A monitoring exercise along the tidal River Thames found that publicly reported live sewage spill durations were later cut by 82.9% after validation, according to an article published on 23 August 2026 by The Conversation. The author argues that the figures published in near real time should be read as provisional, not as confirmed environmental records.
Key takeaways
- England's water companies have been required since January 2025 to publish near real-time sewage discharge data via sites such as Sewage Map and Waterwatch, The Conversation reports.
- Between December 2025 and January 2026, reported spill durations at 91 tidal Thames sites totalled 1,988 hours, later reduced to 339 hours after validation, The Conversation says.
- Anglian Water's publicly reported live discharge durations were cut by 93%, from more than 1,551 hours to 108 hours, The Conversation reports.
- One Anglian Water site near Southend-on-Sea was first recorded as spilling for more than 850 hours, then revised to nine minutes, The Conversation states.
- Event duration monitoring (EDM) records when an overflow starts and stops but does not measure volume, pollutants, dilution or ecological harm, The Conversation writes.
What happened: the data behind the spills
The article, published on 23 August 2026, sets out how storm overflows on combined sewers are designed to release excess water during heavy rainfall and how, since January 2025, water companies in England have been required to publish near real-time discharge data on public dashboards such as Sewage Map and Waterwatch, according to The Conversation.
Beginning in December 2025, the author monitored 91 sites along the tidal River Thames to test whether EDM data could help explain microplastic pollution findings, The Conversation reports. The reported spill durations for the December 2025 to January 2026 window totalled 1,988 hours, or almost 83 days, but were later reduced by 82.9% to 339 hours, around 14 days, after water company validation, The Conversation says.
The largest single-company revision reported was at Anglian Water, where publicly reported live discharge durations fell by 93%, from more than 1,551 hours, about 64 days, to 108 hours, around 4.5 days, according to The Conversation. At one Anglian Water site near Southend-on-Sea in Essex, an apparent spill of more than 850 hours (33 days) was later revised to just nine minutes, The Conversation reports.
The author attributes the revisions to automated sensors being reviewed and corrected by water company staff, citing equipment faults, telemetry problems and precautionary alerts as common causes of overreporting that is then amended, The Conversation writes. The piece also warns the opposite can occur: a spill can be actively discharging but not appear on the monitoring sites, The Conversation notes.
What the data does, and does not, measure
EDM records only when an overflow starts and stops, which is a measure of duration rather than of pollution load, The Conversation reports. The data does not show how much sewage entered a river, lake or coastal water, what pollutants were present, whether the discharge was diluted by rainfall, or what effect it had on aquatic ecosystems, nor does it measure bacterial contamination or indicate whether the water is safe for swimming, according to The Conversation.
The article also argues that a one-hour discharge in a small river during dry weather can have a very different impact from a one-hour discharge into a rain-diluted river, and that several shorter spills can release more pollution than one longer event, even though both patterns would look identical in EDM data, The Conversation writes.
Looking ahead, the author calls for EDM to be combined with measurements of discharge volume, river flow, microbial contamination, chemical pollutants and ecological indicators such as fish and invertebrate populations, The Conversation reports, and flags that climate change, population growth and ageing Victorian sewer infrastructure are all adding pressure to wastewater systems.
Specification read
For a municipal works engineer, the headline figure is 82.9%, the cut between initial live spill hours and post-validation hours across 91 tidal Thames outfalls over roughly a two-month window. On an annualised basis, 1,988 hours of raw reported duration across 91 sites works out to about 21.9 hours per outfall per month, and the revised 339 hours works out to about 3.7 hours per outfall per month, both derived only from figures reported by The Conversation. A single combined-sewer outfall passing, for example, 5 MLD (5,000 m³/day) at a typical event duration of 4 hours would discharge on the order of 800 m³ per spill (general industry range, not from the sources), which is why duration alone is a poor proxy for pollutant load.
The class of project this event bears on is combined-sewer overflow (CSO) screening, storage and primary treatment on a municipal interceptor feeding a biological stage such as A2O, MBR, MBBR or SBR, followed by tertiary treatment and disinfection. The EDM story is squarely about the screening and storage end, where better instrumentation, validated telemetry and flow metering determine whether a plant actually meets its permit. A reader looking at compact or modular CSO and sewage plants of the type covered by our Underground Package Sewage Treatment Plant (WSZ Series) product page and our broader Integrated Sewage Treatment range, including Municipal applications, will recognise the same instrumentation gap: spill duration is logged, but volume, pollutant profile and dilution are not, so consent compliance and receiving-water quality still have to be proved downstream.
If your plant is a municipal works of roughly 1 to 50 MLD (1,000 to 50,000 m³/day) with BOD in the 150 to 350 mg/L range and ammonia above 10 mg/L, and you are under pressure from the Environment Agency on CSO spills and stormwater consent limits, this story signals that live spill dashboards will be the first place your regulator and the public look, so validated event duration monitoring, calibrated flow meters and continuous ammonia/UV transmittance probes are no longer optional ancillaries but part of the consent file.
FAQ
What does EDM actually measure on a CSO? It records the start and stop time of an overflow event, so it gives duration in hours, not volume in cubic metres and not pollutant load in kilograms, The Conversation reports. That is why two spills of the same logged duration can have very different environmental impact.
How unreliable is the live data, in practice? Across 91 tidal Thames outfalls, the author saw total reported spill hours of 1,988 reduced to 339 after validation, a cut of 82.9%, with one Anglian Water site revised from more than 850 hours to nine minutes, The Conversation says. Treat the live feed as provisional, not as an audited record.
What should I add to my instrumentation spec to close this gap? Combine event duration monitoring with discharge volume, river flow, microbial and chemical pollutant sensors, and ecological indicators such as fish and invertebrate surveys, The Conversation writes. For procurement, expect validated flow meters, ammonia analysers, UV transmittance probes and storm tank level instrumentation to be part of the consent file, not optional extras.
What is a realistic budget range to upgrade a small municipal CSO site? The sources give no prices. As a planning anchor only, retrofitting a small CSO site (around 1 to 10 MLD, or 1,000 to 10,000 m³/day) with validated flow metering, telemetry, ammonia and UV transmittance probes typically sits in the low six figures in sterling or dollars (general industry range, not from the sources), with lead times of roughly 8 to 16 weeks for the instrumentation skid and 6 to 12 months for full installation and sign-off.
