DirtMate Centimeter-level machine guidance and fleet telematics for earthmoving
AeroPoints High-accuracy ground points for faster, reliable surveys
Propeller CAD AI-powered earthworks design built directly on your real survey terrain
Drone & hardware integrations Connect Propeller with your existing tools seamlessly
Stockpile inventory numbers become hard to defend when methodology, density, cadence, or provenance drift between surveys. This guide maps the six most common drivers and the structural fixes that make your reported volumes reproducible and traceable.
If your stockpile volumes never quite match your books, and the gap is hard to explain, you are not dealing with a measurement problem. You are dealing with a documentation problem. The measurement was probably fine. The trail behind it was not built to hold up.
This shows up most often in aggregates and mining operations, where the same pile carries a surveyed volume, a book figure, and a plan figure that rarely line up to the decimal. There are six common reasons the number becomes indefensible:
Survey methodology drifts between periods without being noticed. Density assumptions are not validated for the actual material. Survey cadence leaves too big a gap for variance to be traced. Rehandles and pile transfers go unrecorded. Volume figures are disconnected from their source through spreadsheet handoffs. And boundary or base surface definitions change quietly between surveys.
None of these are exotic. All of them are fixable. The first step is knowing which one you’re dealing with.
A stockpile number is defensible when three things are true:
You can explain it. You know what methodology produced it, what density was applied, and what the survey covered.
You can reproduce it. Someone who was not in the room when it was produced can re-run it from the documentation alone, the same way a survey or engineering reviewer would.
You can trace it. The number connects back to its source survey, not to a cell in a spreadsheet that has since been overwritten.
Most operations can explain their stockpile numbers. Fewer can reproduce them. Almost none can trace them through six months of reporting periods without making a call. That gap is where reconciliation disputes, audit findings, and year-end write-downs live.
On any given day, a stockpile has at least three numbers attached to it. The surveyed volume: what the drone model says is physically there, converted to tons. The book inventory: what the ERP or spreadsheet says should be there, based on opening balance plus production minus sales, rehandles, and waste. And the plan system: what the mine or quarry plan says should have moved, based on production targets and sequence.
All three are constructed numbers. All three can be internally consistent. And all three can disagree.
Measuring a stockpile gives you a volume at a point in time. Inventory reporting turns that volume into a number that someone else will rely on, for a royalty calculation, a finance close, an audit, or a JV partner’s books.
The gap between those two jobs is where most stockpile disputes start. The measurement was accurate. The reporting trail was not built to hold up.
Every indefensible number traces back to one or more of these. They compound quietly, and the terrain model looks perfect the whole time. Here is what each one does to your reported volume.
If surveys happen monthly and sales are reported weekly, there is always a period of unconfirmed inventory. If the survey falls on the 15th and the reporting period closes on the 31st, two weeks of production and dispatch are reconciled against a 16-day-old snapshot. That is a timing gap rather than a measurement error, but it shows up as unexplained variance every single period.
Operations that survey every two weeks or weekly can trace variance while the operational context is still live. Operations that survey quarterly are often explaining a gap nobody can reconstruct. Tightening survey cadence is the single cheapest fix on this list.
Method drift is the most persistent source of unexplained variance, and the least visible. If the polygon boundary around a stockpile shifts between surveys, even a few meters at the toe, the volume delta mixes real material movement with boundary geometry. You cannot separate them after the fact.
If the base surface method changes from smart base to reference level, or from reference level to custom base, the volume comparison is no longer measuring the same thing. A change that looks like a minor processing decision can produce a 5–15% difference in reported volume for the same physical pile. Neither of these shows up in the terrain model quality. The model is accurate. The method is inconsistent. The output becomes unreliable.
Surveyed volume is converted to reported tonnage using a density figure. If that figure does not reflect the material being measured, it can distort the reported tonnage. Published bulking factors, such as roughly 1.12 for sand and gravel and around 1.50 for blasted hard rock, can provide a starting point, but they are not site-specific. Moisture, blasting patterns, fragment size, and material grade all move bulk density over time.
If density assumptions are not periodically checked against the actual material, the difference between the conversion figure and site conditions can carry through multiple reporting periods. It appears as inventory variance even when part of the difference comes from the density assumption itself.
Related to density, but often tracked separately: the swell factor applied when material is moved, and the shrinkage factor applied when it is compacted or processed. A loader operator, a stockpile, and a haulage route can each interact with swell differently.
If the swell factor used in reconciliation does not reflect the actual material state, whether bank cubic meters, loose cubic meters, or compacted cubic meters, the tonnage conversion can be consistently overstated or understated across reporting periods.
An unrecorded rehandle can appear as a loss from one pile and a gain in another. If both piles are on the same site and total inventory stays unchanged, the movement can be hard to spot until the individual pile is reviewed for something like a royalty calculation or audit.
Rehandles happen when material moves between product grades, is blended for quality, is shifted to make room, or is transferred from a temporary pile to a permanent one. Recording those movements in the book record keeps the inventory trail intact.
One common break in the chain of evidence happens when a volume is copied out of the measurement platform and into a spreadsheet or report. Once separated from the original survey, the number can lose important context, including the methodology used, the density applied, and the reporting period it represents.
If that context is not carried forward with the value, it becomes harder to trace or verify later. Keeping the source survey and calculation details connected to the reported number makes future review and reconciliation much easier.
The fix for all six drift drivers is structural, not technological. Three habits do most of the work.
Each pile should have a documented volume method, boundary rule, and base surface definition that is enforced across every survey, regardless of who processes it. A consistent methodology makes it far easier to tell whether a variance reflects a real inventory change or just a difference in how the pile was measured.
Where possible, keep stockpile volumes connected to the measurement platform and original survey rather than moving them into spreadsheets without supporting context. Anyone reviewing the number should be able to see where it came from and how it was calculated. Reports that carry the pile, volume, tonnage, material, methodology, and source survey together give you a clearer record for review and reconciliation, and they make sharing that record across the team painless.
Reconciliation does not require a complex process. It requires a consistent one. At the end of every reporting period, document which survey was used and what methodology was applied, record density and conversion assumptions, compare measured and book figures with sources identified, document the reason for any known variance, and note any follow-up actions and owners. That documentation is the audit trail. It does not have to be elaborate. It just has to exist. Tying it to your reporting rhythm keeps it from slipping.
Answer yes or no to each question. It doubles as a quick site check on your reporting process.
Do we survey at least monthly for all production-active piles?
Is the volume calculation method (smart base, reference level, or custom base) documented per pile and enforced consistently?
Does our polygon boundary reflect the current physical extent of each pile after every survey?
Is our bulk density figure based on site-validated data, not a published table reference?
Is our density figure reviewed and updated at least quarterly?
Are swell and shrinkage factors applied consistently on both the survey side and the book side?
Are all rehandles and inter-pile transfers recorded in the book inventory?
Do stockpile volumes stay connected to their source survey, not copied into standalone spreadsheets?
Can a colleague who was not involved re-run last period’s reconciliation from documentation alone?
Are all variance explanations written down before the next reporting period begins?
Watch the webinar (on demand): From measured to defensible: how to reconcile stockpiles with confidence. See the reconciliation workflow in action, including common sources of variance, survey cadence, and ways to create a clearer reporting trail. The session also includes audience Q&A on vertical faces, vegetation cleanup, and export options.
Read the full playbook: The Defensible Stockpile: a reconciliation playbook for aggregates operators. A practical guide to methodology, provenance, and measurement accuracy, with a reconciliation template and audit trail checklist for your team.
Go deeper on the workflow:
How to build a defensible stockpile reconciliation process, a step-by-step workflow with a root-cause investigation guide.
Stockpile reconciliation FAQs, 15 common questions on measurement, density, reporting, and audit trails.
If surveyed tonnage consistently trends higher or lower than book inventory, density may be one factor to investigate. First check for differences in survey timing, boundaries, base surfaces, and material movements. Then compare the density figure being used against a recent material test or other site-specific reference.
Whether drone measurements can be used for royalty or audit purposes depends on the reporting requirements that apply to your operation. Measurement accuracy is only part of the equation. A consistent methodology, clear documentation, and a traceable connection to the source survey all make reported figures easier to review and support. A high-precision ground control workflow with PPK drone mapping helps on the accuracy side, but the documentation trail is what makes a figure defensible.
Variances that cancel out at the total-site level but are large at the individual pile level usually point to boundary or base surface inconsistency. Two piles that share a boundary zone, or one pile with a custom base applied inconsistently, will trade volume with each other across surveys while the site total stays roughly correct. The fix is a methodology review at the pile level, not the site level.
As often as you survey. Reconciliation run monthly or every two weeks catches variance while the operational context is still traceable. Reconciliation run annually surfaces variance that has no living explanation. Match your reconciliation frequency to your survey cadence.
We’re happy to show you how Propeller can power your worksite, and boost productivity.
