186 of 217 workbook wellbores have ordered milestone dates. Their summed calendar intervals are 35,771 days; median 47.5 days. These are elapsed intervals, not rig-days, and include suspension and waiting between milestones.
| Revision | Date | Change and reason |
|---|---|---|
| 1 | 2026-10-19 | Replace mixed drilling-day claims with dated calendar intervals; explain populations and remove unsupported learning conclusions. |
| 2 | 2026-10-20 | Separate historical snapshot counts from the current workbook population and document the retained completion_days name. |
| 3 | 2026-10-20 | Use the issued revision date in the footer to keep regeneration dates consistent. |
| 4 | 2026-10-21 | Add sourced WAR limits with a calendar-specific basis; retain the dated intervals and distinguish depth from a duration denominator. |
For each named lease row in workbook Sheet1, calendar days = TD date minus spud date, with no inclusive endpoint adjustment. Blank lease and TOTALS trailer rows are excluded. Missing or reversed dates are not evaluated; same-day dates give zero. 31 rows lack an ordered pair; 186 also have TVD and appear below. No positive-duration filter or producing-well filter is applied.
The historical 184 figure counted positive legacy DRILLING_DAYS entries in the historical 217-wellbore workbook snapshot; the remaining 33 were zero/missing entries. Current population: 217 workbook wellbores. The historical 11,124-day total mixed milestone spans and WAR-union values and is withdrawn as a rig-time claim. The article's 211 wellbores and 21,944 days cover a different development population (excluding Big Foot), and count drilling plus completion, rather than this page's spud-to-TD interval. Neither the difference in counts nor the difference in totals measures missing rig-time or performance.
Source for the historical counts and definitions: data-basis investigation; source for the article comparison: per-well D&C reconciliation, population table. Exact row-level membership of the article population is not established by this workbook; a paired performance comparison is not evaluated.
The producing-well benchmark is a third population, restricted by its configured
spud window and OGOR-A production. Its spud_to_completion_calendar_days
ends at the first completion operation, not TD. Its completion_days
ends at first oil. Neither benchmark interval is rig-time.
The name completion_days is retained for compatibility
with existing CSV consumers. It means first oil minus first completion in calendar
days, including waiting and suspension; a rename requires a coordinated consumer
migration. Active completion time is not established.
| Field | Wellbores | Median (days) | Sum (days) |
|---|---|---|---|
| Anchor | 13 | 51.0 | 821 |
| Big Foot | 25 | 35.0 | 18,010 |
| Cascade | 9 | 71.0 | 1,566 |
| Chinook | 5 | 28.0 | 317 |
| Jack | 8 | 58.0 | 652 |
| Julia | 9 | 96.0 | 802 |
| Kaskida | 7 | 37.0 | 617 |
| North Platte | 12 | 46.5 | 675 |
| Shenandoah | 17 | 69.0 | 1,238 |
| St Malo | 57 | 38.0 | 8,749 |
| Stones | 22 | 70.0 | 2,110 |
| Tiber | 2 | 107.0 | 214 |
Summed well intervals can overlap in calendar time; this sum is neither portfolio elapsed duration nor rig occupancy.
Synthetic input: spud 2020-01-01, TD 2020-01-08, first completion 2020-03-01, first oil 2020-06-09. TD minus spud = 7 calendar days. Completion minus spud = 31 + 29 = 60 calendar days (leap-year February). First oil minus completion = 31 + 30 + 31 + 8 = 100 calendar days. Thus 60 + 100 = 160 days from spud to first oil. None of these calculations establishes active drilling or completion rig-days. Missing milestones leave their intervals not evaluated.
Method source: Python 3 datetime, timedelta objects (date subtraction).
Weekly boundary quantization: 30 CFR 250.743(a) requires weekly WARs for the region covered by that paragraph, Sunday (12 a.m.) through Saturday (11:59 p.m.). Periods are no longer than seven days; first and last periods may be shorter. When activity transitions are inferred only from weekly status rows, a per-well count is quantized at each boundary by up to one reporting week. This is an inference from the cadence, not a measured error bound on every report date; daily narratives or more frequent reports can refine the boundaries.
Coverage floor: the repository records reports submitted since April 2004 (Makefile, comment above the data target, attributed to issue 1066). This is reported, not confirmed from an agency page; the bulk-file start date remains unconfirmed from the BSEE Raw Data and eWell WAR Field Definitions pages examined here. Historical completeness, including coverage of earlier paper reports, is not established. Missing historical reports shall not be interpreted as zero days.
Depth basis: this page plots true vertical depth (TVD), MAX_WELL_BORE_TVD, as geometry against elapsed calendar time; no depth-normalized ratio is calculated. The legacy measured depth (MD), MAX_BH_TOTAL_MD, is not a duration denominator on this page. MD and TVD are not interchangeable.
Publisher-declared basis: this page reports spud-to-TD calendar days = total-depth date minus spud date, without an inclusive endpoint adjustment. For active rig-time, no definition was found in the sources cited here. The workbook's legacy drilling-day column is not used as a duration input. Missing or reversed dates remain not evaluated; same-day milestones give zero. The interval includes waiting and suspension and does not establish active rig-time, utilisation or learning. Weekly WAR limits describe the reporting source; they do not change this page's date-subtraction calculation. Decision: #1064 decision
Sanity reference: the public publisher-deposited abstract of SPE-87117-MS, Deepwater Success Through Predictable and Distinctive Drilling and Completion Performance, paragraph 4, reports 20–23 days per 10,000 ft for its deepwater examples. The abstract does not establish MD versus TVD or comparability to the legacy figures here. The band is a plausibility reference, not an acceptance threshold; no per-well ratio is reproduced or asserted by this note. Record: https://api.crossref.org/works/10.2118/87117-MS
Sources examined 2026-10-21: