Platinum Group Metals: Applying Best Practices in Metal Production Accounting and Reconciliation
Even if a mine can accurately estimate expected production at the grade control level (in-situ), issues such as blast movement, poor mining practice and inefficient ore/waste classification make for a disconnect between the in-situ figure and estimates from the concentrator. This inhibits effective evaluation of the reserve and grade control models using concentrator estimates. Linking the two ends, mining and concentration, is a neglected area sorely in need of an overhaul.
By Will Jansen, Ph.D.
When Enron rocked the United States with its Titanic-sized financial accounting scandal, miners and metallurgists were probably too busy at mine sites in remote locations to wonder about its ripple effect.
But then Enron sank, Arthur Andersen tanked and the list of scandals, costing billions of dollars, continued to ripple. Followed by Sarbanes-Oxley, the recent credit crunch and weak economy, the currents washed right into the mine—and the mill.
These events placed intense scrutiny on metals production accounting processes as traditional semi-qualitative metal accounting practices provided insufficient credibility and transparency. “There is growing intolerance of organisations that are unsure about the location of thousands of tonnes of ore, which will typically lose 10–15% of their notional value on the way to the concentrator,” writes Professor Rob Morrison in the opening chapter to An Introduction to Metal Accounting and Balancing. “If the measurement and forecasts also contain qualitative (fudge) factors, a very real suspicion might emerge about just how much metal might disappear (by theft or misplacement) before detection becomes likely.”
Fudge factors aside, reporting a precise, accurate figure is common dilemma in most mining groups. For Platinum Group Metals (PGMs) producers, the industry as a whole is characterized by “silos” that determine their data in their own way and share it with reluctance and a general lack of transparency. Therefore, data is often lopsided, in the sense that often cost data is precise and detailed while data about the value of the metal in the process is not. This can result in a skewed view of the business, where cost-based decision-making takes precedence over strategic thinking.
Lonmin, a primary producer of PGMs with mine sites in South Africa, recently faced this issue. Senior management was given numerous data sets for the value of the metal in process, but the difference in data sets was significant. With a fetching price of ZAR300,000/kg of metals and a single silo containing up to 500 kg of the end products, it added up to ZAR150 million.
Therefore, the expensive mud had to be properly valued. The only way to accomplish that was to stop the process, drain, and (as far as possible) empty it. Then, calculate the value of the metals that remained in the process. If it was more than the process managers had estimated, they really wouldn’t know why. If it was less, they would not always know where to start looking for the missing precious metals. In the meantime, they would lose a week’s production. That’s not only a lot of loss production, but a significant loss of money. Hence the need for a Metal Production Accounting system that applies best practices.
Wynand van Dyk, senior technical manager in the process division at Lonmin, defined metallurgical accounting as a complete account of what has happened to the metals that have been put into a process. “While this sounds straightforward, it isn’t—because the process isn’t straightforward,” Van Dyk said. “Each of the six platinum groups metals embedded in the ore requires different process to extract and refine it. There is flex at every stage, and the processes are not all linear—some PGMs go around in circles before emerging, taking longer than others and passing the measurement points more than once. Because it is impossible to measure the stream in its entirety, accounting depends on sampling, which introduces the potential of data inaccuracies.”
As a result, mining companies are beginning to examine the quality of their precious metals accounting systems since they form the basis of company financial accounts and reports. As part of an effort to create that accounting standard in the mineral industry, University of Queensland’s Julius Kruttschnitt Mineral Research Center published the textbook An Introduction to Metal Accounting and Balancing in 2009. While development of a textbook and standard are significant steps forward, the challenge in the precious metals industry is in the application of some of these suggested practices.
Though improving process transparency and precious metals accounting reporting accuracy is a key driver for change from the financial side, perhaps even more important is the potential benefit to operational performance that comes with better measurement, accounting and reconciliation systems. Loss of precious metals, processing of waste and inefficient mining and processing systems are all problems that are faced in an operation on a daily basis which can help be resolved through improvements in metal accounting and reconciliation. The financial consequences of such operational inefficiencies and bad decisions based on poor information are arguably greater than consequences of fraudulent accounting.
Large discrepancies between mine and mill measurement systems that measure the quantity and quality of valuable material currently exist. Some of these are expected because of statistical variation, but inherent random measurement variation is sometimes an insufficient explanation. Though the mill can usually make much better production estimates through their measurement systems (that is, less variation), this knowledge rarely benefits the mine.
Moreover, even if the mine does a reasonably good job of estimating expected production at the grade control level (in-situ), issues such as blast movement, poor mining practice and inefficient ore/waste classification make for a disconnect between the in-situ estimates of the mine and the estimates from the concentrator. This inhibits effective evaluation of the reserve and grade control models using concentrator estimates. Linking the two ends, mining and concentration, is a neglected area sorely in need of an overhaul. Hence the need for improved mine to mill reconciliation: the area of comparing, balancing and adjusting production estimates between mine and mill for consistency in reporting.
Problems with Current Reconciliation Practices
Reconciliation, as it applies to the mining industry, is in a basic sense the comparison of estimates from different sources over a certain period of time. In order to more easily assess the problems in the current practice, I’ve divided them into three categories: general, cultural and technical. As there is not enough space in this article to drill down into all of the issues, this is a broad view.
- There is a lack of standardization in metallurgical accounting (until the P754 code that was recently released) and particularly mine-mill reconciliation. The release of the P754 code is a useful step towards standardization.
- There is a lack of universal applicability of many of the more novel approaches which often exploit site-specific situations (such as large differences in grade to drive a mass balance approach) for improved reconciliation.
- The use of spreadsheets as the basis of metal accounting systems is common. Spreadsheets are inherently difficult to secure, maintain and audit. During many of my visits to mine sites, I’ve witness paper and pencil spreadsheets lying on the ground where they’ve been trampled or on a work table, with coffee spills making some of the data unreadable. In addition, spreadsheets are often difficult for new employees to work with and are poor receptacles for data storage.
- Many metal accounting systems lack a clear audit trail and process transparency, making the sources of production estimates and adjustments unclear.
- A clear and significant cultural divide exists between the perspective (theory/views of measurement) of mining staff responsible for mine production estimates and metallurgical staff responsible for mill production estimates.
- Current reconciliation practice pits two conflicting sets of data against each other in a monthly reconciliation meeting without regard to measurement error or statistical variation.
- Management-set metrics, incentives and production goals are often contrary to mine-mill cooperation.
- The challenge of measurement in mines appears to have led to an “estimation culture” whereas metallurgists can be said to have a “measurement culture,” mainly due to the ease of measurement in the mill compared with the mine.
- There is often a lack of understanding or interest in the benefits of focused reconciliation efforts as evidenced by attitudes of practitioners and levels of effort; usually individual site champions are behind good reconciliation practice.
- There is a general failure to appreciate the potential benefits of more rapid identification of problems and opportunities by senior management.
- There is reliance upon reconciliation factors such as the mine call factor, shaft call factor, etc. as a means for prediction of future mine-mill estimate discrepancies.
- The use and abuse (i.e., unjustified adjustment) of factors in reconciliation such as moisture content, truck/loader factors, densities, etc. is common.
- Arbitrary adjustment techniques are often used for the adjustment of production figures in the case of a mine-mill reconciliation discrepancy.
- There is commonly an inability to determine the most crucial of the numerous areas contributing to reconciliation discrepancy due to the lack of a formal system for hierarchically ranking the impact of the potential causes of mine-mill reconciliation.
- The presence of stockpiles can obscure ore sources as well as unjustifiably “absorb” discrepancies over time.
- Many sites rely on two-product formula calculations for metal balances instead of the more robust “check-in check-out” method supported with redundant data.
- Statistically sound metal accounting approaches based on the confidence in the data are not widely used.
- Untested assumptions regarding the accuracy of various sources of data are often used in the reconciliation process.
Overcoming the Barriers
The first step involves dissolving the traditional mine-mill boundary from a measurement and accounting perspective to avoid the clash of two sets of figures during the reconciliation process. This can be achieved by developing a system that provides a holistic point of view that can examine metrics and incentives that directly or indirectly discourage good accounting/reconciliation practice between departments.
Next, after having defined the system and surveyed the site for the sources of data that are available for mine-mill production accounting and reconciliation, analyze these measurements to obtain more information regarding:
- how the measurements can be classified
- the magnitude and behaviour of error in the measurements
- the measurements’ relationship to each other in terms of redundancy.
The third step is a process hold-up analysis that involves locating points of accumulation for process material over periods of time. Different process hold-ups can have varying effects on the quality of the accounting and reconciliation system. Therefore, developing a method for evaluating and dealing with each process hold-up in an operation is crucial to the effectiveness of the mine-mill metal accounting and reconciliation system.
After having defined the target system and performed thorough analysis of the current measurement network and process hold-ups, step four is to perform a gap analysis to determine if there are any production measurements that are not currently taken that may improve the quality of mine-mill metal accounting and reconciliation. Once these are identified, a cost/benefit analysis should be undertaken to determine the value of adding such measurements into the process.
Next, step five of mine-mill metal accounting and reconciliation methodology requires an audit of behavior on the site at both the operations and management levels to evaluate that behavior’s potential impact on development of a credible accounting and reconciliation system.
The next step is one that proves to be exceptionally challenging: a transition from spreadsheets to technology, which involves integrating the previous gathered data into an electronic database or web-based system. These are not only less prone to spurious error, but they are also more secure. In addition, they act as far better production data storage receptacles than spreadsheet-based systems. An ideal technology solution would possess the following characteristics:
- The ability to manually enter or automatically capture key production accounting data as it become available.
- Automated validation routines for entered/captured data as well as balanced/adjusted data
- The ability to perform various types of data analysis such as Monte Carlo methods, statistical mass balancing, and reconciliation adjustments
- Varying levels of data security, which allow various levels of data viewing, manipulation and sign-off
- Audit function which can track the source of all data used for various production estimates and KPI calculations as well as any adjustments to data and the justification for them
- Ability to report production accounting information data at a number of levels
- Formal exception report as allowed in the P754 Code of Practice (AMIRA, 2007)
Another technology solution is at the heart of step seven. The ability to track ore from source to product is essential for obtaining reliable, detailed information about the deposit, since measurements produced by the processing plant are usually much more accurate than the estimates from the mine. An ore-tracking system that can identify the geographic origin of each parcel of ore as it is measured and processed in the plant can form the basis of a high-resolution reconciliation system. It can better measure the material at specific geographic locations in the mine (i.e., more precise and unbiased estimates attributed to individual ore parcels) and determine estimates for ore/waste misclassification and dilution/ore loss.
A reliable ore tracking system, which relies on RFID, should allow material to be tagged at its source, using markers representing volumes of ore, and later detected as it flows through the system or not detected if sent to a waste dump by accident or design. Even tagged material sent to long-term stockpiles can be detected later in the future when it is eventually processed and then reconciled back to its source.
Step eight begins with the entering and validating of concentrator production measurements before a check-in check-out balance and statistical data adjustment step is applied to the data. Data adjustments and the resulting balanced production estimates are then analyzed before moving on to use this information as the basis for mine production reconciliation (step nine).
Once a statistical mass balance and corresponding data reconciliation has been carried out, there is now a basis for mine production and mine model reconciliation. The key output from step eight to be used in this step is the balanced estimates of feed grade and tonnage to the concentrator. For the concentrator metal balance, it was shown that one can either fix the measured values of the final product(s) during balancing (commercial balance) or one can allow this number to be adjusted during balancing based on the uncertainty in the values (technical balance). A similar decision must also be made at the mine-mill custody transfer point. The feed estimates provided by a concentrator balance are often much more accurate/precise than those estimated by the mine (from precision experimentation during the site survey), so it is recommended that this value be fixed once the concentrator balance is sufficiently defined.
In order to maximize the value of mine-mill reconciliations, sensitivity analyses and clear reporting processes are necessary to finalize the process. Effectively undertaking these two tasks increases the utility and benefit of performing reconciliations, and the results should justify the effort put into the process. Determining the measurement or estimation areas of the process which is most in need of control, monitoring or improvement should be an outcome of a good accounting and reconciliation strategy. Also, quality results are meaningless if not effectively communicated to those who will use this information for reporting and decision-making. Thus, clear reporting protocols and examination of process transparency and auditability are an important step of the mine-mill metal accounting and reconciliation process.
Finally, mine-mill metal accounting and reconciliation systems need a procedure for on-going evaluation, where historical trends are continuously examined over time and where issues highlighted over one or more accounting periods are investigated and corrected. The ability to highlight problem areas effectively and to adapt to changing conditions or personnel are key components to a mine-mill reconciliation process.
The processes involved in metal accounting and mine-mill reconciliation are not simply accounting non-value task that are for the reporting of production figures, despite ripple effect from the financial accounting scandals and compliance. At the end of the day, the reality is that production measurements and means of dealing with those measurements provide a wealth of operational knowledge where the benefits outweigh the risks inherent in change.
Jansen holds a B.S. degree in Mining Engineering and Metallurgy (Summa Cum Laude) from Virginia Tech University, Blackburg, Virginia, USA. He worked for Palabora Mining Co. in South Africa before undertaking a Ph.D. program at the Julius Kruttschnitt Mineral Research Center (JKMRC) at the University of Queensland in Australia. Upon completion of his thesis in the area of metal accounting he took a position with Mincom, first in the in South Africa Johannesburg office as a principal consultant in production accounting before transferring to the Denver, Colorado, USA office as Strategic IMS consultant. This article is based his Ph.D thesis, titled A Strategic Approach to Mine-Mill Reconciliation. His research is also included in the Introduction to Metal Accounting and Balancing textbook published by JKMRC.