F-Factor Dilution Signals and Ore Reserve Correction

Reconciliation data is often left in a spreadsheet that nobody reads until a budget variance makes someone look. By then, ore has been sent to the waste dump, diluted tonnes have been processed at lower grades, and the reserve statement no longer matches what the mine is producing. The F-factor system exists to catch these problems early, but only if someone is reading the numbers correctly. Start with F1 and F2, which show where the metal is lost.

What the F-Factors Measure

The F-factor system compares the reserve model, the grade control model and mill production by means of three ratios. This article uses the standard definitions (Parker 2012). F1 compares the grade control model with the reserve model (or the resource model) for the same mined volume. F2 compares what the mill receives with what the mine delivered to the mill according to the grade control model. F3 is F1 × F2, so it compares what the mill receives with the reserve model. Each factor is calculated for tonnes, grade and metal. Each ratio measures a different part of the process, and together they show where the losses happen.

Most operations watch F3 because it connects directly to revenue. That is the wrong place to start. F3 shows that metal is missing, but not where it was lost. F1 and F2 separate the reserve model from the mining process. A persistent F1 (grade) below 1.0 means the grade control model finds less metal per tonne than the reserve model predicted. A persistent F2 (grade) below 1.0 means the material arriving at the mill has less metal per tonne than the grade control model predicted for the ore delivered. The next step is to find out why.

Three Common Root Causes

A falling grade factor has three common root causes, and each requires a different response. We see this in our work reviewing metal reconciliation programs at copper operations in the Americas. The first is estimation bias in the block model: the model systematically overestimates grade in a mineralized domain, and every tonne mined from that domain confirms the gap. This shows as a low F1 (grade). The second is geological dilution: the orebody boundaries are more irregular than the model represents, and lower-grade material enters the ore stream at the contacts. This also shows mainly in F1, because the blasthole data in the grade control model find the lower grades that the reserve model did not predict. The third is operational dilution: blasting, excavation, and loading practices are adding waste to ore truckloads, or mixing ore and waste in stockpiles. This shows as a low F2 (grade).

Production pressure can make operational dilution worse. When the mine is pushed to meet a higher tonnage target, mining becomes less selective, more waste enters the ore stream, and some ore that needs careful extraction is left behind. Head grade then falls even when the cut-off grade does not change (McCarthy 2011). This often happens when the processing plant has spare capacity and the mine is asked to fill it. If F2 (grade) falls in the same months that the mining rate goes up, check the production rate before you change the block model.

The diagnosis decides the fix, because the corrective action is different in each case. Fixing a poorly conditioned variogram does not help if the cause is a blast pattern that is too wide for a narrow vein contact. Stricter grade control practices do not help if the block model is wrong about where the grade boundary is.

The Copper Dilution Pattern We Have Seen in Peru

In Peruvian open pit copper operations, the geological setting creates specific dilution risks that reconciliation data tends to show before the geologists see them in the field. Porphyry copper deposits often have strong grade continuity in the high-grade core, but the transition zones to the low-grade halo and to lithological contacts are spatially variable at a scale smaller than the SMU. When the block model is built with drillhole spacings appropriate for the central resource, and those same parameters are then applied to the transition zones, the model will show those boundaries with more confidence than the data supports.

The reconciliation numbers in this situation are easy to recognize. F1 (tonnes) stays close to 1.0, meaning the ore tonnes in the grade control model match the reserve model. F1 (grade) slowly falls below 1.0, and the fall is concentrated in specific benches or sectors of the pit. When we map that spatial pattern against the block model domains, the sectors with low F1 (grade) almost always correspond to the transition zones and to contacts with intrusive phases of variable grade.

Why This Changes the Ore Reserve

This pattern affects the Ore Reserve because the transition material is often marginal grade material, with grades close to the cut-off grade. When dilution lowers the effective grade of that material below the cut-off grade, it goes to waste because the ore/waste boundary was drawn with too much confidence in a model that the geology did not support. In this case the Ore Reserve tonnage decreases with no change in price or cost, because dilution was never properly accounted for in the mine design criteria.

Our review of mine design criteria at these operations often shows that the dilution allowance in the reserve model was set as a global factor applied uniformly across the deposit. A global dilution factor applied to a heterogeneous transition zone is too coarse. The F-factor data, read over several quarters, will show you that before any drilling program would.

Act on the F-Factor Trend Early

Once the F-factor trend is clear, decide what to do. The answer depends on where you are in the mine plan and how much of the affected material remains in the reserve inventory.

Diluted Areas in the Short-Term Plan

If the diluted sectors are in the short-term plan, the priority is operational. Review how the grade control boundaries (dig lines) are defined. Check whether blast movement is being tracked and whether the grade control polygons are being adjusted for movement after each blast. Examine stockpile management: material that has been misclassified at the shovel and then blended in a stockpile can no longer be traced. In our grade control and reconciliation work, we keep finding that stockpile losses are underreported, because once material enters a stockpile the reconciliation can only be estimated.

Sectors Not Yet Mined

If the diluted sectors represent material that has not yet been mined, the priority shifts to the reserve model. The Competent Person or Qualified Person, depending on the reporting framework, needs to check whether the dilution allowance in the Ore Reserve model should be revised upward for those domains. A higher dilution allowance applied to material close to the cut-off grade will reduce the reserve tonnage and may reclassify some material from Ore Reserve to Mineral Resource, or remove it from the resource entirely. That result is correct. Deferring it only delays the correction while production continues to confirm the gap.

Effect on the LOM Plan

The F-factor reconciliation also affects the life-of-mine (LOM) plan. If F1 (grade) is persistently below 1.0 in transition zones, then the LOM plan is overstating recoverable metal in those zones throughout the remaining mine life, and the financial model built on that plan is overstating expected revenue. In our Strategy & Reserves reviews, this is one of the first things we examine: whether the reconciliation history has been used to test the assumptions for future years in the LOM plan, or whether the LOM plan still uses the original model parameters that the reconciliation data has already contradicted.

Many operations find this step difficult. The reconciliation report is often treated as a record of the past, while the block model and LOM plan are treated as the plan for the future. The connection between them, using reconciliation to update the assumptions for future years, is the step that often does not happen as a routine. The F-factor also tells you about the future, because the same geological conditions that produced dilution last quarter are still present in the ore not yet mined.

The Error Ends Up in the Reserve Statement

When dilution is uncontrolled and unrecognized, the error goes into the reserve statement without anyone seeing it. The signing professional, whether a Competent Person under JORC Code, or a Qualified Person under NI 43-101 or S-K 1300, is signing for an Ore Reserve estimate that includes a dilution assumption. If the reconciliation data shows that assumption is wrong, the reserve statement is not supportable at its current grade and tonnage.

An overstated reserve leads to a mine plan that schedules more ore than is there, at grades that cannot be achieved. Cut-off grade decisions are made against those grades. Fleet sizing, processing plant throughput targets, and workforce plans all follow from the reserve. When production keeps falling short of the plan, the organization often looks first at operational execution. It should also ask whether the reserve assumptions were ever tested against what the mine was producing.

When the Reserve Model Is Too Conservative

A reserve model that underestimates grade shows the opposite error: F1 (grade) or F3 (grade) is above 1.0 every period, and production beats the plan. This can look like good performance, but a high F1 indicates a conservative reserve model (Parker 2012), and the Ore Reserve is then understated. The cut-off grade and the pit design are based on grades lower than what the mine produces, and investment decisions and public reports use the same low figures. The JORC Code asks the Competent Person to disclose uncertainties that could lead to understatement or overstatement of an Ore Reserve (JORC Code 2012, Clause 35), and the CIM estimation guidelines (CIM 2019) ask for checks of global and local bias and a reconciliation of production against the model. A positive variance that repeats every period needs the same review as a negative one.

Read carefully every period, the F-factor reconciliation connects what the model predicted to what the mine produced. It tests the geological interpretation and the reserve estimation. If you act on it early, the correction is smaller.

References

Parker 2012. Parker, H.M. Reconciliation principles for the mining industry. Mining Technology, Volume 121, Number 3, 2012. https://www.mininggeologygroup.com/wp-content/uploads/2019/11/Parker_Recon-principles-for-the-mining-industry_AusIMM_2012.pdf

JORC Code 2012. Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves (The JORC Code, 2012 Edition). Joint Ore Reserves Committee of the AusIMM, AIG and Minerals Council of Australia, 2012. https://www.jorc.org/docs/JORC_code_2012.pdf

CIM 2019. CIM Estimation of Mineral Resources and Mineral Reserves Best Practice Guidelines. Canadian Institute of Mining, Metallurgy and Petroleum, adopted November 29, 2019. https://mrmr.cim.org/media/1129/cim-mrmr-bp-guidelines_2019.pdf

McCarthy 2011. McCarthy, P. Production rate optimisation: avoiding the temptation of tonnage. In Potvin, Y. (ed), Strategic versus Tactical Approaches in Mining 2011. Australian Centre for Geomechanics, Perth, 2011. https://doi.org/10.36487/ACG_rep/1108_01_McCarthy

If your reconciliation data shows a persistent F1 or F2 trend that your reserve model does not yet explain, contact the Agmines team at agmines.com/contact.

By Alejandro Guerra, Agmines Consulting.