Drill Core Logging Best Practices: From Geotech to Geological Model

I’ve spent more hours in core sheds than I care to count. The smell of diesel from the drill rig, the hum of the saw, the dust in your eyes — it’s not glamorous work. But it’s the most important work in the entire resource estimation pipeline.

Your geological model is built on three things: drillhole collars, downhole surveys, and core logs. The first two are mechanical. The third — core logging — is interpretive. It’s where the geologist’s knowledge translates into data that the model uses. Get it wrong and no amount of geostatistics will save you.

This post covers the core logging workflow I use on Indonesian projects, from receiving the core to delivering a validated log that feeds directly into the geological model.

The core logging workflow

1. Core receipt and handling

Core arrives at the shed in core boxes, typically 3m per box for NQ core (47.6mm diameter). The first thing I do is verify:

• Box labels match the drillhole program — hole ID, box number, depth range
• Depths are marked on the core — every 1m or 3m with a depth block
• Core recovery is calculated and marked — recovery = recovered length / drilled length
• Core is oriented if structural measurements are needed

Common problem in Indonesian projects: Core is delivered wet, and the mud makes it hard to see lithological contacts. Let it dry for at least 2-3 hours before logging. Wet core hides textures, mineralization, and contact relationships.

2. Geological logging

This is the core of the work — pun intended. The geological log captures:

Lithology: What rock type is this? Not just “andesite” or “dacite” — be specific. “Porphyritic andesite with 15% plagioclase phenocrysts (2-5mm) in a fine-grained groundmass.” The level of detail matters because lithology drives domain separation, and domain separation drives the entire estimate.

Alteration: What has happened to this rock since it formed? For Indonesian epithermal and porphyry systems, I log:

• Alteration type (propylitic, argillic, advanced argillic, silicic, potassic)
• Alteration intensity (1-5 scale: 1=faint, 5=complete replacement)
• Alteration minerals visible (quartz, kaolinite, alunite, chlorite, epidote, biotite)
• Distribution (pervasive, vein-controlled, patchy)

Mineralization: What ore minerals are present, and how?

• Sulfide species (pyrite, chalcopyrite, bornite, sphalerite, galena)
• Sulfide percentage (estimated visually, calibrated against assay)
• Occurrence (disseminated, vein, fracture-controlled, massive)
• Grain size (fine <0.5mm, medium 0.5-2mm, coarse >2mm)

Structure: Any visible structural features

• Veins (type, orientation, width, fill)
• Fractures (type, density, orientation if core is oriented)
• Faults (gouge, breccia, slickensides)
• Contacts (gradational, sharp, sheared)

3. Geotechnical logging

This is often skipped or done poorly, but it’s critical for open pit design and underground geomechanics:

• RQD (Rock Quality Designation): Sum of intact pieces longer than 10cm, divided by total run length. RQD >75% is good ground. RQD <25% is very poor.
• Fracture frequency: Count of fractures per meter
• Hardness: 1-5 scale (1=can scratch with fingernail, 5=can’t scratch with knife)
• Weathering: 1-5 scale (1=fresh, 5=complete residual soil)

For Indonesian deposits with deep weathering profiles, the weathering log is especially important. The transition from fresh rock to residual soil can happen over 5-10m or 50-80m, and the boundary determines where open pit mining stops being practical.

4. Sampling

Sampling decisions made during logging affect the entire estimate:

• Sample intervals should respect geological boundaries. Don’t split a sample across a lithological contact.
• Sample length should be consistent within a domain — typically 1m, but 0.5m for narrow veins and 2m for disseminated zones.
• Sample the entire mineralized zone plus 3-5m of hangingwall and footwall. This captures the grade gradient at the boundary.
• Don’t sample obvious waste — but document why you didn’t. “No sample — barren quartzite, logged but not sampled” is better than a gap in the assay record.

The half-core rule: Split the core longitudinally. One half goes to the lab, the other stays in the box as a reference. Use a diamond saw — not a chisel — to ensure the split is clean and representative. For high-grade zones, quarter-core is sometimes used, with one quarter for assay, one for metallurgical testwork, and the half retained.

5. Photography

Photograph every core box before and after logging. Use consistent lighting, a scale bar, and a color reference card. Photos are:

• A permanent record of what the core looked like at the time of logging
• A reference for re-interpretation if the geological model changes
• Evidence for due diligence — if a buyer asks “show me the core from the high-grade zone,” you need to produce the photo

Indonesian-specific tip: High humidity in core sheds causes condensation on camera lenses. Keep silica gel packets in your camera bag and let the lens acclimate for 10 minutes before shooting.

Common logging mistakes

“Copy-paste” logging

The drillhole above had “andesite” from 0-200m, so the one below probably does too. No. Log every meter independently. I’ve seen projects where 80% of the logs were identical because the logger was copying from adjacent holes. The geological model was useless.

Logging from memory

“I logged a similar deposit last year, I know what the alteration looks like.” You don’t. Log what’s in the box in front of you, not what you remember. Mineralogical assemblages vary significantly between deposits, even within the same district.

Inconsistent terminology

Logger A calls it “argillic alteration.” Logger B calls it “clay alteration.” Logger C calls it “advanced argillic.” All three are looking at the same rock. Now your domain model has three domains for one alteration zone.

Fix: Create a project-specific logging code sheet before drilling starts. Every logger uses the same codes. Review logs weekly for consistency.

Not calibrating visual estimates

If you log “5% pyrite,” does that mean 5%? Most geologists underestimate sulfide percentages — a 5% visual estimate is often 8-12% by weight. Calibrate your visual estimates by comparing them to assay results. If your “5% pyrite” consistently assays 10% S, adjust your estimates.

Ignoring the transition zone

The oxide-sulfide transition in Indonesian deposits is where most geological models fail. The boundary is irregular, the mineralogy is mixed, and the metallurgical recovery changes dramatically. Log the transition zone in detail — separate logging codes for “oxide,” “mixed,” and “sulfide,” with sub-codes for the degree of oxidation.

The digital workflow

Modern core logging uses tablets or laptops directly in the core shed. This eliminates transcription errors and enables real-time validation:

• Depth checks — the system flags overlapping or missing intervals
• Code validation — typos are caught immediately
• Visual summaries — strip logs generated on the fly show patterns and inconsistencies

If you’re still logging on paper and transcribing later, you’re adding a full day of work per hole and introducing transcription errors. A ruggedized tablet with logging software pays for itself in a month.

The bottom line

Core logging is not data entry. It’s geological interpretation at the most fundamental level. Every decision you make at the core shed — every contact you draw, every alteration code you assign, every sample interval you define — propagates through the entire resource estimation pipeline and shows up in the final block model.

Take the time. Log consistently. Document your reasoning. Photograph everything. The geologists who do this well produce models that survive from exploration through to production. The ones who don’t spend years trying to explain why the model doesn’t match the mine.