, forecasts, short-term plans
Dig outlines
Survey information
Truck Dispatch information
Weights
Destinations

Stockpile (mill pad) database

Analytical laboratory
Plant data, eg. weightometers, online analysis

39

Data quality and reliability

NO manual entry of data
Data
D t capture
t
is
i automated
t
t d ((eg. PLCs,
PLC lload
d cells,
ll
GPS)
Automated transfer and backup
Dual storage of source data (eg. as resource
model and in integrated database)
redundancy
d d
Business rules for data limits on-entry
validation and automatic flagging of errors

Accessibility
All stakeholders have access to data which is meaningful
to them
No more or less data than each stakeholder requires

Different levels of access depending upon user login

Super user access for system owner
Client-server database access
Web browser access from anywhere and by anyone
via corporate intranet
Allows remote support (eg. from head office)
Allows Managing Director access to key figures and key
output

40

Processing of models & generation of KPIs

All models stored in a database

Access
A
to
t models
d l via
i graphical
hi l query or b
by
interface with survey/block outlines
All major KPIs generated automatically (once data
has been validated)
Ability for ad hoc or one-off queries via a simple
interface

System automation
As much data as possible is gathered
automatically
Truck dispatch, load cells, weightometers, plant
control systems, SCADA systems
User intervention is to validate data and resolve
errors
Avoid tasks which rely on users initiation

41

Best practice in output

Present only as much as is needed!
No more, no less

Different access for different levels voyeur,

voyeur user
user,
administrator
One page management for senior execs
Ability to drill-down for line and middle management
Generate hard copy output at any time minimise
formal reporting
Results always available online
Historical data always available
Graphs and charts scalable over different timeframes
Minimum of text, maximum of graphical display
System status at a glance

Examples of graphical display

42

Example of graphical display

Example of output

43

Easy access to results and targets

Meaningful combinations of graphics and text

44

A useful tool the waterfall graph

Ore Mining - Comformance to plan - June 2004
160
140

P ro d u cctio n (kt)

120
100
80
60
40

Gains
Losses

20

Plan/Actual

0
Planned

Plan not Mined

Equipment
Availability

Reserve
Change

Plant
Availability

Operations
decision

Planning
Issues

Forward analysis

Analyse historical data to generate trends

S
Summarise
i KPI
KPIs and
d llook
k ffor cyclicity
li it
Moving averages
Apply trends to current data
Resource-reserve conversion
Truck factors
Weighbridge factors

45

Actual Mined

Systems integration
Reconciliation uses data from many independent
systems
DATA from these systems is integrated in a
reconciliation database via seamless transfer
Avoid the creation of a single system use best of
breed and replace components as required

46

Useful References
Fouet, T., Riske, R., Morley, C., Cook, A., Conti, D., & Centofanti, J. (2009).
Standardising the Reconciliation Factors Required in Governance
R
Reporting.
ti
S
Seventh
th IInternational
t
ti
l Mi
Mining
i G
Geology
l
C
Conference
f
P
Perth,
th WA
WA, 17
- 19 August 2009. pp 127-139.
Morley, C. (2003). Beyond Reconciliation A Proactive Approach to Using
Mining Data. Fifth Large Open Pit Mining Conference Kalgoorlie, WA, 3 - 5
November 2003. pp 185-191.

47