Save status stays beside the live observation inside the plot. A failed response does not prove that a save failed: the server may have committed it before the connection dropped. The viewer pauses, releases steering/audio and reads the canonical journal without a cache. An exact run/seed/scenario/model and command match confirms the acknowledgement without another write. If the online journal is an exact prefix of the browser route, only an explicit retry appends the remaining actual commands against the checked revision. Server compare-and-set, append-only command prefixes, original-model replay, player scope and immutable finished results remain unchanged. This creates no new branch, drilling, gas, award or campaign well.
The recovery panel compares browser / online command counts and their matching prefix. Different journals are never merged or force-saved. Before loading a different online route, a separate browser-only archival journal must be written and read back successfully. Up to16 copies per attempt are retained without automatic trimming; key collisions, quota, disabled storage or failed read-back block replacement. Compatible longer browser drafts from an older saved revision are recoverable only after exact identity and command-prefix checks. Incompatible / different active drafts are not silently applied, overwritten or removed by another normal save.
Browser copies are not online backups or separately resumable online wells. They can be selected and downloaded as JSON journals for archival inspection/support, not merged over the canonical result. Download requests are not proof that the file reached durable storage; confirm the file yourself. Clearing browser data can remove local drafts/copies. If storage is unavailable, current work may exist only in the open tab. Keep it open, download the journal and retry online saving when possible. No recovery screen, selector or check starts paid generation, device speech or another attempt. New guidance is authoritative text / matching recording unavailable / narration review pending.
Primary interface / protocol evidence opened15October2026: IETF RFC9110 §9.2.2 distinguishes safe retry knowledge from blindly repeating a non-idempotent request; MDN localStorage documents browser policy failures and origin-specific storage; HSE shift handover supports clear task-relevant written information. These motivate conservative software recovery and communication, not field well-control instructions, engineering thresholds or guaranteed browser durability. Geological models and exact simulated operating values are unchanged.
The optional Acquired only button pauses drilling and filters both the main section and Branch logs. Unknown rock stays blank: no undrilled bed polygons, fault planes, water contact, fracture corridor or erosional surface is drawn. Material-coloured dots are sampled at actual retained non-cased survey coordinates under the original frozen model. They are discrete synthetic observations, not inferred contacts between stations. Up to720 markers are shared across branches, keeping endpoints; Branch logs still select every original survey. Failed and abandoned branches remain inspectable. Existing cased bores keep their surveyed path, but receive no new formation readings in this view.
Only stations within ±5m of the displayed N slice get material dots. Off-plane bores remain dashed projections without moving their rock observations into another section. Fit route uses retained surveyed positions and the already-disclosed original programme, never future target windows. The cyan plan remains a programme, and purple remains the player's correlation hypothesis, not an observed reservoir boundary. Actual stationed material agrees with the same versioned sampler used by contact checks. Survey-only plots do not change any drilling, pressure, costs, drainage, saved replay or payout.
This is a reversible display filter, not a blind or competitive assessment: scenario descriptions, live target/pressure labels, guided crew holds and final model results remain simulation aids. Switch it off for the full Structure view and post-well comparison. Automatic full-world fold-hinge narration is withheld while the filter is on; existing event recordings remain where matched. It does not introduce realistic sensor noise, tool offset, surface-gas lag or uncertain measured cuttings. SLB geosteering supports logging-led directional decisions; SLB directional survey distinguishes discrete stations and an approximated interstation trajectory. These concepts motivate the interface only, not its720-marker / ±5m display calibration or field suitability. New guidance is authoritative text, matching recording unavailable / narration review pending. No new paid voice generation or device speech.
The original dashed cyan line is a directional plan, not a guaranteed reservoir centreline. Correlate now lets you choose two actual drilled stations interpreted as the same bed centre, roof or floor, then apply a distinct purple dashed lookahead with an assumed widening band. SLB geosteering supports using geological logging in directional decisions; apparent dip and directional survey distinguish section-dependent interpretation from discrete well observations. They do not validate this game's simple two-point extrapolation or its numerical limits.
Assuming the two picks intersect the same geological marker, their TVD difference divided by horizontal separation defines an interpreted along-track slope. It is not a measured true bed dip or a three-dimensional geological surface. The forecast stays on that assumed heading and original branch; it never samples future material/windows or auto-steers. Minimum30m horizontal tie spacing, maximum30° apparent slope,40–200m lookahead from tie B, a2–20m initial half-width, assumed1–10° slope-range growth and a10m cross-track display corridor are game bounds, not operating limits or statistical confidence. Purple segments outside the viewed N slice are labelled projections. Lookahead is never automatically extended; re-tie after structure, facies/fluid or branch changes.
Hypotheses retain the actual last acquired station/command count at issue. Later checks explicitly separate evidence already known then (retrospective) from genuinely new drilling. Counts describe surveyed trajectory positions inside a band, not productive metres or proof of correctly predicted reservoir. Actual station logs remain authoritative synthetic observations from the frozen model. A two-point interpretation can be wrong even when the trajectory stays inside its band. Latest24 checks link to their own original surveyed stations. No hypothesis moves the bore, changes costs/material/contact/production, or earns a payout; the player still drills every correction.
Up to16 original hypotheses are retained only in this browser, independently of online well reports. Reload validates attempt/model/seed and exact original anchors. Missing unsynced footage leaves notes pending until the matching browser drilling draft is restored; no ties are moved or silently dropped. Invalid notebooks remain untouched, full history refuses further entries, storage failure is labelled session-only, and another tab's changed notebook is not overwritten. Browser-data loss can lose these annotations; they are not server-backed well records. Opening pauses footage, clears steering and cancels recorded playback; closing never resumes automatically. New interpretation guidance is written-only, flagged for narration review; existing recorded cues, model versions and ledgers are unchanged.
The first three advanced horizontal exercises are fictional Foothills gas-well analogues. A fourth exercise uses a fictional northern North Sea tilted gas block. A fifth uses a fictional Viking stratigraphic pinchout. A sixth uses a fictional Leduc-inspired carbonate gas reef. A seventh is an Alberta-inspired angular-unconformity geometry lab, not a named gas field. Existing cased verticals, reservoir windows, local markers, fault displacement where present, pressure, equipment response and costs are simulated. These are not surveyed leases, actual field logs or approved drilling programmes.
AGS / GSC Atlas Chapter 12 documents reef complexes, carbonate/basin relief, dolomitization and variable reservoir character. This supports a distinct carbonate mechanism, not a mapped Leduc lease. This is not Leduc No.1, a historical reconstruction or a measured type log. A local gas-bearing porous dolomite layer, tight limestone cap/baffle, broad asymmetric reef relief and a single sharp flat gas-water contact are illustrative. Actual fluid contacts commonly have transition zones and can tilt; see SLB fluid contact. Capillarity, oil, condensate and sour-gas hazards are outside this exercise.
Manually build and land above the reef flank, follow changing porous interval geometry and correlate the tight baffle. The fixed 20 m contact lattice and depth-clipped fluid boundary supply the same canvas/SVG polygons, material checks, target windows, synthetic logs and actual swept-path contact. Out-of-plane narrowing is clamped at lattice knots before interpolation, avoiding a hidden sampling/drawing kink. Buff stippled dolomite contrasts with brick-textured tight limestone; blue dolomite lies below the cyan fluid boundary. Amber centre/edge lines show a bounded fracture-loss corridor, not a fault: no strata are displaced. The displayed training truth is not an acquired geological survey or a perfectly executable plan.
GR represents radioactivity, not porosity or a hydrocarbon detector; SLB gamma ray log also notes radioactive carbonates. This synthetic local clean carbonate has overlapping low GR in tight limestone, gas-bearing dolomite and wet dolomite. The assumed evaluation suite provides illustrative matrix porosity and resistivity, including a log-scale acquired-station track. Gas remains an immediate relative formation index, not lagged surface mud gas. Rt = 0.12 / (porosity² × water saturation²) is a clean-rock Archie-style game response with fixed Rw, a=1 and exponents m=n=2; see SLB formation-factor relation and water saturation. No field calibration or shale saturation inversion is claimed. Tight limestone can be highly resistive without useful porosity or gas pay. Gas, porosity, Rt, cuttings and surveys must be considered together; the evaluation hold charges interpretation/verification work, not a previously hidden measured log.
SLB lost circulation supports fracture-related loss and reduced fluid-column/well-control concerns. The game's local corridor and 18 m³/h untreated loss potential are invented, not a Leduc field measurement. Actual drilled exposure integrates at ≤0.5 m; loss volume accrues using drilling time in the corridor, with retained mud cost and hole/barrier penalties. Circulation during stationary holds and explicit fluid-column hydraulics are not simulated. Tailored treatment reduces future loss while retaining earlier losses; blanket sealing loses less fluid but carries substantially greater productivity damage. No treatment changes coordinates, contact or surveys.
Wet and tight metres never create gas drainage cells. At TD, the simulated cased selective programme verifies gas access and water exclusion, or a cheaper verified commingled test retains a water-interference factor 1/(1 + actually drilled wet metres/25) on gas rate and recoverable volume. This is a game consequence, not a water-coning forecast or qualified completion design. Isolation does not undo damage, losses or old wet footage. Skipping tests books no performed production test, rate or recoverable gas. Ignoring pressure requires additional charged remedial verification before verified completion; prior damage remains. Unique 80 m cells cap gas credit and the practice payout is zero.
Model 4 dispatches only this new scenario. Advanced models 1/2/3 and Doug 1 remain byte-frozen; old journals/drafts do not upgrade. Live and acquired evidence, choice consequences and final report retain the same original seeded world. Twelve telemetry readings remain next to the plot, including porosity, Rt and lost mud, with a two-row phone layout. Existing generic bed/survey and gas-debrief recordings remain; the specialist holds now have matching local crew recordings. No false fold/fault narration, live voice generation or browser speech is used during play.
AGS / Geological Survey of Canada, Atlas Chapter 21 documents interbedded Viking sand, silt and mud, narrow reservoir bodies, pinchouts and incised-valley / estuarine-fill complexity. Its published central-Alberta reference section, Figure 21.5 illustrates log and lithological variability; no numerical curve has been digitized or copied. SLB's pinch-out definition supports termination by thinning, not a tectonic offset. This exercise does not equate every Viking reservoir with the same channel architecture.
Two thin, low-relief sandstone bodies, their exact thicknesses, a muddy interruption, mild structural relief and cemented streak are synthetic. The upper sand and silty margin taper out and reappear beyond non-pay; the deeper sand is a different target. There are no faults or continuous coals in this local analogue. A target choice changes no coordinates, inclination, surveys, footage or previous contact: the player must actually drill down, or retain and recover the upper sand. The same fixed 20 m contact lattice drives polygons, material, GR, gas, windows and contact. Bed-relative synthetic GR trends distinguish cleaner sandstone, silty margin and shale. Gas is an immediate formation-response index, not a simulated lagged surface measurement or a calibrated fluid analysis.
Model 3 integrates actual swept path at at most 0.5 m spacing to avoid skipping thin beds on fast commands. Previously acquired upper-sand contact survives a later target change. Shale exposure, mud conditioning, bit wear, dogleg-related damage, actual CAD spend and elapsed work persist. Mud holds require actual non-pay readings rather than firing falsely while a bit remains in the deeper sand. Conditional operating thresholds remain game calibration, not recommended field limits.
Verified cased selective access tests only actually contacted selected bodies, retaining separate upper-west, upper-east and deeper-sand drainage identities. These conservative simulation identities do not establish independent measured pressures or communication. Deeper-only completion excludes upper contacts/cells; unverified completion books no performed test, flow or recoverable gas. Reports retain all drilled evidence alongside completion-eligible results. Model 3 is isolated behind explicit version/scenario dispatch; published models 1, 2 and Doug 1 are unchanged and all three advanced versions freeze on publication.
The thin-sand specialist instructions have matching local crew recordings. Existing generic bed/survey briefing and gas debrief recordings are reused; no fold/reverse-fault cue is substituted for stratigraphic thinning, no browser speech or live generation during play. Exact geometry, depths, logs, pressure, equipment response, completion hardware and costs are illustrative, not a measured Viking well or approved programme.
Norwegian Petroleum / Norwegian Offshore Directorate: Huldra describes Middle Jurassic Brent Group gas/condensate sandstone in a rotated fault block at approximately 3,500–3,900 m, with small faults and two principal segments lacking pressure communication. The field originally had HPHT conditions. The Directorate's fact page is the primary field record. This supports a distinct structural/compartment mechanism, not the game's fault position, throw, dip, pressure jump or well path.
The new practice well manually builds and lands in tilted sandstone, crosses a downthrowing normal fault and correlates missing section rather than thrust repetition. Its local synthetic section has sandstone, siltstone and shale but no continuous coal roof/floor; this is not a claim that Brent strata never contain coal. Two block pressures are fixed when the attempt starts. A missed pressure check retains damage and barrier losses; a verified completion requires charged remedial pressure verification if that check was skipped.
At TD, a simulated cased selective completion with verified interval barriers tests both blocks separately. Alternatively, isolate the earlier interval and test only the far block, at lower cost/time but with its own reduced eligible contact, drainage and recoverable gas. Skipping completion isolation/tests books no performed production test or recoverable gas. SLB casing completion and packer definitions support selective perforation and annular sealing principles; the exercise does not specify qualified hardware or imply a mechanical openhole junction is a pressure seal.
Per-block simulated pressure/rate results are kept separate. Total gas rate sums the selected isolated tests; the headline pressure is representative and does not establish communication. Exact geometry, log responses, operational limits, CAD costs and game gas parameters are illustrative. Temperature, condensate behaviour, offshore installation design and an actual Huldra HPHT casing programme are not modelled. The well starts below existing casing; this is not an endorsement of a two-string HPHT programme.
Explicit scenario/version dispatch preserves original model-1 replay for the three Foothills exercises and uses model 2 only for this new scenario. Both published models are frozen. Normal-fault, compartment and completion instructions have matching local crew recordings; the full written evidence remains authoritative. Existing generic recorded briefing/debrief calls may be reused; no reverse-fault voice is substituted for a normal fault, and no browser speech or live generation is used during play.
| Source | What it supports | Game interpretation |
|---|---|---|
| Town of Hinton / Epoch Energy FEED report, 2018, structural geology section | Folded and thrust-repeated strata near Hinton, structural compartments and formation-dependent drilling deviation. | One asymmetric fold and a dipping reverse fault repeat sandstone, siltstone and local coal markers. The exact dimensions and pressures are invented for the exercise. |
| Alberta Geological Survey, structure and architecture of the Western Canada Sedimentary Basin | Competent beds, thrust sheets, shale detachments and contrasting fold styles. | Correlated contacts deform together. Coal and sandstone are part of a continuous local stratigraphic section, rather than unrelated obstacles. |
| SLB, defining directional drilling | Formation-dependent directional response, sidetracks, kickoff methods and equipment limits. | Manual steering, bit/BHA condition and measured separation determine a sidetrack attempt. Exact success criteria, curvature, timing and costs are game calibration. |
| SLB, defining multilateral wells | Stacked and opposed branches, junction support, re-entry and differences in hydraulic isolation. | The first lateral must be secured, the junction verified and re-entry confirmed before a second manually drilled branch. A mechanically supported openhole junction alone is not treated as proof of independent pressure isolation. |
Each choice shows its original model charge, elapsed work, hole/barrier/bit indices, damage multiplier, mud change, target change, verified barriers and retained-path changes. An in-hole return is identified as repositioning with zero new drilling; straight-line displacement is not represented as a drilled length. Chained sidetrack / junction operations retain the next real held issue and show the previous result separately. Resuming or finishing remains explicit.
Decision log reconstructs only the existing fixed-seed command journal under its original frozen model. It preserves the issue as originally issued, the synthetic evidence at that hold, the action actually chosen and its immediate difference. The following interval ends at the next issue or saved-journal end; its costs and raw contact reflect the whole sequence, not proof that one decision alone caused every later change. Original-model contact is not necessarily completion-eligible or producing contact. Pending or suspended issues have no invented action.
Link each issue to its actual retained surveyed station, including failed sidetracks and old branches. This opens the existing own-N evidence inspector; it cannot move a bore, add acquired evidence, test a new interval, reroll geology or resettle a payout. The review yields every128commands, cancels on close/run change and displays at most41 nearby selector options while arrows reach all decisions. It is built on demand, cached by the exact journal and never rebuilt every animation frame. All APIs, ledgers and frozen equations are unchanged.
UK HSE: shift handover / safety-critical communication supports structured task-relevant written information, cross-checking and operator-oriented logs/displays. This is a training-interface adaptation, not an HSE-approved operating procedure. SLB directional drilling remains the primary background for surveys and formation-response interpretation. No new field geology, engineering threshold or measured value is introduced.
Original recorded issue cues can be replayed with exact captions and skip/mute/cancellation. Missing recordings retain authoritative written evidence and a narration-review flag. Event IDs now select the recordings explicitly: dipping North Sea sand is not automatically a fold hinge, and sidetrack planning is not failed separation. The physical fold-hinge call still requires the existing actual dip-reversal test in model1. All seven reviewed clips remain intact; no paid generation or device speech.
Older inclined sandstone, shale and siltstone markers are cut off by an erosional surface before younger shale and a silt marker are deposited above it. Gold dashed boundary = unconformity, not a slipping fault. Fixed20m uncut bed and erosion knots are clipped at their exact intersections: the polygon, window, own-N station, synthetic log and swept contact all use the same geometry, including tips between knots. Eroded upper sand cannot be recovered by following its old projected dip. Target selection does not move the bit; manually descend into the surviving lower bed. Actual younger-cover exposure remains non-pay and reduces the fictional hole-condition index; conditioning changes future susceptibility, never erases earlier exposure or restores contact.
AGS Digital Data2022-0036 (published27February2023) documents oblique underlying units and truncation-defined subcrops. AGS Map573 (26June2014) describes independently modelled units truncated by the unconformity, cross-validated against wells. SLB defines angular unconformity as younger beds separated from eroded older inclined strata. These support the mechanism only. No real mapped unit, sandstone subcrop, source/seal system, gas discovery, southwest dip orientation, lease, measured log or actual well is reproduced. The two gas sands,20–22m thickness,2.7–3.0° apparent game dip, valley relief,2200m reference depth,20m mesh, pressure, mud, curvature, costs and thresholds are synthetic. No karst, weathered interval, blanket coal or pressure communication is asserted.
Original GR stratigraphy is not compressed into the surviving eroded thickness. Gas is the same immediate synthetic formation index, not lagged surface returns. Verified selective completion can test both actually contacted older beds, or isolate earlier upper contact and test only the lower bed; acquired surveys stay retained. Ignored pressure requires charged remedial work without repairing earlier damage. Unverified / suspended work books no performed production test or recoverable gas. Conservative80m drainage cells use the minimum surviving thickness at all20m vertices and both N edges; a partly erased cell may book zero volume despite recorded contact. This deliberately coarse lower-bound geometry proxy is not a reserve estimate or numerical reservoir simulator. Unique cells cap credit and results carry no payout.
Only this new scenario dispatches to model5. Published models1–4 and Doug1 remain unchanged; wrong model/scenario drafts are rejected. Existing seven recordings are retained; the four specialist issues now have matching local crew recordings, with no browser speech or live generation during play. Existing camera, controls, Correlate and original-decision review remain available inside the same viewer.
The two new mirrored exercises build from an existing cased vertical, land a broad synthetic sandstone, drill 220 m beyond the heel, then hold for the horizontal turn briefing. Choosing to begin changes the main view from side section to north-up overhead; the same two steering positions become Turn left / Turn right. Pace controls keep their positions, while Pause / Resume remains separated from the steering dock. After 260 m of actually drilled turn footage the view returns to the side section and Build INC / Drop INC. Compass azimuth is clockwise from fictional grid North: both start at 90° East, then target 60° toward the north-east or 120° toward the south-east. Earlier models retain their original internal direction conventions.
SLB's azimuth definition supports a horizontal direction referenced to North. SLB's directional-drilling overview explains build sections, surveys, dogleg severity in degrees per 30 m and the need to consider BHA, drillstring, logging-tool and casing curvature limits. These support the directional concepts only. The 30° bend, 260 m turn, maximum 0.14° azimuth response per drilled metre, 4° heading tolerance, 35 m design-corridor tolerance, broad sandstone, logs, depths, pressures, wear response and costs are game calibration, not manufacturer ratings, an approved trajectory, a lease map or an anti-collision calculation.
The fixed design curve never follows or replaces the player's actual heading. During overhead steering, left decreases and right increases azimuth; actual inclination holds at the turn's entry value and must be recovered by manual INC steering after side view returns. The bit moves by integrating its own orientation through drilled measured depth, with swept geological/contact samples at intervals no greater than 0.5 m. Heading and E / N displacement are retained in each surveyed station. A held button is limited by drilled metres rather than render-frame or tapping frequency, and accepting a survey never snaps position, heading or contact back to plan.
A missed bend retains actual displacement, adds survey cost and elapsed work, and lowers the persistent damage factor. Up to two optional correction sections require additional charged, manually steered footage; no choice grants free relocation. A gas test can flow from the broad sandstone while the planned drainage corridor is missed, so flow alone cannot earn placement success. The corridor is a survey-design requirement, not a sand boundary or proven drainage extent. Only actual contacted unique sandstone cells enter a verified synthetic production test; suspended or unverified projects book no test, flow or recoverable gas. Scores, campaign wells, market history and other ledgers are unchanged.
Version 6 dispatches only these new scenarios; published models 1–5 and Doug 1 remain byte-identical. Saving during overhead steering and replaying the fixed-seed journal restores the actual active turn, view and surveyed route. Turn instructions now have matching local crew recordings. Existing recordings are preserved, with no browser speech or live generation during play.
Shared carbonate fabrics now distinguish fine micrite, irregular vug motifs, coated ooids and selected dolostone crystal facets without the old generic mirrored carbonate overlay. Fabric sources and limits explain why these are exaggerated surface illustrations, not measured pore size, porosity, mineral fill or proof of connected pay. Original material/water clips, blue water-bearing rock, exact contacts and acquired-only unknown gaps are unchanged.
Existing fault planes and carbonate fracture centres now have restrained rough cut-face shading and broken pale highlights. These are illustrative lighting marks, not measured aperture, mineral fill, a damage-zone width, a new fault or a new loss interval. Relief is clipped inside the original material/water polygons; fracture corridor edges remain boundary markers rather than extra fractures. The original fault symbols, centre lines and exact contacts stay authoritative. Acquired-only views receive no hidden structural detail.
Existing modeled bed contacts also carry intermittent pale cut-face ribbons and sparse rusty surface-colour patches beneath the continuous contact marker. Every patch is a subsection of an original contact segment, including its fold, pinchout, erosion and fault offset; it is not a new bed, vein, mineral assay, oxidation map, pay indicator or log response. The colours and widths are illustrative display conventions guided by the owner-supplied rock reference. Geometry is compiled once with fixed bounds and falls back to the plain contact marker on excessive input. Acquired-only unknown gaps still receive no inferred contact or surface detail.
The live HZ section reuses cached rock grain, pore and coal-cleat fabrics inside the original material polygons. Grain-only shale, silt and sand now use larger deterministic procedural tiles instead of mirrored photographic quadrants; sources and limits are documented here. A uniform display scale prevents vertical exaggeration from stretching particles, while the texture origin stays fixed to world zero. Texture remains illustrative surface detail, not measured grain size, additional acquired evidence, new bedding or a fracture/hazard map. Contact relief follows the same model vertices; texture marks do not define beds. Blue water-bearing rock and existing fluid/fault markers remain distinct. Acquired-only gaps remain unpainted. The separate SVG evidence inspector retains its simplified material colours.
Camera review: Fit section shows the full fictional training structure, existing casing and retained bores. Drag to pan, pinch / wheel / + / − to zoom, and select 1:1 scale to remove display exaggeration. Camera navigation pauses drilling and clears held steering; closing review by Resume returns to the actual bit before any new footage. Follow bit resets the camera without starting drilling. These controls never move a bore or alter a seed, command, contact counter, completion or settled result.
Vertical exaggeration is calculated from the displayed metres-per-pixel scales and labelled as VE; 1:1 uses equal horizontal and vertical metres per pixel. USGS ModelMuse display guidance distinguishes visual exaggeration from model changes. Full-fit geometry is disclosed schematic training truth, not an acquired field log, measured subsurface map or a hidden-geology ranking tool for Doug planning.
The north-up inset now uses equal easting/northing scale and shows retained plugs, the actual bit and surveyed build end / in-target heel (or the true multilateral junction), plus the displayed section extent. Coordinates are a fictional local game grid, not surveyed geographic positions. SLB's directional-survey definition explains discrete position/inclination/direction evidence; the display uses the retained station positions without changing the published models' internal direction convention. A single well's heel is not independently a multilateral junction.
Rendered polygons, contact lines, the inclined fault trace, GR/facies readings and productive-contact checks use the same folded contacts and fault half-planes. The upthrown hanging wall can expose repeated reservoir windows near the inclined plane. The schematic section is shown for the exercise; it does not imply that a real well's subsurface is known this precisely.
The displayed continuous coal roof and floor belong to these fictional Falher-style local sections. They do not imply continuous coal around every Spirit River or Wilrich reservoir. The approximately 65° instability cue is a game threshold informed by local experience, not a universal geomechanical limit.
A return inside an already drilled motherbore is a charged trip, not new drilled footage. The surveyed kickoff, plugged parent, failed attempts, cement and new paths remain visible. A sidetrack has to earn separation by actually drilling; a fresh assembly does not guarantee success without steering. The second lateral is a distinct north fan or a lower stacked target, with no instantaneous move into the lower pay.
Production calculations exclude plugged, failed and suspended branches. Shared drainage cells prevent overlapping completed branches from counting the same connected gas twice. Separate stacked reservoir identities can contribute different pay volumes.
Branch logs pauses drilling and inspects only existing surveyed points on the selected retained bore. GR, gas index, rock, reservoir sheet and pressure are synthetic model responses at that point's own easting, northing and TVD. The slider retains every survey; long charts display at most 480 stations. Bands and curves connect discrete stations, so contacts between them are approximate. Coal's low GR and elevated gas must not be mistaken for sandstone contact. This is a training analogue, not measured LWD or a published type log.
The main section is a fixed-northing slice through the current bit. Solid path spans are within ±5 m of that slice; dashed, faded spans and outlined plugs are projections from another northing. The ±5 m threshold is a display tolerance, not a geological or well-clearance criterion. Branch evidence shows the section at the selected station's own northing and marks that actual station. Inspect each branch's own readings before inferring contact from a projected bore.
The review retains each branch's status, surveyed length, original target/lateral contact counters, connected and shared drainage cells, failed attempts, cement verification and latest operation notes. It exposes AFE variance, elapsed work, remaining bit, damage multiplier and hole/barrier game indices. A mechanically supported junction is not independent pressure isolation; see SLB's multilateral classification above. Inspection never modifies the frozen replay, resamples an attempt's seed, awards footage or recalculates settled results.
Gas initially in place = connected area × gross pay × net-to-gross × porosity × (1 − water saturation) ÷ gas formation-volume factor Bg. Bg converts reservoir volume to standard gas volume. Recoverable gas applies the simulated recovery and damage factors. Flow-test rate uses connected permeability, thickness and pressure response separately; hydrocarbon shows alone do not prove a commercial well.
Volumes are standard cubic metres (Sm³); rates are Sm³/day. Gas indices, pressures, rates, costs and AFE allowances are synthetic. The existing cased vertical is a sunk-cost starting point, stated in the AFE. These exercise reports have no campaign cash payout and do not change saved campaign wells or high scores.
The server fixes each attempt's geology before play. The same Drill Deep player identity retains the seed, append-only command journal and immutable final report. Reloading cannot reroll an attempt. A new attempt is an explicit separate run.
Natural adult male crew calls were generated for these specific situations and saved as local recordings with exact captions. Briefings hold footage until finished or skipped. Mute, replay and written evidence remain available. No browser/device speech is used.
Reviewed 16 October 2026. All depths, fault geometry and operating quantities remain game assumptions.
Recurring acquired-footage checkpoints, clean-run combos and earned boost now span the campaign, offshore platforms, Historic Wells, Kola, Doug and HZ challenges. Boost uses each mode’s existing pace or replay-speed limits. Browser arcade points are separate from online well results and payouts. Saved HZ/Doug models and journals are unchanged.
The Crew radio check panel on each game page plays the real recorded characters with captions, native audio controls and playback status. Game operations pause during this check. Operational calls take priority over reward radio; mute, replay and stopped playback remain available.