Minerals Hub / Innovation & Technology / Reconciling a mine that makes four products
Innovation & Technology · Rutile · 6 min read
Reconciling a mine that makes four products
Reconciliation compares an estimate with what actually came out — which is harder when the model, the plant and the customer all count the rock in different units.
Reviewed by Peter Uppal
The short version
Reconciliation is the practice of comparing what a mine estimated against what it delivered, and the industry does it with a small set of ratio factors computed separately for tonnes, grade and metal. A heavy-mineral operation runs that comparison for several products at once, and it runs it across three different accounting systems: a geological model that may be expressed in mineral percentages, a plant that measures oxide and element grades, and a customer who buys a mineral specification. The arithmetic that converts between those systems is itself an estimate, which means part of any discrepancy sits in the translation rather than in the mine.
What reconciliation actually compares
The word suggests a single check. In practice it is a structured set of them.
The standard framework compares model against model and model against plant, in stages. As one widely used account sets it out, the F1 factor is short range model depletions / long range model depletions — the short-range ore-control model set against the long-range model it depleted, which bears on the accuracy of orebody knowledge in the reserves. The F2 factor is received at mill / delivered to mill — what the mill measured against what the mine recorded as delivered. The paper attaches the failure to keep ore and waste apart to the short-range model rather than to F2, naming inefficiency in the mining process to segregate ore and waste as planned (short range model) as the second of its two principal causes of poor reconciliation. And the F3 factor combines F1 and F2, enabling a comparison of a mine's ability to recover the tonnage, grade and metal content estimated in reserves; in the copy consulted for this article the operator joining the two ratios does not render legibly, so it is not reproduced here. What the paper does say about how they compose is that over a period of time (annual) short range model depletions will cancel with delivered to mill.
Two features of that framework matter for what follows. Each factor is computed for tonnes, grade and metal separately, because the same account notes that generally, both tonnage and grade should be estimated or measured and compared independently. And the framework assumes the two sides of each comparison are measured in the same units. In a single-metal mine they are: grams per tonne in the model, grams per tonne at the mill.
Three counting systems for one rock
That assumption is where a heavy-mineral operation departs from the pattern.
The product is defined mineralogically. A study of one such operation records a concentrate split into three commercial products through various processes in the mineral processing plant, and states that after separation each of the product streams must fulfil predefined specifications not only based on chemical composition but also mineral quantities. Not only chemistry: mineralogy. That is the customer's system of account.
The assay laboratory, meanwhile, measures elements and oxides. The bridge between the two is a calculation. As one description of the method puts it, element-to-mineral conversion is the process of converting bulk chemistry data into mineral grades using known mineral chemistry, on the principle that the mineral grades multiplied by the chemistry of the minerals is equivalent to the bulk rock chemistry. The same source states the constraint that makes this awkward for an assemblage like this one: elements can only be assigned to minerals once, making it difficult for such an approach to differentiate between minerals that share a similar chemistry, and the mass balance approach is limited to deposits with simple mineral assemblages.
Osmond's own reporting shows the two systems side by side. Its Zone 1 bulk channel results are published both as oxide grades and as mineral percentages, and the company states in its 16 July 2026 release that mineral percentages in its reporting are mass-balance estimates rather than direct measurement. That is correct disclosure, and it is also a warning label about what a later reconciliation would be comparing.
If the model counts minerals, the plant counts oxides and the buyer counts specification, a reconciliation is not one comparison. It is a comparison plus two translations, and translations have error bars too.
Where a discrepancy can hide
The standard causes of poor reconciliation apply here as anywhere. The framework above attributes errors to inaccurate orebody knowledge at the time of ore reserve estimation and to the assumptions made as to the accuracy with which ore is selected from waste, and notes sampling bias directly: over sampling of fines in blast hole or grab samples can cause a bias. It also flags a problem this industry is exposed to, since heavy-mineral operations stockpile heavily between mine and plant: reconciliation in the cases where old or large stockpiles are reclaimed or where a large percentage of production is related to stockpile reclaim can be problematic, because stockpiles are assigned the average grade of all the increments added to them.
To those, a multi-product operation adds one more. A shortfall in one product and a surplus in another can be produced by a single misallocation in the conversion step — the same titanium counted as rutile rather than as altered ilmenite, say — while the total titanium reconciles perfectly. The books balance and the products do not.
The limit of what can be said here
That last paragraph is reasoning, and it should be labelled as such. Despite targeted searching, no source could be found stating that reconciliation is harder in a multi-product or polymetallic operation than in a single-metal one — not in mineral sands, not elsewhere. The framework quoted above is written for single-commodity operations; the multi-product descriptions come from a separate literature about plant monitoring; the two have not, as far as this research found, been joined up in print.
That is an unsatisfying place for an article to end, and it is the accurate one. What the sources do establish is each of the pieces: reconciliation is a structured comparison in matched units; heavy-mineral products are specified mineralogically; mineral percentages are commonly derived from chemistry by a calculation with stated limits; and stockpile reclaim complicates the comparison. The conclusion those pieces point toward is available to any operator who assembles them, and it is worth stating that the assembly does not appear to have been published.
Exploration results and mineralogical estimates only. Orión has no JORC-compliant Mineral Resource or Reserve; maiden MRE and Scoping Study pending.
Sources
- Parker (2012), Mining Technology 121(3) — Reconciliation principles for the mining industry (third-party hosted copy, not the publisher's version of record)
- Minerals (MDPI) 11(11) 1253 — Heavy Mineral Sands Mining and Downstream Processing
- Minerals (MDPI) 16(2) 139 — Predictive Modelling of Lithium Mineral Grades from Chemical Assays
- Osmond figure: osm_grade — 19 February 2026 (ASX:OSM release)
Related reading
- The model that decides which mineral you found covers the translation step this article treats as a source of error.
- What a four-product plant asks of its controls is the same multiplication problem in real time rather than after the fact.
- The robot in the sample preparation room deals with the error that enters before any of this begins.
Sources
- PRIMARYParker, H.M., 'Reconciliation principles for the mining industry', Mining Technology (Transactions of the Institutions of Mining and Metallurgy, Section A), vol. 121 no. 3, 2012 (definitions of the F1, F2 and F3 factors; tonnes, grade and metal reported separately; causes of poor reconciliation; stockpile reclaim).
- PRIMARYKoenig and Verryn, 'Heavy Mineral Sands Mining and Downstream Processing: Value of Mineralogical Monitoring Using XRD', Minerals (MDPI) 11(11), article 1253, 2021 (heavy-mineral concentrate split into three commercial products; product streams must meet specifications on chemical composition and mineral quantities; monitoring and blending).
- PRIMARY'Predictive Modelling of Lithium Mineral Grades from Chemical Assays for Geometallurgical Applications', Minerals (MDPI) 16(2), article 139 (element-to-mineral conversion; mass balance principle; elements assignable only once; limitation to simple assemblages). A lithium study, cited for the method.
- PRIMARYosm_grade — ASX:OSM release, 19 February 2026 (Zone 1 bulk channel samples reported as both oxide grades and mineral percentages).
- PRIMARYASX:OSM release, 16 July 2026, 'Mineral Resource Target Area Expanded', Appendix B (mineral percentages stated to be mass-balance estimates rather than direct measurement).
- UNVERIFIEDGAP: no source could be located stating that reconciliation is harder in a polymetallic or multi-product operation than in a single-metal one, in mineral sands or elsewhere, despite targeted searching. The comparison drawn in this article is reasoning from the reconciliation framework and the multi-product descriptions cited, not a sourced finding.Non-public document · no public URL
