14 June 2024


Water company data and performance remains in the spotlight. Most recently, the BBC published an article which highlighted data cited as ’potentially indicating’ dry day storm overflow spills from all water companies in England. The industry strongly rejects the claims, citing flawed methodology and inaccurate interpretation of data. But why the discrepancy? Anglian Water’s Head of spill reduction, Gail Pickles explains why a dry day spill, isn’t always a dry day spill.

We know our customers expect action to stop all spills – no number other than zero will do. We have a huge plan to address this, but we also want to be as transparent as we can to explain our data and the action we are taking.

Storm overflows (CSOs) have been around for decades and were initially introduced to our network as a means of protecting homes and businesses from flooding during periods of heavy or prolonged rainfall or snow melt. Acting as pressure release valves, storm overflows allow excess water into rivers or the sea to prevent sewers from overflowing and backing-up through toilets, drains and manholes. The clue is in the name - storm overflows. The job they do means they should not operate when it’s dry.

We are the only country in Europe to have installed monitors on all our storm overflows. Event duration monitors (EDMs) indicate the start and stop times of when a storm overflow may have been active – and I choose that word deliberately, for a good reason.

Getting accurate information about storm spills is not straightforward. In fact, the technology and EDM monitors we’re using are still relatively new – less than five years old. They are placed in sewer pipes – which are fairly inhospitable environments and it’s not uncommon for them to be ‘activated’ without a spill occurring.

Nearby activity such as high-speed trains, interruptions in network signal or even a very active spider web can trigger an activation. If we see a sensor has been activated when we wouldn’t normally expect it, like in dry weather, we send our teams to check the site and use other telemetry, catchment and weather data to decide whether a spill is in fact genuine. It takes a great deal of time, data and expertise to make sure we’re reporting the most accurate information we can, all of which is shared with our regulator, the Environment Agency in an annual return. And we take a very cautious approach to what we include, if we can’t find enough robust evidence to suggest a spill didn’t happen, it goes in the return.


The BBC ‘dry-day spill’ methodology


Dry day spills are not a standard measure. Should an EDM activation occur on a dry day, its verification depends on three further elements: the recent rainfall; the size, shape & geology of the catchment, and on-site data from a variety of sensors, aside from EDMs. While there is no finalised methodology laid out from the EA on how dry spills should be calculated, we’ve been heavily involved in suggesting how that could be ascertained, taking all of these factors into account.

The geography of that catchment is hugely important in determining whether a spill occurs on a dry day. Water flows downhill so the less hilly an area, the longer it takes to drain through the catchment. This means that in a large catchment area that is relatively flat, rain takes time to drain, so in periods following heavy rainfall, a storm overflow could still operate some days after it has stopped raining. This doesn’t mean the spill has happened on a dry day.

The BBC has applied a 9KM2 area around an EDM monitor to calculate the amount of rain (or lack of rain) which may lead to an activation on a given day. However, geology and the time it takes for water to flow through the system varies considerably from catchment to catchment. For example, for the monitor in Great Billing, Northamptonshire the actual catchment area is 127 KM2, more than 14 times larger than assumed by the BBC, so therefore water would still be flowing through the catchment for a considerable time after the rainfall has stopped.


How accurate is water company data?


We have enough sewer pipe to go around the world twice. As we’ve said, this monitoring technology is still new, and it takes time and a team of operational and data experts to verify the thousands of lines of data we get from EDMs on this network.

We cross reference all this information with the best, localised weather data, as well as other monitors, neither of which the BBC has access to. And our understanding is getting better all the time. Because of the precautionary approach we take, we now know that some of the spills we declared in 2022 were actually not spills at all, but we would always rather over report.


Taking action


Ultimately we want to stop all spills and consign storm overflows to history. Every single one of the 1,471 storm overflows across our network now has a detailed improvement plan that will see discharges significantly reduced. Our shareholders have just fast tracked an addition £100m, to begin this work now.

Earlier this year, we launched our near real time storm overflow map, updated hourly, that shows the most recent activations. From next year, if our business plan is approved by Ofwat, we will be able to move forward with a £1bn package of investment to tackle storm spill over the next five years. 

We want our customers to know that we are taking action and investing in the right solutions which will have the most benefit for the environment now and in the future.