Why API Effluent Breaks Conventional ETPs
Conventional on-site systems and municipal WWTPs "are either ineffective at treating APIs or unable to treat APIs to the levels required to minimize impacts to the environment" (Axine Water Technologies, 2020-02), and that gap has widened rather than closed in the six years since. API effluent reaches the drain through three pathways — consumer excretion, improper disposal, and manufacturing discharge — and a 2018 U.S. Geological Survey report cited by Axine flagged manufacturing facilities as a major, point-source contributor that the plant itself can close. Inside an API plant, the wastewater stream is generated at four pinch points: spent fermentation broth, solvent-recovery condensates, mother liquors from crystallization, and cleaning-in-place (CIP) rinses. Each carries a different mix of high COD, residual solvents, antibiotics, hormones, antiretrovirals, and elevated TDS/sulfate from acid/salt splits, and the composite flow varies hour by hour as campaigns change. Single-stage activated sludge is not a defensible 2026 design for this matrix because it was sized for BOD, not for refractory API residues, low BOD/COD ratios (0.3–0.5), and the toxicity spikes that crash biomass. A multi-barrier train — equalization, primary clarification, biological treatment (MBR or SBR), advanced oxidation, and RO/AC polishing — is the standard engineering response to API discharge limits under China GB 21904-2008, US EPA categorical standards, EU BAT-AEL, and India CPCB draft norms.
2026 API Effluent Design Basis: Loadings and Discharge Targets
A defensible 2026 design starts with a numeric envelope: COD 5,000–25,000 mg/L, BOD₅ 1,500–8,000 mg/L, BOD/COD 0.3–0.5, TSS 500–3,000 mg/L, TDS 5,000–15,000 mg/L, sulfate 1,000–5,000 mg/L, total nitrogen 100–800 mg/L, and pH swings of 2–11 across batch campaigns. Volume baselines run 5–200 m³/day for a single API line and 200–5,000 m³/day for multi-product API plants, with equalization sized for 12–24 h HRT to damp diurnal and campaign variation. Effluent targets for 2026 compliance land at COD ≤250–500 mg/L (China GB 21904-2008 category, second-order COD limit), BOD ≤150 mg/L, total APIs non-detect to <0.1 μg/L depending on jurisdiction, and TDS stripped via an RO reuse loop. Antibiotic classes that drive advanced treatment include tetracyclines, beta-lactams, fluoroquinolones, sulfonamides, macrolides, analgesics, hormones, and antiretrovirals — most are recalcitrant and only partially biodegradable, so biology alone is insufficient.
| Parameter | Typical Influent | Design Effluent | Driver |
|---|---|---|---|
| COD | 5,000–25,000 mg/L | ≤250–500 mg/L | China GB 21904-2008 |
| BOD₅ | 1,500–8,000 mg/L | ≤150 mg/L | China GB 21904-2008 |
| TSS | 500–3,000 mg/L | ≤5 mg/L (post-MBR) | Reuse/RO protection |
| TDS | 5,000–15,000 mg/L | <50 mg/L permeate | RO reuse for CIP/cooling |
| Sulfate | 1,000–5,000 mg/L | Site-specific | Anaerobic/aerobic split |
| Total nitrogen | 100–800 mg/L | ≤30–70 mg/L | EU BAT-AEL / discharge |
| Total APIs | 0.1–50 mg/L (sum) | ND to <0.1 μg/L | EU watch list / site cap |
| pH | 2–11 swings | 6.5–7.5 | Equalization |
Process Train Selection: From Equalization to Polishing
The unit-operation chain below is the default train a process engineer should anchor sizing to for a greenfield or major ETP upgrade in 2026; each stage is sized independently and reviewed against the loadings in the previous section.
- Stage 1 — Equalization & neutralization: 12–24 h HRT, pH control to 6.5–7.5, often with a DAF system for solvent and CIP pre-treatment upstream to strip free oil, FOG, and colloids from fermentation and CIP streams.
- Stage 2 — Primary clarification / DAF: 4–300 m³/h capacity range, removes TSS and floatables that would otherwise overload biology.
- Stage 3 — Biological treatment: an MBR system for pharma effluent biological stage or SBR is preferred over conventional activated sludge; MBR delivers COD ≤200 mg/L and TSS ≤5 mg/L with roughly 60% smaller footprint than CAS at matched load (Zhongsheng field data, 2026).
- Stage 4 — Advanced oxidation (AOP): O₃, O₃/H₂O₂, Fenton, or electrochemical AOP for non-biodegradable API residue polishing; this is the unit that lets the plant meet μg/L API limits.
- Stage 5 — Polishing & reuse: an RO polishing stage for API effluent with 65–75% single-pass recovery (up to 95% with two-pass) plus activated carbon; ClO₂ or ozone for final disinfection.
- Stage 6 (conditional) — ZLD: for high-TDS sites or zero-discharge pharma parks, add a brine concentrator and crystallizer downstream of the RO reject.
| Stage | Equipment | Key Design Parameter | Function |
|---|---|---|---|
| 1 | Equalization tank | 12–24 h HRT | Flow & pH dampening |
| 2 | DAF (ZSQ) | 4–300 m³/h | TSS, FOG, colloids |
| 3 | MBR (integrated) | HRT 24–48 h, MLSS 8,000–12,000 mg/L | COD/BOD/TN removal |
| 4 | AOP (O₃/H₂O₂ or EAOP) | 30–60% TOC removal | API residue destruction |
| 5 | RO + AC + ClO₂ | 65–95% recovery | Polishing & disinfection |
Biological Reactor Sizing: MBR vs SBR for API Loads
MBR and SBR are the two realistic choices for a 2026 API ETP; conventional activated sludge fails on footprint, sludge retention, and effluent TSS. MBR design targets: MLSS 8,000–12,000 mg/L, HRT 24–48 h, SRT 20–40 d, membrane flux 10–18 L/m²·h, with PVDF 0.1 μm flat-sheet or hollow-fiber membranes. The DF series flat-sheet MBR modules cover 80–225 m² per cassette, run at 0.1 μm pore size, consume 10–20× less energy than external cross-flow, and deliver 32–135 m³/day per cassette. SBR design targets: cycle time 6–12 h, MLSS 4,000–6,000 mg/L, decant volume 25–35% — lower capex for batch API plants but a larger footprint and less stable effluent. Use SBR for batch, low-flow (<50 m³/day) plants with highly variable loads; use MBR for continuous, medium/high-flow plants that need reuse-grade effluent and a small footprint. Common sizing mistakes worth flagging in any design review: undersized equalization that lets pH or temperature spikes crash biomass, no pH/temperature trim on antibiotic streams, and no sludge-wasting strategy when influent TDS exceeds ~8,000 mg/L (osmotic stress on biomass).
| Parameter | MBR | SBR | Selection Driver |
|---|---|---|---|
| MLSS | 8,000–12,000 mg/L | 4,000–6,000 mg/L | Loading & footprint |
| HRT | 24–48 h | Equivalent 24–72 h (cycle basis) | Hydraulic variability |
| SRT | 20–40 d | 15–30 d | Sludge stability |
| Membrane flux | 10–18 L/m²·h | N/A | Membrane area |
| Effluent TSS | ≤5 mg/L | 30–80 mg/L | Reuse / RO feed |
| Footprint | Compact | 1.6–2.5× larger | Site constraints |
| Capex | Higher | Lower | Batch vs continuous |
AOP and Membrane Polishing: Pushing APIs to Non-Detect
Biology gets COD down; AOP gets APIs to non-detect. Among AOP options, O₃/H₂O₂ delivers 30–60% TOC removal at moderate capital cost and is the workhorse for effluent polishing; Fenton handles high-COD spikes well but adds sludge burden and iron handling; photocatalytic UV is effective on specific API classes but capital-heavy; electrochemical AOP (EAOP, per Axine) achieves high API destruction with low chemical use and a smaller footprint — a strong fit for water-stressed sites. RO sizing rules: feed from MBR/AOP effluent with SDI <3, 65–75% single-pass recovery, 85–95% with two-pass, permeate TDS <50 mg/L suitable for CIP and cooling-tower make-up. The Zhongsheng RO series is rated up to 95% recovery with PLC automation, suitable for the pharma polishing duty (Zhongsheng product catalog, 2026). For final microbial control, a ClO₂ disinfection before discharge or reuse unit sized 50 g/h to 20,000 g/h gives broad-spectrum kill without the THM precursor profile of chlorine. On water-stressed sites (India, Middle East, inland China), plan ZLD: RO brine → brine concentrator → crystallizer; expect a 35–60% OPEX penalty but zero liquid risk.
| AOP Option | TOC/API Removal | Capex | OPEX Driver | Best Fit |
|---|---|---|---|---|
| O₃ / O₃+H₂O₂ | 30–60% TOC | Moderate | Ozone generator power | General API polishing |
| Fenton | 40–70% COD spike | Low | H₂O₂ + Fe + sludge | High-COD shock loads |
| Photocatalytic UV | Class-specific | High | Lamp replacement | Niche API lists |
| EAOP (Axine) | >90% API | High | Low chemicals, high power | Refractory APIs, ZLD |
2026 CAPEX, OPEX and ZLD Decision Framework
Translating process choices into a defensible budget is the part most top-ranking pages skip. For an MBR + AOP + RO train (no ZLD) in 2026, a 5–50 m³/day single-line API ETP runs CAPEX USD 1.2M–4.5M and OPEX USD 0.9–2.2/m³; a 50–500 m³/day multi-train plant runs CAPEX USD 4.5M–18M and OPEX USD 0.6–1.4/m³ at scale. OPEX breaks down roughly as energy 35–45%, chemicals 15–25%, membranes and parts 10–15%, and labor 15–20% (Zhongsheng field data, 2026). ZLD adds roughly 35–60% to OPEX and 20–40% to CAPEX — mandatory in water-stressed regions and on zero-discharge pharma parks. Use this decision rule: choose ZLD when influent TDS >8,000 mg/L, the site is discharge-restricted, or produced water is needed for reuse at >70% of demand. A useful cross-reference for adjacent cost lines is the vitamin manufacturing wastewater cost benchmark and the broader 2026 industrial wastewater market trends for capex normalization.
| Plant Size | Flow (m³/day) | CAPEX (USD) | OPEX (USD/m³) | ZLD Adders |
|---|---|---|---|---|
| Small | 5–50 | 1.2M–4.5M | 0.9–2.2 | +35–60% OPEX |
| Medium | 50–200 | 4.5M–10M | 0.7–1.4 | +30–50% OPEX |
| Large | 200–500 | 10M–18M | 0.6–1.0 | +20–40% OPEX |
| Mega | >500 (up to 5,000) | 18M–60M+ | 0.5–0.9 | Case-by-case |
2026 Compliance Checklist and Frequently Asked Questions
2026 discharge-limit snapshot for an API ETP: China GB 21904-2008 — COD ≤250 mg/L, BOD ≤150 mg/L, SS ≤60 mg/L, total APIs as per the catalog (with stricter local amendments in pharma parks); US EPA categorical pharmaceutical standards (40 CFR 439) and effluent limitations guidelines for active ingredients; EU BAT-AEL under the BREF for the manufacture of organic fine chemicals (OFCHEM) — COD 30–250 mg/L range, TOC 10–60 mg/L, AOX 1–5 mg/L, and an increasing focus on watch-list APIs; India CPCB draft pharma norms — tightening COD/BOD and adding API-specific caps. A line-of-sight link between biology, AOP, and RO/ClO₂ is what gets you across all four regimes.
FAQ 1 — What influent COD can a properly designed API ETP handle? 5,000–25,000 mg/L is the typical envelope; an MBR + AOP train is designed to bring this to ≤250–500 mg/L with APIs to non-detect on a site-specific basis.
FAQ 2 — MBR or SBR for an API plant? MBR for continuous, medium/high-flow plants needing reuse-grade effluent; SBR for batch, low-flow (<50 m³/day) plants with variable loads and tight capex.
FAQ 3 — Which AOP is best for API residue polishing? O₃/H₂O₂ is the workhorse; Fenton handles COD spikes; EAOP (per Axine) is the strong fit for refractory APIs and water-stressed sites.
FAQ 4 — When is ZLD justified for a pharma ETP? When influent TDS >8,000 mg/L, the site is discharge-restricted, or reuse demand is >70% of total water — expect a 35–60% OPEX uplift.
FAQ 5 — What is the realistic 2026 CAPEX for a 100 m³/day API ETP? Roughly USD 6M–12M for an MBR + AOP + RO train without ZLD, with OPEX USD 0.7–1.4/m³ (Zhongsheng field data, 2026). For a related regulated-stream benchmark, see this hospital wastewater engineering guide.
