Confirm the discharge route before selecting equipment
A petroleum facility sending process wastewater to a POTW must meet the applicable general and specific prohibitions and any categorical, state and local requirements. The 25,000-gpd threshold concerns one SIU criterion; it does not define every industrial user. Identify the actual receiving works and control authority, and obtain the written requirements for the site. Under 403.8, control may be through permit, order or similar means, with additional individual or eligible general control mechanisms for SIUs. Do not infer a universal permit, monitoring timetable or enforcement sequence from the industry label.
Regulatory Chain: CWA, 40 CFR Part 403, MAHL, and the MWRD Permit
The citation chain a Commerce City engineer can hand to a regulator runs Clean Water Act §307(b) (33 U.S.C. § 1251 et seq.) → EPA General Pretreatment Regulations at 40 CFR Part 403 → EPA's Local Limits Development Guidance (Chapter 2, MAHL Decision Tree) → MWRD-adopted Technically Based Local Limits (TBLL) → individual Industrial User permit. The MAHL decision tree walks five steps: Step 1 identifies pollutants of concern (POCs), Step 2 collects headworks data, Step 3 calculates Maximum Allowable Headworks Loadings (MAHLs) for each POC, Step 4 designates and implements local limits, and Step 5 addresses collection-system concerns such as corrosion, flammability, and worker safety (per EPA Local Limits Development Guidance, 2004/2021 update). Four inputs drive every MAHL: NPDES effluent limits on the receiving POTW, Colorado water-quality standards for the South Platte, Part 503 biosolids numerical criteria, and NIOSH/ecology protection factors (per the 2020 St. Joseph, MO TBLL evaluation by Black & Veatch).
The Industrial User (IU) trigger is the first gate: a facility that discharges process wastewater to a POTW, or that contributes ≥25,000 gpd of non-domestic waste, qualifies. Most petroleum bulk plants are noncategorical SIUs; dedicated refinery and fuel-blending operations may be categorical under 40 CFR Part 419 (per St. Joseph, 2020 TBLL). The distinction matters for permit stringency but not for the daily reporting burden. What does matter is the difference between MAHL and MAIL: MAHL is the headworks ceiling the POTW can absorb without violating any of its downstream permits; MAIL is the per-IU allocation the POTW prints on a permit, derived by dividing MAHL by all contributing industrial flow plus uncontrolled domestic load. For a 2026 MWRD permit, the daily-max HEM number on the page is MAIL, not MAHL — and the engineer should know which ceiling the MWRD used to back-calculate it. Colorado Discharge Permit System (CDPS) pretreatment coordination overlays federal 40 CFR Part 403 with state authority, which is why MWRD's TBLL adoption goes through both EPA Region 8 review and CDPHE sign-off.
Build the monitoring and response plan from written requirements
Obtain the actual control mechanism, pollutant list, sampling location, methods, frequencies, report dates and notification duties. Under 403.12(g)(3), grab samples are required for oil and grease and volatile organics, among the listed parameters. Permitted laboratory compositing under appropriate preservation and approved protocols is different from prescribing a field flow-proportional BTEX or HEM composite. Have the laboratory and authority confirm the sampling plan. Retain actual results and chain of custody; do not estimate missing laboratory results as if measured. The 24-hour violation notice and repeat-analysis provisions in 403.12(g)(2) have their own conditions and exceptions; do not confuse them with the immediate potential-problem notification in 403.12(f).
The Four-Stage Pretreatment Train — and Why the Order Is Non-Negotiable
A petroleum terminal pretreatment train has four stages, and the sequence is dictated by droplet-size distribution logic, not vendor preference. Stage 1 is source segregation: keeping hydrocarbon-contaminated streams out of clean stormwater shrinks the volume hitting the treatment train by 40–70% in field retrofits and converts most of the remaining flow from "design problem" to "design choice" (HydropureWater field data, 2026). The hardware is segregated laterals on product-handling pads, covered and locked dump valves on coalescers, drip pans under loading arms, and dedicated oil/water sewering on truck loading islands.
Stage 2 is primary oil/water separation — either an API gravity separator sized for ≥30 minutes residence at peak flow, or a Corrugated Plate Interceptor (CPI) with 1–2 in plate spacing and corrugation near 45°. This stage handles free oil at ≥60–150 µm droplet size, the band that accounts for most of the mass loading from tank-bottom water and coalescer dumps. Stage 3 is emulsified-oil polishing with a Dissolved Air Flotation (DAF) system, where micro-bubbles generated at 60–90 psig saturation pressure float oil droplets down to 10–25 µm. Surface hydraulic loading in oilfield service runs 2–5 gpm/ft², and air-to-solids ratio (ASR) carries a 20–30% safety margin to absorb slug loads. Stage 4 is biological or adsorption polishing — Moving Bed Biofilm Reactor (MBBR) or activated sludge for ammonia and sulfide, granular activated carbon where the train has to swing to <20 mg/L for a water-reuse loop.
Why a single-technology approach fails: tank-bottom water and coalescer dumps carry free oil at ≥150 µm, which blankets DAF micro-bubbles and crashes ASR; wash-rack water is emulsified by surfactants with droplet sizes below 50 µm, which a CPI alone cannot break. The robust path is CPI or API primary, then DAF polish, sized with margin on hydraulic and ASR. A Dissolved Air Flotation (DAF) system is the reference unit for the Stage 3 polish on Commerce City tank-bottom and wash-rack streams.
Choosing the Primary Separator: API, CPI, Coalescer, and DAF Compared
The four primary technologies sit in different performance bands and are not interchangeable. The table below maps droplet size, surface loading, best-fit scenario, and the key limitation for each:
| Technology | Target droplet size | Surface loading / spec | Best-fit scenario | Key limitation |
|---|---|---|---|---|
| API gravity separator | ≥150 µm (free oil) | ≥30 min residence at peak flow | High-throughput marine/terminal, large flow swings | Cannot break emulsions; large footprint |
| CPI (corrugated plate interceptor) | 60–150 µm (free oil) | 1–2 in plate spacing, ~45° corrugation | Small-to-mid terminal with steady flow; retrofit into existing concrete vault | Rarely meets <100 mg/L HEM on emulsified waste; plate fouling risk |
| Coalescer (plate or multimedia) | 10–25 µm (emulsified/colloidal) | 5–10 gpm/ft² (vendor-specific) | Polishing stage or low-flow sites with strict <50 mg/L needs | Higher O&M; media replacement 1–3 yr |
| DAF as primary | 10–25 µm (emulsified) | 2–5 gpm/ft² surface; ASR 20–30% margin | Truck-loading rack with emulsified oils; as primary only where free oil is pre-strained | Slug-sensitive without upstream primary; needs air saturation system |
Field guidance: a DAF alone without a primary gravity stage fails under slug loads from coalescer dumps, because free oil blankets the bubble surface and crashes ASR. A CPI alone rarely meets a 100 mg/L HEM limit on emulsified wash-rack water — it removes free oil efficiently but does not address sub-60 µm droplets (HydropureWater field data, 2026). The robust path for a U.S. petroleum bulk plant is CPI or API as primary, then DAF as the emulsified-oil polisher, sized with a 20–30% safety margin on hydraulic and air-to-solids loading. Engineers weighing DAF against conventional clarification in a different service can review the DAF vs clarifier decision guide for additional context on the technology trade.
Commerce City–Specific Design Wrinkles: Altitude, Freeze, and Watershed Reuse
Two local factors derate any DAF or outdoor separator design along the South Platte corridor, and neither appears in a generic U.S. explainer. The first is altitude: Commerce City sits at approximately 5,280 ft elevation, where atmospheric pressure is roughly 83% of sea-level standard. Henry's Law governs DAF air saturation, and lower partial pressure means less dissolved air per unit of saturator volume at the same pressure and temperature. The field consequence is that a DAF sized at sea level will deliver proportionally less air-to-solids ratio at altitude, eroding the 20–30% ASR safety margin. The mitigation is to oversize the saturation tank, raise saturation pressure within pump and vessel ratings, or add a booster compressor on the recycle line to preserve the design ASR envelope.
The second factor is freeze protection. Outdoor API and CPI units along the South Platte need to be designed for design ambient around -20°F, with buried or insulated vaults, heat-traced dump valves, and enclosed skimmer housings. A frozen dump valve fails closed during a cold snap, then releases as a slug load on the first thaw — that slug hits the DAF with free oil at ≥150 µm and crashes ASR in a single event. The standard mitigation is heat trace on all dump-valve bodies, insulation on exposed oil/water piping, and a sample-port design that does not freeze closed at -20°F. The physics and the failure mode are the same as for any outdoor wastewater train in a continental-interior climate, but the design air temperature is colder than most U.S. explainers assume (HydropureWater field data, 2026).
A third local factor is watershed reuse. The South Platte basin is on the industrial water-scarcity watchlist, and stricter 50 mg/L HEM ceilings are plausible by 2027–2028 (per the ACS ES&T Engineering 2021 review on U.S. industrial water scarcity and reuse). Stormwater segregation also matters more in Colorado than in coastal climates: spring snowmelt combined with summer convective storms can produce 3–5× design hydraulic peaks in a single event, and the slug load on a primary separator during peak flow is the most common cause of HEM carryover in field retrofits. A staged equalization basin ahead of the train is the standard mitigation.
Three Numbers That Drive a Defensible Design
Three numbers — plus one Commerce City–specific fourth — must be locked in before talking to a vendor. First, peak instantaneous flow in gpm or m³/h, not the daily average. A coalescer dump or tank drop can spike 3–5× the daily mean, and the primary separator must hold its Reynolds-number cap and Froude-number floor at that peak (HydropureWater field data, 2026). Second, daily O&G load in lb/day or kg/day, calculated from tank turnover, wash-rack volume, and loading-arm drip rates. Third, target residual O&G in mg/L, taken from the MWRD permit ceiling and ideally set 20–30% below it as a safety margin so a single upset does not push the terminal into NOV territory.
The fourth number, specific to Commerce City and the South Platte corridor, is design air temperature, used to size heat tracing, vault insulation, and enclosure heating on outdoor API/CPI/DAF equipment. Standard ASHRAE design data for the Denver Metro area places the 99% heating design temperature near -10°F, but operators along the South Platte routinely spec to -20°F to absorb inversions and cold-pool events. The cost difference between -10°F and -20°F design ambient is small in heat-trace loading, but the operational difference during a polar vortex is the difference between a train that runs and a train that slug-loads the DAF on the first thaw. For chemical conditioning ahead of the DAF, an automatic chemical dosing system sized to the design peak flow keeps coagulant and flocculant ratios stable through the slug.
Significant noncompliance: use the actual federal criteria
40 CFR 403.8(f)(2)(viii) defines significant noncompliance (SNC) for the public-notification provision. The chronic and TRC criteria apply to SIUs; the stated interference, imminent-endangerment and programme-impact criteria can also reach other industrial users. A single exceedance can still be a violation even when a statistical SNC threshold is not reached.
| Criterion | Federal test |
|---|---|
| Chronic violations | 66% or more of measurements for the same parameter in a six-month period exceed a numeric pretreatment standard or requirement. |
| TRC violations | 33% or more of measurements for the same parameter in six months equal or exceed the applicable limit multiplied by 1.4 for BOD, TSS and fats/oil/grease, or 1.2 for other pollutants except pH. |
| Other criteria | Specified interference/pass-through, imminent endangerment, a compliance-schedule milestone still unmet 90 days after its date, required reports not provided within 45 days after the due date, inaccurate noncompliance reporting, or other programme-adversely-affecting violations determined by the POTW. |
The 45-day criterion is not permission to file late. Reporting deadlines and earlier notification duties remain enforceable. Federal SNC is not a universal sequence of one late report, two NOVs and a hearing, and filing on the fifteenth does not guarantee protection. Obtain the POTW’s enforcement response plan and any actual order; corrective deadlines and remedies come from the applicable requirements, not a generic 30/60-day cure window.
Build the monitoring and response plan from written requirements
Obtain the actual control mechanism, pollutant list, sampling location, methods, frequencies, report dates and notification duties. Under 403.12(g)(3), grab samples are required for oil and grease and volatile organics, among the listed parameters. Permitted laboratory compositing under appropriate preservation and approved protocols is different from prescribing a field flow-proportional BTEX or HEM composite. Have the laboratory and authority confirm the sampling plan. Retain actual results and chain of custody; do not estimate missing laboratory results as if measured. The 24-hour violation notice and repeat-analysis provisions in 403.12(g)(2) have their own conditions and exceptions; do not confuse them with the immediate potential-problem notification in 403.12(f).
Build the monitoring and response plan from written requirements
Obtain the actual control mechanism, pollutant list, sampling location, methods, frequencies, report dates and notification duties. Under 403.12(g)(3), grab samples are required for oil and grease and volatile organics, among the listed parameters. Permitted laboratory compositing under appropriate preservation and approved protocols is different from prescribing a field flow-proportional BTEX or HEM composite. Have the laboratory and authority confirm the sampling plan. Retain actual results and chain of custody; do not estimate missing laboratory results as if measured. The 24-hour violation notice and repeat-analysis provisions in 403.12(g)(2) have their own conditions and exceptions; do not confuse them with the immediate potential-problem notification in 403.12(f).
Frequently asked questions
Does one late report automatically mean SNC?
A late report can violate its deadline immediately, but federal SNC is determined under 403.8(f)(2)(viii), not by a universal one-report/two-NOV sequence. The required-report criterion is failure to provide it within 45 days after its due date; other SNC criteria and earlier notification duties remain independent.
Should HEM and BTEX be collected as field composites?
Do not prescribe that generically. The pretreatment rule specifies grab samples for oil/grease and volatile organics, with defined laboratory-compositing provisions. Confirm the actual permit, method and preservation with the authority and laboratory.
What should an equipment acceptance specification contain?
Provide the actual limits, sample basis and measured feed range. Agree residual-pollutant tests, throughput, chemical and energy use, sludge handling, alarms and safe containment of off-spec water. Keep compliance responsibilities and response deadlines explicit.