Campaigns · · 6 min read
Sewage spills with no rain: what River Hub shows about dry-weather spilling on the Dart
Nearly 3,000 sewage discharges into the Dart since 2020 happened when it hadn’t rained. The first in our series on what River Hub can show, and why it needs to go national.
This is the first in a series of posts showing the kind of analysis River Hub makes possible. River Hub is our volunteer-built tool that brings together the public data on sewage discharges into the Dart, and now the Teign. Each post takes one question that the public data can’t easily answer, shows what River Hub finds, and explains what we are doing about it. We start with one of the clearest warning signs there is: sewage spilling when it hasn’t rained.
Why a dry spill matters
Storm overflows are a safety valve. In heavy rain, when rainwater floods into the sewers, they let diluted sewage escape into the river so that it doesn’t back up into homes and streets. Their permits allow this in exceptional wet weather, not as a matter of routine. So when an overflow discharges on a dry day, there is no storm to explain it. It points to something else: a blockage, a failed pump or a works that can’t cope with its everyday flow. Whatever the cause, it is hard to reconcile with the permit.
How we decide a spill was dry
We use the Environment Agency’s own one-day test. Each spill is matched to the nearest Environment Agency rain gauge, and it counts as dry if that gauge recorded no more than 0.25 mm of rain on the day of the spill and on the day before. That is less rain than a light drizzle. For the five overflows below, the matched gauge is between 30 metres and 4.3 km away, and the table shows which gauge each one uses.
Every dry spill on River Hub has its own evidence page. It shows when the discharge started and stopped, which rain gauge it was matched to, and the rainfall on that day and the days around it. Anyone can check our classification, and that includes South West Water and the Environment Agency.
What we found
Since 2020 the overflows on the Dart have recorded nearly 3,000 dry-weather discharges. These are individual discharges as logged by the monitors on each overflow, not the Environment Agency’s annual “counted spills”, which group discharges into 12- and 24-hour blocks. Most come from five overflows at four treatment works, and only one of those works has a date for improvement.
Overflow | Dry spills since 2020 | Longest recent dry spills | Matched rain gauge | What South West Water has planned |
|---|---|---|---|---|
1,775 | 47 hours from 24 Feb 2026; 43 hours from 10 Dec 2025 | Austins Bridge, 4.3 km | No scheme before 2030; pencilled in for 2035 to 2040 | |
260 | 12 hours from 5 Mar 2026 | Austins Bridge, on site (30 m) | No more than 10 spills a year and 2 per bathing season, plus screening and UV treatment, by 30 April 2028 | |
154 | About 6½ days from 24 Dec 2025; about 5½ days from 1 Mar 2025 | Holne Priddons Farm, 2.0 km | No scheme before 2030 | |
84 | 4 hours from 27 Feb 2026, then repeated short spills | Austins Bridge, on site (30 m) | As Kilbury above, by 30 April 2028 | |
78 | Almost 6 days from 26 May 2026; 77 hours from 19 Mar 2026 | Harbertonford Slipperstone, 2.6 km | No scheme before 2030 |
Some of these spills lasted for days, with no rain to drive them: almost six days at Harbertonford from late May this year, and six and a half days at Scorriton over Christmas 2025. Rattery’s total is dominated by one year. 1,735 of its dry spills came in 2022, which points to a sustained problem that year, either at the works or in its monitor, and is exactly what an investigation should explain.
The dates South West Water has set are about cutting spills in wet weather, to meet the government’s target of about ten a year by 2045 to 2050. A dry spill is a different problem. Permits only allow spills in exceptional wet weather, so dry spilling should be investigated and stopped now, not left to an improvement scheme that may be 10 or 15 years away.
Dry spills year by year
Dry spilling is not just a Rattery problem. Leave out Rattery’s 2022 and there are still around 1,250 dry-weather discharges since 2020, from 31 different overflows. Kilbury’s storm-tank overflow has spilled dry every year.
Overflow | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 (to date) | Total |
|---|---|---|---|---|---|---|---|---|
Rattery, storm overflow | 0 | 0 | 1,735 | 14 | 19 | 0 | 7 | 1,775 |
Kilbury, storm-tank overflow | 48 | 9 | 23 | 84 | 55 | 18 | 23 | 260 |
Scorriton, storm overflow | 0 | 18 | 97 | 14 | 9 | 16 | 0 | 154 |
Kilbury, storm overflow | 0 | 0 | 4 | 47 | 33 | 0 | 0 | 84 |
Harbertonford, storm-tank overflow | 0 | 0 | 0 | 62 | 6 | 1 | 9 | 78 |
Harberton, storm overflow | 0 | 0 | 38 | 32 | 3 | 0 | 2 | 75 |
Harberton, storm-tank overflow | 0 | 0 | 20 | 15 | 17 | 15 | 0 | 67 |
Totnes, storm-tank overflow | 0 | 2 | 3 | 15 | 30 | 11 | 0 | 61 |
Broadhempston, storm-tank overflow | 0 | 34 | 0 | 3 | 19 | 0 | 0 | 56 |
Dartington School, combined sewer overflow | 0 | 3 | 37 | 0 | 5 | 4 | 0 | 49 |
21 other overflows | 34 | 28 | 98 | 78 | 53 | 15 | 18 | 324 |
All 31 overflows | 82 | 94 | 2,055 | 364 | 249 | 80 | 59 | 2,983 |
Of the rest of the top ten, Dartington School’s overflow is due to meet its 10-a-year limit by 30 April 2028. Harberton’s two overflows and Broadhempston have no scheme before 2030. Planned dates are from our earlier posts on what South West Water’s Dart investigations found and the ten-spill promise.
What we have asked the Environment Agency
On 7 October 2026 we wrote to the Environment Agency with these findings and links to the evidence for 15 example dry spills. We asked two things.
First, is the Agency already investigating dry-weather spilling at these works, and if so, how far has it got?
Second, under the Environmental Information Regulations, we asked for the “flow to full treatment” settings for Rattery, Kilbury, Scorriton and Harbertonford, and for any spill investigation reports it holds on them. Flow to full treatment is the amount of sewage a works must fully treat before its overflow is allowed to operate. It is the key number for telling whether a works is spilling too early. This is information about emissions to the environment, so it cannot be withheld on grounds of commercial confidentiality.
We have also offered to share our event-by-event data with the Agency. We’ll report what comes back.
Why this needs to go national
SWW and the EA do not report on dry spills to the public. The annual spill returns give one total per overflow for the year. They don’t say which spills happened in dry weather, and they don’t match spills against rainfall. The regulators’ first joint report, published on 29 October, doesn’t go below company level at all. To find a dry spill, someone has to take each individual discharge, find the right rain gauge, check the rainfall on the day and the day before, and keep doing it as new data comes in. That is a lot of work to do by hand for one river, and nobody can do it by hand for every river.
River Hub does it automatically. It started on the Dart and now runs on the Teign too, where Friends of the River Teign have their own version. Nothing in the method is specific to Devon. Every overflow in England has a monitor, every company now publishes its spill data, and the Environment Agency’s rain gauges cover the whole country.
Doing this nationally would make three things possible:
- Every river group could see its own dry spills, with the evidence to back them up, instead of relying on yearly totals.
- Regulators and journalists could compare works and companies on a like-for-like basis, using the Agency’s own test.
- The overflows that spill without rain could be found and fixed sooner, wherever they are, instead of only where a local group has the time and skills to dig.
We are now working to make River Hub available to other rivers. If you’re part of a river group and would like it for your river, or you’d like to help us build it, please get in touch at hello@friendsofthedart.org.
Coming up in the series
Future posts will look at emergency overflows for which no data is published, overflows that spill before their treatment works is full, whether the works can cope with the homes they serve, and which problem overflows have no improvement planned.
