The study, published on 27 July 2026 in the journal Energy & Environment Nexus by Mei Ma’s team at the Chinese Academy of Sciences, is notable for its unusually comprehensive scope. Rather than measuring a handful of known chemicals, the team assessed the total biological activity of wastewater samples using effect-based tools that capture the combined action of entire chemical mixtures. This matters because oxidation and photochemical reactions during advanced treatment can generate transformation products whose biological properties differ sharply from those of their parent compounds. Some of these byproducts may be cytotoxic, genotoxic, or endocrine-disrupting, yet routine chemical analysis—built to detect known target compounds—can miss their combined effects entirely.

The researchers worked at a typical wastewater treatment plant in Tianjin, China, supplemented by two pilot-scale treatment systems. Samples were collected from 11 locations spanning the full treatment train: the plant inlet, preliminary sedimentation, anaerobic-anoxic-oxic (A2O) biological treatment, secondary sedimentation, ultraviolet disinfection, ozonation, coagulation, and biological activated carbon filtration. Organic compounds were isolated from each sample using solid-phase extraction, concentrating the dissolved chemistry so that even trace activity could be measured.

The heart of the method lay in a panel of recombinant yeast assays. Each assay carried a human hormone receptor—estrogen receptor alpha, androgen receptor, or thyroid receptor beta—wired to a reporter system, allowing the team to quantify both receptor-agonistic effects, in which chemicals mimic a hormone, and receptor-antagonistic effects, in which chemicals block a hormone from acting. Bioactivity was expressed as equivalent concentrations of reference chemicals, giving the results a tangible toxicological meaning. Crucially, the team also derived an effect-based trigger value of 7.97 micrograms of flutamide equivalents per liter for androgen receptor-antagonistic activity, providing a benchmark against which treated effluents could be judged.

For estrogenic activity, the treatment train performed largely as intended. Estrogen receptor-agonistic activity was prominent in untreated and preliminarily treated wastewater but declined progressively through the plant, falling below the detection limit after secondary sedimentation. Estrogen receptor-antagonistic activity followed a similar trajectory, dropping from 58.4 micrograms per liter in the influent to roughly 7.8 to 9.2 micrograms per liter after treatment. These results confirm that conventional biological treatment, particularly the A2O process, removes most detectable estrogen-linked endocrine activity before water ever reaches the advanced polishing stages.

Thyroid-related activity told a similar, though slightly more complicated, story. Thyroid receptor-antagonistic activity fell dramatically from 0.56 milligrams per liter to 0.0085 milligrams per liter through the main plant. In the pilot systems, ozonation—and ozonation followed by biological activated carbon filtration—further suppressed this activity, demonstrating that oxidative processes can genuinely help control certain endocrine endpoints. Yet coagulation, a standard physicochemical step, produced a slight rebound in thyroid receptor antagonism. The researchers suggest this may occur because residual endocrine-disrupting chemicals compete with other organic matter during removal, escaping the coagulation process while bulk organic material is stripped away.

The most alarming findings concerned the androgen receptor. No androgen receptor-agonistic activity was detected anywhere in the treatment train—good news in itself—but antagonistic activity, which can interfere with male hormone signaling, was present throughout. The A2O biological process reduced it markedly, from 156.6 to 12.4 micrograms of flutamide equivalents per liter, a roughly twelvefold decrease that brought the effluent comfortably below the trigger value. What happened next undermined that progress. Both ultraviolet irradiation and ozonation increased anti-androgenic activity, pushing the treated effluents above the newly established threshold of 7.97 micrograms per liter. Even biological activated carbon filtration applied after ozonation did not fully reverse the increase.

The researchers propose a chemical explanation for this counterintuitive result. Advanced oxidation and UV-driven photochemistry rarely mineralize organic contaminants completely; instead, they transform parent compounds into partially degraded products through reactions such as hydroxylation, dealkylation, decarboxylation, and deamination. These transformation products may retain or acquire the ability to bind and block the androgen receptor, even as the original pollutants disappear from chemical inventories. In other words, a treatment plant can appear to be succeeding by every conventional metric—removing parent pollutants, reducing chemical oxygen demand, eliminating pathogens—while its effluent grows biologically more hazardous along a specific endocrine pathway.

The broader implication is that treatment performance cannot be judged solely by pollutant removal or by measuring a single biological endpoint. UV irradiation and ozonation remain valuable technologies for pathogen control and contaminant degradation, and the study does not argue for abandoning them. Instead, it calls for optimizing their operating conditions and downstream treatment configurations to prevent the formation or persistence of hazardous transformation products. It also argues for embedding multi-endpoint, effect-based monitoring into routine wastewater management, so that risks invisible to chemical analysis—such as mixture effects that simultaneously activate or inhibit different hormone receptors—can be detected before reclaimed water is discharged or reused.

As water reuse expands from arid regions into mainstream urban planning, the study offers a timely reminder that purification is not the same as safety. Effect-based trigger values like the one derived for anti-androgenic activity could become a template for regulators seeking to balance purification efficiency against ecological and human health protection. For now, the message for engineers and water utilities is clear: the cleanest-looking effluent is not necessarily the safest, and the next generation of wastewater reclamation systems must be designed with the full spectrum of hormone receptor activity in view.

Subject of Research: Endocrine-disrupting effects of transformation products formed during advanced wastewater treatment

Article Title: Advanced wastewater treatments may intensify hidden endocrine risks

Article References: Advanced wastewater treatments may intensify hidden endocrine risks. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: wastewater treatment, endocrine disruption, ozonation, ultraviolet irradiation, androgen receptor, estrogen receptor, thyroid receptor, water reclamation, transformation products, effect-based monitoring, A2O treatment, flutamide equivalents

The team, led by Ferlian Vida Satriaji and Jheng-Jie Jiang of Chung Yuan Christian University together with collaborators in Taiwan, the Czech Republic, and Japan, tracked nine target PFAS compounds through five points in the treatment process: the raw influent, primary sedimentation, final sedimentation, the membrane bioreactor, and the final effluent. Their facility employed an anaerobic-anoxic-oxic biological process followed by membrane filtration, a configuration widely regarded as state of the art for municipal wastewater recycling. Sampling was unusually thorough. Twelve monthly grab-sampling events ran from June 2018 to May 2019, and a separate intensive campaign collected daily samples over seven consecutive days from 21 to 27 May 2019, allowing the researchers to separate seasonal patterns from short-term variability.

Seven of the nine target compounds were detected above their compound-specific method detection limits in at least one sample. Only the two longest-chain compounds examined, perfluoroundecanoic acid (PFUnA) and perfluorododecanoic acid (PFDoA), were never detected, a finding consistent with the phase-out and declining use of long-chain PFAS chemistry in consumer and industrial products over the past two decades. The workhorse compounds of the legacy PFAS era, perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS), dominated the profile. Together they accounted for 70.4 percent of the estimated annual mass of PFAS leaving the plant, underscoring how these two legacy chemicals continue to define the PFAS burden of municipal wastewater even years after their production was curtailed in much of the world.

The headline numbers from the monthly dataset tell a subtle story. Mean concentrations of the seven quantified compounds, summed as sigma-PFAS, were 20.84 nanograms per liter in the influent, rose to 29.62 nanograms per liter after the membrane bioreactor, and stood at 26.21 nanograms per liter in the final effluent. In other words, the plant did not remove PFAS on balance; concentrations in the treated water were actually higher than in the incoming wastewater. This counterintuitive pattern, observed repeatedly at treatment plants around the world, is usually attributed to the transformation of precursor chemicals. Fluorinated precursors that were not directly measured in this study can be converted by microbial and chemical activity into the stable perfluoroalkyl acids that the analytical methods capture, so the mass of measurable PFAS can grow even as the total fluorinated burden stays constant or slowly declines.

Statistical analysis sharpened that picture. Friedman tests, a nonparametric method suited to repeated measurements across matched treatment stages, revealed significant differences in summed PFAS concentrations among the five sampling points in both the monthly dataset and the seven-day campaign, with p-values of 0.002 and 0.013 respectively. Yet the direct comparison between influent and final effluent, the pair that matters most for assessing what a plant sends to receiving waters, was not statistically significant in either campaign. The stage-wise differences were real, but the net change across the whole plant was small enough to be lost within the variability of the data. That distinction matters for regulators and engineers alike: a treatment train can reshuffle PFAS concentrations internally without meaningfully reducing the load discharged to rivers or reuse systems.

Translating concentrations into mass gives the study its practical punch. The researchers estimated that the facility discharges approximately 0.314 kilograms of the seven quantified PFAS compounds into the environment each year through its final effluent. That figure may sound modest, but it represents a continuous, diffuse input of essentially nondegradable chemicals into a receiving watershed, accumulating alongside contributions from every other upstream discharge. Because PFAS do not break down, each kilogram released persists and can migrate through surface water, groundwater, soil, and eventually drinking water supplies. Mass discharge estimates of this kind are increasingly recognized as the critical metric for prioritizing which sources to control first, since a plant with moderate concentrations but enormous flow volumes can contribute more pollution than a smaller source with higher concentrations.

Risk assessment in the study offered reassurance, with caveats. The team calculated compound-specific risk quotients by comparing measured effluent concentrations against ecotoxicological benchmarks drawn from resources such as the United States Environmental Protection Agency’s ECOTOX knowledgebase. The resulting quotients ranged from 1.30 times ten to the minus nine to 1.65 times ten to the minus five, all classified as negligible under the criteria applied. Such calculations, however, only cover the compounds measured and the endpoints selected. PFAS mixtures can act additively or synergistically, regulatory toxicity thresholds continue to tighten as new health data emerge, and the risk landscape can shift dramatically if precursor transformation downstream generates additional perfluoroalkyl acids after discharge. A negligible quotient today is not a guarantee of negligible risk under tomorrow’s benchmarks.

The authors are candid about the limits of their mass balance. Precursor compounds, sludge, suspended solids, and total organic fluorine were not measured, which means the observed differences between treatment stages cannot be used to definitively establish precursor transformation, phase partitioning between water and solids, or a complete accounting of fluorinated mass through the plant. Sludge is a particularly important blind spot. PFAS sorb to solids and biosolids, and when sludge is land-applied or landfilled, those chemicals can re-enter the environment through entirely separate pathways. The oxidizable precursor assay, a technique increasingly used to convert hidden precursors into measurable perfluoroalkyl acids, would be a natural next step for facilities seeking a fuller picture, as recent screening work on mixed liquor and effluents has demonstrated.

Why does any of this matter beyond one plant in Taoyuan? Water resource recycling centers are the linchpin of circular water economies, especially in water-stressed regions like Taiwan, where reclaimed water is being promoted for industrial and environmental uses. If PFAS pass through reclamation essentially untouched, then every liter of recycled water carries a legacy chemical load, and every year of operation adds to the environmental stock. Conventional biological treatment and membrane bioreactors were simply never designed to break carbon-fluorine bonds. Emerging technologies such as granular activated carbon, ion exchange, reverse osmosis, and destructive techniques including supercritical water oxidation and electrochemical treatment show promise, but they carry substantial costs and energy demands, and they concentrate rather than eliminate the problem unless destruction is included.

The study’s greatest value may lie in its template. A full year of monthly sampling plus an intensive daily campaign, applied across five treatment stages with compound-specific detection limits and nonparametric statistics, provides exactly the kind of facility-specific evidence that regulators need to design monitoring programs and that engineers need to justify targeted PFAS treatment investments. As governments worldwide move to restrict PFAS manufacture, designate them as hazardous substances, and set enforceable limits in effluents and drinking water, studies like this one quantify the scale of the challenge: the forever chemicals we have already released keep cycling back through our pipes, and the plants built to clean our water are, for now, largely transparent to them. Closing that gap is one of the defining environmental engineering tasks of the coming decade.

Subject of Research: Fate and mass discharge of per- and polyfluoroalkyl substances across treatment stages at a municipal water resource recycling center

Article Title: Treatment-stage variation and final-effluent mass discharge of per- and polyfluoroalkyl substances at a municipal water resource recycling center

Article References: Satriaji, F. V., Li, M.-C., Jachimowicz, P., Solayman, H. M., Mori, A. M. J. P., Le Tong, C. T., Widyastuti, A., Nguyen, T.-B., Horie, Y., Chao, H.-R., & Jiang, J.-J. (2026). Treatment-stage variation and final-effluent mass discharge of per- and polyfluoroalkyl substances at a municipal water resource recycling center. Environmental Geochemistry and Health, 48(14), Article 576. https://doi.org/10.1007/s10653-026-03471-7

Image Credits: AI Generated

DOI: 10.1007/s10653-026-03471-7

Keywords: PFAS, forever chemicals, wastewater treatment, membrane bioreactor, water recycling, PFOA, PFOS, mass discharge, effluent, ecological risk, Taiwan, water reclamation