Yield decline on BR-UP-01 · MES-26073
INV-104HYPOTHESES_UNDER_REVIEWThree candidate explanations, each with the evidence that supports or contradicts it. Nothing here is an approved root cause; the deviation Q-DEV-031 remains the system of record.
Yield trend
Calculated81.2 → 75.9
MES-26071 → 26073
Feed droop
Observation−14.2%
after T+264.4h
Head pressure peak
Observation1.48bar
ceiling 1.20 bar
Hypotheses
1 / 3
supported · 2 contradicted
Mechanical tubing fatigue / slip on PM-204
Supported · not establishedOver-occlusion after the WO-8819 clamp adjustment, on tubing near the end of its rated life, raised discharge pressure and reduced delivered volume per revolution.
Supporting · 5
R² = 0.941 between feed loss and head pressure across 3 batches (n = 237 hourly points, imputed excluded). Correlation, not causal proof.
CalculatedWO-8819 raised clamp torque 0.42 → 0.55 N·m ~48 h before the spike; dynamic slip check not performed.
ObservationSOP-UP-014 §7.2 requires a slip verification after any occlusion adjustment.
ObservationTubing at 412 h of a 400–500 h OEM service life; OEM notes higher occlusion shortens life and raises pressure.
ObservationPT-204 pressure transmitter calibrated Oct 3, as-found +0.3% FS — the pressure reading itself is credible.
Observation
Contradicting · 1
MES-26072 declined 2.8 pts with no clamp adjustment and normal pressure — tubing slip does not explain that batch.
Observation
What would change our mind
- Gravimetric slip test on the removed tubing at 0.55 N·m shows delivery within ±3% of setpoint.
- Fresh tubing at 0.42 N·m reproduces the same pressure rise at the same RPM.
- Feed-loss-adjusted yield model leaves most of the 6.1-pt gap unexplained.
Raw feed media lot variation (FM-2207)
ContradictedA nutrient-composition difference in the feed lot reduced productivity.
Contradicting · 3
CoA within specification: osmolality 312 mOsm/kg (290–330), glucose 198 g/L (190–210).
ObservationSame lot FM-2207 fed MES-26072, which showed no T+264h feed drop.
ObservationTiter trend diverges only after the feed-flow drop, not from feed start.
Observation
What would change our mind
- Trace-element or hydrolysate assay shows FM-2207 outside historical range (not on CoA today).
- An alternate-lot batch reproduces the decline with normal feed delivery.
Dissolved-oxygen sensor drift
ContradictedA drifting DO probe caused control to under-aerate, limiting cell productivity.
Contradicting · 2
Redundant probes DO_A and DO_B agree within 0.6% over T+200–280h.
ObservationDO stays inside the 35–45% band on all three batches.
Observation
What would change our mind
- Off-line DO calibration check shows both probes biased in the same direction (common-mode drift).
Test plans
Counter-evidence tests designed to confirm or rule out each hypothesis
TP-01Hypothesis APlanned · not executedOwner: K. Daniels (Maintenance)Gravimetric slip test of removed tubing TS24-PT-5521 at 0.42 and 0.55 N·m, 184 RPM, 5 min each
Acceptance criterion: Delivery within ±3% of setpoint at 0.42 N·m; >5% shortfall at 0.55 N·m supports A
TP-02Hypothesis BPlanned · not executedOwner: M. Okafor (QC)Retain-sample trace-element panel on FM-2207 vs two prior lots
Acceptance criterion: All elements within ±2 SD of historical lots keeps B contradicted
What remains unknown
- UnknownWhether feed loss alone explains the full 6.1-pt yield gap on MES-26073.
- UnknownWhy MES-26072 declined 2.8 pts without a pressure event.
- ImputedTelemetry for 42 minutes at T+182.0h (PI buffer overflow). Imputed values are excluded from every statistic.
- UnknownPhysical condition of the removed tubing — not yet inspected; no camera or photo evidence connected.
Status is separate from approval
Your companion
Here with you in BioDemo