How a theory applied to real world can increase the production performances? Sampling constitutes the main source of measurement error. Among all the components of the overall measurement error, the sampling error is generally the largest. The estimate of the sampling error necessitates to build heterogeneity models for the different kinds of processed materials. Discover in this paper how Eurasian Resources Group («ERG») applies the Theory Of Sampling (TOS) to redesign its sampling plan and improve its measurement system. An article presented at the 9th World Conference on Sampling and Blending (WCSB9) on May, Beijing (China).

How sampling error calculation can enhance the measurement system in a chromite concentrator

Abstract

The ERG chromite concentrator performs many individual measurements on the different streams to assess plant performance and to produce the metal accounting report. In addition to measuring flowrates, numerous samples are taken for analysis, mainly to obtain the Cr2O3 content.

The accuracy of such measurements is a key factor for confidence in the performance calculation and, more critically, in the metal accounting. Continuous enhancement of the measurement system is the only way to improve the production with relevant data. The calculation of the measurement error, including the sampling error, allows to estimate the confidence level for each measurement as it is currently done, and to improve the procedures and practice.

Among all the components of the overall measurement error, the sampling error is generally the largest, and, more precisely, the fundamental sampling error. Following the Theory Of Sampling (TOS), the variance of this component is proportional to the inverse of the sample mass and to the Intrinsic Heterogeneity (IH), which is estimated based on a heterogeneity model representing, as far as possible, the heterogeneity of the ore. Such a model can be derived from various sources of information including geological description of the ore deposit, mineralogical studies or process data.

The present paper describes how to use this information to build a relevant model and how this model is used to calculate the IH for each step of the sampling plant and, consequently, to calculate the overall measurement error. The same approach is used to redesign the sampling plan for measurement improvement. The case of some specific streams illustrates this approach.

Contents 

• Chromite ore and processing plants
  • Chromite ore
  • Crushing
  • Concentrators
  • Pelletizing plants
• Heterogeneity model
• Measurement error calculation
  • Sample preparation procedures
  • Chemical analysis
  • Uncertainty budget
• Application to two products
  • Rich ore sampling
  • Gravity tailings sampling
• Conclusion

Authors 

• Stephane Brochot, PhD. – CASPEO
• Philippe Wavrer, PhD. – CASPEO
• Sergey Opanasenko – ERG