Thursday, 26 January 2012

Productivity and Mine Planning - Part 2


Before Christmas I started writing about the issue of mine planning.  This entry follows that same theme.  This week, I will again visit the area of mine plans not delivering quality information for appropriate decision making.  Why haven’t shareholders and the stock exchanges held Boards of Directors accountable for their poor decisions on how to proceed with mining a particular resource or whether to proceed at all.  When shareholders and stock exchanges start holding mining companies accountable for the decisions they make, the people doing the planning might need to start explaining themselves.  Development engineers may need to dust off their CV’s and mine planning consultants might need to become acquainted with their Professional Indemnity insurers. 

The scheduling side of Mine Planning Tools is letting the industry down badly.  As new logging and monitoring technologies are being developed, the amount of and complexity of the data is becoming overwhelming and decisions are often made based on only a small portion of the available data.  In most cases production estimates are higher than what is achieved.  Is it a human trait to be optimistic or is it pressure to produce results which are good enough to gain shareholder or Executive Management approval?  We work in an industry which is wildly optimistic about what could be achieved and then prepared to accept mediocrity in what is delivered.

Most of the inputs are based on ‘ideal’ values or values that are given by OEM’s or third party experts.   The biggest mistake made by planners and their managers is not linking the detail of the plan to the requirement of the stage of planning.  A good example is a long term plan which plans a shovel or excavator down to payload, wait on truck, fill time, swing time, etc. and then ties it all together in some impressive-looking Monte Carlo simulation.  Surely for a long term plan you should use realistic annual production numbers.  Detailed analysis follows.  Then we have the problems with planners not using enough detail for short term plans.  Now this is a fine line.  We can collect data ad nauseum but then something changes in the pit so the ongoing optimisation of the plan becomes a balance between collecting and using data and the dynamics of the pit.

The following are actual reasons why overly ambitious production rates have been used from personal experience;
  • Dig depths and face heights not considered, 
  • Variation in seam dip not considered,
  • Planning done in 2D and then merged to 3D,
  • Scheduling using maximum potential rate for KPI’s and productivity rather than what can be achieved over a longer period,
  • Scaling performance from equipment of different capacity,
  • Overestimating hours of work,
  • Not considering fleet interactions,
  • Not understanding operational limitations, eg. Double side loading vs single side loading
  • Not understanding densities and bucket fill

While most mining executives have encountered plans which have gone pear shaped they haven’t always understood why.  Well the answer is in many occasions the poor use of realistic production rates.  The following are actual examples from our work in the past for loading units and trucks:

  • Truck fleet actual operating hours 21% below plan.
  • Electric rope shovel actual annual output up to 33% below plan. 
  • BER (payload / bucket capacity) 25% below plan
  • Payload for trucks being assumed at design load, when on-site performance was 14% underloading (limited by tray volume)
  • Dragline swing time used in plan was 14 seconds when actual time was 22 seconds.
  • In 2008 the average shortfall in dragline coal uncovered was one million tonnes per dragline, (for 20% of draglines there was a shortfall of over two millions of coal)

In each of these cases (partly the reason I chose them) the actual performance is not vastly different from the worldwide average for the make and model. 

You can’t plan effectively without accurate inputs.  You can’t make good decisions without good planning.  You don’t have to accept inaccurate inputs.  Just use the data available.  Benchmark against industry standards.  It seems too obvious.

Graham Lumley 
BE(Min)Hons, MBA, DBA, FAUSIMM(CP), MMICA, MAICD, RPEQ

No comments:

Post a Comment