Asset Optimization Systems: 5 Lessons From 10 Years in Mining
Asset Optimization Systems: 5 Lessons From 10 Years in Mining
Asset Optimization Systems: 5 Lessons From 10 Years in Mining
Greg Johnson
Schneider Electric Mining, Minerals and Metals
Greg1.Johnson@schneider-electric.com
KEYWORDS
Asset Optimization, Efficiency, Downtime, Production, Maintenance
INTRODUCTION
The mining business is not getting easier any time soon. Major market trends are creating
structural changes that are generating major challenges to mining companies. The basics of efficiency and
productivity and the systems to achieve these goals however are still of major importance.
As Dr. Tony Filmer (2009) has stated, As a price taker in a commodity industry, each mining and
metals business is seeking to sustainably reduce unit costs and maximize tonnage from capital assets.
Business success is based on accessing the best resource base, and then utilizing this resource at lowest
cost. A recent study by McKinsey (2012) however has shown that while investment in assets in Australia
for example has increased remarkably in recent years, performance has not followed that investment. The
amount of capital per worked hour is 25 % higher than it was six years ago but workers on average are
producing only 7% more output per hour.
Mining companies are continually looking for ways to use information and systems to address
these issues but typically face problems of many disparate pieces of information, data duplication and time
wastage in data gathering and a lack of a global perspective on plant information trends. One recent typical
review at a coal miner identified problems in the operational systems landscape as:
- applications landscape is complex and fragmented;
- there are significant data quality issues;
- considerable effort is spent moving data between applications; and
- access to data is neither timely or easy
While it is possible and indeed necessary to work on the overall information systems architecture,
the one system that can provide immediate benefits across the complete mining supply chain is the asset
optimization system (sometimes called the downtime system or the delay accounting system). Getting
this system right will result in many benefits in all areas of mining. Asset performance is critical to
production output, energy efficiency, recovery, environmental performance and schedule performance.
This paper will first review the theory and background to asset optimization in mining and then
review two case studies from real implementations over the last 10 years. The last part of the paper
presents five generalized lessons for success.
THEORY OF ASSET OPTIMIZATION IN MINING
What is Asset Optimization?
Lean management principles tell us that we need to continually seek out and eliminate waste. In
an asset intensive operation like mining, asset performance is critical and eliminating asset performance
losses are a major part of overall performance improvement. The goal is continuous operation and this is
achieved when all losses are eliminated and asset performance is optimized.
The method for determining asset performance losses was established by Taiichi Ohno, one of the
founders of the Toyota Production System as reported by Smalley (2006), Taiichi Ohno had people stand
at the problem machine for the entire 8-h shift and record the production plan versus actual amount in
small increments, such as 15 minutes to 1 h. At the end of the shift, all the losses and the actual reasons
why were identified in a Pareto chart.
Stops and delays are causes of losses because every time a process is stopped and started, losses
are accrued:
Ramp-up to production speed
Start-up waste
Extra wear on machine
Overtime to catch up on production
Energy losses
Fixed Costs are amortised over less output
Demurrage or train delay penalties
Dunstan et al. (2006) in a review of the applicability of Lean to mining have laid out some of the
key differences between manufacturing (as in the Toyota Production System) and mining operations. Some
of the key differences include:
In mining then, while the principles laid out by Taiichi Ohno still hold, there are some unique
considerations. Because production is continuous, slow running is equally important as stoppages. Causes
of losses can come from many areas including materials, people, poor practices and even the weather (for
instance, rain causes the top speed of the mobile fleet to reduce). The distances are large, so it is impossible
to visually detect loss events. There must be strong integration to the automatic control systems to
automatically detect issues.
Presentation of information should include Pareto Analysis, Gantt Chart Analysis and Pie Chart Analysis
(Figure 2), as well as flexible delivery of information by Report, Dashboard or Mobile App (Figure 3)
a Six Sigma common improvement language and methodology was introduced across what was then BHP,
to generate consistency, rigor, discipline and data-analysis into the decision-making processes.
Using the OE methodology, in 2001, Cannington chartered a project to reduce unscheduled
stoppages at its processing plant and achieve AUD$1 million in savings over a two-year period. Having
determined that the usability and richness of information available from its legacy systems would not
suffice in its current state, Cannington worked with Citect (now part of Schneider Electric) to implement
new supporting systems.
An OE coach who had completed training in the use of the Six Sigma quality process and tools
led the project. In keeping with a key Cannington business strategy; make decisions, based on
knowledge, the project was aimed at allowing users to drill down and analyze root causes of downtime,
and prioritize maintenance or improvement projects to optimize uptime. With the right information, the site
could more accurately measure the availability of different process areas, and analyze equipment
performance and production patterns to begin improving plant process performance. Using this
methodology, Cannington could compare availability of different process areas, and then drill-down into
each area to identify the main causes of downtime within that area. This would then allow isolation and
resolution of the largest bottlenecks before moving on to the next one.
A key specification was that the new system be operator error proof so that automated reporting
systems would not report on spurious data. This meant most data was gathered automatically and that
mannually entered data used for automated analysis, was entered via pick-lists rather than free-format text.
If an operator entered data such as Blocked Chute, or different words to describe the same problem then
the reports would be meaningless. Consistency of information was of critical importance to Cannington
and was a key criterion of the new system.
When a loss event was captured, an automated notification was sent to operators to prompt them
to enter information immediately. Data entry was much faster and more accurate using picklists. This
process helped alleviate many data consistency problems that Cannington had experienced with previous
systems. Once the operator completed data entry, it was available for confirmation by a supervisor. Once
confirmed, the event was locked and an audit trail created for any further changes to the event details. The
identity of the operator and supervisor are automatically captured by the system for validation. When
updating the details of an event, operators can select Cause locations/equipment from the plant hierarchy,
making navigation and data entry very simple. Cause codes and Classifications are then selected from
picklists filtered for that location, The concept of classifications and causes is shown in Figure 4.
Issue
Perceived Issue
Data Required
Proposed Solution
Issue
Perceived Issue
Data Required
Proposed Solution
How does the asset
optimization system
help?
The Asset Optimization system is not just a fault recording system, but also an identifier of
business improvement opportunites. Following an understanding of Lean concepts, it is the
driver of asset performance improvement.
2.
Spreadsheets do not work sustainably. Whilst being a flexible and powerful analysis tool, there
are too many problems to rely on spreadsheets in a production environment. Some of these
problems include manual entry errors, version control, lack of transparency, and no audit trail for
manual changes.
3.
When implementing Asset Optimization, take the opportunity to re-integrate the production,
maintenance and energy departments. One of the barriers to a more co-operative relationship has
been disconnected information systems. New systems remove the gap and can be used to
provide the elusive one source of the truth to drive global improvement.
4.
Advanced Planning and Scheduling will not work without reliable capability. Whilst new tools
incorporating the latest optimization methods are available to globally optimize schedules, these
tools will be less than successful if the assets in the supply chain do not perform reliably.
5.
New technology can do many things, technology is easy, but organisational change is hard.
Implementation of an Asset Optimization system will change work practices in the organization.
There is a need to invest solidly in change management.
The continuous improvement cycle assumes that the change implemented by a team becomes
permanent and the basis for a new round of improvements. Unfortunately, this is not always the case. A
key obstacle is sliding back to the old ways. This unravels the projects stated benefit and removes the
basis for further improvement. This sliding back is particular problem with manual business processes particularly when the change involves people who were not members of the team. How does one transfer
the understanding and acceptance of a change to others? This is further exasperated if the other people are
beyond line of sight in another depart or shift. On the other hand, software systems provide automated
business processes that broadly communicate and enforce the change. All the people affected see the
change and the effect can be made across multiple business systems.
Spreadsheets Dont Work Sustainably
Whilst spreadsheets are a flexible and powerful analysis tool, there are too many problems
(manual entry errors, version control, lack of transparency, too easy to manipulate) to rely on spreadsheets
in a production environment. The issues with spreadsheets can be illustrated by noting the manual shift log
shown in Figure 5. Times are rounded to the nearest 15 minutes and some descriptions make sense only to
the writer (trouble same as start of shift). In a typical effort at analysis, these sheets would be collected
and transposed to a spreadsheet, thus inviting further loss of precision and accuracy through the
transposition process. As spreadsheets are combined with other spreadsheets for further analysis
(downtime, production, energy consumption for example), timeliness and data trustworthiness are lost.
This leads to further losses in challenging the data.
Figure 6 Shift reporting available automatically and immediately after the end of shift
Use One Truth to Drive People Integration
Though not a universal rule, the relationship between maintenance and operations can often be
less than harmonious. At a review meeting analyzing why targets have not been met, the answer from
operations might be delays caused by equipment breakdowns and slow maintenance response. The
Maintenance viewpoint might be poor operating practices and lack of access for preventive work. The
asset optimization system can be used as the one source of the truth to forge a new working relationship
focused on real improvement as discussed in the coal mining case study above.
Other Optimization Depends on Asset Performance
Advanced Planning and Scheduling systems are starting to make an impact on mining operations
(for example, see www.solveitsoftware.com). An example of an optimized schedule is shown in Figure 7.
These systems offer tremendous improvement in co-ordination across an operations, but may be for nought
without reliable performance from the operations. This consideration also applied to other kinds of
optimization (such as that aimed at recovery or spare parts optimization for example). Consider that asset
optimization systems should be deployed first, as the basis for all performance lies in asset performance.
Time Frame
Shift
Day
Week
Month
Quarter
There is now 10 years experience with Asset Optimization systems in mining. Significant benefits
are shown in many operations around the world. Implemented in conjunction with a continuous
improvement process, these systems have proven a consistent benefit of 5% to 10% improvement in
production.
Leading miners are now extending these systems up and downstream, to provide information on
the bottlenecks and losses in the entire mining delivery chain. With a consistent process and consistent
system across the value chain, operators, engineers and managers can start identifting and implementing
the next round of improvements and productivity gains.
1.
2.
3.
4.
5.
Putting these lessons into action is one step in the goal of utilizing the resource at lowest cost.
REFERENCES
Dunstan, K., Lavin, B., & Sandford R. (2006). The Application of lean manufacturing in a mining
environment. In International Mine Management Conference, pp. 145154, October 2006.
Filmer, A,O. (2009). Business improvement in the mining and metals industry. The Journal of the Southern
African Institute of Mining and Metallurgy 109, 621628
McKinsey Global Institute. (2012). Beyond the boom: Australias productivity imperative.
Smalley, A. (2006). Creating Basic Stability, Retrieved from
http://www.superfactory.com/articles/basic_stability.aspx. Available 1/10/06