Thursday, 4 December 2008

Improving Process Data Quality in Mineral Processing Plants

Advanced mathematical and statistical tools facilitate the detection and speed up the correction of process data used in the computation of the metallurgical and production mass balances. But this is not the only benefit.

Statistical Data Reconciliation

More and more world leaders in mineral processing install advanced computer software for improving the quality of their production data and the reliability and availability of their metallurgical and production reports. The direct benefits of more accurate data and better availability are a faster and better decision making in terms of process operation and production management. Both the metallurgical staff and the management staff use such tools for detecting and correcting measurement errors before they result in bad decision makings and accounting errors. Most of the time, despite the simplicity of the physical laws of mass conservation, reaching a coherent mass balance in a mineral processing plant is a tedious task. The reasons are numerous:

• The heterogeneous nature of the streams results in non representative samples,;

• The lack of instrumentation or the lack of precision of the instrumentation when available results in non reliable process variable values;

• The various and diversified data sources (databases and manual entries) render complex the collection of the required data;

• Too often, human errors are also introduced in the data collection process.

When measurement errors are not detected, they propagate through the computation of recoveries, yields and other key performance indicators. Bad decision makings inevitably result.

The data uncertainties must therefore be taken into account for detecting, sizing and correcting measurement errors. As a consequence, statistics and statistical methods are a prerequisite to any data reconciliation procedure and mass balance computation. This significantly increases the complexity of converging towards a coherent and reliable production mass balance for most metallurgical people. However, advanced mathematical and statistical tools have been developed some twenty or thirty years ago for solving that particular problem and solutions exist for addressing the specificities of the mineral processing industry. While some major mineral producers have been using such solutions for many years now, the decreasing costs of information technologies are now making those solutions available to any mineral processing plant. Furthermore, the originally marketed solutions have significantly been improved profiting from the advances in information technologies.

The Benefits

The direct benefits of statistical data reconciliation and of a faster and wider distribution of metallurgical and production performances are easily identifiable: more accurate, more reliable and more easily available process data inevitably result in faster and better decision makings and finally contribute to increasing the overall profitability of any mineral processing plant. Statistical data reconciliation also indirectly contributes to improving sampling and sampling strategies, laboratory sample analysis and instrumentation maintenance procedures. Finally the accumulation, sharing and retention over the years of the process know-how by the plant personnel is also greatly improved.

To read full article, please visit

No comments: