NSTA National Data Repository (NDR)

It is where the United Kingdom's offshore petroleum record lives by law rather than by choice. The North Sea Transition Authority (NSTA) operates the National Data Repository inside its Digital Energy Platform as the statutory archive for offshore petroleum-related licence information, with carbon dioxide appraisal and storage licence records being added over time alongside the conventional record.

Organisation is the archive's best idea. Every item hangs off a unique Project ID following the fixed pattern CCCCYYYYtypeNNNN - company, year, a project-type code (well = well description, seis = seismic survey, plus mhaz, rems and inpt), and an incrementing number. Decode one and you know who lodged the data, when, and roughly what it is before a single file opens. Files inside a project carry Classification Tags, and each information type carries one of three availability states - Available, Unavailable or Not Acquired - which makes the archive a completeness registry as much as a file store.

Scale sits at petabyte class, commonly cited around 1.6 PB - a figure that circulates widely even though the portal itself declines to publish one. Within Datadory's catalog of 3,932 datasets across 159 viable industries, this record scores 8/10: authority and depth rather than breadth. Get a sample of this dataset and judge the rows first.

Start with the grammar, because that is what a statutory archive really sells: addressability. One Project ID, decoded:

# the anatomy - one Project ID groups everything lodged for a well, survey or report set
project_id       : CCCC2020seis0001
                   CCCC=company  YYYY=year  type=well|seis|mhaz|rems|inpt  NNNN=sequence

# fields riding on every project
file_id          : individual file handle within the project
c_tag            : Classification Tag - what kind of data item the file is
completeness     : Available | Unavailable | Not Acquired
geometry         : footprint against Quadrant, Block and Field polygons + hexagon density bins
project_metadata : date completed | main data type | licences | contractor | description | survey IDs

# the completeness registry, read as three states
Available    = file lodged and present
Unavailable  = acquired but never lodged
Not Acquired = never acquired at all

# formats carried inside projects
SEG-Y   LAS   DLIS   TIFF   CSV   PDF

# shape of the archive
organisation=Project ID -> File ID     grain=project level down to single files
geography=UKCS quadrants, blocks, fields   scale=petabyte class (commonly cited ~1.6 PB)

Read the anatomy rather than any single value. The identifier itself answers three questions - which company lodged it, which year, and what kind of project - before the metadata fields add completion date, contractor, licences and survey IDs. The three-state completeness line is the part most archives cannot offer: Unavailable means the requirement was acquired and then never lodged, a compliance gap with a name, while Not Acquired means the evidence never existed. Treating those two as the same 'missing' throws away the registry's entire point.

Six documented fields form the core, split between addressing and accounting:

• Project ID - the composite key whose segments encode company, year, type and sequence; the join key every other record hangs from.
• File ID - the per-file handle beneath a project, the finest grain the archive indexes.
• Classification Tag (C Tag) - what each file actually is inside the project inventory.
• Completeness status - Available, Unavailable or Not Acquired; the audit layer.
• Geometry / spatial location - the project's footprint against Quadrant, Block and Field polygons plus hexagon density bins, so spatial filtering is native rather than bolted on.
• Project metadata fields - date completed, main data type, licences, contractor, description and survey ID(s).

Fields beyond the spine fold under additional fields on request below and are confirmed against live records when your sample is cut - including the expansion of the three type codes the public documentation leaves abbreviated, which is stated here rather than papered over.

• Geography: the United Kingdom Continental Shelf, and nothing but it - which is the point. Every project resolves against UK offshore Quadrant, Block and Field polygons, with hexagon density bins summarising where the archive runs deepest. A jurisdiction-wide ledger under one indexing scheme beats a dozen national fragments.
• Temporal: a continuous statutory archive rather than a series of versioned cuts. Records span the depth of the UK offshore drilling era, each carrying its own completion dates in project metadata, and the shelf keeps widening - carbon dioxide appraisal and storage licence records now arrive beside the petroleum record.
• Granularity: project-level entries for wells, seismic surveys, hazard surveys and reports, each resolving to individual File IDs with classification tags and status. One schema serves the portfolio question and the single-well question alike.

Set against the wider catalog - 3,932 datasets averaging 7.81 - this record scores 8/10, carried by verified field documentation and statutory provenance rather than dimension count.

API, files, or your warehouse. Daily, weekly, or hourly.

Name the wells, quadrants, project types or years and the sample returns cut exactly to that scope - a single well's lodged record or a quadrant-wide sweep. Cadence gets settled only after the sample validates, and changing it later is a settings conversation, not a re-integration project.

Every delivery ships the field dictionary above unchanged, with the three completeness states arriving as a typed enumeration rather than free text, so an audit group-by never turns into a cleanup pass. Because the Project ID grammar encodes company, year and type, your pipeline can route and partition on the key alone before any parsing begins.

• Decommissioning teams start from the well description projects: what was drilled, how it was completed, which contractor holds the record - and whether the evidence set reads Available or Unavailable before a P&A campaign is priced.
• New-entry and farm-in analysts audit the lodged record for target assets, using the three-state registry as a diligence checklist rather than a vendor's assurance.
• Geoscientists and reprocessing shops reach for legacy seismic survey projects in SEG-Y and give vintage acquisition a second life on modern processing.
• CCS developers screen carbon dioxide appraisal and storage licence records as they join the archive, getting a statutory baseline for storage-site work.
• ML engineers assemble log-response and seismic-characterisation corpora from LAS, DLIS and SEG-Y material that arrives keyed to real well identities.
• Regulatory researchers and journalists cite the record itself - a statutory archive outranks any aggregation when the question is what officially happened in a basin.

Get a sample of this dataset scoped to whichever of those jobs is yours.

Journalists, academics and students lead the ranking at 3 of 3 - a citable statutory source with provenance no aggregator can match. Market researchers and consultants match them at 3 of 3, sizing and screening UKCS activity from official licence and well records. Data scientists and ML engineers sit at 2 of 3 with SEG-Y and LAS training material keyed to real wells, and developers and builders at 2 of 3 for metadata products built on the Project ID grammar. Investors and quant researchers close at 1 of 3 - diligence material rather than a trading feed. Persona-by-persona detail lives on the Oil & Gas Equipment & Services hub.

The NSTA family splits by job. This archive holds the documents and measurements - well descriptions, seismic volumes, reports - lodged under statutory duty. The NSTA structured datasets registry holds the current-state geography - 922 layers tracing wellbore origins, blocks, subareas and field outlines. Use the registry to map and screen; use this archive for document-level depth on a specific well.

Against the globals, the contrast is depth versus breadth. OGIM v2.5.1 maps roughly 6.7 million infrastructure features worldwide but carries no statutory UK survey archive at all, and OPD counts platforms from orbit without reading a single well file. When the question is what did that well actually log, this is the only stop in the catalog.

Three things, stated upfront.

First, provenance of this page: the six field definitions come from the publisher's own documentation, captured during the August 2026 research pass. Live-row examples ride in with your sample rather than being asserted here, and the expansion of the mhaz, rems and inpt type codes is confirmed against records at sampling rather than guessed.

Second, scope by named cut. A petabyte-class archive rewards precision: name the wells, quadrants, project types or years and the sample arrives shaped to the decision you are feeding, not as an undifferentiated continent of files.

Third, pair it deliberately. The NSTA structured datasets registry supplies current-state geography to screen against, OGIM supplies the global surround, and this archive supplies the evidence beneath. Start at Get a sample of this dataset; the daily, weekly, or hourly decision comes after the sample validates.