Workbook ReadMe

WORKBOOK STATUS

Working consolidation model, published for technical review.

This workbook compiles Quantifying Antiquity calculations and operation-specific worksheets into a single integrated model. It is published as a working model alongside the current Working White Paper Draft - named at B77, which is the single place the paper version is recorded - and the Interactive Constructability Model. Version numbers record workbook assembly iterations.

MODEL PURPOSE

This workbook evaluates the conventional Great Pyramid construction scenario as an integrated logistics and material-balance system. It does not propose an alternative construction method. It tests whether the conventional labor/material framework closes under auditable construction-management constraints.

MATERIAL-BALANCE RULE

The same crop, fiber, copper, timber, rope, fuel, boat capacity, transport capacity, water, or labor day cannot be spent twice. If consumed by the pyramid system, it is unavailable to the non-pyramid economy during that period.

Off-site personnel are not assumed to reside at Giza. However, the modeled share of their labor, land output, crop yield, fiber yield, copper output, timber output, transport capacity, and household support is assigned to the pyramid supply chain. These quantities are treated as project-dedicated production burden.

CURRENT MODEL LEVELS

CURRENT INCLUDED OPERATIONS

Haulage, manpower, water carriers, track wetting volume, rope plant, fiber production/prep/transport, quarry, copper chain, route capacity, food grain system, grain transport, milling/baking capacity check, oven fuel, shaping/fitting/setting (Wave 1), timber labor (Wave 3), ramp construction/maintenance/removal (Wave 4), Aswan granite expedition (Wave 5, bodies only), river/harbor corridor (Wave 6, bodies only), and selected support-system calculations.

KEY BOUNDARY NOTE

The current M3 result should be treated as a partial integrated lower-bound system-closure test, not a final maximum estimate. Several major operations remain pending or only partially included.

REQUIRED-FUNCTION AUDIT

A burden cannot be deleted from the model while leaving its function intact. If labor, rope, timber, water, food, tools, transport, or another support burden is reduced, the workbook must identify the replacement mechanism, alternate resource, productivity increase, schedule change, or scope reduction.

RESOURCE-LOADED PROGRAM MODEL

This workbook treats the conventional scenario as a resource-loaded WBS/program model. Isolated feasibility demonstrations do not prove system closure; each required operation must be loaded with labor, material, replacement, logistics, support, and state-capacity burden at the required cadence.

UNIVERSAL HUMAN SUPPORT REQUIREMENTS

Every counted laborer, dependent, and project-supported person must carry baseline human support requirements: food, water, rest/absence, mortality/replacement where applicable, and household support. HumanRequirements_20 now exposes the current domestic/family water gap separately from the main M3 totals because loading it into the Giza carrier model creates a feedback loop requiring iterative/algebraic treatment.

Stabilization controls added in v0.29

Before adding additional operations, the workbook now uses four control sheets to prevent uneven treatment of workers, dependents, off-site production, and recursive support burdens: PersonCategoryMatrix_22, FeedbackLoopRegister_23, BurdenMaturity_24, and ModelModes_25. A new operation should not enter the M3 totals until its labor, food, water, rest/absence, replacement, household/dependent, location, and feedback treatment are explicitly assigned or flagged as an open item.

WATER INFRASTRUCTURE ALTERNATIVES

WaterInfraAlt_27 tests claims that canals, depots, wells, cisterns, animals, sled tanks, boats, or closer water access reduced hand-carried domestic/process water. These alternatives are not allowed to reduce M3 totals unless the replacement route, labor, material, maintenance, source trail, and feedback burden are assigned.

The current main model remains the hand-carry feedback solution with bounded youth light-support credit. Water alternatives are held as replacement-burden tests, not automatic reductions.

FOOD FEEDBACK CONTROL

FoodFeedback_28 now tests the agricultural farmer-household loop with annual-yield farmers corrected to roster factor 1.0. Annual t/ha/yr crop yield already embeds the agricultural calendar, so applying the construction crew roster-and-replacement multiplier to grain or fiber farmers would double-count availability loss.

The corrected loop computes and should be compared against the direct-cropland lower-bound. Standing processing, transport, facility, quarry, water, and construction crews retain the full roster-and-replacement multiplier.

SANITATION / BASIC MEDICAL SUPPORT CONTROL

SanitationMedical_29 adds a first-order diagnostic for village sanitation, waste handling, cleaning/refuse, and basic medical/injury-care support. It recognizes these as required functions for the current on-site village population but holds them outside the main M3 total until labor, materials, source trail, and feedback treatment are solved.

PROCESSING / FUEL / CONTAINERS CONTROL

ProcessingFuelContainers_30 adds a first-order diagnostic for milling/baking/brewing labor, oven/brewery fuel, minimum daily water/grain/fuel handling inventory, and container replacement sensitivity. It does not rewrite the main M3 total.

The control prevents treating delivered grain and water as final usable support without processing labor, fuel, containers, storage, distribution, replacement, and feedback treatment.

TIMBER SYSTEM CONTROL

TimberSystem_31 carries the structural-timber dependency diagnostic covering sleds, ramp/track surfaces, anchors, levers, cribbing, boats/harbor systems, workshops, and replacement stock. It distinguishes structural timber from fuel wood already quantified for smelting and food processing.

Timber LABOR is integrated into M3: TimberLabor_40 (Wave 3) converts the structural, ramp-planking, and barge-hull timber streams into felling, sawing, shipping, and haul labor, entering Model_04 off-site. The MATERIAL diagnostic (stock, service life, species suitability, supply ceiling) remains control-only; no timber closure test exists yet. Species/grade sourcing and use-class service life remain open items.

RAMP SYSTEM CONTROL

RampSystem_32 adds a construction-route diagnostic covering ramp geometry/volume, material movement, surface condition, traffic capacity, maintenance, phasing, and removal/rework. Base ramp construction/maintenance/removal labor is integrated into M3 as Wave 4; ramp surface, geometry and phasing remain open.

The ramp module prevents treating ramps as cost-free explanatory geometry. Any ramp reduction must preserve lift/access/traffic functions through a priced replacement mechanism.

AGRICULTURAL ROSTER CORRECTION / CAPACITY-BAND CONTROL

v0.38 corrected grain and fiber farmers to roster factor 1.0. This lowered the headline M3 burden in favor of the conventional model and removed an anti-generous double count. The system-stress finding remains because the corrected population still falls between the 300k and 500k mobilization-capacity bands.

v0.39 replaces single-cutoff language with mobilization-capacity bands. The 150,000 value is retained only as a working upper stress screen within the sensitivity table; it is not a sourced historical ceiling, and 9.375% is arithmetic only.

SHAPING / SETTING CONTROL

ShapingSetting_33 shaping, fitting, setting, and alignment crews are integrated into M3 (Wave 1). Casing scope, productivity sourcing, and copper-wear coupling remain open levers.

Future integration must add productivity, copper wear, timber/cribbing dependencies, labor roster/household support, water/food burden, and feedback treatment before changing M3.

WORKING VERSION STATUS

Current workbook

QA_Integrated_Model_v0_92_USER_SOURCE.xlsx

Current matching white paper

Working White Paper v3.9 (wp39), published at quantifyingantiquity.com, built on QA_Integrated_Model_v0_92_USER_SOURCE.xlsx. This cell names the PUBLISHED paper and the workbook it ships with, so the pairing is stated rather than inferred - a paper and its workbook are one release. wp39 was reconciled figure by figure against v0.90; v0.91 and v0.92 add only text repairs that moved no reported value, so every figure in the paper holds against the workbook named here. This cell moves only when a paper is published - not when a working draft is revised.

Development status

Working model, published for technical review alongside the ICM. Code repository is source of record; the workbook is generated as a versioned build artifact.

Headline result

533,234 project-dedicated people; 33.33% of the 1.6M Butzer comparator; above the 500,000 top band - functionally kingdom-scale reorganization.

Standing verdicts

Test 1: ex-farmer labor at 61.6% of the non-farm male labor pool on the concurrent/standing basis ruled in MR-08 - EXCEEDS SINGLE-PROJECT MOBILIZATION CEILING (2.47x the 25% benchmark). The pre-correction rostered/standing figure of 70.4% is retained as a transparency line. Demographic closure: sustained mobilization at 0.95x the Weni maximum on the concurrent basis ruled in MR-09, 0.825x-1.140x across the mortality band - DOES NOT CLOSE; the finding is Weni-scale mobilization held continuously, not categorical exceedance.

Capacity framing

Use mobilization-capacity bands. 150,000 is only a working upper stress screen. 9.375% is arithmetic only: 150,000 / 1,600,000.

Integration waves

Wave 1 shaping/setting; Wave 3 timber labor; Wave 4 ramp system; Wave 5 Aswan granite (bodies only); Wave 6 river/harbor corridor (bodies only). All integrated into M3.

Most recent passes

v0.60 ex-farmer Test 1 numerator ruling (F24); v0.61 DemographicClosure_46 verdict layer; v0.62 hardening levers (imported-labor share, plank service-life steelman); v0.63 RiverHarbor_47 (Wave 6); v0.64 source-register hardening with sight-verified page pins; v0.65 register close-out, check repair and ceiling-lever wiring (Palermo cedar page-pinned and badge-split; QAChecks_08 M3 reconciliation corrected to Model_04 D133 + RiverHarbor_47 B29; 25% mobilization ceiling wired from ClosureSensitivity_45 B5 into the NationalClosure_41 verdict cells); v0.66 male-cohort terminology and provenance repair - sex-neutral "able-bodied" labels replaced per ruling 2026-07-10, SRC-MALE-COHORT registered; no formula, value or verdict change. v0.67 check-and-report layer repair: QAChecks_08 B121 rebuilt to test the 1.6M and 2.0M grid blocks against the live ceiling (was testing the 2.5M global minimum and returning a false FAIL); Findings_09 F6 repointed from Model_04 B51 / Inputs_03 B67 (a smelting labor coefficient) to Model_04 B65, correcting the published copper multiple from 7.46x to 7.76x and aligning it with SensitivityBreakEven_06 B10 and the paper. No burden, roster, feedback loop, M3 component or verdict changed. v0.68 badge and provenance hardening: seven inputs re-badged off SOURCED where the note recorded engineering judgment, a planning value or an unverified attribution (metabolic efficiency, mechanical work per man-day, desert water need, rest days, leave/festival days, Chatham ropemakers, block-mass point selection); median male age at death re-badged DERIVED as the midpoint of the sourced 30-35 cemetery range with the replacement-multiplier floor at age 35 disclosed; vertical lift re-badged CONTROL after a full-workbook scan confirmed zero consumers and that the climb is priced through the B51 incline multiplier, which is not derived from it; village capacity relabelled to disclose its dual role as capacity comparator and village-population-ratio numerator; rope project lifecycle wired from a hardcoded 30 years to the build-duration lever B81; Bir Nasib and mortality citations carried into the input notes and register; SRC-MORTALITY and SRC-ROPE-CHATHAM confidence lowered to Medium with explicit pin tasks; mortality band added to DemographicClosure_46. No headline value, no verdict change. v0.69 methods-rulings register staged as a Tier-0 documentation working file; not promoted, folded into v0.70. v0.70 comparator-basis ruling and conformance: MethodsRulings_48 added as a pure-sink register of closed methodology questions, seeded with MR-01 (ground-haul 500 m lever retained, 2026-07-08) and MR-07 (timber comparator basis, ruled 2026-07-24, reaffirmed 2026-08-01) and carrying a plain-language timber-chain preamble and worked loss-stack example; IERRampTimber_38 B32 repointed from B26 to B30, moving the cedar-comparator ratio from the installed-volume basis to the felled basis and the published multiple from 9.74x to 25.23x, matching TimberSystem_31 B40 which was already on that basis; both comparator row labels rewritten to state their volume basis explicitly; SRC-CEDAR-PROCESSING-STATE registered at Low confidence, recording that the Palermo entry specifies ships of cedar wood without stating processing state, badged UNATTESTED with the primary reading task open; IERRampTimber_38 row 49 policy pointer added. One formula change, one value change (B32 only). No roster, burden, feedback loop, M3 component or verdict changed. v0.71 Gate 3 break-even and hardcode promotion: SensitivityBreakEven_06 gains the granite shore-transfer line - the lifting footprint at which bearing pressure meets saturated-bank capacity (5.33 m2 against a current 1.5 m2 assumption, a factor of 3.56, or 44% of the beam's own bearing face), so the Gate 3 finding is stated as a threshold rather than an assumption; GraniteAswan_42 B23 promoted from a hardcoded 15,000 literal to Inputs_03 B194 via Assumptions_02 E95, and B20's hardcoded 8 x 1.5 beam face promoted to Inputs_03 B195/B196; B23 badge corrected from SOURCE to ASSUMPTION - the value is defensible and generous but carried no register entry - and SRC-BANK-BEARING registered at Low confidence with the citation task OPEN. Two formula rewrites, both value-neutral; zero numeric change to any pre-existing cell; Gate 3 status unchanged. v0.72 Gate 3 rescoped to the Giza delivery end (ruling MR-03): Aswan sits at the First Cataract, where granite bedrock underlies the city and extends along the eastern bank at the quarry and the floodplain is almost nonexistent, so soft saturated alluvium does not describe the loading end. The bearing-pressure test and its break-even now apply at Giza, where the beam is transferred off the barge onto floodplain and harbor-basin ground. The Aswan-end position is stated affirmatively in GraniteAswan_42 row 25 rather than dropped: the burden there is unchanged in kind - an unattested lifting mechanism and its geometry, carried by Gates 1, 2 and 4 - and proposed load-at-grade routes are recorded as the open pricing task SRC-GRANITE-LOADING-PROPOSALS. Value unchanged at 15,000 kg/m2; scope and description only. Zero numeric change anywhere in the workbook. v0.73 Gate 4b transfer stability: the beam has to get ONTO the hull, and a floating hull moves as it is loaded. GraniteAswan_42 rows 44-53 price the deck-edge drop at the loading edge as 4W/A - parallel sinkage plus the trim or heel contribution - which is 0.410 m on the modelled 60 x 13 m hull against a generous 0.10 m transfer tolerance, 4.10 times over. End-loading and side-loading derive to the SAME value, so no choice of loading method escapes it; verified against a numerical hydrostatic solve with no small-angle assumption, agreeing within 0.32%. SensitivityBreakEven_06 rows 20-24 carry the break-even: holding the loading edge within tolerance needs 3,200 m2 of waterplane, or 246 m of hull at the modelled beam, against the largest attested Egyptian vessel at roughly 1,323 m2. Restraint the mooring must hold is priced at 32 t per beam. Hull length, hull beam and transfer tolerance added as Dashboard levers (Inputs_03 B197-B199), all set generous. SRC-BARGE-DIMENSIONS registered at Low confidence - no Old Kingdom heavy-cargo hull is attested with dimensions. Ruling MR-04 anchors transfer stability to BargeTimber_43!B12 (218.0 t solved) rather than Inputs_03!B168 (90 t asserted): the 4W/A identity holds only while the unloaded end stays immersed, and on the 90 t basis it understates by 233%. A three-tier validity guard reports which band the model is in. Gate 4 rows 27-29 still read B168 and are NOT YET CONFORMED. Zero numeric change to any pre-existing cell. v0.74 Gate 4b corrections following external review: restraint friction basis moved from Inputs_03!B166 (0.40, unimproved ground) to B4 (0.30, wet track, badged favourable to the conventional scenario) per ruling MR-05, since the transfer runs across a prepared ramp and a timber deck - restraint falls from 32 t to 24 t per beam, the generous reading, and now closes against the stated push force. The attested hull comparator is promoted from prose to levers (Inputs_03 B200/B201, 63 x 21 m) with waterplane, deck-edge drop and tolerance multiple computed live at GraniteAswan_42 rows 54-56, so no figure on the granite explainer is a hardcode. One value change (B53, 32 -> 24). v0.75 cross-reference repair: ReadMe_00 B77 bumped from v2.5 (wp25) to v2.9 (wp29). The cell had gone stale across four paper releases - v0.70 corrected it once and no later pass re-scanned it, the same failure mode the version-string scan was built to catch. Three MethodsRulings_48 affected-cell entries also stopped citing versioned paper filenames (wp23, wp27, wp29) and now cite sections, and a fifth register rule records why. ReadMe_00 A3 had named White Paper Draft v1.8 since that release, eleven versions stale and never scanned because it sits in prose rather than a labelled field; it now points at B77 rather than carrying a version of its own, so it cannot go stale again. No model content changed. v0.76 granite wording and scope corrections, no calculation touched. Two notes had described the Aswan-Giza corridor as a "~900 km round trip" when Inputs_03!B193 (890 km, SOURCE) is each way; corrected to ~900 km each way / ~1,800 km round trip. B193 itself is unchanged and remains the computational input - prose reports ~900 km because the Nile branch serving Giza ran closer to the complex in the Old Kingdom and the course has shifted, so a to-the-kilometre ancient figure would imply false precision. GraniteAswan_42 E34 no longer implies the 1,601-person corps rides the river corridor: it works the ~5 km overland legs at each end, while the corridor is crewed and priced separately in RiverHarbor_47. The beam basis is now stated consistently in three places as a conservative modelled subset of 30 major granite beams in the ~50-80 t class, generous-low against the forty to fifty in the relieving chambers alone. Two objections added: an overland-haul rescue does not remove the shore-to-vessel transfer, only shortens the floating leg, and the burden of supplying a credible dry route to the west-bank plateau sits with the critic proposing it; and the 890/900 precision question is answered directly. Zero numeric change. v0.77 publication-reference alignment. ReadMe_00 B77 now names the PUBLISHED white paper draft, v3.0 (wp30), and its scope is stated in the cell: it moves on publication, not on every working revision. That definition is the fix for a recurring failure - B77 went stale across four releases before v0.75 and again immediately after, because it had been tracking whichever draft happened to be current when the workbook last shipped. No model content changed. v0.78 timber delivery equivalence wired live. The shipload figures the project has published were derived outside the workbook; TimberLabor_40 rows 25-33 now compute them from B7 with cedar per-ship volume as a Dashboard lever (Inputs_03 B202, Assumptions_02 E103, generous-high - a LOWER value raises every count, so the published figures are the conservative end). Annual shiploads, cumulative shiploads, the comparator ship count and the multiples are all live. SensitivityBreakEven_06 rows 26-29 add the delivery break-even. Rationale is stated on the sheet: Egypt has no structural softwood stand, so procurement is delivery-limited rather than volume-limited and the ship is the operative unit. Zero numeric change to any pre-existing cell. v0.79 comparator-basis fix and two disclosures, raised by external review before the shipload figures were bound to an explainer. TimberLabor_40!B31 had derived the forty-ship record as comparator volume divided by the per-ship lever, locking the two: lowering per-ship volume raised the "record" to 54.9 ships and the headline multiple B32 reduced algebraically to B7/B146, invariant to the lever. Ruling MR-06 makes the ship count the sourced quantity (Inputs_03!B203 = 40, SOURCE), leaves per-ship volume as the ASSUMPTION lever, and derives comparator volume as count x volume; value-neutral at DEFAULT. A pure-sink steelman mirror at TimberLabor_40 rows 35-43 records that at 2.0x plank service life the ANNUAL requirement falls below the record while the build-long requirement stands - so the annual lead holds at DEFAULT and the build-long figure is the one that survives the steelman. WoodFuelCredit_44 rows 29-30 state that mortar and gypsum firing fuel is not quantified anywhere in the model, so the 2,205 km2 understates the woodland requirement in the conventional account's favour. v0.80 publication of white paper draft v3.2 (wp32). ReadMe_00 B77 moved from v3.0 to v3.2 on publication, which is the only event that moves it. No model content changed. v0.81 external-review corrections, five value changes and three new rulings. MR-08 puts Test 1 on a dimensionally consistent basis: the rostered numerator embedded a 20-year mortality flow and was being divided by a single-moment standing pool. Concurrent-over-standing and cumulative-over-cumulative return the identical share, so the correction is unambiguous; Test 1 moves from 70.5% to 61.7% and the verdict is unchanged against the 25% benchmark. MR-09 fires the Weni verdict on the concurrent basis only - Weni levied once, so the cumulative comparison was dimensionally invalid; the finding is restated as mobilization at the scale of the maximum documented levy held continuously, rather than categorically exceeding it. MR-10 separates Snefru-ship EQUIVALENTS of total timber throughput from the literal imported-cedar requirement, which the imported-share lever puts at 30.09 shiploads a year, below the record annually and 15.05x over the build. Gate 4 refloat conformed from the asserted 90 t to the solved 218 t per MR-04, which had been outstanding. The 25% ceiling is retained as a comparative benchmark and the requirement is now also stated directly as a break-even a conventional scenario must demonstrate. A conjunction block at DemographicClosure_46 rows 125-135 states that the failures are simultaneous: a demographic rescue is denominated in people and touches nothing in timber, copper, granite transfer, barge stability, refloat or settlement. Two objections added on population rescue and seasonal labor. v0.81 propagation pass (same version, in place): the MR-08 correction had been applied to the NationalClosure_41 headline but not to everything downstream. ClosureSensitivity_45!B4 still read the rostered numerator, so the whole grid was computed on the superseded basis; it now reads the concurrent figure and the grid moves with it - the 1.6M extreme corner becomes 28.2%, and two 2.0M corners fall below the benchmark at 24.2% and 22.6%. The population blocks are relabelled to the boxed-canyon treatment: 1.6M is the sourced anchor, 2.0M an out-of-period favorable stress test, 2.5M a later-era extreme. QAChecks_08 B121 now tests the sourced anchor block only, since the 2.0M corners legitimately fall below on the corrected basis. The DemographicClosure_46 mortality-band commentary is restated on the concurrent basis: the band is 0.825x-1.140x, not the rostered 0.83x-1.52x, and the exceedance-at-the-midpoint claim is withdrawn. Findings_09 F24 and F25 are rebuilt live, ReadMe_00 B80 corrected, and the MR-04 status closed now that Gate 4 rows 27-29 conform. A third review pass closed MR-04 completely: the status column had been updated when Gate 4 conformed but the reasoning and affected-cells columns still read "NOT yet conformed", leaving the authoritative register contradicting the model. A fifth register rule now requires a ruling to be closed in every column in the same edit. ClosureSensitivity_45 A2, the section 2 wave list and the model-architecture table header were also brought current. White paper draft v3.3 (wp33) is published alongside this workbook, and ReadMe_00 B77 moves with it. v0.82 integration-status reconciliation, raised by external review of white paper draft v3.4. DiagnosticIntegration_35 carried statuses that contradicted the wave assignments and, in one case, itself: ShapingSetting_33 was marked NOT READY in column H while column I read YES and the module sheet read INTEGRATED INTO M2/M3. Statuses are now stated per module - shaping/setting labor integrated as Wave 1 with the copper/tool diagnostic (1.3797 t/yr) explicitly not integrated, timber labor integrated as Wave 3 with the material screen control-only, ramp integrated as Wave 4, IERRampTimber_38 control-only - and a PARTIAL tier is added because a two-way ready/not-ready split could not express a module with integrated labor and an unintegrated material chain. Findings_09 F14 and F16 keep their historical statements and gain a dated status-update clause. Two count cells move (B18, B19); no burden, roster, feedback loop, M3 component or verdict changed. v0.83 household-factor repair (ruling MR-11), raised by external review. Inputs_03!B87 "People fed per worker" was wired to Assumptions_02!E40 "Sled length" - identical at DEFAULT 4, divergent at STEELMAN 3.5 - so a sledge dimension silently drove family size, food demand, sanitation burden and the national population result through 22 consumers in 9 sheets. B87 is now a literal 4, relabelled "Dependents per project worker", note retained. Twenty-one consumers computing a worker-plus-dependent population now take (1+B87); VillageSufficiency_39!B19 keeps B87 alone because it wants dependents. Off-site population rises from 201,383 to 251,033 and M3 from 483,769 to 533,419, because off-site workers had been excluded from the population they belong to. The 5.95 settlement treatment, the water-feedback solution, the on-site roster and Test 1 are untouched. No assumption row added, no lever introduced, Inputs_03!B49 not activated. v0.84 four verified corrections. MR-12: MR-08 propagated into the population architecture - a concurrent multiplier (Model_04!B115 = coverage only) and a concurrent on-site ledger (F116:G122) sit beside the cumulative-distinct ledger; village population B123 and the water-feedback solution now read the concurrent figures, removing a 20-year mortality flow from a single-moment population and from the daily water coefficient. The 5.95 settlement ratio is unchanged. MR-13: Model_04!B48 is quarry-only, matching its label and Exclusions_11; the shaping/setting crew is removed from the quarry copper-user count so the separate shaping copper diagnostic is no longer double-charged, and the 7.76x Bir Nasib multiple becomes 7.16x. WoodFuelCredit_44!B8 converts its two volume streams at 0.7 t/m3 before adding the tonne-based barge stream. VillageSufficiency_39!B8:B10 read the computed sanitation, medical and cleaning headcounts instead of a service rate and two blanks; container labor is stated as unquantified. ShapingSetting_33 stale publication text is brought into line with the v0.49 Wave 1 ruling. All four raised by external review. v0.85 Aswan granite mass-transport burden (MR-14) and verdict conformance. Total Aswan granite of 8,000 t (Hemeda & Sonbol 2020, SOURCED Medium) is registered; the 30 x 80 t module is now the HEAVY-LOAD SUBSET within it, with the reconciliation visible - 8,000 = 2,400 subset + 5,600 remaining, the remainder a floor because the subset is priced at design maximum. Throughput for the full mass is priced on the sourced Merer schedule at GraniteAswan_42 rows 57-82 at one design-mass piece per voyage - 100 voyages against 30, 3.79 fleet-seasons against 1.14, 182,019 crew-days, 182 t grain; the finding is more operating seasons on the same fleet rather than more concurrent crews. The 1,601 heavy-beam crew is not scaled by mass. Quarry extraction, haul and load/unload of the remaining mass are UNPRICED (no granite-specific rate); shaping, fitting and setting are EXCLUDED (no component geometry) - both conservative and stated. Weni verdict strings gain a below-0.9 tier and now read 0.77x at DEFAULT on the propagated concurrent basis; the sensitivity-grid verdict and QAChecks_08 B121 are re-scoped to the DEFAULT-lever cell at the sourced anchor, since two extreme corners cleared on the corrected basis and exceptionlessness was never the claim. Payload denominator corrected before acceptance: an initial build divided granite by the laden hull mass at BargeTimber_43!B9 (101.2 t, including rigging, crew and provisions) and returned 80 voyages; the cargo denominator is the 80 t design piece, and voyages are 100.

Current open modules

SanitationMedical_29, ProcessingFuelContainers_30, TimberSystem_31 (material screen), RampSystem_32 open levers, IERRampTimber_38, and WaterInfraAlt_27 are readiness-gated diagnostics/controls outside M3. FoodFeedback_28 is a separate published fork (reconciled ceiling ~880,300 / 55.0%).

Diagnostic governance

DiagnosticIntegration_35 governs non-additive diagnostic exposures and M3 readiness gates.

Validation

Python engine validated 155/155 outputs against the recalculated workbook; full automated test suite passing (count reported on /readme from the repository).

Falsifiability

Reversal conditions for each standing verdict are specified in advance in the falsifiability statement. Any proposed burden reduction can be entered through the model levers and the verdict re-read.

Mortality band - live mirrored sensitivity block

MORTALITY BAND (DemographicClosure_46 rows 44-123). The mortality band is a LIVE MIRRORED SENSITIVITY BLOCK, not manual output. It duplicates the 61-cell dependency chain from Inputs_03 B88 (median male age at death) to the verdict cells with live formulas, once per scenario: column B at age 30, column C at age 35, column D at the live age lever. It is a PURE SINK - no formula anywhere outside those rows reads any cell inside them, and the block feeds no burden, roster, feedback loop, M3 component or verdict. It is not a user input area; every cell is a formula and the only mortality input remains Inputs_03 B88. Column D exists as a DRIFT GUARD: because it mirrors the chain at the live lever it must equal the live model exactly, and DemographicClosure_46 B117 tests that, surfaced in the check panel at QAChecks_08 row 122. THE DRIFT GUARD DETECTS STALE MIRROR FORMULAS BUT DOES NOT REPAIR THEM: on FAIL the mirror must be regenerated from the source chain. The mirrored chain must also be regenerated after ANY structural change to the labor or demographic dependency path - adding a department row or a new feedback term changes the 61-cell cut set, and the guard may still pass while the endpoint columns are wrong.