The brutal math of cokemaking wastewater — and the plant designs that beat it

Raw coke‑oven wastewater is among industry’s most concentrated. Lin et al. (2018) report influent COD (chemical oxygen demand, a bulk measure of organics) ≈5,200 mg/L and NH₃–N (ammonia‑nitrogen) ≈650 mg/L, with phenol ≈945 mg/L and thiocyanate ≈512 mg/L (www.mdpi.com). Compliance often targets COD <100 mg/L, BOD (biochemical oxygen demand) <30 mg/L and NH₃–N <5–10 mg/L; multi‑stage sequences have delivered final COD ~90 mg/L and NH₃ ~11 mg/L (patents.google.com).

The engineering response is layered: pretreatment to tame shocks and oils, front‑loaded ammonia removal, high‑rate biology (anaerobic then aerobic), and tertiary polishing to strip away refractory organics and salts. Case studies and reviews draw the blueprint (iwaponline.com).

Influent profile and discharge targets

Loads and variability in cokemaking effluent routinely exceed municipal norms by orders of magnitude, including turbidity and oils. To meet strict outfalls, plants design for COD <100 mg/L, BOD <30 mg/L and NH₃–N <5–10 mg/L; full‑scale reports show COD ~90 mg/L and NH₃ ~11 mg/L, with volatile phenol = 0 (patents.google.com).

Equalization and oil/tar removal

Upstream equalization (6–12‑hour basin volume) buffers pH swings, flow surges, and heat loads (for example, from intermittent coke quenching), stabilizing influent strength for downstream units.

Mechanical separation follows: API‑style oil traps or lamella settlers remove free oils and fines; compact trains often specify high‑rate modules such as a lamella settler. One full‑scale sequence used an initial oil trap followed by air‑driven microbubble flotation to coalesce tar/oil into a floating slick for skimming (patents.google.com). Typical flotation/settling removes >80–90% of free oil; dissolved/colloidal solids and coke fines are screened or settled, with plants opting for continuous devices such as an automatic screen where debris loads are high.

When emulsified oils persist or load variability is extreme, operators pair gravity traps with dissolved air flotation; high capture rates and short detention are hallmarks of a DAF unit. As a category, primary physical separation systems set up the downstream biology for success.

Front‑loaded ammonia removal

Coke effluent often benefits from ammonia removal in pretreatment. Chemical precipitation can be used: adding alkali (e.g., Na₂CO₃ or lime) plus Mg²⁺ and PO₄³⁻ crystallizes ammonium as struvite (ammonium‑magnesium phosphate). In one example, dosing MgCl₂ and KH₂PO₄ at a 1.2:1:1 molar ratio at pH≈9 precipitated NH₄MgPO₄·6H₂O and cut NH₄⁺ by >80% (patents.google.com). Metering alkali and magnesium salts is typically handled with a dosing pump to maintain control.

Alternatively or additionally, packed‑tower ammonia stripping is employed. At pH ≈10.8–11.5, ammonium converts to free NH₃, which volatilizes into an air stream in a counter‑current tower (nepis.epa.gov). Well‑designed towers with ≥30‑minute residence at ~20–30 °C routinely achieve 90–95% ammonium removal; operating at pH 11.5 and 20 °C achieved ≈95% ammonia‑N removal in an EPA‑cited design (nepis.epa.gov). The stripped ammonia is recovered (often as dilute ammonium sulfate) or vented with air treatment, and effluent pH is neutralized. In practice, pretreatment can cut ammonia by orders of magnitude (e.g., 6,230→1,080 mg/L NH₃‑N; patents.google.com) and remove 20–40% of COD via precipitation/coagulation and flotation.

Anaerobic–anoxic–aerobic biology

Residual organics and ammonia are then treated biologically. A common sequence is A²/O (anaerobic–anoxic–oxic): an anaerobic reactor (e.g., UASB, upflow anaerobic sludge blanket) followed by anoxic denitrification and aerobic nitrification. The anaerobic stage typically removes 40–70% of remaining COD (biogas production captures energy from volatile acids and simple aromatics). Subsequent aerobic treatment, often in activated sludge, oxidizes most of what remains (iwaponline.com). Plants implementing this train may deploy packaged anaerobic digestion systems and conventional activated sludge tanks, or combine biology and separation in membrane bioreactors (MBR).

Performance is documented: a demonstration with raw COD ~1,230 mg/L achieved 92.3% COD removal to ~98 mg/L and NH₃‑N removal from 278.6 to 1.7 mg/L (97.8%) in an A²/O system (iwaponline.com). In practice, schemes combining anaerobic pretreatment with nitrifying activated sludge (or MBR) routinely achieve ∼90–99% COD removal and ∼95–99% ammonia removal (iwaponline.com), with effluent BOD typically <10–20 mg/L. Advanced schemes can incorporate anammox for energy‑efficient nitrogen removal, though conventional nitrification remains most common.

Nitrification design parameters

Aerobic tanks are sized for biomass retention and oxygen transfer to nitrify: dissolved oxygen is typically held at 2–4 mg/L, and sludge ages often exceed 15–20 days to support slow‑growing nitrifiers. Anoxic zones upstream of aeration enable denitrification using residual BOD from upstream or recycled organic (e.g., methanol or internal recirculation), keeping total nitrogen in check (iwaponline.com).

Advanced oxidation for refractory organics

To hit very stringent discharge or reuse standards, tertiary polishing targets refractory organics, trace phenols, color, and the last ammonia. Advanced oxidation processes (AOPs) such as ozone or catalytic oxidation (O₃ + UV/H₂O₂, including Fenton‑type reactions) are used. Na et al. (2017) reported catalytic ozonation cutting biotreated COD from ~78 mg/L to ~26 mg/L (67% removal) and NH₄–N from ~19.7 to ~4.9 mg/L (75% removal) (iwaponline.com). Ozone dosing in the range of a few grams O₃ per gram COD is typical; UV reactors used in AOP trains are akin to those in ultraviolet disinfection systems.

Adsorptive polishing with carbon

Granular activated carbon (GAC) filters or powdered activated carbon (PAC) additions adsorb remaining organics and color. A recent pilot showed that ~4 g/L of a coal‑based PAC removed ~48% of remaining COD from biologically pretreated coke wastewater (www.mdpi.com), preferentially targeting humic‑ and fulvic‑like fractions (www.mdpi.com). GAC beds are typically sized for several hours of empty‑bed contact time; many plants specify high‑activity media in activated carbon systems to pair with AOP.

Membranes for solids, salts, and PAHs

Membrane filtration locks in final clarity and salt control. Ultrafiltration (UF) removes suspended solids and colloids and is widely adopted as pretreatment to high‑pressure steps; industrial lines often deploy UF modules upstream of desalination. Nanofiltration (NF) or reverse osmosis (RO) then reject dissolved organics and salts; NF can also reduce hardness at lower pressure, making nanofiltration a candidate where partial desalting is sufficient.

One sequence ran sand filtration followed by RO on biologically treated coke effluent: the RO step cut total PAHs (polycyclic aromatic hydrocarbons) from ~94.7 μg/L to 15.0 μg/L (≈84% removal) in permeate, with COD in the single digits (iwaponline.com). Plants often front this train with dual‑media sand filtration and specify brackish‑water RO skids rated for high organics exposure, such as brackish water RO systems. UF/NF/RO configurations are part of broader membrane system portfolios used in industrial water reuse.

Performance benchmarks and compliance

By combining these barriers, strict limits are met: multi‑step plants have reported final effluent ~COD <100 mg/L, BOD <30 mg/L, NH₃–N <5–10 mg/L, and non‑detectable phenols/cyanides (patents.google.com; iwaponline.com). For example, a full‑scale sequence yielded COD ~90 mg/L and NH₃ ~11 mg/L (BOD ~35 mg/L) with volatile phenol = 0 (patents.google.com). Such performance meets or exceeds standards in most jurisdictions; Indonesia’s regulatory limits for phenol (<1 mg/L) and ammoniacal nitrogen (<5–10 mg/L) would be attained.

Key design outcomes and benchmarks include: system COD removal ≈90–99%, ammonia removal ≈95–99%, and proportional effluent concentrations of tens of mg/L or below (iwaponline.com; iwaponline.com). These figures guide sizing (tank volumes) and equipment (aeration, stripping tower dimensions, membrane surface area) in engineering the plant.

Sources and referenced designs

Comprehensive reviews and case studies provide data on coke‑oven effluent and treatment performance (www.mdpi.com; iwaponline.com). Pretreatment design (pH adjustment, stripping) follows well‑established guidelines (nepis.epa.gov; nepis.epa.gov). Pilot/full‑scale metrics (COD/N removal in A²O systems, advanced polishing removals) are documented (iwaponline.com; iwaponline.com; www.mdpi.com), informing tank volumes and equipment selection.

References: Lin et al. (2018) Int. J. Environ. Res. Public Health, 15(3), 441 (www.mdpi.com); Zhang & Huang (2015) Water Sci. Technol. review (iwaponline.com); U.S. EPA Ammonia Stripping Fact Sheet (2000) (nepis.epa.gov; nepis.epa.gov); Na et al. (2017) Water Sci. Technol. (cited in WST review) (iwaponline.com); Xia et al. (2022) Water, 14(15), 2446 (www.mdpi.com); Patent CN103936240A (Chen et al., 2014) (patents.google.com).