NPDES / Permit Compliance Reporting Bot
AI 90%Live in demoDischarge monitoring reports are assembled by hand in most utilities: lab LIMS exports, SCADA averages and operator logs stitched into a spreadsheet under deadline. The Compliance Bot assembles the DMR continuously, shows the running permit margin for every parameter, and predicts exceedance risk days before the compliance sample is pulled.
Coastal Bend Regional Water Authority · Gulf Coast Metro Service Area · TX
Every permitted parameter, continuously calculated
Rolling averages are computed as the permit defines them — not reconstructed at month end.
| Parameter | Permit limit | Current | Headroom | State |
|---|---|---|---|---|
| NH3-N (30-day avg) | 1.8 mg/L | 1.24 mg/L | 31% | Watch |
| BOD5 (30-day avg) | 25 mg/L | 8.6 mg/L | 66% | In range |
| TSS (30-day avg) | 30 mg/L | 9.1 mg/L | 70% | In range |
| Total P (monthly) | 1.0 mg/L | 0.61 mg/L | 39% | Watch |
| E. coli (geo-mean) | 126 MPN | 18 MPN | 86% | In range |
| Cl2 residual (min) | 0.2 mg/L | 0.94 mg/L | 78% | In range |
| Turbidity, WTP-1 (95th) | 0.30 NTU | 0.21 NTU | 30% | Watch |
| pH range | 6.0 – 9.0 | 7.3 | 82% | In range |
NH3-N 30-day average trajectory
Exceedance risk
NH3-N trending toward 30-day limit
AI 89%Aeration efficiency loss on Blower #2 has reduced nitrification margin. Forecast 30-day average reaches 1.44 mg/L against a 1.8 mg/L limit — 80% of limit — if the derate persists through the reporting period.
Risk window: next 11 days
WTP-1 filter turbidity 95th percentile tightening
AI 84%Filter 3 run length has shortened 14%; 95th-percentile turbidity has moved from 0.16 to 0.21 NTU against the 0.30 NTU action level.
Risk window: next 30 days
Draft DMR with per-value provenance
Every number traces back to the sample or SCADA tag that produced it. A licensed operator signs the final report — the bot never submits.
| Report field | Value | Source | State |
|---|---|---|---|
| Flow, monthly average | 12.41 MGD | SCADA totalizer FIT-101 | Verified |
| BOD5, 30-day average | 8.6 mg/L | LIMS · 12 of 12 samples | Verified |
| TSS, 30-day average | 9.1 mg/L | LIMS · 12 of 12 samples | Verified |
| NH3-N, 30-day average | 1.24 mg/L | LIMS · 11 of 12 samples | 1 pending |
| Total P, monthly | 0.61 mg/L | LIMS · 4 of 4 samples | Verified |
| E. coli, geometric mean | 18 MPN/100mL | LIMS · 20 of 20 samples | Verified |
| Percent removal, BOD | 96.4% | Calculated (influent/effluent) | Verified |
| Bypass / SSO events | 0 | Operator log + level telemetry | Verified |
Guardrails
- Never submits on the utility's behalf — a licensed operator signs every report.
- Values flagged as estimated are labeled and excluded from the certified draft.
- Full immutable change log for every edited value.
Data in · results out
- LIMS lab results (BOD, TSS, NH3-N, TP, E. coli)
- Continuous analyzers: turbidity, chlorine residual, DO, pH
- Effluent and bypass flow totalizers
- Operator log entries and bypass / SSO events
- Permit limit set and sampling calendar
- Live permit margin per parameter with rolling averages
- Pre-sample exceedance risk warnings
- Draft DMR package with audit trail
- Sampling calendar compliance tracker
- Exceedance root-cause narrative for the regulator file
DeploymentRequires LIMS export access (CSV, SQL or API) plus effluent analyzer tags. Typical stand-up 3–4 weeks including permit encoding review.
Other active bots
Energy Optimization Bot
Tracks blower and pump energy against real-time flow demand and prices out savings from optimized run schedules, DO trim, and blower sequencing.
Typical: 8–14% reduction in kWh/MG
Chemical Dosing Optimizer Bot
Optimizes chlorine, polymer and alum dosing against real-time flow and quality targets rather than fixed setpoints, holding compliance margin at the lowest feasible dose.
Typical: 6–11% chemical spend reduction
Demo environment · values shown are realistic estimates, not audited financials