Coke-plant wastewater’s ammonia problem meets three hard choices

Coke-oven wastewater routinely arrives “hot”—not by temperature, but by nitrogen. One study reports raw NH₃–N (ammoniacal nitrogen, a measure of ammonia as nitrogen) of ≈500–740 mg/L [mdpi.com] [mdpi.com].

The co‑contaminants are just as punishing: nitrifying bacteria stall if phenol exceeds ~200 mg/L or free cyanide (CN) tops ~0.2 mg/L [pubmed.ncbi.nlm.nih.gov]. Yet regulators often expect ~5–10 mg/L NH₃–N at the end of the pipe (as in Indonesian refinery standards) [id.scribd.com].

That tension forces coke plants toward three options: thermal steam stripping, biological nitrification–denitrification (conversion of NH₄⁺ to NO₂⁻/NO₃⁻ aerobically, then to N₂ anaerobically), or breakpoint chlorination (oxidation of ammonia to nitrogen gas using chlorine until the “breakpoint”). Each can work; none is universal.

Technology metrics and requirements

Steam (thermal) stripping typically removes 85–98% per pass, with bottoms liquor often in the single‑digit mg/L NH₃ range. It requires high pH (usually via NaOH), steam at roughly 50–80 °C or higher, and a packed column; ammonia is distilled as a gas, then recovered as ammonium hydroxide (NH₄OH) or salts. The trade‑off is large energy (steam) use [mdpi.com] [nepis.epa.gov] [nepis.epa.gov].

Biological nitrification–denitrification removes 80–95% (if non‑inhibited), often achieving <50 mg/L NH₃–N in well‑designed systems. It relies on aerobic/anaerobic bioreactors, long solids retention time (SRT, the average time biomass remains in the system), and organic carbon for denitrification. It produces N₂ gas and biomass (sludge), but fails when toxic organics are high (phenol >200 mg/L or free CN >0.2 mg/L) [nepis.epa.gov] [pubmed.ncbi.nlm.nih.gov].

Breakpoint chlorination removes ~90–95% with a high chlorine dose and can approach 0 mg/L NH₃–N. It needs a chlorine source (Cl₂/NaOCl), Ca(OH)₂ to maintain pH ≈6–8, and only minutes of contact time. It converts ammonia to N₂ but consumes about 8–10 g Cl₂ per g NH₃–N, with risks of chlorinated byproducts (including NCl₃ and organochlorines) and high chemical cost [nepis.epa.gov] [nepis.epa.gov].

Steam stripping performance and cost

In practice, steam stripping (often 50–80 °C or above) elevates pH, liberates NH₃ gas, and drives it counter‑current to liquid. Reported removals span ~55% at lower temperature/shorter time to >98% at high pH/longer time: Ferraz et al. achieved ≈98% after 24 h at pH 11 [mdpi.com], while Liu et al. reported 63% in 6 h at 70 °C, pH 8.5, L/G ≈27 L/m³ [mdpi.com].

A coke‑plant pilot cut 487 mg/L NH₃–N to ~7 mg/L in the bottoms—an 86% removal per pass and 96.8% overall removal—using NaOH or lime to lift pH above ~10 and steam to strip, with the vapor condensed for recovery as ammonium salts or NH₄OH [nepis.epa.gov].

Economics hinge on heat. An EPA analysis (ammonia‑plant condensate, 760 L/min at 1000 ppm NH₃–N) estimated operating cost at about $0.0012 per L treated, or $1.49 per metric ton of NH₃ removed (fuel, electricity, labor included). Ion exchange or precipitation alternatives ran ~$3–4 per ton N in the same review [nepis.epa.gov] [nepis.epa.gov]. Steam consumption for ~90% removal of a 1000 ppm stream can be on the order of 5–10 kg steam per kg NH₃ removed; operational data cite ~5,440 kg/h steam at that loading and 760 L/min feed [nepis.epa.gov] [nepis.epa.gov].

Steam cost itself was quoted at ≈$25–30 per 10^5 gallons of condensate [nepis.epa.gov]. Plants often recover heat or route the ammonia‑rich off‑gas to tall furnaces for destruction (furnace stack injection) [nepis.epa.gov]. Where sidestreams once turned to ion exchange, the category remains a niche comparator (ion exchange systems).

Biological nitrification–denitrification limits

Conventional activated sludge and biofilm designs convert NH₄⁺→NO₂⁻/NO₃⁻ aerobically and then NO₃⁻→N₂ under anoxic conditions. In municipal loaders, >90% ammonia removal is routine. In coke wastewater, pilots show ~80–90% when toxicity is managed [nepis.epa.gov].

One multistage coke‑plant setup reported ~90% NH₃ removal with >99.9% phenol removal (EPA, 1973) [nepis.epa.gov]. Oxygen demand is substantial—≈4.6 mg O₂ per mg‑N oxidized [nepis.epa.gov]—and denitrification needs available carbon.

The catch: nitrifiers are highly sensitive. Free CN above 0.2 mg/L or phenol above ~200 mg/L “seriously inhibited” nitrification; thiocyanate and p‑cresol also disrupt it [pubmed.ncbi.nlm.nih.gov]. Raw effluent usually needs pretreatment—phenol removal by distillation or adsorption—before a nitrogen stage. Many plants use granular media such as activated carbon to trim inhibitory compounds ahead of biology.

Despite the sensitivity, biology has advantages: no chemical sludge beyond biomass, modest energy (aeration), and N₂ as the end product. Well‑designed systems can produce effluent on the order of 5–10 mg/L NH₃–N. Lab data show even unacclimated nitrification up to ~350 mg/L NH₃–N, but coke wastewater typically needs acclimation or front‑end removal [pubmed.ncbi.nlm.nih.gov]. Operators often choose classic suspended‑growth lines like activated sludge, while biofilm carriers align with MBBR systems.

Breakpoint chlorination chemistry and risk

Breakpoint chlorination pushes ammonia through a series of oxidative steps to nitrogen gas. In buffered water at 20 mg/L NH₃–N, tests achieved essentially complete conversion to N₂, with residual NO₂⁻ below 2 mg/L [nepis.epa.gov].

At pH 6–8, a Cl₂:NH₃–N weight ratio ≈8:1 worked in pure water, rising toward 10:1 for raw wastes. Pilot work on filtered secondary effluent showed >95% conversion at controlled pH (>6) with low chlorine residuals. In one run, 20 mg/L NH₃–N dropped to near zero, with only trace NO₂⁻/NO₃⁻ (≤2 mg/L) and NCl₃ detected [nepis.epa.gov] [nepis.epa.gov].

But chlorine also attacks organics. EPA reviewers warned against chlorination in coal/coke contexts due to toxic chlorinated phenol formation and other halogenated aromatics [nepis.epa.gov]. The process consumes roughly 8–10 g Cl₂ per g NH₃–N, plus caustic for pH control, and any residual chlorine or chloramines must be quenched (e.g., sulfur dioxide or activated carbon) [nepis.epa.gov]. Where it is used as a polisher, precise chemical metering via a dosing pump and quenching with a dechlorination agent are standard operational controls.

Comparative outcomes and typical hybrids

Steam stripping can remove ~85–95+% NH₃–N in one pass, with a documented case reducing 487 mg/L to ~7 mg/L in the bottoms stream and 96.8% overall removal [nepis.epa.gov]. It needs pH ~11 and heat; waste‑heat integration can make it the lowest‑cost option for pure ammonia streams [nepis.epa.gov] [nepis.epa.gov].

Biological trains excel on moderate loads (<200–300 mg/L) with minimal inhibitors and adequate carbon, often achieving >90% nitrogen removal and effluent around 5–10 mg/L NH₃–N when well designed. If inhibitors spike, nitrification may stall; many plants therefore strip or precipitate high‑strength ammonia up front and biologically polish the remainder. Where effluent reuse quality is desired, some operators turn to membrane‑coupled bioreactors such as MBR units as part of the polishing step.

Breakpoint chlorination can drive ammonia to near zero, but high chlorine dose and byproduct liabilities make it a last‑barrier option, generally after organics are reduced. In most industrial cases, its reagent cost and byproduct oversight are prohibitive [nepis.epa.gov] [nepis.epa.gov].

Decision criteria for coke-plant effluent

• Ammonia concentration: For very high NH₃ (≫200–300 mg/L), thermal stripping is usually most effective; after stripping (or precipitation), residual ammonia is often low enough for biological polishing [mdpi.com]. For lower‑strength (~<100 mg/L), biological nitrification–denitrification is often preferred if detoxification is done.
• Co‑contaminants/inhibitors: If phenols or cyanides are high, prioritize removing those (e.g., phenol distillation, activated carbon) before nitrification. If an inhibitor‑free sidestream with ammonia exists (e.g., blowdown liquor or absorber bleed), it can go directly to stripping. Breakpoint chlorination is generally avoided when organic loads are significant.
• Carbon availability: Denitrification requires an organic or methanol source. Coking wastewater often has low biodegradable BOD after phenol removal. If COD/N is low, external carbon or alternate denitrification (e.g., anammox with nitritation) would be needed for full N removal; where no carbon is available, this favors chemical stripping.
• Effluent NH₃ limits: For moderate limits (≤10–20 mg/L NH₃–N), single‑stage nitrification or a steam‑stripper tail‑water may suffice. Tighter limits (≈5 mg/L NH₃ as in Indonesian refining standards) may require combined approaches. Only breakpoint chlorination or a very high‑performance biofilter can reliably push below ~1–2 mg/L [id.scribd.com].
• Resource and economics: With abundant steam or waste heat, stripping is cost‑effective [nepis.epa.gov] [nepis.epa.gov]. If power is cheap and footprint is available, biology may be cheaper long‑term. Capital often favors biology or chlorination (tanks) over high‑pressure steam equipment, though O&M can invert that.

In practice, hybrids dominate. Many coke plants strip ammonia from “lean water” (COG scrubber bleed) by distillation/stripping to recover ammonium salts, then send the remaining flow to biological treatment; a small chlorine dose at the end (breakpoint or breakpoint+carbon) can ensure compliance when needed [nepis.epa.gov] [nepis.epa.gov]. Biological packages for such duty are cataloged under biological digestion.

Data sources and documentation

Data here are drawn from coke‑oven and industrial wastewater studies and reviews, including EPA reports and peer‑reviewed papers: nepis.epa.gov, mdpi.com, nepis.epa.gov, regulatory context id.scribd.com, and toxicity thresholds pubmed.ncbi.nlm.nih.gov. All figures (removal efficiencies, Cl/N ratios, costs) are cited from those sources: nepis.epa.gov, nepis.epa.gov, nepis.epa.gov, and pubmed.ncbi.nlm.nih.gov.