Event Duration Monitors
On 30 December 2023 DEFRA published a press release Storm overflows monitoring hits 100% target – Every storm overflow across England’s water network is now monitored it says: Event Duration Monitors [EDMs] increase transparency by measuring how, when and for how long a storm overflow is in operation. This shows the public when discharges are happening…
Storm overflow data has been available for a few years, but only as annual returns which isn’t very helpful if you want to know if a particular water body might be safe to swim in today. There’s been a lot in the press about this so water companies have been prevailed upon to start publishing real-time data from their EDMs.
Anglian Water's storm overflow map
On 9 May 2024 Anglian Water published a news item: Information at your fingertips: Our new storm overflow map which says: The map shows where our storm overflow monitors are located and provides near real time information on any activations. Updated hourly, it gives customers, stakeholders and those passionate about their local environment, a way to see how our network is operating, so they can make an informed decision about how they interact with their local river or bathing water.
The overflow map itself shows all 1471 Anglian Water's overflow locations, and more information if you click on one of them.
revIvel heard about this map a bit earlier than the official launch and sent a question to a contact at Anglian Water:
The preamble to the new storm overflow map said on 30th April: The full data for each sensor can be found on storm overflow information page. I couldn't find that page and looking again today, 2 May, the preamble text seems to have changed to: History for monitors that have been triggered since then is displayed in the box that shows when you click on an icon. but the box seems to only show the Most recent activation. Where can I find the activation history for a site? (and I mean this real-time data, not the annual return spreadsheets which are just a belated summary).
The reply on 10 May was: the visualisation system can currently only display the history for the most recent activation and does not have the ability to store a complete history for each sensor. We publish the history of each sensor, each year.
In other words, they're not publishing any real time history even though they must have it.
Letchworth Sewage Treatment Works
This works discharges into Pix Brook which rises in Letchworth, runs through the middle of Stotfold and joins the Ivel close to the railway bridge at Henlow. Today, 24 May, if you click on Letchworth STW on the map, it says: Most recent activation: Started 22 May 2024 21:30, Ended 22 May 2024 21:30 Duration: 0 days, 0 hours, 0 minutes.
On that Wednesday 22 May it rained a lot, a proper downpour, all day. Looking at the map now, a user will probably think that Anglian Water are doing quite well to have had no storm overflow at this sewage works after significantly more than an inch of rain fell in the Letchworth area that day.
But did they really have no overflows?
Because the map only shows the most recent EDM activation data, a revIvel member interested in overflow history recorded the readings of a few overflow sites which drain into the River Ivel. Not long after it started raining on 22 May, several overflows were activated. All went well for a while, and then the member started noticing strange things happening to some of them, notably Pirton, Hitchin (Ashbrook) and Letchworth STW. A section of the log for Letchworth STW is reproduced here.
The first two columns are the row number (for our reference below) and the time the log was made. The other four columns are the data collected from the map, all times are BST.
The first two rows show how things had been proceeding for some time as one might expect, the overflow had started at 12:02 and if you had clicked on the map at 18:23 you would have seen the overflow had been going on for 6 hours and 21 minutes.
But if you clicked on the map a few minutes later at 18:31 you would have seen something different; now the map is saying (row 3) the overflow started at 16:46 and has been going on for just 1 hour and 44 minutes; nearly four hours and 44 minutes of overflow had vanished!
It's not like the overflow stopped and started in the time between logging rows 2 & 3 because then you would expect to see a duration of just a few minutes in row 3, but no, the start time was way before, at 16:46.
At 19:26 this happened again, the overflow had reached 2 hours and 33 minutes duration (row 10) and then the start time was bumped forwards to 17:31 (row 11) meaning the duration dropped to 1 hour and 55 minutes - another 45 minutes of overflow had vanished!
At 19:54 (row 15) two things happened: On the map you would have seen there was no current overflow, and it said the most recent overflow lasted for just two minutes, by now a total of 6 hours and 11 minutes of overflow had vanished.
It gets wierder: Between 20:01 and 21:18 (rows 16 - 27) someone clicking on the map would have seen the last overflow as lasting variously 2 minutes, 0 minutes or 3 minutes, and then at 21:25 (row 28) it is overflowing again, but this is not a new overflow because it said it started at 19:31 (ie between rows 10 & 11) which was before the last overflow stopped. This suggests the overflow had not actually stopped at all but had been going on continuously since 12:02, that's 9 hours and 23 minutes duration, rather than the 1 hour and 53 minutes claimed.
The overflow had stopped by 22:56 (row 41) but said the last overflow lasted just under 2 minutes. Even though there's no evidence of any further overflows, at 23:31 (row 46) the data said the most recent overflow lasted 0 minutes, 0 seconds (in the raw data it is actually 115 micro seconds which is a bit unlikely), and that is what we see two days later on 24 May, and perhaps will continue to see until the next time it rains.
It appears no less than 10 hours and 54 minutes of overflow from Letchworth STW on Wednesday 22 May totally vanished!
Is Anglian Water telling the truth about overflows?
It doesn't look like it, does it?
The main purpose of making water companies publish this stuff is to identify so-called "dry spills" which are illegal. A ‘dry spill’ is when a storm overflow is used on a ‘dry day’ – which is usually defined as no rainfall above 0.25mm on that day and the preceding day (24 hours). On a day when there was something between 30 and 40 mm of rain in the area, the spill detailed above was almost certainly within the discharge licence, but if Anglian Water can't record legal spills correctly, what chance of anyone ever seeing an illegal one?
An Event Duration Monitor is a simple thing, either there is an overflow or there is not, and all it has to do is report when that overflow started and when it stopped. It would have been better had the government required the water companies to measure, record and publish overflow volume as well, but unfortunately they didn't. Nevertheless, what the EDMs are doing is better than nothing, and the IT involved in this sort of thing is not rocket science, especially not for a £multi-billion company like Anglian Water.
It appears their EDMs are reporting the current status about 1 hour and 40 minutes after the fact, which could explain some of the apparent vanishing of overflow, but very short events are obviously nonsense. With their Storm overflow map, Anglian Water seem to have implemented something which should be simple, could be helpful to lots of people, but appears to be totally unfit for purpose in its present form.
Not only that, at least in the case of overflows from Letchworth on 22 May as evidenced here, it would appear to be an outright deception: Contrary to what Anglian Water say, a casual user trying to make an informed decision about how they interact with their local river to see if it was OK for them and/or their children and/or their dog to swim in Pix Brook that evening might have thought it was fine when in reality it could easily have been chock full of toxic raw sewage.
Whether Anglian Water's many other EDMs do the same thing is difficult to say at this time, but with systems like this it makes operational and financial sense to set up both hardware and software in a consistent manner, so it would not be a surprise to find all their EDM reporting is similarly deceptive.
For casual users, an overflow total time for the day so far, or over a rolling 24 hour period would likely be far more informative than what we see today. For revIvel and the many other other people and organizations who have an interest in the state of our rivers, access to the full overflow history of each site would give a proper meaning to the 'transparency' which DEFRA would like to see. It would also help to mitigate any allegation that the figures contained in annual returns are 'massaged', as massaged they must be with a data muddle like we see in this example.
It would be interesting to know, after proper review, the duration Anglian Water think there was an overflow from Letchworth STW on that rainy Wednesday 22 May: Was it something close to 10 hours and 54 minutes or was it nearer 0 hours and 0 minutes?
And, dear Anglian Water readers; What about publishing overflow volumes too?