Pharma’s tiniest wastewater stream is its riskiest — and regulators want it cleaned to parts‑per‑billion

High‑potency pharmaceutical waste — think antibiotics and cytotoxics — is officially hazardous B3 waste under Indonesian law, yet specific effluent limits for these “pharma pollutants” are still absent (pusakom.pom.go.id). Industry studies flag trace antibiotics such as ciprofloxacin and tetracycline in effluents (pusakom.pom.go.id).

A 2023 survey of 74 Indonesian drug plants found about 5% of outlets outside the legal pH 6–9 window and roughly 14% with dissolved oxygen (DO) under 4 mg/L (pusakom.pom.go.id), signaling neutralization gaps. Globally, WHO/UNEP guidance (Sept 2024) calls for binding manufacturing‑site discharge targets because of “high levels of antibiotics in water bodies downstream of manufacturing sites,” advocating PNECs (Predicted No‑Effect Concentrations) often below 1 ppb (www.unwater.org; axinewater.com).

A USGS study cited by industry sources found API concentrations from drug plants far above consumer wastewater (axinewater.com). The result: a realistic pretreatment goal is compliance with Indonesian pH 6–9 (Permen LHK 68/2016) alongside polishing organics and micro‑pollutants down toward emerging ng/L‑level API targets (pusakom.pom.go.id; axinewater.com; www.cidrap.umn.edu).

Regulatory targets and risk profile

Neutral pH and removal of suspended solids/metals are table stakes; the harder lift is reducing active pharmaceutical ingredients (APIs) to PNECs that can be sub‑ppb (www.unwater.org; axinewater.com). For context, manufacturing effluent often exceeds urban sewage concentrations (www.cidrap.umn.edu).

Design criteria therefore include pH neutrality (6–9 per Permen LHK 68/2016), suspended/metal solids removal, and polishing/degradation of organics to PNECs (pusakom.pom.go.id; axinewater.com; www.cidrap.umn.edu).

Waste stream characteristics and design basis

Deactivation and cleaning of cytotoxic/antibiotic equipment can use strong NaOH or acids, yielding extremely high or low pH, high conductivity/TDS, and concentrated API residues. Literature reports pharmaceutical wastewater with very high organics — azithromycin production effluent at ~1625 mg/L TOC (total organic carbon) before treatment (www.sciencedirect.com), and solvent residues such as benzene or dichloromethane and toxic intermediates like aromatic amines (www.mdpi.com).

For design, assume small flows (e.g., less than 10 m³/day) with COD/BOD in the thousands of mg/L and API concentrations above 100 mg/L, based on cases reported in the literature (www.sciencedirect.com; www.mdpi.com). Even small volumes carry disproportionate hazard and must be treated to micro‑concentration levels.

Neutralization reactor and dosing control

The first unit operation is automated pH adjustment to Indonesian standards (final effluent 6–9), typically to a setpoint near 7–8 (pusakom.pom.go.id). Inline pH probes drive dosing of acid (H₂SO₄ or HCl) or base (NaOH) to achieve >95% neutralization efficiency; batch reactors or continuous stirred tanks are both viable, with flow‑proportional verification of stability afterward.

Plants favor precise chemical feed gear — for example, an accurate dosing pump to avoid overshoot — and sometimes lime [Ca(OH)₂] to raise pH while precipitating CaCO₃ and flocculating ions (EPA fundamentals: nepis.epa.gov).

Coagulation, precipitation, and settling

After neutralization, chemical precipitation/coagulation strips suspended organics, colloids, and residual inorganics. Ferric chloride (FeCl₃) or alum, typically at 50–150 mg/L FeCl₃ following jar testing, forms hydroxide flocs at optimal pH that “sweep” particles and dissolved species from solution (EPA: nepis.epa.gov). Plants commonly source coagulants and tune doses in rapid/slow mixing zones before settling.

Polymer programs can enhance floc formation; commercial flocculants are added downstream of coagulants. A settling basin or compact lamella settler can remove more than 90% of turbidity and over 80% of suspended COD; combined lime/iron programs reduce metal hardness via CaCO₃ and Fe(OH)₃ formation (EPA: nepis.epa.gov). Resulting sludge is managed as B3 waste. Well‑optimized coagulation can eliminate more than 75% of remaining COD and TSS prior to carbon polishing (nepis.epa.gov).

Carbon adsorption and multi‑hour EBCT

Granular activated carbon (GAC) provides final polishing for APIs, solvents, and amines. Two GAC beds in series are recommended to maximize empty‑bed contact time (EBCT, the hydraulic residence time inside a carbon bed) and stretch run length; activated carbon’s high surface area underpins broad organic uptake (EPA, Carbon Adsorption Manual: nepis.epa.gov).

Multi‑hour EBCTs are often necessary in high‑strength pharma streams; one study treating ~1625 mg/L TOC achieved about 87% TOC removal at ~5 hours EBCT (www.sciencedirect.com). In batch tests at 24 hours, GAC removed roughly 52% of TOC from a 912 mg/L synthetic mix (www.mdpi.com), with initial rapid adsorption in the first hour and slower uptake thereafter (www.mdpi.com).

Adsorption capacity varies; one report cites a maximum of ~522 mgC per g GAC (www.mdpi.com). At 1625 mg/L influent, about 87% removal over 5 hours required ~8 g/L GAC dose (see [33]). As an engineering illustration, a 5 m³/day stream might employ around 2–4 m³ of GAC (about 1,250–2,500 kg) split across two 2 m³ vessels, with backwashing and US‑gas sparging provisions for particulate control. Plants typically pair media with certified activated carbon supply chains to ensure performance and compliant disposal.

Hexane-blinded: Removal performance differs by molecule; reports show GAC can remove 13–90% of different drugs, with diclofenac reaching about 98% under ideal conditions (www.mdpi.com). In realistic waters, overall COD removals are approximately 50–70% (EPA: nepis.epa.gov). If feed COD/BOD is 1,000–2,000 mg/L, post‑carbon values may fall to a few hundred mg/L (a 50–80% drop); verification must include target APIs near PNECs (<1 µg/L) (axinewater.com). Even 95% removal can leave traces, so multiple carbon stages and strict source control are common.

Operations track GAC exhaustion via API or surrogates (e.g., TOC). Spent carbon is API‑contaminated and handled as B3 waste; on‑site thermal regeneration is possible but costly, so most plants replace media and ship used GAC off‑site as hazardous (axinewater.com). To protect the beds, fine filtration should precede adsorption — for example, a cartridge filter installed in a sanitary 316L stainless steel housing.

Optional advanced oxidation polishing

Where ultra‑low PNECs are mandatory, advanced oxidation (e.g., ozone/H₂O₂) or electrochemical oxidation can follow GAC. One vendor reports >99% destruction of APIs via EAOP with no chemical waste (axinewater.com), though such systems are outside this basic design scope.

Expected outcomes and compliance envelope

A well‑sized neutralization/precipitation step can remove nearly 90–95% of inorganics and settleable solids. GAC polishing contributes a further 50–85% TOC/COD removal for concentrated pharmaceutical effluents, per peer‑reviewed studies (www.sciencedirect.com; www.mdpi.com). In one pilot, 1,625 mg/L TOC was cut to ~212 mg/L after ~87% removal (www.sciencedirect.com).

Applied to a 2,000 mg/L COD stream, the design target is under ~300 mg/L COD (about 85% removal) with pH ≈ 7 and heavy metals to non‑detect. Indonesian discharge limits require pH 6–9 and metals below thresholds; benchmarking to Permen LHK 5/2014 suggests COD limits of ~250–1,000 mg/L for most industries, which this train is designed to meet. Practically, polishing also raises downstream DO because high‑demand organics are removed. Benchmarks from the literature include 50–90% organic removal, pH = 7, DO > 4 mg/L, and TDS < 2 g/L (pusakom.pom.go.id; www.cidrap.umn.edu).

Critically, if a potent drug enters at 10 mg/L, roughly 99.99% removal is needed to reach 1 ppb (axinewater.com). Carbon alone rarely achieves this, so multiple stages and/or upstream oxidation plus strict source controls are often combined.

Monitoring, redundancy, and modular control

Continuous instrumentation — flow meters, pH probes, and periodic API assays via LC‑MS/MS — underpins performance assurance. Jar tests and pilots dial in coagulant doses (EPA: nepis.epa.gov), while GAC beds are audited for breakthrough. Supporting controls and skids fall under water treatment ancillaries for valves, mixers, and interlocks.

Designs build a 20–30% neutralization capacity margin over peak acid/base loads, and incorporate modular tanks so excess flow — for example, during CIP (clean‑in‑place) flushes — can be diverted or held, preventing bypass of untreated streams.

Business calculus and policy trendlines

On‑site treatment avoids trucking and incinerating liquid wastes at 800–1,200 °C — a common pharmaceutical practice that adds CO₂ and hazard transport — and can convert hazardous flows into water reusable for cooling or cleaning (axinewater.com).

Pressure is mounting: by 2025 the global API market is ~US$250 billion, and scrutiny of pharmaceutical effluent is intensifying (axinewater.com; www.cidrap.umn.edu). WHO/UNEP now explicitly advocates discharge targets (PNECs) for manufacturers (www.unwater.org; www.cidrap.umn.edu). For example, an antibiotic maker must meet emerging WHO PNEC targets (<1–100 ng/L for various drugs), otherwise they cannot legally sell in regulated markets.

System architecture summary

In summary, a dedicated pretreatment line — automated pH neutralization, chemical precipitation/flocculation, and granular activated carbon polishing — sized to small but extreme flows can de‑contaminate high‑potency deactivation effluent. The train is intended to meet Indonesian discharge limits (COD, BOD, TSS under regulated maxima; pH 6–9; heavy metals below thresholds) and push APIs toward ng/L, with literature benchmarks of 50–90% organic removal, pH = 7, DO > 4 mg/L, and TDS < 2 g/L (pusakom.pom.go.id; www.cidrap.umn.edu). EPA design fundamentals for precipitation and carbon adsorption provide the core engineering basis (nepis.epa.gov; nepis.epa.gov).