Aller au contenu
SGS logo SGS Geological Services

Drillhole Data Validation

Why Validation Matters

Validation is a key step before interpretation, compositing, and estimation.

A simple rule applies: garbage in, garbage out.

If source data is inconsistent, overlapping, orphaned, or physically suspect, downstream results can be biased or wrong.

Where Validation Runs

GENESIS validates drillhole data in two places:

  1. During import (loader validation + import rules)
  2. From the Custom Tables Validate menu (database validation on data currently stored in GENESIS)

These two layers are related, but not identical.

Import Validation vs Database Validation

Import validation (during load)

Import workflows apply an extra validation/filtering layer before records are committed.

Examples:

  • invalid rows can be rejected
  • duplicate/conflicting rows can be skipped
  • issues are reported in import results/validation log

Because rejected rows are not inserted, later database validation cannot see them.

Database validation (Validate menu)

The Validate tools analyze what is already inside the GENESIS database.

So if bad rows were skipped at import time, this validation will not report those skipped rows.

Validate Menu Options (Custom Tables)

In Custom Tables:

  • Validate > Fixed Tables validates all fixed drillhole tables.
  • Validate > Selected Tables validates only selected tables.

What Validation Checks

Depending on table type, validation checks include:

1) Duplicate holes

Detects holes that have identical start/end trace geometry.

2) Interval overlap checks (From-To tables)

Finds overlapping intervals in the same hole.

  • For lithology, checks are level-aware.

3) From/To consistency

Flags records where FromDepth >= ToDepth.

4) Gap checks (when enabled)

Finds depth gaps between consecutive intervals in the same hole (and same lithology level when relevant).

5) Duplicate depth checks (Depth tables)

Finds duplicate depth records for the same hole.

6) Orphan hole references

Finds rows where HoleName does not exist in collars.

7) Suspect deviations

Checks deviation progression by depth and flags abrupt/suspect changes in azimuth/dip behavior. Also includes duplicate depth checks in deviation data.

8) Negative value checks

Checks numeric variables for negative values where negatives are usually invalid.

Notes:

  • Missing-value sentinels are excluded from this warning logic.
  • Some fields (such as coordinates and dip) are excluded from generic negative checks.
  • Legacy deviation azimuth missing code (-999) is also treated as a special case.

Validation Output

Validation produces a structured report with sections by rule/table.

Typical outcome:

  • No Error Found, or
  • detailed warnings/errors with hole and depth/location context.
  1. Import data.
  2. Review import summary and import validation output.
  3. Run Validate > Fixed Tables.
  4. Fix issues in source or tables.
  5. Reimport/revalidate before modeling tasks (intervals, composites, estimation).