Why this matters
What a sewage spill actually means for normal people — and why the data often confuses everyone
A combined sewer overflow (usually shortened to CSO) is a release point in the sewer network that can discharge a mix of rainwater and untreated wastewater into rivers during heavy pressure events. In plain English: when systems are overloaded, water companies can let any water in the sewer out so sewage does not back up into homes and streets. It's diluted wastewater, but this still has an impact.
The central problem is not only pollution. It is visibility.
Why UK sewage data feels contradictory
Open data has only recently been established, started by Thames Water. You will now often see two claims at once: “spills are improving” and “rivers are still under pressure”. Both can be true. Year-on-year numbers move with weather, sensor uptime, and local network behaviour. A wet year can make performance look worse even if operational response improves. A dry year can flatter results that are not structurally fixed.
At first glance, a lower spill total looks like a clear win. It is not always that simple. Without context, raw totals can mislead communities and policymakers in both directions.
They show event counts without explaining rainfall context, monitor outages, or changes in local catchment behaviour.
Live status, event timing, nearby rain and river context, and explainers on methodology so trends can be interpreted properly, alongside live alerts for those who care most.
How our system works in plain English
We ingest public event data, keep the raw timeline, and surface it in forms real people can act on: map view, site pages, downloads, and alerts. You can check whether an overflow is active now, when it began, and what happened at that site over time. We also place events next to river-level and rainfall signals to avoid the “single number” trap.
In practical terms, this means a parent planning a weekend river walk can make a better judgement call. A local campaign group can point to repeat patterns at a site rather than one-off anecdotes. A journalist can inspect the same raw backbone that powers the visuals.
We do not claim to replace regulators. We do make the public record faster to understand and highlight the limitations. It's tough to have a bulletproof system.
Data limits (and why we publish them openly)
- Sensor networks can go offline, which creates blind spots.
- Public feeds can update with delay during incidents.
- Event data shows discharge timing, not full ecological impact.
- A site can look better or worse depending on weather year.
- Sites can be adversley affected by groundwater, which makes a site look like it's just spilling sewage
That transparency is intentional. Trust is not built by pretending data is perfect; it is built by showing what the data can and cannot prove.
A real-world use case
Imagine a site that had an upgrade, but continued to spill at similar rates. A quick headline would blame. Our deeper view might reveal that total spill hours increase, but the site under equal pressure conditions performed better when you facor in rainfall, river levels, soil saturation, groundwater and all the things that add strain to a sewer network. This is exactly the type of nuance that simple league tables miss.
Suggested reading path:
