Why, oh why has the Environment Agency asked Affinity Water to turn abstraction pumps back on in the River Chess catchment?

In the late 20-teens John Lawson came to me with a great idea that could end over-abstraction in many chalk streams, especially those near London. He explained how in the next ten years or so Affinity Water would be building a pipeline to connect their southern region (south of the Thames) where they have more than enough water, with their northern, where they do not. This pipeline, said John, would enable the wholesale reduction of groundwater abstraction in the chalk stream tributaries of the Colne: iconic streams like the Ver and Chess that have been far too heavily abstracted for over half-a-century. And maybe the chalk streams of the Lea too.

If you turn the groundwater pumps off, a lot of the water you “leave in the ground” so to speak, comes back as flow in the stream. With a pipeline you could abstract the water at the lower end of the catchment instead, store it in reservoirs (of which there are several in the London area) and take the water back to the places it came from, to be used as public water supply. The difference being that this way, the rivers actually get to flow. We called John’s brilliantly simple idea “Chalk Streams First” because it gives the river first use of its water. With the support of a coalition of eNGOs we started trying to interest the water companies, the Environment Agency and Ofwat-Rapid (Regulators Alliance for Progressing Infrastructure Development).

Rapid was interested from the start. Paul Hickey, who directs Rapid, is passionate about ensuring we actually deliver on our environmental ambition. The Environment Agency was interested too. Even Affinity Water took to the idea, especially once the Environment Agency indicated that it might allow some variations on the theme and flexibility with licence relocation favoured by Affinity’s very clever technical guru, Doug Hunt.

The introduction of the Grand Union Canal transfer scheme promised to underwrite any losses to public supply (ie. disparity between what you no longer abstract at the top end and what you get back at the lower end of the streams) and thus Affinity Water started to build abstraction relocation into their business plan. They will address the Colne chalk streams to start with, but in due course all the chalk streams of the Lea could also be included. The first shots, the prequel shots in fact, were fired in 2020 when Affinity Water voluntarily shut down their Alma Road abstraction on the River Chess.

Through these same years the CaBA chalk stream strategy has gained momentum, with support from Defra, the water industry and all stakeholders. The Colne version of Chalk Streams First promised to become a national flagship for how to realign abstraction, put the environment first, but still take account of public supply. Literally everyone liked the idea. Who in their right mind wouldn’t?

So why, in the name of all that is Holy, has the Environment Agency asked Affinity Water to resume abstracting from Alma Road at a rate of up to 7 million litres per day, without much in the way of explanation (at first) or consultation (thus far)? The decision itself seems bizarre. The way it has been handled given how the plight of our chalk streams has touched the national consciousness and has been debated in Parliament, is clumsy, to put it politely.

In the interests of fairness, I ought to say that the Environment Agency has now explained that this request was made of Affinity Water in order to conduct a five-year modelling exercise to study the relationship between abstraction, groundwater levels and fluvial flood risk in the Chess catchment. Note the words I have placed in italics.

The River Chess has historically suffered from excessive abstraction which has reduced flows in the river and sometimes caused it to dry up altogether in its upper reaches. As with other Chilterns streams groundwater abstraction climbed massively through the 20th century, in the Chess from a minimal 2.5 Ml/d in the 1920s to a peak of over 20 Ml/d between 2008 and 2018, almost 38% of the average aquifer recharge, placing it amongst the highest impacted chalk streams in the country (in the more impacted, such as the Ivel and Darent, abstraction exceeds 50% of recharge).

The cessation of the Chartridge and Alma Road abstractions has reduced the overall catchment abstraction to more like 25% of aquifer recharge: still far too high, but enough to show noticeable benefits.

The River Chess Association report that otters, water voles, brown trout, water crowfoot, mayfly and rare invertebrates such as the winterbourne stonefly have all returned to Chesham. In fact nothing monitors improving river health more effectively than invertebrates. The Association has been recording river-dwelling invertebrates in Chesham since 2009. In recent years species previously unseen in Chesham have been recorded, including mayfly (Ephemera Danica), caseless caddis (Rhyacophilidae ), turkey brown (Paraleptophlebia submarginata), and the nationally rare winterbourne stonefly (Nemoura lacustris).  

Personally, I remember taking photographs in Chesham of a dry river in May 2017 and of a flowing river full of ranunculus in August 2022. The Chess stood for hope.

So why toss that all away? The stated explanation seems feeble. When asked by the River Chess Association what reasoning and data were behind the decision, the Environment Agency replied:

“The Environment Agency used their current understanding of the relationship between groundwater abstractions, groundwater levels, and river flows in the Chess catchment. This included information from two reports … which concluded that there is a relationship between groundwater abstraction and river flows. Based on the conclusions from both reports, a potential increase in fluvial flood risk [arising from a reduction of abstraction – my clarification not EAs] could not be ruled out. Implementing temporary adaptive abstraction, as set out in the operating agreement, minimises the potential impacts of abstraction reductions on fluvial flood risk until these impacts are better understood and managed.”

This states the obvious – that there is a relationship between abstraction, groundwater levels and flows – and presents it as an explanation. Of course there is a relationship! That’s why we want the abstraction to be reduced, to increase flows in the river and thus river health. In as much as it is an explanation it seems to be saying that the resumption of abstraction will be used to assess if abstraction can be used to reduce flows in the river, and via this reduce the risk of flooding.

Taken at face value this is very strange reasoning. The idea appears to be to use abstraction to reduce flows in the river. Despite what the EA state about adaptive abstraction* in the operating agreement I wonder a) if repurposing an abstraction licence from its use for public water supply to a different use of so-called flood-risk mitigation is within the remit of the licence and b) whether it is entirely legal under WFD legislation to deliberately reduce the flows in the river in order to theoretically reduce flood risk.

(*adaptive abstraction essentially comprises the variation of pumping rates across time, but I’ve only ever heard of the idea as a means to reduce ecological damage, which is the unfortunate by-product of the public water use, the reason why the licence exists. The EA’s idea here is actually putting the abstraction to a entirely different use than intended by the existing licence)

But these queries aside, this plan is not even a good way to reduce fluvial flood risk. Of course flooding is related to flows (and flows to groundwater levels), but in a chalk stream fluvial flooding is much more likely to be influenced by things such as impoundments, culverts, drainage, ditching and land use in the upper catchment. The EA would be far, far better off looking at these issues in order to mitigate fluvial flooding.

And that aside, using groundwater abstraction as a temporary measure to reduce fluvial flooding is like blowing the other way in order to slow down a tanker. The impact of groundwater abstraction accumulates over time and its impact on flows is geared via its impact on groundwater levels. By the time you realise you might have to reduce groundwater levels to reduce flows it is too late. You could only reasonably make this idea work if you run the abstraction all the time and reduce flows all the time, which is exactly what groundwater abstraction does.

Besides, where are they going to put all the water? Pump it downstream?

To me this feels like a nonsense explanation.

As anyone with a Twitter account knows, the sewage works at Chesham spills groundwater ingress sewage all the time when groundwater levels are high. In other words the groundwater spills through cracks in the pipes and overwhelms the sewage works. It is almost certain that the increase in groundwater levels that has followed the reduction in abstraction has increased the groundwater sewage spills (that and some very wet winters).

Is this really about modelling something we know all about already (the relationship between abstraction, groundwater levels and flows, which it is perfectly possible to accurately model), or is it a designed to see if the groundwater ingress flooding can be reduced by resuming abstraction?

You decide. Maybe I’m being too cynical. But if my suspicions are correct the EA would be trying to play tunes on the theme of ecological damage, resuming one form of damage to reduce another and I’m not sure that’s such a great idea. Or maybe Defra is exploring ways to meet its own stormwater reduction plan targets for chalk streams? These pesky groundwater ingress discharges are going to be a hard nut to crack.

If the issue really is fluvial flooding, where is the risk occurring exactly? And why not consult the Chess Association, and the Chilterns Society / chalk streams project to explore how the flood risk could be addressed without pumping all the groundwater away? I will be happy to publish any reply or further explanation from the EA.