This report covers only the second OBDLink segment of the day (filename CSVLog_20260711_060115, 06:01:15 local start), running from just west of Williams, AZ through arrival in Salida, CO — 42,690 s (11.86 hr) elapsed, crossing from MST (Arizona) into MDT (Colorado) partway through. The first segment — Prescott departure, Chino Valley, the approach into Williams, and a ~45-minute accident-related delay, all recorded before an OBDLink dropout around 04:50 local — has not yet been uploaded and is not reflected in any total below. Conditions were hot throughout (peak 100.4°F OAT near Durango), with an early tailwind (~5–11 mph, driver-confirmed) giving way to a gusty right crosswind with a modest headwind approaching Salida.
Totals below cover the full uploaded file (06:01:15 local start through arrival), including brief idle/stopped periods. State-B duty-cycle and Section 03 incremental-fuel figures use the idle-filtered (gear 70/128/130 removed) moving-only dataset per standing convention.
| Metric | Value | Note |
|---|---|---|
| Total fuel | 44.36 gal | measured — Trip Fuel PID, cross-checked against rate-integration (44.36 gal, 0.02% agreement) |
| Total distance | 574.03 mi | measured — Trip Distance PID |
| Raw fuel economy | 12.94 mpg | combined driving + ProPower + brief idle + grades; not propulsion-only |
| ProPower delivered (moving) | 16.75 kWh AC | 5 charging windows; feeds Section 03 baseline calc |
| ProPower delivered (while stopped) | ~1.42 kWh AC | additional, during fuel/rest stops — excluded from Sec. 03 conversion (idle fuel dynamics differ), noted here only |
| State B (ProPower active >4 kW) duty | 25.7% | of moving time — BMS-cycled, not a deliberate test segment (see Sec. 03 note) |
| Truck HV battery SOC start → end | 34.4% → 53.6% | F-150's own hybrid battery, not the LightShip's 77 kWh pack |
| LightShip SOC | 100% at departure only | driver-reported (Atlas display); no end-of-day reading captured verbally this leg — flag, follow up |
All five windows below are BMS-driven ProPower cycling, not a deliberate Phase A/B test design — segment boundaries follow the LightShip's own battery-management setpoint logic (per the Thermal Load Management instruction added 2026-07-10), not route markers. Durations (10–73 min) fall within the expected "up to ~1 hour per cycle" pattern, with the longest window slightly exceeding that as a data point to note, not a fault.
| Window (local start) | Duration | kWh AC | Measured gal (÷7.70) | Modeled gal (÷8.54 theoretical) | Δ meas–model |
|---|---|---|---|---|---|
| 08:04 MST | 28.6 min | 2.81 | 0.365 | 0.329 | +10.9% |
| 09:50 MST | 35.0 min | 3.45 | 0.448 | 0.404 | +10.9% |
| 13:39 MDT | 73.1 min | 7.16 | 0.930 | 0.838 | +10.9% |
| 15:56 MDT | 24.0 min | 2.35 | 0.305 | 0.275 | +10.9% |
| 17:56 MDT | 10.0 min* | 0.98 | 0.127 | 0.115 | +10.9% |
| Total | 170.7 min | 16.75 | 2.175 | 1.962 | +10.9% |
*Window 5 continued for ~14.5 more minutes / 1.4 kWh after the vehicle went idle at arrival — shown in Section 02 as the "delivered while stopped" line, not here.
The "measured" column is the established 7.70 kWh AC/gal baseline (KB doc 01, ~22.7% full-chain efficiency) applied to this leg's delivered energy — it is not independently re-derived here, consistent with the paired-elevation methodology's requirement for a controlled ON/OFF comparison this leg doesn't provide. The "modeled" column applies the 8.54 kWh/gal theoretical (no-loss) figure for comparison. The constant +10.9% gap between columns is arithmetic (same ratio applied throughout, 8.54/7.70), not a new finding — it restates the known measured-vs-theoretical efficiency gap.
| Term | Value |
|---|---|
| ProPower coefficient (b) | 0.110 gal/hr per kW |
| Implied rate at 5.8 kW | 0.638 gal/hr |
| R² | 0.587 |
| Mass basis | estimated ~7,053 kg (15,550 lb) — CAT scale documentation pending upload |
First attempt was a direct 60–64 mph ON-vs-OFF fuel-rate comparison — result was ≈0 gal/hr (even slightly negative), because this leg's OFF-state rows in that speed band averaged ~934 ft higher altitude than the ON-state rows, a grade confound overwhelming the signal (the same failure mode documented for the June round trip's return leg). Altitude-binned stratification didn't resolve it either — per-bin differences swung from −1.92 to +1.14 gal/hr, since altitude alone doesn't capture instantaneous grade.
Switched to an OLS regression using this leg's full moving dataset: fuel_rate = a + b·(ProPower kW) + c·(grade power, from smoothed TrackLogger altitude rate) + d·(speed³ proxy). This isolates the ProPower coefficient by letting grade vary as its own term rather than relying on route symmetry — no round trip required. Result: 0.638 gal/hr at 5.8 kW, landing inside the measured baseline's 0.55–0.75 gal/hr range and close to its 0.64–0.65 gal/hr center. R²=0.587 leaves real unexplained variance (no hybrid-battery-SOC term, uncalibrated speed³ aero proxy pending the Cd·A regression, altitude-rate noise even after smoothing) — this is a same-leg exploratory corroboration of the measured baseline, not an independent re-measurement of it, and should be presented to LightShip as such.
| Quantity | Value |
|---|---|
| Mean head/tail component, whole leg (HRRR-blend × TrackLogger heading) | +2.0 mph (net headwind) |
| Wind-adjusted expected yield | not computed — pending Cd·A regression |
| Residual after wind adjustment | not applicable this leg |
The blended wind estimate (3-anchor inverse-distance weighting: Williams AZ, Cortez/Durango, Salida) shows a net leg-average headwind of ~2 mph, masking a real shift — early tailwind of 5–11 mph across Arizona/Utah/western Colorado (matches driver-reported position callouts closely) giving way to a gusty right crosswind with a modest headwind near Salida (also matches the driver's own late-leg report). A rigorous wind-adjusted yield model requires the Cd·A regression from the Method-2 speed-hold protocol, not yet run — this section reports the wind observation only, deliberately not fabricating an adjusted-yield number without that input.
No State A activity is reported for this leg. TrekDrive was deliberately minimized per the hot-weather thermal management plan, to avoid stacking battery debt on top of the day's thermal load. Separately, LightShip's TrekDrive axle motor and battery sit on the LightShip's own CAN network, not the F-150's — so State A engagement isn't directly observable in this OBDLink capture even when it occurs. This is the expected, planned outcome for this leg, not a data gap.
| Segment | Dist mi | Battery kWh | Fuel saved gal | Net vs fuel-equiv |
|---|---|---|---|---|
| None recorded this leg. | ||||
Rich-λ mask: λ < 0.95, absolute load > 65%, RPM > 1200 (excludes tip-out/DFCO transients), sustained ≥5 consecutive rows (≥10 s at the 2.0 s frame rate), restricted to OAT > 86°F. Fifteen brief episodes met all criteria, clustering into four geographic groups below — all on sustained mountain grades, not the flat hot-plains condition the protocol is really built for (that's Broomfield→Oshkosh). Total confirmed penalty is small but real.
| Segment | Rows λ<thr (sustained) | Peak OAT °F | Mean rich λ | Est. penalty gal |
|---|---|---|---|---|
| US-160 near Kayenta, AZ/UT border | 17 / 34 s | 96.8 | 0.910 | 0.0145 |
| US-160 Cortez–Durango grades, CO | 61 / 122 s | 96.8 | 0.871 | 0.0653 |
| US-160 Durango–Pagosa Springs grades, CO | 34 / 68 s | 96.8 | 0.889 | 0.0309 |
| US-160/US-285 approach to Wolf Creek Pass, CO | 5 / 10 s | 91.4 | 0.902 | 0.0030 |
| Total | 117 / 234 s (~3.9 min) | 100.4† | — | 0.114 |
†Leg-wide OAT peak (100.4°F); no single sustained-rich episode occurred at the exact peak moment — episodes cluster at 91–97°F under high grade load.
All four clusters are climbing-grade events, not State-A-vs-State-B-at-temperature comparisons — TrekDrive was minimized this leg (Section 04), so there's no same-temperature no-assist/assist pair to contrast here. IAT2 peaked at 156.2°F, well above the 140°F Tier-3 throttle-closure threshold, for 133 rows (0.7% of moving time) — see Section 06. The hot-plains protocol remains the priority target once past Denver.
Driver observations recorded during this leg's drive. These are testimony, not telemetry — preserved substantially verbatim, neither measured nor modeled.
Cross-referenced directly in Sections 05, 06, and 08 where noted above.