Building Trust on Site, Over Time
On active mine sites, trust is not a value…
Read More
→
Report Date: 2026-01-21
Core sheds fail in predictable ways. The failure rarely looks dramatic on the day. It shows up later as rework, missing intervals, confused samples, and data nobody trusts.
I have worked in sheds where the drill rig was the easy part and the shed was the bottleneck. The common pattern is not “bad geology”. It is weak flow control, unclear ownership, and small deviations that become normal.
Rule: If the shed cannot produce clean, traceable data at pace, the drill program is already behind. You just have not paid for it yet.
The core shed is a machine, not a room
A core shed is a high-volume processing line. Input is rock. Output is logged, sampled, traceable data that survives audit. Treating the shed like a quiet workspace creates the wrong incentives. People optimise for comfort instead of throughput and integrity.
When Rock Force deploys to remote programs, we assume the shed will be under-resourced, hot, dusty, and short on time. That is normal. The job is to make the workflow survive those constraints without lowering standards.
The compound interest of error
A one percent failure rate in a 300 m/day campaign is not one percent loss. Shed errors compound because they propagate across systems.
| Failure at source | Next failure | Where it lands |
|---|---|---|
| Mislabeled block or tray | Wrong interval logged | Assay attached to the wrong depth |
| Illegible meter marks | Ambiguous data entry | Reconciliation and resampling |
| Poor orientation control | Structural features misread | Bad geotech interpretation and decisions |
“Precision is not an act. It is a habit. A shortcut taken on Tuesday is a failure discovered in the audit six months later.”
Preparation is production. If the core is not cleaned, oriented, and marked before logging starts, the shed has already accepted downstream risk.
The cost of shortcuts
Most sheds have the same friction points. The difference is whether they are treated as “just how it is” or treated as defects with owners.
| Operational failure | Immediate consequence | Downstream impact |
|---|---|---|
| Poor core orientation | Logger stops to realign intervals | Meterage rate drops and structural confidence degrades |
| Illegible meter marks | Ambiguity at entry | Assay mismatching risk increases |
| Disorganised sampling | Cut sheet errors | Re-splits, delays, and cost blowouts |
| Dirty or unwashed core | Features missed | False negatives in structure and alteration logging |
Workflow architecture: the Rock Force standard
Entropy wins unless you design against it. We run a simple model: linear flow and hard handoffs. Core moves one way from receiving to racking. Work does not double back. Every handoff has a check.
Linear flow: receive → wash → orient/mark → log → sample → dispatch → rack. If you see backtracking, you have a layout problem.
We enforce a “Clean, Mark, Log” protocol because it puts the failure modes up front, where they are cheapest to correct.
Clean, Mark, Log
Clean: Wash and inspect until fractures, fabric, and contacts are visible. Do not “log through dirt” because it saves minutes. You will spend those minutes later, except now it is a resample and an argument.
Mark: Orientation, meter marks, and sample intervals must be readable at arm’s length. If someone needs to guess a number, it is already wrong.
Log: Log into the system you will report from. If you must transcribe, treat it as a controlled step with a second-person check.
The physical setup
Ergonomics and light quality are data quality. Tables must be set to reduce fatigue. Lighting must be consistent and high quality so lithology and alteration calls are not made under glare or shadow.
If a geologist is squinting, they are guessing. If they are guessing, the database is lying.
The data pipeline
Paper notes are not evil. Uncontrolled transcription is. If you have to go paper-first, then you need explicit controls: fixed templates, required fields, and a daily reconciliation loop before the shift ends.
Where possible, we push direct entry to reduce copy steps and keep validation close to observation. The goal is not “digital”. The goal is traceable.
Essential disciplines for modern logging teams
Modern site work is not just identifying minerals. It is managing flow, custody, and decision-grade data under time pressure.
| Discipline | What “good” looks like | How it fails |
|---|---|---|
| QA/QC vigilance | Standards and blanks inserted to plan, logged, and verified daily | “We will catch it later” becomes permanent drift |
| RQD consistency | Same measurement rules every day, regardless of meterage pressure | Speed overrides method and trends become noise |
| Communication loops | Rig ↔ shed ↔ office close-outs with named owners and timestamps | Issues become verbal-only and die at shift change |
Conclusion
Good geology is wasted if the shed produces bad data. The core shed deserves the same respect as any production plant, because its output drives decisions that cost real money.
Rock Force does not treat shed performance as “labour hire”. We treat it as an operational system. When the shed runs clean, your program runs clean. When it does not, every downstream decision inherits risk.
