Core Logging Best Practices: A Field Guide to Reliable Structural Data

Core logging best practices are the routines that keep a drill core log complete, consistent and traceable. In practice, that means controlling depth and recovery in the core shack, checking the orientation reference before measuring, recording structural measurements with one documented convention, linking each measurement to the hole’s trajectory data, and reviewing the log while the core is still available.

These routines matter because the core log is the primary record that resource geologists, geotechnical engineers and modelers build on. The JORC Code (2012 Edition) reflects this: its Table 1 checklist asks whether core has been geologically and geotechnically logged to a level of detail that supports Mineral Resource estimation, mining studies and metallurgical studies.

This guide concentrates on structural logging of oriented core, from the core box to a reviewable dip and dip direction, and on the checks that keep it consistent with the geologic and geotechnical log. For the fundamentals, start with our guide to core logging.

Key takeaways

• Resolve depth, recovery and core loss before you measure anything.
• Treat the orientation line as data with a confidence level, not as a given.
• Store the measurement convention with the data, not only in a manual.
• A dip and dip direction is only as reliable as the trajectory used to calculate it. Record which trajectory you used.
• Review structural data while the core is still on the table.

What a reliable core log contains

A core log is a sequence of observations recorded against depth. Whether the focus is geological core logging, geotechnical core logging or structural logging, a detailed log for mineral exploration usually has three layers, each serving a different downstream user.

Log layer	What it records	Why it matters downstream
Geologic logging	Lithology, alteration, mineralization, contacts and veins	Defines geological domains and supports the interpretation behind a resource model
Geotechnical logging	Core recovery, RQD, fracture frequency, fracture condition, weathering and strength	Characterizes rock mass quality for mine and excavation design
Structural logging	The type and orientation of faults, fractures, veins, bedding, foliation and other planar features	Constrains structural interpretation and provides the joint orientation data used in rock slope design

All three layers share one requirement: observation must stay separate from interpretation. A 2026 U.S. Geological Survey method for logging drill core, written for its core library at the Idaho National Laboratory, describes the log as a series of observations that aims to minimize interpretation, and it identifies consistency as the most important attribute of a log. If a feature is described in one part of the hole, it should be described the same way everywhere else.

Interpretation still belongs in the dataset. It simply needs its own fields, so that a later reviewer can tell what the logger saw from what the logger concluded.

Before you log: core handling and depth control

Structural measurements inherit every error in depth and core order. The core logging procedure below adds a few minutes per box and prevents problems that are very difficult to fix later.

1. Confirm box order and way-up. Check that boxes run shallow to deep and that every piece sits the right way up, in the correct downhole direction, in its row. The USGS method recommends using as many orientation and depth indicators as possible, because core can be misplaced or mislabeled.
2. Reconcile run blocks and meter marks. Compare the depths on the run blocks with the driller’s records and mark meter intervals on the core or the box. Resolve discrepancies before logging starts, not after.
3. Record recovery and core loss. Calculate core recovery for each run and record the from-to interval of every core loss. Rubbly zones and lithologic contacts deserve extra attention, because they often have low recovery and are hard to measure.
4. Separate mechanical breaks from natural fractures. Drilling-induced breaks are typically clean and fresh, without weathering, oxidation or mineral coatings. Mark them so they are not logged as natural fractures.
5. Photograph every box before logging. Use consistent lighting and include a label with the hole ID and depth interval in each photo. Wet core photos improve the visibility of textures and structures, and dry photos help with color and archive consistency.
6. Keep the original image files. Store unedited originals as a backup before any renaming, cropping or color correction.

JORC Table 1 lists core photography alongside its logging criteria, so the photo set is part of the evidence trail, not a courtesy.

Check the orientation reference before you measure

An orientation tool marks or records the bottom-of-hole (or top-of-hole) position of the core at the end of a run. The core is then reassembled and a core orientation line is drawn along the run. In oriented core logging, every beta angle you measure depends on that line. For how the orientation itself is obtained, see our article on core orientation methods and tools.

Core orientation is a multi-step process, and each step can introduce error. Cylwik, Ryan and Cicchini (2011) list the error sources introduced by the orientation process: measuring the in situ orientation, marking it on the core, reassembling the run and drawing the reference line. These errors affect the beta angle. The alpha angle is affected only by how carefully it is physically measured.

Practical checks before you measure:

• Compare consecutive runs. Where two consecutive orientation marks can be compared across a reassembled run, measure and record the angular mismatch. Cylwik and co-authors call this the difference angle and use it to quantify part of the orientation error.
• Apply a written tolerance. Decide in advance what mismatch is acceptable and what happens when a run exceeds it, for example treating the run as low confidence or unoriented.
• Do not carry the line through broken intervals. If pieces cannot be fitted together, the line cannot be reliably extended. Flag the interval.
• Record a confidence flag for every run. It travels with every structure measured in that run.

For unoriented or low-confidence intervals, alpha angles are still worth recording. They constrain the possible orientation of a structure even when beta is unreliable.

Structural logging step by step

With depth and orientation checked, log each structure in the same sequence.

1. Locate the structure and record its depth. Apply one depth rule throughout, for example the depth at which the structure crosses the core axis.
2. Classify it from a controlled list. Use agreed codes such as fault, fracture, joint, vein, veinlet, bedding, foliation, cleavage, shear zone or contact, each defined in the logging manual.
3. Measure the alpha angle. Alpha is the acute angle between the structural plane and the core axis, from 0° (parallel to the axis) to 90° (perpendicular to it).
4. Measure the beta angle. Beta is the angle around the core circumference between the core orientation line and the reference point on the structure’s ellipse, read in the direction and from the reference point that your project convention defines.
5. Record the attributes your program needs. Typical examples include infilling, aperture, roughness and vein composition.
6. Photograph the structure. Keep the orientation line and the depth label visible in the frame.
7. Carry the run’s orientation confidence onto the record.

Document the convention with the data

There is no single convention for core angles. Cylwik and co-authors note that many conventions exist for measuring the angles needed to orient joints. A beta convention is only fully defined by four elements together:

• Start point: where the measurement begins (the reference line, or a point on the structure’s ellipse).
• End point: where it ends.
• Rotation direction: clockwise or counterclockwise, and the viewing direction (for example, looking downhole).
• Reference-line type: whether the reference is the bottom-of-hole line, the top-of-hole line, or a separate reference line whose angle to the orientation mark is recorded.

Conventions that look different can describe the same angle. Cylwik and co-authors measure beta counterclockwise, looking in the direction of drilling, from the downhole tip of the ellipse to the reference line. Measuring clockwise, looking downhole, from the reference line to the downhole tip gives the same value, because both the direction and the start and end points are reversed. A change to any single element on its own, however, changes the number. Mixing conventions without recording them is what corrupts a dataset.

Store these choices with every dataset:

• the reference-line type (bottom-of-hole, top-of-hole or a separate reference line);
• the start point, end point and rotation direction used for beta;
• how alpha is defined and read;
• the depth rule;
• the sign convention for hole inclination (in many mining datasets, negative values indicate downward holes).

Common measurement errors

• Measuring from the wrong line. Cut lines and saw lines are not orientation lines. Use distinct colors or symbols for each.
• Reading the wrong point on the ellipse. If the convention specifies the downhole apex, reading the opposite apex gives a beta that is 180° off.
• Coarse rounding. Read angles carefully, to the nearest degree where the method allows. Cylwik and co-authors stress careful measurement with precision instruments and open communication with the driller.
• Overlooking low-alpha sensitivity. For structures at a small angle to the core axis, a small beta error can shift the calculated dip and dip direction substantially.
• Skipping damaged low-alpha structures. Cylwik and co-authors report that joints at low alpha angles are commonly damaged by drilling and are typically bypassed during logging unless loggers are instructed otherwise. Tell loggers to record them, with a condition flag, rather than skip them.
• Logging through lost orientation without a flag. Data from those runs can look perfectly normal in the database.

From alpha and beta to dip and dip direction: why trajectory matters

Alpha and beta describe a plane relative to the drill hole. To express that plane in real-world coordinates, you also need the hole’s azimuth and inclination at the depth of the structure. Stigsson and Munier (2013) describe this as a calculation with four parameters, and show that the uncertainty of the resulting orientation combines the uncertainty of all four.

This has two practical consequences.

The trajectory you use is part of the measurement. Drill holes deviate from their planned path. If you calculate dip and dip direction from the planned azimuth and inclination where the hole has deviated, that difference passes directly into the structural data. Once downhole survey data are available and validated, recalculating with the measured trajectory is the more defensible choice.

Raw data must survive the calculation. Keep alpha, beta, depth and the convention in the database alongside the calculated dip and dip direction. Record which trajectory was used (planned or surveyed) and which survey file version. That allows recalculation when final surveys arrive, and it allows a reviewer to reproduce any value.

Two further points deserve a line in the logging procedure:

• Between survey stations, the hole orientation is interpolated, so survey spacing affects the trajectory assigned to each structure.
• Define how depths below the deepest survey station are handled, rather than leaving it to software defaults.

JORC Table 1 also asks about the accuracy and quality of the collar and downhole surveys used to locate drill holes, so survey quality is a reporting matter as well as a structural one.

Geotechnical essentials: recovery, RQD and fracture frequency

Structural and geotechnical logs describe the same core, so they need to agree on depth intervals and on which breaks are natural.

Parameter	What it measures	Watch for
Core recovery	Length of core recovered compared with the drilled run length	Unrecorded core loss and run blocks placed at the wrong depth
Rock Quality Designation (RQD)	Percentage of a run made up of sound core pieces at least 100 mm long, measured along the core centerline	Mechanical breaks counted as natural, and inconsistent measurement of piece length
Fracture frequency	Number of natural fractures per unit length of core	Drilling-induced breaks included in the count

RQD dates from the 1960s. In their 1989 review for the U.S. Army Corps of Engineers, Deere and Deere described it as a modified core recovery percentage that counts only sound pieces of core 100 mm or longer, measured along the centerline, and noted that NQ and larger core diameters are appropriate. They also cautioned that RQD is not a design parameter that stands alone and should be read together with the geology.

Decide and document how mechanical breaks are handled before logging starts, because that rule directly changes both RQD and fracture frequency.

QA/QC for core logs

Logging involves judgment. In a 2021 study of legacy drill hole logs, Joshi and co-authors describe geological logging as subjective, shaped by the purpose of the study, the lack of a common standard between projects and companies, and each geologist’s personal biases. QA/QC cannot remove judgment, but it can make it consistent and visible.

Check	When	What it catches
Logging manual with controlled code lists and definitions	Before drilling starts; versioned when it changes	Different terms for the same feature between loggers
Calibration logging: two loggers log the same interval and compare	At program start, when new loggers join, and periodically	Drift in how individuals apply the codes
Stereonet review of new structural data, plotted as if the hole were vertical	Daily, while the core is still available	Runs with bad orientation, which plot as circles centered on the net
Screening of data by logger, rock type or RQD	Weekly or per hole	Person-specific or domain-specific bias
Consecutive-run orientation mismatch	Every run where it can be measured	Error sources introduced by the orientation process
Validation at data entry: interval gaps and overlaps, value ranges, code lists	At entry	Transcription and keying errors
Photo check of a sample of measurements	Weekly or per hole	Wrong reference line or wrong apex
Change log on edited records	Always	Untracked corrections

The stereonet and screening checks follow Cylwik and co-authors, who recommend inspecting oriented-core data during the drilling program, while quality control can still be performed. The data entry checks correspond to JORC Table 1’s data-verification and database-integrity criteria, which cover documentation of primary data, data entry and verification procedures, and measures to prevent transcription or keying errors between collection and use.

Manual vs. digital structural logging: what changes and what doesn’t

Moving from a protractor or goniometer and a paper sheet to a digital workflow does not change the geometry, the need for a valid orientation line or the need for QA/QC. It changes how the record is captured and how easily it can be checked. That shift is the subject of our overview of digital core logging.

Aspect	Manual measurement and log sheet	Image-linked digital workflow
What is stored	Numbers on a sheet or spreadsheet	Numbers stored with an image of the structure and its orientation reference
How a value is reviewed	Return to the core, if it is still accessible	Reopen the image, then re-check or reprocess the measurement
Data transfer	A separate transcription step	Captured at the point of measurement, so a separate transcription step can be avoided
Trajectory link	A separate calculation step	Calculated in the workflow; the workflow should record which trajectory was used (planned or surveyed)
Still required	A valid orientation line, intact core, a trained logger and QA/QC	The same

Image-assisted logging apps are one way to build this kind of record. SPT’s StructMaster™ structural logging app, for example, photographs a structure together with its orientation reference, detects the structure and the orientation line, and calculates alpha and beta. Optional manual adjustment and a fully manual mode are available. The image stays with the measurement, so a saved log can be reopened, corrected or reprocessed. The app includes eleven default geological structure types.

For dip and dip direction, it uses the hole’s planned azimuth and inclination by default. After synchronizing with SmartCloud™, it can use the hole’s measured downhole survey data instead, and the user chooses which. StructMaster™ does not orient the core: it relies on the orientation line produced upstream.

For a closer look at that workflow, see digital oriented core logging with StructMaster™.

Limitations and technical considerations

• No logging workflow can recover an orientation that was never valid. Manual or digital, beta angles are only as good as the orientation line.
• Broken, rubbly or ground core limits what can be measured. Where pieces cannot be fitted together, record what you can (depth, structure type, alpha) and flag the orientation as unavailable.
• Low-alpha structures are under-sampled for two separate reasons. A drill hole samples like a scanline: structures at a small angle to the hole are intersected less often, and those nearly parallel to it are rarely intersected at all (a blind zone). On top of that sampling bias, damaged low-alpha joints may be bypassed during logging. Interpret their orientations and frequency with care.
• Survey quality limits the quality of calculated orientations. Uncertainty in hole azimuth and inclination carries into every dip and dip direction.
• Image-based measurement depends on image capture. Lighting, camera alignment and a clearly visible orientation line all affect the result, and an image does not make poorly oriented core valid.
• Merged datasets need a common convention. Combining programs that used different alpha, beta or reference-line conventions, without converting them, corrupts the dataset.

Next step

If your team is evaluating an image-linked structural logging workflow, you can request a quote for StructMaster™ or download the StructMaster™ brochure (PDF).

Sources

• JORC (2012). Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves (The JORC Code), 2012 Edition. Table 1, Sections 1 and 3. jorc.org. A revised Code is under review; the 2012 Edition is the version referenced here.
• Dietz, H. (2026). Standardized method for logging drill core at the Idaho National Laboratory, Idaho. U.S. Geological Survey Techniques and Methods 5-E1. doi:10.3133/tm5E1
• Cylwik, S., Ryan, T.M., and Cicchini, P.F. (2011). Error quantification in oriented-core data and its influence on rock slope design. Slope Stability 2011, Vancouver.
• Stigsson, M., and Munier, R. (2013). Orientation uncertainty goes bananas: An algorithm to visualise the uncertainty sample space on stereonets for oriented objects measured in boreholes. Computers & Geosciences 56, 56-61. doi:10.1016/j.cageo.2013.03.001
• Deere, D.U., and Deere, D.W. (1989). Rock Quality Designation (RQD) after twenty years. U.S. Army Corps of Engineers Contract Report GL-89-1.
• Joshi, R., Madaiah, K., Jessell, M., Lindsay, M., and Pirot, G. (2021). dh2loop 1.0: an open-source Python library for automated processing and classification of geological logs. Geoscientific Model Development 14, 6711-6740. doi:10.5194/gmd-14-6711-2021