SPK-07

Workforce, Supply Chain & Transition Continuity Kernel

The United States is attempting to run two full-scale strategic missile programs simultaneously — sustaining Minuteman III through 2050 while fielding Sentinel — with an industrial base, workforce, and infrastructure sized for one. The Air Force assessed 2050 operation as "feasible" only if "entire subsystems" including propulsion and guidance are replaced.[2] Neither replacement program exists. Meanwhile, the propellant is aging toward a viability horizon that the Air Force's own assessment says requires action. Guidance spares are depleting — the Air Force is already reducing test launches to conserve parts.[13] The security certification pathway for Sentinel is suspended with no schedule. There is no viable path that both maintains 400 operational ICBMs and respects the Air Force's own 2050 feasibility conditions under current funding. WST-K identifies the only feasible configuration and the decisions that must be made by ~2028 — or the statutory requirement of 400 operational ICBMs[5] becomes physically impossible to maintain.

Zenodo DOI: 10.5281/zenodo.18933132

WST-K constraint architecture: WCRM++, CISG++, DCTC++, HCM, and PTLFM components with cross-domain propagation chains
0

Two Programs. One Industrial Base. Zero Margin.

The January 2024 Nunn-McCurdy breach[1] — 81% cost growth to $140.9 billion — obscured a parallel crisis. GAO-25-108466[2] confirmed Minuteman III operation through 2050 is "feasible" but identified three unmitigated risk categories: parts obsolescence, workforce capacity constraints, and industrial base fragility. These are not independent — they create compounding failure modes where workforce shortages accelerate parts cannibalization, parts shortages force unsustainable maintenance tempos, and both delay Sentinel, extending MMIII requirements further.

ICBM wings logged over 2.4 million maintenance hours over five years — a 30% increase.[3] A further 25% increase is projected[4] over the remainder of the decade.

ParameterValueSource
Statutory ICBM Floor400 operational10 U.S.C. §9062(n)[5]
Launch Facilities Required≥450§9062(n)[5]
Sentinel IOC TargetEarly 2030sBreaking Defense, Feb 2026[40]
MMIII Must Operate Until~2050GAO-25-108466[2]
The 50-silo offline limit used throughout this analysis is a mathematical deduction: 450 LF mandate minus 400 deployed floor = 50 maximum offline at any time. The Air Force drew down from 450 to 400 deployed ICBMs in 2017[6] to meet New START. New START expired March 2026.
What WST-K is not: WST-K validates whether proposed configurations satisfy capacity-closed constraints. It does not generate optimal staffing plans, estimate costs (UC-BCK domain), optimize schedules (SI-CK domain), or assess whether 400 ICBMs provide adequate deterrence against specific threat scenarios (policy domain). WST-K ensures legal compliance and identifies physical impossibilities.

Three Positions That Cannot All Be True

The Air Force has made three statements to Congress over five years. Each is documented. Each was made by named officials. Together, they are logically incompatible — any two being true makes the third false.

Position 1 (2021): Minuteman III Cannot Be Extended

STRATCOM Commander Admiral Charles Richard stated[45] at a Defense Writers Group roundtable in January 2021: "You cannot life-extend Minuteman III… You're quickly getting to the point where you can't do it at all." AFGSC Commander General Timothy Ray told the Senate Armed Services Committee[46] in May 2021 that a SLEP would cost $38 billion more than building GBSD. This was the justification for Sentinel: the existing system is beyond saving, so fund the replacement.

Position 2 (2024-2025): Minuteman III Can Operate Until 2050

When Sentinel breached at 81% and Congress asked what happens if Sentinel is late, the Air Force reversed. GAO-25-108466 reports[2] the Air Force assessed that operating Minuteman III through 2050 is "feasible" — but only if "entire subsystems" including propulsion and guidance are replaced. GAO characterized this as carrying "significant risk"[43] with "significant unknowns" in aging component performance.

Position 3 (2025-2026): The Required Replacements Will Not Be Funded

AFGSC Commander Lt. Gen. Gebara stated:[41] "A long-term SLEP still does not make sense for Minuteman. What is going to happen is Minuteman sustainment to keep it viable until Sentinel is delivered." No PRP-2. No NS50_RECAP. No budget line. No analysis of alternatives. No program of record. The Air Force chose "sustainment" — repairing what breaks with existing spares — over the subsystem replacement its own feasibility assessment says is required.

The Logical Structure:

A: Minuteman III cannot be life-extended (the argument for funding Sentinel).
B: Minuteman III can operate until 2050 (the argument that Sentinel delays are manageable).
C: The subsystem replacements required for 2050 operation will not be funded (the budget reality).

If A and C are true — the system can't be extended and replacements won't be funded — then B is false. Minuteman III cannot reach 2050.

If B and C are true — it can reach 2050 but replacements won't be funded — then the "feasible" assessment is wrong, because GAO documented that feasibility is conditional on those replacements.

If A and B are true — it can't be extended but also can reach 2050 — then C must be false: the replacements must be funded. But they are not.

All three positions are on the public record. They cannot simultaneously hold. The remainder of this page quantifies when the contradiction becomes a physical impossibility.

The word "feasible" is doing precise work. It means "not physically impossible." It does not mean planned. It does not mean funded. It does not mean likely. The Air Force is telling Congress that Minuteman III can last until 2050 in the same way that a journey is "feasible" when the vehicle exists but no one has bought fuel.

The PRP SAR[7] and Air Force completion announcements[8] confirm the original Propulsion Replacement Program is finished with no follow-on. The Air Force has explicitly ruled out a new SLEP,[44] choosing sustainment only — while the warm line produces 15 motors per year[10] against a need for 50-100, and guidance spares are depleting with test launches already being reduced to conserve inventory.[13]

What "sustainment" means versus what "SLEP" means

The distinction between "sustainment" and a formal Service Life Extension Program (SLEP) is not semantic — it is bureaucratic and consequential. A SLEP exceeding $100 million requires Headquarters Air Force approval through a formal submission process, triggering cost reporting, Congressional notification, and program oversight structures. "Sustainment" is managed through existing Weapon System Sustainment (WSS) accounts without the same visibility thresholds.

By characterizing the approach as sustainment rather than a SLEP, the Air Force avoids triggering the institutional accountability mechanisms that would force explicit acknowledgment of the scope of investment required. The warm line ($40M/year) and Boeing depot repair contract ($1.6B) continue under sustainment. But what GAO says is needed — replacement of propulsion and guidance subsystems — is a SLEP-scale effort that does not exist as a program.

Lt. Gen. Gebara's language is precise: sustainment "to keep it viable until Sentinel is delivered." This frames the problem as a bridge, not a destination. But if Sentinel IOC is the early 2030s and the propellant viability horizon begins in 2028, the bridge must carry traffic before it is built.

1

The Propellant Viability Horizon

The Propulsion Replacement Program delivered exactly 601 booster sets from FY2000-FY2009[7] at rates up to 8 per month. The propellant binder — a polybutadiene-family system — undergoes progressive aging through oxidative crosslinking, plasticizer migration, and interfacial debonding.

Binder identity unknown: The specific formulation (PBAN or HTPB) is not confirmed in open sources.[7] The PRP SAR notes "new materials and manufacturing processes to replace unavailable or environmentally prohibited materials." Open propellant aging literature indicates PBAN demonstrates slower oxidative crosslinking and longer service lives (decades) than HTPB (typically 13-20 years to critical failure criteria). This is the single highest-value piece of classified information for calibrating the entire analysis.

While degradation is progressive, the consequence exhibits threshold-like behavior: once propellant integrity drops below a critical level, crack formation under ignition pressure can rapidly increase burn surface area, potentially causing overpressure exceeding case burst rating.

Scenarioτ_propOldest Motors Hit ThresholdYoungest Motors
If HTPB (shorter life)20-25 years2018-2025 (may already be approaching)2027-2034
If PBAN (longer life)30-40 years2028-20402037-2049
The Air Force's own position resolves the ambiguity: GAO-25-108466 reports the Air Force assessed that SRM replacement is a condition of 2050 operation.[2] If the propellant timeline were not binding, replacement would not be a stated condition. A second Propulsion Replacement Program (PRP-2) is required. No such program currently exists. No budget line. No analysis of alternatives. No program of record. The Air Force identified the requirement and did not initiate it.

Motor Case Aging — PRP-2 Cost

The PRP SAR explicitly states Stage 3 composite motor cases "must be replaced"[7] during remanufacture — filament-wound composite cases cannot be recaptured. Stage 1 D6AC steel cases were successfully recaptured in the original PRP. The original PRP cost ~$2.1B[9] (~$3.7-4.0B inflation-adjusted). WST-K estimates PRP-2 at $8-12B (hybrid: Stage 1 repour + new Stage 3 cases).

Warm Line Capacity

The Solid Rocket Motor Warm Line produces up to 30 PRP-configured motors over a two-year period[10] — 5 full missile sets (15 motors) per year. To re-core ~400 missiles (1,200 motors) at warm-line rate: ~80 years. Required PRP-2 rate: 50-100 motors/year, requiring full production-line restart.

Propellant aging mechanism detail

Polybutadiene-family binders undergo oxidative crosslinking as atmospheric oxygen reacts with C=C double bonds, creating additional covalent crosslinks that increase the glass transition temperature and reduce strain-to-failure. The material transitions from rubbery (compliant) to glassy (brittle). Under ignition pressure (>1,000 psi in milliseconds), embrittled propellant cannot stretch to accommodate case expansion — it cracks, creating additional burn surface area and potential catastrophic overpressurization. WST-K models this as a conservative planning threshold (τ_prop ≈ 25-30 years from pour date) because the consequence is catastrophic and irreversible.

2

Guidance Depletion: Inventory Trending to Zero

The NS-50 Missile Guidance Set was installed under the Guidance Replacement Program from 1999-2008, designed for service through 2020.[11] The GRP produced 652 NS-50 units,[12] of which 450 are deployed and a maximum of 202 were available as spares, war reserve, and test expendables.

With no production line and ~16 years of test launches (historically 4-6 per year) and field failures since production ended, the remaining spares pool is substantially reduced. Simple arithmetic: 64-96 units consumed in testing alone, before operational failures and depot repair limits.

Behavioral evidence of scarcity: The Air Force has begun reducing operational test launch frequency specifically to conserve spares[13] — trading reliability validation for inventory preservation. GAO-20-296 found in 2020[14] that the weapon system would be "unable to meet full mission requirements after 2026" due to "expected attrition of current field assets."

The Air Force's 2050 feasibility conditions explicitly identify guidance replacement as a necessity.[2] No published sustainment plan demonstrates guidance inventory remains positive through transition completion.

NS50_RECAP: Required but Nonexistent

WST-K identifies the need for a guidance recapitalization program. No such program currently exists. The Air Force's current approach is limited to depot-level repair under a $1.6B Boeing sustainment contract (2021) and component-level remanufacture at Ogden ALC — approaches that extend the spares pool but do not add new production units.

3

The Clearance Bottleneck

Security clearance processing times have deteriorated significantly:

PeriodTS ProcessingTrendSource
FY23 Q3146 daysBaselineClearanceJobs/PAC PMO[15]
FY24 Q3241 days+65%
FY24 Q4249 daysPlateaued high

DCSA investigative inventory peaked at 291,200 cases (September 2024),[16] reduced to 222,700 by April 2025. The investigative workforce was significantly reduced from the OPM-era peak. GAO-26-108838 (February 2026)[17] confirmed TS processing times "consistently trended longer from fiscal year 2022 to 2025." The NBIS migration created additional friction — GAO-25-107325[18] found contractors manually entering information into two IT systems.

CNWDI: Final Clearance or Nothing

Access to Critical Nuclear Weapon Design Information requires FINAL clearance per 32 CFR 117.20.[19] No interim access. No waiver for contractors. This applies to physics package design, internal weapon design, and fuzing/firing set engineering — not standard mechanical warhead integration (RV mating), which requires S/RD access at a lower bar.

Engineers awaiting final clearance contribute zero productive hours to CNWDI-touching tasks specifically. Programs can partially mitigate by assigning non-CNWDI work during the 12-18 month adjudication window, but when CNWDI engineering is on the critical path, this mitigation does not resolve the schedule constraint.

The Training Wall

Component training durations documented in CFETP 2M0X3 — BMT (~2 months), technical school (~2.5 months), 5-level upgrade (12 months), 7-level upgrade (12 months) — combined with typical SSgt promotion timelines (~4-5 years) yield approximately 5-7 years from accession to fully qualified Team Chief. This timeline is not compressible — rank requirements impose promotion physics that training acceleration cannot overcome. The Team Chiefs needed for 2030 were recruited in 2024.

4

Nuclear-Critical Software: The Triple Compound Bottleneck

Three independent constraints compound on the same talent pool:

1. CNWDI gating: Only final-cleared, CNWDI-authorized personnel touch weapons-design code.

2. NSCCA (Nuclear Safety Cross-Check Analysis): Nuclear-critical software is NOT subject to Two-Person Concept (TPC), which applies to physical operations only (AFI 91-114; AFI 91-101).[20] Instead, AFMAN 91-119[21] requires independent verification by a structurally separate organization with "technical, managerial, and financial independence." For the ICBM force, this is performed under a $360M sole-source contract (FA8207-21-D-1001, Peraton, March 2021). One organization builds the software; a completely separate organization independently evaluates it. Both require independently cleared, PRP-certified, domain-qualified engineers from the same constrained pool. WST-K models this as ~2× demand (planning parameter).

3. Churn: Nearly half of cleared professionals leave within 3 years. With ~14 months to productivity and ~36-month median tenure, net productive years per hire ≈ 22 months. Efficiency ≈ 50-60%.

Combined effect: WST-K estimates nuclear-critical software positions require approximately 3-4× the naive FTE count — ~2× NSCCA overhead × ~2× churn inefficiency. This is a planning heuristic sensitive to both inputs; no published DoD figure exists for this subpopulation.
Geographic friction compounds the multiplier. F.E. Warren (WY), Malmstrom (MT), and Minot (ND) impose severe recruiting penalties for software talent. WST-K models effective workforce availability at remote missile bases at 50-70% of nominal planning assumptions due to PRP lifestyle restrictions and geographic isolation. The 3-4× multiplier applies to the reduced pool — not the national market.
Why NSCCA is worse than the paper originally assumed

The original analysis cited Two-Person Integrity (TPI) — two engineers at the same desk reviewing each other's code. The corrected mechanism is far more demanding: two entire organizations — structurally firewalled, separately contracted, independently cleared — must produce parallel engineering assessments of the same software. The $360M Peraton contract demonstrates the scale. The NSCCA evaluator faces identical recruitment, clearance, and retention challenges as the prime developer, competing for the same scarce talent pool. The talent pool competition effect is larger than the original analysis assumed.

5

Security Forces: A Massive Military Movement

Security Forces represent the largest manpower component of the missile wings:

WingLocationSF StrengthSource
90th SFGF.E. Warren, WY~1,250-1,400AF unit articles
91st SFGMinot, ND~1,000 AD + 140 ANG ≈ 1,140Commander bio;[22] NGB News[23]
341st SFGMalmstrom, MT>1,200341st MSOS article, 2021[24]
Total~3,540-3,740

The Sentinel transition requires removing 400 Minuteman III weapons and emplacing replacements across 32,000 square miles — characterized by AFGSC officials as "a massive military movement".[2] Strategic-distance movements use Prime Nuclear Airlift Force (PNAF) via C-17.[25] Ground convoys cover the tactical last mile — multi-domain operations[26] integrating CRF defenders, TRF recapture teams, and continuous MH-139A Grey Wolf helicopter escort.[27]

WST-K uses ~800 tactical ground movements and ~40 SF per convoy as planning parameters (actual counts classified). AFGSC Commander Gen. Bussiere testified[28] (April 2024) that "AFGSC Security Forces manning requirements will continue to increase due to nuclear modernization and transition." GAO-25-108466 reports AFGSC plans a 5% SF increase[2] for peak transition year FY2030. WST-K analysis indicates ~15-20% is required (modeling output, not official requirement).

6

Industrial Base: Three Layers of Fragility

Layer 1 — Prime Assembly Sites

Northrop Grumman operates two distinct SRM facilities:[29] Promontory, UT (large first-stage motors, Sentinel Stage-1 testing[30]) and Bacchus (Magna), UT (upper-stage and submarine-launched motors, including all three stages of Trident D5[38]). A Promontory outage does NOT halt Trident production. The April 2025 explosion at Promontory[31] destroyed a propellant-ingredient building but caused no program impact.

Layer 2 — The Documented Single Point of Failure

AMPAC Cedar City, UT is the sole DoD-approved domestic producer of ammonium perchlorate[32] — the oxidizer for ALL large solid rocket motors. The facility has experienced a prior fatal explosion (1997).[33] The primary capacity expansion — a $100M NewMarket investment[34] — builds additional production at the same site, increasing capacity but concentrating geographic risk. An AMPAC outage halts ALL large SRM production — both land and sea deterrent legs simultaneously.

Layer 3 — Collapsed Sub-Tier Network

The SRM supplier base shrank from ~5,000 to ~1,000 over two decades.[35] The National Defense Industrial Strategy Implementation Plan (2024)[36] acknowledges DoD "lacks a deeper understanding of the shared industrial base critical for solid rocket motor production."

Rayon/CRP Gap

NARC (sole domestic rayon supplier) ceased production September 1997.[37] DPA Title III awards (September 2025): ICF Mercantile $9.3M (rayon precursor), Americarb $12.6M (carbonization). Americarb first-phase production targeted October 2026, full expansion by September 2027. Qualification for strategic SRM applications requires additional multi-year testing. A potential gap period exists between legacy stockpile exhaustion and new-source qualification.

Rad-Hard Electronics

Honeywell Plymouth, MN is a critical trusted supplier of strategic rad-hard microelectronics, sustained by a $25.8M DPA investment for 90nm production. SkyWater Technology's RH90 platform (~$269M DoD funding) is in production maturity phase but has not been qualified for named strategic nuclear systems. Plymouth's loss would create multi-year disruption even though alternatives exist at various qualification stages.

7

The Security Certification Gap

AFMAN 91-118[39] and 91-119 require demonstration that security systems meet nuclear surety standards, but specify functional performance criteria — "criteria are not design solutions and are not intended to restrict the designer in the methods and techniques used." Certification can be achieved through models, simulations, calculations, and alternative testing codified in a Certification Requirements Plan.

The Air Force designated the Physical Security System Test Facility (PSSTF) at Dugway as the method for Sentinel. Construction is suspended[2] with no schedule. The Air Force is "exploring alternatives" but no acceptable substitute has been identified. Until the certification pathway is resolved, Sentinel missiles cannot be operationally certified regardless of hardware readiness.

Constructive opportunity: The governing manuals explicitly permit alternative certification approaches. The Air Force could potentially satisfy security certification through simulation, modeling, and alternative testing under AFMAN 91-118/119 rather than waiting for PSSTF construction.
8

The Threaded Continuity Regime

After modeling all constraints simultaneously, WST-K identifies one feasible configuration — the Threaded Continuity Regime (TCR). It requires ALL of the following:

ConditionStatusNotes
Sentinel IOC slip from post-restructure baseline ≤5 yearsTBD at MS-B (end 2026)Baseline "early 2030s"
PRP-2 initiated by ~2028No program of recordAF's own 2050 conditions require it
NS50_RECAP initiated late 2020sNo program of recordAF's own 2050 conditions require it
Security certification pathway resolvedPSSTF suspended, no alternativeAFMAN 91-118 permits alternatives
SF manning +15-20%5% planned (GAO-25-108466)WST-K modeling output
Software hiring ~3-4× nominalNot addressedWST-K planning heuristic
DPA rayon/CRP on scheduleFunded; in progressAmericarb first phase Oct 2026

Slip Classes

WST-K measures delay from the post-restructure Sentinel IOC baseline (estimated ~2032-2033, to be confirmed at Milestone B):

ClassΔT from BaselineConsequenceRequired Actions
Manageable≤ 2 yearsStandard MMIII sustainmentIncremental maintenance
Major2-5 yearsMMIII SLEP requiredPRP-2, NS50_RECAP initiation
Crisis> 5 yearsStrategic reviewFull reassessment; cliffs engaged

TCR Feasibility

The original PRP delivered 601 booster sets over ~9 years.[7] Current DPA investments expand SRM capacity. PRP-2 at required scale is feasible if funded. PSSTF is medium-scale MILCON with no technical barriers — or alternative certification under AFMAN 91-118 is possible. SF growth of 15-20% is aggressive but historically achievable.

Outside TCR — no feasible configuration exists:
If ΔT >5 years: propellant cliffs engaged, NS-50 exhausted. HALT.
If PRP-2 not funded: motors age out 2028-2039. D_count < 400 becomes inevitable.
If NS50_RECAP not funded: guidance inventory reaches zero.
If certification pathway not resolved: Sentinel IOC blocked indefinitely.
9

Hard Cliffs: Deadlines Physics Cannot Negotiate

CliffEarliestLatestTrigger
Propellant oldest (HTPB scenario)~2023~2030t > τ_prop(oldest, HTPB)
Propellant oldest (PBAN scenario)~2028~2040t > τ_prop(oldest, PBAN)
NS-50 guidance exhaustion~2028~2035Inv_NS50(t) = 0
Propellant youngest (any binder)~2034~2049t > τ_prop(youngest)
Security certification (if IOC ~2033)~2031~2032IOC minus τ_cert

All cliffs intersect the Sentinel delay window. The cost of inaction: a strategic deterrent that cannot maintain 400 operational ICBMs — not because of policy choice, but because of physics.

2028 is the decision point. PRP-2 must begin by then under any plausible binder scenario. NS50_RECAP must be underway. The certification pathway must be resolved. The Air Force's own 2050 feasibility assessment[2] identifies propulsion replacement and guidance replacement as necessary conditions — conditions the Air Force stated and did not fund. These are not technical problems awaiting engineering solutions. They are funding decisions awaiting policy action.

Cross-Domain Constraint Propagation

WST-K's core contribution: making compounding failure modes visible. Each chain shows how a constraint in one domain propagates to create failures in others.

Sentinel slip MMIII sustainment packages generated workforce/supply overload infeasibility
SF shortage conversion rate throttled schedule slip extended MMIII requirements
Hiring surge DCSA backlog inflated longer clearances for everyone slower hiring (self-defeating loop)
Propellant cliff PRP-2 required rayon/CRP demand DPA qualification timeline (chemistry propagates through materials, industrial base, and acquisition)
Certification pathway undefined SF cannot certify IOC impossible regardless of hardware readiness
NS-50 depletion flight testing and operations compete for last units (already observable in reduced test frequency)
CNWDI + NSCCA + churn = triple compound bottleneck: only final-cleared touch weapons code, a separate org must independently evaluate it, half the hires leave before peak productivity 3-4× multiplier

Acceptance Criteria

WST-K defines nine testable conditions. None are currently met.

AC-1
Statutory Floor: D_count(t) ≥ 400 computationally verified for all t in the transition period.
● UNMET
AC-2
Concurrency Bound: Offline_silos(t) ≤ 50 for all t.
● UNMET
AC-3
Workforce Closure: H_demand(s,r,t) ≤ H_avail(s,r,t) for all sites, roles, times.
● UNMET
AC-4
Supply Chain Closure: Consumption(k,t) ≤ Available(k,t) for all critical items.
● UNMET
AC-5
NS-50 Inventory Positive: Inv_NS50(t) > 0 until last MMIII retirement (monitored under uncertainty).
● UNMET
AC-6
Propellant Viability: Fleet_viable_MMIII(t) ≥ Motors_required(t) for all t.
● UNMET
AC-7
Security Certification: Sentinel IOC occurs only after certification pathway (PSSTF or accepted alternative) complete.
● UNMET
AC-8
Slip-Trigger Coupling: Sentinel slip >2 years automatically generates MMIII sustainment packages.
● UNMET
AC-9
HALT Monitoring: HCM generates HALT flags when cliff conditions reached.
● UNMET

Sources

  1. DoD Sentinel Nunn-McCurdy Review, July 2024. defense.gov
  2. GAO-25-108466, "ICBM Modernization," September 2025. gao.gov
  3. Gen. Bussiere SASC testimony, March 2023 (30% maintenance increase). armed-services.house.gov
  4. Atlantic Council, May 2024 (25% further increase). atlanticcouncil.org
  5. 10 U.S.C. §9062(n), FY26 NDAA. afmissileers.org
  6. AFGSC "AF meets New START," June 2017. afgsc.af.mil
  7. PRP SAR, December 2009. esd.whs.mil
  8. "Minuteman III replacement program wraps up," Arnold AFB, July 2009. arnold.af.mil
  9. DoD IG PRP acquisition, 2000. media.defense.gov
  10. Hill AFB SRMWL, December 2011. hill.af.mil
  11. GRP/NS-50 design life. globalsecurity.org
  12. DoD IG Report 97-199, "Guidance Replacement Program," July 1997 (652 NS-50 units produced). Archived; no stable web link.
  13. AFGSC test launch frequency reduction, November 2025. afgsc.af.mil
  14. GAO-20-296, "ICBMs to Fall Short," March 2020. gao.gov
  15. ClearanceJobs/PAC PMO data (249 days TS). clearancejobs.com
  16. DCSA Personnel Vetting Initiative. dcsa.mil
  17. GAO-26-108838, "Personnel Vetting," February 2026. gao.gov
  18. GAO-25-107325, "Trusted Workforce 2.0," May 2025. gao.gov
  19. 32 CFR 117.20 (CNWDI). ecfr.gov
  20. AFI 91-114, Safety Rules for the ICBM System (Jan 2023). e-publishing.af.mil; DAFI 91-101, Air Force Nuclear Weapons Surety Program. e-publishing.af.mil
  21. AFMAN 91-119, Safety Design and Evaluation Criteria for Nuclear Weapon Systems Software (NSCCA requirement). bits.de (1999 version)
  22. 91st SFG Commander bio. minot.af.mil
  23. 219th SFS ANG. nationalguard.mil
  24. 341st MSOS, 2021. malmstrom.af.mil
  25. AFI 13-527 (PNAF). e-publishing.af.mil
  26. CRF convoy operations. dvidshub.net
  27. MH-139A Grey Wolf. afrc.af.mil
  28. Gen. Bussiere SASC testimony, April 2024. armed-services.senate.gov
  29. Northrop Grumman SRM fact sheet. northropgrumman.com
  30. Sentinel Stage-1 test. af.mil
  31. Promontory explosion, April 2025. apnews.com
  32. DoD IG AMPAC audit, July 2020. dodig.mil
  33. AMPAC 1997 incident. lasvegassun.com
  34. AMPAC expansion. breakingdefense.com
  35. GAO-18-45, SRM suppliers. gao.gov
  36. NDIS Implementation Plan, 2024. govinfo.gov
  37. NARC rayon. ntrs.nasa.gov
  38. Trident D5 at Bacchus. naval-technology.com
  39. AFMAN 91-118. globalspec.com
  40. Sentinel IOC "early 2030s." breakingdefense.com
  41. Gebara: "long-term SLEP still does not make sense." airandspaceforces.com
  42. Arms Control Today: GAO feasible with replacement only. armscontrol.org
  43. Breaking Defense: "feasible, but with risks." breakingdefense.com
  44. Air & Space Forces: Minuteman III weapon overview (SLEP ruled out). airandspaceforces.com
  45. Adm. Charles Richard, Defense Writers Group roundtable, January 5, 2021. stratcom.mil
  46. Gen. Timothy Ray SASC testimony, May 12, 2021 ($38B SLEP cost delta). afgsc.af.mil