Why Petrochemical Wastewater Defies Conventional COD Removal
Petrochemical wastewater is biologically hostile by design. Characterization of a real high-salinity petrochemical stream in the 2018 Jorfi et al. study (Chem. Biochem. Eng. Q., 2018) found COD of 370–2,900 mg/L (average 1,271), BOD5 of 50–128 mg/L (average 89), BOD5/COD of only 0.05–0.09, TDS of 15,000–45,400 mg/L, TSS of 130–240 mg/L, and pH 7.8–8.9.
That BOD5/COD ratio sits well below 0.1, the range in which standard activated sludge is engineered to operate, and the chloride-bearing TDS pushes the osmotic stress on biomass into the failure zone documented for petroleum refining. A 2017 review of full-scale biological plants (inTechOpen, 2017) explicitly warns that activated sludge can fail on petrochemical wastewater with COD above 10 g/L and on aromatic streams containing phenol and its derivatives, and that oil, fat, and grease can float sludge out of the basin. The same review notes that load shock from multiple refining units, plus high VOC emissions from open aeration tanks, makes a single fixed biological recipe unstable across a normal operating year. Any 2026 design that begins with "just add an aeration basin" should be rejected before jar testing starts.
Unit Processes Compared: What Actually Removes COD in 2026
The 2017 inTechOpen review of full-scale biological plants is the most useful single reference for a head-to-head view because it documents the operating envelope, not just a headline removal percentage. Aerobic refinery systems are reported at 70–98% COD removal, anaerobic systems at 70–93%, and contact/extended activated sludge at 89–95% at a feed-to-microorganism ratio of 0.38. MBBR biofilm reactors delivered 74% COD removal at an organic loading rate of 4.2 kg COD/m³·d, while BAF, MSBR, and HF-UF MBR systems exceeded 80% COD removal at higher loadings. The same review documents HyVAB biofilm reactors running at full scale (240 m³/d, feed COD 7–35 g/L) at 94% soluble and 85% total COD removal with air aeration, and holding over 90% soluble COD removal at organic loading rates approaching 30 kg COD/m³·d and 15 h HRT once oxygen aeration was introduced. Coagulation data point the same direction: ferric chloride achieved 75.5% COD removal at pH 5.6 on a 2,776 mg/L COD stream, while PAC-PAM combinations typically remove 30–70% of COD depending on wastewater composition. The 2024 HydroChemix commercial guide reports that AOPs achieve 80–95% COD removal on refractory streams, and the inTechOpen review records that ozonation at 100–200 mg O3/h raised BOD/COD from 20% to 35% in 30 min on ABS petrochemical wastewater. On the high-salinity extreme, Jorfi et al. (Chem. Biochem. Eng. Q., 2018) reported 89.4% COD removal on real petrochem feed (TDS ~25,000 mg/L) by photo-assisted peroxi-coagulation at pH 3, 1.5 V, 2 cm electrode gap, 420 min reaction time, with a pseudo-first-order rate constant of 0.0048 min⁻¹ and a specific energy consumption of 1.73 kWh/kg COD removed.
| Unit process | Reported COD removal | Operating envelope in petrochem service | Primary source |
|---|---|---|---|
| Aerobic activated sludge (refinery) | 70–98% | Fails above ~10 g/L COD; sensitive to oil/grease and shock load | inTechOpen review (2017) |
| Anaerobic (UASB, etc.) | 70–93% | Light petroleum extraction most degradable; heavy oil 20–60% | inTechOpen review (2017) |
| Contact/extended activated sludge | 89–95% | F/M ratio 0.38 | inTechOpen review (2017) |
| MBBR biofilm | 74% | 4.2 kg COD/m³·d OLR | inTechOpen review (2017) |
| MBR (BAF, MSBR, HF-UF MBR) | >80% | Compact footprint, handles higher OLR | inTechOpen review (2017) |
| HyVAB biofilm (full scale) | 94% soluble, 85% total COD | Up to ~30 kg COD/m³·d, 15 h HRT; 82% methane in biogas | inTechOpen review (2017) |
| Ferric chloride coagulation | 75.5% | pH 5.6, 2,776 mg/L COD feed | inTechOpen review (2017) |
| PAC-PAM coagulation | 30–70% | Dose-dependent; pretreatment or tertiary polish | HydroChemix guide (2024) |
| Ozonation (AOP) | BOD/COD raised 20%→35% | 100–200 mg O3/h, 30 min on ABS wastewater | inTechOpen review (2017) |
| Advanced oxidation (AOPs, general) | 80–95% | Refractory and toxic organics | HydroChemix guide (2024) |
| Photo-assisted peroxi-coagulation | 89.4% | pH 3, 1.5 V, 2 cm gap, 420 min, TDS ~25,000 mg/L, 1.73 kWh/kg COD | Jorfi et al. (2018) |
Building a 2026 Process Train for Petrochemical COD
No single unit process clears petrochem influent to a typical 100 mg/L discharge permit on its own. The train that consistently does so in the published record stacks pretreatment, biological reduction, and polishing in that order, and the 2024 HydroChemix refinery case is the cleanest published example of the pattern. Stage 1 is oil, grease, and suspended solids removal: a DAF or API separator ahead of biology and membranes, sized to the 130–240 mg/L TSS band measured by Jorfi et al. (2018) and to the floatable-oil risk flagged in the 2017 inTechOpen review. A purpose-built petrochemical DAF for oil and grease removal typically slots in here. Stage 2 is equalization and nutrient balancing to dampen the load shock the inTechOpen review identifies as a recurrent cause of biomass failure. Stage 3 is coagulation plus biological reduction: in the HydroChemix refinery case (2,500 m³/d, 4,500 mg/L COD, 150 mg/L phenol, 100 mg/L permit) the PAC-PAM stage removed 45% of COD and 60% of phenols, the biological stage removed a further 55% of COD, and the activated carbon polish brought the final effluent to 85 mg/L, a 75% overall removal. An automated coagulant and pH dosing skid is the practical way to hold that 45% coagulation step stable across influent swings. Stage 4 is polishing: an MBR for petrochemical COD polishing gives the >80% COD removal documented in the 2017 review while shrinking footprint relative to conventional activated sludge, and a downstream AOP such as the Jorfi et al. (2018) peroxi-coagulation configuration is a credible slot for residual refractory load once the stream is biologically stabilized. The point of the train is that each stage is sized to a specific failure mode the previous one cannot handle, which is what the raw influent data in the first section demands.
Sizing, Energy, and Cost Reality Check
Budget conversations in 2026 still run on three numbers: energy per kilogram of COD removed, chemical cost per cubic meter, and footprint per kilogram of daily load. The only published petrochem-specific energy benchmark in the supplied research is the 1.73 kWh/kg COD removed reported by Jorfi et al. (2018) at the 89.4% operating point on a TDS ~25,000 mg/L feed; any AOP being proposed for the polishing slot should be sanity-checked against that figure or against the buyer's own pilot data. On chemicals, the HydroChemix guide (2024) gives a worked example: 1,000 m³/d at 3,000 mg/L COD, PAC at 600 mg/L ($0.35/kg) plus PAM at 4 mg/L ($2.80/kg), yields a daily chemical spend of about $222. Because coagulation-only removal sits in the 30–70% band the same guide reports, a biological stage is mandatory to reach refinery-level permits, and the chemical line item must be sized against the residual load after biology, not the raw influent. On footprint, the inTechOpen review (2017) shows HyVAB biofilm reactors sustaining over 90% soluble COD removal at organic loading rates approaching 30 kg COD/m³·d and 15 h HRT, which is the order of magnitude that allows a much smaller tankage envelope than conventional activated sludge for the same daily COD mass. The permit ceiling is consistent across the literature: Jorfi et al. (2018) cite the Iran environmental discharge limit of 100 mg/L COD, and the HydroChemix refinery case study targets the same 100 mg/L, with a measured 85 mg/L effluent. Designs must clear both that COD ceiling and any local TDS, chloride, or ammonia cap before the polishing stage is selected.
| Metric | Published value | Source |
|---|---|---|
| Energy per kg COD removed (PPC AOP) | 1.73 kWh/kg COD | Jorfi et al. (2018) |
| Chemical spend, 1,000 m³/d at 3,000 mg/L COD | ~$222/d (PAC 600 mg/L + PAM 4 mg/L) | HydroChemix guide (2024) |
| HyVAB biofilm OLR ceiling | ~30 kg COD/m³·d at 15 h HRT | inTechOpen review (2017) |
| Refinery case overall COD removal (coag + bio + carbon) | 75% (4,500 → 85 mg/L) | HydroChemix guide (2024) |
| Permit ceiling cited | 100 mg/L COD | Jorfi et al. (2018); HydroChemix guide (2024) |
2026 Specification Checklist for Procurement
The fastest way to turn the article above into a working bid document is to anchor each line item to a number or a parameter the research actually documents. Influent characterization should mirror Jorfi et al. (2018): demand daily composite data for COD, BOD5, BOD5/COD, TDS, TSS, pH, oil and grease, phenols, and temperature before any vendor selection. The target effluent must be stated numerically — a 100 mg/L COD ceiling is consistent with both the Jorfi et al. (2018) reference and the HydroChemix (2024) refinery case — alongside any local TDS, chloride, and ammonia caps. The process train scope in the bid should explicitly require oil and grease removal (DAF or equivalent), a named biological reactor type with documented salinity and phenol tolerance, a chemical dosing skid, and a polishing step sized against the 80–95% AOP removal band or the >80% MBR band reported in the 2017 inTechOpen review. Ask each vendor for operating data: specific energy in kWh/kg COD removed (the Jorfi et al. 1.73 kWh/kg figure is the only petrochem-specific benchmark in the supplied set), chemical consumption, sludge yield, and tolerance to influent TDS and phenol shocks aligned with the failure modes listed in the 2017 review. Finally, require compliance documentation that maps the offered train to local discharge regulations and the plant's pretreatment permit before any purchase order is released.
Frequently Asked Questions
Should we use AOP or biology as the primary COD step for a high-salinity petrochemical stream?
Use biology as the bulk reducer and AOP as the polish. The 2017 inTechOpen review reports that full-scale biological plants (HyVAB biofilm at 94% soluble COD removal, aerobic at 70–98%, MBR at >80%) handle the bulk of the load once toxicity and shock are managed, while AOPs at 80–95% removal per the 2024 HydroChemix guide are sized for the residual refractory fraction. Jorfi et al. (2018) hit 89.4% COD removal with photo-assisted peroxi-coagulation on TDS ~25,000 mg/L feed at 1.73 kWh/kg COD removed, which is realistic as a polishing step, not as a standalone on raw influent.
What is a realistic 2026 budget for chemicals on a 1,000 m³/d petrochemical wastewater train?
The only published petrochem-relevant worked example in the supplied research is the 2024 HydroChemix guide: 1,000 m³/d at 3,000 mg/L COD using PAC at 600 mg/L ($0.35/kg) and PAM at 4 mg/L ($2.80/kg) gives roughly $222 per day in coagulant and flocculant. Because PAC-PAM alone typically removes 30–70% of COD, a biological stage is required to reach the 100 mg/L permit cited in both the Jorfi et al. (2018) and HydroChemix (2024) cases, so the chemical line item should be sized against the residual load after biology. Request vendor-specific dose curves and current unit prices for your local influent before locking the operating budget.
How do I pick a biological reactor that will not fail on petrochemical influent?
Match the reactor to the failure modes the 2017 inTechOpen review documents: activated sludge fails above ~10 g/L COD and on phenol-type aromatics, while oil and grease cause sludge washout. The same review reports HyVAB biofilm reactors sustaining over 90% soluble COD removal at organic loading rates approaching 30 kg COD/m³·d and 15 h HRT, and MBR variants (BAF, MSBR, HF-UF MBR) clearing >80% COD at higher OLRs in a smaller footprint. Ask each bidder for documented tolerance to your specific TDS, phenol, and oil-and-grease envelope, not a generic removal curve.
What should we require from vendors before issuing a purchase order?
Require influent and effluent characterization data on the parameters Jorfi et al. (2018) measured (COD, BOD5, BOD5/COD, TDS, TSS, pH, oil and grease, phenols), a specific energy figure in kWh/kg COD removed (the Jorfi et al. 2018 benchmark is 1.73 kWh/kg for the AOP slot), chemical consumption with unit prices tied to your flow, sludge yield, and a compliance map from the offered train to the local discharge regulation and pretreatment permit. Treat any vendor that cannot supply all four as a non-compliant bidder, regardless of headline removal percentage.
