Ensuring metal accounting practices meet the requirements of modern corporate governance requires codes of practice, standard operating procedures and planned task observations. AMIRA International facilitated the development of AMIRA P754 (Ref 1), a standardized Code of Best Practice for metal accounting and reconciliation.
Meeting the AMIRA Code and Theory of Sampling principles in a self-owned processing facility is often challenging in itself. But how can these principles be applied in a third-party toll treatment facility treating multiple ore sources in a continuous blended feed rather than in campaigns?
Through its 10 core principles, the AMIRA Code addresses critical aspects of metal accounting, from data collection and validation through to reporting and auditing. A key principle is the metal accounting system must be based on accurate measurements of mass and metal content across defined ‘Check-in Check-out’ boundaries using best practices as defined in the Code. In a typical Carbon in Pulp (CIP) or Carbon in Leach (CIL) gold plant, such accounting boundaries might include mill feed, leach feed, leach residue, elution circuit feed and gold bullion produced.
Accurately determining the metal content of delivered run-of-mine (ROM) ore is a key challenge given ore sources with varying characteristics are blended and must be separately accounted for at the coarse delivery size (e.g. primary crushed).
Wet solid flowrates are readily measured via weighbridge or conveyor weightometer, subject to approved calibration procedures. Dry mass determination requires measurement of moisture content of ore on delivery (at best, rough estimates).
Stop-belt sampling and automated cross-belt sampling are two techniques used to collect primary samples of ROM ore for head grade determination; both have strengths and weaknesses. Stop-belt sampling interrupts ore flow, while automated cross-belt samplers do not. Automated cross-belt samplers, however, are more prone to yield biased samples if the complete size range is not quantitatively recovered. In both cases, samples should be tonnage rather than time-based. This results in the generation of large primary samples that need to be crushed and split to yield representative aliquots for laboratory analysis.
Given the blended feed, recovered gold is typically allocated in proportion to the expected recoverability of each ore source. This allocation is based on the delivered feed content and a leach factor derived from bottle roll tests, which account for varying leach characteristics among ore sources. It is essential the bottle roll procedure is aligned with the actual plant process; for example, with/ without carbon in leach, similar grind sizes, cyanide concentrations and leach times.
Finally, full disclosure of metal accounts must be stated in the toll treatment agreement, while maintaining the confidentiality of other participants.
Ref 1: AMIRA International; P754: Metal Accounting; Code of Practice and Guidelines: Release 3; February 2007
