Real Volve DLIS composite logs, ingested incrementally into an analysis-ready Delta table. Two deliveries: LWD composites for 15/9-F-9 (1204 frames) and 15/9-F-11 (1637 frames) on 11 March, then the wireline composite for 15/9-F-15 C (19791 frames) on 12 March, which is six channels wider and forces an ALTER TABLE ADD COLUMN before it can land. DISCOVER registers each drop, recognising DLIS from its storage unit label rather than its extension; the channel names come from the CHANNEL and FRAME sets inside the file, which is the only place a DLIS frame's meaning is written down. Three real properties of the data drive the demo: the second delivery is wider than the first, DEPTH is recorded in tenths of an inch in two files and millimetres in the third, and every channel writes -999.25 where no reading was taken. The loader resolves all three on the way in and keys on the source file, so the demo runs one delivery twice and asserts the count did not move. Finishes with a petrophysical evaluation: 280 net pay samples between 3126 and 3220 m, which is the Hugin Formation. Data from the OSDU Forum open test data; see ATTRIBUTION.md.
-- ============================================================================
-- Volve Composite Log Ingestion - Setup Script
-- ============================================================================
-- A petrophysicist is building an analysis-ready log table for the Volve
-- field. The logs arrive as DLIS composites in two waves, from two different
-- tool strings:
--
-- 11 March LWD composites for 15/9-F-9 and 15/9-F-11. Six channels:
-- DEPTH, GR, RDEP, RMED, ROP, BS. Recorded while drilling.
-- 12 March The wireline composite for 15/9-F-15 C. Twelve channels: the
-- six above plus CALI, DEN, DENC, PEF, NEU and AC, which is the
-- triple-combo suite a petrophysical evaluation needs.
--
-- The second wave is wider than the first, so the curated table has to grow
-- when it lands. That is not a contrived demo step: it is what happens every
-- time a wireline run follows an LWD pass.
--
-- The data is real. These are the Volve composite logs Equinor released and
-- the OSDU Forum redistributes in its open test data; see ATTRIBUTION.md in
-- the parent folder. Nothing here was written for the demo, which is why the
-- awkward parts are present: the depth channel is recorded in tenths of an
-- inch in two of the files and in millimetres in the third, and every channel
-- uses -999.25 for a reading that was not taken.
--
-- 1. lwd_composite external, DISCOVER over the LWD drop
-- 2. wireline_composite external, DISCOVER over the wireline drop
-- 3. well_logs DELTA, the curated target, depths in metres
-- ============================================================================
-- ----------------------------------------------------------------------------
-- STEP 1: Zone and schema
-- ----------------------------------------------------------------------------
CREATE ZONE IF NOT EXISTS {{zone_name}} TYPE EXTERNAL
COMMENT 'External tables - demo datasets and file-backed data';
CREATE SCHEMA IF NOT EXISTS {{zone_name}}.petrophysics
COMMENT 'Volve composite logs read in place from DLIS, curated into Delta';
-- ----------------------------------------------------------------------------
-- STEP 2: Register both drops with DISCOVER
-- ----------------------------------------------------------------------------
-- DLIS is identified from the storage unit label and the visible-record
-- structure, not from the file extension. No column is named by hand here:
-- the channels come from the CHANNEL and FRAME sets inside each file, which
-- is the only place they exist.
--
-- The two tool strings are registered separately because they are separate
-- deliveries with separate channel sets, and keeping them apart is what lets
-- the loader treat the wider one as the schema change it is.
-- ----------------------------------------------------------------------------
DROP EXTERNAL TABLE IF EXISTS {{zone_name}}.petrophysics.lwd_composite;
DISCOVER {{zone_name}}.petrophysics.lwd_composite
PATH '{{data_subdir}}/landing/lwd'
WITH (FILE_METADATA = true);
DROP EXTERNAL TABLE IF EXISTS {{zone_name}}.petrophysics.wireline_composite;
DISCOVER {{zone_name}}.petrophysics.wireline_composite
PATH '{{data_subdir}}/landing/wireline'
WITH (FILE_METADATA = true);
-- ----------------------------------------------------------------------------
-- STEP 3: The curated Delta target
-- ----------------------------------------------------------------------------
-- Created with the LWD channel set only, because on 11 March that is all
-- there is. The wireline columns are added in queries.sql at the point the
-- wireline delivery lands, which is the honest order of events.
--
-- depth_m is the reason this table exists. The DLIS files record depth in
-- whichever unit the logging contractor's system used, and two of these three
-- disagree;
-- (excerpt; run the demo for the full script)## What this demo does Real Volve DLIS composite logs, ingested incrementally into an analysis-ready Delta table. Two deliveries: LWD composites for 15/9-F-9 (1204 frames) and 15/9-F-11 (1637 frames) on 11 March, then the wireline composite for 15/9-F-15 C (19791 frames) on 12 March, which is six channels wider and forces an ALTER TABLE ADD COLUMN before it can land. DISCOVER registers each drop, recognising DLIS from its storage unit label rather than its extension; the channel names come from the CHANNEL and FRAME sets inside the file, which is the only place a DLIS frame's meaning is written down. Three real properties of the data drive the demo: the second delivery is wider than the first, DEPTH is recorded in tenths of an inch in two files and millimetres in the third, and every channel writes -999.25 where no reading was taken. The loader resolves all three on the way in and keys on the source file, so the demo runs one delivery twice and asserts the count did not move. Finishes with a petrophysical evaluation: 280 net pay samples between 3126 and 3220 m, which is the Hugin Formation. Data from the OSDU Forum open test data; see ATTRIBUTION.md. ## What it creates - `{{zone_name}}.petrophysics.lwd_composite` (external table) - `{{zone_name}}.petrophysics.wireline_composite` (external table) - `{{zone_name}}.petrophysics.well_logs` (delta table) ## Running it The demo ships `setup.sql`, `queries.sql` and `cleanup.sql` under `demos/subsurface/dlis-wireline-petrophysics`. The queries carry their own `ASSERT`s, so a run that completes has verified its own numbers. Data: 3 files, 1.2 MB.