Why Mines Fail. Presented by: Richard J. Lambert, MBA, P.E., P.Eng. Seeking Better Mining Outcomes in the Next Cycle.

Transcription

1 Why Mines Fail Seeking Better Mining Outcomes in the Next Cycle Presented by: Richard J. Lambert, MBA, P.E., P.Eng. March 6, 2017

2 About RPA Advice to the mining industry for 30 years The specialty firm of choice for resource and reserve work All stages: exploration and resource evaluation scoping, prefeasibility and feasibility studies financing, permitting, construction, operation, closure and rehabilitation. Clients are financial institutions, governments, major mining companies, exploration and development firms, law firms, individual investors, and private equity ventures Offices in Canada, United States and United Kingdom Head office in Toronto 100% employee owned 2

3 Outline Why Do Mines Fail? Common Threads Fatal Flaws Non-Fatal Flaws that can be overcome 3

4 Why do Mines Fail? Poor Management Built to be more than it really was Mineral Resource not there Prices (up or down) Level of Study (moving too fast) Insufficient funds 4

5 Failed/Troubled Projects Since 1985 RPA review identified 75 projects that failed to meet expectations in a material way. 35 Grade Related 10 Mining Method 16 Processing 8 Capital Cost 7 Operating Cost 6 Social/Permitting/Government 5 Prices/Timing 5

6 Common Threads Head Grade - Resource Reserve models. Basic errors or incorrect assumptions in resource estimation. Productivity and misapplication of mining method Mining extraction overestimated and mining dilution underestimated. Processing, including metallurgy, equipment selection, and ramp up. Capital costs underestimated. Operating costs underestimated. People and Technology. Lack of experience / Team size Substituting software for judgment. External Pressures. Time and cost constraints. 6

7 The results? 7

8 Resource Head Grade Geology understand the deposit, do not let the software drive the interpretation and replace judgement Collect enough data Database integrity QA/QC. Capping. Specific Gravity or Bulk Density? Cut-off grades break even or incremental? Checkerboarding Low grade deposits and accuracy of estimation. Reconciliation. Grade smearing 8

9 The Effect of Cut-Off Grade = 1.0 g/t LEGEND: Grey = Blocks Below Cut-Off Red = Blocks Above Cut-Off 9

10 3.0 g/t LEGEND: Grey = Blocks Below Cut-Off Red = Blocks Above Cut-Off 10

11 5.0 g/t LEGEND: Grey = Blocks Below Cut-Off Red = Blocks Above Cut-Off 11

12 Dilution Mining method and production rate. Minimum width. Ground conditions. Equipment selection. Mining skills. Technical and supervisory skills. Grade Control and sampling. Assay turnaround. Most underestimated of all technical parameters. 12

13 Dilution & Extraction Mineralization Outline Internal Dilution

14 Dilution & Extraction Mineralization Outline Drift Planned Loss Planned Dilution Designed Stope Planned Dilution Planned Loss Drift 14

15 Dilution & Extraction Mineralization Outline Drift Actual Stope Unplanned Dilution Designed Stope Unplanned Dilution Drift 15

16 Minimum Mining Width Vein width vs. equipment requirements dilution. Block size for resource model. Narrow vein with big blocks or bulk mining with small blocks. 16

17 Minimum Mining Width 17

18 Selective Unit Open Pit Mine 2.5 x 2.5 m 5 x 5 m Considerations include: - Ability to locate individual blocks in BH pattern - Excavate entire block cleanly loading unit - From a shot-rock pile that has moved 18

19 Head Grades - Actual Production vs PFS and FS Studies Based on 70 Projects 19

20 Productivity Not enough workplaces = less flexibility in production plan = higher dilution. Mine operators will generally always deliver the required tonnes, but sometimes there is no ore in them. Zero base use cycle times, not factors. Lay out entire mine at the start, including all development. Schedule in detail monthly, quarterly, annually, life of mine. Detailed 3 month plan for operations. Underground 50 metres vertical per year. Open pit 8 to 12 benches per year. 20

21 Processing Metallurgical assumptions. Enough testwork to establish recoveries, con grades and impurities. Bench scale, Locked cycle tests, Grinding tests, Pilot Plant. Bulk Sample. Non Standard or new technology. Plant ramp up. Can range from two weeks to never. Should forecast four to six months at the FS level, for standard technology. Allow for experienced startup team and many people. Ensure assay lab is running and can turnaround samples quickly. 21

22 Capital Costs RPA database from EMJ Gypton Jan 2002 and recent work by RCF from RCF also gathered historical data from Mining Engineering Bullock Apr Close to 200 projects. General average overrun 20% to 25%. Much higher in times of high metal demand and prices. Not just mining analysis of other projects pre-2002 showed an average overrun of 28% - going back 90 years. 22

23 Capital Cost Overruns - Experience 300% Capital Cost Overruns in Mine Construction 250% 200% Cost Overrun 150% 100% 50% 0% $10 $100 $1,000-50% Original Project Estimate ($M) 23

24 Capital Cost Overruns By Project 180% 160% 140% Actual Capital Cost Compared to Budget 120% 100% 80% 60% 40% 20% 0% -20%

25 Costs Example How Reliable are Study Estimates 73% Increase of Capital Costs from Prefeasibility Study (PFS) to Updated Feasibility Study (UFS). Operating Cost increase not as dramatic, but still 37% higher than PFS. Capital Costs Billions $3.0 $2.5 $2.0 $1.5 $1.0 $0.5 $- Large OP Gold in Canada Oct-09 PFS Jun-10 FS Oct-12 UFS Oct-13 ACT INIT CAP SUST CAP Source of cost escalation largely in labour, consumables, equipment. This is not atypical for recent mining projects, across all commodities, mine types, locations, and company sizes. OPEX ($/t processed) $25 $20 $15 $10 $5 $- Operating Costs and Grade - OP Gold in Canada Oct-09 PFS Jun-10 FS Oct-12 UFS Feb-14 ACT Head Grade (gpt) OpEx ($/t processed) Head Grade (gpt) 25

26 Changes in Study Estimates Initial Capital Cost (Millions) $500 $450 $400 $350 $300 $250 $200 $150 $100 $50 $- Capital Cost Escalation - Mid Size OP Gold in W. Africa Oct-06 PEA Sep-07 PFS May-08 FS Sep-09 UFS Aug-10 ACT Initial Capital Costs (Millions) $250 $200 $150 $100 $50 $- Mid Size OP & UG Gold/Silver in Mexico Feb-06 PEA Aug-07 FS Dec-07 UFS Mar-09 ACT 15% increase in CAPEX from PEA to Actual 52% increase in CAPEX from PEA to Actual 26

27 Operating Costs Larger is lower unit cost. Smaller is higher unit cost. Complexity is higher cost. Remote areas are higher cost. The key cost drivers will change from area to area, mine to mine and process to process. What was key for a remote Canadian operation may not be key to a remote operation in another country. 27

28 Where Do Operating Cost Estimates Go Wrong? Productivity assumptions in every separate unit operation Tonnes per manhour Tonnes per hour of equipment Metres of development per day Productivity less at high elevations Understanding fixed vs variable costs. Missing the impact of administration on a unit cost basis especially in low tonnage operations (administration is fixed not variable). Believing that a zero base budget (or cost estimate from first principles) is correct just because there is a huge amount of detail in the calculations. 28

29 Where Do Operating Cost Estimates Go Wrong? Omitting items (administration, benefits, fees, head office, warehousing, support functions). Factoring based on other operations with similar methods. Not including costs for high end products that may be needed winter diesel, biodiesel, galvanized bolts, stainless steel products. Over complicated spreadsheets resulting in logic and arithmetic errors that defy auditing Failure to keep track of edits and changes to costs including the rationale for such changes. Believing that a fixed price contract passes the risk to the contractor or supplier. 29

30 Fixed Price Contracts and Cost Control 30

31 People and Technology Knowledge and experience gap. Lack of skilled people at all levels. Majors have abandoned the traditional role of training engineers and managers. Reliance on software in replacement of judgment = precision over accuracy. Statistical analysis is overrated. Should I really feel more comfortable if someone tells me the project is P90? Process related analysis can lead to a false sense of security, e.g. risk assessments. Team Size Matters - Project complexity is the primary driver Teams are commonly too small (especially for mega projects). Very common to see vacant positions on the org chart. 31

32 Non-Fatal Areas That Can Be Overcome Schedule Delay Ramp-Up Delay Discussed Previously Capital Cost Overruns Management and People Level of Study Individually these problem areas can usually be overcome, when combined, may again be fatal. 32

33 Schedule Delay The primary purpose of the schedule developed by the EPCM firm is to provide a target for project completion. The Master Schedule typically excludes any contingency. 33

34 Ramp-Up From: Terry McNulty Mining Engineering Oct

35 More Observations 1. Precision vs Accuracy - believing that detail makes it all correct. 2. Believing that cheap labour means lower cost. 3. Not benchmarking. 4. Truisms that aren t. We can do better than everyone else we have a better team. If that s all the money we can raise, then that s what we will build it for. The contractors should carry the risk or - don t worry, we have fixed price contracts. The EPCM people have promised us their best team. 35

36 Summary Objective: Provide background on why mining projects fail Describe fatal and non-fatal failures Describe problem areas to understand impact Outcome Understanding risks for mining projects Know when to get out 36

37 Richard J. Lambert, P.E. Tel: +1 (303) Toronto Denver London Vancouver Quebec City