Refineries Are Detoxing Their Wastewater: Strip the Sour, Then Buffer the Bugs

Petroleum refinery wastewater (“effluent”) is a volatile mix. Studies flag ammonia, hydrogen sulfide (H₂S), phenols, sulfur compounds and trace heavy metals as prime inhibitors of microorganisms in biological plants (pmc.ncbi.nlm.nih.gov) (www.digitalrefining.com). Typical refinery effluent can run at COD ≈2,500 mg/L (COD, chemical oxygen demand, is a measure of oxidizable load), with high sulfide and refractory organics such as polycyclic aromatics and phenols; one dataset cites ~8.2 mg/L SO₄²⁻ in the mix (pmc.ncbi.nlm.nih.gov).

Without effective pretreatment, these loads “can create an overload condition” in the biological plant and even force production cuts (www.pall.com). The operating logic has become clear: remove the most inhibitory contaminants upstream, so activated sludge or biofilm units aren’t shocked by toxic spikes or chronic inhibition.

Physical oil and solids separation

First stop is physical separation to cut oil/grease and suspended solids. API/CPI separators, dissolved air flotation (DAF), and hydrocyclones strip out free and emulsified oil plus grit before anything reaches heat exchangers or microbes (www.pall.com). Pall notes that hydrocarbon carryover into the sour-water stripper can cascade into “overload condition[s]” in the wastewater plant (www.pall.com).

Many facilities deploy primary separation systems at the headworks to intercept debris and oil. Where continuous debris removal helps protect downstream pumps, an automatic screen is often paired with DAF. Free-oil removal skids, such as oil-removal units, and flotation cells like a DAF unit cut tens to hundreds of mg/L of total petroleum hydrocarbons before any biological stage. The practical goal is to prevent oil shunts and solid-bound toxins from coating flocs or introducing inhibitory surface-active compounds.

Sour-water stripping of H₂S and NH₃

One of a refinery’s most toxic internal streams is sour water — condensates rich in NH₃ (ammonia) and H₂S (hydrogen sulfide). A sour-water stripper (SWS, a steam-stripping column with pH control) volatilizes and removes these contaminants, typically >90–95% of free ammonia and H₂S, recovering them as acid gas (www.pall.com). This step prevents sulfide and ammonia shocks to the bioreactor and conditions the water for reuse or discharge. The caveat is upstream oil control: analysts warn the sour-water feed must be de-oiled to avoid fouling the stripper and bleeding oil-rich water downstream — a path to severe effluent violations (www.pall.com) (www.pall.com).

Because SWS performance hinges on precise pH control and stable reagent feed, many plants standardize chemical addition hardware like a dosing pump to meter caustic or acid accurately.

Handling acidic and caustic wastes

Refineries also generate acidic wash waters and spent caustic (from alkylation and naphthenate scrubbing) with extremely high COD and phenolic/mercaptan toxicity. These streams are typically not sent straight to the biotank: the toxic components “must be reduced before the effluent can be subjected to conventional treatment facilities” (www.digitalrefining.com). Spent caustic can contain 1–2% dispersed oil/organics and emulsified naphthenates (www.digitalrefining.com), and many refineries divert it to specialized oxidation or incineration (using CO₂ [34] or peroxide) rather than the main wastewater line. Neutralization and pH correction must be managed carefully to avoid releasing high loads of sulfide gas — pH <3 can liberate H₂S/mercaptans (www.digitalrefining.com).

Targeted polishing and standards

Beyond the front-end, targeted polishing removes residual inhibitors or meets final limits. Metal precipitation is used for inhibitory heavy metals (Cr, Cu, Ni); adsorption or oxidation can tackle stubborn organics. A common polish is granular activated carbon (GAC): passing secondary effluent through a carbon tower removed an additional ~36% of residual COD and ~34% of ammonia in one case study, with effluent meeting discharge/reuse standards (www.sinoucarbon.com) (www.sinoucarbon.com). Plants often source media through activated carbon systems and combine polishing with clarification. Where settling capacity is a constraint, a clarifier upstream of carbon can stabilize solids loading; coagulant/flocculant programs (coagulants, flocculants) are common adjuncts.

Regulatory targets drive this rigor. Indonesia’s environmental rules (e.g., Peraturan Pemerintah No. 82/2001 and sector-specific decrees) set BOD₅ (biochemical oxygen demand over 5 days), COD and phenol limits for refinery wastewater on the order of 10^1–10^2 mg/L. Globally, best practices combine physical/chemical pretreatment with high-rate biological processes to deliver overall COD removals >90%, BOD removals similarly high, and effluent phenols typically <0.5–1 mg/L.

A benchmark multistage biochemical pilot removed >99% of sulfide: an initial stage dropped sulfide from 8.2 to 1.9 mg/L (≈77%), with later steps essentially eliminating it; influent COD ~2,554 mg/L fell to ~110–190 mg/L after full treatment (pmc.ncbi.nlm.nih.gov). These figures underscore how early-stage removal unlocks downstream biology.

Powdered activated carbon in bioreactors

Even with strong pretreatment, trace toxins persist. Dosing powdered activated carbon (PAC; very fine activated carbon particles) into the bioreactor has emerged as a pragmatic stabilizer: PAC adsorbs hydrophobic or poorly-biodegradable organics, shielding microbes and smoothing shock loads (www.sinoucarbon.com) (nepis.epa.gov). In practice, PAC is added to the aeration basin (the mixed liquor) of an activated sludge system, continuously or in pulses during upsets. The result is a hybrid adsorption–biodegradation system that protects biomass and incrementally boosts removal.

Mechanistically, the gains come from adsorption and biological synergy: PAC offers high surface area for toxins, while also serving as micro-carriers that improve floc formation and activity. This often translates to better sludge settleability (measured by SVI, sludge volume index) and higher apparent biomass (MLSS, mixed liquor suspended solids), with aeration system oxygen (DO, dissolved oxygen) tracking more stable (www.sinoucarbon.com) (nepis.epa.gov).

Measured effects in full-scale and bench trials

One refinery case saw influent COD spike to 2,162 mg/L — far above its 1,200 mg/L design — with secondary effluent COD initially at 79.4 mg/L (49% removal). After a single PAC addition, the next day’s effluent COD fell to 44.9 mg/L (69% removal), a roughly 43% further drop in COD. Simultaneously, DO rose from 1.89 to 2.13 mg/L, MLSS jumped from 1,377 to 3,958 mg/L, and SVI halved from 232 mL/g to 111 mL/g (www.sinoucarbon.com). EPA-sponsored research likewise found that PAC addition “did result in increased sludge compactability and sludge settling rates” (nepis.epa.gov).

Controlled studies document selective toxin removal: dosing 100 mg/L PAC produced >90% removal of recalcitrant aromatics — including 1,2-dichlorobenzene, trichlorobenzene, and lindane — largely unaffected by ordinary biomass (nepis.epa.gov). In full-scale refinery trials, PAC reduced effluent COD by 18% from control, corresponding to a ~2% improvement in overall COD removal efficiency; oil and grease improved by 0.5–2.0 mg/L (nepis.epa.gov) (nepis.epa.gov). By contrast, PAC had little effect on phenol, phosphorus or total organic carbon in those tests (nepis.epa.gov).

The headline takeaway: plants using PAC have recorded concrete, quantifiable gains — for some pollutants, 50–80% better effluent — alongside more stable operation (less sludge bulking, fewer upsets) and higher safety margins for meeting permits. In one example, PAC converted a marginal effluent (COD = 79 mg/L) into a very compliant one (COD = 45 mg/L) (www.sinoucarbon.com).

Dosage ranges and handling economics

Literature notes PAC doses from the tens to low hundreds of mg/L — for example, 12–100 mg/L on an influent basis (nepis.epa.gov). Some plants add PAC to the secondary aeration tank roughly “once every half month,” triggered by high loads (www.sinoucarbon.com). Spent PAC accumulates in pressurized sludge dewatering cake and is then incinerated; operators note that if carbon could be regenerated and recycled, “operating cost can be greatly reduced” (www.sinoucarbon.com).

There are trade-offs: unused PAC must be separated with the sludge, and slight increases in effluent suspended solids can occur if PAC breaks through (nepis.epa.gov). Robust clarification capacity — including a well-sized secondary clarifier — helps capture carbon fines.

Operational targets and integration

The integrated playbook is now standard: upstream pretreatment (SWS for NH₃/H₂S, oil/water separation, isolation of caustic/acid wastes) to remove the worst toxins, followed by an enhanced biological stage that may include PAC. Effective SWS typically strips ~90% of free NH₃/H₂S, greatly alleviating ammonia toxicity and sulfide inhibition for the bugs. After pretreatment and PAC, operators verify that effluent COD/BOD are consistently below target (e.g., <100–200 mg/L) and that phenols, sulfides, and ammonia stay low. An advanced front-end — analogous to sour-water stripping — enabled one pilot to cut sulfide from 8.2 to 1.9 mg/L (≈77%) and drive COD to ~110–190 mg/L after full treatment (pmc.ncbi.nlm.nih.gov).

When these blocks are in place, biological systems deliver >90% COD removal routinely, with stable sludge behavior and fewer compliance scares. For context and sourcing: all numeric results and process impacts are backed by the cited references, including EPA research (nepis.epa.gov) (nepis.epa.gov), recent case studies (www.sinoucarbon.com) (pmc.ncbi.nlm.nih.gov), and technical reviews (pmc.ncbi.nlm.nih.gov) (www.digitalrefining.com). All numeric results and process impacts are backed by cited references (see metadata below).