Wastewater-based epidemiology

01Overview and value chain

Markers: [EC: WS/T 10042-2025 / wastewater surveillance & public-health intelligence | OECD: environmental-biotech | Regulator: FDA (US), EMA (EU), NMPA (CN)]

Wastewater-based epidemiology (WBE) is a non-invasive public-health intelligence service that screens whole cities and closed communities (universities, prisons, care homes, airports) for circulating viral and bacterial pathogens, antimicrobial-resistance (AMR) markers and psychoactive substances, by analysing representative composite samples drawn from the sewer network. Every resident shedding viral particles in faeces or urine contributes to the signal daily, so a single 24-hour sample at a treatment plant summarises the infection burden of an entire catchment — detecting viral-load rises 7 to 14 days before they surface as hospital admissions, and at a per-pathogen cost orders of magnitude below individual clinical testing. Biobot Analytics returns quantified pathogen results in 3–5 days from collection across 50+ health signals; the Verily-backed WastewaterSCAN network covers 147 plants across 40 US states and more than 37 million people; Bio-Rad’s QX600 droplet generator partitions each reaction into ~20,000 monodisperse nanodroplets with a process limit of detection of 2,160–2,680 copies/L for SARS-CoV-2; and long-term monitoring at Amsterdam Schiphol airport quantified SARS-CoV-2 in 98.6% (273/277) of samples over two and a half pandemic years.

The key directions of wastewater-based epidemiology are:

1. Viral particle concentration: municipal sewage is extremely dilute and loaded with inhibitors; electronegative-membrane filtration, PEG precipitation and centrifugal ultrafiltration (Amicon 100 kDa) selectively concentrate fragile enveloped RNA viruses from hundreds of millilitres down to 1–2 mL, routinely achieving 100–133× concentration factors.
2. Absolute quantification by droplet-digital PCR (ddPCR): partitioning the eluate into ~20,000 monodisperse nanodroplets makes each droplet an isolated Poisson reactor, delivering absolute genome-copy counts without calibration curves and tolerating the humic-acid and heavy-metal inhibitors that defeat RT-qPCR.
3. Metagenomic next-generation sequencing (mNGS): deep shotgun and hybrid-capture sequencing (Illumina NovaSeq) profiles every circulating strain, resolves rare and emerging sublineages, and maps the city-wide resistome — hybrid capture lifts the viral read fraction above 19% and can flag novel pathogens targeted panels miss.
4. PMMoV normalization and predictive analytics: raw pathogen concentrations are divided by the pepper mild mottle virus (PMMoV) faecal marker to cancel out storm-water dilution, then fed into GIS dashboards whose AI models forecast clinical outbreak curves 1–2 weeks ahead of physician reports.

Sectoral value chain

[24h composite autosampler at WWTP] ──> [Ultrafiltration / concentration] ──> [RNA/DNA extraction]
                                                                                   │
                                                                        (ddPCR / deep NGS)
                                                                                   │
   [Municipal containment response] <─── [Predictive Biobot / WWSCAN dashboard] <── [PMMoV normalization]

Value chain levels

Cross-cutting technologies of the sector:

• Digital PCR (ddPCR/dPCR): Bio-Rad QX600 (~20,000 droplets/reaction) and QIAGEN QIAcuity platforms are cross-platform equivalent for wastewater surveillance, enabling multi-lab networks.
• Hybrid-capture & shotgun metagenomics: Twist Comprehensive Viral Research Panel + Illumina NovaSeq lift viral read fractions >19% and detect >1,300 viral species per catchment.
• Settled-solids methodology: WastewaterSCAN’s analysis of sewage solids (sludge) raises analytical sensitivity ~10× versus liquid influent and stabilizes intra-site variance.

02US

The United States operates the most developed commercial WBE ecosystem, coordinated by the CDC National Wastewater Surveillance System (NWSS) in partnership with commercial data startups and academic consortia, with the White House ONDCP now using sewage data for national substance-use intelligence.

Biobot commercial leadership, CDC NWSS, Verily WastewaterSCAN, NIH RADx pipeline

• Biobot Analytics (Cambridge, MA): MIT-alumni founded startup (2017, $91.5M total funding across 15 rounds, ~$20M annual revenue) running the national Biobot Portal; awarded the White House ONDCP contract in January 2026 to deliver nationwide substance-use intelligence across more than 20 chemical targets, building on its 2020–2024 NIDA SBIR programme.
• CDC National Wastewater Surveillance System (NWSS): the federal programme aggregates commercial-operator and academic-lab data into a single early-warning system, with CDC NWSS endorsing both Bio-Rad and QIAGEN digital-PCR platforms for multi-lab surveillance.
• Verily WastewaterSCAN (Stanford/Emory): the Alphabet life-sciences subsidiary partners with Stanford and Emory universities to run a settled-solids network spanning 147 plants across 40 states and >37 million people, reporting SARS-CoV-2 N-gene RNA in ~94% of February 2026 samples (median 31,869 copies/gram).

03CN

China is building industrial-scale wastewater epidemiology under a centralised China CDC programme, codified in the new national technical standard WS/T 10042-2025 and rolled out across major megacities and industrial zones.

China CDC national sewage surveillance, WS/T 10042-2025 standard, Tsinghua methods, industrial-zone AMR monitoring

• National China CDC network: the China CDC Institute of Environmental Health convened a 15-city multi-pathogen sewage surveillance seminar in Zhuhai (November 2025) with 50 experts, and field-validated provincial rollout in Shaanxi (Xi’an, Baoji, Yulin, Ankang), targeting cholera, typhoid and respiratory pathogens.
• WS/T 10042-2025 national standard: the National Disease Control Bureau’s Technical specification for surveillance of SARS-CoV-2 in wastewater — co-drafted by China CDC, Tsinghua University and HKU — took effect on 1 June 2026, standardising sampling points, viral enrichment, nucleic-acid detection and data reporting nationwide.
• Industrial-park and zoonotic monitoring: provincial CDCs extend WBE to poultry and industrial zones to track antimicrobial resistance and pre-empt zoonotic outbreaks, with Tsinghua environmental biotechnology groups developing high-efficiency nano-adsorbents and anti-fouling membrane materials for in-flow virus capture.

04EU

The European Union was the first jurisdiction to mandate wastewater surveillance by law for large cities, creating a pan-European early-warning network anchored by the Joint Research Centre (JRC) and the Dutch KWR Water Research Institute.

EU JRC monitoring mandate, KWR Water Research, SCORE consortium, Horizon Europe WBE projects

• KWR Water Research Institute (Netherlands): the world’s pioneering WBE R&D centre, which first detected SARS-CoV-2 in Amsterdam Schiphol airport sewage; its long-term study quantified the virus in 98.6% (273/277) of samples and resolved variants in 91.0% (161/177) of sequenced samples between February 2020 and September 2022.
• JRC legislative mandate: under EU directives, every city above 100,000 population must perform weekly genomic monitoring of municipal sewage for pathogens and AMR markers, coordinated through the JRC into a single European database.
• SCORE consortium & airport surveillance: the Sewage Analysis CORe group Europe harmonises cross-EU analysis for illicit-drug and pharmaceutical consumption, while German (Robert Koch Institute / TU Darmstadt) and Dutch teams extend the model to airplane and airport wastewater as a traveller-based surveillance gateway.

05Leading companies and research institutes

06Tech stack and innovations

The WBE stack fuses low-tech sampling logistics with some of the most sensitive molecular diagnostics and computational epidemiology in routine public-health use.

1. Electronegative-membrane filtration & PEG precipitation:
   • At mildly acidic pH (3.5–5.5) viral capsids take on a net positive charge and bind electrostatically to cellulose-ester membranes bearing negative functional groups; a mildly alkaline (pH 9.0) elution buffer then releases the concentrated virus into 1–2 mL.
   • PEG 8000 (8%) with 0.3 M NaCl incubated at +4 °C precipitates viral and protein particles for high-speed centrifugal recovery, giving the 100–133× concentration factors needed to lift pathogens from the 100–100,000 copies/L raw range into the detection window.
2. Droplet-digital PCR (ddPCR):
   • The Bio-Rad QX600 splits each 20 µL reaction into ~20,000 monodisperse nanodroplets; after end-point PCR a six-channel optical reader classifies droplets at 1,000/second and Poisson statistics convert the positive-to-negative ratio into absolute copies per µL with a 95% CI, bypassing calibration-curve and inhibitor artefacts.
   • A 2026 head-to-head showed Bio-Rad ddPCR and QIAGEN QIAcuity dPCR are quantitatively equivalent for wastewater (mean differences ≤0.12 log copies/L, R² >0.93), validating multi-platform surveillance networks; Bio-Rad processed samples 32% faster (305 vs 435 minutes per 96-well plate).
3. PMMoV normalization & predictive analytics:
   • Pepper mild mottle virus is a plant RNA virus ingested with pepper and chilli sauces and shed in human faeces at remarkably stable concentrations, making it an ideal faecal-strength marker: dividing raw pathogen copies by the matched PMMoV reading cancels storm-water dilution.
   • Biobot’s and WastewaterSCAN’s cloud GIS portals then fuse normalized loads with clinical baselines and geography, delivering 7–14-day outbreak forecasts; the Gujarat WBE programme (2,130 samples) achieved Pearson r = 0.898 between wastewater and clinical COVID-19 and flagged the JN.1 lineage ~1 week before clinical isolation.

07Value chains and production pipelines

Industrial pipeline of metropolitan respiratory-virus (SARS-CoV-2, influenza A/B) wastewater surveillance via weekly sampling at 10 WWTPs, ultrafiltration, absolute ddPCR quantification on the QX600 and integration into the municipal Biobot early-warning portal (cGMP / cGDP / ISO 17025)

┌───────────────────────────┐      ┌───────────────────────────┐
│ 1. 24-hour composite      │ ───> │ 2. Ultrafiltration and    │
│    autosampling at WWTP   │      │    concentration           │
└───────────────────────────┘      └───────────────────────────┘
                                                  │
                                                  ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 4. Droplet generation and │ <─── │ 3. Automated magnetic-bead│
│    ddPCR amplification    │      │    RNA extraction (MagMax) │
└───────────────────────────┘      └───────────────────────────┘
               │
               ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 5. Absolute copy reading  │ ───> │ 6. PMMoV normalization and│
│    on the QX600 reader    │      │    AI-portal export        │
└───────────────────────────┘      └───────────────────────────┘

Stage 1: 24-hour composite autosampling at the WWTP

Flow-proportional Teledyne ISCO autosamplers installed at the inlet of 10 key plants draw 50 mL of raw influent every 15 minutes, compiling a 1.5-litre 24-hour composite held at +4 °C; couriers move the refrigerated samples to the analytical lab within 4 hours of collection.

Stage 2: ultrafiltration and concentration

A 200 mL aliquot is centrifuged at 4,000 × g for 20 minutes to settle coarse solids, the supernatant is passed through a 0.22 µm polyethersulfone vacuum filter, and the filtrate is loaded onto Amicon Ultra-15 100 kDa columns centrifuged at 3,000 × g for 30 minutes, concentrating viral particles 133× down to a 1.5 mL retentate.

Stage 3: automated magnetic-bead RNA extraction

1.0 mL of the viral concentrate is transferred to a KingFisher Flex deep-well plate; MagMax lysis buffer (guanidinium isothiocyanate) disrupts capsids and denatures inhibitors, a Phi6 bacteriophage process control benchmarks recovery efficiency, and MagMax magnetic nanobeads bind, wash and elute purified RNA into 100 µL of TE buffer.

Stage 4: droplet generation and ddPCR amplification

10 µL of RNA is combined with One-Step RT-ddPCR reagent and a multiplex FAM/HEX/Cy5.5 TaqMan panel targeting SARS-CoV-2 N, influenza A M, influenza B HA and the PMMoV marker; the QX600 Droplet Generator partitions the 20 µL mix into 20,000 nanodroplets, which are then thermal-cycled through 40 rounds of RT-PCR.

Stage 5: absolute copy reading on the QX600 reader

The post-PCR plate is loaded into the QX600 Droplet Reader, which streams droplets single-file through a laser at 1,000 droplets/second across six spectral channels; QuantaSoft clusters the droplets and, from the positive-to-negative ratio plus Poisson statistics, reports absolute copies per µL of eluate with a 95% confidence interval.

Stage 6: PMMoV normalization and AI-portal export

Absolute concentrations are divided by the matched PMMoV reading to cancel dilution, then pushed to the municipal Biobot Portal; an AI model fuses the normalized load with geography and clinical history and issues a 7–14-day outbreak forecast, auto-alerting the health department to stage beds and target mobile vaccination in the threatened districts.