Simon Collins, TSG Consulting, Good Models are More Than Screen Deep: The Case of Newmont Mining’s Boddington Gold expansion

“Good models are more than screen deep” Newmont Boddington Gold Case Study

Bulk Materials Handling 2014 conference 01-05

30th April 2014

© 2012 TSG Consulting All Rights Reserved Simulation | Optimisation | Strategic Planning

1 •  CAPEX Optimsation

2 •  Plant commissioning

performance evaluation

3 •  Operating &

maintenance optimisation

Three recent & succesful NBG studies

Developing confidence in the NBG DES

model

Site knowledge NBG

Model integrity TSG

Cross functional

study team

System response

System insights

Iterative learning

Introduction

Page 2 INT-MKT-PRE-BMH2014ConferencePresentation-01-05.pptx


TSG and NBG | study timeline

The following represents the different studies that have been conducted over the past 14 years at the start date of the study.


Timeline 2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14

Preliminary study carried

out to evaluate initial flow

sheet. Plant only model

Model updated to represent

new flow sheet

Model expanded to

include mining to undertake

trucking evaluation

Mine interface study

Contribution to financial modeling

Model update and additional

scenario analysis

Expanded works study

Optimization study

Mine plant interface study

with revised equipment and

operating strategy

Bottleneck analysis, upgrade

options and variability evaluation

Operations modeling


Model animations



Detailed truck and shovel model implemented during the latest studies Enabled understanding of crusher utilisation and variability in supply of material – proper understanding of coarse ore stockpile sizing and the impact on system productivity

NBG mining operation | animation



NBG mining operation | animation



Process model was updated from previous work – major capital options testing required the largest model updates Significant effort to insert detailed bin level controls, up-to-date failure data, smart screen splits, conveyor capacity limits

NBG processing plant | animation



NBG processing plant | animation



Summary of results



CAPEX Optimisation | summary

Option 1: Minor CAPEX process improvement options Option 2: Large CAPEX possibilities allowing higher mill production rates Iterative process to arrive at a final set of cases – increasing levels of system response insights and new improvement ideas


Process improvements capital options at varying mill rate assumptions


Plant commissioning performance evaluation | summary

Detailed mining model development enabled rehandle approaches to be tested Confirmation and optimisation of truck and shovel strategies supplying the processing facility


Mining truck and shovels optimisation Effect of equipment upgrades with and without

an operating mine


Operating & maintenance optimisation | summary

Testing of maintenance strategies to minimise system disturbances and maximise production More than 15 maintenance schedules tested Minor upgrade timing effects


Planned maintenance scenarios testing Timing of equipment upgrades


NBG expertise

Significant expansion project management and site management experience. This is a key element to a successful study outcome

Data extraction and preparation: NBG team members excelled during this phase of the project and enabled model validation Documentation trail for internal critique

Understanding of the decision making processes whether they be automated or manual decisions. One team member had significant control engineering experience on site

Understanding how the supply chain should respond to change. Very knowledgeable team members with substantial operations experience. This enabled detailed validation work

TSG expertise

TSG has conducted more than 500 supply chain studies in almost 3 decades and is the company’s core business. Successful studies require a high level of attention to detail for both modelling and reporting of pertinent insights

Optimal input of data into the DES model and highlighting areas where data quality is key to a successful outcome Data and logic document implemented for the model

Iterative and collaborative work with the NBG team to ensure model functionality and model output tools were accurately reflecting the real system

Assisting with the development and implementation of sensitivity cases . Displaying meaningful outputs to allow system level assessment

DES model performance | developing confidence

Collaborative effort to arrive at a quality study outcome



NBG assigned significant resourcing to validate model against known plant performance Detailed checking of sensitivity case model outputs built confidence in expansion option case results

DES model performance | example outputs


•  System component KPI’s •  Annual production •  Equipment states •  Average equipment rates or truck cycle times

•  Equipment output time series & distributions

Conveyor rate limit

3, 2, 1 & 0 units offline Major planned outage

No feed to conveyor 03

Average ball mill performance for a “1-year” model run


•  System response testing •  System productivity with increasing mill rates

(see right) •  System productivity with increasing truck

numbers •  System productivity with increasing COS size

•  Highlight point at which upgrades are no longer cost effective •  For the system as defined •  Other model changes could shift the point of

cost effectiveness •  Combinations of upgrades are often effective •  Use of detailed outputs to assess where

bottleneck predominates – then propose changes to equipment upgrades or control strategies

DES model performance | system response

System response evaluation provides another layer of confidence that model is performing as expected “Good models are more than screen deep”


32

33

34

35

36

37

38

Mean Mill Rate -‐300



Mean Mill Rate Mean Mill Rate + 100

Mean Mill Rate + 200




Mill Rate Sensitivity

Mtpa

Expected diminishing system return from increasing mill rate


-1 -0.5 0 0.5 1 -1

0

133

34

35

36

37

38

Thro

ughp

ut (M

tpa)

•  DES modelling a powerful tool to understand supply chains

•  NBG and TSG efforts resulted in a robust supply chain model •  Provides confidence for capital upgrade

decision making •  Directly supplies financial analysis (IRR

determination)

•  Study process generates improvement ideas •  Facilitates a system level understanding •  Can highlight shortfalls in data collection

systems •  Can highlight areas where higher system level

control strategies can be improved

Conclusions | “Good models are more than screen deep”


Design of Experiments: efficient system response understanding


Questions

The end



Appendix | DES general description



Dynamic Modelling •  Predicts how a system performance will vary with time. •  1 minute time steps typically (525,600 minutes = 1 year)

Monte Carlo Simulation (random numbers) •  Variability is incorporated in events such as breakdowns

and failures, grade fluctuations, load sizes. •  Combining the randomness of each operation through

modelling enables us to quantify the expected variation of the physical system and to predict performance with a level of confidence.

Discrete Entities and Processes •  Each physical item (truck, mill, pump, ship) is modelled

as a discrete element with its own uniquely defined set of properties or attributes (speed, operating rate, reliability, carrying capacity).

•  Rules are defined to describe the interaction of the entities in the system.

•  The models are animated.

Discrete event simulation (DES) | general description

DES promotes system understanding and is inclusive of system input variability – results fall on a distribution DES has three defining elements:

Sub Process

Sub Process

Sub Process

Sub Process

Sub Process

Data

Model Boundary

Sub Process

Sub Process

Queue

Surge Pile

Delay

?

Operational Decisions

Dependent Processes

Logic

Understanding Inputs

Feedback

Variation

Performance

Simon Collins, TSG Consulting, Good Models are More Than Screen Deep: The Case of Newmont Mining’s Boddington Gold expansion

Engineering

rights reserved

good models

mill rate

equipment

maintenance

model

system

defined