The pharma water problem is getting pricier. GAC and RO are the fix.

Pharmaceutical effluents are a chemical grab bag — acids/bases, metal salts, unreacted reagents and low‑level active pharmaceutical ingredient (API) residues from reactor clean‑outs, plus fermentation and extraction wastes laced with organic nutrients, solvents like ethanol, acetone and IPA, and stabilizers (Manufacturing Chemist) (Manufacturing Chemist). These streams carry high chemical oxygen demand (COD; a measure of oxidizable organics) and diverse micropollutants (antibiotics, hormones, etc.) at μg–mg/L levels (micrograms to milligrams per liter), which standard biological treatment struggles to remove.

At the same time, water scarcity and rising prices are forcing a rethink. Global water supply gaps could reach ~40% by 2030 (Manufacturing Chemist), while water costs have ticked up ~4% per year (Manufacturing Chemist).

The result: more manufacturers are pursuing reuse and even zero‑liquid discharge (ZLD) — rerouting water into process, cooling, or cleaning uses — to reduce cost and risk (Manufacturing Chemist) (Manufacturing Chemist). That push hinges on robust tertiary polishing — most often granular activated carbon and reverse osmosis.

Plants typically add these steps after clarification and biological treatment; for solids separation, a clarifier is a common front‑end before polishing. For the RO barrier, packaged membrane systems are widely deployed in industrial and municipal reuse trains.

Granular activated carbon adsorption

GAC (granular activated carbon; highly porous carbon media) adsorbs dissolved organics onto an immense internal surface. Fixed‑bed designs in wastewater polishing are sized on empty‑bed contact time (EBCT; the volume of the carbon bed divided by flow), typically 15–30 minutes (PMC). Longer EBCT not only boosts adsorption but allows some biodegradation within biofilms on the carbon (PMC) (PMC).

Data underline the sizing point: an 18‑minute EBCT GAC column removed on average ~74% of a mixed suite of pharmaceuticals (fexofenadine, oxazepam, cetirizine, DEET, fluconazole), while a 6‑minute EBCT version removed ~56% (MDPI – Water 2020). In practice, GAC is installed as a polishing step after clarification/biological treatment, and spent media are routinely regenerated off‑site to cut environmental impact (PubMed). Tech buyers typically specify coal‑ or coconut‑based media; an activated carbon module is the standard building block.

GAC removal efficiency and bed life

In a year‑long pilot on real water, initial removal across seven representative organics averaged about 80% (MDPI – Water 2020). Over tens of thousands of bed volumes (BV; one BV is a throughput equal to the packed bed volume), removal gradually declined as adsorption sites saturated: combined removal fell from ~90–95% at ~20,000 BV to ~80% at ~75,000 BV (MDPI – Water 2020).

Compound‑specific behavior matters. Lamotrigine removal exceeded 99% and carbamazepine ~90% by the end of the study (MDPI – Water 2020), while less‑adsorbable compounds like DEET and oxazepam broke through earlier yet still saw roughly 0.5–0.7 log removal. Across systems, typical GAC polishing reduces pharmaceutical concentrations by roughly 70–90% when well designed (MDPI – Water 2020) (MDPI – Water 2020).

Service life is measured in BV. A meta‑analysis shows GAC columns commonly exceed 10,000–20,000 BV before even 20% breakthrough for compounds like diclofenac (PubMed). In the Water (2020) pilot, columns at 20–75k BV still maintained 80–95% removal (MDPI – Water 2020). That informs design: to sustain >80% removal, engineers target at least 20,000 BV (~1.8×10^4) between carbon changeouts and hold EBCT in the 15–30 minute range (MDPI – Water 2020), then monitor breakthrough with indicators such as carbamazepine or UV254.

The practical math is straightforward: an 80% reduction turns a 100 μg/L influent into ~20 μg/L effluent of that compound. Many installations use GAC post‑treatment to comply with emerging micropollutant guidelines (e.g., Sweden’s advanced WWTP requirements). The measurable outcome: most pharmaceutically active compounds drop to 10–30% of their initial concentration, while very persistent ones like carbamazepine and lamotrigine are often >90% removed (MDPI – Water 2020).

Reverse osmosis design and pretreatment

RO (reverse osmosis; high‑pressure, semi‑permeable membrane separation) rejects virtually all dissolved ions and many organics. Well‑engineered systems achieve 95–99% total dissolved solids (TDS; the sum of dissolved ions) rejection (AXEON), with designs commonly targeting ~98.5% salt rejection at 60–75% recovery (i.e., 60–75% of feed becomes permeate) (AXEON) (AXEON).

Pharma feeds can have high scaling potential (from salts and clean‑in‑place, or CIP, chemicals), so pretreatment is standard: microfiltration/UF (ultrafiltration) and antiscalants are common. Plants often stage an ultrafiltration barrier ahead of the RO racks and dose specialty membrane antiscalants to protect the array. For brackish feeds, dedicated brackish-water RO configurations are typical practice.

RO performance on salts and organics

RO drives salinity to near‑potable levels. In one study on mixed alumina‑wastewater, final permeate measured aluminum <3 mg/L, sodium ≈145 mg/L and chloride ≈193 mg/L (versus much higher feeds) (PMC). Vendor data indicate 98–99% salt removal is routine (AXEON), making it reasonable to expect a 3,000 mg/L TDS brackish feed to yield ~30–60 mg/L permeate. If ultra‑low TDS is required, multi‑pass RO or blending with fresh water is used.

The membrane barrier also strips pharmaceuticals. Systematic reviews report >90% removal for most drugs (PMC). Compiled data show >99% ibuprofen removal, ~95% diclofenac, >93% for sulfonamide antibiotics, and >99% for macrolide antibiotics and bisphenol A (PMC). In lab tests, a low‑pressure RO membrane rejected 73–99% of amoxicillin/ampicillin under varying conditions (PMC).

RO is a separation, not a reaction process: contaminants that do not pass are concentrated in the reject stream. That’s why many buyers pair high‑rejection spiral elements — for example, Filmtec/DuPont RO membranes — with a concentrate management step.

Reuse yields and ZLD pathways

For high reuse ambitions, RO permeate approaches potable quality. In Indonesia, Kekha Pharma (PT. Kalbe Farma) installed RO on its purified water loop and analyzed the RO reject; measurements showed that reject met Indonesian effluent standards and was classed “Kelas A (Baik Sekali)” — essentially potable quality after minimal treatment (Jurnal Pelita Bangsa). By inference, the RO permeate was even purer. More generally, well‑run RO yields permeate with very low TDS (<50 mg/L) and negligible organics, and some reuse guidelines consider RO‑treated water suitable for filling boilers or even for indirect potable reuse, pending disinfection. A low‑energy option for that barrier is ultraviolet disinfection.

Standalone RO typically recovers 60–75% of influent as high‑quality permeate, with the balance as brine concentrate. To push recovery higher, plants treat the concentrate: one pharmaceutical research lab added a vacuum evaporator after RO and recovered >90% of its wastewater as condensate for reuse (Manufacturing Chemist). Integrated “zero discharge” trains (RO + multi‑effect or mechanical evaporation) have cut waste volumes by up to 98%, dramatically reducing disposal cost (Manufacturing Chemist).

The key outcomes are quantifiable: up to ~2–3 logs of salt and micropollutant removal; permeate conductivity often <50 μS/cm (representing >98% TDS removal) (AXEON); >90% rejection of targeted pharmaceuticals (PMC) (PMC); and high yield (60–75%). For a 2,000 mg/L TDS pharma stream, pressure‑driven RO can produce ~20–40 mg/L TDS permeate with >99% of APIs removed, enabling non‑critical reuse. The remaining concentrate (25–40% of flow) will run roughly 3× the feed TDS and carry the rejected organics; common management is evaporation or deep disposal.

Integrating GAC and RO barriers

GAC and RO are complementary. A GAC bed knocks down hydrophobic organics and reduces COD, protecting the membrane array from fouling; RO then removes dissolved salts, metals and residual micropollutants to finish the polish. Selection lines up with objectives: strict salt discharge points toward RO first, while refined micropollutant control argues for a GAC stage. For the GAC route, an 18‑minute EBCT delivered ~74% drug removal in testing (MDPI – Water 2020). For the RO route, ~98% salt rejection is a workable planning basis (AXEON), with recoveries tuned based on fouling risk and pretreatment quality. Plants often add biological digestion upstream and reserve polishing for these advanced barriers.

Adoption, regulation and economics

Worldwide, operators are adding advanced polishing to pharmaceutical effluent. In the EU and beyond, municipal WWTPs are retrofitting GAC or ozonation to address API discharge. In pharma, RO and evaporators are increasingly used to recover process water. In Indonesia, facilities must meet strict effluent BOD/COD and other standards, even though pharmaceuticals are not yet specifically regulated; the Kalbe Farma case shows RO can deliver to the top regulatory class (Jurnal Pelita Bangsa). Market analyses project strong growth in GAC and membrane use for reuse — a trend mirrored in orders for modular RO/NF/UF systems.

Costs are real: RO draws high‑pressure energy; GAC needs periodic reactivation. But avoided sewer fees, lower raw water purchases and minimized penalty risk can offset. One lab’s decision to add vacuum evaporation after RO enabled >90% water recovery (Manufacturing Chemist). In round numbers, data‑driven GAC+RO design can deliver >70% reuse of treated effluent and >90% reduction in pollutant loads (salts and organics), according to the sources cited above. Plants round out installations with supporting ancillaries for monitoring and control.

Reference sources

Data and statements above are supported by recent literature and reports (PMC) (Jurnal Pelita Bangsa) (PMC) (PMC) (MDPI – Water 2020) (MDPI – Water 2020) (MDPI – Water 2020) (AXEON) (AXEON) (Manufacturing Chemist) (Manufacturing Chemist) (Manufacturing Chemist) (full references provided).