facility_financials.yaml now carries the
researched OKC numbers (OG&E $0.085/kWh, OKC Utilities $0.00582/gal at the
dripper, Athena Pro $0.06/gal mixed feed, OK cultivation labor $16/hr) — but
wholesale_price_per_lb is intentionally null so the new
state-by-state pricing override hierarchy fires. Watch the amber
"state inference (OK: $700/lb)" pills next to every $-figure that uses
wholesale price; if Michael ever supplies a survey number it would flip them
green ("owner-supplied"); the red "national avg — needs owner override" tier
would fire only if the state were missing from state_pricing.yaml.
OK ~$700/lb is ~35% below the $1,081 national default — this is the textbook
"state inference matters" case for AROYA.
Owner-facing translation layer — operator metrics into check-writer dollars. Every $-number is haircut for confidence and flags (assumed) inputs. Cost-down + risk-down lead; yield-uplift is haircut hardest.
Predictability of revenue. Current CV = 60.5%. Halving it recovers $3.9k/cycle haircut 0.4 · med state inference (OK: $700/lb).
| Owner | Rec | Kind | Δ$/cycle | Confidence | Time-to-realize | Why |
|---|---|---|---|---|---|---|
| Michael | Triple P1 shot duration + push feed EC (Cycle 41 stretch) 16868 · Cycle 41 Wk 2–3 stretch |
yield_lift | $19k/cycle state inference (OK: $700/lb) |
haircut 0.2 · low | 1 cycle (Cycle 42) | P1 shot 60s → 180–240s, 4–6 shots in first 2h after P0 ends. P0 delay 2–3h. Feed EC 2.6–2.8 mS. Anchors stretch and builds the EC stack. Phase target: 20–30 absolute pp overnight dryback; current trajectory is well below. |
| Michael | Restore 'Anchoring (Heavy DB)' phase in 13-phase template Recipe template · Cycle 42+ stretch |
yield_lift | $19k/cycle state inference (OK: $700/lb) |
haircut 0.2 · low | 1 cycle (Cycle 42) | Heavy P3 dryback (25–35 absolute pp on coco) at the very start of stretch is what anchors generative steering. Cycle 41's failure here is the textbook case. |
| Michael | Hold Cycle 40 stack, plan ripening pull-back Wk 7 12403 · Cycle 40 Wk 6 Pump/Ripen entry |
risk_down | $3.9k/cycle state inference (OK: $700/lb) |
haircut 0.4 · med | Current cycle (1 cycle) | Substrate EC stack 5.25 mS up from 1.98 — phase target 6.0–8.0 mS through ripen. Drop feed EC to 2.0 mS Wk 7. Deepen P3 dryback into 25+ absolute pp. |
| Brittany | Raise lights-on VPD into 1.0–1.2 kPa (Cycle 41) 16868 · Cycle 41 stretch |
cost_down | $480/cycle | haircut 0.2 · low | Within 5–7 days | Drop RH 4–5%, raise air-temp 1–2°F if HVAC has headroom. Fixes the climate so P0 dryback can establish; without VPD, irrigation tuning can't land. Phase target: P0 build of 1–3 absolute pp before first P1 shot. |
| Michael | Tighten per-phase pore_EC guardrails (recipe config) Recipe guardrails · all flower phases |
cost_down | $480/cycle | haircut 0.4 · med | Immediate (config push) | Stretch target_min 5.0/max 12.0; Bulking 3.5/7.0; Ripening 6.0/8.0; Flush 1.0/4.0. Today everything sits at 3.0/5.0 across flower phases — zero alerts fired in 9 days. Catches EC stack collapse in P2 before it becomes a yield event. |
| Process | CO₂ facility setpoint audit (4 of 8 rooms over-enriching) CO₂ facility setpoint |
cost_down | $320/cycle | haircut 0.4 · med | Immediate (config) | 4 of 8 rooms 1500–1900 ppm (target 1000–1200). Empty turnover 12406 still being enriched. Either one setpoint is wrong or multiple rooms are mis-configured. |
| Process | AUTOMAIT install + recipe 147112 live on 16868 facility-wide |
capex_enable | 9.6 mo payback | haircut 0.4 · med | 1–2 cycles post-install | Manual 13-phase template can't ratchet per-phase like AUTOMAIT can. Cycle 41 lead failure is the textbook case for sensor-triggered P1/P2 stack control. Resolve May 6 internal-consult runoff-threshold blocker first. |
Wholesale $/lb resolved via state inference (tier 2 of 3). Facility state = OK → $700/lb (state_pricing.yaml). Cultivator override (tier 1) would replace this; national default (tier 3, $1081/lb) would overstate Oklahoma realized $/lb by ~35%. Owner survey still required before external publish to upgrade to owner-supplied.
Wholesale $/lb override hierarchy — this report:
6 figures derived from state inference for Oklahoma (OK: $700/lb).
Defaulted inputs in this section (lowest confidence in section: low, 5 defaulted, 14 $-figures generated):
harvests_per_year → defaulted to 5.0 (assumed)kwh_per_lb → defaulted to — (assumed)feed_volume_gal_per_cycle → defaulted to — (assumed)runoff_pct → defaulted to — (assumed)labor_hours_per_cycle → defaulted to — (assumed)Default sources (re-validate annually):
Confidence haircuts applied: high 0.7 · med 0.4 · low 0.2 — baked into every $-figure above.
See aroya-reporting/lib/financial_roi.py framing rules.